diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,134472 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 19206,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 531.315111998153,
+      "learning_rate": 3.466204506065858e-08,
+      "loss": 5.0363,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 292.2964725663925,
+      "learning_rate": 6.932409012131716e-08,
+      "loss": 4.6321,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 264.51170744342073,
+      "learning_rate": 1.0398613518197575e-07,
+      "loss": 4.7518,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 65.54403336268102,
+      "learning_rate": 1.386481802426343e-07,
+      "loss": 4.7551,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 71.62633373356861,
+      "learning_rate": 1.733102253032929e-07,
+      "loss": 4.9667,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 93.71744734115397,
+      "learning_rate": 2.079722703639515e-07,
+      "loss": 4.4864,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 79.13891063463899,
+      "learning_rate": 2.426343154246101e-07,
+      "loss": 4.3112,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 130.9824844691712,
+      "learning_rate": 2.772963604852686e-07,
+      "loss": 4.7113,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 62.15746440823066,
+      "learning_rate": 3.119584055459272e-07,
+      "loss": 4.5315,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.06252314526499,
+      "learning_rate": 3.466204506065858e-07,
+      "loss": 4.1675,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.181877274700796,
+      "learning_rate": 3.8128249566724436e-07,
+      "loss": 4.7039,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 77.56295637293104,
+      "learning_rate": 4.15944540727903e-07,
+      "loss": 4.7308,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 90.59912217138084,
+      "learning_rate": 4.5060658578856156e-07,
+      "loss": 4.268,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.27261208385014,
+      "learning_rate": 4.852686308492202e-07,
+      "loss": 4.0223,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 53.092113647032356,
+      "learning_rate": 5.199306759098788e-07,
+      "loss": 4.0547,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 194.42885862477203,
+      "learning_rate": 5.545927209705372e-07,
+      "loss": 4.1719,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 192.72886528486674,
+      "learning_rate": 5.89254766031196e-07,
+      "loss": 3.9482,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.923874944802805,
+      "learning_rate": 6.239168110918544e-07,
+      "loss": 4.0473,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 243.21030215791566,
+      "learning_rate": 6.58578856152513e-07,
+      "loss": 3.7208,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 199.76016694550418,
+      "learning_rate": 6.932409012131716e-07,
+      "loss": 3.5057,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 219.9332157291184,
+      "learning_rate": 7.279029462738301e-07,
+      "loss": 3.8037,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 224.8460521206362,
+      "learning_rate": 7.625649913344887e-07,
+      "loss": 3.6361,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.96817437593409,
+      "learning_rate": 7.972270363951473e-07,
+      "loss": 3.451,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 176.07245204045535,
+      "learning_rate": 8.31889081455806e-07,
+      "loss": 3.1902,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 196.34398136614035,
+      "learning_rate": 8.665511265164646e-07,
+      "loss": 2.9131,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.43285611524579,
+      "learning_rate": 9.012131715771231e-07,
+      "loss": 2.8942,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 130.1069982354433,
+      "learning_rate": 9.358752166377817e-07,
+      "loss": 2.521,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.298152068413035,
+      "learning_rate": 9.705372616984403e-07,
+      "loss": 2.6266,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.695189025738316,
+      "learning_rate": 1.0051993067590988e-06,
+      "loss": 2.5869,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.64249612592944,
+      "learning_rate": 1.0398613518197575e-06,
+      "loss": 2.4856,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.675691242861983,
+      "learning_rate": 1.074523396880416e-06,
+      "loss": 2.4556,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.53407161761756,
+      "learning_rate": 1.1091854419410745e-06,
+      "loss": 2.318,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 96.18604569702083,
+      "learning_rate": 1.1438474870017332e-06,
+      "loss": 2.3074,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 141.04263290664068,
+      "learning_rate": 1.178509532062392e-06,
+      "loss": 2.1149,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.26582216230617,
+      "learning_rate": 1.2131715771230504e-06,
+      "loss": 2.0832,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.986545105175942,
+      "learning_rate": 1.2478336221837089e-06,
+      "loss": 2.2623,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.66247728466575,
+      "learning_rate": 1.2824956672443676e-06,
+      "loss": 2.1439,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.463273681199404,
+      "learning_rate": 1.317157712305026e-06,
+      "loss": 1.9362,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 84.87951823415129,
+      "learning_rate": 1.3518197573656846e-06,
+      "loss": 2.054,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.529187478472057,
+      "learning_rate": 1.3864818024263433e-06,
+      "loss": 2.0543,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 57.58101304987958,
+      "learning_rate": 1.4211438474870018e-06,
+      "loss": 1.9546,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 59.388993477434354,
+      "learning_rate": 1.4558058925476603e-06,
+      "loss": 2.0146,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.39079349038478,
+      "learning_rate": 1.490467937608319e-06,
+      "loss": 2.0469,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.011224687411563,
+      "learning_rate": 1.5251299826689774e-06,
+      "loss": 1.8853,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 57.65040422903312,
+      "learning_rate": 1.5597920277296362e-06,
+      "loss": 1.8947,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.209375496951544,
+      "learning_rate": 1.5944540727902946e-06,
+      "loss": 1.8545,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.473361066444234,
+      "learning_rate": 1.6291161178509536e-06,
+      "loss": 1.7729,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 59.65041968269149,
+      "learning_rate": 1.663778162911612e-06,
+      "loss": 1.9303,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 47.739525868265204,
+      "learning_rate": 1.6984402079722705e-06,
+      "loss": 1.8462,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 64.66907700879658,
+      "learning_rate": 1.7331022530329292e-06,
+      "loss": 1.8848,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 59.48751044620968,
+      "learning_rate": 1.7677642980935877e-06,
+      "loss": 1.8173,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.627369497177005,
+      "learning_rate": 1.8024263431542462e-06,
+      "loss": 1.7882,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 55.94019378697077,
+      "learning_rate": 1.837088388214905e-06,
+      "loss": 1.8156,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.79310617919457,
+      "learning_rate": 1.8717504332755634e-06,
+      "loss": 1.7695,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.858413094829237,
+      "learning_rate": 1.906412478336222e-06,
+      "loss": 1.6991,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.585260260538933,
+      "learning_rate": 1.9410745233968806e-06,
+      "loss": 1.6335,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 65.95190152730467,
+      "learning_rate": 1.9757365684575393e-06,
+      "loss": 1.8287,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.052890780604574,
+      "learning_rate": 2.0103986135181976e-06,
+      "loss": 1.5994,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 52.96283885189271,
+      "learning_rate": 2.0450606585788563e-06,
+      "loss": 1.7196,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 56.70262797924363,
+      "learning_rate": 2.079722703639515e-06,
+      "loss": 1.7748,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.195107480088836,
+      "learning_rate": 2.1143847487001733e-06,
+      "loss": 1.6626,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 63.72151010848557,
+      "learning_rate": 2.149046793760832e-06,
+      "loss": 1.7699,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 38.40982890234492,
+      "learning_rate": 2.1837088388214907e-06,
+      "loss": 1.6679,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 43.77033326258082,
+      "learning_rate": 2.218370883882149e-06,
+      "loss": 1.8076,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.80828537183788,
+      "learning_rate": 2.2530329289428077e-06,
+      "loss": 1.6574,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.0141134889743,
+      "learning_rate": 2.2876949740034664e-06,
+      "loss": 1.7012,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.550052719832058,
+      "learning_rate": 2.322357019064125e-06,
+      "loss": 1.495,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 42.31678881948093,
+      "learning_rate": 2.357019064124784e-06,
+      "loss": 1.6645,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.22241947059129,
+      "learning_rate": 2.391681109185442e-06,
+      "loss": 1.6002,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.846153629350137,
+      "learning_rate": 2.4263431542461008e-06,
+      "loss": 1.7184,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.693782384986555,
+      "learning_rate": 2.4610051993067595e-06,
+      "loss": 1.5383,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 65.46057084460385,
+      "learning_rate": 2.4956672443674178e-06,
+      "loss": 1.5847,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 48.29185609030196,
+      "learning_rate": 2.530329289428076e-06,
+      "loss": 1.609,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.895269717082964,
+      "learning_rate": 2.564991334488735e-06,
+      "loss": 1.4175,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.28353952262731,
+      "learning_rate": 2.599653379549394e-06,
+      "loss": 1.5396,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 41.48833822367313,
+      "learning_rate": 2.634315424610052e-06,
+      "loss": 1.537,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.32863144952529,
+      "learning_rate": 2.668977469670711e-06,
+      "loss": 1.3886,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.59213100046043,
+      "learning_rate": 2.703639514731369e-06,
+      "loss": 1.6539,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.78870746070221,
+      "learning_rate": 2.7383015597920283e-06,
+      "loss": 1.5099,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.31520248932208,
+      "learning_rate": 2.7729636048526865e-06,
+      "loss": 1.5438,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 42.59625680645732,
+      "learning_rate": 2.8076256499133452e-06,
+      "loss": 1.5118,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 56.6037912144551,
+      "learning_rate": 2.8422876949740035e-06,
+      "loss": 1.6787,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.51296773654188,
+      "learning_rate": 2.8769497400346622e-06,
+      "loss": 1.6533,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 65.8743678692378,
+      "learning_rate": 2.9116117850953205e-06,
+      "loss": 1.4466,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.660212235000567,
+      "learning_rate": 2.9462738301559796e-06,
+      "loss": 1.5002,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.080726230549912,
+      "learning_rate": 2.980935875216638e-06,
+      "loss": 1.5862,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.809914237432615,
+      "learning_rate": 3.0155979202772966e-06,
+      "loss": 1.4045,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.38411280641693,
+      "learning_rate": 3.050259965337955e-06,
+      "loss": 1.4151,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.09621377509721,
+      "learning_rate": 3.084922010398614e-06,
+      "loss": 1.3729,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 56.80399612113042,
+      "learning_rate": 3.1195840554592723e-06,
+      "loss": 1.4924,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.03162578163297,
+      "learning_rate": 3.154246100519931e-06,
+      "loss": 1.2896,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.05831648463537,
+      "learning_rate": 3.1889081455805893e-06,
+      "loss": 1.5209,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.381577279632964,
+      "learning_rate": 3.223570190641248e-06,
+      "loss": 1.2987,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 50.67155855969887,
+      "learning_rate": 3.258232235701907e-06,
+      "loss": 1.4132,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.14601135720065,
+      "learning_rate": 3.2928942807625654e-06,
+      "loss": 1.3058,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 52.753150739630186,
+      "learning_rate": 3.327556325823224e-06,
+      "loss": 1.5529,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 41.63375309679603,
+      "learning_rate": 3.3622183708838824e-06,
+      "loss": 1.3607,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 38.060557282953845,
+      "learning_rate": 3.396880415944541e-06,
+      "loss": 1.5151,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 36.341843029383874,
+      "learning_rate": 3.4315424610051994e-06,
+      "loss": 1.4832,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 32.71183195182694,
+      "learning_rate": 3.4662045060658585e-06,
+      "loss": 1.3898,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.364513544716985,
+      "learning_rate": 3.5008665511265168e-06,
+      "loss": 1.2867,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.96477526522823,
+      "learning_rate": 3.5355285961871755e-06,
+      "loss": 1.3614,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 54.31042409667524,
+      "learning_rate": 3.5701906412478338e-06,
+      "loss": 1.3225,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.162779470946084,
+      "learning_rate": 3.6048526863084925e-06,
+      "loss": 1.4249,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.29934761904572,
+      "learning_rate": 3.6395147313691507e-06,
+      "loss": 1.2458,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.47241906103389,
+      "learning_rate": 3.67417677642981e-06,
+      "loss": 1.4181,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.256401884767726,
+      "learning_rate": 3.708838821490468e-06,
+      "loss": 1.4879,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 43.925025114520146,
+      "learning_rate": 3.743500866551127e-06,
+      "loss": 1.3717,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.95344397624258,
+      "learning_rate": 3.778162911611785e-06,
+      "loss": 1.2605,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.758083549766795,
+      "learning_rate": 3.812824956672444e-06,
+      "loss": 1.4066,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 53.9073924565535,
+      "learning_rate": 3.8474870017331025e-06,
+      "loss": 1.3514,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 41.10458204572215,
+      "learning_rate": 3.882149046793761e-06,
+      "loss": 1.2649,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 53.05955697604764,
+      "learning_rate": 3.91681109185442e-06,
+      "loss": 1.4632,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.46621436256183,
+      "learning_rate": 3.951473136915079e-06,
+      "loss": 1.4173,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.335895509505711,
+      "learning_rate": 3.986135181975737e-06,
+      "loss": 1.2472,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 53.528985041276094,
+      "learning_rate": 4.020797227036395e-06,
+      "loss": 1.3109,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.74688266972805,
+      "learning_rate": 4.055459272097054e-06,
+      "loss": 1.5135,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.26028178330096,
+      "learning_rate": 4.090121317157713e-06,
+      "loss": 1.3563,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.316768802456025,
+      "learning_rate": 4.124783362218371e-06,
+      "loss": 1.3292,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.574670927485485,
+      "learning_rate": 4.15944540727903e-06,
+      "loss": 1.1467,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 45.81860769341073,
+      "learning_rate": 4.194107452339689e-06,
+      "loss": 1.3843,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 57.43559259894182,
+      "learning_rate": 4.228769497400347e-06,
+      "loss": 1.4919,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.415087089635171,
+      "learning_rate": 4.263431542461005e-06,
+      "loss": 1.1586,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.678235459508098,
+      "learning_rate": 4.298093587521664e-06,
+      "loss": 1.3752,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.934836609796854,
+      "learning_rate": 4.332755632582323e-06,
+      "loss": 1.294,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.2067769266012,
+      "learning_rate": 4.367417677642981e-06,
+      "loss": 1.2223,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.158718020945006,
+      "learning_rate": 4.40207972270364e-06,
+      "loss": 1.2844,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.88473113813146,
+      "learning_rate": 4.436741767764298e-06,
+      "loss": 1.3532,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.647074162323214,
+      "learning_rate": 4.471403812824957e-06,
+      "loss": 1.3752,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 36.23125015937893,
+      "learning_rate": 4.506065857885615e-06,
+      "loss": 1.2189,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 41.133569313035245,
+      "learning_rate": 4.540727902946274e-06,
+      "loss": 1.4709,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 36.33225915991336,
+      "learning_rate": 4.575389948006933e-06,
+      "loss": 1.4186,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.414972508619215,
+      "learning_rate": 4.6100519930675915e-06,
+      "loss": 1.3723,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.35428797151585,
+      "learning_rate": 4.64471403812825e-06,
+      "loss": 1.3848,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.34666713235231,
+      "learning_rate": 4.679376083188908e-06,
+      "loss": 1.3439,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.078741925934583,
+      "learning_rate": 4.714038128249568e-06,
+      "loss": 1.3974,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.90329138232109,
+      "learning_rate": 4.7487001733102254e-06,
+      "loss": 1.1737,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.292796210039388,
+      "learning_rate": 4.783362218370884e-06,
+      "loss": 1.3312,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.7018850322623,
+      "learning_rate": 4.818024263431543e-06,
+      "loss": 1.2078,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 45.284279681800335,
+      "learning_rate": 4.8526863084922016e-06,
+      "loss": 1.2895,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 51.891835734719514,
+      "learning_rate": 4.88734835355286e-06,
+      "loss": 1.4174,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.47660619461644,
+      "learning_rate": 4.922010398613519e-06,
+      "loss": 1.2463,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.76010427665964,
+      "learning_rate": 4.956672443674177e-06,
+      "loss": 1.381,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.70995866748017,
+      "learning_rate": 4.9913344887348355e-06,
+      "loss": 1.3176,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.20985658428473,
+      "learning_rate": 5.025996533795494e-06,
+      "loss": 1.2154,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.78268072622544,
+      "learning_rate": 5.060658578856152e-06,
+      "loss": 1.3954,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.29081364869007,
+      "learning_rate": 5.095320623916812e-06,
+      "loss": 1.2138,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.82941003268194,
+      "learning_rate": 5.12998266897747e-06,
+      "loss": 1.2528,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.308513069270575,
+      "learning_rate": 5.164644714038128e-06,
+      "loss": 1.3094,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.716104024319435,
+      "learning_rate": 5.199306759098788e-06,
+      "loss": 1.197,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.569028238421126,
+      "learning_rate": 5.2339688041594464e-06,
+      "loss": 1.2321,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.039305431241836,
+      "learning_rate": 5.268630849220104e-06,
+      "loss": 1.225,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.957562751900497,
+      "learning_rate": 5.303292894280763e-06,
+      "loss": 1.2773,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.24003345089031,
+      "learning_rate": 5.337954939341422e-06,
+      "loss": 1.1255,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.444971755777517,
+      "learning_rate": 5.37261698440208e-06,
+      "loss": 1.1667,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 46.517308003223164,
+      "learning_rate": 5.407279029462738e-06,
+      "loss": 1.3639,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.84407460430838,
+      "learning_rate": 5.441941074523397e-06,
+      "loss": 1.3066,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 27.237931197168404,
+      "learning_rate": 5.4766031195840565e-06,
+      "loss": 1.2181,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.470842864262124,
+      "learning_rate": 5.511265164644714e-06,
+      "loss": 1.2999,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 51.05946680302711,
+      "learning_rate": 5.545927209705373e-06,
+      "loss": 1.4571,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.331443568035873,
+      "learning_rate": 5.580589254766031e-06,
+      "loss": 1.17,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 31.660909220397603,
+      "learning_rate": 5.6152512998266905e-06,
+      "loss": 1.1699,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 39.83129881241251,
+      "learning_rate": 5.649913344887349e-06,
+      "loss": 1.3412,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 41.582122805500425,
+      "learning_rate": 5.684575389948007e-06,
+      "loss": 1.2571,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 34.92599689727244,
+      "learning_rate": 5.719237435008666e-06,
+      "loss": 1.3948,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 35.588642926181166,
+      "learning_rate": 5.7538994800693245e-06,
+      "loss": 1.3808,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.375995972310854,
+      "learning_rate": 5.788561525129983e-06,
+      "loss": 1.2056,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.049289160851054,
+      "learning_rate": 5.823223570190641e-06,
+      "loss": 1.1249,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.980834700967822,
+      "learning_rate": 5.8578856152513006e-06,
+      "loss": 1.2299,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.81297797838982,
+      "learning_rate": 5.892547660311959e-06,
+      "loss": 1.2217,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.62331531360481,
+      "learning_rate": 5.927209705372617e-06,
+      "loss": 1.1349,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 38.67789488822055,
+      "learning_rate": 5.961871750433276e-06,
+      "loss": 1.1445,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.04467601914481,
+      "learning_rate": 5.996533795493935e-06,
+      "loss": 1.234,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 39.50498866391282,
+      "learning_rate": 6.031195840554593e-06,
+      "loss": 1.3125,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.497032026697326,
+      "learning_rate": 6.065857885615252e-06,
+      "loss": 1.3066,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 33.757574274360955,
+      "learning_rate": 6.10051993067591e-06,
+      "loss": 1.1588,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.29514032396307,
+      "learning_rate": 6.135181975736569e-06,
+      "loss": 1.2298,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 40.78940463621289,
+      "learning_rate": 6.169844020797228e-06,
+      "loss": 1.3352,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.524942823101373,
+      "learning_rate": 6.204506065857886e-06,
+      "loss": 1.1822,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.69774783822274,
+      "learning_rate": 6.239168110918545e-06,
+      "loss": 1.2495,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 41.3084844464238,
+      "learning_rate": 6.273830155979203e-06,
+      "loss": 1.1862,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 34.43259940959318,
+      "learning_rate": 6.308492201039862e-06,
+      "loss": 1.2432,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 38.24744721071966,
+      "learning_rate": 6.34315424610052e-06,
+      "loss": 1.1918,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 41.74737424574073,
+      "learning_rate": 6.377816291161179e-06,
+      "loss": 1.1328,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.256152917656,
+      "learning_rate": 6.412478336221838e-06,
+      "loss": 1.1292,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.65204682336752,
+      "learning_rate": 6.447140381282496e-06,
+      "loss": 1.3824,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 35.39689166268264,
+      "learning_rate": 6.481802426343155e-06,
+      "loss": 1.2992,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 43.78390060477098,
+      "learning_rate": 6.516464471403814e-06,
+      "loss": 1.3357,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.22914734988391,
+      "learning_rate": 6.551126516464472e-06,
+      "loss": 1.225,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.597944332714018,
+      "learning_rate": 6.585788561525131e-06,
+      "loss": 1.2756,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 419.35612647258307,
+      "learning_rate": 6.620450606585789e-06,
+      "loss": 1.2129,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 46.917815459962284,
+      "learning_rate": 6.655112651646448e-06,
+      "loss": 1.2689,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 31.56025284047614,
+      "learning_rate": 6.689774696707106e-06,
+      "loss": 1.1644,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.313856039421796,
+      "learning_rate": 6.724436741767765e-06,
+      "loss": 1.1856,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.530156384407103,
+      "learning_rate": 6.759098786828423e-06,
+      "loss": 1.2193,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.97730140906862,
+      "learning_rate": 6.793760831889082e-06,
+      "loss": 1.177,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.392897250545163,
+      "learning_rate": 6.828422876949741e-06,
+      "loss": 1.231,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.92835037394705,
+      "learning_rate": 6.863084922010399e-06,
+      "loss": 1.146,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.552219517077233,
+      "learning_rate": 6.8977469670710574e-06,
+      "loss": 1.0737,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.903193126268295,
+      "learning_rate": 6.932409012131717e-06,
+      "loss": 1.1864,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 38.42820331679975,
+      "learning_rate": 6.967071057192375e-06,
+      "loss": 1.1963,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.780253042602195,
+      "learning_rate": 7.0017331022530336e-06,
+      "loss": 1.2246,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 41.88507083311122,
+      "learning_rate": 7.036395147313691e-06,
+      "loss": 1.253,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.588234394474906,
+      "learning_rate": 7.071057192374351e-06,
+      "loss": 1.1283,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.2968956254753365,
+      "learning_rate": 7.10571923743501e-06,
+      "loss": 1.2647,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.7423780177881625,
+      "learning_rate": 7.1403812824956675e-06,
+      "loss": 1.1301,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 40.2782497605626,
+      "learning_rate": 7.175043327556327e-06,
+      "loss": 1.416,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.16224681685655,
+      "learning_rate": 7.209705372616985e-06,
+      "loss": 1.1316,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.03994387862284,
+      "learning_rate": 7.244367417677644e-06,
+      "loss": 1.0575,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 45.38313356254118,
+      "learning_rate": 7.2790294627383015e-06,
+      "loss": 1.2639,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.74022686936235,
+      "learning_rate": 7.313691507798961e-06,
+      "loss": 1.2833,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.19193833892053,
+      "learning_rate": 7.34835355285962e-06,
+      "loss": 1.271,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.62994222002236,
+      "learning_rate": 7.383015597920278e-06,
+      "loss": 1.1843,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.961412282025826,
+      "learning_rate": 7.417677642980936e-06,
+      "loss": 1.1769,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.229385803664805,
+      "learning_rate": 7.452339688041596e-06,
+      "loss": 1.1415,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 49.63053238151394,
+      "learning_rate": 7.487001733102254e-06,
+      "loss": 1.2606,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.18979464060009,
+      "learning_rate": 7.521663778162912e-06,
+      "loss": 1.0297,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.33706410383136,
+      "learning_rate": 7.55632582322357e-06,
+      "loss": 1.2468,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.466290187188605,
+      "learning_rate": 7.59098786828423e-06,
+      "loss": 1.2497,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 31.785716426582937,
+      "learning_rate": 7.625649913344888e-06,
+      "loss": 1.148,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.186234469992598,
+      "learning_rate": 7.660311958405546e-06,
+      "loss": 1.1191,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.136424431825276,
+      "learning_rate": 7.694974003466205e-06,
+      "loss": 1.1053,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.5331661009021,
+      "learning_rate": 7.729636048526865e-06,
+      "loss": 1.1752,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.825170021241153,
+      "learning_rate": 7.764298093587522e-06,
+      "loss": 1.1805,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 49.98371478649682,
+      "learning_rate": 7.79896013864818e-06,
+      "loss": 1.2036,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.76367878778229,
+      "learning_rate": 7.83362218370884e-06,
+      "loss": 1.1512,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.343654337053652,
+      "learning_rate": 7.868284228769498e-06,
+      "loss": 1.2117,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 33.74923088998774,
+      "learning_rate": 7.902946273830157e-06,
+      "loss": 1.1138,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.786056645264466,
+      "learning_rate": 7.937608318890815e-06,
+      "loss": 1.1355,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 48.2760855668748,
+      "learning_rate": 7.972270363951475e-06,
+      "loss": 1.2969,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.061946352264094,
+      "learning_rate": 8.006932409012133e-06,
+      "loss": 1.2115,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.30631954006009,
+      "learning_rate": 8.04159445407279e-06,
+      "loss": 1.0828,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.30636139112351,
+      "learning_rate": 8.076256499133448e-06,
+      "loss": 1.2671,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.827200155898385,
+      "learning_rate": 8.110918544194108e-06,
+      "loss": 1.2521,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.46692480831601,
+      "learning_rate": 8.145580589254767e-06,
+      "loss": 1.1863,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 33.95893658100235,
+      "learning_rate": 8.180242634315425e-06,
+      "loss": 1.1368,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.481549417206551,
+      "learning_rate": 8.214904679376083e-06,
+      "loss": 1.2642,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.55777034389227,
+      "learning_rate": 8.249566724436743e-06,
+      "loss": 1.1898,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 31.537696253165546,
+      "learning_rate": 8.2842287694974e-06,
+      "loss": 1.1685,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 40.98763564530311,
+      "learning_rate": 8.31889081455806e-06,
+      "loss": 1.1621,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.768020223733394,
+      "learning_rate": 8.353552859618718e-06,
+      "loss": 1.0423,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.991154264281352,
+      "learning_rate": 8.388214904679377e-06,
+      "loss": 1.1565,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 35.33399157297096,
+      "learning_rate": 8.422876949740035e-06,
+      "loss": 1.1692,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.435295857435555,
+      "learning_rate": 8.457538994800693e-06,
+      "loss": 1.0559,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.817123491290523,
+      "learning_rate": 8.492201039861353e-06,
+      "loss": 1.1404,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 35.386770881490634,
+      "learning_rate": 8.52686308492201e-06,
+      "loss": 1.2227,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 45.657356716165744,
+      "learning_rate": 8.56152512998267e-06,
+      "loss": 1.2336,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.792431903517928,
+      "learning_rate": 8.596187175043328e-06,
+      "loss": 1.1526,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 43.45146147895134,
+      "learning_rate": 8.630849220103988e-06,
+      "loss": 1.3344,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 40.00593644734271,
+      "learning_rate": 8.665511265164645e-06,
+      "loss": 1.3203,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 27.483074029490002,
+      "learning_rate": 8.700173310225303e-06,
+      "loss": 1.1013,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.86979396650612,
+      "learning_rate": 8.734835355285963e-06,
+      "loss": 1.2232,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 61.6369277483482,
+      "learning_rate": 8.769497400346622e-06,
+      "loss": 1.3066,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.427511507502366,
+      "learning_rate": 8.80415944540728e-06,
+      "loss": 1.0331,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.41941191543784,
+      "learning_rate": 8.838821490467938e-06,
+      "loss": 1.193,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.89093912658737,
+      "learning_rate": 8.873483535528596e-06,
+      "loss": 1.188,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.7963108025890575,
+      "learning_rate": 8.908145580589255e-06,
+      "loss": 1.0219,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 39.48217111225304,
+      "learning_rate": 8.942807625649913e-06,
+      "loss": 1.287,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.591255719236734,
+      "learning_rate": 8.977469670710573e-06,
+      "loss": 0.993,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.271428492673127,
+      "learning_rate": 9.01213171577123e-06,
+      "loss": 1.0541,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.960546409549913,
+      "learning_rate": 9.04679376083189e-06,
+      "loss": 1.114,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 23.37091576688421,
+      "learning_rate": 9.081455805892548e-06,
+      "loss": 1.0278,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.44645666671875,
+      "learning_rate": 9.116117850953206e-06,
+      "loss": 1.0882,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 29.394409995098258,
+      "learning_rate": 9.150779896013866e-06,
+      "loss": 1.098,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.84000688314728,
+      "learning_rate": 9.185441941074525e-06,
+      "loss": 1.1333,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 31.90842895091659,
+      "learning_rate": 9.220103986135183e-06,
+      "loss": 1.1737,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 29.000285479151675,
+      "learning_rate": 9.25476603119584e-06,
+      "loss": 1.1174,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 39.56155290117829,
+      "learning_rate": 9.2894280762565e-06,
+      "loss": 1.1217,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 133.62511300634418,
+      "learning_rate": 9.324090121317158e-06,
+      "loss": 1.2347,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.648703538269523,
+      "learning_rate": 9.358752166377816e-06,
+      "loss": 1.2441,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.332707601116592,
+      "learning_rate": 9.393414211438476e-06,
+      "loss": 1.1396,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.044852545885977,
+      "learning_rate": 9.428076256499135e-06,
+      "loss": 1.2006,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.57984200583586,
+      "learning_rate": 9.462738301559793e-06,
+      "loss": 1.1745,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 37.63920262944199,
+      "learning_rate": 9.497400346620451e-06,
+      "loss": 1.1589,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.540283066738453,
+      "learning_rate": 9.532062391681109e-06,
+      "loss": 0.986,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 39.58208785113726,
+      "learning_rate": 9.566724436741768e-06,
+      "loss": 1.2153,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.028840029223187,
+      "learning_rate": 9.601386481802428e-06,
+      "loss": 1.179,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 27.1549390118637,
+      "learning_rate": 9.636048526863086e-06,
+      "loss": 1.1244,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.940692187597406,
+      "learning_rate": 9.670710571923744e-06,
+      "loss": 1.3229,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 39.666642069642,
+      "learning_rate": 9.705372616984403e-06,
+      "loss": 1.1767,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 47.50090416624944,
+      "learning_rate": 9.740034662045061e-06,
+      "loss": 1.2691,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.807914091835052,
+      "learning_rate": 9.77469670710572e-06,
+      "loss": 1.0357,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.087099286974343,
+      "learning_rate": 9.809358752166378e-06,
+      "loss": 1.208,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.14497004182539,
+      "learning_rate": 9.844020797227038e-06,
+      "loss": 1.1656,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.562239852531384,
+      "learning_rate": 9.878682842287696e-06,
+      "loss": 1.2537,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.507637794902266,
+      "learning_rate": 9.913344887348354e-06,
+      "loss": 1.0488,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30.378105736734124,
+      "learning_rate": 9.948006932409013e-06,
+      "loss": 1.222,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 33.37978941506294,
+      "learning_rate": 9.982668977469671e-06,
+      "loss": 1.1733,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 32.412566171836765,
+      "learning_rate": 1.001733102253033e-05,
+      "loss": 1.2059,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 28.658334281858885,
+      "learning_rate": 1.0051993067590988e-05,
+      "loss": 1.1613,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.339859763827214,
+      "learning_rate": 1.0086655112651646e-05,
+      "loss": 1.2065,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.60806040298398,
+      "learning_rate": 1.0121317157712304e-05,
+      "loss": 1.1927,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 28.020750138656787,
+      "learning_rate": 1.0155979202772965e-05,
+      "loss": 1.1303,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 32.55648462321204,
+      "learning_rate": 1.0190641247833623e-05,
+      "loss": 1.2168,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 28.584010805483356,
+      "learning_rate": 1.0225303292894281e-05,
+      "loss": 1.1082,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.66448217384437,
+      "learning_rate": 1.025996533795494e-05,
+      "loss": 1.0592,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 33.284408458864455,
+      "learning_rate": 1.0294627383015599e-05,
+      "loss": 1.0929,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 31.98890294225414,
+      "learning_rate": 1.0329289428076256e-05,
+      "loss": 1.169,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.00410883601317,
+      "learning_rate": 1.0363951473136914e-05,
+      "loss": 1.058,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.910588839190368,
+      "learning_rate": 1.0398613518197575e-05,
+      "loss": 1.1519,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 36.227928446836465,
+      "learning_rate": 1.0433275563258233e-05,
+      "loss": 1.0383,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.57893303187256,
+      "learning_rate": 1.0467937608318893e-05,
+      "loss": 1.2146,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.806770479481848,
+      "learning_rate": 1.050259965337955e-05,
+      "loss": 1.1068,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.19032893913023,
+      "learning_rate": 1.0537261698440209e-05,
+      "loss": 1.0901,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 28.416192711982273,
+      "learning_rate": 1.0571923743500866e-05,
+      "loss": 1.0974,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.870990538838257,
+      "learning_rate": 1.0606585788561526e-05,
+      "loss": 1.0428,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.31028593040287,
+      "learning_rate": 1.0641247833622184e-05,
+      "loss": 1.0996,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.12638582913621,
+      "learning_rate": 1.0675909878682843e-05,
+      "loss": 1.0697,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.229273797125238,
+      "learning_rate": 1.0710571923743503e-05,
+      "loss": 1.1433,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.102329681320533,
+      "learning_rate": 1.074523396880416e-05,
+      "loss": 1.105,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.304508250537534,
+      "learning_rate": 1.0779896013864819e-05,
+      "loss": 1.1613,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 45.92040450216617,
+      "learning_rate": 1.0814558058925477e-05,
+      "loss": 1.1904,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.0916480782526,
+      "learning_rate": 1.0849220103986136e-05,
+      "loss": 1.0855,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 43.67952947804621,
+      "learning_rate": 1.0883882149046794e-05,
+      "loss": 1.3718,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.554504791956532,
+      "learning_rate": 1.0918544194107452e-05,
+      "loss": 0.9619,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.608862729388665,
+      "learning_rate": 1.0953206239168113e-05,
+      "loss": 1.0781,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 41.32110408068733,
+      "learning_rate": 1.0987868284228771e-05,
+      "loss": 1.1264,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.537403696445477,
+      "learning_rate": 1.1022530329289429e-05,
+      "loss": 1.1479,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.88435630109961,
+      "learning_rate": 1.1057192374350088e-05,
+      "loss": 1.1415,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.390235825883956,
+      "learning_rate": 1.1091854419410746e-05,
+      "loss": 1.0666,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.016075798543312,
+      "learning_rate": 1.1126516464471404e-05,
+      "loss": 1.0004,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.738052474932843,
+      "learning_rate": 1.1161178509532062e-05,
+      "loss": 1.0154,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 38.4046694746161,
+      "learning_rate": 1.1195840554592723e-05,
+      "loss": 1.2249,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.28829747954234,
+      "learning_rate": 1.1230502599653381e-05,
+      "loss": 1.1102,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.448439893190514,
+      "learning_rate": 1.1265164644714039e-05,
+      "loss": 1.0582,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.83991091682964,
+      "learning_rate": 1.1299826689774698e-05,
+      "loss": 1.1886,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 36.654051427770966,
+      "learning_rate": 1.1334488734835356e-05,
+      "loss": 1.1055,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.0320110581874,
+      "learning_rate": 1.1369150779896014e-05,
+      "loss": 0.9618,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 33.79684069416141,
+      "learning_rate": 1.1403812824956672e-05,
+      "loss": 1.2244,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.228386863544692,
+      "learning_rate": 1.1438474870017332e-05,
+      "loss": 1.077,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 31.836119884132362,
+      "learning_rate": 1.1473136915077991e-05,
+      "loss": 1.0394,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.110325038671736,
+      "learning_rate": 1.1507798960138649e-05,
+      "loss": 1.2516,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.72559738791931,
+      "learning_rate": 1.1542461005199308e-05,
+      "loss": 1.1126,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.478969825496325,
+      "learning_rate": 1.1577123050259966e-05,
+      "loss": 0.9767,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.48757251280962,
+      "learning_rate": 1.1611785095320624e-05,
+      "loss": 1.2302,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 43.73903702776825,
+      "learning_rate": 1.1646447140381282e-05,
+      "loss": 1.153,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 36.286243693056086,
+      "learning_rate": 1.1681109185441942e-05,
+      "loss": 1.1822,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.34518100764557,
+      "learning_rate": 1.1715771230502601e-05,
+      "loss": 1.0814,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.58957807542444,
+      "learning_rate": 1.175043327556326e-05,
+      "loss": 1.1042,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.286081019374777,
+      "learning_rate": 1.1785095320623919e-05,
+      "loss": 1.0859,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 32.51010524518503,
+      "learning_rate": 1.1819757365684576e-05,
+      "loss": 1.0435,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.181531261302148,
+      "learning_rate": 1.1854419410745234e-05,
+      "loss": 1.107,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 35.47835515156875,
+      "learning_rate": 1.1889081455805894e-05,
+      "loss": 1.0314,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 24.60361534656248,
+      "learning_rate": 1.1923743500866552e-05,
+      "loss": 1.0581,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.603447091746283,
+      "learning_rate": 1.195840554592721e-05,
+      "loss": 1.0639,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 42.005933401381604,
+      "learning_rate": 1.199306759098787e-05,
+      "loss": 1.119,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 31.325769802994927,
+      "learning_rate": 1.2027729636048529e-05,
+      "loss": 1.0979,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 46.88281582331888,
+      "learning_rate": 1.2062391681109186e-05,
+      "loss": 1.1392,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 31.996595149395624,
+      "learning_rate": 1.2097053726169844e-05,
+      "loss": 1.0662,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 32.306233032599955,
+      "learning_rate": 1.2131715771230504e-05,
+      "loss": 1.0391,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 33.385801572063706,
+      "learning_rate": 1.2166377816291162e-05,
+      "loss": 1.2242,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.43179488371909,
+      "learning_rate": 1.220103986135182e-05,
+      "loss": 1.1773,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 33.50602676407605,
+      "learning_rate": 1.2235701906412477e-05,
+      "loss": 1.0898,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.631638215145294,
+      "learning_rate": 1.2270363951473139e-05,
+      "loss": 1.0632,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 25.926041056193505,
+      "learning_rate": 1.2305025996533797e-05,
+      "loss": 0.9388,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.932012060913973,
+      "learning_rate": 1.2339688041594456e-05,
+      "loss": 1.2701,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.65525816696688,
+      "learning_rate": 1.2374350086655114e-05,
+      "loss": 0.9718,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.096018642350085,
+      "learning_rate": 1.2409012131715772e-05,
+      "loss": 0.9923,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.929265322745625,
+      "learning_rate": 1.244367417677643e-05,
+      "loss": 1.1121,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.227952983086723,
+      "learning_rate": 1.247833622183709e-05,
+      "loss": 1.1241,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.766410493745894,
+      "learning_rate": 1.2512998266897749e-05,
+      "loss": 1.1074,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.994232906235982,
+      "learning_rate": 1.2547660311958407e-05,
+      "loss": 1.1329,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.510957715316916,
+      "learning_rate": 1.2582322357019066e-05,
+      "loss": 1.0699,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.935369772262348,
+      "learning_rate": 1.2616984402079724e-05,
+      "loss": 1.1374,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 36.271485986247896,
+      "learning_rate": 1.2651646447140382e-05,
+      "loss": 1.0808,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.99020265257391,
+      "learning_rate": 1.268630849220104e-05,
+      "loss": 1.0781,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.36522216302166,
+      "learning_rate": 1.27209705372617e-05,
+      "loss": 1.0592,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 31.96383324342438,
+      "learning_rate": 1.2755632582322357e-05,
+      "loss": 0.9815,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 33.51334153868519,
+      "learning_rate": 1.2790294627383017e-05,
+      "loss": 1.1268,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.693184440634166,
+      "learning_rate": 1.2824956672443676e-05,
+      "loss": 1.0338,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.48413772334245,
+      "learning_rate": 1.2859618717504334e-05,
+      "loss": 1.2143,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.667324067868737,
+      "learning_rate": 1.2894280762564992e-05,
+      "loss": 1.0773,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 32.613217908327364,
+      "learning_rate": 1.2928942807625652e-05,
+      "loss": 1.0994,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 25.86133089764893,
+      "learning_rate": 1.296360485268631e-05,
+      "loss": 1.0735,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 31.189272337112506,
+      "learning_rate": 1.2998266897746967e-05,
+      "loss": 1.1056,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.21329440551253,
+      "learning_rate": 1.3032928942807628e-05,
+      "loss": 1.0915,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 29.346472574260808,
+      "learning_rate": 1.3067590987868286e-05,
+      "loss": 1.1536,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 45.79309598200069,
+      "learning_rate": 1.3102253032928944e-05,
+      "loss": 1.1109,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 29.733980734491936,
+      "learning_rate": 1.3136915077989602e-05,
+      "loss": 1.0806,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.973894157979053,
+      "learning_rate": 1.3171577123050262e-05,
+      "loss": 1.093,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.54901906636627,
+      "learning_rate": 1.320623916811092e-05,
+      "loss": 1.0221,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.908550257544434,
+      "learning_rate": 1.3240901213171577e-05,
+      "loss": 1.0472,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.727359550673427,
+      "learning_rate": 1.3275563258232235e-05,
+      "loss": 1.0048,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 29.224659504713923,
+      "learning_rate": 1.3310225303292896e-05,
+      "loss": 1.1251,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.47984058913266,
+      "learning_rate": 1.3344887348353554e-05,
+      "loss": 1.0694,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.247752152418364,
+      "learning_rate": 1.3379549393414212e-05,
+      "loss": 1.2244,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.877945774559905,
+      "learning_rate": 1.3414211438474872e-05,
+      "loss": 1.0694,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.140336946003078,
+      "learning_rate": 1.344887348353553e-05,
+      "loss": 1.2015,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.9027450262441,
+      "learning_rate": 1.3483535528596187e-05,
+      "loss": 1.086,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.171382283179494,
+      "learning_rate": 1.3518197573656845e-05,
+      "loss": 1.1288,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.796142295773222,
+      "learning_rate": 1.3552859618717506e-05,
+      "loss": 1.1133,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.9444241544439,
+      "learning_rate": 1.3587521663778164e-05,
+      "loss": 1.1072,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.169641143832457,
+      "learning_rate": 1.3622183708838824e-05,
+      "loss": 1.0657,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.603979212792716,
+      "learning_rate": 1.3656845753899482e-05,
+      "loss": 1.2969,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 32.377532200853615,
+      "learning_rate": 1.369150779896014e-05,
+      "loss": 1.0666,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.266813074204492,
+      "learning_rate": 1.3726169844020797e-05,
+      "loss": 1.0658,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.32559842190941,
+      "learning_rate": 1.3760831889081457e-05,
+      "loss": 1.0985,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.371715400867,
+      "learning_rate": 1.3795493934142115e-05,
+      "loss": 1.3105,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 41.63190000424927,
+      "learning_rate": 1.3830155979202774e-05,
+      "loss": 1.0433,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.90569959847415,
+      "learning_rate": 1.3864818024263434e-05,
+      "loss": 1.0908,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.6998805764918,
+      "learning_rate": 1.3899480069324092e-05,
+      "loss": 1.0031,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.075689023941184,
+      "learning_rate": 1.393414211438475e-05,
+      "loss": 1.0568,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.3133373278367,
+      "learning_rate": 1.3968804159445408e-05,
+      "loss": 1.1699,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.4763675646989,
+      "learning_rate": 1.4003466204506067e-05,
+      "loss": 1.2245,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 31.103725701166116,
+      "learning_rate": 1.4038128249566725e-05,
+      "loss": 1.0939,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.757187190093457,
+      "learning_rate": 1.4072790294627383e-05,
+      "loss": 1.0749,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.835036799977743,
+      "learning_rate": 1.4107452339688044e-05,
+      "loss": 1.031,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 42.772867923933184,
+      "learning_rate": 1.4142114384748702e-05,
+      "loss": 1.1827,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.04203269843442,
+      "learning_rate": 1.417677642980936e-05,
+      "loss": 1.2603,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.941152088039125,
+      "learning_rate": 1.421143847487002e-05,
+      "loss": 1.0402,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.012488225595373,
+      "learning_rate": 1.4246100519930677e-05,
+      "loss": 1.0436,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.68960934660663,
+      "learning_rate": 1.4280762564991335e-05,
+      "loss": 1.0056,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.378956623070927,
+      "learning_rate": 1.4315424610051993e-05,
+      "loss": 1.1198,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.51518555812069,
+      "learning_rate": 1.4350086655112654e-05,
+      "loss": 1.1016,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.430786335206683,
+      "learning_rate": 1.4384748700173312e-05,
+      "loss": 1.0082,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.425432187951778,
+      "learning_rate": 1.441941074523397e-05,
+      "loss": 1.0608,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.56158873892592,
+      "learning_rate": 1.445407279029463e-05,
+      "loss": 1.0575,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.993745699464292,
+      "learning_rate": 1.4488734835355287e-05,
+      "loss": 1.1353,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 33.698564188109685,
+      "learning_rate": 1.4523396880415945e-05,
+      "loss": 1.1354,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.738939505418845,
+      "learning_rate": 1.4558058925476603e-05,
+      "loss": 1.0725,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 30.872998443969035,
+      "learning_rate": 1.4592720970537263e-05,
+      "loss": 1.1156,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.35365116093291,
+      "learning_rate": 1.4627383015597922e-05,
+      "loss": 1.0658,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.853792234356817,
+      "learning_rate": 1.466204506065858e-05,
+      "loss": 1.0421,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 32.3637560867717,
+      "learning_rate": 1.469670710571924e-05,
+      "loss": 1.13,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.375143306873333,
+      "learning_rate": 1.4731369150779897e-05,
+      "loss": 1.1187,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.315598036728183,
+      "learning_rate": 1.4766031195840555e-05,
+      "loss": 1.0749,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 31.578187560218176,
+      "learning_rate": 1.4800693240901213e-05,
+      "loss": 1.0663,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 29.277862572733177,
+      "learning_rate": 1.4835355285961873e-05,
+      "loss": 1.0366,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.773854849425298,
+      "learning_rate": 1.4870017331022532e-05,
+      "loss": 1.0361,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 30.07149383967753,
+      "learning_rate": 1.4904679376083192e-05,
+      "loss": 1.1342,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 28.905491040550434,
+      "learning_rate": 1.493934142114385e-05,
+      "loss": 1.2152,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.66915981630242,
+      "learning_rate": 1.4974003466204507e-05,
+      "loss": 1.0779,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 31.81603142861209,
+      "learning_rate": 1.5008665511265165e-05,
+      "loss": 1.0872,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 34.11991082135092,
+      "learning_rate": 1.5043327556325825e-05,
+      "loss": 0.9974,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.143410269225893,
+      "learning_rate": 1.5077989601386483e-05,
+      "loss": 1.1808,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.641715173600016,
+      "learning_rate": 1.511265164644714e-05,
+      "loss": 1.1308,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 33.98886923001349,
+      "learning_rate": 1.5147313691507802e-05,
+      "loss": 1.1661,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.572967687726788,
+      "learning_rate": 1.518197573656846e-05,
+      "loss": 1.0132,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.362442319213617,
+      "learning_rate": 1.5216637781629117e-05,
+      "loss": 1.0149,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.28730137556571,
+      "learning_rate": 1.5251299826689775e-05,
+      "loss": 1.0923,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.457901932680777,
+      "learning_rate": 1.5285961871750435e-05,
+      "loss": 0.9266,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.42685430960279,
+      "learning_rate": 1.532062391681109e-05,
+      "loss": 1.0687,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.110720276179347,
+      "learning_rate": 1.535528596187175e-05,
+      "loss": 1.022,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.072955662218703,
+      "learning_rate": 1.538994800693241e-05,
+      "loss": 1.0953,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.11605472079033,
+      "learning_rate": 1.542461005199307e-05,
+      "loss": 0.9869,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 34.431652101331636,
+      "learning_rate": 1.545927209705373e-05,
+      "loss": 1.3131,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 33.46294148670095,
+      "learning_rate": 1.5493934142114385e-05,
+      "loss": 0.9587,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.8286323566764,
+      "learning_rate": 1.5528596187175045e-05,
+      "loss": 1.1004,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.15366308486822,
+      "learning_rate": 1.5563258232235705e-05,
+      "loss": 0.9616,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.821871415050428,
+      "learning_rate": 1.559792027729636e-05,
+      "loss": 0.8986,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 35.99595845901309,
+      "learning_rate": 1.563258232235702e-05,
+      "loss": 1.2051,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.096743531257545,
+      "learning_rate": 1.566724436741768e-05,
+      "loss": 0.9922,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.701941811616413,
+      "learning_rate": 1.570190641247834e-05,
+      "loss": 0.9759,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.0741338296344,
+      "learning_rate": 1.5736568457538996e-05,
+      "loss": 1.0567,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2711761486261555,
+      "learning_rate": 1.5771230502599655e-05,
+      "loss": 1.1068,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.120557194163602,
+      "learning_rate": 1.5805892547660315e-05,
+      "loss": 0.9328,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.54513505382531,
+      "learning_rate": 1.584055459272097e-05,
+      "loss": 1.1424,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 29.169606155817124,
+      "learning_rate": 1.587521663778163e-05,
+      "loss": 0.9743,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.73216498446794,
+      "learning_rate": 1.5909878682842286e-05,
+      "loss": 0.9859,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.04924244626764,
+      "learning_rate": 1.594454072790295e-05,
+      "loss": 1.0824,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.32800409759019,
+      "learning_rate": 1.5979202772963606e-05,
+      "loss": 1.0437,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.947383935631013,
+      "learning_rate": 1.6013864818024265e-05,
+      "loss": 1.0657,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 42.90413702383083,
+      "learning_rate": 1.6048526863084925e-05,
+      "loss": 1.1631,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.11730306306142,
+      "learning_rate": 1.608318890814558e-05,
+      "loss": 1.2399,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.04111791854184,
+      "learning_rate": 1.611785095320624e-05,
+      "loss": 0.952,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.07962514994669,
+      "learning_rate": 1.6152512998266897e-05,
+      "loss": 0.9327,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.38630126346466,
+      "learning_rate": 1.618717504332756e-05,
+      "loss": 1.0544,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 32.06929738788105,
+      "learning_rate": 1.6221837088388216e-05,
+      "loss": 1.1331,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.664537084714688,
+      "learning_rate": 1.6256499133448875e-05,
+      "loss": 0.9988,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.206259630000375,
+      "learning_rate": 1.6291161178509535e-05,
+      "loss": 1.0925,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 36.306049239295874,
+      "learning_rate": 1.632582322357019e-05,
+      "loss": 1.0819,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.65730606476609,
+      "learning_rate": 1.636048526863085e-05,
+      "loss": 1.0467,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 31.947233780879255,
+      "learning_rate": 1.639514731369151e-05,
+      "loss": 1.0235,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.346397050271698,
+      "learning_rate": 1.6429809358752166e-05,
+      "loss": 0.9507,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 32.20872637450095,
+      "learning_rate": 1.6464471403812826e-05,
+      "loss": 1.1141,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 40.187479039923865,
+      "learning_rate": 1.6499133448873485e-05,
+      "loss": 1.0833,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.63407629745231,
+      "learning_rate": 1.6533795493934145e-05,
+      "loss": 1.0913,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.55700920799157,
+      "learning_rate": 1.65684575389948e-05,
+      "loss": 1.0217,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.303006357296574,
+      "learning_rate": 1.660311958405546e-05,
+      "loss": 0.9185,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 34.83074933968313,
+      "learning_rate": 1.663778162911612e-05,
+      "loss": 1.1405,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.905102922878694,
+      "learning_rate": 1.6672443674176776e-05,
+      "loss": 1.0552,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.5374601706932,
+      "learning_rate": 1.6707105719237436e-05,
+      "loss": 1.0506,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 30.868805725347034,
+      "learning_rate": 1.6741767764298095e-05,
+      "loss": 1.2388,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 31.809013872207533,
+      "learning_rate": 1.6776429809358755e-05,
+      "loss": 1.1573,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.584330183913444,
+      "learning_rate": 1.681109185441941e-05,
+      "loss": 1.0159,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.359652479793855,
+      "learning_rate": 1.684575389948007e-05,
+      "loss": 0.9319,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.035349084560117,
+      "learning_rate": 1.688041594454073e-05,
+      "loss": 1.192,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.091984735728435,
+      "learning_rate": 1.6915077989601386e-05,
+      "loss": 0.993,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.26835354784831,
+      "learning_rate": 1.6949740034662046e-05,
+      "loss": 1.1226,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.006399020570456,
+      "learning_rate": 1.6984402079722705e-05,
+      "loss": 0.9572,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.67672143114982,
+      "learning_rate": 1.7019064124783365e-05,
+      "loss": 1.0041,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 35.59097800039987,
+      "learning_rate": 1.705372616984402e-05,
+      "loss": 1.0966,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.007039133085687,
+      "learning_rate": 1.708838821490468e-05,
+      "loss": 1.0351,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.6685605548639,
+      "learning_rate": 1.712305025996534e-05,
+      "loss": 0.9102,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.69384850275509,
+      "learning_rate": 1.7157712305025996e-05,
+      "loss": 1.0426,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 29.480033542751137,
+      "learning_rate": 1.7192374350086656e-05,
+      "loss": 1.0882,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 40.07920094211409,
+      "learning_rate": 1.7227036395147316e-05,
+      "loss": 1.0148,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 29.233060679818937,
+      "learning_rate": 1.7261698440207975e-05,
+      "loss": 1.0744,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 35.182566386669606,
+      "learning_rate": 1.729636048526863e-05,
+      "loss": 0.9997,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 30.269976589123253,
+      "learning_rate": 1.733102253032929e-05,
+      "loss": 1.1004,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.245753203984723,
+      "learning_rate": 1.736568457538995e-05,
+      "loss": 1.0803,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 39.04873017456366,
+      "learning_rate": 1.7400346620450606e-05,
+      "loss": 1.1467,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.792476515138087,
+      "learning_rate": 1.7435008665511266e-05,
+      "loss": 1.0947,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 30.2379304968761,
+      "learning_rate": 1.7469670710571926e-05,
+      "loss": 1.1267,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 33.65685682891974,
+      "learning_rate": 1.7504332755632585e-05,
+      "loss": 1.1169,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.56601044270328,
+      "learning_rate": 1.7538994800693245e-05,
+      "loss": 1.0628,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.60060010011299,
+      "learning_rate": 1.75736568457539e-05,
+      "loss": 1.0078,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.11059699605151,
+      "learning_rate": 1.760831889081456e-05,
+      "loss": 1.046,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 34.78563216246366,
+      "learning_rate": 1.7642980935875217e-05,
+      "loss": 1.2045,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 27.789317608451398,
+      "learning_rate": 1.7677642980935876e-05,
+      "loss": 0.9659,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 27.727447236338673,
+      "learning_rate": 1.7712305025996536e-05,
+      "loss": 1.0327,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.80408909837495,
+      "learning_rate": 1.7746967071057192e-05,
+      "loss": 1.0152,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.355020428703057,
+      "learning_rate": 1.7781629116117855e-05,
+      "loss": 1.0644,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.682545495976207,
+      "learning_rate": 1.781629116117851e-05,
+      "loss": 1.0329,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.989554582298343,
+      "learning_rate": 1.785095320623917e-05,
+      "loss": 1.1096,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.79673932655621,
+      "learning_rate": 1.7885615251299827e-05,
+      "loss": 1.1577,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.84256013993326,
+      "learning_rate": 1.7920277296360486e-05,
+      "loss": 1.0689,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.533577428342877,
+      "learning_rate": 1.7954939341421146e-05,
+      "loss": 0.8793,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.370691730497956,
+      "learning_rate": 1.7989601386481802e-05,
+      "loss": 1.0736,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.610226379684057,
+      "learning_rate": 1.802426343154246e-05,
+      "loss": 0.987,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.44208765087892,
+      "learning_rate": 1.805892547660312e-05,
+      "loss": 1.0504,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.144404591753734,
+      "learning_rate": 1.809358752166378e-05,
+      "loss": 1.0604,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.877103069616282,
+      "learning_rate": 1.812824956672444e-05,
+      "loss": 0.9898,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.737747217762195,
+      "learning_rate": 1.8162911611785096e-05,
+      "loss": 1.0038,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.25951547791112,
+      "learning_rate": 1.8197573656845756e-05,
+      "loss": 1.0588,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 27.599355801073518,
+      "learning_rate": 1.8232235701906412e-05,
+      "loss": 0.9713,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 33.122993457048636,
+      "learning_rate": 1.826689774696707e-05,
+      "loss": 1.0201,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.223397469833518,
+      "learning_rate": 1.830155979202773e-05,
+      "loss": 1.0946,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 29.740458575239316,
+      "learning_rate": 1.833622183708839e-05,
+      "loss": 0.9838,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.341621904414357,
+      "learning_rate": 1.837088388214905e-05,
+      "loss": 1.0443,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.34550028831831,
+      "learning_rate": 1.8405545927209706e-05,
+      "loss": 1.0417,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.180889743363394,
+      "learning_rate": 1.8440207972270366e-05,
+      "loss": 1.1049,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.271695696375218,
+      "learning_rate": 1.8474870017331022e-05,
+      "loss": 1.0915,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.83622464957578,
+      "learning_rate": 1.850953206239168e-05,
+      "loss": 1.1916,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.2894505708719,
+      "learning_rate": 1.854419410745234e-05,
+      "loss": 1.073,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 28.2656482092286,
+      "learning_rate": 1.8578856152513e-05,
+      "loss": 1.0285,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.59298381914133,
+      "learning_rate": 1.861351819757366e-05,
+      "loss": 1.1132,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.60061650304393,
+      "learning_rate": 1.8648180242634316e-05,
+      "loss": 1.1276,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 35.98301939371637,
+      "learning_rate": 1.8682842287694976e-05,
+      "loss": 1.159,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4888981427578205,
+      "learning_rate": 1.8717504332755632e-05,
+      "loss": 1.0028,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 36.99781273324386,
+      "learning_rate": 1.8752166377816292e-05,
+      "loss": 1.1377,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.939578263032633,
+      "learning_rate": 1.878682842287695e-05,
+      "loss": 0.999,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.85660085708705,
+      "learning_rate": 1.882149046793761e-05,
+      "loss": 1.0134,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.241163366830648,
+      "learning_rate": 1.885615251299827e-05,
+      "loss": 1.0081,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.474464093786764,
+      "learning_rate": 1.8890814558058927e-05,
+      "loss": 1.1213,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 33.11335194940052,
+      "learning_rate": 1.8925476603119586e-05,
+      "loss": 1.0378,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 31.546985247203498,
+      "learning_rate": 1.8960138648180246e-05,
+      "loss": 1.0594,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.353861552798367,
+      "learning_rate": 1.8994800693240902e-05,
+      "loss": 1.0908,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.9522994570028,
+      "learning_rate": 1.902946273830156e-05,
+      "loss": 0.9971,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 26.431519891655817,
+      "learning_rate": 1.9064124783362217e-05,
+      "loss": 1.1073,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.854707471528116,
+      "learning_rate": 1.909878682842288e-05,
+      "loss": 1.0473,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.821719178716894,
+      "learning_rate": 1.9133448873483537e-05,
+      "loss": 0.987,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.670015765342832,
+      "learning_rate": 1.9168110918544196e-05,
+      "loss": 1.0873,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.380559817983787,
+      "learning_rate": 1.9202772963604856e-05,
+      "loss": 0.9886,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.144432556811733,
+      "learning_rate": 1.9237435008665512e-05,
+      "loss": 0.9142,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 26.895634272720514,
+      "learning_rate": 1.927209705372617e-05,
+      "loss": 1.1352,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.938473895183744,
+      "learning_rate": 1.9306759098786828e-05,
+      "loss": 1.1945,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.095345260358535,
+      "learning_rate": 1.9341421143847487e-05,
+      "loss": 1.0434,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 41.15274760623393,
+      "learning_rate": 1.9376083188908147e-05,
+      "loss": 1.1775,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.886517844676984,
+      "learning_rate": 1.9410745233968806e-05,
+      "loss": 1.0895,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.219768592031212,
+      "learning_rate": 1.9445407279029466e-05,
+      "loss": 1.022,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.961321974964306,
+      "learning_rate": 1.9480069324090122e-05,
+      "loss": 1.0518,
+      "step": 562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 32.005822631718004,
+      "learning_rate": 1.951473136915078e-05,
+      "loss": 1.0212,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.961686493739755,
+      "learning_rate": 1.954939341421144e-05,
+      "loss": 1.0504,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.25877929944018,
+      "learning_rate": 1.9584055459272097e-05,
+      "loss": 1.0929,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.163975500572413,
+      "learning_rate": 1.9618717504332757e-05,
+      "loss": 1.0706,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 26.611705318472513,
+      "learning_rate": 1.9653379549393416e-05,
+      "loss": 0.9836,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 28.184492445566853,
+      "learning_rate": 1.9688041594454076e-05,
+      "loss": 0.9916,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.68156266800803,
+      "learning_rate": 1.9722703639514732e-05,
+      "loss": 1.1573,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 28.378879126518083,
+      "learning_rate": 1.975736568457539e-05,
+      "loss": 0.9497,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.356736233473026,
+      "learning_rate": 1.979202772963605e-05,
+      "loss": 0.9251,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.592310467899786,
+      "learning_rate": 1.9826689774696707e-05,
+      "loss": 1.0297,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.14714174210939,
+      "learning_rate": 1.9861351819757367e-05,
+      "loss": 1.0315,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 28.635279627490927,
+      "learning_rate": 1.9896013864818026e-05,
+      "loss": 1.0889,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.14251563807322,
+      "learning_rate": 1.9930675909878686e-05,
+      "loss": 1.0212,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.605607340399068,
+      "learning_rate": 1.9965337954939342e-05,
+      "loss": 1.055,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.861064176734494,
+      "learning_rate": 2e-05,
+      "loss": 0.9411,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 32.708221310020775,
+      "learning_rate": 1.9999999857802926e-05,
+      "loss": 1.0222,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.77872904734455,
+      "learning_rate": 1.9999999431211706e-05,
+      "loss": 1.0681,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.068217319997437,
+      "learning_rate": 1.999999872022635e-05,
+      "loss": 1.1066,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 31.490787994547425,
+      "learning_rate": 1.9999997724846883e-05,
+      "loss": 1.0383,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 34.98645871334549,
+      "learning_rate": 1.9999996445073327e-05,
+      "loss": 1.1206,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 35.632614390703424,
+      "learning_rate": 1.9999994880905726e-05,
+      "loss": 1.0542,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.23781468122964,
+      "learning_rate": 1.9999993032344115e-05,
+      "loss": 1.0387,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.61966249849483,
+      "learning_rate": 1.9999990899388556e-05,
+      "loss": 1.0255,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.504192646772623,
+      "learning_rate": 1.9999988482039104e-05,
+      "loss": 1.0037,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.648244447373916,
+      "learning_rate": 1.999998578029583e-05,
+      "loss": 1.2731,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.92577731908766,
+      "learning_rate": 1.999998279415881e-05,
+      "loss": 1.0188,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.714237684191083,
+      "learning_rate": 1.999997952362813e-05,
+      "loss": 1.0955,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.875496663575905,
+      "learning_rate": 1.999997596870388e-05,
+      "loss": 0.9669,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.41305678438287,
+      "learning_rate": 1.9999972129386165e-05,
+      "loss": 0.9065,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.794859932565107,
+      "learning_rate": 1.999996800567509e-05,
+      "loss": 1.0044,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.429124072996856,
+      "learning_rate": 1.999996359757078e-05,
+      "loss": 1.0774,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.424359437948585,
+      "learning_rate": 1.9999958905073352e-05,
+      "loss": 1.0994,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.345026243969027,
+      "learning_rate": 1.9999953928182942e-05,
+      "loss": 1.0255,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 30.414133266324512,
+      "learning_rate": 1.9999948666899695e-05,
+      "loss": 1.0921,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.384991495585798,
+      "learning_rate": 1.9999943121223753e-05,
+      "loss": 1.1104,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.39591885961205,
+      "learning_rate": 1.999993729115528e-05,
+      "loss": 1.0308,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 36.55302360842166,
+      "learning_rate": 1.999993117669444e-05,
+      "loss": 1.1033,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.528707666170977,
+      "learning_rate": 1.999992477784141e-05,
+      "loss": 1.002,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 39.0482325048875,
+      "learning_rate": 1.999991809459637e-05,
+      "loss": 0.9924,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 31.36379419691972,
+      "learning_rate": 1.9999911126959503e-05,
+      "loss": 1.0878,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.230094375873414,
+      "learning_rate": 1.9999903874931017e-05,
+      "loss": 0.9643,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.097369602337533,
+      "learning_rate": 1.9999896338511117e-05,
+      "loss": 1.0073,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.538517270030404,
+      "learning_rate": 1.999988851770001e-05,
+      "loss": 1.0647,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.087550676595903,
+      "learning_rate": 1.9999880412497927e-05,
+      "loss": 1.1512,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.644576876286553,
+      "learning_rate": 1.9999872022905094e-05,
+      "loss": 1.0329,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.57077115583161,
+      "learning_rate": 1.9999863348921748e-05,
+      "loss": 1.0519,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.806895674814548,
+      "learning_rate": 1.999985439054814e-05,
+      "loss": 1.0923,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.1697596113921,
+      "learning_rate": 1.9999845147784526e-05,
+      "loss": 1.0938,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 34.984784936143015,
+      "learning_rate": 1.999983562063116e-05,
+      "loss": 1.0604,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.31641290786309,
+      "learning_rate": 1.999982580908832e-05,
+      "loss": 1.0975,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.700540192601025,
+      "learning_rate": 1.9999815713156285e-05,
+      "loss": 1.1786,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.784075607083317,
+      "learning_rate": 1.9999805332835344e-05,
+      "loss": 1.1711,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.566909129873377,
+      "learning_rate": 1.9999794668125784e-05,
+      "loss": 0.979,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 32.344774729394906,
+      "learning_rate": 1.9999783719027913e-05,
+      "loss": 1.0891,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.350961898141318,
+      "learning_rate": 1.9999772485542048e-05,
+      "loss": 1.0251,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 30.323243079901548,
+      "learning_rate": 1.99997609676685e-05,
+      "loss": 1.0448,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 26.559370164481408,
+      "learning_rate": 1.99997491654076e-05,
+      "loss": 1.019,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.116834511656986,
+      "learning_rate": 1.9999737078759684e-05,
+      "loss": 1.0316,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 38.23014262882626,
+      "learning_rate": 1.999972470772509e-05,
+      "loss": 1.0279,
+      "step": 621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.264683982109712,
+      "learning_rate": 1.999971205230418e-05,
+      "loss": 1.0609,
+      "step": 622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 37.16865727262475,
+      "learning_rate": 1.999969911249731e-05,
+      "loss": 0.9348,
+      "step": 623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.669900577725162,
+      "learning_rate": 1.9999685888304844e-05,
+      "loss": 1.0294,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.08810207445628,
+      "learning_rate": 1.9999672379727165e-05,
+      "loss": 1.0945,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.214393464839559,
+      "learning_rate": 1.999965858676465e-05,
+      "loss": 0.9841,
+      "step": 626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.103198322971277,
+      "learning_rate": 1.9999644509417694e-05,
+      "loss": 1.0875,
+      "step": 627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.915001681033875,
+      "learning_rate": 1.99996301476867e-05,
+      "loss": 0.9758,
+      "step": 628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.30079704900265,
+      "learning_rate": 1.9999615501572073e-05,
+      "loss": 1.0322,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 38.04965149385254,
+      "learning_rate": 1.999960057107423e-05,
+      "loss": 1.0074,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.414347805622835,
+      "learning_rate": 1.9999585356193597e-05,
+      "loss": 0.9241,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.90992076557785,
+      "learning_rate": 1.9999569856930604e-05,
+      "loss": 1.0194,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.110853662612588,
+      "learning_rate": 1.9999554073285695e-05,
+      "loss": 1.0284,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.778630374216906,
+      "learning_rate": 1.999953800525932e-05,
+      "loss": 1.1036,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 24.904941753025636,
+      "learning_rate": 1.9999521652851933e-05,
+      "loss": 0.9953,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.34261364640905,
+      "learning_rate": 1.9999505016063998e-05,
+      "loss": 1.012,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.114153859611994,
+      "learning_rate": 1.9999488094895992e-05,
+      "loss": 0.9224,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.201733390049078,
+      "learning_rate": 1.9999470889348394e-05,
+      "loss": 1.0225,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.33693891034599,
+      "learning_rate": 1.9999453399421692e-05,
+      "loss": 0.9935,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.755063072714822,
+      "learning_rate": 1.9999435625116383e-05,
+      "loss": 0.8937,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.135516419968262,
+      "learning_rate": 1.999941756643298e-05,
+      "loss": 1.0474,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.210619093269454,
+      "learning_rate": 1.9999399223371988e-05,
+      "loss": 0.9777,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.874397763131537,
+      "learning_rate": 1.999938059593393e-05,
+      "loss": 1.0197,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.09541794840785,
+      "learning_rate": 1.999936168411934e-05,
+      "loss": 1.0258,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.123838546802762,
+      "learning_rate": 1.999934248792875e-05,
+      "loss": 0.9936,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.733233766116115,
+      "learning_rate": 1.9999323007362708e-05,
+      "loss": 0.9803,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 26.073300094137274,
+      "learning_rate": 1.9999303242421773e-05,
+      "loss": 0.9174,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.286722855667904,
+      "learning_rate": 1.9999283193106504e-05,
+      "loss": 0.944,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 32.56092169939891,
+      "learning_rate": 1.9999262859417466e-05,
+      "loss": 1.019,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.95125505355276,
+      "learning_rate": 1.9999242241355247e-05,
+      "loss": 0.8869,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 32.132816929611685,
+      "learning_rate": 1.9999221338920424e-05,
+      "loss": 1.1021,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.434045022076624,
+      "learning_rate": 1.9999200152113595e-05,
+      "loss": 0.938,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 32.50571419843277,
+      "learning_rate": 1.9999178680935366e-05,
+      "loss": 1.0048,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.556043944205204,
+      "learning_rate": 1.9999156925386344e-05,
+      "loss": 0.9398,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.766440333330717,
+      "learning_rate": 1.9999134885467148e-05,
+      "loss": 0.9585,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.354974463671109,
+      "learning_rate": 1.9999112561178402e-05,
+      "loss": 1.074,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.99983671056059,
+      "learning_rate": 1.9999089952520746e-05,
+      "loss": 1.1193,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.083429746626358,
+      "learning_rate": 1.9999067059494824e-05,
+      "loss": 0.9585,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.075384931081132,
+      "learning_rate": 1.999904388210128e-05,
+      "loss": 0.9859,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.781617919943116,
+      "learning_rate": 1.9999020420340782e-05,
+      "loss": 0.9583,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.966704287289733,
+      "learning_rate": 1.999899667421399e-05,
+      "loss": 1.098,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.534238631631656,
+      "learning_rate": 1.999897264372158e-05,
+      "loss": 1.0995,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.068356514703574,
+      "learning_rate": 1.999894832886424e-05,
+      "loss": 0.984,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.9964647538874,
+      "learning_rate": 1.9998923729642657e-05,
+      "loss": 0.998,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.393814403269237,
+      "learning_rate": 1.999889884605753e-05,
+      "loss": 0.9782,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 24.982013960016147,
+      "learning_rate": 1.9998873678109575e-05,
+      "loss": 1.0013,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.385080431888543,
+      "learning_rate": 1.99988482257995e-05,
+      "loss": 1.0315,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.403720561477744,
+      "learning_rate": 1.9998822489128028e-05,
+      "loss": 0.937,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 32.37210180407703,
+      "learning_rate": 1.9998796468095897e-05,
+      "loss": 1.0377,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.621343531451927,
+      "learning_rate": 1.999877016270384e-05,
+      "loss": 0.999,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.807004663220358,
+      "learning_rate": 1.9998743572952608e-05,
+      "loss": 0.9875,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.530692159822845,
+      "learning_rate": 1.999871669884296e-05,
+      "loss": 1.0033,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.477830139147436,
+      "learning_rate": 1.999868954037566e-05,
+      "loss": 0.9291,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.30354990763825,
+      "learning_rate": 1.9998662097551475e-05,
+      "loss": 0.9682,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.701506649907408,
+      "learning_rate": 1.9998634370371192e-05,
+      "loss": 1.0579,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.09844673750356,
+      "learning_rate": 1.9998606358835596e-05,
+      "loss": 1.0579,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.042153693159076,
+      "learning_rate": 1.9998578062945483e-05,
+      "loss": 1.1038,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.87003977935117,
+      "learning_rate": 1.999854948270166e-05,
+      "loss": 0.9282,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1471.0013591518905,
+      "learning_rate": 1.999852061810494e-05,
+      "loss": 1.2118,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 35.0744775570728,
+      "learning_rate": 1.9998491469156137e-05,
+      "loss": 1.0116,
+      "step": 680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 28.13299186170543,
+      "learning_rate": 1.999846203585609e-05,
+      "loss": 0.9564,
+      "step": 681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 23.11076190518904,
+      "learning_rate": 1.9998432318205632e-05,
+      "loss": 0.9886,
+      "step": 682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.001125157958516,
+      "learning_rate": 1.9998402316205606e-05,
+      "loss": 1.0623,
+      "step": 683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 28.818401774857087,
+      "learning_rate": 1.9998372029856866e-05,
+      "loss": 1.0083,
+      "step": 684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.18902011512905,
+      "learning_rate": 1.9998341459160277e-05,
+      "loss": 0.9179,
+      "step": 685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 35.677553504388904,
+      "learning_rate": 1.9998310604116704e-05,
+      "loss": 1.0216,
+      "step": 686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.768781962357206,
+      "learning_rate": 1.999827946472703e-05,
+      "loss": 0.9849,
+      "step": 687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 31.64937660972853,
+      "learning_rate": 1.999824804099213e-05,
+      "loss": 0.9782,
+      "step": 688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.918096941199437,
+      "learning_rate": 1.9998216332912908e-05,
+      "loss": 1.0171,
+      "step": 689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.61013836708593,
+      "learning_rate": 1.9998184340490264e-05,
+      "loss": 0.885,
+      "step": 690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 22.565527580076296,
+      "learning_rate": 1.9998152063725107e-05,
+      "loss": 1.0084,
+      "step": 691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.157218866202182,
+      "learning_rate": 1.9998119502618353e-05,
+      "loss": 0.9549,
+      "step": 692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.942288719847813,
+      "learning_rate": 1.9998086657170925e-05,
+      "loss": 0.9984,
+      "step": 693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.81983038671551,
+      "learning_rate": 1.9998053527383766e-05,
+      "loss": 1.1064,
+      "step": 694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.702851369532596,
+      "learning_rate": 1.999802011325781e-05,
+      "loss": 0.9075,
+      "step": 695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 31.219935790826188,
+      "learning_rate": 1.9997986414794012e-05,
+      "loss": 1.0162,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.062142120681791,
+      "learning_rate": 1.999795243199333e-05,
+      "loss": 0.9574,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.163135072367126,
+      "learning_rate": 1.9997918164856728e-05,
+      "loss": 0.9591,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 22.240807100852727,
+      "learning_rate": 1.9997883613385184e-05,
+      "loss": 0.9251,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 28.1818260352272,
+      "learning_rate": 1.999784877757968e-05,
+      "loss": 0.9941,
+      "step": 700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.27154914499387,
+      "learning_rate": 1.99978136574412e-05,
+      "loss": 0.9307,
+      "step": 701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 25.716527018880733,
+      "learning_rate": 1.999777825297075e-05,
+      "loss": 1.0928,
+      "step": 702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.930286027268007,
+      "learning_rate": 1.9997742564169335e-05,
+      "loss": 0.9775,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 32.953649056374836,
+      "learning_rate": 1.999770659103797e-05,
+      "loss": 1.0377,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 31.720956645746604,
+      "learning_rate": 1.999767033357768e-05,
+      "loss": 1.136,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 23.675923263181794,
+      "learning_rate": 1.9997633791789496e-05,
+      "loss": 0.9852,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.159710022438254,
+      "learning_rate": 1.999759696567445e-05,
+      "loss": 0.9578,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.615629888993382,
+      "learning_rate": 1.9997559855233596e-05,
+      "loss": 0.983,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 33.146591727613576,
+      "learning_rate": 1.999752246046799e-05,
+      "loss": 1.1633,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.228318464618354,
+      "learning_rate": 1.9997484781378694e-05,
+      "loss": 1.0657,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.935915363694495,
+      "learning_rate": 1.9997446817966776e-05,
+      "loss": 1.0634,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 23.081959929447248,
+      "learning_rate": 1.9997408570233322e-05,
+      "loss": 1.0599,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.53073736736404,
+      "learning_rate": 1.9997370038179414e-05,
+      "loss": 1.1398,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.904704015913204,
+      "learning_rate": 1.9997331221806152e-05,
+      "loss": 0.9441,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.297954312924443,
+      "learning_rate": 1.999729212111464e-05,
+      "loss": 0.9148,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.77771079126166,
+      "learning_rate": 1.9997252736105985e-05,
+      "loss": 0.9018,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.4628075742039,
+      "learning_rate": 1.9997213066781312e-05,
+      "loss": 0.9501,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 25.887196421443424,
+      "learning_rate": 1.9997173113141747e-05,
+      "loss": 1.0388,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.88366988841301,
+      "learning_rate": 1.9997132875188427e-05,
+      "loss": 0.8158,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 37.05347981778112,
+      "learning_rate": 1.9997092352922495e-05,
+      "loss": 1.176,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 31.72231868155152,
+      "learning_rate": 1.999705154634511e-05,
+      "loss": 1.2076,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.056440302541347,
+      "learning_rate": 1.999701045545742e-05,
+      "loss": 1.0438,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 28.075445910281495,
+      "learning_rate": 1.9996969080260604e-05,
+      "loss": 1.1965,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.17635262973931,
+      "learning_rate": 1.999692742075584e-05,
+      "loss": 1.0819,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 32.90122458402218,
+      "learning_rate": 1.99968854769443e-05,
+      "loss": 0.9422,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 25.10317667173797,
+      "learning_rate": 1.999684324882719e-05,
+      "loss": 1.031,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.35435118631165,
+      "learning_rate": 1.9996800736405702e-05,
+      "loss": 1.0282,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 33.61465806374499,
+      "learning_rate": 1.9996757939681052e-05,
+      "loss": 1.1021,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 29.424962678920473,
+      "learning_rate": 1.9996714858654456e-05,
+      "loss": 1.095,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 23.8259090862828,
+      "learning_rate": 1.9996671493327135e-05,
+      "loss": 0.9476,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.119920837066687,
+      "learning_rate": 1.9996627843700325e-05,
+      "loss": 0.9104,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 27.554752234049488,
+      "learning_rate": 1.9996583909775267e-05,
+      "loss": 1.0307,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.391961434436325,
+      "learning_rate": 1.999653969155321e-05,
+      "loss": 1.0621,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 44.55982382056265,
+      "learning_rate": 1.999649518903541e-05,
+      "loss": 1.0835,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.406889639247517,
+      "learning_rate": 1.9996450402223137e-05,
+      "loss": 0.9581,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.53286648400594,
+      "learning_rate": 1.9996405331117662e-05,
+      "loss": 1.0734,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.201748348929918,
+      "learning_rate": 1.999635997572027e-05,
+      "loss": 1.0315,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 31.566733431409812,
+      "learning_rate": 1.9996314336032243e-05,
+      "loss": 1.0203,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.12963293665336,
+      "learning_rate": 1.9996268412054887e-05,
+      "loss": 0.9966,
+      "step": 739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.609661739748793,
+      "learning_rate": 1.9996222203789504e-05,
+      "loss": 0.9146,
+      "step": 740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.767601571223842,
+      "learning_rate": 1.9996175711237406e-05,
+      "loss": 1.0671,
+      "step": 741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 38.12139600385158,
+      "learning_rate": 1.9996128934399923e-05,
+      "loss": 0.9169,
+      "step": 742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 36.19068369218754,
+      "learning_rate": 1.999608187327838e-05,
+      "loss": 1.0309,
+      "step": 743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 38.01410823168409,
+      "learning_rate": 1.9996034527874117e-05,
+      "loss": 0.951,
+      "step": 744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 35.589617285805026,
+      "learning_rate": 1.9995986898188477e-05,
+      "loss": 1.0413,
+      "step": 745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 29.23197348215027,
+      "learning_rate": 1.9995938984222815e-05,
+      "loss": 1.0136,
+      "step": 746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 34.717249418117866,
+      "learning_rate": 1.99958907859785e-05,
+      "loss": 1.1298,
+      "step": 747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 30.097320466640863,
+      "learning_rate": 1.99958423034569e-05,
+      "loss": 0.9868,
+      "step": 748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 30.448825783240018,
+      "learning_rate": 1.9995793536659388e-05,
+      "loss": 1.1305,
+      "step": 749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.343293259970014,
+      "learning_rate": 1.9995744485587356e-05,
+      "loss": 0.9858,
+      "step": 750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.89076918989473,
+      "learning_rate": 1.99956951502422e-05,
+      "loss": 0.9851,
+      "step": 751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 39.1549199883858,
+      "learning_rate": 1.999564553062532e-05,
+      "loss": 1.0863,
+      "step": 752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.20988632682434,
+      "learning_rate": 1.9995595626738128e-05,
+      "loss": 0.9575,
+      "step": 753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 28.49339147910342,
+      "learning_rate": 1.9995545438582044e-05,
+      "loss": 0.9938,
+      "step": 754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.020720661897847,
+      "learning_rate": 1.9995494966158494e-05,
+      "loss": 0.9864,
+      "step": 755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 27.414747939478183,
+      "learning_rate": 1.9995444209468916e-05,
+      "loss": 1.0082,
+      "step": 756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.755829020368772,
+      "learning_rate": 1.999539316851475e-05,
+      "loss": 0.9256,
+      "step": 757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.231491864276187,
+      "learning_rate": 1.999534184329745e-05,
+      "loss": 0.9442,
+      "step": 758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.269750093204454,
+      "learning_rate": 1.9995290233818475e-05,
+      "loss": 0.9794,
+      "step": 759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.29033883226424,
+      "learning_rate": 1.9995238340079295e-05,
+      "loss": 1.0591,
+      "step": 760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 31.637327954914582,
+      "learning_rate": 1.9995186162081384e-05,
+      "loss": 1.1424,
+      "step": 761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 28.39544905269339,
+      "learning_rate": 1.9995133699826222e-05,
+      "loss": 1.1801,
+      "step": 762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 37.082381837224915,
+      "learning_rate": 1.999508095331531e-05,
+      "loss": 1.0525,
+      "step": 763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 31.00075740331179,
+      "learning_rate": 1.9995027922550137e-05,
+      "loss": 0.9518,
+      "step": 764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 28.91964572175146,
+      "learning_rate": 1.999497460753222e-05,
+      "loss": 1.0819,
+      "step": 765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.666236160603669,
+      "learning_rate": 1.9994921008263072e-05,
+      "loss": 0.8689,
+      "step": 766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.07462339621955,
+      "learning_rate": 1.9994867124744216e-05,
+      "loss": 0.8584,
+      "step": 767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 29.995975931807806,
+      "learning_rate": 1.9994812956977183e-05,
+      "loss": 1.0508,
+      "step": 768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.138358958316012,
+      "learning_rate": 1.9994758504963522e-05,
+      "loss": 0.9026,
+      "step": 769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 41.30043257083839,
+      "learning_rate": 1.9994703768704773e-05,
+      "loss": 1.0871,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.84981941206687,
+      "learning_rate": 1.9994648748202493e-05,
+      "loss": 0.9018,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.056270868284336,
+      "learning_rate": 1.9994593443458252e-05,
+      "loss": 0.9941,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.231511167349787,
+      "learning_rate": 1.999453785447362e-05,
+      "loss": 0.9491,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.790377846130333,
+      "learning_rate": 1.999448198125018e-05,
+      "loss": 0.957,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 32.21334604847917,
+      "learning_rate": 1.9994425823789517e-05,
+      "loss": 1.1117,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 40.859264248174526,
+      "learning_rate": 1.999436938209323e-05,
+      "loss": 1.0816,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.681576869897665,
+      "learning_rate": 1.9994312656162928e-05,
+      "loss": 0.9506,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.761793379861027,
+      "learning_rate": 1.9994255646000217e-05,
+      "loss": 0.9511,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 23.009516921325208,
+      "learning_rate": 1.999419835160672e-05,
+      "loss": 1.1356,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 29.058749703471978,
+      "learning_rate": 1.999414077298407e-05,
+      "loss": 0.9984,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.967857301390442,
+      "learning_rate": 1.9994082910133903e-05,
+      "loss": 0.9311,
+      "step": 781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.28353516797878,
+      "learning_rate": 1.9994024763057865e-05,
+      "loss": 0.9967,
+      "step": 782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 28.361055374804256,
+      "learning_rate": 1.9993966331757607e-05,
+      "loss": 0.9965,
+      "step": 783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.399942802295453,
+      "learning_rate": 1.9993907616234796e-05,
+      "loss": 1.1351,
+      "step": 784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 27.1752552356069,
+      "learning_rate": 1.9993848616491097e-05,
+      "loss": 1.0399,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.665572190710133,
+      "learning_rate": 1.9993789332528193e-05,
+      "loss": 0.9757,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.87341129638875,
+      "learning_rate": 1.9993729764347763e-05,
+      "loss": 0.9788,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.404118302845683,
+      "learning_rate": 1.9993669911951504e-05,
+      "loss": 0.9547,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 23.788661705048366,
+      "learning_rate": 1.999360977534112e-05,
+      "loss": 0.9624,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 26.020252454309638,
+      "learning_rate": 1.999354935451832e-05,
+      "loss": 1.0428,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.643653534481263,
+      "learning_rate": 1.999348864948482e-05,
+      "loss": 1.0729,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 29.491574148911653,
+      "learning_rate": 1.9993427660242356e-05,
+      "loss": 1.0208,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.411371135016022,
+      "learning_rate": 1.999336638679265e-05,
+      "loss": 1.0212,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.16741036518218,
+      "learning_rate": 1.999330482913745e-05,
+      "loss": 0.9471,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 26.060560856406358,
+      "learning_rate": 1.9993242987278508e-05,
+      "loss": 1.0415,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.240362704137674,
+      "learning_rate": 1.999318086121758e-05,
+      "loss": 0.8658,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 33.59682973451131,
+      "learning_rate": 1.9993118450956434e-05,
+      "loss": 1.0131,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.413119608867582,
+      "learning_rate": 1.9993055756496845e-05,
+      "loss": 0.8949,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.063337245324533,
+      "learning_rate": 1.9992992777840596e-05,
+      "loss": 1.0127,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 32.5013716032295,
+      "learning_rate": 1.999292951498948e-05,
+      "loss": 0.9796,
+      "step": 800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.4678596495198,
+      "learning_rate": 1.9992865967945295e-05,
+      "loss": 0.8518,
+      "step": 801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.86441435108581,
+      "learning_rate": 1.9992802136709842e-05,
+      "loss": 0.9583,
+      "step": 802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 33.19805786961399,
+      "learning_rate": 1.999273802128495e-05,
+      "loss": 1.0568,
+      "step": 803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 30.22292863501949,
+      "learning_rate": 1.9992673621672427e-05,
+      "loss": 1.0313,
+      "step": 804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.81481619456222,
+      "learning_rate": 1.9992608937874115e-05,
+      "loss": 1.0087,
+      "step": 805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.22193855322229,
+      "learning_rate": 1.999254396989185e-05,
+      "loss": 1.0421,
+      "step": 806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 27.403442756654155,
+      "learning_rate": 1.9992478717727478e-05,
+      "loss": 1.0045,
+      "step": 807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 28.41463838234398,
+      "learning_rate": 1.999241318138286e-05,
+      "loss": 0.9212,
+      "step": 808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.336751760909568,
+      "learning_rate": 1.9992347360859858e-05,
+      "loss": 0.9983,
+      "step": 809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.56880793631396,
+      "learning_rate": 1.9992281256160337e-05,
+      "loss": 0.9342,
+      "step": 810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.966927639020398,
+      "learning_rate": 1.9992214867286182e-05,
+      "loss": 0.8636,
+      "step": 811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.393939609188461,
+      "learning_rate": 1.9992148194239287e-05,
+      "loss": 0.9692,
+      "step": 812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.863124719263677,
+      "learning_rate": 1.999208123702154e-05,
+      "loss": 1.0059,
+      "step": 813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 25.308754605893263,
+      "learning_rate": 1.999201399563485e-05,
+      "loss": 1.0584,
+      "step": 814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.94069254585308,
+      "learning_rate": 1.9991946470081124e-05,
+      "loss": 0.9221,
+      "step": 815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.48242480544358,
+      "learning_rate": 1.9991878660362285e-05,
+      "loss": 1.0481,
+      "step": 816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 28.650056717809818,
+      "learning_rate": 1.9991810566480264e-05,
+      "loss": 0.9757,
+      "step": 817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.304980313698515,
+      "learning_rate": 1.9991742188436992e-05,
+      "loss": 0.9911,
+      "step": 818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.18954040059897,
+      "learning_rate": 1.999167352623442e-05,
+      "loss": 1.0247,
+      "step": 819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.615591504789602,
+      "learning_rate": 1.99916045798745e-05,
+      "loss": 0.9229,
+      "step": 820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.042097360141371,
+      "learning_rate": 1.999153534935919e-05,
+      "loss": 0.9028,
+      "step": 821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.36765037774426,
+      "learning_rate": 1.999146583469046e-05,
+      "loss": 0.9344,
+      "step": 822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.649725520853114,
+      "learning_rate": 1.9991396035870282e-05,
+      "loss": 0.9665,
+      "step": 823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.752910489352573,
+      "learning_rate": 1.999132595290065e-05,
+      "loss": 1.0209,
+      "step": 824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.927958070642642,
+      "learning_rate": 1.9991255585783547e-05,
+      "loss": 1.0627,
+      "step": 825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.393859252958695,
+      "learning_rate": 1.9991184934520987e-05,
+      "loss": 0.9001,
+      "step": 826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.021183249807795,
+      "learning_rate": 1.9991113999114966e-05,
+      "loss": 0.9242,
+      "step": 827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.814606630902126,
+      "learning_rate": 1.9991042779567512e-05,
+      "loss": 1.0353,
+      "step": 828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.908477894047138,
+      "learning_rate": 1.9990971275880643e-05,
+      "loss": 0.883,
+      "step": 829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.140590699180752,
+      "learning_rate": 1.9990899488056396e-05,
+      "loss": 0.9375,
+      "step": 830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.82090466464421,
+      "learning_rate": 1.9990827416096813e-05,
+      "loss": 0.9946,
+      "step": 831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.062615461106372,
+      "learning_rate": 1.9990755060003945e-05,
+      "loss": 0.939,
+      "step": 832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.180641766823623,
+      "learning_rate": 1.9990682419779844e-05,
+      "loss": 0.9904,
+      "step": 833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.594210069190982,
+      "learning_rate": 1.9990609495426582e-05,
+      "loss": 0.8201,
+      "step": 834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.217124454000132,
+      "learning_rate": 1.9990536286946227e-05,
+      "loss": 0.9544,
+      "step": 835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.726451528781027,
+      "learning_rate": 1.9990462794340864e-05,
+      "loss": 0.9834,
+      "step": 836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.424770998493283,
+      "learning_rate": 1.9990389017612587e-05,
+      "loss": 0.9333,
+      "step": 837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 31.88116710317783,
+      "learning_rate": 1.999031495676349e-05,
+      "loss": 1.0023,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.989587294002025,
+      "learning_rate": 1.999024061179568e-05,
+      "loss": 0.9795,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.427036379858826,
+      "learning_rate": 1.9990165982711266e-05,
+      "loss": 0.9978,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.00849579116745,
+      "learning_rate": 1.999009106951238e-05,
+      "loss": 0.932,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.269670381901356,
+      "learning_rate": 1.9990015872201147e-05,
+      "loss": 1.0195,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.72704759716728,
+      "learning_rate": 1.998994039077971e-05,
+      "loss": 0.9288,
+      "step": 843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.77555586998277,
+      "learning_rate": 1.9989864625250206e-05,
+      "loss": 0.91,
+      "step": 844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.9001177653043,
+      "learning_rate": 1.99897885756148e-05,
+      "loss": 0.9871,
+      "step": 845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.365308538926303,
+      "learning_rate": 1.998971224187565e-05,
+      "loss": 0.9918,
+      "step": 846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.15001487553259,
+      "learning_rate": 1.998963562403493e-05,
+      "loss": 0.9024,
+      "step": 847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 30.6216208734604,
+      "learning_rate": 1.9989558722094813e-05,
+      "loss": 1.1355,
+      "step": 848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.538922120278034,
+      "learning_rate": 1.998948153605749e-05,
+      "loss": 0.9668,
+      "step": 849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.142611519432652,
+      "learning_rate": 1.9989404065925157e-05,
+      "loss": 0.9824,
+      "step": 850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.998781814548558,
+      "learning_rate": 1.9989326311700016e-05,
+      "loss": 0.9729,
+      "step": 851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.815911225495,
+      "learning_rate": 1.9989248273384278e-05,
+      "loss": 1.1024,
+      "step": 852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.568448812719364,
+      "learning_rate": 1.998916995098016e-05,
+      "loss": 1.0269,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 37.97529794878723,
+      "learning_rate": 1.9989091344489896e-05,
+      "loss": 1.1285,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.093833307130453,
+      "learning_rate": 1.9989012453915718e-05,
+      "loss": 0.9447,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.977681582043218,
+      "learning_rate": 1.998893327925987e-05,
+      "loss": 1.0056,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.978096634156046,
+      "learning_rate": 1.9988853820524596e-05,
+      "loss": 0.9743,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.30994013082935,
+      "learning_rate": 1.998877407771217e-05,
+      "loss": 1.0,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.29930427913624,
+      "learning_rate": 1.9988694050824847e-05,
+      "loss": 1.0081,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.667376292754565,
+      "learning_rate": 1.998861373986491e-05,
+      "loss": 0.8904,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 25.86367086933168,
+      "learning_rate": 1.9988533144834642e-05,
+      "loss": 1.1446,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.355136942558662,
+      "learning_rate": 1.9988452265736335e-05,
+      "loss": 0.9395,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 27.30893269829037,
+      "learning_rate": 1.998837110257229e-05,
+      "loss": 0.961,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.567106449316867,
+      "learning_rate": 1.9988289655344814e-05,
+      "loss": 0.9266,
+      "step": 864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.23541728308947,
+      "learning_rate": 1.998820792405622e-05,
+      "loss": 0.8713,
+      "step": 865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.765799393213108,
+      "learning_rate": 1.998812590870884e-05,
+      "loss": 0.9929,
+      "step": 866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 27.135648330436652,
+      "learning_rate": 1.9988043609304995e-05,
+      "loss": 0.9062,
+      "step": 867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.160797839762104,
+      "learning_rate": 1.9987961025847037e-05,
+      "loss": 1.0473,
+      "step": 868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 28.63223994100243,
+      "learning_rate": 1.998787815833731e-05,
+      "loss": 0.9395,
+      "step": 869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 37.465919881683654,
+      "learning_rate": 1.9987795006778173e-05,
+      "loss": 0.9055,
+      "step": 870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.221696657207115,
+      "learning_rate": 1.9987711571171987e-05,
+      "loss": 1.0072,
+      "step": 871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 34.105648468036904,
+      "learning_rate": 1.9987627851521126e-05,
+      "loss": 1.0029,
+      "step": 872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 28.93263817186818,
+      "learning_rate": 1.998754384782797e-05,
+      "loss": 1.0123,
+      "step": 873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 24.908467970607894,
+      "learning_rate": 1.998745956009491e-05,
+      "loss": 0.9502,
+      "step": 874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.42165320838223,
+      "learning_rate": 1.9987374988324345e-05,
+      "loss": 0.9002,
+      "step": 875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.078146314795195,
+      "learning_rate": 1.9987290132518677e-05,
+      "loss": 0.9438,
+      "step": 876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 33.27848961893949,
+      "learning_rate": 1.9987204992680322e-05,
+      "loss": 0.9485,
+      "step": 877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.26684856394545,
+      "learning_rate": 1.9987119568811693e-05,
+      "loss": 1.0242,
+      "step": 878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.897137022108495,
+      "learning_rate": 1.998703386091523e-05,
+      "loss": 0.868,
+      "step": 879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.083151933156167,
+      "learning_rate": 1.9986947868993366e-05,
+      "loss": 0.9719,
+      "step": 880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.88446296027853,
+      "learning_rate": 1.998686159304855e-05,
+      "loss": 0.8459,
+      "step": 881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.437631533617076,
+      "learning_rate": 1.9986775033083225e-05,
+      "loss": 0.861,
+      "step": 882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.935485884994645,
+      "learning_rate": 1.9986688189099865e-05,
+      "loss": 0.9921,
+      "step": 883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.621648396422263,
+      "learning_rate": 1.9986601061100934e-05,
+      "loss": 0.9493,
+      "step": 884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.00615028190211,
+      "learning_rate": 1.9986513649088915e-05,
+      "loss": 0.9605,
+      "step": 885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.535952044605807,
+      "learning_rate": 1.9986425953066284e-05,
+      "loss": 0.9192,
+      "step": 886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 28.33890047192677,
+      "learning_rate": 1.9986337973035542e-05,
+      "loss": 1.1137,
+      "step": 887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 28.633470350257046,
+      "learning_rate": 1.9986249708999194e-05,
+      "loss": 1.0084,
+      "step": 888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 30.547819182250514,
+      "learning_rate": 1.998616116095974e-05,
+      "loss": 0.8784,
+      "step": 889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.728594514284588,
+      "learning_rate": 1.9986072328919707e-05,
+      "loss": 0.929,
+      "step": 890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.206547964184086,
+      "learning_rate": 1.9985983212881618e-05,
+      "loss": 1.0093,
+      "step": 891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 27.868262481702807,
+      "learning_rate": 1.9985893812848007e-05,
+      "loss": 0.9966,
+      "step": 892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.120365907755325,
+      "learning_rate": 1.998580412882142e-05,
+      "loss": 0.9585,
+      "step": 893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.40060378094494,
+      "learning_rate": 1.9985714160804406e-05,
+      "loss": 0.951,
+      "step": 894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 63.25595737001749,
+      "learning_rate": 1.9985623908799518e-05,
+      "loss": 1.0075,
+      "step": 895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.184525813103157,
+      "learning_rate": 1.9985533372809335e-05,
+      "loss": 0.9598,
+      "step": 896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.37430623943003,
+      "learning_rate": 1.9985442552836417e-05,
+      "loss": 1.0131,
+      "step": 897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.614081869721268,
+      "learning_rate": 1.9985351448883358e-05,
+      "loss": 0.8407,
+      "step": 898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 40.66036947009204,
+      "learning_rate": 1.9985260060952742e-05,
+      "loss": 1.1067,
+      "step": 899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.771311306654475,
+      "learning_rate": 1.9985168389047175e-05,
+      "loss": 0.9295,
+      "step": 900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.97576504675759,
+      "learning_rate": 1.9985076433169258e-05,
+      "loss": 0.9276,
+      "step": 901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 33.185904152894416,
+      "learning_rate": 1.9984984193321608e-05,
+      "loss": 0.9348,
+      "step": 902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.139107118380018,
+      "learning_rate": 1.9984891669506847e-05,
+      "loss": 1.0378,
+      "step": 903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.13446842949091,
+      "learning_rate": 1.9984798861727614e-05,
+      "loss": 1.1061,
+      "step": 904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 33.56381521010012,
+      "learning_rate": 1.9984705769986538e-05,
+      "loss": 0.9512,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.831056176741296,
+      "learning_rate": 1.998461239428627e-05,
+      "loss": 0.9442,
+      "step": 906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.284614438729086,
+      "learning_rate": 1.9984518734629467e-05,
+      "loss": 1.0857,
+      "step": 907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.470259374024849,
+      "learning_rate": 1.998442479101879e-05,
+      "loss": 0.8212,
+      "step": 908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.079067228749622,
+      "learning_rate": 1.9984330563456918e-05,
+      "loss": 1.0071,
+      "step": 909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.548201359288534,
+      "learning_rate": 1.998423605194652e-05,
+      "loss": 0.9289,
+      "step": 910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.867643656089143,
+      "learning_rate": 1.9984141256490294e-05,
+      "loss": 0.8817,
+      "step": 911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.75903194116444,
+      "learning_rate": 1.9984046177090926e-05,
+      "loss": 0.9227,
+      "step": 912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.776527642640172,
+      "learning_rate": 1.998395081375113e-05,
+      "loss": 0.9352,
+      "step": 913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.641440914128953,
+      "learning_rate": 1.9983855166473613e-05,
+      "loss": 0.9707,
+      "step": 914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.1738760286084,
+      "learning_rate": 1.998375923526109e-05,
+      "loss": 0.9254,
+      "step": 915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.244305857370346,
+      "learning_rate": 1.99836630201163e-05,
+      "loss": 0.9369,
+      "step": 916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.632949080259692,
+      "learning_rate": 1.9983566521041973e-05,
+      "loss": 1.0031,
+      "step": 917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.584330628760394,
+      "learning_rate": 1.9983469738040852e-05,
+      "loss": 0.9407,
+      "step": 918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.7209982779378,
+      "learning_rate": 1.9983372671115696e-05,
+      "loss": 0.8765,
+      "step": 919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.68512370905215,
+      "learning_rate": 1.998327532026926e-05,
+      "loss": 0.9519,
+      "step": 920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.818700823930502,
+      "learning_rate": 1.9983177685504312e-05,
+      "loss": 1.0332,
+      "step": 921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.28754939570171,
+      "learning_rate": 1.998307976682363e-05,
+      "loss": 0.9153,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.84776960948699,
+      "learning_rate": 1.998298156423e-05,
+      "loss": 0.9509,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.2049697740919,
+      "learning_rate": 1.9982883077726218e-05,
+      "loss": 0.9338,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.359839754106535,
+      "learning_rate": 1.998278430731508e-05,
+      "loss": 0.9355,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.97638165661955,
+      "learning_rate": 1.9982685252999393e-05,
+      "loss": 0.8382,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.32619890080394,
+      "learning_rate": 1.998258591478198e-05,
+      "loss": 0.959,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.89415149625322,
+      "learning_rate": 1.998248629266566e-05,
+      "loss": 0.8957,
+      "step": 928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 31.943558460292774,
+      "learning_rate": 1.9982386386653273e-05,
+      "loss": 0.9798,
+      "step": 929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 24.017375857712963,
+      "learning_rate": 1.9982286196747653e-05,
+      "loss": 0.8992,
+      "step": 930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 42.161691614895396,
+      "learning_rate": 1.9982185722951655e-05,
+      "loss": 1.0017,
+      "step": 931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.98815800644647,
+      "learning_rate": 1.9982084965268134e-05,
+      "loss": 0.9948,
+      "step": 932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.64485681263662,
+      "learning_rate": 1.9981983923699956e-05,
+      "loss": 0.9194,
+      "step": 933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.91240020274571,
+      "learning_rate": 1.9981882598249996e-05,
+      "loss": 0.9707,
+      "step": 934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 35.70244139970563,
+      "learning_rate": 1.998178098892113e-05,
+      "loss": 0.9954,
+      "step": 935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.285258117678723,
+      "learning_rate": 1.998167909571626e-05,
+      "loss": 0.9304,
+      "step": 936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.17099899625401,
+      "learning_rate": 1.998157691863827e-05,
+      "loss": 0.9262,
+      "step": 937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.25734694190907,
+      "learning_rate": 1.998147445769007e-05,
+      "loss": 1.0516,
+      "step": 938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.78016361673385,
+      "learning_rate": 1.9981371712874576e-05,
+      "loss": 0.8487,
+      "step": 939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 25.60884712680659,
+      "learning_rate": 1.9981268684194712e-05,
+      "loss": 0.8976,
+      "step": 940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.543851541489303,
+      "learning_rate": 1.9981165371653404e-05,
+      "loss": 1.0043,
+      "step": 941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.880149045154408,
+      "learning_rate": 1.998106177525359e-05,
+      "loss": 0.9128,
+      "step": 942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.143342445306903,
+      "learning_rate": 1.998095789499822e-05,
+      "loss": 1.0114,
+      "step": 943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.71753376324862,
+      "learning_rate": 1.9980853730890248e-05,
+      "loss": 0.9462,
+      "step": 944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.723056235586299,
+      "learning_rate": 1.998074928293263e-05,
+      "loss": 0.853,
+      "step": 945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.305508550183559,
+      "learning_rate": 1.9980644551128347e-05,
+      "loss": 0.947,
+      "step": 946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.7884693475016,
+      "learning_rate": 1.9980539535480365e-05,
+      "loss": 0.9528,
+      "step": 947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.684553332254552,
+      "learning_rate": 1.998043423599168e-05,
+      "loss": 0.9229,
+      "step": 948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.45037526225891,
+      "learning_rate": 1.9980328652665285e-05,
+      "loss": 0.93,
+      "step": 949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.342525647370522,
+      "learning_rate": 1.9980222785504178e-05,
+      "loss": 0.8307,
+      "step": 950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.949869235113345,
+      "learning_rate": 1.9980116634511377e-05,
+      "loss": 0.9576,
+      "step": 951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 25.881216256166415,
+      "learning_rate": 1.9980010199689893e-05,
+      "loss": 0.9439,
+      "step": 952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.92329329418154,
+      "learning_rate": 1.997990348104276e-05,
+      "loss": 1.0542,
+      "step": 953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.721333365378953,
+      "learning_rate": 1.997979647857301e-05,
+      "loss": 0.9466,
+      "step": 954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.140817454402058,
+      "learning_rate": 1.9979689192283683e-05,
+      "loss": 1.0371,
+      "step": 955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.79666640560971,
+      "learning_rate": 1.997958162217784e-05,
+      "loss": 0.9012,
+      "step": 956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.34264524598088,
+      "learning_rate": 1.9979473768258528e-05,
+      "loss": 0.9864,
+      "step": 957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.877202936832614,
+      "learning_rate": 1.9979365630528824e-05,
+      "loss": 0.8326,
+      "step": 958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.292802280839137,
+      "learning_rate": 1.9979257208991796e-05,
+      "loss": 1.0353,
+      "step": 959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.998349341868792,
+      "learning_rate": 1.9979148503650528e-05,
+      "loss": 0.9579,
+      "step": 960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.46610453602181,
+      "learning_rate": 1.997903951450812e-05,
+      "loss": 0.8989,
+      "step": 961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.396557091882116,
+      "learning_rate": 1.9978930241567662e-05,
+      "loss": 1.072,
+      "step": 962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.311744914247736,
+      "learning_rate": 1.9978820684832268e-05,
+      "loss": 0.8628,
+      "step": 963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.23925123443074,
+      "learning_rate": 1.997871084430505e-05,
+      "loss": 0.9007,
+      "step": 964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.155163351411675,
+      "learning_rate": 1.997860071998913e-05,
+      "loss": 0.8953,
+      "step": 965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.784837092670617,
+      "learning_rate": 1.997849031188765e-05,
+      "loss": 0.9702,
+      "step": 966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.56300106113026,
+      "learning_rate": 1.9978379620003737e-05,
+      "loss": 0.9037,
+      "step": 967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 30.88380557334971,
+      "learning_rate": 1.9978268644340545e-05,
+      "loss": 0.9354,
+      "step": 968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 24.44037004968756,
+      "learning_rate": 1.997815738490123e-05,
+      "loss": 0.9041,
+      "step": 969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.061041783148795,
+      "learning_rate": 1.997804584168896e-05,
+      "loss": 0.8689,
+      "step": 970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.00842304682578,
+      "learning_rate": 1.9977934014706898e-05,
+      "loss": 1.0025,
+      "step": 971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.164024982279454,
+      "learning_rate": 1.997782190395823e-05,
+      "loss": 1.036,
+      "step": 972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.160050762989524,
+      "learning_rate": 1.9977709509446148e-05,
+      "loss": 1.0333,
+      "step": 973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 28.18236742319079,
+      "learning_rate": 1.9977596831173843e-05,
+      "loss": 0.9852,
+      "step": 974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.215368577298406,
+      "learning_rate": 1.9977483869144522e-05,
+      "loss": 0.8775,
+      "step": 975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.592001121889975,
+      "learning_rate": 1.9977370623361393e-05,
+      "loss": 1.0141,
+      "step": 976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.88561976656229,
+      "learning_rate": 1.997725709382768e-05,
+      "loss": 0.9635,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 24.50692665461813,
+      "learning_rate": 1.9977143280546616e-05,
+      "loss": 1.0061,
+      "step": 978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 25.888740135974892,
+      "learning_rate": 1.997702918352143e-05,
+      "loss": 1.0239,
+      "step": 979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.65698507822647,
+      "learning_rate": 1.997691480275537e-05,
+      "loss": 0.8671,
+      "step": 980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 24.87193977279463,
+      "learning_rate": 1.997680013825169e-05,
+      "loss": 0.8638,
+      "step": 981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.821167913617344,
+      "learning_rate": 1.997668519001365e-05,
+      "loss": 0.9667,
+      "step": 982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 29.648484524776272,
+      "learning_rate": 1.9976569958044523e-05,
+      "loss": 0.985,
+      "step": 983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.75908343510108,
+      "learning_rate": 1.997645444234758e-05,
+      "loss": 0.9432,
+      "step": 984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.14695877945067,
+      "learning_rate": 1.997633864292611e-05,
+      "loss": 0.9516,
+      "step": 985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.29064336987488,
+      "learning_rate": 1.99762225597834e-05,
+      "loss": 0.9803,
+      "step": 986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.519886680918606,
+      "learning_rate": 1.997610619292276e-05,
+      "loss": 1.0261,
+      "step": 987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.018074294943958,
+      "learning_rate": 1.9975989542347495e-05,
+      "loss": 1.0793,
+      "step": 988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.608446275899777,
+      "learning_rate": 1.9975872608060927e-05,
+      "loss": 0.8461,
+      "step": 989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.69574199675815,
+      "learning_rate": 1.9975755390066373e-05,
+      "loss": 1.0362,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.82663372196337,
+      "learning_rate": 1.9975637888367173e-05,
+      "loss": 1.0071,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.896183397888986,
+      "learning_rate": 1.9975520102966667e-05,
+      "loss": 1.0048,
+      "step": 992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.794927310425322,
+      "learning_rate": 1.9975402033868207e-05,
+      "loss": 0.9567,
+      "step": 993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 27.72274353370509,
+      "learning_rate": 1.9975283681075142e-05,
+      "loss": 0.9253,
+      "step": 994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.81541402234388,
+      "learning_rate": 1.9975165044590853e-05,
+      "loss": 0.9425,
+      "step": 995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.675916738713358,
+      "learning_rate": 1.99750461244187e-05,
+      "loss": 1.0206,
+      "step": 996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.905428859595194,
+      "learning_rate": 1.997492692056207e-05,
+      "loss": 0.9438,
+      "step": 997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 28.522451249444654,
+      "learning_rate": 1.9974807433024357e-05,
+      "loss": 0.9945,
+      "step": 998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.718876927446388,
+      "learning_rate": 1.9974687661808954e-05,
+      "loss": 0.9148,
+      "step": 999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.091939756214508,
+      "learning_rate": 1.997456760691927e-05,
+      "loss": 0.9874,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.11514146058491,
+      "learning_rate": 1.9974447268358715e-05,
+      "loss": 0.9015,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.804668716079288,
+      "learning_rate": 1.9974326646130716e-05,
+      "loss": 0.8472,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.656475030728036,
+      "learning_rate": 1.9974205740238704e-05,
+      "loss": 0.8391,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.558633175644697,
+      "learning_rate": 1.9974084550686116e-05,
+      "loss": 0.9494,
+      "step": 1004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 27.818598239427306,
+      "learning_rate": 1.9973963077476394e-05,
+      "loss": 1.0851,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.655754652375885,
+      "learning_rate": 1.9973841320613e-05,
+      "loss": 0.9687,
+      "step": 1006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.40657534730464,
+      "learning_rate": 1.997371928009939e-05,
+      "loss": 1.0059,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.40173965431018,
+      "learning_rate": 1.9973596955939043e-05,
+      "loss": 1.0217,
+      "step": 1008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.04782350799938,
+      "learning_rate": 1.997347434813543e-05,
+      "loss": 0.9063,
+      "step": 1009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.56124047007055,
+      "learning_rate": 1.997335145669204e-05,
+      "loss": 0.8219,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.831400271602874,
+      "learning_rate": 1.997322828161237e-05,
+      "loss": 1.0246,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.31427635365123,
+      "learning_rate": 1.9973104822899923e-05,
+      "loss": 0.9491,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.734833288421257,
+      "learning_rate": 1.9972981080558208e-05,
+      "loss": 1.1044,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.49551648289499,
+      "learning_rate": 1.9972857054590744e-05,
+      "loss": 1.0689,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.384798855558653,
+      "learning_rate": 1.997273274500106e-05,
+      "loss": 0.8931,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.769802069531156,
+      "learning_rate": 1.997260815179269e-05,
+      "loss": 1.1256,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.464246187071343,
+      "learning_rate": 1.997248327496918e-05,
+      "loss": 1.0696,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.956797254439103,
+      "learning_rate": 1.9972358114534074e-05,
+      "loss": 0.9252,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.760436841003276,
+      "learning_rate": 1.9972232670490945e-05,
+      "loss": 0.8813,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.583151045220244,
+      "learning_rate": 1.997210694284335e-05,
+      "loss": 0.8999,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.44323274196326,
+      "learning_rate": 1.9971980931594865e-05,
+      "loss": 0.9655,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.494483652381803,
+      "learning_rate": 1.9971854636749075e-05,
+      "loss": 0.9329,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.267103554759874,
+      "learning_rate": 1.9971728058309572e-05,
+      "loss": 0.987,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.118992572482618,
+      "learning_rate": 1.9971601196279963e-05,
+      "loss": 0.9983,
+      "step": 1024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.842380598831724,
+      "learning_rate": 1.997147405066384e-05,
+      "loss": 0.9695,
+      "step": 1025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.65298669391427,
+      "learning_rate": 1.9971346621464833e-05,
+      "loss": 0.9535,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 27.512256742170667,
+      "learning_rate": 1.9971218908686563e-05,
+      "loss": 0.9578,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 24.574149753059544,
+      "learning_rate": 1.9971090912332657e-05,
+      "loss": 0.9132,
+      "step": 1028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 29.33446986848635,
+      "learning_rate": 1.9970962632406757e-05,
+      "loss": 1.0906,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.516292745642946,
+      "learning_rate": 1.9970834068912515e-05,
+      "loss": 0.9737,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.62443621688516,
+      "learning_rate": 1.9970705221853584e-05,
+      "loss": 0.9541,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.22644135185706,
+      "learning_rate": 1.9970576091233633e-05,
+      "loss": 0.9547,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.726208484949435,
+      "learning_rate": 1.9970446677056325e-05,
+      "loss": 0.8472,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.621540227646772,
+      "learning_rate": 1.997031697932535e-05,
+      "loss": 0.9316,
+      "step": 1034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.326850876810315,
+      "learning_rate": 1.9970186998044392e-05,
+      "loss": 0.8996,
+      "step": 1035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.65304831806845,
+      "learning_rate": 1.9970056733217147e-05,
+      "loss": 0.9552,
+      "step": 1036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.16398263318951,
+      "learning_rate": 1.996992618484732e-05,
+      "loss": 0.8852,
+      "step": 1037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.376545159769574,
+      "learning_rate": 1.9969795352938625e-05,
+      "loss": 0.9859,
+      "step": 1038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.187092733003354,
+      "learning_rate": 1.9969664237494785e-05,
+      "loss": 1.0501,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.335407337507426,
+      "learning_rate": 1.9969532838519524e-05,
+      "loss": 0.865,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.160217237182188,
+      "learning_rate": 1.996940115601658e-05,
+      "loss": 1.1142,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.337951351272025,
+      "learning_rate": 1.99692691899897e-05,
+      "loss": 1.0136,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 27.48985530503833,
+      "learning_rate": 1.9969136940442636e-05,
+      "loss": 0.9256,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.359953718842061,
+      "learning_rate": 1.996900440737915e-05,
+      "loss": 0.9117,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.059047708332876,
+      "learning_rate": 1.996887159080301e-05,
+      "loss": 0.9515,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.98626266797299,
+      "learning_rate": 1.9968738490717992e-05,
+      "loss": 0.9373,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.413081707680554,
+      "learning_rate": 1.9968605107127886e-05,
+      "loss": 0.9226,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 26.178540115576087,
+      "learning_rate": 1.996847144003648e-05,
+      "loss": 0.897,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.96063426285729,
+      "learning_rate": 1.9968337489447578e-05,
+      "loss": 0.9423,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.224057065412012,
+      "learning_rate": 1.996820325536499e-05,
+      "loss": 0.8617,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.948927645674216,
+      "learning_rate": 1.996806873779253e-05,
+      "loss": 0.9121,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 26.049640858786617,
+      "learning_rate": 1.9967933936734027e-05,
+      "loss": 0.9288,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.044905889730735,
+      "learning_rate": 1.996779885219332e-05,
+      "loss": 0.8825,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 30.567094353191365,
+      "learning_rate": 1.9967663484174235e-05,
+      "loss": 0.8961,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.90970061560877,
+      "learning_rate": 1.9967527832680637e-05,
+      "loss": 0.9373,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 26.129488848548302,
+      "learning_rate": 1.996739189771638e-05,
+      "loss": 0.9245,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 25.387900679932635,
+      "learning_rate": 1.9967255679285324e-05,
+      "loss": 0.9912,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.449547174792245,
+      "learning_rate": 1.9967119177391348e-05,
+      "loss": 0.9457,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.37084016139191,
+      "learning_rate": 1.9966982392038335e-05,
+      "loss": 1.0184,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 77.16469358476006,
+      "learning_rate": 1.996684532323017e-05,
+      "loss": 0.9994,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.957716178638062,
+      "learning_rate": 1.9966707970970756e-05,
+      "loss": 0.9131,
+      "step": 1061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.762517296446227,
+      "learning_rate": 1.9966570335264e-05,
+      "loss": 0.8824,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.06801353583546,
+      "learning_rate": 1.9966432416113812e-05,
+      "loss": 0.8856,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.171373317220162,
+      "learning_rate": 1.9966294213524117e-05,
+      "loss": 0.8615,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.024850750346925,
+      "learning_rate": 1.9966155727498843e-05,
+      "loss": 0.9013,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.164824136200476,
+      "learning_rate": 1.9966016958041933e-05,
+      "loss": 1.0358,
+      "step": 1066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 41.681919746720965,
+      "learning_rate": 1.996587790515733e-05,
+      "loss": 0.8274,
+      "step": 1067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.945059858878466,
+      "learning_rate": 1.9965738568848986e-05,
+      "loss": 0.9953,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.953576014056228,
+      "learning_rate": 1.996559894912087e-05,
+      "loss": 1.0221,
+      "step": 1069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.972018525935646,
+      "learning_rate": 1.996545904597695e-05,
+      "loss": 0.9417,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.388072707490695,
+      "learning_rate": 1.9965318859421203e-05,
+      "loss": 0.9903,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.47781751645356,
+      "learning_rate": 1.9965178389457617e-05,
+      "loss": 0.8525,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.860240417558686,
+      "learning_rate": 1.9965037636090187e-05,
+      "loss": 0.9391,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 26.294374092027503,
+      "learning_rate": 1.9964896599322917e-05,
+      "loss": 0.9244,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.660765057313604,
+      "learning_rate": 1.9964755279159816e-05,
+      "loss": 0.8199,
+      "step": 1075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 31.299620579344733,
+      "learning_rate": 1.9964613675604904e-05,
+      "loss": 0.9738,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.760483900897643,
+      "learning_rate": 1.996447178866221e-05,
+      "loss": 0.9592,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.650392983694106,
+      "learning_rate": 1.9964329618335766e-05,
+      "loss": 0.9399,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.959922380667738,
+      "learning_rate": 1.9964187164629617e-05,
+      "loss": 0.976,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 25.05245508419107,
+      "learning_rate": 1.996404442754781e-05,
+      "loss": 0.9548,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 26.07033700544596,
+      "learning_rate": 1.9963901407094415e-05,
+      "loss": 0.9631,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.19212974742126,
+      "learning_rate": 1.996375810327349e-05,
+      "loss": 1.0318,
+      "step": 1082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.661445227363606,
+      "learning_rate": 1.9963614516089112e-05,
+      "loss": 0.9228,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.85009642256209,
+      "learning_rate": 1.9963470645545365e-05,
+      "loss": 1.0432,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.32518228904275,
+      "learning_rate": 1.9963326491646346e-05,
+      "loss": 0.9614,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.71690331977722,
+      "learning_rate": 1.9963182054396144e-05,
+      "loss": 0.8802,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 30.1355175023656,
+      "learning_rate": 1.9963037333798876e-05,
+      "loss": 1.0874,
+      "step": 1087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.292365728501006,
+      "learning_rate": 1.9962892329858655e-05,
+      "loss": 0.876,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.214769426839712,
+      "learning_rate": 1.99627470425796e-05,
+      "loss": 0.8733,
+      "step": 1089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.379554352322735,
+      "learning_rate": 1.9962601471965854e-05,
+      "loss": 0.9827,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 29.110130115979604,
+      "learning_rate": 1.9962455618021545e-05,
+      "loss": 0.8946,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.795895266959587,
+      "learning_rate": 1.9962309480750827e-05,
+      "loss": 1.029,
+      "step": 1092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.176858967088037,
+      "learning_rate": 1.9962163060157854e-05,
+      "loss": 1.0959,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 36.20100217748054,
+      "learning_rate": 1.9962016356246794e-05,
+      "loss": 0.9817,
+      "step": 1094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.19508124130565,
+      "learning_rate": 1.9961869369021814e-05,
+      "loss": 0.8772,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.21362086243986,
+      "learning_rate": 1.9961722098487098e-05,
+      "loss": 0.9472,
+      "step": 1096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.058492890904425,
+      "learning_rate": 1.9961574544646835e-05,
+      "loss": 0.8941,
+      "step": 1097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 195.05581634123945,
+      "learning_rate": 1.9961426707505217e-05,
+      "loss": 0.8981,
+      "step": 1098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 39.32973305071126,
+      "learning_rate": 1.996127858706645e-05,
+      "loss": 1.0174,
+      "step": 1099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.445888327195416,
+      "learning_rate": 1.996113018333475e-05,
+      "loss": 0.9646,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.171377947074845,
+      "learning_rate": 1.996098149631433e-05,
+      "loss": 0.9686,
+      "step": 1101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.874052487351488,
+      "learning_rate": 1.9960832526009427e-05,
+      "loss": 1.0018,
+      "step": 1102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.598162584973974,
+      "learning_rate": 1.996068327242427e-05,
+      "loss": 0.9975,
+      "step": 1103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.004610504949117,
+      "learning_rate": 1.9960533735563113e-05,
+      "loss": 0.8323,
+      "step": 1104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.582257565653794,
+      "learning_rate": 1.99603839154302e-05,
+      "loss": 0.7476,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.541556438261914,
+      "learning_rate": 1.9960233812029798e-05,
+      "loss": 0.8683,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 38.115180145076536,
+      "learning_rate": 1.996008342536617e-05,
+      "loss": 0.9944,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.129276843311849,
+      "learning_rate": 1.9959932755443596e-05,
+      "loss": 0.8668,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.878704671929537,
+      "learning_rate": 1.995978180226636e-05,
+      "loss": 0.9611,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 30.11822019357628,
+      "learning_rate": 1.9959630565838758e-05,
+      "loss": 1.0163,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.82556571920938,
+      "learning_rate": 1.9959479046165088e-05,
+      "loss": 1.0279,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.952955102622985,
+      "learning_rate": 1.9959327243249662e-05,
+      "loss": 0.8513,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.68419848419876,
+      "learning_rate": 1.9959175157096794e-05,
+      "loss": 0.9806,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.483035655291626,
+      "learning_rate": 1.995902278771081e-05,
+      "loss": 0.9821,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.279112038316622,
+      "learning_rate": 1.9958870135096044e-05,
+      "loss": 0.9228,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.95422021734671,
+      "learning_rate": 1.9958717199256833e-05,
+      "loss": 0.9523,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.930983981026705,
+      "learning_rate": 1.9958563980197536e-05,
+      "loss": 0.9921,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.897593610216855,
+      "learning_rate": 1.9958410477922504e-05,
+      "loss": 0.8384,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 25.011398549303884,
+      "learning_rate": 1.9958256692436103e-05,
+      "loss": 0.9378,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 30.645943474306943,
+      "learning_rate": 1.995810262374271e-05,
+      "loss": 0.9036,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.178524056826843,
+      "learning_rate": 1.99579482718467e-05,
+      "loss": 0.7769,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.82688153633873,
+      "learning_rate": 1.9957793636752466e-05,
+      "loss": 0.8656,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 28.657159114881118,
+      "learning_rate": 1.9957638718464405e-05,
+      "loss": 1.0352,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.372462772966394,
+      "learning_rate": 1.995748351698693e-05,
+      "loss": 1.0062,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 31.516940010486838,
+      "learning_rate": 1.995732803232444e-05,
+      "loss": 0.9846,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.428904983246213,
+      "learning_rate": 1.995717226448137e-05,
+      "loss": 1.0084,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 26.96075121589446,
+      "learning_rate": 1.995701621346215e-05,
+      "loss": 0.9589,
+      "step": 1127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.407298769632664,
+      "learning_rate": 1.9956859879271207e-05,
+      "loss": 0.891,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.547919443225897,
+      "learning_rate": 1.9956703261913e-05,
+      "loss": 0.8514,
+      "step": 1129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.761614780186456,
+      "learning_rate": 1.995654636139197e-05,
+      "loss": 0.9503,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.186137096558895,
+      "learning_rate": 1.9956389177712592e-05,
+      "loss": 1.0059,
+      "step": 1131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.500806487315653,
+      "learning_rate": 1.9956231710879325e-05,
+      "loss": 1.0018,
+      "step": 1132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.397915764093298,
+      "learning_rate": 1.9956073960896652e-05,
+      "loss": 0.8797,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.37684573662613,
+      "learning_rate": 1.9955915927769066e-05,
+      "loss": 0.9778,
+      "step": 1134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.25531568966331,
+      "learning_rate": 1.995575761150105e-05,
+      "loss": 1.0669,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.679965697748836,
+      "learning_rate": 1.995559901209711e-05,
+      "loss": 0.896,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.073898200835217,
+      "learning_rate": 1.995544012956176e-05,
+      "loss": 0.9498,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.17758573778675,
+      "learning_rate": 1.995528096389952e-05,
+      "loss": 0.9395,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.8285652424402,
+      "learning_rate": 1.995512151511491e-05,
+      "loss": 0.9703,
+      "step": 1139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.66749442919373,
+      "learning_rate": 1.995496178321247e-05,
+      "loss": 1.0551,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.76111052344706,
+      "learning_rate": 1.9954801768196738e-05,
+      "loss": 0.8594,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.19758400764559,
+      "learning_rate": 1.9954641470072267e-05,
+      "loss": 0.8637,
+      "step": 1142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.223655637944525,
+      "learning_rate": 1.9954480888843614e-05,
+      "loss": 1.0662,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.395337053731495,
+      "learning_rate": 1.995432002451535e-05,
+      "loss": 0.9162,
+      "step": 1144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.209186781300918,
+      "learning_rate": 1.995415887709205e-05,
+      "loss": 0.8881,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 24.438681956827622,
+      "learning_rate": 1.995399744657829e-05,
+      "loss": 0.9011,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.813041791707164,
+      "learning_rate": 1.9953835732978668e-05,
+      "loss": 0.8796,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 24.719100036988042,
+      "learning_rate": 1.995367373629778e-05,
+      "loss": 0.9395,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.08496977604367,
+      "learning_rate": 1.9953511456540233e-05,
+      "loss": 0.9645,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.237033522061687,
+      "learning_rate": 1.9953348893710645e-05,
+      "loss": 0.91,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.488883503013803,
+      "learning_rate": 1.9953186047813637e-05,
+      "loss": 0.8988,
+      "step": 1151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 36.55713919113992,
+      "learning_rate": 1.9953022918853837e-05,
+      "loss": 0.9923,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 30.13020995109024,
+      "learning_rate": 1.9952859506835887e-05,
+      "loss": 1.0171,
+      "step": 1153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.313885660675833,
+      "learning_rate": 1.995269581176444e-05,
+      "loss": 0.925,
+      "step": 1154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.380561423525133,
+      "learning_rate": 1.9952531833644145e-05,
+      "loss": 0.9817,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 33.27406470233171,
+      "learning_rate": 1.9952367572479664e-05,
+      "loss": 1.102,
+      "step": 1156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.73252246633459,
+      "learning_rate": 1.995220302827567e-05,
+      "loss": 0.9488,
+      "step": 1157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 34.5026125177882,
+      "learning_rate": 1.9952038201036848e-05,
+      "loss": 1.0748,
+      "step": 1158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 28.870464497190003,
+      "learning_rate": 1.995187309076788e-05,
+      "loss": 0.8552,
+      "step": 1159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.031554636165179,
+      "learning_rate": 1.9951707697473465e-05,
+      "loss": 0.8569,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 26.63769626711829,
+      "learning_rate": 1.99515420211583e-05,
+      "loss": 1.0523,
+      "step": 1161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.890918687515374,
+      "learning_rate": 1.9951376061827104e-05,
+      "loss": 0.9819,
+      "step": 1162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 22.280379834779016,
+      "learning_rate": 1.995120981948459e-05,
+      "loss": 0.966,
+      "step": 1163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.07628401442733,
+      "learning_rate": 1.9951043294135498e-05,
+      "loss": 0.8894,
+      "step": 1164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.540979091455273,
+      "learning_rate": 1.995087648578455e-05,
+      "loss": 0.8823,
+      "step": 1165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.13708005658681,
+      "learning_rate": 1.9950709394436497e-05,
+      "loss": 1.0229,
+      "step": 1166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.64973108975881,
+      "learning_rate": 1.9950542020096094e-05,
+      "loss": 0.9317,
+      "step": 1167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.215047049702857,
+      "learning_rate": 1.995037436276809e-05,
+      "loss": 0.8126,
+      "step": 1168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.32932582071813,
+      "learning_rate": 1.9950206422457263e-05,
+      "loss": 0.9556,
+      "step": 1169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.990490524916908,
+      "learning_rate": 1.9950038199168387e-05,
+      "loss": 1.0087,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 26.641587413955936,
+      "learning_rate": 1.9949869692906245e-05,
+      "loss": 0.9149,
+      "step": 1171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 26.106088595271515,
+      "learning_rate": 1.994970090367563e-05,
+      "loss": 0.9971,
+      "step": 1172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.41581929549556,
+      "learning_rate": 1.994953183148134e-05,
+      "loss": 0.856,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.672523506627265,
+      "learning_rate": 1.9949362476328184e-05,
+      "loss": 1.0399,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.695419951962345,
+      "learning_rate": 1.9949192838220982e-05,
+      "loss": 0.8773,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.926550004091787,
+      "learning_rate": 1.9949022917164556e-05,
+      "loss": 0.8676,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 30.105655016483126,
+      "learning_rate": 1.9948852713163737e-05,
+      "loss": 0.9021,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.316633414260945,
+      "learning_rate": 1.9948682226223366e-05,
+      "loss": 0.9636,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.736809104705028,
+      "learning_rate": 1.9948511456348295e-05,
+      "loss": 0.9362,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.08183740911144,
+      "learning_rate": 1.9948340403543375e-05,
+      "loss": 1.0331,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.57154819026003,
+      "learning_rate": 1.9948169067813473e-05,
+      "loss": 0.983,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.184887788579513,
+      "learning_rate": 1.9947997449163467e-05,
+      "loss": 1.0024,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.90526965392405,
+      "learning_rate": 1.9947825547598226e-05,
+      "loss": 0.9809,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 29.46472945046384,
+      "learning_rate": 1.994765336312265e-05,
+      "loss": 0.8987,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.38233590675501,
+      "learning_rate": 1.994748089574163e-05,
+      "loss": 0.9454,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.310221537703704,
+      "learning_rate": 1.9947308145460073e-05,
+      "loss": 0.9608,
+      "step": 1186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.30178722481752,
+      "learning_rate": 1.9947135112282894e-05,
+      "loss": 0.9571,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.044473009121155,
+      "learning_rate": 1.9946961796215006e-05,
+      "loss": 0.8899,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.507704191568795,
+      "learning_rate": 1.994678819726135e-05,
+      "loss": 0.8034,
+      "step": 1189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.730711127433754,
+      "learning_rate": 1.994661431542685e-05,
+      "loss": 0.9218,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.246142138176882,
+      "learning_rate": 1.994644015071646e-05,
+      "loss": 0.8613,
+      "step": 1191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.45738185901497,
+      "learning_rate": 1.994626570313513e-05,
+      "loss": 0.9275,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.320499622610207,
+      "learning_rate": 1.9946090972687827e-05,
+      "loss": 1.0286,
+      "step": 1193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.75997874661781,
+      "learning_rate": 1.994591595937951e-05,
+      "loss": 0.7884,
+      "step": 1194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.398642040283757,
+      "learning_rate": 1.994574066321516e-05,
+      "loss": 0.9726,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.392093268692573,
+      "learning_rate": 1.9945565084199766e-05,
+      "loss": 0.9629,
+      "step": 1196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.446296983902577,
+      "learning_rate": 1.994538922233832e-05,
+      "loss": 0.9321,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.243092139911507,
+      "learning_rate": 1.994521307763582e-05,
+      "loss": 0.8464,
+      "step": 1198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 32.68458859395483,
+      "learning_rate": 1.994503665009728e-05,
+      "loss": 0.953,
+      "step": 1199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 25.726231668744806,
+      "learning_rate": 1.9944859939727714e-05,
+      "loss": 0.9653,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.352486735112418,
+      "learning_rate": 1.994468294653215e-05,
+      "loss": 0.7899,
+      "step": 1201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.72271390253157,
+      "learning_rate": 1.994450567051562e-05,
+      "loss": 0.9373,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.202867844926203,
+      "learning_rate": 1.9944328111683166e-05,
+      "loss": 0.9778,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.555628186642041,
+      "learning_rate": 1.9944150270039837e-05,
+      "loss": 0.9264,
+      "step": 1204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 31.11643087054771,
+      "learning_rate": 1.994397214559069e-05,
+      "loss": 0.8862,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.276235170561964,
+      "learning_rate": 1.9943793738340796e-05,
+      "loss": 0.9889,
+      "step": 1206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.994496461120395,
+      "learning_rate": 1.9943615048295224e-05,
+      "loss": 1.0607,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 31.373176717403403,
+      "learning_rate": 1.9943436075459056e-05,
+      "loss": 0.9986,
+      "step": 1208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.367489900837658,
+      "learning_rate": 1.994325681983738e-05,
+      "loss": 0.8561,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.57649557531624,
+      "learning_rate": 1.9943077281435302e-05,
+      "loss": 0.9088,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.22798190056659,
+      "learning_rate": 1.994289746025792e-05,
+      "loss": 0.9386,
+      "step": 1211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.051370076626796,
+      "learning_rate": 1.994271735631035e-05,
+      "loss": 0.9943,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.49291742688194,
+      "learning_rate": 1.9942536969597713e-05,
+      "loss": 0.9867,
+      "step": 1213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.380714742365978,
+      "learning_rate": 1.9942356300125146e-05,
+      "loss": 1.01,
+      "step": 1214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.769314347622814,
+      "learning_rate": 1.994217534789778e-05,
+      "loss": 1.009,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 25.32979480338763,
+      "learning_rate": 1.994199411292076e-05,
+      "loss": 0.97,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.86239682642862,
+      "learning_rate": 1.9941812595199246e-05,
+      "loss": 0.9607,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.97029570902465,
+      "learning_rate": 1.99416307947384e-05,
+      "loss": 0.8529,
+      "step": 1218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.543302660121476,
+      "learning_rate": 1.9941448711543386e-05,
+      "loss": 0.9853,
+      "step": 1219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.608061158043856,
+      "learning_rate": 1.994126634561939e-05,
+      "loss": 0.9182,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.731217702565235,
+      "learning_rate": 1.994108369697159e-05,
+      "loss": 0.9641,
+      "step": 1221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 33.05863261724831,
+      "learning_rate": 1.9940900765605187e-05,
+      "loss": 1.0102,
+      "step": 1222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 39.6450780495734,
+      "learning_rate": 1.994071755152538e-05,
+      "loss": 1.0147,
+      "step": 1223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.845016571956475,
+      "learning_rate": 1.9940534054737383e-05,
+      "loss": 1.0116,
+      "step": 1224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.703273159160023,
+      "learning_rate": 1.9940350275246412e-05,
+      "loss": 0.8897,
+      "step": 1225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.373733486995256,
+      "learning_rate": 1.9940166213057693e-05,
+      "loss": 0.9425,
+      "step": 1226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.455188548883385,
+      "learning_rate": 1.9939981868176462e-05,
+      "loss": 1.0323,
+      "step": 1227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 28.822589758446107,
+      "learning_rate": 1.9939797240607966e-05,
+      "loss": 0.9633,
+      "step": 1228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.82033621737185,
+      "learning_rate": 1.9939612330357446e-05,
+      "loss": 0.9004,
+      "step": 1229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.329713536063725,
+      "learning_rate": 1.993942713743017e-05,
+      "loss": 0.9775,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.118670545951,
+      "learning_rate": 1.9939241661831398e-05,
+      "loss": 0.8798,
+      "step": 1231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 27.92895927653587,
+      "learning_rate": 1.9939055903566403e-05,
+      "loss": 0.826,
+      "step": 1232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.968752351958887,
+      "learning_rate": 1.9938869862640476e-05,
+      "loss": 0.9785,
+      "step": 1233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.587930629160464,
+      "learning_rate": 1.9938683539058906e-05,
+      "loss": 0.9025,
+      "step": 1234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.091215742957534,
+      "learning_rate": 1.9938496932826988e-05,
+      "loss": 0.8837,
+      "step": 1235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.754134579278293,
+      "learning_rate": 1.993831004395003e-05,
+      "loss": 1.0415,
+      "step": 1236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.773719017875198,
+      "learning_rate": 1.9938122872433353e-05,
+      "loss": 0.7607,
+      "step": 1237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.629002864946615,
+      "learning_rate": 1.9937935418282274e-05,
+      "loss": 0.8634,
+      "step": 1238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.14983628556474,
+      "learning_rate": 1.9937747681502122e-05,
+      "loss": 0.878,
+      "step": 1239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 28.180830389525347,
+      "learning_rate": 1.993755966209824e-05,
+      "loss": 1.002,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.552035578828285,
+      "learning_rate": 1.9937371360075977e-05,
+      "loss": 0.9009,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.728857726390583,
+      "learning_rate": 1.9937182775440683e-05,
+      "loss": 1.0715,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.58531103284727,
+      "learning_rate": 1.9936993908197725e-05,
+      "loss": 0.9315,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 37.56450138694873,
+      "learning_rate": 1.9936804758352473e-05,
+      "loss": 1.0043,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.127500708486227,
+      "learning_rate": 1.9936615325910308e-05,
+      "loss": 0.8986,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.70305438398933,
+      "learning_rate": 1.9936425610876614e-05,
+      "loss": 0.9206,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.75094908712745,
+      "learning_rate": 1.993623561325679e-05,
+      "loss": 0.9542,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.105175353425047,
+      "learning_rate": 1.9936045333056235e-05,
+      "loss": 0.854,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.642600995924038,
+      "learning_rate": 1.9935854770280362e-05,
+      "loss": 0.8935,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.381684001037932,
+      "learning_rate": 1.9935663924934596e-05,
+      "loss": 0.8901,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.971888514001433,
+      "learning_rate": 1.993547279702436e-05,
+      "loss": 0.8007,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.821220718931485,
+      "learning_rate": 1.9935281386555087e-05,
+      "loss": 0.9687,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.53065080363474,
+      "learning_rate": 1.9935089693532225e-05,
+      "loss": 0.9721,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.75973181592774,
+      "learning_rate": 1.9934897717961223e-05,
+      "loss": 0.8984,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.57298354615714,
+      "learning_rate": 1.9934705459847544e-05,
+      "loss": 0.9365,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.953908912711572,
+      "learning_rate": 1.9934512919196647e-05,
+      "loss": 0.8488,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 28.33725745547857,
+      "learning_rate": 1.9934320096014022e-05,
+      "loss": 0.9324,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.281497199769635,
+      "learning_rate": 1.9934126990305142e-05,
+      "loss": 0.8084,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.136670117142064,
+      "learning_rate": 1.99339336020755e-05,
+      "loss": 0.9868,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.091622239433082,
+      "learning_rate": 1.99337399313306e-05,
+      "loss": 0.8913,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.635426654212395,
+      "learning_rate": 1.9933545978075942e-05,
+      "loss": 0.8554,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.474915783037556,
+      "learning_rate": 1.9933351742317053e-05,
+      "loss": 0.983,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 29.234804159467206,
+      "learning_rate": 1.993315722405945e-05,
+      "loss": 0.966,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.672568471062904,
+      "learning_rate": 1.9932962423308666e-05,
+      "loss": 0.8412,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.449711114972242,
+      "learning_rate": 1.9932767340070243e-05,
+      "loss": 0.9385,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.669071223225533,
+      "learning_rate": 1.9932571974349724e-05,
+      "loss": 0.8169,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.992868998436204,
+      "learning_rate": 1.9932376326152672e-05,
+      "loss": 0.9201,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.272789298085467,
+      "learning_rate": 1.9932180395484644e-05,
+      "loss": 0.9691,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.28794206515459,
+      "learning_rate": 1.993198418235122e-05,
+      "loss": 0.9517,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.36011207674172,
+      "learning_rate": 1.993178768675797e-05,
+      "loss": 1.0202,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.541887003351604,
+      "learning_rate": 1.9931590908710495e-05,
+      "loss": 0.9676,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.141704428640352,
+      "learning_rate": 1.9931393848214376e-05,
+      "loss": 0.8703,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.695246094783986,
+      "learning_rate": 1.9931196505275233e-05,
+      "loss": 1.0039,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.346715071676691,
+      "learning_rate": 1.993099887989867e-05,
+      "loss": 0.8852,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.672661702505575,
+      "learning_rate": 1.9930800972090307e-05,
+      "loss": 0.9104,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.618545684106746,
+      "learning_rate": 1.9930602781855772e-05,
+      "loss": 0.8934,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 37.476588063711446,
+      "learning_rate": 1.9930404309200706e-05,
+      "loss": 0.9254,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.0360428722522,
+      "learning_rate": 1.9930205554130748e-05,
+      "loss": 0.9899,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 33.663050132821624,
+      "learning_rate": 1.9930006516651555e-05,
+      "loss": 0.9813,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.84790344808927,
+      "learning_rate": 1.9929807196768785e-05,
+      "loss": 0.9439,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.98229335110451,
+      "learning_rate": 1.9929607594488104e-05,
+      "loss": 0.8303,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 30.035470062652568,
+      "learning_rate": 1.9929407709815197e-05,
+      "loss": 0.9388,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.215968611692105,
+      "learning_rate": 1.992920754275574e-05,
+      "loss": 0.8859,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.943706562053954,
+      "learning_rate": 1.992900709331543e-05,
+      "loss": 0.9961,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.8796046794665,
+      "learning_rate": 1.9928806361499965e-05,
+      "loss": 0.8896,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.615686040819751,
+      "learning_rate": 1.9928605347315057e-05,
+      "loss": 0.9176,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.386918984886165,
+      "learning_rate": 1.992840405076642e-05,
+      "loss": 0.8916,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 24.019404373913307,
+      "learning_rate": 1.992820247185978e-05,
+      "loss": 1.0235,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.637710906025376,
+      "learning_rate": 1.992800061060087e-05,
+      "loss": 0.9788,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.1480366265285,
+      "learning_rate": 1.992779846699543e-05,
+      "loss": 0.9326,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.699055472836726,
+      "learning_rate": 1.992759604104921e-05,
+      "loss": 0.9894,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 28.074772851204248,
+      "learning_rate": 1.9927393332767965e-05,
+      "loss": 1.0129,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.953500482332803,
+      "learning_rate": 1.9927190342157462e-05,
+      "loss": 1.0021,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.84006127194746,
+      "learning_rate": 1.9926987069223475e-05,
+      "loss": 0.9293,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.89379450621844,
+      "learning_rate": 1.9926783513971777e-05,
+      "loss": 0.9532,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 35.378460087236796,
+      "learning_rate": 1.9926579676408168e-05,
+      "loss": 1.0128,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.318854205624405,
+      "learning_rate": 1.992637555653844e-05,
+      "loss": 0.9178,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.897438525977122,
+      "learning_rate": 1.9926171154368393e-05,
+      "loss": 0.8841,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.500584955485103,
+      "learning_rate": 1.992596646990385e-05,
+      "loss": 1.0633,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 34.49422380219095,
+      "learning_rate": 1.992576150315062e-05,
+      "loss": 0.7963,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.81966630425843,
+      "learning_rate": 1.9925556254114547e-05,
+      "loss": 0.8769,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 29.164915812711097,
+      "learning_rate": 1.9925350722801458e-05,
+      "loss": 0.9607,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.58943896263914,
+      "learning_rate": 1.9925144909217197e-05,
+      "loss": 0.9324,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.13568356508364,
+      "learning_rate": 1.9924938813367626e-05,
+      "loss": 0.8545,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.45426296385659,
+      "learning_rate": 1.99247324352586e-05,
+      "loss": 1.0189,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 31.72773203400263,
+      "learning_rate": 1.9924525774895986e-05,
+      "loss": 0.9538,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.50644130479907,
+      "learning_rate": 1.9924318832285667e-05,
+      "loss": 1.0431,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.018262046216933,
+      "learning_rate": 1.9924111607433528e-05,
+      "loss": 0.9644,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.072629528814268,
+      "learning_rate": 1.992390410034546e-05,
+      "loss": 0.9031,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 24.638114951632378,
+      "learning_rate": 1.992369631102736e-05,
+      "loss": 1.0506,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.19323599795892,
+      "learning_rate": 1.9923488239485146e-05,
+      "loss": 0.9868,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.022933296867507,
+      "learning_rate": 1.992327988572473e-05,
+      "loss": 1.0442,
+      "step": 1312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 38.60077075343159,
+      "learning_rate": 1.9923071249752038e-05,
+      "loss": 0.9415,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.935327725801237,
+      "learning_rate": 1.9922862331573007e-05,
+      "loss": 0.8662,
+      "step": 1314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.549857823175042,
+      "learning_rate": 1.9922653131193575e-05,
+      "loss": 0.9445,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.42482200842557,
+      "learning_rate": 1.9922443648619696e-05,
+      "loss": 0.8671,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 31.910562303238194,
+      "learning_rate": 1.992223388385732e-05,
+      "loss": 0.8879,
+      "step": 1317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.252798493002622,
+      "learning_rate": 1.9922023836912418e-05,
+      "loss": 0.8011,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.150754687231103,
+      "learning_rate": 1.992181350779096e-05,
+      "loss": 0.8793,
+      "step": 1319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.31881387900814,
+      "learning_rate": 1.9921602896498934e-05,
+      "loss": 0.8197,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.93742356016639,
+      "learning_rate": 1.9921392003042322e-05,
+      "loss": 0.9119,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.878665614920763,
+      "learning_rate": 1.9921180827427126e-05,
+      "loss": 0.8363,
+      "step": 1322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.5208541018877675,
+      "learning_rate": 1.9920969369659354e-05,
+      "loss": 1.0458,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.1772940981394,
+      "learning_rate": 1.9920757629745015e-05,
+      "loss": 0.9305,
+      "step": 1324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.442677599117722,
+      "learning_rate": 1.992054560769013e-05,
+      "loss": 0.9774,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.854559087963043,
+      "learning_rate": 1.9920333303500734e-05,
+      "loss": 0.9629,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 35.30302763377855,
+      "learning_rate": 1.992012071718286e-05,
+      "loss": 0.9763,
+      "step": 1327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.153192085498706,
+      "learning_rate": 1.9919907848742556e-05,
+      "loss": 0.9926,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.260157248420533,
+      "learning_rate": 1.991969469818588e-05,
+      "loss": 0.9913,
+      "step": 1329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.96376624247682,
+      "learning_rate": 1.9919481265518885e-05,
+      "loss": 0.9022,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 34.18917028402202,
+      "learning_rate": 1.9919267550747647e-05,
+      "loss": 0.9388,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.98196083705158,
+      "learning_rate": 1.991905355387824e-05,
+      "loss": 0.8616,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.595385975033494,
+      "learning_rate": 1.9918839274916757e-05,
+      "loss": 0.8426,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.307162634521717,
+      "learning_rate": 1.9918624713869287e-05,
+      "loss": 0.895,
+      "step": 1334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.081120207404,
+      "learning_rate": 1.9918409870741928e-05,
+      "loss": 0.7825,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.539929219347595,
+      "learning_rate": 1.9918194745540798e-05,
+      "loss": 0.8481,
+      "step": 1336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 33.29577674826702,
+      "learning_rate": 1.991797933827201e-05,
+      "loss": 0.895,
+      "step": 1337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.000071211179446,
+      "learning_rate": 1.9917763648941692e-05,
+      "loss": 1.023,
+      "step": 1338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.508815965063587,
+      "learning_rate": 1.991754767755598e-05,
+      "loss": 0.8633,
+      "step": 1339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.74053182565702,
+      "learning_rate": 1.991733142412101e-05,
+      "loss": 0.9433,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.644052303855188,
+      "learning_rate": 1.9917114888642937e-05,
+      "loss": 0.8993,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.775108462957714,
+      "learning_rate": 1.9916898071127915e-05,
+      "loss": 0.9787,
+      "step": 1342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.003199432061866,
+      "learning_rate": 1.9916680971582116e-05,
+      "loss": 0.9631,
+      "step": 1343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.829538243320078,
+      "learning_rate": 1.9916463590011713e-05,
+      "loss": 0.8365,
+      "step": 1344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 24.878333483592566,
+      "learning_rate": 1.9916245926422883e-05,
+      "loss": 0.9273,
+      "step": 1345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.257384510446926,
+      "learning_rate": 1.9916027980821823e-05,
+      "loss": 0.8607,
+      "step": 1346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.07316436252357,
+      "learning_rate": 1.9915809753214725e-05,
+      "loss": 0.8915,
+      "step": 1347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.959282926977835,
+      "learning_rate": 1.99155912436078e-05,
+      "loss": 0.741,
+      "step": 1348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.978979063158075,
+      "learning_rate": 1.991537245200726e-05,
+      "loss": 0.8688,
+      "step": 1349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 27.744051462182586,
+      "learning_rate": 1.991515337841933e-05,
+      "loss": 1.12,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.25120064531158,
+      "learning_rate": 1.9914934022850233e-05,
+      "loss": 0.8152,
+      "step": 1351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.29446485142575,
+      "learning_rate": 1.9914714385306217e-05,
+      "loss": 0.9278,
+      "step": 1352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.049437390735815,
+      "learning_rate": 1.9914494465793522e-05,
+      "loss": 0.8964,
+      "step": 1353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.324527902453276,
+      "learning_rate": 1.9914274264318402e-05,
+      "loss": 0.9232,
+      "step": 1354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.591530142767086,
+      "learning_rate": 1.9914053780887122e-05,
+      "loss": 0.8671,
+      "step": 1355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.597951831475005,
+      "learning_rate": 1.9913833015505957e-05,
+      "loss": 0.9686,
+      "step": 1356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.237243762700256,
+      "learning_rate": 1.9913611968181178e-05,
+      "loss": 0.925,
+      "step": 1357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.727077098296117,
+      "learning_rate": 1.9913390638919074e-05,
+      "loss": 0.879,
+      "step": 1358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 27.987888497407347,
+      "learning_rate": 1.9913169027725936e-05,
+      "loss": 1.0334,
+      "step": 1359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.733868300949965,
+      "learning_rate": 1.9912947134608073e-05,
+      "loss": 0.9026,
+      "step": 1360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.173910454258483,
+      "learning_rate": 1.9912724959571793e-05,
+      "loss": 0.9208,
+      "step": 1361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.230580679171098,
+      "learning_rate": 1.9912502502623413e-05,
+      "loss": 0.9103,
+      "step": 1362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.052439439329678,
+      "learning_rate": 1.991227976376926e-05,
+      "loss": 0.824,
+      "step": 1363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.76117529959484,
+      "learning_rate": 1.991205674301567e-05,
+      "loss": 0.8669,
+      "step": 1364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.210222016837943,
+      "learning_rate": 1.9911833440368982e-05,
+      "loss": 0.7947,
+      "step": 1365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 33.605481570448795,
+      "learning_rate": 1.9911609855835553e-05,
+      "loss": 0.9422,
+      "step": 1366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.18952214004595,
+      "learning_rate": 1.9911385989421736e-05,
+      "loss": 0.8542,
+      "step": 1367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.163667112580974,
+      "learning_rate": 1.9911161841133898e-05,
+      "loss": 0.8978,
+      "step": 1368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.03465533772165,
+      "learning_rate": 1.9910937410978418e-05,
+      "loss": 0.8659,
+      "step": 1369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.56129051678673,
+      "learning_rate": 1.9910712698961673e-05,
+      "loss": 0.8997,
+      "step": 1370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.7784257478019,
+      "learning_rate": 1.991048770509006e-05,
+      "loss": 0.9045,
+      "step": 1371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.464371663980677,
+      "learning_rate": 1.991026242936997e-05,
+      "loss": 0.8788,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.28065532590489,
+      "learning_rate": 1.9910036871807817e-05,
+      "loss": 1.0487,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.378756413259477,
+      "learning_rate": 1.9909811032410012e-05,
+      "loss": 0.8824,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 24.43542564349211,
+      "learning_rate": 1.9909584911182977e-05,
+      "loss": 0.9517,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.59946889494999,
+      "learning_rate": 1.9909358508133145e-05,
+      "loss": 0.9472,
+      "step": 1376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.575767925465115,
+      "learning_rate": 1.990913182326695e-05,
+      "loss": 0.9322,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.283341835010557,
+      "learning_rate": 1.990890485659085e-05,
+      "loss": 0.877,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.015921680538895,
+      "learning_rate": 1.9908677608111287e-05,
+      "loss": 0.9034,
+      "step": 1379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 32.386463181298566,
+      "learning_rate": 1.9908450077834732e-05,
+      "loss": 1.018,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.62281652293308,
+      "learning_rate": 1.9908222265767653e-05,
+      "loss": 0.9683,
+      "step": 1381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.893433320441366,
+      "learning_rate": 1.9907994171916524e-05,
+      "loss": 0.9804,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.677590592003387,
+      "learning_rate": 1.9907765796287844e-05,
+      "loss": 0.9403,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.765449230010383,
+      "learning_rate": 1.9907537138888097e-05,
+      "loss": 0.9905,
+      "step": 1384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.90626779981043,
+      "learning_rate": 1.990730819972379e-05,
+      "loss": 0.8958,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.605297923163597,
+      "learning_rate": 1.9907078978801432e-05,
+      "loss": 0.8924,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.121787663737475,
+      "learning_rate": 1.9906849476127545e-05,
+      "loss": 0.8506,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 25.914299010726207,
+      "learning_rate": 1.9906619691708652e-05,
+      "loss": 0.9328,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.406456402733113,
+      "learning_rate": 1.9906389625551294e-05,
+      "loss": 0.9968,
+      "step": 1389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.848600183369733,
+      "learning_rate": 1.990615927766201e-05,
+      "loss": 0.9835,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.99942841962401,
+      "learning_rate": 1.9905928648047346e-05,
+      "loss": 0.9886,
+      "step": 1391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.034098767161495,
+      "learning_rate": 1.990569773671387e-05,
+      "loss": 0.9693,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 35.71487357663943,
+      "learning_rate": 1.9905466543668143e-05,
+      "loss": 1.1704,
+      "step": 1393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.45938206197676,
+      "learning_rate": 1.9905235068916745e-05,
+      "loss": 0.9234,
+      "step": 1394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.495776853003562,
+      "learning_rate": 1.9905003312466255e-05,
+      "loss": 0.9915,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.212968858485853,
+      "learning_rate": 1.9904771274323262e-05,
+      "loss": 0.8888,
+      "step": 1396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.54383243693455,
+      "learning_rate": 1.990453895449437e-05,
+      "loss": 0.9925,
+      "step": 1397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.750532358874487,
+      "learning_rate": 1.9904306352986183e-05,
+      "loss": 0.8366,
+      "step": 1398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.479072838232593,
+      "learning_rate": 1.990407346980532e-05,
+      "loss": 0.959,
+      "step": 1399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 28.19172560606223,
+      "learning_rate": 1.99038403049584e-05,
+      "loss": 0.8447,
+      "step": 1400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 27.135755440848115,
+      "learning_rate": 1.9903606858452056e-05,
+      "loss": 0.879,
+      "step": 1401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.06888160182277,
+      "learning_rate": 1.9903373130292928e-05,
+      "loss": 0.9948,
+      "step": 1402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.61414992076343,
+      "learning_rate": 1.9903139120487654e-05,
+      "loss": 0.8504,
+      "step": 1403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.871522748279556,
+      "learning_rate": 1.9902904829042902e-05,
+      "loss": 0.9806,
+      "step": 1404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.369601845853666,
+      "learning_rate": 1.9902670255965333e-05,
+      "loss": 0.9354,
+      "step": 1405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.74389234094852,
+      "learning_rate": 1.9902435401261608e-05,
+      "loss": 0.8639,
+      "step": 1406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.603213032638298,
+      "learning_rate": 1.990220026493842e-05,
+      "loss": 0.9023,
+      "step": 1407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 20.97678157365228,
+      "learning_rate": 1.9901964847002442e-05,
+      "loss": 1.014,
+      "step": 1408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.73880095593121,
+      "learning_rate": 1.990172914746038e-05,
+      "loss": 0.9541,
+      "step": 1409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.593196662706202,
+      "learning_rate": 1.9901493166318934e-05,
+      "loss": 0.8117,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 20.13659180417085,
+      "learning_rate": 1.9901256903584812e-05,
+      "loss": 0.8954,
+      "step": 1411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.86775771340296,
+      "learning_rate": 1.9901020359264738e-05,
+      "loss": 1.0768,
+      "step": 1412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.893536761352216,
+      "learning_rate": 1.9900783533365437e-05,
+      "loss": 0.8369,
+      "step": 1413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 30.969588877520685,
+      "learning_rate": 1.9900546425893644e-05,
+      "loss": 0.851,
+      "step": 1414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 30.364502580785054,
+      "learning_rate": 1.99003090368561e-05,
+      "loss": 0.9364,
+      "step": 1415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 25.64482005553018,
+      "learning_rate": 1.990007136625956e-05,
+      "loss": 0.9606,
+      "step": 1416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.860730601099355,
+      "learning_rate": 1.989983341411078e-05,
+      "loss": 0.8902,
+      "step": 1417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.945570604122867,
+      "learning_rate": 1.9899595180416532e-05,
+      "loss": 1.036,
+      "step": 1418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.585343567181834,
+      "learning_rate": 1.9899356665183584e-05,
+      "loss": 0.9409,
+      "step": 1419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.592404568031984,
+      "learning_rate": 1.989911786841873e-05,
+      "loss": 0.8498,
+      "step": 1420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.34782284981256,
+      "learning_rate": 1.9898878790128752e-05,
+      "loss": 0.862,
+      "step": 1421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.330109461349725,
+      "learning_rate": 1.9898639430320448e-05,
+      "loss": 0.9395,
+      "step": 1422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.958950574966494,
+      "learning_rate": 1.9898399789000633e-05,
+      "loss": 0.8653,
+      "step": 1423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 27.022031537393666,
+      "learning_rate": 1.9898159866176115e-05,
+      "loss": 0.8856,
+      "step": 1424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.055090281468015,
+      "learning_rate": 1.9897919661853723e-05,
+      "loss": 0.9617,
+      "step": 1425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.808843135298243,
+      "learning_rate": 1.9897679176040284e-05,
+      "loss": 0.8153,
+      "step": 1426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.98273301223038,
+      "learning_rate": 1.9897438408742644e-05,
+      "loss": 0.9232,
+      "step": 1427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.467110019854843,
+      "learning_rate": 1.9897197359967642e-05,
+      "loss": 0.899,
+      "step": 1428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.321039026185428,
+      "learning_rate": 1.9896956029722138e-05,
+      "loss": 1.0245,
+      "step": 1429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.247209941626622,
+      "learning_rate": 1.9896714418012993e-05,
+      "loss": 0.8231,
+      "step": 1430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.348569622001488,
+      "learning_rate": 1.9896472524847078e-05,
+      "loss": 0.961,
+      "step": 1431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 37.56406826166652,
+      "learning_rate": 1.9896230350231274e-05,
+      "loss": 0.8571,
+      "step": 1432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 20.53945002910316,
+      "learning_rate": 1.9895987894172474e-05,
+      "loss": 0.8621,
+      "step": 1433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.920721265341765,
+      "learning_rate": 1.989574515667756e-05,
+      "loss": 1.022,
+      "step": 1434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.206127859641747,
+      "learning_rate": 1.9895502137753448e-05,
+      "loss": 0.8861,
+      "step": 1435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.513001768723406,
+      "learning_rate": 1.989525883740704e-05,
+      "loss": 0.8813,
+      "step": 1436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 25.080606141750987,
+      "learning_rate": 1.989501525564526e-05,
+      "loss": 0.9389,
+      "step": 1437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.902371726278396,
+      "learning_rate": 1.9894771392475036e-05,
+      "loss": 1.035,
+      "step": 1438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.802421383683235,
+      "learning_rate": 1.98945272479033e-05,
+      "loss": 0.9444,
+      "step": 1439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.20849078913029,
+      "learning_rate": 1.9894282821936995e-05,
+      "loss": 0.7965,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.171240186323416,
+      "learning_rate": 1.989403811458308e-05,
+      "loss": 0.9591,
+      "step": 1441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.5019219963036,
+      "learning_rate": 1.9893793125848506e-05,
+      "loss": 1.0449,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.663329008585723,
+      "learning_rate": 1.9893547855740243e-05,
+      "loss": 0.9246,
+      "step": 1443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.112498416673166,
+      "learning_rate": 1.9893302304265267e-05,
+      "loss": 0.8921,
+      "step": 1444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.039729486857773,
+      "learning_rate": 1.989305647143056e-05,
+      "loss": 0.8282,
+      "step": 1445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.691181532359877,
+      "learning_rate": 1.9892810357243116e-05,
+      "loss": 0.8882,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.225087008849833,
+      "learning_rate": 1.989256396170993e-05,
+      "loss": 0.8138,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.41466549670376,
+      "learning_rate": 1.9892317284838016e-05,
+      "loss": 1.0354,
+      "step": 1448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.987971303075874,
+      "learning_rate": 1.9892070326634383e-05,
+      "loss": 0.9045,
+      "step": 1449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 33.14052844179611,
+      "learning_rate": 1.9891823087106057e-05,
+      "loss": 0.9687,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.543345362147758,
+      "learning_rate": 1.9891575566260068e-05,
+      "loss": 0.8688,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 31.82596494286104,
+      "learning_rate": 1.9891327764103456e-05,
+      "loss": 0.8913,
+      "step": 1452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.102029975724584,
+      "learning_rate": 1.989107968064327e-05,
+      "loss": 0.8815,
+      "step": 1453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.800652465672165,
+      "learning_rate": 1.9890831315886566e-05,
+      "loss": 0.9337,
+      "step": 1454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.492450865502047,
+      "learning_rate": 1.98905826698404e-05,
+      "loss": 0.9639,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 36.36599028160405,
+      "learning_rate": 1.989033374251185e-05,
+      "loss": 0.9794,
+      "step": 1456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.531828198249837,
+      "learning_rate": 1.9890084533907998e-05,
+      "loss": 0.836,
+      "step": 1457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.454867441883128,
+      "learning_rate": 1.9889835044035925e-05,
+      "loss": 0.9107,
+      "step": 1458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 29.831723663035085,
+      "learning_rate": 1.988958527290273e-05,
+      "loss": 0.9699,
+      "step": 1459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 30.106522068630024,
+      "learning_rate": 1.9889335220515512e-05,
+      "loss": 0.902,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.6693855491818,
+      "learning_rate": 1.9889084886881387e-05,
+      "loss": 0.843,
+      "step": 1461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.156992833080583,
+      "learning_rate": 1.9888834272007475e-05,
+      "loss": 0.8507,
+      "step": 1462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.58564273789724,
+      "learning_rate": 1.9888583375900897e-05,
+      "loss": 0.8502,
+      "step": 1463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.48048829094658,
+      "learning_rate": 1.9888332198568794e-05,
+      "loss": 0.8605,
+      "step": 1464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.514160969565783,
+      "learning_rate": 1.988808074001831e-05,
+      "loss": 0.8924,
+      "step": 1465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.93567949722883,
+      "learning_rate": 1.9887829000256592e-05,
+      "loss": 0.8573,
+      "step": 1466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.38912034665384,
+      "learning_rate": 1.9887576979290803e-05,
+      "loss": 0.8493,
+      "step": 1467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 26.441573061468294,
+      "learning_rate": 1.9887324677128107e-05,
+      "loss": 0.8848,
+      "step": 1468
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.710959270815522,
+      "learning_rate": 1.988707209377568e-05,
+      "loss": 1.0571,
+      "step": 1469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.66331877496578,
+      "learning_rate": 1.9886819229240708e-05,
+      "loss": 0.763,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.84368962246273,
+      "learning_rate": 1.9886566083530382e-05,
+      "loss": 0.9172,
+      "step": 1471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.960246015749057,
+      "learning_rate": 1.9886312656651897e-05,
+      "loss": 0.9751,
+      "step": 1472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.7503540177721,
+      "learning_rate": 1.9886058948612466e-05,
+      "loss": 0.827,
+      "step": 1473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.23471947012008,
+      "learning_rate": 1.98858049594193e-05,
+      "loss": 0.8813,
+      "step": 1474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 27.443001082697744,
+      "learning_rate": 1.9885550689079624e-05,
+      "loss": 0.9107,
+      "step": 1475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.963675917420925,
+      "learning_rate": 1.988529613760067e-05,
+      "loss": 1.0011,
+      "step": 1476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.62294501476367,
+      "learning_rate": 1.9885041304989682e-05,
+      "loss": 0.8256,
+      "step": 1477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.379305682327313,
+      "learning_rate": 1.9884786191253895e-05,
+      "loss": 0.9107,
+      "step": 1478
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.226568236750126,
+      "learning_rate": 1.988453079640057e-05,
+      "loss": 0.9533,
+      "step": 1479
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.931625383750685,
+      "learning_rate": 1.9884275120436976e-05,
+      "loss": 0.8969,
+      "step": 1480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 28.762895368604905,
+      "learning_rate": 1.988401916337038e-05,
+      "loss": 0.947,
+      "step": 1481
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.57032215038423,
+      "learning_rate": 1.9883762925208058e-05,
+      "loss": 0.8551,
+      "step": 1482
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.22208747140622,
+      "learning_rate": 1.98835064059573e-05,
+      "loss": 0.8877,
+      "step": 1483
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 37.12283804808617,
+      "learning_rate": 1.9883249605625403e-05,
+      "loss": 1.0072,
+      "step": 1484
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.543397514422557,
+      "learning_rate": 1.988299252421967e-05,
+      "loss": 0.9646,
+      "step": 1485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 31.81226664028502,
+      "learning_rate": 1.9882735161747407e-05,
+      "loss": 0.8968,
+      "step": 1486
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.148444937919745,
+      "learning_rate": 1.9882477518215938e-05,
+      "loss": 0.8184,
+      "step": 1487
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.605271616238163,
+      "learning_rate": 1.9882219593632595e-05,
+      "loss": 0.8806,
+      "step": 1488
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.401248724863805,
+      "learning_rate": 1.9881961388004702e-05,
+      "loss": 0.8844,
+      "step": 1489
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.85256456659565,
+      "learning_rate": 1.988170290133961e-05,
+      "loss": 0.8483,
+      "step": 1490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.243625712281947,
+      "learning_rate": 1.9881444133644668e-05,
+      "loss": 0.918,
+      "step": 1491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.86633504313018,
+      "learning_rate": 1.9881185084927234e-05,
+      "loss": 0.914,
+      "step": 1492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.082115977569647,
+      "learning_rate": 1.9880925755194677e-05,
+      "loss": 0.9888,
+      "step": 1493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.567946105511787,
+      "learning_rate": 1.9880666144454372e-05,
+      "loss": 0.8667,
+      "step": 1494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.103235278591296,
+      "learning_rate": 1.9880406252713703e-05,
+      "loss": 0.8576,
+      "step": 1495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 32.05414053766156,
+      "learning_rate": 1.988014607998006e-05,
+      "loss": 0.9628,
+      "step": 1496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.314446426447713,
+      "learning_rate": 1.987988562626084e-05,
+      "loss": 1.0049,
+      "step": 1497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.489690572149616,
+      "learning_rate": 1.9879624891563456e-05,
+      "loss": 0.8725,
+      "step": 1498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.839622826800472,
+      "learning_rate": 1.9879363875895317e-05,
+      "loss": 0.9889,
+      "step": 1499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.44151166791262,
+      "learning_rate": 1.987910257926385e-05,
+      "loss": 0.9919,
+      "step": 1500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.1580910476791,
+      "learning_rate": 1.9878841001676484e-05,
+      "loss": 0.846,
+      "step": 1501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.442634325784066,
+      "learning_rate": 1.9878579143140656e-05,
+      "loss": 0.8613,
+      "step": 1502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.65284426453401,
+      "learning_rate": 1.987831700366382e-05,
+      "loss": 0.9577,
+      "step": 1503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.49152889666891,
+      "learning_rate": 1.9878054583253426e-05,
+      "loss": 0.8942,
+      "step": 1504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.020052578138305,
+      "learning_rate": 1.9877791881916935e-05,
+      "loss": 0.8903,
+      "step": 1505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.929546441871096,
+      "learning_rate": 1.9877528899661826e-05,
+      "loss": 0.8664,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.08767658742868,
+      "learning_rate": 1.9877265636495574e-05,
+      "loss": 0.8503,
+      "step": 1507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.611948125225855,
+      "learning_rate": 1.987700209242566e-05,
+      "loss": 0.9032,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.24670966874891,
+      "learning_rate": 1.9876738267459587e-05,
+      "loss": 0.8426,
+      "step": 1509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.202767041853868,
+      "learning_rate": 1.9876474161604858e-05,
+      "loss": 0.8817,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.355531537159216,
+      "learning_rate": 1.9876209774868975e-05,
+      "loss": 0.9919,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.44822621871114,
+      "learning_rate": 1.9875945107259467e-05,
+      "loss": 0.9514,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.600634870616602,
+      "learning_rate": 1.987568015878386e-05,
+      "loss": 0.8976,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 42.65773799033624,
+      "learning_rate": 1.9875414929449684e-05,
+      "loss": 0.9613,
+      "step": 1514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.868428440579983,
+      "learning_rate": 1.9875149419264483e-05,
+      "loss": 0.8894,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.1866930186212,
+      "learning_rate": 1.987488362823581e-05,
+      "loss": 0.8675,
+      "step": 1516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.6598789347875,
+      "learning_rate": 1.9874617556371224e-05,
+      "loss": 0.8718,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.132711136285618,
+      "learning_rate": 1.987435120367829e-05,
+      "loss": 0.9302,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 28.074359770835063,
+      "learning_rate": 1.987408457016459e-05,
+      "loss": 0.8548,
+      "step": 1519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.889135313023882,
+      "learning_rate": 1.9873817655837695e-05,
+      "loss": 0.877,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 20.15093699153894,
+      "learning_rate": 1.9873550460705203e-05,
+      "loss": 0.7814,
+      "step": 1521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 41.35476704008727,
+      "learning_rate": 1.987328298477471e-05,
+      "loss": 0.8904,
+      "step": 1522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.969927016775248,
+      "learning_rate": 1.9873015228053828e-05,
+      "loss": 0.8934,
+      "step": 1523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 20.84439129393353,
+      "learning_rate": 1.9872747190550167e-05,
+      "loss": 0.987,
+      "step": 1524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.562192149270235,
+      "learning_rate": 1.987247887227135e-05,
+      "loss": 0.8728,
+      "step": 1525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 20.610440172458183,
+      "learning_rate": 1.9872210273225012e-05,
+      "loss": 0.845,
+      "step": 1526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 28.046246185266774,
+      "learning_rate": 1.9871941393418788e-05,
+      "loss": 0.8761,
+      "step": 1527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.19092754233467,
+      "learning_rate": 1.9871672232860323e-05,
+      "loss": 0.9944,
+      "step": 1528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.301384599738446,
+      "learning_rate": 1.9871402791557276e-05,
+      "loss": 0.9345,
+      "step": 1529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.007479441077614,
+      "learning_rate": 1.987113306951731e-05,
+      "loss": 0.9111,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.403992620522057,
+      "learning_rate": 1.9870863066748092e-05,
+      "loss": 0.8233,
+      "step": 1531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.482426596277037,
+      "learning_rate": 1.9870592783257303e-05,
+      "loss": 0.8811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.977830127577914,
+      "learning_rate": 1.987032221905263e-05,
+      "loss": 0.8781,
+      "step": 1533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.989255471340897,
+      "learning_rate": 1.9870051374141765e-05,
+      "loss": 0.7822,
+      "step": 1534
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.703784181026588,
+      "learning_rate": 1.9869780248532416e-05,
+      "loss": 0.963,
+      "step": 1535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.42849824799235,
+      "learning_rate": 1.986950884223229e-05,
+      "loss": 0.8673,
+      "step": 1536
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.104362050169783,
+      "learning_rate": 1.9869237155249105e-05,
+      "loss": 0.9335,
+      "step": 1537
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.762347184048217,
+      "learning_rate": 1.9868965187590588e-05,
+      "loss": 0.9469,
+      "step": 1538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.174756447752902,
+      "learning_rate": 1.9868692939264474e-05,
+      "loss": 0.9615,
+      "step": 1539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.08999698213284,
+      "learning_rate": 1.9868420410278506e-05,
+      "loss": 0.9476,
+      "step": 1540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.4324296667565,
+      "learning_rate": 1.9868147600640433e-05,
+      "loss": 0.8804,
+      "step": 1541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.979844326593657,
+      "learning_rate": 1.9867874510358017e-05,
+      "loss": 0.8384,
+      "step": 1542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.081608419897385,
+      "learning_rate": 1.9867601139439023e-05,
+      "loss": 0.8873,
+      "step": 1543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.779687846062483,
+      "learning_rate": 1.9867327487891224e-05,
+      "loss": 0.9797,
+      "step": 1544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.391452958535215,
+      "learning_rate": 1.9867053555722406e-05,
+      "loss": 0.8472,
+      "step": 1545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.469499017498077,
+      "learning_rate": 1.9866779342940354e-05,
+      "loss": 0.9956,
+      "step": 1546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.310513414299503,
+      "learning_rate": 1.986650484955287e-05,
+      "loss": 0.9349,
+      "step": 1547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.279051936045528,
+      "learning_rate": 1.986623007556776e-05,
+      "loss": 0.8944,
+      "step": 1548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.777165605094,
+      "learning_rate": 1.9865955020992837e-05,
+      "loss": 0.9278,
+      "step": 1549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.286095031490465,
+      "learning_rate": 1.9865679685835924e-05,
+      "loss": 0.8482,
+      "step": 1550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.06437172765661,
+      "learning_rate": 1.9865404070104857e-05,
+      "loss": 0.8321,
+      "step": 1551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.569456568911974,
+      "learning_rate": 1.9865128173807465e-05,
+      "loss": 0.9967,
+      "step": 1552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.787221183987047,
+      "learning_rate": 1.98648519969516e-05,
+      "loss": 0.978,
+      "step": 1553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.99072945819066,
+      "learning_rate": 1.9864575539545115e-05,
+      "loss": 0.8469,
+      "step": 1554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.420302041685954,
+      "learning_rate": 1.986429880159587e-05,
+      "loss": 0.9421,
+      "step": 1555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.393284233906577,
+      "learning_rate": 1.986402178311174e-05,
+      "loss": 0.9912,
+      "step": 1556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.083219956199336,
+      "learning_rate": 1.98637444841006e-05,
+      "loss": 1.0749,
+      "step": 1557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.371095586728668,
+      "learning_rate": 1.9863466904570338e-05,
+      "loss": 0.9149,
+      "step": 1558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.69632820410229,
+      "learning_rate": 1.9863189044528847e-05,
+      "loss": 0.8413,
+      "step": 1559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.669799542191065,
+      "learning_rate": 1.986291090398403e-05,
+      "loss": 0.9167,
+      "step": 1560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.127502376430776,
+      "learning_rate": 1.9862632482943796e-05,
+      "loss": 0.9223,
+      "step": 1561
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 31.630367421031174,
+      "learning_rate": 1.9862353781416062e-05,
+      "loss": 0.8649,
+      "step": 1562
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.87502752758015,
+      "learning_rate": 1.9862074799408757e-05,
+      "loss": 0.8531,
+      "step": 1563
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 39.3331372197143,
+      "learning_rate": 1.9861795536929814e-05,
+      "loss": 0.9515,
+      "step": 1564
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.271972655549494,
+      "learning_rate": 1.9861515993987174e-05,
+      "loss": 0.8518,
+      "step": 1565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.494631913935663,
+      "learning_rate": 1.986123617058879e-05,
+      "loss": 0.801,
+      "step": 1566
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.919234674150502,
+      "learning_rate": 1.9860956066742613e-05,
+      "loss": 1.007,
+      "step": 1567
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.858569040426747,
+      "learning_rate": 1.986067568245662e-05,
+      "loss": 0.8046,
+      "step": 1568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.534347097329405,
+      "learning_rate": 1.9860395017738778e-05,
+      "loss": 0.82,
+      "step": 1569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.73581702617474,
+      "learning_rate": 1.9860114072597064e-05,
+      "loss": 0.8114,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.757802254792043,
+      "learning_rate": 1.985983284703948e-05,
+      "loss": 0.8302,
+      "step": 1571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.301606886146196,
+      "learning_rate": 1.9859551341074016e-05,
+      "loss": 0.9559,
+      "step": 1572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.235940291350914,
+      "learning_rate": 1.985926955470868e-05,
+      "loss": 0.894,
+      "step": 1573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 25.379316689200724,
+      "learning_rate": 1.9858987487951485e-05,
+      "loss": 0.9196,
+      "step": 1574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.96398671967239,
+      "learning_rate": 1.9858705140810453e-05,
+      "loss": 0.9733,
+      "step": 1575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 33.444085546192184,
+      "learning_rate": 1.985842251329361e-05,
+      "loss": 0.9762,
+      "step": 1576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.0310834093762,
+      "learning_rate": 1.9858139605409005e-05,
+      "loss": 0.9813,
+      "step": 1577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.530349919890558,
+      "learning_rate": 1.985785641716467e-05,
+      "loss": 0.9046,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.473781409203127,
+      "learning_rate": 1.985757294856867e-05,
+      "loss": 0.891,
+      "step": 1579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 24.782652740656243,
+      "learning_rate": 1.985728919962906e-05,
+      "loss": 0.969,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 27.928155643095884,
+      "learning_rate": 1.985700517035391e-05,
+      "loss": 0.9972,
+      "step": 1581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.171645769485439,
+      "learning_rate": 1.98567208607513e-05,
+      "loss": 0.779,
+      "step": 1582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.15141236688541,
+      "learning_rate": 1.9856436270829315e-05,
+      "loss": 0.8682,
+      "step": 1583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 30.34893832160513,
+      "learning_rate": 1.985615140059605e-05,
+      "loss": 0.8446,
+      "step": 1584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.90170651294565,
+      "learning_rate": 1.9855866250059604e-05,
+      "loss": 0.9663,
+      "step": 1585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.814766740364764,
+      "learning_rate": 1.9855580819228087e-05,
+      "loss": 0.9283,
+      "step": 1586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 25.826641225339994,
+      "learning_rate": 1.985529510810962e-05,
+      "loss": 0.9141,
+      "step": 1587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 24.461194642870954,
+      "learning_rate": 1.9855009116712318e-05,
+      "loss": 0.9014,
+      "step": 1588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.67600278512863,
+      "learning_rate": 1.985472284504433e-05,
+      "loss": 0.9292,
+      "step": 1589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.464522902281754,
+      "learning_rate": 1.9854436293113782e-05,
+      "loss": 0.9108,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 27.952304762485202,
+      "learning_rate": 1.9854149460928832e-05,
+      "loss": 0.7976,
+      "step": 1591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.89888782104047,
+      "learning_rate": 1.9853862348497638e-05,
+      "loss": 0.8497,
+      "step": 1592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.30094980944647,
+      "learning_rate": 1.9853574955828364e-05,
+      "loss": 0.919,
+      "step": 1593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 29.379491346000943,
+      "learning_rate": 1.985328728292918e-05,
+      "loss": 0.8353,
+      "step": 1594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 24.51259496756525,
+      "learning_rate": 1.985299932980827e-05,
+      "loss": 0.9382,
+      "step": 1595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.055782837566728,
+      "learning_rate": 1.9852711096473824e-05,
+      "loss": 1.0492,
+      "step": 1596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.092632275464698,
+      "learning_rate": 1.9852422582934037e-05,
+      "loss": 0.8701,
+      "step": 1597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.837682850336902,
+      "learning_rate": 1.9852133789197117e-05,
+      "loss": 0.7924,
+      "step": 1598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 24.70587231115513,
+      "learning_rate": 1.9851844715271272e-05,
+      "loss": 0.9711,
+      "step": 1599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 26.768006772182794,
+      "learning_rate": 1.985155536116473e-05,
+      "loss": 0.913,
+      "step": 1600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.889000930105173,
+      "learning_rate": 1.9851265726885716e-05,
+      "loss": 0.9164,
+      "step": 1601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 29.56957839094889,
+      "learning_rate": 1.9850975812442465e-05,
+      "loss": 0.9676,
+      "step": 1602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.131166672112233,
+      "learning_rate": 1.9850685617843225e-05,
+      "loss": 0.929,
+      "step": 1603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.227839084474525,
+      "learning_rate": 1.985039514309625e-05,
+      "loss": 0.85,
+      "step": 1604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.506373025882,
+      "learning_rate": 1.9850104388209802e-05,
+      "loss": 0.87,
+      "step": 1605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.749856350545816,
+      "learning_rate": 1.9849813353192142e-05,
+      "loss": 0.7983,
+      "step": 1606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.040139069914062,
+      "learning_rate": 1.9849522038051553e-05,
+      "loss": 0.8483,
+      "step": 1607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.499774206716868,
+      "learning_rate": 1.9849230442796324e-05,
+      "loss": 0.9599,
+      "step": 1608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.094841699068112,
+      "learning_rate": 1.9848938567434737e-05,
+      "loss": 0.9141,
+      "step": 1609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.47469423698453,
+      "learning_rate": 1.98486464119751e-05,
+      "loss": 0.899,
+      "step": 1610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.056744094228375,
+      "learning_rate": 1.9848353976425723e-05,
+      "loss": 0.876,
+      "step": 1611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.517360578017563,
+      "learning_rate": 1.984806126079492e-05,
+      "loss": 0.8606,
+      "step": 1612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.184089523301797,
+      "learning_rate": 1.9847768265091015e-05,
+      "loss": 0.8799,
+      "step": 1613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.220808366496115,
+      "learning_rate": 1.984747498932234e-05,
+      "loss": 0.9153,
+      "step": 1614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.191952781195642,
+      "learning_rate": 1.984718143349724e-05,
+      "loss": 0.9018,
+      "step": 1615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.696585650972292,
+      "learning_rate": 1.9846887597624054e-05,
+      "loss": 0.8602,
+      "step": 1616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.645189871587714,
+      "learning_rate": 1.984659348171115e-05,
+      "loss": 0.9109,
+      "step": 1617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.80176946317184,
+      "learning_rate": 1.9846299085766888e-05,
+      "loss": 0.7923,
+      "step": 1618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.40837045415924,
+      "learning_rate": 1.9846004409799636e-05,
+      "loss": 0.8333,
+      "step": 1619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.21026819611962,
+      "learning_rate": 1.984570945381778e-05,
+      "loss": 0.9582,
+      "step": 1620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.057845668281168,
+      "learning_rate": 1.9845414217829708e-05,
+      "loss": 0.8639,
+      "step": 1621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.807790616028218,
+      "learning_rate": 1.9845118701843816e-05,
+      "loss": 0.8879,
+      "step": 1622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.871412538595905,
+      "learning_rate": 1.9844822905868507e-05,
+      "loss": 0.9686,
+      "step": 1623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.205071110481367,
+      "learning_rate": 1.984452682991219e-05,
+      "loss": 0.9883,
+      "step": 1624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.853528600595627,
+      "learning_rate": 1.984423047398329e-05,
+      "loss": 0.9944,
+      "step": 1625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.91780237453702,
+      "learning_rate": 1.9843933838090236e-05,
+      "loss": 0.8593,
+      "step": 1626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.217180569043126,
+      "learning_rate": 1.9843636922241463e-05,
+      "loss": 0.9105,
+      "step": 1627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.632563895879928,
+      "learning_rate": 1.984333972644541e-05,
+      "loss": 0.818,
+      "step": 1628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.235173435921375,
+      "learning_rate": 1.984304225071054e-05,
+      "loss": 0.8356,
+      "step": 1629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.182657224892584,
+      "learning_rate": 1.98427444950453e-05,
+      "loss": 0.7999,
+      "step": 1630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.394914966399828,
+      "learning_rate": 1.9842446459458166e-05,
+      "loss": 0.8587,
+      "step": 1631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.2234754879083,
+      "learning_rate": 1.9842148143957612e-05,
+      "loss": 0.9409,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.972849386188827,
+      "learning_rate": 1.9841849548552124e-05,
+      "loss": 0.8701,
+      "step": 1633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.238464828067983,
+      "learning_rate": 1.9841550673250192e-05,
+      "loss": 0.8618,
+      "step": 1634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.860763723830463,
+      "learning_rate": 1.9841251518060314e-05,
+      "loss": 0.9003,
+      "step": 1635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.63649729993102,
+      "learning_rate": 1.9840952082991e-05,
+      "loss": 0.8897,
+      "step": 1636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.24084487468591,
+      "learning_rate": 1.984065236805077e-05,
+      "loss": 0.8665,
+      "step": 1637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.135258355377267,
+      "learning_rate": 1.984035237324814e-05,
+      "loss": 0.8578,
+      "step": 1638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 28.447627242906936,
+      "learning_rate": 1.9840052098591645e-05,
+      "loss": 0.9812,
+      "step": 1639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.872861871021367,
+      "learning_rate": 1.9839751544089827e-05,
+      "loss": 0.8685,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.810502189781224,
+      "learning_rate": 1.983945070975123e-05,
+      "loss": 0.834,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.15290131322366,
+      "learning_rate": 1.983914959558441e-05,
+      "loss": 0.9041,
+      "step": 1642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.15181768591629,
+      "learning_rate": 1.983884820159793e-05,
+      "loss": 0.878,
+      "step": 1643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.195868435407395,
+      "learning_rate": 1.9838546527800364e-05,
+      "loss": 0.8713,
+      "step": 1644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 34.692844200851354,
+      "learning_rate": 1.983824457420029e-05,
+      "loss": 0.984,
+      "step": 1645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.70194492425561,
+      "learning_rate": 1.983794234080629e-05,
+      "loss": 0.9004,
+      "step": 1646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.841944503640804,
+      "learning_rate": 1.9837639827626973e-05,
+      "loss": 0.8657,
+      "step": 1647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 29.375172863613965,
+      "learning_rate": 1.983733703467093e-05,
+      "loss": 0.9512,
+      "step": 1648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 25.117005789234447,
+      "learning_rate": 1.983703396194678e-05,
+      "loss": 0.9423,
+      "step": 1649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 22.010961772423034,
+      "learning_rate": 1.9836730609463134e-05,
+      "loss": 0.8295,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.48029181875695,
+      "learning_rate": 1.983642697722863e-05,
+      "loss": 0.8895,
+      "step": 1651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.08662586900136,
+      "learning_rate": 1.9836123065251892e-05,
+      "loss": 0.8811,
+      "step": 1652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.468787104680578,
+      "learning_rate": 1.9835818873541572e-05,
+      "loss": 0.8907,
+      "step": 1653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 42.75017693549901,
+      "learning_rate": 1.983551440210631e-05,
+      "loss": 1.0122,
+      "step": 1654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 37.603883522249845,
+      "learning_rate": 1.9835209650954775e-05,
+      "loss": 1.016,
+      "step": 1655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.902028193039648,
+      "learning_rate": 1.9834904620095636e-05,
+      "loss": 0.8202,
+      "step": 1656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.05601838766581,
+      "learning_rate": 1.9834599309537557e-05,
+      "loss": 0.8901,
+      "step": 1657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.000108851990827,
+      "learning_rate": 1.9834293719289227e-05,
+      "loss": 0.9947,
+      "step": 1658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 52.55156834831746,
+      "learning_rate": 1.9833987849359338e-05,
+      "loss": 1.0453,
+      "step": 1659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.137705460376477,
+      "learning_rate": 1.9833681699756592e-05,
+      "loss": 0.8133,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 24.64539404819129,
+      "learning_rate": 1.9833375270489682e-05,
+      "loss": 0.9252,
+      "step": 1661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 29.720110765826092,
+      "learning_rate": 1.983306856156734e-05,
+      "loss": 0.8509,
+      "step": 1662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.061173508564595,
+      "learning_rate": 1.9832761572998274e-05,
+      "loss": 0.8002,
+      "step": 1663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.487527124277147,
+      "learning_rate": 1.9832454304791224e-05,
+      "loss": 0.9104,
+      "step": 1664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.2215853506947,
+      "learning_rate": 1.9832146756954925e-05,
+      "loss": 0.9163,
+      "step": 1665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.120777669608227,
+      "learning_rate": 1.9831838929498123e-05,
+      "loss": 0.855,
+      "step": 1666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 33.39743699387739,
+      "learning_rate": 1.983153082242957e-05,
+      "loss": 0.8794,
+      "step": 1667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.19784093810489,
+      "learning_rate": 1.983122243575804e-05,
+      "loss": 0.7899,
+      "step": 1668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.899923464553627,
+      "learning_rate": 1.9830913769492288e-05,
+      "loss": 0.896,
+      "step": 1669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.8511313622163,
+      "learning_rate": 1.9830604823641103e-05,
+      "loss": 0.985,
+      "step": 1670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 31.44816884401661,
+      "learning_rate": 1.9830295598213263e-05,
+      "loss": 0.8494,
+      "step": 1671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.58686807623665,
+      "learning_rate": 1.9829986093217575e-05,
+      "loss": 0.9452,
+      "step": 1672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 29.097783068819563,
+      "learning_rate": 1.9829676308662826e-05,
+      "loss": 0.9155,
+      "step": 1673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.029132243761168,
+      "learning_rate": 1.9829366244557837e-05,
+      "loss": 0.8357,
+      "step": 1674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.770385605535186,
+      "learning_rate": 1.982905590091142e-05,
+      "loss": 0.8228,
+      "step": 1675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.883739146354582,
+      "learning_rate": 1.98287452777324e-05,
+      "loss": 0.9237,
+      "step": 1676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.522797857862518,
+      "learning_rate": 1.982843437502962e-05,
+      "loss": 0.9413,
+      "step": 1677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.670813278734217,
+      "learning_rate": 1.982812319281191e-05,
+      "loss": 0.7638,
+      "step": 1678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 22.824570396700686,
+      "learning_rate": 1.982781173108813e-05,
+      "loss": 0.8331,
+      "step": 1679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.102055059619059,
+      "learning_rate": 1.982749998986713e-05,
+      "loss": 0.9727,
+      "step": 1680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.613612673391664,
+      "learning_rate": 1.982718796915778e-05,
+      "loss": 0.9464,
+      "step": 1681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.58938208456188,
+      "learning_rate": 1.9826875668968956e-05,
+      "loss": 0.8991,
+      "step": 1682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.565050239136012,
+      "learning_rate": 1.9826563089309534e-05,
+      "loss": 0.8076,
+      "step": 1683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 33.96244031969681,
+      "learning_rate": 1.982625023018841e-05,
+      "loss": 1.0186,
+      "step": 1684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.330985058321044,
+      "learning_rate": 1.982593709161447e-05,
+      "loss": 0.8724,
+      "step": 1685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.11666479201521,
+      "learning_rate": 1.9825623673596634e-05,
+      "loss": 0.7933,
+      "step": 1686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.77564466044492,
+      "learning_rate": 1.9825309976143807e-05,
+      "loss": 0.897,
+      "step": 1687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.416525951954807,
+      "learning_rate": 1.982499599926491e-05,
+      "loss": 0.798,
+      "step": 1688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 26.160292368889735,
+      "learning_rate": 1.9824681742968876e-05,
+      "loss": 0.812,
+      "step": 1689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 34.88096240990854,
+      "learning_rate": 1.9824367207264638e-05,
+      "loss": 0.9091,
+      "step": 1690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 36.20317577149408,
+      "learning_rate": 1.9824052392161148e-05,
+      "loss": 0.9105,
+      "step": 1691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.68920813017233,
+      "learning_rate": 1.982373729766735e-05,
+      "loss": 0.8965,
+      "step": 1692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.33353376344843,
+      "learning_rate": 1.9823421923792213e-05,
+      "loss": 0.9168,
+      "step": 1693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 25.1289374411941,
+      "learning_rate": 1.98231062705447e-05,
+      "loss": 0.9557,
+      "step": 1694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 22.92384134919337,
+      "learning_rate": 1.9822790337933795e-05,
+      "loss": 0.9023,
+      "step": 1695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.142957059410243,
+      "learning_rate": 1.9822474125968478e-05,
+      "loss": 0.7806,
+      "step": 1696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.00885413543742,
+      "learning_rate": 1.9822157634657742e-05,
+      "loss": 0.8202,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.582682493009642,
+      "learning_rate": 1.9821840864010588e-05,
+      "loss": 0.8614,
+      "step": 1698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.35402551523379,
+      "learning_rate": 1.9821523814036026e-05,
+      "loss": 0.8824,
+      "step": 1699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.520273603716277,
+      "learning_rate": 1.9821206484743073e-05,
+      "loss": 0.8615,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.00684525162505,
+      "learning_rate": 1.9820888876140752e-05,
+      "loss": 0.8243,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.61042011017325,
+      "learning_rate": 1.9820570988238096e-05,
+      "loss": 0.8826,
+      "step": 1702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.07278778761485,
+      "learning_rate": 1.982025282104415e-05,
+      "loss": 0.7674,
+      "step": 1703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.739100708209826,
+      "learning_rate": 1.981993437456795e-05,
+      "loss": 0.8416,
+      "step": 1704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.387826597579927,
+      "learning_rate": 1.9819615648818568e-05,
+      "loss": 0.8929,
+      "step": 1705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 24.54800838449656,
+      "learning_rate": 1.9819296643805058e-05,
+      "loss": 0.947,
+      "step": 1706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 27.16360430129431,
+      "learning_rate": 1.98189773595365e-05,
+      "loss": 0.9132,
+      "step": 1707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.937256023756593,
+      "learning_rate": 1.9818657796021963e-05,
+      "loss": 0.8758,
+      "step": 1708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.491781477037854,
+      "learning_rate": 1.9818337953270544e-05,
+      "loss": 0.9212,
+      "step": 1709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 22.04077979195327,
+      "learning_rate": 1.981801783129134e-05,
+      "loss": 0.8816,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.636449283785787,
+      "learning_rate": 1.981769743009345e-05,
+      "loss": 0.8137,
+      "step": 1711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 29.107738028061473,
+      "learning_rate": 1.9817376749685994e-05,
+      "loss": 0.8802,
+      "step": 1712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 26.23181844435767,
+      "learning_rate": 1.981705579007808e-05,
+      "loss": 1.1021,
+      "step": 1713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.399339037603546,
+      "learning_rate": 1.9816734551278843e-05,
+      "loss": 0.8471,
+      "step": 1714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.343173807355683,
+      "learning_rate": 1.981641303329742e-05,
+      "loss": 0.938,
+      "step": 1715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.64286212425664,
+      "learning_rate": 1.981609123614295e-05,
+      "loss": 0.9954,
+      "step": 1716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 30.577373505150504,
+      "learning_rate": 1.981576915982459e-05,
+      "loss": 0.9168,
+      "step": 1717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 22.085219672148874,
+      "learning_rate": 1.9815446804351495e-05,
+      "loss": 0.8749,
+      "step": 1718
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.99618885817167,
+      "learning_rate": 1.9815124169732834e-05,
+      "loss": 0.8716,
+      "step": 1719
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 24.651759864107493,
+      "learning_rate": 1.9814801255977786e-05,
+      "loss": 0.9354,
+      "step": 1720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 23.447190975021126,
+      "learning_rate": 1.9814478063095533e-05,
+      "loss": 1.0228,
+      "step": 1721
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 27.748623575943913,
+      "learning_rate": 1.9814154591095262e-05,
+      "loss": 0.8155,
+      "step": 1722
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.932456202019589,
+      "learning_rate": 1.9813830839986177e-05,
+      "loss": 0.8719,
+      "step": 1723
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.80784905725702,
+      "learning_rate": 1.9813506809777483e-05,
+      "loss": 0.9,
+      "step": 1724
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.68492081003693,
+      "learning_rate": 1.9813182500478395e-05,
+      "loss": 0.9751,
+      "step": 1725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.15240916369025,
+      "learning_rate": 1.981285791209814e-05,
+      "loss": 0.8592,
+      "step": 1726
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.619245552542825,
+      "learning_rate": 1.9812533044645945e-05,
+      "loss": 0.9925,
+      "step": 1727
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 32.54050154142809,
+      "learning_rate": 1.981220789813105e-05,
+      "loss": 0.9115,
+      "step": 1728
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.369549080638826,
+      "learning_rate": 1.9811882472562703e-05,
+      "loss": 0.9084,
+      "step": 1729
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.557472124537792,
+      "learning_rate": 1.9811556767950158e-05,
+      "loss": 0.9147,
+      "step": 1730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.197237848997407,
+      "learning_rate": 1.981123078430268e-05,
+      "loss": 0.8611,
+      "step": 1731
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 23.070902991884388,
+      "learning_rate": 1.9810904521629534e-05,
+      "loss": 0.813,
+      "step": 1732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.363541672221512,
+      "learning_rate": 1.9810577979940006e-05,
+      "loss": 0.8972,
+      "step": 1733
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.100392094047965,
+      "learning_rate": 1.9810251159243378e-05,
+      "loss": 0.9072,
+      "step": 1734
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 23.17521346846764,
+      "learning_rate": 1.9809924059548944e-05,
+      "loss": 0.8894,
+      "step": 1735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.59224159932445,
+      "learning_rate": 1.980959668086601e-05,
+      "loss": 0.9422,
+      "step": 1736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 26.431800013074078,
+      "learning_rate": 1.9809269023203887e-05,
+      "loss": 0.8817,
+      "step": 1737
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.204420327753652,
+      "learning_rate": 1.9808941086571888e-05,
+      "loss": 0.8591,
+      "step": 1738
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 24.24064240124899,
+      "learning_rate": 1.9808612870979346e-05,
+      "loss": 0.8312,
+      "step": 1739
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.829441547252312,
+      "learning_rate": 1.9808284376435588e-05,
+      "loss": 0.7915,
+      "step": 1740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.05777680683614,
+      "learning_rate": 1.9807955602949962e-05,
+      "loss": 0.8875,
+      "step": 1741
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.82942523608552,
+      "learning_rate": 1.9807626550531816e-05,
+      "loss": 0.9477,
+      "step": 1742
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.526141031429328,
+      "learning_rate": 1.9807297219190508e-05,
+      "loss": 0.865,
+      "step": 1743
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.71654978111612,
+      "learning_rate": 1.9806967608935403e-05,
+      "loss": 0.8723,
+      "step": 1744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.122209608729346,
+      "learning_rate": 1.980663771977588e-05,
+      "loss": 0.8894,
+      "step": 1745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.36792613612658,
+      "learning_rate": 1.9806307551721313e-05,
+      "loss": 0.9009,
+      "step": 1746
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.668873128133317,
+      "learning_rate": 1.9805977104781096e-05,
+      "loss": 0.858,
+      "step": 1747
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.70445764376114,
+      "learning_rate": 1.9805646378964626e-05,
+      "loss": 0.8944,
+      "step": 1748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 24.62086584393321,
+      "learning_rate": 1.9805315374281314e-05,
+      "loss": 0.8313,
+      "step": 1749
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 26.815314747600738,
+      "learning_rate": 1.9804984090740563e-05,
+      "loss": 0.9241,
+      "step": 1750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 31.07602689554703,
+      "learning_rate": 1.9804652528351806e-05,
+      "loss": 0.8719,
+      "step": 1751
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.43906421023547,
+      "learning_rate": 1.980432068712446e-05,
+      "loss": 0.9143,
+      "step": 1752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.522436014708436,
+      "learning_rate": 1.9803988567067975e-05,
+      "loss": 0.9915,
+      "step": 1753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 22.50613093822696,
+      "learning_rate": 1.9803656168191786e-05,
+      "loss": 0.901,
+      "step": 1754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.236064247078225,
+      "learning_rate": 1.9803323490505352e-05,
+      "loss": 0.8827,
+      "step": 1755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.635590551628425,
+      "learning_rate": 1.9802990534018135e-05,
+      "loss": 0.8642,
+      "step": 1756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.513616355557605,
+      "learning_rate": 1.9802657298739598e-05,
+      "loss": 0.9179,
+      "step": 1757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.952587424952117,
+      "learning_rate": 1.9802323784679225e-05,
+      "loss": 0.8081,
+      "step": 1758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 34.995843557082345,
+      "learning_rate": 1.9801989991846495e-05,
+      "loss": 0.9588,
+      "step": 1759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.433058171333993,
+      "learning_rate": 1.9801655920250903e-05,
+      "loss": 0.9047,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 27.19085417697937,
+      "learning_rate": 1.9801321569901955e-05,
+      "loss": 0.8589,
+      "step": 1761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.490056017876276,
+      "learning_rate": 1.980098694080915e-05,
+      "loss": 0.7968,
+      "step": 1762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.330236244397957,
+      "learning_rate": 1.9800652032982013e-05,
+      "loss": 1.0385,
+      "step": 1763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.55127119426707,
+      "learning_rate": 1.9800316846430064e-05,
+      "loss": 0.9348,
+      "step": 1764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.18259629186154,
+      "learning_rate": 1.979998138116284e-05,
+      "loss": 0.9195,
+      "step": 1765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 20.289223537753408,
+      "learning_rate": 1.9799645637189875e-05,
+      "loss": 0.9477,
+      "step": 1766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.693346282413629,
+      "learning_rate": 1.979930961452072e-05,
+      "loss": 0.783,
+      "step": 1767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 25.326541681708854,
+      "learning_rate": 1.9798973313164934e-05,
+      "loss": 0.881,
+      "step": 1768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 31.889067523030125,
+      "learning_rate": 1.979863673313208e-05,
+      "loss": 0.9513,
+      "step": 1769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.741223782349362,
+      "learning_rate": 1.979829987443173e-05,
+      "loss": 0.881,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.742427767299679,
+      "learning_rate": 1.9797962737073456e-05,
+      "loss": 0.8532,
+      "step": 1771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 26.50449467949628,
+      "learning_rate": 1.9797625321066863e-05,
+      "loss": 1.0252,
+      "step": 1772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.619867725984435,
+      "learning_rate": 1.9797287626421534e-05,
+      "loss": 0.8524,
+      "step": 1773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 26.785282296602777,
+      "learning_rate": 1.9796949653147073e-05,
+      "loss": 0.8381,
+      "step": 1774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.71690646061289,
+      "learning_rate": 1.97966114012531e-05,
+      "loss": 0.9848,
+      "step": 1775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.441185843013914,
+      "learning_rate": 1.9796272870749225e-05,
+      "loss": 0.8482,
+      "step": 1776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.0942923208159,
+      "learning_rate": 1.9795934061645084e-05,
+      "loss": 0.8812,
+      "step": 1777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.787196562122961,
+      "learning_rate": 1.9795594973950305e-05,
+      "loss": 0.9134,
+      "step": 1778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 25.1654330487038,
+      "learning_rate": 1.979525560767454e-05,
+      "loss": 0.9075,
+      "step": 1779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.38385296156733,
+      "learning_rate": 1.9794915962827427e-05,
+      "loss": 0.7823,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.93106760878871,
+      "learning_rate": 1.9794576039418643e-05,
+      "loss": 0.8569,
+      "step": 1781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.028411733726802,
+      "learning_rate": 1.979423583745784e-05,
+      "loss": 0.8344,
+      "step": 1782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.801500937619682,
+      "learning_rate": 1.9793895356954704e-05,
+      "loss": 0.8412,
+      "step": 1783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.849405129691057,
+      "learning_rate": 1.9793554597918908e-05,
+      "loss": 0.9314,
+      "step": 1784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.8297964628564,
+      "learning_rate": 1.9793213560360153e-05,
+      "loss": 0.8966,
+      "step": 1785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.357410765379026,
+      "learning_rate": 1.9792872244288132e-05,
+      "loss": 0.8239,
+      "step": 1786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.137299185445123,
+      "learning_rate": 1.9792530649712552e-05,
+      "loss": 0.8157,
+      "step": 1787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.96843058682852,
+      "learning_rate": 1.979218877664313e-05,
+      "loss": 0.8441,
+      "step": 1788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.4538187763964,
+      "learning_rate": 1.9791846625089588e-05,
+      "loss": 0.7619,
+      "step": 1789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 29.816475872016984,
+      "learning_rate": 1.9791504195061655e-05,
+      "loss": 0.8856,
+      "step": 1790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.022937439097497,
+      "learning_rate": 1.9791161486569072e-05,
+      "loss": 0.9113,
+      "step": 1791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.92024545155266,
+      "learning_rate": 1.979081849962158e-05,
+      "loss": 0.984,
+      "step": 1792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.34857719844794,
+      "learning_rate": 1.979047523422894e-05,
+      "loss": 0.8896,
+      "step": 1793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.357809398328623,
+      "learning_rate": 1.9790131690400913e-05,
+      "loss": 0.9505,
+      "step": 1794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.204969814461755,
+      "learning_rate": 1.9789787868147265e-05,
+      "loss": 0.7692,
+      "step": 1795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.921512275834388,
+      "learning_rate": 1.9789443767477778e-05,
+      "loss": 0.9257,
+      "step": 1796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.092290704193367,
+      "learning_rate": 1.978909938840224e-05,
+      "loss": 0.8839,
+      "step": 1797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.555757245780182,
+      "learning_rate": 1.9788754730930436e-05,
+      "loss": 1.0024,
+      "step": 1798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.588832850024062,
+      "learning_rate": 1.9788409795072175e-05,
+      "loss": 0.8748,
+      "step": 1799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.08744310269793,
+      "learning_rate": 1.978806458083727e-05,
+      "loss": 0.9688,
+      "step": 1800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.231640853636327,
+      "learning_rate": 1.978771908823553e-05,
+      "loss": 0.8112,
+      "step": 1801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.850053553523697,
+      "learning_rate": 1.978737331727679e-05,
+      "loss": 0.9087,
+      "step": 1802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.854705950518575,
+      "learning_rate": 1.9787027267970873e-05,
+      "loss": 0.8585,
+      "step": 1803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 26.570283879099517,
+      "learning_rate": 1.9786680940327626e-05,
+      "loss": 0.9648,
+      "step": 1804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.773776774758392,
+      "learning_rate": 1.97863343343569e-05,
+      "loss": 0.762,
+      "step": 1805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.047349295762505,
+      "learning_rate": 1.978598745006855e-05,
+      "loss": 0.8808,
+      "step": 1806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.56834834060168,
+      "learning_rate": 1.978564028747244e-05,
+      "loss": 0.889,
+      "step": 1807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.922200495363604,
+      "learning_rate": 1.9785292846578446e-05,
+      "loss": 0.8565,
+      "step": 1808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 27.401746957361137,
+      "learning_rate": 1.9784945127396446e-05,
+      "loss": 0.9528,
+      "step": 1809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.761854402797546,
+      "learning_rate": 1.9784597129936333e-05,
+      "loss": 0.753,
+      "step": 1810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.52069704734337,
+      "learning_rate": 1.9784248854208e-05,
+      "loss": 0.9209,
+      "step": 1811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.558703475252177,
+      "learning_rate": 1.9783900300221354e-05,
+      "loss": 0.7836,
+      "step": 1812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 20.251583693824134,
+      "learning_rate": 1.9783551467986307e-05,
+      "loss": 0.873,
+      "step": 1813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 26.622178118444687,
+      "learning_rate": 1.978320235751278e-05,
+      "loss": 0.9453,
+      "step": 1814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.12187551550468,
+      "learning_rate": 1.97828529688107e-05,
+      "loss": 0.9581,
+      "step": 1815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.286965899324226,
+      "learning_rate": 1.9782503301890004e-05,
+      "loss": 0.9024,
+      "step": 1816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.827516362290183,
+      "learning_rate": 1.9782153356760635e-05,
+      "loss": 0.8556,
+      "step": 1817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.34215166787253,
+      "learning_rate": 1.978180313343255e-05,
+      "loss": 0.8911,
+      "step": 1818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.600726753311086,
+      "learning_rate": 1.9781452631915705e-05,
+      "loss": 0.8825,
+      "step": 1819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.253950137496588,
+      "learning_rate": 1.9781101852220074e-05,
+      "loss": 0.8998,
+      "step": 1820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.370409260145085,
+      "learning_rate": 1.978075079435562e-05,
+      "loss": 0.9154,
+      "step": 1821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.678942426605634,
+      "learning_rate": 1.978039945833234e-05,
+      "loss": 0.9338,
+      "step": 1822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.99321255846877,
+      "learning_rate": 1.978004784416022e-05,
+      "loss": 0.9495,
+      "step": 1823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.751243995365353,
+      "learning_rate": 1.977969595184926e-05,
+      "loss": 0.895,
+      "step": 1824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.825625049592,
+      "learning_rate": 1.9779343781409466e-05,
+      "loss": 0.8204,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.513645984582833,
+      "learning_rate": 1.9778991332850857e-05,
+      "loss": 0.8813,
+      "step": 1826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.849047481639463,
+      "learning_rate": 1.9778638606183453e-05,
+      "loss": 1.0108,
+      "step": 1827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.838968960727696,
+      "learning_rate": 1.977828560141729e-05,
+      "loss": 0.8283,
+      "step": 1828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.336703018712903,
+      "learning_rate": 1.9777932318562407e-05,
+      "loss": 0.9114,
+      "step": 1829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 23.559767149740388,
+      "learning_rate": 1.9777578757628845e-05,
+      "loss": 0.9584,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.806822963396797,
+      "learning_rate": 1.9777224918626662e-05,
+      "loss": 0.8323,
+      "step": 1831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.578962190177567,
+      "learning_rate": 1.9776870801565924e-05,
+      "loss": 0.9118,
+      "step": 1832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 23.522935496446518,
+      "learning_rate": 1.97765164064567e-05,
+      "loss": 0.8278,
+      "step": 1833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.73905184759165,
+      "learning_rate": 1.977616173330907e-05,
+      "loss": 0.8432,
+      "step": 1834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.574097636000438,
+      "learning_rate": 1.9775806782133114e-05,
+      "loss": 0.7777,
+      "step": 1835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 32.37142113747569,
+      "learning_rate": 1.9775451552938937e-05,
+      "loss": 0.9256,
+      "step": 1836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 23.42497999339028,
+      "learning_rate": 1.9775096045736634e-05,
+      "loss": 0.8323,
+      "step": 1837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 21.867717619627314,
+      "learning_rate": 1.9774740260536318e-05,
+      "loss": 0.7863,
+      "step": 1838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.43668407151477,
+      "learning_rate": 1.977438419734811e-05,
+      "loss": 0.8583,
+      "step": 1839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 23.98656454871355,
+      "learning_rate": 1.977402785618213e-05,
+      "loss": 0.9369,
+      "step": 1840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.480920694582508,
+      "learning_rate": 1.9773671237048514e-05,
+      "loss": 0.8054,
+      "step": 1841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.789256845271577,
+      "learning_rate": 1.977331433995741e-05,
+      "loss": 0.9995,
+      "step": 1842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.08763435503969,
+      "learning_rate": 1.977295716491896e-05,
+      "loss": 0.8417,
+      "step": 1843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 42.645517418034,
+      "learning_rate": 1.9772599711943323e-05,
+      "loss": 0.9394,
+      "step": 1844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.063569339318114,
+      "learning_rate": 1.977224198104067e-05,
+      "loss": 0.864,
+      "step": 1845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 21.15715535845419,
+      "learning_rate": 1.9771883972221174e-05,
+      "loss": 0.8319,
+      "step": 1846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 27.82915968741137,
+      "learning_rate": 1.977152568549501e-05,
+      "loss": 0.9725,
+      "step": 1847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 33.0771888236441,
+      "learning_rate": 1.9771167120872375e-05,
+      "loss": 0.8645,
+      "step": 1848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.287435605199015,
+      "learning_rate": 1.977080827836346e-05,
+      "loss": 0.9815,
+      "step": 1849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.977668332700398,
+      "learning_rate": 1.9770449157978475e-05,
+      "loss": 0.7676,
+      "step": 1850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.51914198257357,
+      "learning_rate": 1.9770089759727634e-05,
+      "loss": 0.8214,
+      "step": 1851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 22.3756017737208,
+      "learning_rate": 1.976973008362115e-05,
+      "loss": 0.8112,
+      "step": 1852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 22.287185008240183,
+      "learning_rate": 1.976937012966926e-05,
+      "loss": 0.7605,
+      "step": 1853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.322853428520098,
+      "learning_rate": 1.97690098978822e-05,
+      "loss": 0.9402,
+      "step": 1854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 38.27820211694408,
+      "learning_rate": 1.9768649388270213e-05,
+      "loss": 0.8972,
+      "step": 1855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.129150240127114,
+      "learning_rate": 1.976828860084355e-05,
+      "loss": 0.87,
+      "step": 1856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.448595222631848,
+      "learning_rate": 1.9767927535612473e-05,
+      "loss": 0.8432,
+      "step": 1857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.16610402139183,
+      "learning_rate": 1.9767566192587253e-05,
+      "loss": 0.8021,
+      "step": 1858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.94836865062978,
+      "learning_rate": 1.9767204571778163e-05,
+      "loss": 0.9226,
+      "step": 1859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.758907658204592,
+      "learning_rate": 1.976684267319549e-05,
+      "loss": 0.8218,
+      "step": 1860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.600530844694067,
+      "learning_rate": 1.9766480496849527e-05,
+      "loss": 0.8165,
+      "step": 1861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.589273431906122,
+      "learning_rate": 1.9766118042750566e-05,
+      "loss": 0.7289,
+      "step": 1862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.449142527324344,
+      "learning_rate": 1.9765755310908925e-05,
+      "loss": 0.8735,
+      "step": 1863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.13288460527496,
+      "learning_rate": 1.9765392301334912e-05,
+      "loss": 0.8423,
+      "step": 1864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.610665193675995,
+      "learning_rate": 1.976502901403886e-05,
+      "loss": 0.9759,
+      "step": 1865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.8569561545966,
+      "learning_rate": 1.9764665449031093e-05,
+      "loss": 0.9654,
+      "step": 1866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.353380535986272,
+      "learning_rate": 1.976430160632195e-05,
+      "loss": 0.8497,
+      "step": 1867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 30.508107354647823,
+      "learning_rate": 1.9763937485921785e-05,
+      "loss": 0.9529,
+      "step": 1868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 22.969245304229776,
+      "learning_rate": 1.976357308784095e-05,
+      "loss": 0.8159,
+      "step": 1869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 37.45412320798896,
+      "learning_rate": 1.9763208412089802e-05,
+      "loss": 0.785,
+      "step": 1870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.87231431358377,
+      "learning_rate": 1.9762843458678723e-05,
+      "loss": 0.8465,
+      "step": 1871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.83016601028346,
+      "learning_rate": 1.976247822761809e-05,
+      "loss": 0.8687,
+      "step": 1872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.515231235357167,
+      "learning_rate": 1.976211271891828e-05,
+      "loss": 0.9696,
+      "step": 1873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.620431983707377,
+      "learning_rate": 1.97617469325897e-05,
+      "loss": 0.8897,
+      "step": 1874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 28.371820075238965,
+      "learning_rate": 1.9761380868642745e-05,
+      "loss": 1.0011,
+      "step": 1875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.977174386233976,
+      "learning_rate": 1.9761014527087828e-05,
+      "loss": 0.9175,
+      "step": 1876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.64465399427102,
+      "learning_rate": 1.9760647907935372e-05,
+      "loss": 0.7933,
+      "step": 1877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.79532450577986,
+      "learning_rate": 1.9760281011195793e-05,
+      "loss": 0.8207,
+      "step": 1878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.920172704359203,
+      "learning_rate": 1.9759913836879535e-05,
+      "loss": 0.8858,
+      "step": 1879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.638220166399208,
+      "learning_rate": 1.9759546384997042e-05,
+      "loss": 0.9311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.781828814539178,
+      "learning_rate": 1.975917865555875e-05,
+      "loss": 0.8638,
+      "step": 1881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 21.054452944353176,
+      "learning_rate": 1.9758810648575132e-05,
+      "loss": 0.8943,
+      "step": 1882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.283434521357467,
+      "learning_rate": 1.9758442364056643e-05,
+      "loss": 0.873,
+      "step": 1883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.699712329327701,
+      "learning_rate": 1.975807380201377e-05,
+      "loss": 0.8444,
+      "step": 1884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.492083412886533,
+      "learning_rate": 1.9757704962456978e-05,
+      "loss": 0.9192,
+      "step": 1885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 27.698089892390986,
+      "learning_rate": 1.9757335845396768e-05,
+      "loss": 0.898,
+      "step": 1886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 35.97715392895058,
+      "learning_rate": 1.9756966450843635e-05,
+      "loss": 0.9176,
+      "step": 1887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.065809891214382,
+      "learning_rate": 1.9756596778808085e-05,
+      "loss": 0.862,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.38781764166271,
+      "learning_rate": 1.9756226829300628e-05,
+      "loss": 0.8988,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.457850544725364,
+      "learning_rate": 1.9755856602331788e-05,
+      "loss": 0.8768,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.788459257811533,
+      "learning_rate": 1.9755486097912092e-05,
+      "loss": 0.8156,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 23.2677302948353,
+      "learning_rate": 1.9755115316052076e-05,
+      "loss": 0.7801,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.212946392362642,
+      "learning_rate": 1.9754744256762293e-05,
+      "loss": 0.8462,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 24.414810513443292,
+      "learning_rate": 1.9754372920053284e-05,
+      "loss": 0.9012,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.37392721590177,
+      "learning_rate": 1.975400130593562e-05,
+      "loss": 0.8598,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 25.597424221282452,
+      "learning_rate": 1.975362941441986e-05,
+      "loss": 0.9248,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.675500822089871,
+      "learning_rate": 1.975325724551659e-05,
+      "loss": 0.9456,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.55818149337485,
+      "learning_rate": 1.9752884799236387e-05,
+      "loss": 0.7981,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 25.309976139340897,
+      "learning_rate": 1.9752512075589844e-05,
+      "loss": 0.9044,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.143745146314007,
+      "learning_rate": 1.9752139074587563e-05,
+      "loss": 0.8209,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.12003667117563,
+      "learning_rate": 1.9751765796240153e-05,
+      "loss": 0.8165,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.426752350094178,
+      "learning_rate": 1.9751392240558228e-05,
+      "loss": 0.9043,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.94066089000282,
+      "learning_rate": 1.9751018407552417e-05,
+      "loss": 0.9336,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.89384598324315,
+      "learning_rate": 1.9750644297233342e-05,
+      "loss": 0.8236,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.065600030567253,
+      "learning_rate": 1.9750269909611645e-05,
+      "loss": 0.9207,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.548708685621445,
+      "learning_rate": 1.9749895244697978e-05,
+      "loss": 0.8082,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.02616298028141,
+      "learning_rate": 1.9749520302502993e-05,
+      "loss": 0.7811,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.92885723191646,
+      "learning_rate": 1.9749145083037354e-05,
+      "loss": 0.8268,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 21.275090474531414,
+      "learning_rate": 1.9748769586311732e-05,
+      "loss": 0.7387,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.556192712972138,
+      "learning_rate": 1.9748393812336807e-05,
+      "loss": 0.798,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.336131973739484,
+      "learning_rate": 1.9748017761123265e-05,
+      "loss": 0.8965,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.347891191708836,
+      "learning_rate": 1.97476414326818e-05,
+      "loss": 0.925,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 26.474903759362682,
+      "learning_rate": 1.9747264827023113e-05,
+      "loss": 0.8234,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.728841426987962,
+      "learning_rate": 1.974688794415792e-05,
+      "loss": 0.9392,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.138012152042396,
+      "learning_rate": 1.9746510784096928e-05,
+      "loss": 0.7939,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.234423786012925,
+      "learning_rate": 1.974613334685088e-05,
+      "loss": 0.8185,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.95480920560591,
+      "learning_rate": 1.9745755632430496e-05,
+      "loss": 0.925,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 21.165180219028983,
+      "learning_rate": 1.9745377640846523e-05,
+      "loss": 0.9045,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.59620959701154,
+      "learning_rate": 1.974499937210971e-05,
+      "loss": 0.8652,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.839764076434742,
+      "learning_rate": 1.974462082623082e-05,
+      "loss": 0.7006,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 34.00717524065735,
+      "learning_rate": 1.9744242003220607e-05,
+      "loss": 0.9767,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 38.31057661931976,
+      "learning_rate": 1.9743862903089857e-05,
+      "loss": 0.9409,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 25.645945820431308,
+      "learning_rate": 1.9743483525849345e-05,
+      "loss": 0.8992,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.038228969081157,
+      "learning_rate": 1.974310387150986e-05,
+      "loss": 0.8825,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.159937852966443,
+      "learning_rate": 1.9742723940082203e-05,
+      "loss": 0.8062,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.16880505023091,
+      "learning_rate": 1.9742343731577173e-05,
+      "loss": 0.8642,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.390617242801474,
+      "learning_rate": 1.9741963246005592e-05,
+      "loss": 0.8687,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.389926695388457,
+      "learning_rate": 1.974158248337827e-05,
+      "loss": 0.8399,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.723959075932196,
+      "learning_rate": 1.9741201443706042e-05,
+      "loss": 0.7871,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.139668398846528,
+      "learning_rate": 1.9740820126999743e-05,
+      "loss": 0.9488,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 26.27604148544649,
+      "learning_rate": 1.9740438533270218e-05,
+      "loss": 0.8128,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.72623610921065,
+      "learning_rate": 1.9740056662528318e-05,
+      "loss": 0.8657,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 24.8832880342997,
+      "learning_rate": 1.9739674514784906e-05,
+      "loss": 0.8405,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.337330406431807,
+      "learning_rate": 1.973929209005085e-05,
+      "loss": 0.918,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 30.514234999501607,
+      "learning_rate": 1.9738909388337023e-05,
+      "loss": 0.9271,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.198109752315794,
+      "learning_rate": 1.973852640965431e-05,
+      "loss": 0.7544,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.291033052491546,
+      "learning_rate": 1.9738143154013603e-05,
+      "loss": 0.9277,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 23.874831114790798,
+      "learning_rate": 1.9737759621425797e-05,
+      "loss": 0.8836,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.85217318481541,
+      "learning_rate": 1.973737581190181e-05,
+      "loss": 0.8899,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.54350282921674,
+      "learning_rate": 1.973699172545255e-05,
+      "loss": 0.7964,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 26.825471069158436,
+      "learning_rate": 1.973660736208894e-05,
+      "loss": 0.9965,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 30.974556993406168,
+      "learning_rate": 1.973622272182191e-05,
+      "loss": 0.9037,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.06262056315065,
+      "learning_rate": 1.9735837804662404e-05,
+      "loss": 0.9677,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.281524040440718,
+      "learning_rate": 1.9735452610621366e-05,
+      "loss": 0.7922,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 23.790383033869862,
+      "learning_rate": 1.9735067139709753e-05,
+      "loss": 0.8615,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 28.5259578759439,
+      "learning_rate": 1.973468139193852e-05,
+      "loss": 0.9008,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 30.496352395834368,
+      "learning_rate": 1.9734295367318648e-05,
+      "loss": 0.9198,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.299607119627156,
+      "learning_rate": 1.9733909065861107e-05,
+      "loss": 0.8284,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.589582219757954,
+      "learning_rate": 1.9733522487576884e-05,
+      "loss": 0.8235,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 25.360756131942562,
+      "learning_rate": 1.973313563247698e-05,
+      "loss": 0.8538,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 23.774548061178905,
+      "learning_rate": 1.9732748500572395e-05,
+      "loss": 0.8256,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.7938465135125,
+      "learning_rate": 1.973236109187413e-05,
+      "loss": 0.8148,
+      "step": 1952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.9068540534322,
+      "learning_rate": 1.973197340639321e-05,
+      "loss": 0.8526,
+      "step": 1953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.386475881564493,
+      "learning_rate": 1.973158544414066e-05,
+      "loss": 0.8164,
+      "step": 1954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.91021109918913,
+      "learning_rate": 1.9731197205127514e-05,
+      "loss": 0.7654,
+      "step": 1955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.011544735068918,
+      "learning_rate": 1.973080868936481e-05,
+      "loss": 0.8016,
+      "step": 1956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.43234175575187,
+      "learning_rate": 1.9730419896863603e-05,
+      "loss": 0.7588,
+      "step": 1957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.63898731879432,
+      "learning_rate": 1.9730030827634942e-05,
+      "loss": 0.9247,
+      "step": 1958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.99724500329837,
+      "learning_rate": 1.9729641481689897e-05,
+      "loss": 0.8635,
+      "step": 1959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.48537461833763,
+      "learning_rate": 1.972925185903954e-05,
+      "loss": 0.8497,
+      "step": 1960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.691299742929573,
+      "learning_rate": 1.9728861959694957e-05,
+      "loss": 0.8473,
+      "step": 1961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.4133431263369,
+      "learning_rate": 1.9728471783667226e-05,
+      "loss": 0.7965,
+      "step": 1962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.797636299579123,
+      "learning_rate": 1.9728081330967448e-05,
+      "loss": 0.8817,
+      "step": 1963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.551570678047835,
+      "learning_rate": 1.972769060160673e-05,
+      "loss": 0.9161,
+      "step": 1964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.364141629164624,
+      "learning_rate": 1.9727299595596178e-05,
+      "loss": 0.9401,
+      "step": 1965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.65152058387248,
+      "learning_rate": 1.9726908312946918e-05,
+      "loss": 0.7555,
+      "step": 1966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.519633299946516,
+      "learning_rate": 1.9726516753670078e-05,
+      "loss": 0.9234,
+      "step": 1967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.187621768784943,
+      "learning_rate": 1.972612491777679e-05,
+      "loss": 0.8401,
+      "step": 1968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.113322729939835,
+      "learning_rate": 1.9725732805278198e-05,
+      "loss": 0.8784,
+      "step": 1969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.121764926298493,
+      "learning_rate": 1.9725340416185456e-05,
+      "loss": 0.7902,
+      "step": 1970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.984586533215857,
+      "learning_rate": 1.9724947750509718e-05,
+      "loss": 0.8074,
+      "step": 1971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.058733451536096,
+      "learning_rate": 1.9724554808262157e-05,
+      "loss": 0.9017,
+      "step": 1972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 32.59790955222826,
+      "learning_rate": 1.9724161589453948e-05,
+      "loss": 1.061,
+      "step": 1973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.22872112332218,
+      "learning_rate": 1.9723768094096268e-05,
+      "loss": 0.8682,
+      "step": 1974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.163242602012352,
+      "learning_rate": 1.9723374322200315e-05,
+      "loss": 0.8198,
+      "step": 1975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.033946309405977,
+      "learning_rate": 1.9722980273777284e-05,
+      "loss": 0.7993,
+      "step": 1976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 28.124299212043177,
+      "learning_rate": 1.9722585948838378e-05,
+      "loss": 0.7894,
+      "step": 1977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.55154649185102,
+      "learning_rate": 1.972219134739482e-05,
+      "loss": 0.8277,
+      "step": 1978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.439044298984843,
+      "learning_rate": 1.9721796469457827e-05,
+      "loss": 0.8154,
+      "step": 1979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.263510828712764,
+      "learning_rate": 1.9721401315038628e-05,
+      "loss": 0.867,
+      "step": 1980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.413829884031742,
+      "learning_rate": 1.972100588414846e-05,
+      "loss": 0.8369,
+      "step": 1981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.421277764785366,
+      "learning_rate": 1.9720610176798574e-05,
+      "loss": 0.9176,
+      "step": 1982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.591302858287282,
+      "learning_rate": 1.972021419300022e-05,
+      "loss": 0.8975,
+      "step": 1983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.44606981628151,
+      "learning_rate": 1.9719817932764662e-05,
+      "loss": 0.7829,
+      "step": 1984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 25.60036458020331,
+      "learning_rate": 1.9719421396103167e-05,
+      "loss": 0.7589,
+      "step": 1985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.488065013806025,
+      "learning_rate": 1.9719024583027012e-05,
+      "loss": 0.9107,
+      "step": 1986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.800749859434116,
+      "learning_rate": 1.9718627493547486e-05,
+      "loss": 0.744,
+      "step": 1987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.87122036874998,
+      "learning_rate": 1.9718230127675877e-05,
+      "loss": 0.8792,
+      "step": 1988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.451930796982744,
+      "learning_rate": 1.971783248542349e-05,
+      "loss": 0.7478,
+      "step": 1989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 23.369709856981036,
+      "learning_rate": 1.9717434566801633e-05,
+      "loss": 0.9079,
+      "step": 1990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.6883050569356,
+      "learning_rate": 1.9717036371821615e-05,
+      "loss": 0.8945,
+      "step": 1991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.590627946723554,
+      "learning_rate": 1.9716637900494775e-05,
+      "loss": 0.739,
+      "step": 1992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 32.17525068408096,
+      "learning_rate": 1.9716239152832434e-05,
+      "loss": 0.9047,
+      "step": 1993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.223546792008175,
+      "learning_rate": 1.9715840128845934e-05,
+      "loss": 0.746,
+      "step": 1994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 26.474822254232627,
+      "learning_rate": 1.9715440828546626e-05,
+      "loss": 0.9029,
+      "step": 1995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.990517024875519,
+      "learning_rate": 1.9715041251945863e-05,
+      "loss": 0.7252,
+      "step": 1996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 23.633635011972196,
+      "learning_rate": 1.9714641399055007e-05,
+      "loss": 0.8276,
+      "step": 1997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 30.676383652935883,
+      "learning_rate": 1.9714241269885436e-05,
+      "loss": 0.9074,
+      "step": 1998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.616173736210115,
+      "learning_rate": 1.9713840864448528e-05,
+      "loss": 0.8446,
+      "step": 1999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.878151480522458,
+      "learning_rate": 1.9713440182755665e-05,
+      "loss": 0.8643,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.67374095374624,
+      "learning_rate": 1.9713039224818247e-05,
+      "loss": 0.8124,
+      "step": 2001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.68957343006568,
+      "learning_rate": 1.9712637990647674e-05,
+      "loss": 0.8725,
+      "step": 2002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.224064555261798,
+      "learning_rate": 1.971223648025536e-05,
+      "loss": 0.8174,
+      "step": 2003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.320630612895478,
+      "learning_rate": 1.9711834693652722e-05,
+      "loss": 0.8064,
+      "step": 2004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.548316005124242,
+      "learning_rate": 1.9711432630851182e-05,
+      "loss": 0.8951,
+      "step": 2005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.668361165424816,
+      "learning_rate": 1.9711030291862187e-05,
+      "loss": 0.8711,
+      "step": 2006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.07097755932726,
+      "learning_rate": 1.9710627676697167e-05,
+      "loss": 0.8192,
+      "step": 2007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.01017858545481,
+      "learning_rate": 1.9710224785367575e-05,
+      "loss": 0.8742,
+      "step": 2008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.438593319146708,
+      "learning_rate": 1.9709821617884873e-05,
+      "loss": 0.9124,
+      "step": 2009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.11446710507202,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.8485,
+      "step": 2010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.509390048068454,
+      "learning_rate": 1.9709014454506e-05,
+      "loss": 0.8336,
+      "step": 2011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.8540006601597,
+      "learning_rate": 1.9708610458632783e-05,
+      "loss": 0.9212,
+      "step": 2012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.426905372360444,
+      "learning_rate": 1.970820618665237e-05,
+      "loss": 0.8895,
+      "step": 2013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.478918339061252,
+      "learning_rate": 1.9707801638576246e-05,
+      "loss": 0.8906,
+      "step": 2014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.198459180637528,
+      "learning_rate": 1.9707396814415927e-05,
+      "loss": 0.8595,
+      "step": 2015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 23.220622442841528,
+      "learning_rate": 1.970699171418292e-05,
+      "loss": 0.8944,
+      "step": 2016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.090429001609277,
+      "learning_rate": 1.9706586337888744e-05,
+      "loss": 0.9746,
+      "step": 2017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 37.80171079558046,
+      "learning_rate": 1.9706180685544933e-05,
+      "loss": 0.9084,
+      "step": 2018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.693300957325114,
+      "learning_rate": 1.9705774757163022e-05,
+      "loss": 0.7714,
+      "step": 2019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 36.16578761922622,
+      "learning_rate": 1.9705368552754554e-05,
+      "loss": 0.9715,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 30.37952921600066,
+      "learning_rate": 1.970496207233108e-05,
+      "loss": 1.0402,
+      "step": 2021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.97949704174473,
+      "learning_rate": 1.9704555315904164e-05,
+      "loss": 0.7923,
+      "step": 2022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.425513981446954,
+      "learning_rate": 1.9704148283485374e-05,
+      "loss": 0.8088,
+      "step": 2023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.737425028472487,
+      "learning_rate": 1.9703740975086282e-05,
+      "loss": 0.9466,
+      "step": 2024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.688795316457627,
+      "learning_rate": 1.970333339071847e-05,
+      "loss": 0.788,
+      "step": 2025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.67685110749523,
+      "learning_rate": 1.9702925530393535e-05,
+      "loss": 0.8963,
+      "step": 2026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.37720486713858,
+      "learning_rate": 1.970251739412307e-05,
+      "loss": 0.8601,
+      "step": 2027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.424242169193125,
+      "learning_rate": 1.9702108981918693e-05,
+      "loss": 0.8474,
+      "step": 2028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.64494758115085,
+      "learning_rate": 1.970170029379201e-05,
+      "loss": 0.9745,
+      "step": 2029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.170713737276607,
+      "learning_rate": 1.9701291329754645e-05,
+      "loss": 0.7781,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 26.473651958448663,
+      "learning_rate": 1.970088208981823e-05,
+      "loss": 0.8505,
+      "step": 2031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 25.099027657055245,
+      "learning_rate": 1.9700472573994403e-05,
+      "loss": 0.8165,
+      "step": 2032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.389292650100625,
+      "learning_rate": 1.970006278229481e-05,
+      "loss": 0.8138,
+      "step": 2033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.028228690501816,
+      "learning_rate": 1.9699652714731106e-05,
+      "loss": 0.8045,
+      "step": 2034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.785452550579784,
+      "learning_rate": 1.9699242371314952e-05,
+      "loss": 0.8292,
+      "step": 2035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.201463608848353,
+      "learning_rate": 1.9698831752058023e-05,
+      "loss": 0.863,
+      "step": 2036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.340812277280527,
+      "learning_rate": 1.969842085697199e-05,
+      "loss": 0.938,
+      "step": 2037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.485254576668932,
+      "learning_rate": 1.969800968606854e-05,
+      "loss": 0.8792,
+      "step": 2038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.50361377495418,
+      "learning_rate": 1.9697598239359368e-05,
+      "loss": 0.9366,
+      "step": 2039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.004593364380916,
+      "learning_rate": 1.9697186516856177e-05,
+      "loss": 0.8729,
+      "step": 2040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.27001093393149,
+      "learning_rate": 1.9696774518570673e-05,
+      "loss": 0.81,
+      "step": 2041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.275109917781354,
+      "learning_rate": 1.9696362244514576e-05,
+      "loss": 0.8853,
+      "step": 2042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.747320829069553,
+      "learning_rate": 1.9695949694699604e-05,
+      "loss": 0.8787,
+      "step": 2043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.324589701361806,
+      "learning_rate": 1.96955368691375e-05,
+      "loss": 0.9621,
+      "step": 2044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.62062350748662,
+      "learning_rate": 1.9695123767839995e-05,
+      "loss": 0.7259,
+      "step": 2045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.868806449305948,
+      "learning_rate": 1.9694710390818844e-05,
+      "loss": 0.8744,
+      "step": 2046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.289745679185543,
+      "learning_rate": 1.9694296738085802e-05,
+      "loss": 0.8401,
+      "step": 2047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.319830790116374,
+      "learning_rate": 1.9693882809652626e-05,
+      "loss": 0.8339,
+      "step": 2048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.35813448628382,
+      "learning_rate": 1.96934686055311e-05,
+      "loss": 0.8886,
+      "step": 2049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.498450702459422,
+      "learning_rate": 1.9693054125732995e-05,
+      "loss": 0.9545,
+      "step": 2050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.41585665636211,
+      "learning_rate": 1.96926393702701e-05,
+      "loss": 0.7897,
+      "step": 2051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.60940300575881,
+      "learning_rate": 1.9692224339154215e-05,
+      "loss": 0.8198,
+      "step": 2052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.23249166491358,
+      "learning_rate": 1.9691809032397135e-05,
+      "loss": 0.7522,
+      "step": 2053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.991811089588687,
+      "learning_rate": 1.9691393450010675e-05,
+      "loss": 0.8621,
+      "step": 2054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 26.00682109812143,
+      "learning_rate": 1.969097759200666e-05,
+      "loss": 0.9021,
+      "step": 2055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.863507928510778,
+      "learning_rate": 1.969056145839691e-05,
+      "loss": 0.8484,
+      "step": 2056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.320363565333135,
+      "learning_rate": 1.9690145049193257e-05,
+      "loss": 0.8124,
+      "step": 2057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 24.337468244865462,
+      "learning_rate": 1.968972836440755e-05,
+      "loss": 1.0166,
+      "step": 2058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.886848211054595,
+      "learning_rate": 1.9689311404051638e-05,
+      "loss": 0.9355,
+      "step": 2059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.150662392893672,
+      "learning_rate": 1.9688894168137372e-05,
+      "loss": 0.8225,
+      "step": 2060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.100422275118852,
+      "learning_rate": 1.9688476656676628e-05,
+      "loss": 0.8357,
+      "step": 2061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.912121041966605,
+      "learning_rate": 1.9688058869681277e-05,
+      "loss": 0.7751,
+      "step": 2062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.385547289120343,
+      "learning_rate": 1.96876408071632e-05,
+      "loss": 0.9013,
+      "step": 2063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.674005406574164,
+      "learning_rate": 1.968722246913428e-05,
+      "loss": 0.8803,
+      "step": 2064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.651857143774105,
+      "learning_rate": 1.9686803855606422e-05,
+      "loss": 0.8493,
+      "step": 2065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.83345569066038,
+      "learning_rate": 1.968638496659153e-05,
+      "loss": 0.8327,
+      "step": 2066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.981675972967828,
+      "learning_rate": 1.9685965802101517e-05,
+      "loss": 0.8561,
+      "step": 2067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.980884605410015,
+      "learning_rate": 1.96855463621483e-05,
+      "loss": 1.0125,
+      "step": 2068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.170432441982427,
+      "learning_rate": 1.9685126646743814e-05,
+      "loss": 0.7933,
+      "step": 2069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 25.31564892738416,
+      "learning_rate": 1.9684706655899988e-05,
+      "loss": 1.0028,
+      "step": 2070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 27.402884714954798,
+      "learning_rate": 1.968428638962877e-05,
+      "loss": 0.8449,
+      "step": 2071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.701878523820394,
+      "learning_rate": 1.9683865847942116e-05,
+      "loss": 0.8012,
+      "step": 2072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.698104536417045,
+      "learning_rate": 1.968344503085198e-05,
+      "loss": 0.9933,
+      "step": 2073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.286533086458956,
+      "learning_rate": 1.968302393837033e-05,
+      "loss": 0.8337,
+      "step": 2074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.971950636545188,
+      "learning_rate": 1.9682602570509147e-05,
+      "loss": 0.8593,
+      "step": 2075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.314689908288504,
+      "learning_rate": 1.968218092728041e-05,
+      "loss": 0.9049,
+      "step": 2076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.8035843946017,
+      "learning_rate": 1.968175900869611e-05,
+      "loss": 0.8038,
+      "step": 2077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.762809406054114,
+      "learning_rate": 1.968133681476825e-05,
+      "loss": 0.9007,
+      "step": 2078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.792393175444175,
+      "learning_rate": 1.9680914345508834e-05,
+      "loss": 0.8621,
+      "step": 2079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.112575483961884,
+      "learning_rate": 1.9680491600929874e-05,
+      "loss": 0.9009,
+      "step": 2080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.51078749449631,
+      "learning_rate": 1.9680068581043396e-05,
+      "loss": 0.8398,
+      "step": 2081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.781832085251615,
+      "learning_rate": 1.9679645285861433e-05,
+      "loss": 0.8318,
+      "step": 2082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.82291140287787,
+      "learning_rate": 1.9679221715396018e-05,
+      "loss": 0.8048,
+      "step": 2083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 32.944377575626525,
+      "learning_rate": 1.9678797869659204e-05,
+      "loss": 0.9733,
+      "step": 2084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.676625373293733,
+      "learning_rate": 1.9678373748663037e-05,
+      "loss": 0.8363,
+      "step": 2085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.685482164690686,
+      "learning_rate": 1.967794935241958e-05,
+      "loss": 0.95,
+      "step": 2086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 21.79509052642431,
+      "learning_rate": 1.967752468094091e-05,
+      "loss": 0.8782,
+      "step": 2087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.101179041048873,
+      "learning_rate": 1.9677099734239098e-05,
+      "loss": 0.9321,
+      "step": 2088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.373758789247308,
+      "learning_rate": 1.967667451232623e-05,
+      "loss": 0.8857,
+      "step": 2089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.729766105341012,
+      "learning_rate": 1.9676249015214396e-05,
+      "loss": 0.7793,
+      "step": 2090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.513667037997,
+      "learning_rate": 1.9675823242915705e-05,
+      "loss": 0.8488,
+      "step": 2091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.170709642632474,
+      "learning_rate": 1.967539719544226e-05,
+      "loss": 0.8632,
+      "step": 2092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.873762529006765,
+      "learning_rate": 1.967497087280618e-05,
+      "loss": 0.8697,
+      "step": 2093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.146375161618987,
+      "learning_rate": 1.9674544275019583e-05,
+      "loss": 0.8427,
+      "step": 2094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.39465103304834,
+      "learning_rate": 1.9674117402094612e-05,
+      "loss": 0.7769,
+      "step": 2095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.994700505673816,
+      "learning_rate": 1.96736902540434e-05,
+      "loss": 0.8662,
+      "step": 2096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.297744865219787,
+      "learning_rate": 1.9673262830878094e-05,
+      "loss": 0.8444,
+      "step": 2097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.50302667844549,
+      "learning_rate": 1.9672835132610856e-05,
+      "loss": 0.8278,
+      "step": 2098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.231258831369892,
+      "learning_rate": 1.967240715925384e-05,
+      "loss": 0.9185,
+      "step": 2099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.586851130579518,
+      "learning_rate": 1.9671978910819225e-05,
+      "loss": 0.8729,
+      "step": 2100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.596507171474503,
+      "learning_rate": 1.9671550387319188e-05,
+      "loss": 0.7587,
+      "step": 2101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 21.54986398351388,
+      "learning_rate": 1.9671121588765913e-05,
+      "loss": 0.8134,
+      "step": 2102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 21.44040746963521,
+      "learning_rate": 1.96706925151716e-05,
+      "loss": 0.9036,
+      "step": 2103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.642517391678766,
+      "learning_rate": 1.967026316654845e-05,
+      "loss": 0.9083,
+      "step": 2104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.546774618435716,
+      "learning_rate": 1.966983354290867e-05,
+      "loss": 0.8881,
+      "step": 2105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.580669462025064,
+      "learning_rate": 1.9669403644264485e-05,
+      "loss": 0.8202,
+      "step": 2106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9494752545919876,
+      "learning_rate": 1.966897347062811e-05,
+      "loss": 0.8271,
+      "step": 2107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 28.918628136324678,
+      "learning_rate": 1.9668543022011792e-05,
+      "loss": 0.8817,
+      "step": 2108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.775240789369356,
+      "learning_rate": 1.9668112298427764e-05,
+      "loss": 0.9015,
+      "step": 2109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.737897706509454,
+      "learning_rate": 1.9667681299888277e-05,
+      "loss": 0.9715,
+      "step": 2110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.843163249344535,
+      "learning_rate": 1.966725002640559e-05,
+      "loss": 0.8514,
+      "step": 2111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 29.238157946085167,
+      "learning_rate": 1.9666818477991967e-05,
+      "loss": 0.8035,
+      "step": 2112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.25716850174795,
+      "learning_rate": 1.9666386654659683e-05,
+      "loss": 0.7557,
+      "step": 2113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.080429387928188,
+      "learning_rate": 1.9665954556421014e-05,
+      "loss": 0.8524,
+      "step": 2114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.46195452223196,
+      "learning_rate": 1.9665522183288253e-05,
+      "loss": 0.9517,
+      "step": 2115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.22997900690526,
+      "learning_rate": 1.9665089535273698e-05,
+      "loss": 0.7616,
+      "step": 2116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.71597199941219,
+      "learning_rate": 1.966465661238965e-05,
+      "loss": 0.7954,
+      "step": 2117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.747367255209674,
+      "learning_rate": 1.966422341464842e-05,
+      "loss": 0.819,
+      "step": 2118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.337880005125236,
+      "learning_rate": 1.966378994206233e-05,
+      "loss": 0.8782,
+      "step": 2119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 21.6161545450506,
+      "learning_rate": 1.9663356194643704e-05,
+      "loss": 0.7597,
+      "step": 2120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.478990774292374,
+      "learning_rate": 1.9662922172404887e-05,
+      "loss": 0.8034,
+      "step": 2121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 21.93094670261835,
+      "learning_rate": 1.966248787535821e-05,
+      "loss": 0.8329,
+      "step": 2122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.203284656659473,
+      "learning_rate": 1.9662053303516036e-05,
+      "loss": 0.9106,
+      "step": 2123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.648103820474752,
+      "learning_rate": 1.9661618456890714e-05,
+      "loss": 0.7647,
+      "step": 2124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.951290720050416,
+      "learning_rate": 1.9661183335494617e-05,
+      "loss": 0.8088,
+      "step": 2125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.791425707541396,
+      "learning_rate": 1.9660747939340116e-05,
+      "loss": 0.8276,
+      "step": 2126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.346590899726824,
+      "learning_rate": 1.9660312268439593e-05,
+      "loss": 0.8467,
+      "step": 2127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 31.015011462333717,
+      "learning_rate": 1.9659876322805448e-05,
+      "loss": 0.852,
+      "step": 2128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.85371623488054,
+      "learning_rate": 1.965944010245006e-05,
+      "loss": 0.7844,
+      "step": 2129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.3166608686285,
+      "learning_rate": 1.9659003607385857e-05,
+      "loss": 0.731,
+      "step": 2130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.230175135283112,
+      "learning_rate": 1.9658566837625236e-05,
+      "loss": 0.8378,
+      "step": 2131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.268960548845886,
+      "learning_rate": 1.9658129793180625e-05,
+      "loss": 0.8499,
+      "step": 2132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 21.393021354586935,
+      "learning_rate": 1.965769247406445e-05,
+      "loss": 0.8268,
+      "step": 2133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 24.690563458490438,
+      "learning_rate": 1.9657254880289155e-05,
+      "loss": 0.8692,
+      "step": 2134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.730910040437305,
+      "learning_rate": 1.965681701186718e-05,
+      "loss": 0.8778,
+      "step": 2135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 34.76191543504597,
+      "learning_rate": 1.9656378868810978e-05,
+      "loss": 0.9182,
+      "step": 2136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.923119795595042,
+      "learning_rate": 1.965594045113301e-05,
+      "loss": 0.8778,
+      "step": 2137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 35.14006837969686,
+      "learning_rate": 1.965550175884574e-05,
+      "loss": 0.9905,
+      "step": 2138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.81176056135414,
+      "learning_rate": 1.9655062791961654e-05,
+      "loss": 0.9221,
+      "step": 2139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.778697089127505,
+      "learning_rate": 1.9654623550493227e-05,
+      "loss": 0.7828,
+      "step": 2140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.075896231015909,
+      "learning_rate": 1.9654184034452955e-05,
+      "loss": 0.7573,
+      "step": 2141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.250945541957204,
+      "learning_rate": 1.9653744243853333e-05,
+      "loss": 0.7862,
+      "step": 2142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.914711142341105,
+      "learning_rate": 1.9653304178706877e-05,
+      "loss": 0.8773,
+      "step": 2143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.355644536765228,
+      "learning_rate": 1.96528638390261e-05,
+      "loss": 0.8611,
+      "step": 2144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.78372677028951,
+      "learning_rate": 1.9652423224823515e-05,
+      "loss": 0.8258,
+      "step": 2145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.400282524238868,
+      "learning_rate": 1.965198233611166e-05,
+      "loss": 0.8081,
+      "step": 2146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.44229291155315,
+      "learning_rate": 1.9651541172903076e-05,
+      "loss": 0.8738,
+      "step": 2147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.966595726243217,
+      "learning_rate": 1.965109973521031e-05,
+      "loss": 0.8329,
+      "step": 2148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.382324163627537,
+      "learning_rate": 1.965065802304591e-05,
+      "loss": 0.7689,
+      "step": 2149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.824299573279532,
+      "learning_rate": 1.965021603642244e-05,
+      "loss": 0.8131,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.696418335757354,
+      "learning_rate": 1.9649773775352475e-05,
+      "loss": 0.7632,
+      "step": 2151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.731280950027507,
+      "learning_rate": 1.9649331239848587e-05,
+      "loss": 0.8744,
+      "step": 2152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.206521976165025,
+      "learning_rate": 1.9648888429923364e-05,
+      "loss": 0.7901,
+      "step": 2153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.19688452062193,
+      "learning_rate": 1.9648445345589395e-05,
+      "loss": 0.8523,
+      "step": 2154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.333005006512693,
+      "learning_rate": 1.9648001986859288e-05,
+      "loss": 0.7507,
+      "step": 2155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.647414526418764,
+      "learning_rate": 1.9647558353745645e-05,
+      "loss": 0.8648,
+      "step": 2156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.029609753829504,
+      "learning_rate": 1.964711444626109e-05,
+      "loss": 0.8985,
+      "step": 2157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 35.82979041602923,
+      "learning_rate": 1.9646670264418242e-05,
+      "loss": 0.8636,
+      "step": 2158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.343582252070068,
+      "learning_rate": 1.9646225808229734e-05,
+      "loss": 0.7577,
+      "step": 2159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.348562087888574,
+      "learning_rate": 1.9645781077708208e-05,
+      "loss": 0.7974,
+      "step": 2160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.810396517951595,
+      "learning_rate": 1.9645336072866312e-05,
+      "loss": 0.8364,
+      "step": 2161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.723141038253587,
+      "learning_rate": 1.9644890793716697e-05,
+      "loss": 0.9236,
+      "step": 2162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.359374407699494,
+      "learning_rate": 1.9644445240272033e-05,
+      "loss": 0.9071,
+      "step": 2163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.453060436232484,
+      "learning_rate": 1.964399941254499e-05,
+      "loss": 0.87,
+      "step": 2164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.467608515584082,
+      "learning_rate": 1.964355331054824e-05,
+      "loss": 0.9243,
+      "step": 2165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.248844364392504,
+      "learning_rate": 1.9643106934294476e-05,
+      "loss": 0.8163,
+      "step": 2166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.122933555051425,
+      "learning_rate": 1.9642660283796395e-05,
+      "loss": 0.952,
+      "step": 2167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.93321172529305,
+      "learning_rate": 1.9642213359066694e-05,
+      "loss": 0.8702,
+      "step": 2168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.994449727132253,
+      "learning_rate": 1.9641766160118086e-05,
+      "loss": 1.0528,
+      "step": 2169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 29.239068988388972,
+      "learning_rate": 1.964131868696329e-05,
+      "loss": 0.8495,
+      "step": 2170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.44738303728948,
+      "learning_rate": 1.964087093961503e-05,
+      "loss": 0.9238,
+      "step": 2171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 26.948207012360225,
+      "learning_rate": 1.9640422918086038e-05,
+      "loss": 0.8579,
+      "step": 2172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.89967049998885,
+      "learning_rate": 1.963997462238906e-05,
+      "loss": 0.7927,
+      "step": 2173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.8201842213537,
+      "learning_rate": 1.963952605253684e-05,
+      "loss": 0.7746,
+      "step": 2174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.545274448273872,
+      "learning_rate": 1.963907720854214e-05,
+      "loss": 0.9211,
+      "step": 2175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.93813130975505,
+      "learning_rate": 1.9638628090417727e-05,
+      "loss": 0.8126,
+      "step": 2176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.84867541245477,
+      "learning_rate": 1.9638178698176364e-05,
+      "loss": 0.8616,
+      "step": 2177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.82145434641559,
+      "learning_rate": 1.963772903183084e-05,
+      "loss": 0.844,
+      "step": 2178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.43618542481379,
+      "learning_rate": 1.963727909139394e-05,
+      "loss": 0.782,
+      "step": 2179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.283206081398543,
+      "learning_rate": 1.9636828876878456e-05,
+      "loss": 0.8548,
+      "step": 2180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.47930205510977,
+      "learning_rate": 1.96363783882972e-05,
+      "loss": 0.7673,
+      "step": 2181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 24.63215972502464,
+      "learning_rate": 1.963592762566298e-05,
+      "loss": 0.8863,
+      "step": 2182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 27.53223359410934,
+      "learning_rate": 1.9635476588988615e-05,
+      "loss": 0.8786,
+      "step": 2183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 29.538060661131315,
+      "learning_rate": 1.963502527828693e-05,
+      "loss": 0.8706,
+      "step": 2184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 57.84225302005477,
+      "learning_rate": 1.9634573693570767e-05,
+      "loss": 0.8658,
+      "step": 2185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.793039716665785,
+      "learning_rate": 1.963412183485296e-05,
+      "loss": 0.7691,
+      "step": 2186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.26578418496224,
+      "learning_rate": 1.9633669702146366e-05,
+      "loss": 0.8299,
+      "step": 2187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.478741276842545,
+      "learning_rate": 1.963321729546384e-05,
+      "loss": 0.8549,
+      "step": 2188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.167273619461376,
+      "learning_rate": 1.963276461481825e-05,
+      "loss": 0.9106,
+      "step": 2189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 56.31049650419636,
+      "learning_rate": 1.963231166022247e-05,
+      "loss": 0.861,
+      "step": 2190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.201976639815516,
+      "learning_rate": 1.9631858431689382e-05,
+      "loss": 0.8063,
+      "step": 2191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.645263961030828,
+      "learning_rate": 1.963140492923187e-05,
+      "loss": 0.8363,
+      "step": 2192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.49710724500291,
+      "learning_rate": 1.963095115286284e-05,
+      "loss": 0.8837,
+      "step": 2193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.65235460289088,
+      "learning_rate": 1.963049710259519e-05,
+      "loss": 0.9756,
+      "step": 2194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 29.373686438471058,
+      "learning_rate": 1.963004277844184e-05,
+      "loss": 0.8135,
+      "step": 2195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.171045728122307,
+      "learning_rate": 1.9629588180415707e-05,
+      "loss": 0.8234,
+      "step": 2196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.958338018415482,
+      "learning_rate": 1.9629133308529717e-05,
+      "loss": 0.9029,
+      "step": 2197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.054987007780277,
+      "learning_rate": 1.962867816279681e-05,
+      "loss": 0.8789,
+      "step": 2198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.724933992763166,
+      "learning_rate": 1.9628222743229927e-05,
+      "loss": 0.8741,
+      "step": 2199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.22813346884525,
+      "learning_rate": 1.9627767049842022e-05,
+      "loss": 0.9541,
+      "step": 2200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.565110368362504,
+      "learning_rate": 1.9627311082646053e-05,
+      "loss": 0.8735,
+      "step": 2201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.17061207401581,
+      "learning_rate": 1.962685484165499e-05,
+      "loss": 0.7636,
+      "step": 2202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.33147770307174,
+      "learning_rate": 1.9626398326881805e-05,
+      "loss": 0.8396,
+      "step": 2203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.692971774283702,
+      "learning_rate": 1.9625941538339484e-05,
+      "loss": 0.8401,
+      "step": 2204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.696540156952329,
+      "learning_rate": 1.962548447604102e-05,
+      "loss": 0.8128,
+      "step": 2205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.082897626280321,
+      "learning_rate": 1.9625027139999404e-05,
+      "loss": 0.7894,
+      "step": 2206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.58972804448063,
+      "learning_rate": 1.962456953022765e-05,
+      "loss": 0.7516,
+      "step": 2207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 24.880836772182267,
+      "learning_rate": 1.9624111646738767e-05,
+      "loss": 0.9179,
+      "step": 2208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.28436038367192,
+      "learning_rate": 1.962365348954578e-05,
+      "loss": 0.8,
+      "step": 2209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.63477719802138,
+      "learning_rate": 1.9623195058661716e-05,
+      "loss": 0.8046,
+      "step": 2210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.092068807439325,
+      "learning_rate": 1.9622736354099612e-05,
+      "loss": 0.7285,
+      "step": 2211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 25.19775477201366,
+      "learning_rate": 1.9622277375872518e-05,
+      "loss": 0.8251,
+      "step": 2212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.554167467807773,
+      "learning_rate": 1.9621818123993482e-05,
+      "loss": 0.9185,
+      "step": 2213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 29.398751526067016,
+      "learning_rate": 1.962135859847557e-05,
+      "loss": 0.9394,
+      "step": 2214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.06735841237785,
+      "learning_rate": 1.9620898799331848e-05,
+      "loss": 0.9402,
+      "step": 2215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 30.64704539316894,
+      "learning_rate": 1.962043872657539e-05,
+      "loss": 0.989,
+      "step": 2216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.050664734708178,
+      "learning_rate": 1.9619978380219282e-05,
+      "loss": 0.8904,
+      "step": 2217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.489511197724184,
+      "learning_rate": 1.9619517760276618e-05,
+      "loss": 0.889,
+      "step": 2218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.021809114490775,
+      "learning_rate": 1.9619056866760497e-05,
+      "loss": 0.8951,
+      "step": 2219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.09823309624137,
+      "learning_rate": 1.9618595699684026e-05,
+      "loss": 0.8278,
+      "step": 2220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.472222781964913,
+      "learning_rate": 1.9618134259060318e-05,
+      "loss": 0.8743,
+      "step": 2221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.254923119860816,
+      "learning_rate": 1.96176725449025e-05,
+      "loss": 0.9064,
+      "step": 2222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.05791563768528,
+      "learning_rate": 1.9617210557223697e-05,
+      "loss": 0.8709,
+      "step": 2223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.543076767281683,
+      "learning_rate": 1.9616748296037052e-05,
+      "loss": 0.8798,
+      "step": 2224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.666738148373785,
+      "learning_rate": 1.9616285761355717e-05,
+      "loss": 0.895,
+      "step": 2225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.746350567830284,
+      "learning_rate": 1.9615822953192833e-05,
+      "loss": 0.8131,
+      "step": 2226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.436412622403108,
+      "learning_rate": 1.9615359871561574e-05,
+      "loss": 0.7679,
+      "step": 2227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.810434706482013,
+      "learning_rate": 1.9614896516475107e-05,
+      "loss": 0.79,
+      "step": 2228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 35.932532497907,
+      "learning_rate": 1.96144328879466e-05,
+      "loss": 0.8941,
+      "step": 2229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.831174199343533,
+      "learning_rate": 1.9613968985989253e-05,
+      "loss": 1.0721,
+      "step": 2230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 25.270256096950266,
+      "learning_rate": 1.961350481061625e-05,
+      "loss": 0.9382,
+      "step": 2231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.38343000500924,
+      "learning_rate": 1.9613040361840792e-05,
+      "loss": 0.9236,
+      "step": 2232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.443885365676206,
+      "learning_rate": 1.961257563967609e-05,
+      "loss": 0.759,
+      "step": 2233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.715840613201532,
+      "learning_rate": 1.9612110644135365e-05,
+      "loss": 0.8116,
+      "step": 2234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.429731530895157,
+      "learning_rate": 1.961164537523183e-05,
+      "loss": 0.9412,
+      "step": 2235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.716173067865874,
+      "learning_rate": 1.9611179832978725e-05,
+      "loss": 0.8309,
+      "step": 2236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 28.72478583669382,
+      "learning_rate": 1.961071401738929e-05,
+      "loss": 0.9226,
+      "step": 2237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.306935361831256,
+      "learning_rate": 1.961024792847677e-05,
+      "loss": 0.8484,
+      "step": 2238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.03491843244981,
+      "learning_rate": 1.960978156625442e-05,
+      "loss": 0.8104,
+      "step": 2239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.021771311788267,
+      "learning_rate": 1.96093149307355e-05,
+      "loss": 0.8226,
+      "step": 2240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.19508629061962,
+      "learning_rate": 1.960884802193329e-05,
+      "loss": 0.8646,
+      "step": 2241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.193580278916478,
+      "learning_rate": 1.960838083986106e-05,
+      "loss": 0.8639,
+      "step": 2242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.843368560416156,
+      "learning_rate": 1.9607913384532102e-05,
+      "loss": 0.886,
+      "step": 2243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.13145673744327,
+      "learning_rate": 1.9607445655959707e-05,
+      "loss": 0.8167,
+      "step": 2244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.90926852762867,
+      "learning_rate": 1.9606977654157176e-05,
+      "loss": 0.763,
+      "step": 2245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.687757201124638,
+      "learning_rate": 1.9606509379137823e-05,
+      "loss": 0.7763,
+      "step": 2246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.68120718669408,
+      "learning_rate": 1.960604083091496e-05,
+      "loss": 0.877,
+      "step": 2247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.947481220770838,
+      "learning_rate": 1.9605572009501918e-05,
+      "loss": 0.7808,
+      "step": 2248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.670105893666836,
+      "learning_rate": 1.9605102914912025e-05,
+      "loss": 0.8162,
+      "step": 2249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.617112052826652,
+      "learning_rate": 1.960463354715863e-05,
+      "loss": 0.8944,
+      "step": 2250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 28.60815847009406,
+      "learning_rate": 1.9604163906255068e-05,
+      "loss": 0.8596,
+      "step": 2251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.786423520382826,
+      "learning_rate": 1.9603693992214707e-05,
+      "loss": 0.8678,
+      "step": 2252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.845369301044503,
+      "learning_rate": 1.9603223805050903e-05,
+      "loss": 0.8616,
+      "step": 2253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.754836853200725,
+      "learning_rate": 1.9602753344777036e-05,
+      "loss": 0.8292,
+      "step": 2254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 26.716550201073716,
+      "learning_rate": 1.960228261140648e-05,
+      "loss": 0.9063,
+      "step": 2255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.451365058301914,
+      "learning_rate": 1.960181160495262e-05,
+      "loss": 0.8574,
+      "step": 2256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.877343921425027,
+      "learning_rate": 1.960134032542886e-05,
+      "loss": 0.8011,
+      "step": 2257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.849301502671569,
+      "learning_rate": 1.9600868772848595e-05,
+      "loss": 0.8001,
+      "step": 2258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.364172779712053,
+      "learning_rate": 1.9600396947225238e-05,
+      "loss": 0.8061,
+      "step": 2259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.25579954351772,
+      "learning_rate": 1.959992484857221e-05,
+      "loss": 0.8852,
+      "step": 2260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.53819251681691,
+      "learning_rate": 1.9599452476902934e-05,
+      "loss": 0.7397,
+      "step": 2261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.778921804544083,
+      "learning_rate": 1.9598979832230846e-05,
+      "loss": 0.8344,
+      "step": 2262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.35354442268991,
+      "learning_rate": 1.9598506914569383e-05,
+      "loss": 0.789,
+      "step": 2263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.017006439206416,
+      "learning_rate": 1.9598033723932004e-05,
+      "loss": 0.8317,
+      "step": 2264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 28.88808198024425,
+      "learning_rate": 1.959756026033216e-05,
+      "loss": 0.8481,
+      "step": 2265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.361581158227146,
+      "learning_rate": 1.9597086523783312e-05,
+      "loss": 0.9174,
+      "step": 2266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.32743899042107,
+      "learning_rate": 1.959661251429894e-05,
+      "loss": 0.9045,
+      "step": 2267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 43.97896255880699,
+      "learning_rate": 1.959613823189252e-05,
+      "loss": 1.0044,
+      "step": 2268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.26575233791744,
+      "learning_rate": 1.9595663676577543e-05,
+      "loss": 0.8442,
+      "step": 2269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.3437053194103,
+      "learning_rate": 1.9595188848367507e-05,
+      "loss": 0.7827,
+      "step": 2270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.962052756168983,
+      "learning_rate": 1.959471374727591e-05,
+      "loss": 0.7672,
+      "step": 2271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 29.265136007937627,
+      "learning_rate": 1.959423837331627e-05,
+      "loss": 0.9122,
+      "step": 2272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 26.90766591102539,
+      "learning_rate": 1.95937627265021e-05,
+      "loss": 0.9845,
+      "step": 2273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.351767860703873,
+      "learning_rate": 1.9593286806846935e-05,
+      "loss": 0.6728,
+      "step": 2274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 28.274015619945366,
+      "learning_rate": 1.95928106143643e-05,
+      "loss": 0.9657,
+      "step": 2275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.033006236249136,
+      "learning_rate": 1.9592334149067745e-05,
+      "loss": 0.7861,
+      "step": 2276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 27.176641004022738,
+      "learning_rate": 1.9591857410970818e-05,
+      "loss": 0.8808,
+      "step": 2277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.834020162842634,
+      "learning_rate": 1.9591380400087077e-05,
+      "loss": 0.9017,
+      "step": 2278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.08037626915052,
+      "learning_rate": 1.9590903116430087e-05,
+      "loss": 0.8966,
+      "step": 2279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.240594464788924,
+      "learning_rate": 1.9590425560013427e-05,
+      "loss": 0.8006,
+      "step": 2280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.109508313127826,
+      "learning_rate": 1.9589947730850668e-05,
+      "loss": 0.7795,
+      "step": 2281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.75527198715787,
+      "learning_rate": 1.958946962895541e-05,
+      "loss": 0.7473,
+      "step": 2282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.926261129035797,
+      "learning_rate": 1.9588991254341243e-05,
+      "loss": 0.8775,
+      "step": 2283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.210590480261835,
+      "learning_rate": 1.9588512607021777e-05,
+      "loss": 0.7519,
+      "step": 2284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.38915544885124,
+      "learning_rate": 1.958803368701062e-05,
+      "loss": 0.9165,
+      "step": 2285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.4473599510166,
+      "learning_rate": 1.9587554494321394e-05,
+      "loss": 0.7747,
+      "step": 2286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.650569556725742,
+      "learning_rate": 1.958707502896773e-05,
+      "loss": 0.8813,
+      "step": 2287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.36865528908284,
+      "learning_rate": 1.9586595290963254e-05,
+      "loss": 0.8279,
+      "step": 2288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.36321122746258,
+      "learning_rate": 1.958611528032162e-05,
+      "loss": 0.7433,
+      "step": 2289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.60395808911962,
+      "learning_rate": 1.9585634997056472e-05,
+      "loss": 0.9214,
+      "step": 2290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.629458455531346,
+      "learning_rate": 1.9585154441181475e-05,
+      "loss": 0.7866,
+      "step": 2291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.64594494261864,
+      "learning_rate": 1.9584673612710293e-05,
+      "loss": 0.8852,
+      "step": 2292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.88986593059227,
+      "learning_rate": 1.95841925116566e-05,
+      "loss": 0.811,
+      "step": 2293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.920206414764746,
+      "learning_rate": 1.9583711138034075e-05,
+      "loss": 0.9009,
+      "step": 2294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 36.281520411250064,
+      "learning_rate": 1.958322949185642e-05,
+      "loss": 0.8237,
+      "step": 2295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.502897981405205,
+      "learning_rate": 1.9582747573137315e-05,
+      "loss": 0.9108,
+      "step": 2296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.900124518863047,
+      "learning_rate": 1.958226538189048e-05,
+      "loss": 0.8748,
+      "step": 2297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.353509063519404,
+      "learning_rate": 1.958178291812962e-05,
+      "loss": 0.9391,
+      "step": 2298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.069704368787633,
+      "learning_rate": 1.9581300181868462e-05,
+      "loss": 0.7822,
+      "step": 2299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.563794908993,
+      "learning_rate": 1.958081717312073e-05,
+      "loss": 0.9021,
+      "step": 2300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.7725747077412,
+      "learning_rate": 1.9580333891900164e-05,
+      "loss": 0.8624,
+      "step": 2301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.52007982411239,
+      "learning_rate": 1.9579850338220502e-05,
+      "loss": 0.8048,
+      "step": 2302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 31.01456514921061,
+      "learning_rate": 1.9579366512095504e-05,
+      "loss": 0.8553,
+      "step": 2303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.39003111946264,
+      "learning_rate": 1.9578882413538928e-05,
+      "loss": 0.8755,
+      "step": 2304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.87317373622345,
+      "learning_rate": 1.9578398042564534e-05,
+      "loss": 0.8392,
+      "step": 2305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 28.639158294850937,
+      "learning_rate": 1.9577913399186106e-05,
+      "loss": 0.8533,
+      "step": 2306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.981001838335168,
+      "learning_rate": 1.9577428483417424e-05,
+      "loss": 0.8814,
+      "step": 2307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.767328364359244,
+      "learning_rate": 1.957694329527228e-05,
+      "loss": 0.9141,
+      "step": 2308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.60064072758327,
+      "learning_rate": 1.9576457834764468e-05,
+      "loss": 0.8803,
+      "step": 2309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.785497650368177,
+      "learning_rate": 1.9575972101907798e-05,
+      "loss": 0.8541,
+      "step": 2310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.31662293800399,
+      "learning_rate": 1.9575486096716086e-05,
+      "loss": 0.8953,
+      "step": 2311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.52787343403029,
+      "learning_rate": 1.957499981920315e-05,
+      "loss": 0.7918,
+      "step": 2312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.865508904039164,
+      "learning_rate": 1.957451326938282e-05,
+      "loss": 0.8077,
+      "step": 2313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 25.315322527875505,
+      "learning_rate": 1.957402644726893e-05,
+      "loss": 0.8776,
+      "step": 2314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.644909527203188,
+      "learning_rate": 1.9573539352875333e-05,
+      "loss": 0.9013,
+      "step": 2315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.987844389972054,
+      "learning_rate": 1.9573051986215873e-05,
+      "loss": 0.7833,
+      "step": 2316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.81496192809406,
+      "learning_rate": 1.9572564347304418e-05,
+      "loss": 0.7088,
+      "step": 2317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.478514597867123,
+      "learning_rate": 1.957207643615483e-05,
+      "loss": 0.8507,
+      "step": 2318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.714268848564387,
+      "learning_rate": 1.9571588252780992e-05,
+      "loss": 0.8358,
+      "step": 2319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.02872563318518,
+      "learning_rate": 1.9571099797196778e-05,
+      "loss": 0.7872,
+      "step": 2320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.76587062007009,
+      "learning_rate": 1.957061106941609e-05,
+      "loss": 0.8648,
+      "step": 2321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.02049166892051,
+      "learning_rate": 1.9570122069452817e-05,
+      "loss": 0.8226,
+      "step": 2322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.835083095121952,
+      "learning_rate": 1.9569632797320876e-05,
+      "loss": 0.7784,
+      "step": 2323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.276543367213247,
+      "learning_rate": 1.9569143253034172e-05,
+      "loss": 0.8056,
+      "step": 2324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.884539515346667,
+      "learning_rate": 1.9568653436606632e-05,
+      "loss": 0.9212,
+      "step": 2325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.54858209589304,
+      "learning_rate": 1.9568163348052186e-05,
+      "loss": 0.7636,
+      "step": 2326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.923779145843948,
+      "learning_rate": 1.9567672987384774e-05,
+      "loss": 0.9018,
+      "step": 2327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.46457753946557,
+      "learning_rate": 1.9567182354618338e-05,
+      "loss": 0.9344,
+      "step": 2328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 29.285314327737893,
+      "learning_rate": 1.956669144976683e-05,
+      "loss": 0.8852,
+      "step": 2329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.760738511625377,
+      "learning_rate": 1.9566200272844216e-05,
+      "loss": 0.716,
+      "step": 2330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.21331647187897,
+      "learning_rate": 1.956570882386446e-05,
+      "loss": 0.8791,
+      "step": 2331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.41563576630927,
+      "learning_rate": 1.9565217102841543e-05,
+      "loss": 0.8382,
+      "step": 2332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.750899110356364,
+      "learning_rate": 1.9564725109789447e-05,
+      "loss": 0.7417,
+      "step": 2333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.760458215179696,
+      "learning_rate": 1.9564232844722166e-05,
+      "loss": 0.7939,
+      "step": 2334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.114660816668017,
+      "learning_rate": 1.9563740307653693e-05,
+      "loss": 0.8228,
+      "step": 2335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.373569290927932,
+      "learning_rate": 1.9563247498598045e-05,
+      "loss": 0.8,
+      "step": 2336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.201653685477662,
+      "learning_rate": 1.956275441756923e-05,
+      "loss": 0.8496,
+      "step": 2337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.45357286711351,
+      "learning_rate": 1.9562261064581272e-05,
+      "loss": 0.8678,
+      "step": 2338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.698160175413193,
+      "learning_rate": 1.956176743964821e-05,
+      "loss": 0.9822,
+      "step": 2339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.29147152809738,
+      "learning_rate": 1.9561273542784067e-05,
+      "loss": 0.8554,
+      "step": 2340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.468092203891214,
+      "learning_rate": 1.9560779374002902e-05,
+      "loss": 0.9422,
+      "step": 2341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.90812846481586,
+      "learning_rate": 1.9560284933318763e-05,
+      "loss": 0.9153,
+      "step": 2342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.881066794651561,
+      "learning_rate": 1.9559790220745715e-05,
+      "loss": 0.7362,
+      "step": 2343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.834799798238674,
+      "learning_rate": 1.9559295236297822e-05,
+      "loss": 0.8588,
+      "step": 2344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.44682119945334,
+      "learning_rate": 1.9558799979989165e-05,
+      "loss": 0.8622,
+      "step": 2345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.7818948774698,
+      "learning_rate": 1.9558304451833828e-05,
+      "loss": 0.9244,
+      "step": 2346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 32.15240395829633,
+      "learning_rate": 1.955780865184591e-05,
+      "loss": 0.8391,
+      "step": 2347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.52928040606101,
+      "learning_rate": 1.95573125800395e-05,
+      "loss": 0.8631,
+      "step": 2348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.014948379107024,
+      "learning_rate": 1.955681623642871e-05,
+      "loss": 0.7669,
+      "step": 2349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.3606479985429365,
+      "learning_rate": 1.955631962102766e-05,
+      "loss": 0.9177,
+      "step": 2350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.70675317168515,
+      "learning_rate": 1.955582273385047e-05,
+      "loss": 0.7743,
+      "step": 2351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.959628858998226,
+      "learning_rate": 1.955532557491127e-05,
+      "loss": 0.7911,
+      "step": 2352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.870408728266078,
+      "learning_rate": 1.95548281442242e-05,
+      "loss": 0.7857,
+      "step": 2353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.12068729290322,
+      "learning_rate": 1.9554330441803406e-05,
+      "loss": 0.8328,
+      "step": 2354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.39010542509534,
+      "learning_rate": 1.9553832467663048e-05,
+      "loss": 0.8622,
+      "step": 2355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.484470397819127,
+      "learning_rate": 1.955333422181728e-05,
+      "loss": 0.9214,
+      "step": 2356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.757632950277799,
+      "learning_rate": 1.955283570428027e-05,
+      "loss": 0.8439,
+      "step": 2357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.790934236685473,
+      "learning_rate": 1.955233691506621e-05,
+      "loss": 0.9286,
+      "step": 2358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.639654567005557,
+      "learning_rate": 1.9551837854189274e-05,
+      "loss": 0.8256,
+      "step": 2359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.339632657461078,
+      "learning_rate": 1.9551338521663656e-05,
+      "loss": 0.8338,
+      "step": 2360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.92224329452618,
+      "learning_rate": 1.9550838917503557e-05,
+      "loss": 0.937,
+      "step": 2361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 24.22799428542671,
+      "learning_rate": 1.9550339041723188e-05,
+      "loss": 1.0115,
+      "step": 2362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.104370162075357,
+      "learning_rate": 1.954983889433676e-05,
+      "loss": 0.7641,
+      "step": 2363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 26.204035727222347,
+      "learning_rate": 1.9549338475358505e-05,
+      "loss": 0.9896,
+      "step": 2364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.977562247710003,
+      "learning_rate": 1.9548837784802646e-05,
+      "loss": 0.7483,
+      "step": 2365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.580521959104235,
+      "learning_rate": 1.9548336822683426e-05,
+      "loss": 0.8389,
+      "step": 2366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.841327111116088,
+      "learning_rate": 1.9547835589015095e-05,
+      "loss": 0.7908,
+      "step": 2367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.442799154015507,
+      "learning_rate": 1.9547334083811906e-05,
+      "loss": 0.9086,
+      "step": 2368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 23.129577828623827,
+      "learning_rate": 1.954683230708812e-05,
+      "loss": 0.8035,
+      "step": 2369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.687118091034925,
+      "learning_rate": 1.9546330258858007e-05,
+      "loss": 0.9282,
+      "step": 2370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.267640544320162,
+      "learning_rate": 1.9545827939135846e-05,
+      "loss": 0.8739,
+      "step": 2371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.876459420089844,
+      "learning_rate": 1.9545325347935923e-05,
+      "loss": 0.8541,
+      "step": 2372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.80787492667919,
+      "learning_rate": 1.9544822485272528e-05,
+      "loss": 0.8464,
+      "step": 2373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.987684051973616,
+      "learning_rate": 1.954431935115997e-05,
+      "loss": 0.876,
+      "step": 2374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.834476586173817,
+      "learning_rate": 1.9543815945612548e-05,
+      "loss": 0.8347,
+      "step": 2375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.49566676745921,
+      "learning_rate": 1.9543312268644586e-05,
+      "loss": 0.9176,
+      "step": 2376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.844859224781588,
+      "learning_rate": 1.9542808320270405e-05,
+      "loss": 0.7221,
+      "step": 2377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.313890675370274,
+      "learning_rate": 1.954230410050434e-05,
+      "loss": 0.8976,
+      "step": 2378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.046858249402987,
+      "learning_rate": 1.9541799609360727e-05,
+      "loss": 0.8804,
+      "step": 2379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.667118069864024,
+      "learning_rate": 1.9541294846853914e-05,
+      "loss": 0.7766,
+      "step": 2380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.285024985122114,
+      "learning_rate": 1.954078981299826e-05,
+      "loss": 0.7479,
+      "step": 2381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.918423588302655,
+      "learning_rate": 1.954028450780812e-05,
+      "loss": 0.8779,
+      "step": 2382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.615739365173884,
+      "learning_rate": 1.9539778931297874e-05,
+      "loss": 0.7884,
+      "step": 2383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.350674587510465,
+      "learning_rate": 1.9539273083481894e-05,
+      "loss": 0.8458,
+      "step": 2384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.364339834697589,
+      "learning_rate": 1.953876696437457e-05,
+      "loss": 0.7645,
+      "step": 2385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.216220994129927,
+      "learning_rate": 1.953826057399029e-05,
+      "loss": 0.833,
+      "step": 2386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.280495522490607,
+      "learning_rate": 1.953775391234346e-05,
+      "loss": 0.8504,
+      "step": 2387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.523350656292592,
+      "learning_rate": 1.953724697944849e-05,
+      "loss": 0.83,
+      "step": 2388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.70498830699112,
+      "learning_rate": 1.9536739775319793e-05,
+      "loss": 0.7669,
+      "step": 2389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.407102977740667,
+      "learning_rate": 1.9536232299971797e-05,
+      "loss": 0.7656,
+      "step": 2390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.86788595577168,
+      "learning_rate": 1.953572455341893e-05,
+      "loss": 0.9123,
+      "step": 2391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 23.757444997301512,
+      "learning_rate": 1.9535216535675636e-05,
+      "loss": 0.8602,
+      "step": 2392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.303482902670957,
+      "learning_rate": 1.9534708246756364e-05,
+      "loss": 0.8665,
+      "step": 2393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 26.802081116741597,
+      "learning_rate": 1.9534199686675567e-05,
+      "loss": 0.8638,
+      "step": 2394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.25971291935905,
+      "learning_rate": 1.9533690855447706e-05,
+      "loss": 0.7855,
+      "step": 2395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 39.078096568216615,
+      "learning_rate": 1.9533181753087255e-05,
+      "loss": 0.8725,
+      "step": 2396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.16279272140576,
+      "learning_rate": 1.953267237960869e-05,
+      "loss": 0.9653,
+      "step": 2397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 23.145180519751115,
+      "learning_rate": 1.9532162735026503e-05,
+      "loss": 0.8258,
+      "step": 2398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.526106660730825,
+      "learning_rate": 1.9531652819355184e-05,
+      "loss": 0.795,
+      "step": 2399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.05502578024802,
+      "learning_rate": 1.953114263260923e-05,
+      "loss": 0.7586,
+      "step": 2400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.07608855269181,
+      "learning_rate": 1.9530632174803158e-05,
+      "loss": 0.8031,
+      "step": 2401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.335398720951122,
+      "learning_rate": 1.953012144595148e-05,
+      "loss": 0.8231,
+      "step": 2402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.26518463444549,
+      "learning_rate": 1.9529610446068727e-05,
+      "loss": 0.7897,
+      "step": 2403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.46569208530518,
+      "learning_rate": 1.9529099175169425e-05,
+      "loss": 0.8447,
+      "step": 2404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.051389937276333,
+      "learning_rate": 1.952858763326812e-05,
+      "loss": 0.7447,
+      "step": 2405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.646038065046703,
+      "learning_rate": 1.9528075820379353e-05,
+      "loss": 0.732,
+      "step": 2406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.79895137956602,
+      "learning_rate": 1.9527563736517683e-05,
+      "loss": 0.788,
+      "step": 2407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.013451145876225,
+      "learning_rate": 1.952705138169768e-05,
+      "loss": 0.8152,
+      "step": 2408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.993917869946696,
+      "learning_rate": 1.95265387559339e-05,
+      "loss": 0.7993,
+      "step": 2409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 27.754758625511705,
+      "learning_rate": 1.9526025859240936e-05,
+      "loss": 0.8487,
+      "step": 2410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.465520094814222,
+      "learning_rate": 1.952551269163337e-05,
+      "loss": 0.9287,
+      "step": 2411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.208621756966533,
+      "learning_rate": 1.9524999253125792e-05,
+      "loss": 0.7842,
+      "step": 2412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.480105149736534,
+      "learning_rate": 1.952448554373281e-05,
+      "loss": 0.9469,
+      "step": 2413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.46800276512284,
+      "learning_rate": 1.9523971563469026e-05,
+      "loss": 0.9319,
+      "step": 2414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 28.447086995510453,
+      "learning_rate": 1.9523457312349067e-05,
+      "loss": 0.832,
+      "step": 2415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 30.189566063717486,
+      "learning_rate": 1.9522942790387552e-05,
+      "loss": 0.9581,
+      "step": 2416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.866945686719145,
+      "learning_rate": 1.9522427997599113e-05,
+      "loss": 0.7988,
+      "step": 2417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.18456878190828,
+      "learning_rate": 1.9521912933998392e-05,
+      "loss": 0.9309,
+      "step": 2418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.084679660748574,
+      "learning_rate": 1.9521397599600042e-05,
+      "loss": 0.788,
+      "step": 2419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.8868244206621,
+      "learning_rate": 1.952088199441871e-05,
+      "loss": 0.8429,
+      "step": 2420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.5815178602056,
+      "learning_rate": 1.9520366118469064e-05,
+      "loss": 0.9069,
+      "step": 2421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.245618282799214,
+      "learning_rate": 1.9519849971765775e-05,
+      "loss": 0.8289,
+      "step": 2422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.110846508041238,
+      "learning_rate": 1.951933355432352e-05,
+      "loss": 0.9091,
+      "step": 2423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.453560795753642,
+      "learning_rate": 1.9518816866156994e-05,
+      "loss": 0.844,
+      "step": 2424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.633751101179815,
+      "learning_rate": 1.951829990728088e-05,
+      "loss": 0.8443,
+      "step": 2425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.87846786728831,
+      "learning_rate": 1.951778267770988e-05,
+      "loss": 0.8341,
+      "step": 2426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.67956006674215,
+      "learning_rate": 1.9517265177458716e-05,
+      "loss": 0.8208,
+      "step": 2427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 27.744037822142193,
+      "learning_rate": 1.9516747406542097e-05,
+      "loss": 0.789,
+      "step": 2428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.707457612097446,
+      "learning_rate": 1.9516229364974748e-05,
+      "loss": 0.8049,
+      "step": 2429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.061050355247907,
+      "learning_rate": 1.95157110527714e-05,
+      "loss": 0.9428,
+      "step": 2430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.28179708653034,
+      "learning_rate": 1.95151924699468e-05,
+      "loss": 0.8558,
+      "step": 2431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.939199239990796,
+      "learning_rate": 1.9514673616515686e-05,
+      "loss": 0.8189,
+      "step": 2432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.907680305183728,
+      "learning_rate": 1.9514154492492827e-05,
+      "loss": 0.7942,
+      "step": 2433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.527868800517016,
+      "learning_rate": 1.9513635097892975e-05,
+      "loss": 0.9862,
+      "step": 2434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 26.291971377888153,
+      "learning_rate": 1.951311543273091e-05,
+      "loss": 0.859,
+      "step": 2435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.189825695882313,
+      "learning_rate": 1.9512595497021406e-05,
+      "loss": 0.8974,
+      "step": 2436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.843778903543672,
+      "learning_rate": 1.951207529077925e-05,
+      "loss": 0.8477,
+      "step": 2437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 27.592198517817323,
+      "learning_rate": 1.9511554814019237e-05,
+      "loss": 0.8692,
+      "step": 2438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 29.270361096022043,
+      "learning_rate": 1.951103406675617e-05,
+      "loss": 0.8998,
+      "step": 2439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 27.333147997392555,
+      "learning_rate": 1.951051304900486e-05,
+      "loss": 0.8525,
+      "step": 2440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.34717099325237,
+      "learning_rate": 1.9509991760780117e-05,
+      "loss": 0.8892,
+      "step": 2441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.99948084662135,
+      "learning_rate": 1.9509470202096774e-05,
+      "loss": 0.8871,
+      "step": 2442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.181937820280993,
+      "learning_rate": 1.9508948372969664e-05,
+      "loss": 0.9362,
+      "step": 2443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.371574580522392,
+      "learning_rate": 1.950842627341362e-05,
+      "loss": 0.8514,
+      "step": 2444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.015313346849275,
+      "learning_rate": 1.95079039034435e-05,
+      "loss": 0.8166,
+      "step": 2445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.168714169644574,
+      "learning_rate": 1.9507381263074153e-05,
+      "loss": 0.7802,
+      "step": 2446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.710394869090457,
+      "learning_rate": 1.9506858352320444e-05,
+      "loss": 0.8928,
+      "step": 2447
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.27983543026391,
+      "learning_rate": 1.9506335171197243e-05,
+      "loss": 0.8349,
+      "step": 2448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.263049157155898,
+      "learning_rate": 1.9505811719719435e-05,
+      "loss": 0.8699,
+      "step": 2449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.720897475693256,
+      "learning_rate": 1.95052879979019e-05,
+      "loss": 0.8266,
+      "step": 2450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.566849225297409,
+      "learning_rate": 1.9504764005759534e-05,
+      "loss": 0.8193,
+      "step": 2451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.10746439220069,
+      "learning_rate": 1.950423974330724e-05,
+      "loss": 0.7747,
+      "step": 2452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 26.526712596026236,
+      "learning_rate": 1.9503715210559928e-05,
+      "loss": 0.8887,
+      "step": 2453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.950628753504276,
+      "learning_rate": 1.9503190407532514e-05,
+      "loss": 0.8675,
+      "step": 2454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.83666826684625,
+      "learning_rate": 1.9502665334239926e-05,
+      "loss": 0.7574,
+      "step": 2455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.02049736775882,
+      "learning_rate": 1.9502139990697095e-05,
+      "loss": 0.8922,
+      "step": 2456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.486613397471082,
+      "learning_rate": 1.9501614376918963e-05,
+      "loss": 0.8107,
+      "step": 2457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.76034383741656,
+      "learning_rate": 1.9501088492920472e-05,
+      "loss": 0.7128,
+      "step": 2458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.59133177161005,
+      "learning_rate": 1.9500562338716584e-05,
+      "loss": 0.892,
+      "step": 2459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.671197644036523,
+      "learning_rate": 1.9500035914322263e-05,
+      "loss": 0.7916,
+      "step": 2460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.787109359369694,
+      "learning_rate": 1.9499509219752473e-05,
+      "loss": 0.8417,
+      "step": 2461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.881299648109724,
+      "learning_rate": 1.94989822550222e-05,
+      "loss": 0.8222,
+      "step": 2462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 31.415744304819388,
+      "learning_rate": 1.9498455020146428e-05,
+      "loss": 0.8926,
+      "step": 2463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 29.347522763218166,
+      "learning_rate": 1.9497927515140155e-05,
+      "loss": 0.9163,
+      "step": 2464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.668232378260434,
+      "learning_rate": 1.9497399740018376e-05,
+      "loss": 0.7106,
+      "step": 2465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.630541805620698,
+      "learning_rate": 1.9496871694796106e-05,
+      "loss": 0.7943,
+      "step": 2466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.891356055337113,
+      "learning_rate": 1.949634337948836e-05,
+      "loss": 0.8403,
+      "step": 2467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.755547365964734,
+      "learning_rate": 1.9495814794110165e-05,
+      "loss": 0.6969,
+      "step": 2468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.946137380188254,
+      "learning_rate": 1.949528593867655e-05,
+      "loss": 0.8803,
+      "step": 2469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.55641275230682,
+      "learning_rate": 1.9494756813202558e-05,
+      "loss": 0.8794,
+      "step": 2470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.42324417424368,
+      "learning_rate": 1.9494227417703237e-05,
+      "loss": 0.8305,
+      "step": 2471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 38.59207416980971,
+      "learning_rate": 1.949369775219364e-05,
+      "loss": 0.9413,
+      "step": 2472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.52672246064103,
+      "learning_rate": 1.9493167816688837e-05,
+      "loss": 0.8335,
+      "step": 2473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 37.97518438284426,
+      "learning_rate": 1.949263761120389e-05,
+      "loss": 0.9332,
+      "step": 2474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.705251099045753,
+      "learning_rate": 1.9492107135753884e-05,
+      "loss": 0.7934,
+      "step": 2475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.23079161444644,
+      "learning_rate": 1.9491576390353906e-05,
+      "loss": 0.9349,
+      "step": 2476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.89753302696428,
+      "learning_rate": 1.9491045375019043e-05,
+      "loss": 0.8761,
+      "step": 2477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.637885229490248,
+      "learning_rate": 1.949051408976441e-05,
+      "loss": 0.7511,
+      "step": 2478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.165437200175504,
+      "learning_rate": 1.9489982534605098e-05,
+      "loss": 0.7051,
+      "step": 2479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.88011490371271,
+      "learning_rate": 1.948945070955624e-05,
+      "loss": 0.7667,
+      "step": 2480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.006794537053363,
+      "learning_rate": 1.9488918614632953e-05,
+      "loss": 0.6903,
+      "step": 2481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.41213134631574,
+      "learning_rate": 1.9488386249850372e-05,
+      "loss": 0.8073,
+      "step": 2482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 24.74164763735822,
+      "learning_rate": 1.9487853615223636e-05,
+      "loss": 0.8618,
+      "step": 2483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 28.02158039608608,
+      "learning_rate": 1.9487320710767898e-05,
+      "loss": 0.9661,
+      "step": 2484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.41546536046868,
+      "learning_rate": 1.9486787536498304e-05,
+      "loss": 0.8838,
+      "step": 2485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.02531547412535,
+      "learning_rate": 1.9486254092430022e-05,
+      "loss": 0.9066,
+      "step": 2486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.248648046662286,
+      "learning_rate": 1.9485720378578225e-05,
+      "loss": 0.8172,
+      "step": 2487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.50523668204783,
+      "learning_rate": 1.948518639495809e-05,
+      "loss": 0.7969,
+      "step": 2488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.12783320336865,
+      "learning_rate": 1.94846521415848e-05,
+      "loss": 0.9122,
+      "step": 2489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 30.81649489649323,
+      "learning_rate": 1.9484117618473553e-05,
+      "loss": 0.8958,
+      "step": 2490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.140569533430803,
+      "learning_rate": 1.948358282563955e-05,
+      "loss": 0.8575,
+      "step": 2491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.109909442061948,
+      "learning_rate": 1.9483047763097994e-05,
+      "loss": 0.8475,
+      "step": 2492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.933604053862425,
+      "learning_rate": 1.9482512430864113e-05,
+      "loss": 0.8546,
+      "step": 2493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.782840037220737,
+      "learning_rate": 1.9481976828953126e-05,
+      "loss": 0.848,
+      "step": 2494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.515427145793463,
+      "learning_rate": 1.9481440957380258e-05,
+      "loss": 0.8328,
+      "step": 2495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.258223283970217,
+      "learning_rate": 1.948090481616076e-05,
+      "loss": 0.9235,
+      "step": 2496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.32457242172472,
+      "learning_rate": 1.9480368405309876e-05,
+      "loss": 0.8043,
+      "step": 2497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 28.00501584092179,
+      "learning_rate": 1.9479831724842862e-05,
+      "loss": 0.8732,
+      "step": 2498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.98044102978033,
+      "learning_rate": 1.9479294774774974e-05,
+      "loss": 0.8157,
+      "step": 2499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.99142914663275,
+      "learning_rate": 1.9478757555121493e-05,
+      "loss": 0.8674,
+      "step": 2500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.768989350516703,
+      "learning_rate": 1.9478220065897687e-05,
+      "loss": 0.7731,
+      "step": 2501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.897985533324526,
+      "learning_rate": 1.947768230711885e-05,
+      "loss": 0.9334,
+      "step": 2502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 27.057438971284018,
+      "learning_rate": 1.9477144278800275e-05,
+      "loss": 0.8583,
+      "step": 2503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.44126465952835,
+      "learning_rate": 1.9476605980957258e-05,
+      "loss": 0.8965,
+      "step": 2504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.897611667484993,
+      "learning_rate": 1.947606741360511e-05,
+      "loss": 0.7554,
+      "step": 2505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 32.64812153674515,
+      "learning_rate": 1.9475528576759147e-05,
+      "loss": 0.8532,
+      "step": 2506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 33.427141424698995,
+      "learning_rate": 1.9474989470434695e-05,
+      "loss": 0.7965,
+      "step": 2507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.03586781371496,
+      "learning_rate": 1.9474450094647087e-05,
+      "loss": 0.7687,
+      "step": 2508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.1414637793719,
+      "learning_rate": 1.947391044941166e-05,
+      "loss": 0.7306,
+      "step": 2509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.98787337028632,
+      "learning_rate": 1.947337053474376e-05,
+      "loss": 0.9005,
+      "step": 2510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.887055852476287,
+      "learning_rate": 1.9472830350658748e-05,
+      "loss": 0.7767,
+      "step": 2511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.22012925811569,
+      "learning_rate": 1.9472289897171978e-05,
+      "loss": 0.8296,
+      "step": 2512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 27.639276062421043,
+      "learning_rate": 1.9471749174298827e-05,
+      "loss": 0.8719,
+      "step": 2513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.703361496140543,
+      "learning_rate": 1.9471208182054668e-05,
+      "loss": 0.735,
+      "step": 2514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.538841045503059,
+      "learning_rate": 1.947066692045489e-05,
+      "loss": 0.8326,
+      "step": 2515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.81425147168844,
+      "learning_rate": 1.9470125389514884e-05,
+      "loss": 0.813,
+      "step": 2516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.290948836591507,
+      "learning_rate": 1.9469583589250055e-05,
+      "loss": 0.8913,
+      "step": 2517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.048454680311696,
+      "learning_rate": 1.9469041519675806e-05,
+      "loss": 0.8645,
+      "step": 2518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 24.305522683022993,
+      "learning_rate": 1.9468499180807554e-05,
+      "loss": 0.8648,
+      "step": 2519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.836333264507807,
+      "learning_rate": 1.9467956572660727e-05,
+      "loss": 0.7702,
+      "step": 2520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.05402738118547,
+      "learning_rate": 1.9467413695250753e-05,
+      "loss": 0.9407,
+      "step": 2521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.635108099384144,
+      "learning_rate": 1.946687054859307e-05,
+      "loss": 0.9345,
+      "step": 2522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.09553264643005,
+      "learning_rate": 1.9466327132703128e-05,
+      "loss": 0.8608,
+      "step": 2523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.853735737980077,
+      "learning_rate": 1.946578344759638e-05,
+      "loss": 0.8349,
+      "step": 2524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.318348150735257,
+      "learning_rate": 1.946523949328829e-05,
+      "loss": 0.9876,
+      "step": 2525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.541723103066253,
+      "learning_rate": 1.9464695269794322e-05,
+      "loss": 0.8498,
+      "step": 2526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.278090788039147,
+      "learning_rate": 1.9464150777129956e-05,
+      "loss": 0.848,
+      "step": 2527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.57808443785332,
+      "learning_rate": 1.9463606015310684e-05,
+      "loss": 0.8322,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.491282169821478,
+      "learning_rate": 1.9463060984351988e-05,
+      "loss": 0.8076,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.797416678848354,
+      "learning_rate": 1.946251568426938e-05,
+      "loss": 0.8709,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.77472292459336,
+      "learning_rate": 1.9461970115078356e-05,
+      "loss": 0.8743,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.31539705440156,
+      "learning_rate": 1.9461424276794435e-05,
+      "loss": 0.8091,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 28.366568187486834,
+      "learning_rate": 1.9460878169433147e-05,
+      "loss": 1.0113,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 32.48096707042649,
+      "learning_rate": 1.9460331793010018e-05,
+      "loss": 0.9509,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.78370720239767,
+      "learning_rate": 1.9459785147540585e-05,
+      "loss": 0.9013,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.215463192254308,
+      "learning_rate": 1.9459238233040393e-05,
+      "loss": 0.7608,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 29.078353551943156,
+      "learning_rate": 1.9458691049525003e-05,
+      "loss": 0.7252,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.396372730432347,
+      "learning_rate": 1.9458143597009974e-05,
+      "loss": 0.7558,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.206016682786856,
+      "learning_rate": 1.9457595875510874e-05,
+      "loss": 0.8054,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.420907792507922,
+      "learning_rate": 1.9457047885043275e-05,
+      "loss": 0.7454,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 31.388612438205207,
+      "learning_rate": 1.945649962562277e-05,
+      "loss": 0.8577,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.179107481660534,
+      "learning_rate": 1.9455951097264947e-05,
+      "loss": 0.8072,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.118587854484836,
+      "learning_rate": 1.9455402299985408e-05,
+      "loss": 0.8226,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 34.098309812510784,
+      "learning_rate": 1.9454853233799756e-05,
+      "loss": 0.8972,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.84101018093514,
+      "learning_rate": 1.9454303898723613e-05,
+      "loss": 0.8312,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.21854358745712,
+      "learning_rate": 1.9453754294772593e-05,
+      "loss": 0.8243,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.23783987916674,
+      "learning_rate": 1.9453204421962334e-05,
+      "loss": 0.846,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.92782513595387,
+      "learning_rate": 1.945265428030847e-05,
+      "loss": 0.8237,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.45996879303278,
+      "learning_rate": 1.945210386982665e-05,
+      "loss": 0.9023,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.74723628226207,
+      "learning_rate": 1.9451553190532523e-05,
+      "loss": 0.7803,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.662748966228154,
+      "learning_rate": 1.945100224244175e-05,
+      "loss": 0.8573,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 26.167088374355984,
+      "learning_rate": 1.9450451025570006e-05,
+      "loss": 0.8045,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.369647147470108,
+      "learning_rate": 1.9449899539932962e-05,
+      "loss": 0.8749,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.154262547420428,
+      "learning_rate": 1.9449347785546303e-05,
+      "loss": 0.8593,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.46712800738409,
+      "learning_rate": 1.9448795762425722e-05,
+      "loss": 0.8296,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.195440759834735,
+      "learning_rate": 1.9448243470586913e-05,
+      "loss": 0.8126,
+      "step": 2556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.662371809438405,
+      "learning_rate": 1.944769091004559e-05,
+      "loss": 0.8672,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.915135451291725,
+      "learning_rate": 1.9447138080817466e-05,
+      "loss": 0.8059,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 33.48498326513195,
+      "learning_rate": 1.9446584982918256e-05,
+      "loss": 0.9164,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.587054317708027,
+      "learning_rate": 1.9446031616363702e-05,
+      "loss": 0.9321,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.981016468882483,
+      "learning_rate": 1.9445477981169528e-05,
+      "loss": 0.7125,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.959233508270154,
+      "learning_rate": 1.944492407735149e-05,
+      "loss": 0.7989,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.3023220892182,
+      "learning_rate": 1.9444369904925337e-05,
+      "loss": 0.9101,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 28.232542782831064,
+      "learning_rate": 1.9443815463906826e-05,
+      "loss": 0.8185,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.32370665246752,
+      "learning_rate": 1.9443260754311728e-05,
+      "loss": 0.8727,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.90503022940415,
+      "learning_rate": 1.944270577615582e-05,
+      "loss": 0.8984,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.48925276003296,
+      "learning_rate": 1.944215052945488e-05,
+      "loss": 0.8324,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.314506234030222,
+      "learning_rate": 1.9441595014224702e-05,
+      "loss": 0.8118,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.248241293835397,
+      "learning_rate": 1.944103923048109e-05,
+      "loss": 0.8533,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.064490976242134,
+      "learning_rate": 1.944048317823984e-05,
+      "loss": 0.8003,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.62154523721757,
+      "learning_rate": 1.9439926857516777e-05,
+      "loss": 0.8221,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.815765695892782,
+      "learning_rate": 1.9439370268327712e-05,
+      "loss": 0.9321,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.541433321944204,
+      "learning_rate": 1.9438813410688478e-05,
+      "loss": 0.7801,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.184646929430254,
+      "learning_rate": 1.9438256284614915e-05,
+      "loss": 0.8541,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.487637588092507,
+      "learning_rate": 1.943769889012286e-05,
+      "loss": 0.9258,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.381051461654067,
+      "learning_rate": 1.9437141227228175e-05,
+      "loss": 0.8346,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.422308540720923,
+      "learning_rate": 1.943658329594671e-05,
+      "loss": 0.9246,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.992872255931847,
+      "learning_rate": 1.9436025096294337e-05,
+      "loss": 0.8081,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.661583718384255,
+      "learning_rate": 1.9435466628286933e-05,
+      "loss": 0.7142,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.723917551601362,
+      "learning_rate": 1.9434907891940376e-05,
+      "loss": 0.7743,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.244948795001786,
+      "learning_rate": 1.943434888727056e-05,
+      "loss": 0.8379,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.210386887312627,
+      "learning_rate": 1.9433789614293376e-05,
+      "loss": 0.8381,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.886106102683641,
+      "learning_rate": 1.9433230073024737e-05,
+      "loss": 0.8676,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.00643830713997,
+      "learning_rate": 1.943267026348055e-05,
+      "loss": 0.7641,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.577011405528797,
+      "learning_rate": 1.943211018567674e-05,
+      "loss": 0.745,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.638061927585586,
+      "learning_rate": 1.9431549839629235e-05,
+      "loss": 0.7064,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.29325120004328,
+      "learning_rate": 1.943098922535397e-05,
+      "loss": 0.7488,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.640773912233634,
+      "learning_rate": 1.9430428342866888e-05,
+      "loss": 0.7827,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.03750906162206,
+      "learning_rate": 1.942986719218394e-05,
+      "loss": 0.7597,
+      "step": 2589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.373635234477586,
+      "learning_rate": 1.9429305773321085e-05,
+      "loss": 0.834,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 26.200292549633932,
+      "learning_rate": 1.9428744086294293e-05,
+      "loss": 0.8667,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.750456096009135,
+      "learning_rate": 1.942818213111953e-05,
+      "loss": 0.8944,
+      "step": 2592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.175852747481397,
+      "learning_rate": 1.9427619907812788e-05,
+      "loss": 0.8616,
+      "step": 2593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.656538642126716,
+      "learning_rate": 1.9427057416390048e-05,
+      "loss": 0.8284,
+      "step": 2594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 24.168777550937367,
+      "learning_rate": 1.942649465686731e-05,
+      "loss": 0.9372,
+      "step": 2595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.447199351089612,
+      "learning_rate": 1.9425931629260578e-05,
+      "loss": 0.8351,
+      "step": 2596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.96381596581886,
+      "learning_rate": 1.9425368333585862e-05,
+      "loss": 0.8431,
+      "step": 2597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.22198750003726,
+      "learning_rate": 1.9424804769859188e-05,
+      "loss": 0.8937,
+      "step": 2598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.90344687896506,
+      "learning_rate": 1.942424093809658e-05,
+      "loss": 0.8646,
+      "step": 2599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.041467011567903,
+      "learning_rate": 1.9423676838314067e-05,
+      "loss": 0.8619,
+      "step": 2600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.060994326514724,
+      "learning_rate": 1.9423112470527703e-05,
+      "loss": 0.9244,
+      "step": 2601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.994162290491946,
+      "learning_rate": 1.9422547834753532e-05,
+      "loss": 0.9067,
+      "step": 2602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.774520219794255,
+      "learning_rate": 1.942198293100761e-05,
+      "loss": 0.9444,
+      "step": 2603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.34939937930872,
+      "learning_rate": 1.9421417759306006e-05,
+      "loss": 0.8641,
+      "step": 2604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.431907047704055,
+      "learning_rate": 1.9420852319664797e-05,
+      "loss": 0.8796,
+      "step": 2605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 28.351289316323932,
+      "learning_rate": 1.9420286612100054e-05,
+      "loss": 0.8465,
+      "step": 2606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.141287664490484,
+      "learning_rate": 1.9419720636627874e-05,
+      "loss": 0.7885,
+      "step": 2607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.238178940393663,
+      "learning_rate": 1.9419154393264346e-05,
+      "loss": 0.7726,
+      "step": 2608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.774143510626377,
+      "learning_rate": 1.9418587882025578e-05,
+      "loss": 0.8392,
+      "step": 2609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.471464637756934,
+      "learning_rate": 1.941802110292768e-05,
+      "loss": 0.8838,
+      "step": 2610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.16966640337463,
+      "learning_rate": 1.941745405598677e-05,
+      "loss": 0.6908,
+      "step": 2611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.489898343229815,
+      "learning_rate": 1.941688674121898e-05,
+      "loss": 0.8888,
+      "step": 2612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.863847662501758,
+      "learning_rate": 1.941631915864044e-05,
+      "loss": 0.7215,
+      "step": 2613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.77807440109385,
+      "learning_rate": 1.9415751308267284e-05,
+      "loss": 0.8139,
+      "step": 2614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.827306565822177,
+      "learning_rate": 1.9415183190115678e-05,
+      "loss": 0.8615,
+      "step": 2615
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.65043099442133,
+      "learning_rate": 1.9414614804201764e-05,
+      "loss": 0.8808,
+      "step": 2616
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.045503889813485,
+      "learning_rate": 1.9414046150541712e-05,
+      "loss": 0.8402,
+      "step": 2617
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 24.587246551792816,
+      "learning_rate": 1.94134772291517e-05,
+      "loss": 0.7904,
+      "step": 2618
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.43876005868107,
+      "learning_rate": 1.9412908040047894e-05,
+      "loss": 0.8568,
+      "step": 2619
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.123789396234184,
+      "learning_rate": 1.94123385832465e-05,
+      "loss": 0.8212,
+      "step": 2620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.576751786741397,
+      "learning_rate": 1.9411768858763695e-05,
+      "loss": 0.8884,
+      "step": 2621
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.457738781891761,
+      "learning_rate": 1.941119886661569e-05,
+      "loss": 0.7221,
+      "step": 2622
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.87818287717947,
+      "learning_rate": 1.9410628606818696e-05,
+      "loss": 0.7998,
+      "step": 2623
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 26.376370772903343,
+      "learning_rate": 1.9410058079388933e-05,
+      "loss": 0.8713,
+      "step": 2624
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.514787003821464,
+      "learning_rate": 1.940948728434262e-05,
+      "loss": 0.7811,
+      "step": 2625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.618198144833002,
+      "learning_rate": 1.940891622169599e-05,
+      "loss": 0.8024,
+      "step": 2626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.111914775500992,
+      "learning_rate": 1.940834489146529e-05,
+      "loss": 0.8077,
+      "step": 2627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.8638801579698,
+      "learning_rate": 1.9407773293666764e-05,
+      "loss": 0.813,
+      "step": 2628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.381499617072896,
+      "learning_rate": 1.940720142831667e-05,
+      "loss": 0.8414,
+      "step": 2629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.59328042022629,
+      "learning_rate": 1.940662929543127e-05,
+      "loss": 0.8662,
+      "step": 2630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.48884477740405,
+      "learning_rate": 1.9406056895026837e-05,
+      "loss": 0.887,
+      "step": 2631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 24.64496365595773,
+      "learning_rate": 1.9405484227119646e-05,
+      "loss": 0.7966,
+      "step": 2632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.36209585684469,
+      "learning_rate": 1.9404911291725985e-05,
+      "loss": 0.8656,
+      "step": 2633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.859547374754367,
+      "learning_rate": 1.9404338088862152e-05,
+      "loss": 0.8752,
+      "step": 2634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.376115107822649,
+      "learning_rate": 1.940376461854444e-05,
+      "loss": 0.827,
+      "step": 2635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.630232947310596,
+      "learning_rate": 1.940319088078917e-05,
+      "loss": 0.826,
+      "step": 2636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.305490575839116,
+      "learning_rate": 1.9402616875612645e-05,
+      "loss": 0.8631,
+      "step": 2637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.794256984693206,
+      "learning_rate": 1.9402042603031202e-05,
+      "loss": 0.8284,
+      "step": 2638
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 46.730669959461885,
+      "learning_rate": 1.9401468063061165e-05,
+      "loss": 0.7978,
+      "step": 2639
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.69849547185854,
+      "learning_rate": 1.9400893255718874e-05,
+      "loss": 0.852,
+      "step": 2640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.864571354155228,
+      "learning_rate": 1.9400318181020682e-05,
+      "loss": 0.7427,
+      "step": 2641
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.975606694813866,
+      "learning_rate": 1.9399742838982937e-05,
+      "loss": 0.8502,
+      "step": 2642
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.738481673925747,
+      "learning_rate": 1.9399167229622008e-05,
+      "loss": 0.8133,
+      "step": 2643
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.74722582957297,
+      "learning_rate": 1.9398591352954258e-05,
+      "loss": 0.8737,
+      "step": 2644
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.801209265631115,
+      "learning_rate": 1.939801520899607e-05,
+      "loss": 0.8404,
+      "step": 2645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.29072114103525,
+      "learning_rate": 1.9397438797763825e-05,
+      "loss": 0.8134,
+      "step": 2646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.701624742579497,
+      "learning_rate": 1.9396862119273918e-05,
+      "loss": 0.805,
+      "step": 2647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.658906172625787,
+      "learning_rate": 1.939628517354275e-05,
+      "loss": 0.8634,
+      "step": 2648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.305061373737125,
+      "learning_rate": 1.939570796058673e-05,
+      "loss": 0.8961,
+      "step": 2649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.0478111361923,
+      "learning_rate": 1.939513048042227e-05,
+      "loss": 0.8788,
+      "step": 2650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.228470704640216,
+      "learning_rate": 1.9394552733065797e-05,
+      "loss": 0.8253,
+      "step": 2651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.60915031174712,
+      "learning_rate": 1.939397471853374e-05,
+      "loss": 0.8131,
+      "step": 2652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.542882152081166,
+      "learning_rate": 1.9393396436842537e-05,
+      "loss": 0.8406,
+      "step": 2653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.779320073894784,
+      "learning_rate": 1.9392817888008632e-05,
+      "loss": 0.8619,
+      "step": 2654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.323311619922016,
+      "learning_rate": 1.939223907204848e-05,
+      "loss": 0.8435,
+      "step": 2655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.03902363482122,
+      "learning_rate": 1.9391659988978546e-05,
+      "loss": 0.8012,
+      "step": 2656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.83149713891515,
+      "learning_rate": 1.9391080638815295e-05,
+      "loss": 0.7867,
+      "step": 2657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.66758118725523,
+      "learning_rate": 1.9390501021575203e-05,
+      "loss": 0.8648,
+      "step": 2658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.167484054660793,
+      "learning_rate": 1.938992113727476e-05,
+      "loss": 0.7431,
+      "step": 2659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.54112188354704,
+      "learning_rate": 1.9389340985930447e-05,
+      "loss": 0.8308,
+      "step": 2660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.94534245710849,
+      "learning_rate": 1.938876056755877e-05,
+      "loss": 0.8617,
+      "step": 2661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.41685642053697,
+      "learning_rate": 1.9388179882176237e-05,
+      "loss": 0.7977,
+      "step": 2662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.65406110649243,
+      "learning_rate": 1.938759892979936e-05,
+      "loss": 0.9213,
+      "step": 2663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.703268422098926,
+      "learning_rate": 1.9387017710444662e-05,
+      "loss": 0.8146,
+      "step": 2664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.150328875169098,
+      "learning_rate": 1.9386436224128668e-05,
+      "loss": 0.7223,
+      "step": 2665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.439974196960556,
+      "learning_rate": 1.938585447086792e-05,
+      "loss": 0.8488,
+      "step": 2666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.982365029709545,
+      "learning_rate": 1.9385272450678966e-05,
+      "loss": 0.7891,
+      "step": 2667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.709679270988122,
+      "learning_rate": 1.938469016357835e-05,
+      "loss": 0.7952,
+      "step": 2668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.691875524818492,
+      "learning_rate": 1.938410760958263e-05,
+      "loss": 0.8298,
+      "step": 2669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.818480974198497,
+      "learning_rate": 1.9383524788708387e-05,
+      "loss": 0.8908,
+      "step": 2670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.902835379961296,
+      "learning_rate": 1.9382941700972188e-05,
+      "loss": 0.7914,
+      "step": 2671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.726659806356395,
+      "learning_rate": 1.938235834639061e-05,
+      "loss": 0.8642,
+      "step": 2672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.178781863414006,
+      "learning_rate": 1.9381774724980253e-05,
+      "loss": 0.8986,
+      "step": 2673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.916274957607293,
+      "learning_rate": 1.9381190836757712e-05,
+      "loss": 0.7313,
+      "step": 2674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.844817039102114,
+      "learning_rate": 1.938060668173959e-05,
+      "loss": 0.8021,
+      "step": 2675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.15385254455125,
+      "learning_rate": 1.93800222599425e-05,
+      "loss": 0.8223,
+      "step": 2676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 25.65244877768913,
+      "learning_rate": 1.9379437571383067e-05,
+      "loss": 0.7941,
+      "step": 2677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.381240955530274,
+      "learning_rate": 1.9378852616077915e-05,
+      "loss": 0.8056,
+      "step": 2678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.230558983335847,
+      "learning_rate": 1.9378267394043678e-05,
+      "loss": 0.8657,
+      "step": 2679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.617910331165994,
+      "learning_rate": 1.9377681905297007e-05,
+      "loss": 0.9116,
+      "step": 2680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.23549219837821,
+      "learning_rate": 1.937709614985455e-05,
+      "loss": 0.9101,
+      "step": 2681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.55923682796077,
+      "learning_rate": 1.937651012773296e-05,
+      "loss": 0.8404,
+      "step": 2682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 24.6342227567183,
+      "learning_rate": 1.9375923838948907e-05,
+      "loss": 0.8992,
+      "step": 2683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.826724038442235,
+      "learning_rate": 1.9375337283519067e-05,
+      "loss": 0.7664,
+      "step": 2684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.281880268499513,
+      "learning_rate": 1.9374750461460117e-05,
+      "loss": 0.7582,
+      "step": 2685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.387846112902928,
+      "learning_rate": 1.9374163372788748e-05,
+      "loss": 0.8163,
+      "step": 2686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.980031414030144,
+      "learning_rate": 1.9373576017521657e-05,
+      "loss": 0.8424,
+      "step": 2687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.226036465604434,
+      "learning_rate": 1.9372988395675547e-05,
+      "loss": 0.7509,
+      "step": 2688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.09676331821948,
+      "learning_rate": 1.9372400507267132e-05,
+      "loss": 0.9732,
+      "step": 2689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.227840404579943,
+      "learning_rate": 1.9371812352313125e-05,
+      "loss": 0.8197,
+      "step": 2690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.38703806842231,
+      "learning_rate": 1.9371223930830264e-05,
+      "loss": 0.848,
+      "step": 2691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 30.472192629947614,
+      "learning_rate": 1.9370635242835272e-05,
+      "loss": 1.0408,
+      "step": 2692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.427443078613916,
+      "learning_rate": 1.9370046288344894e-05,
+      "loss": 0.784,
+      "step": 2693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.309832968515117,
+      "learning_rate": 1.9369457067375884e-05,
+      "loss": 0.8382,
+      "step": 2694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.275171236141343,
+      "learning_rate": 1.9368867579944994e-05,
+      "loss": 0.7785,
+      "step": 2695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.936757177918732,
+      "learning_rate": 1.936827782606899e-05,
+      "loss": 0.802,
+      "step": 2696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.114803473183333,
+      "learning_rate": 1.9367687805764647e-05,
+      "loss": 0.9395,
+      "step": 2697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.621051054466136,
+      "learning_rate": 1.936709751904874e-05,
+      "loss": 0.8338,
+      "step": 2698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.16405750203337,
+      "learning_rate": 1.936650696593806e-05,
+      "loss": 0.9549,
+      "step": 2699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.09112454742946,
+      "learning_rate": 1.9365916146449405e-05,
+      "loss": 0.872,
+      "step": 2700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.780151092780315,
+      "learning_rate": 1.936532506059957e-05,
+      "loss": 0.938,
+      "step": 2701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.783845257351054,
+      "learning_rate": 1.936473370840537e-05,
+      "loss": 0.9177,
+      "step": 2702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.529436400045263,
+      "learning_rate": 1.936414208988362e-05,
+      "loss": 0.8504,
+      "step": 2703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.865737750064977,
+      "learning_rate": 1.936355020505115e-05,
+      "loss": 0.9391,
+      "step": 2704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.11100780236159,
+      "learning_rate": 1.9362958053924786e-05,
+      "loss": 0.8241,
+      "step": 2705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.930240394988935,
+      "learning_rate": 1.9362365636521377e-05,
+      "loss": 0.8171,
+      "step": 2706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.082579719991905,
+      "learning_rate": 1.9361772952857762e-05,
+      "loss": 0.8348,
+      "step": 2707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.997880650018738,
+      "learning_rate": 1.93611800029508e-05,
+      "loss": 0.7933,
+      "step": 2708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 25.995473620715693,
+      "learning_rate": 1.9360586786817355e-05,
+      "loss": 0.8625,
+      "step": 2709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.979720218080338,
+      "learning_rate": 1.9359993304474302e-05,
+      "loss": 0.8743,
+      "step": 2710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.680965757860424,
+      "learning_rate": 1.935939955593851e-05,
+      "loss": 0.8663,
+      "step": 2711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 30.115294095895603,
+      "learning_rate": 1.9358805541226872e-05,
+      "loss": 0.8895,
+      "step": 2712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.827004143683194,
+      "learning_rate": 1.9358211260356282e-05,
+      "loss": 0.8517,
+      "step": 2713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.58248676962657,
+      "learning_rate": 1.9357616713343633e-05,
+      "loss": 0.857,
+      "step": 2714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.93136013431971,
+      "learning_rate": 1.935702190020584e-05,
+      "loss": 0.8615,
+      "step": 2715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.677871383670924,
+      "learning_rate": 1.9356426820959817e-05,
+      "loss": 0.8615,
+      "step": 2716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.538824027642022,
+      "learning_rate": 1.935583147562249e-05,
+      "loss": 0.9242,
+      "step": 2717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.46955974493218,
+      "learning_rate": 1.9355235864210792e-05,
+      "loss": 1.0344,
+      "step": 2718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.194503384287984,
+      "learning_rate": 1.9354639986741653e-05,
+      "loss": 0.9864,
+      "step": 2719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 24.302758743922624,
+      "learning_rate": 1.9354043843232028e-05,
+      "loss": 0.9412,
+      "step": 2720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.898592952387137,
+      "learning_rate": 1.935344743369887e-05,
+      "loss": 0.8174,
+      "step": 2721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.005242735707103,
+      "learning_rate": 1.9352850758159136e-05,
+      "loss": 0.8357,
+      "step": 2722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 28.29533521537663,
+      "learning_rate": 1.9352253816629796e-05,
+      "loss": 0.8839,
+      "step": 2723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.290798532477393,
+      "learning_rate": 1.9351656609127833e-05,
+      "loss": 0.758,
+      "step": 2724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.259434209592417,
+      "learning_rate": 1.9351059135670222e-05,
+      "loss": 0.827,
+      "step": 2725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 27.97107893874645,
+      "learning_rate": 1.9350461396273963e-05,
+      "loss": 0.8841,
+      "step": 2726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.08908614714832,
+      "learning_rate": 1.934986339095605e-05,
+      "loss": 0.8445,
+      "step": 2727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.594918757657876,
+      "learning_rate": 1.934926511973349e-05,
+      "loss": 0.7765,
+      "step": 2728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.439456423537294,
+      "learning_rate": 1.9348666582623302e-05,
+      "loss": 0.7948,
+      "step": 2729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.495766741869282,
+      "learning_rate": 1.9348067779642506e-05,
+      "loss": 0.8457,
+      "step": 2730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.42110042932668,
+      "learning_rate": 1.9347468710808128e-05,
+      "loss": 0.8191,
+      "step": 2731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.988561329125673,
+      "learning_rate": 1.9346869376137206e-05,
+      "loss": 0.8307,
+      "step": 2732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.340376920452577,
+      "learning_rate": 1.9346269775646793e-05,
+      "loss": 0.8083,
+      "step": 2733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.13230733836942,
+      "learning_rate": 1.9345669909353934e-05,
+      "loss": 0.8327,
+      "step": 2734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.948771506687264,
+      "learning_rate": 1.9345069777275685e-05,
+      "loss": 0.7678,
+      "step": 2735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.722435713386524,
+      "learning_rate": 1.934446937942912e-05,
+      "loss": 0.7494,
+      "step": 2736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.34788575734836,
+      "learning_rate": 1.9343868715831313e-05,
+      "loss": 0.7902,
+      "step": 2737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.809935082445637,
+      "learning_rate": 1.9343267786499346e-05,
+      "loss": 0.7964,
+      "step": 2738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.479036582074325,
+      "learning_rate": 1.9342666591450307e-05,
+      "loss": 0.8675,
+      "step": 2739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.61778473520191,
+      "learning_rate": 1.9342065130701297e-05,
+      "loss": 0.8963,
+      "step": 2740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.35420822044292,
+      "learning_rate": 1.9341463404269422e-05,
+      "loss": 0.8754,
+      "step": 2741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.1031477576541,
+      "learning_rate": 1.934086141217179e-05,
+      "loss": 0.7699,
+      "step": 2742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.799582350019515,
+      "learning_rate": 1.9340259154425524e-05,
+      "loss": 0.7482,
+      "step": 2743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 26.268775707852193,
+      "learning_rate": 1.933965663104775e-05,
+      "loss": 0.878,
+      "step": 2744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.606657994922337,
+      "learning_rate": 1.9339053842055606e-05,
+      "loss": 0.8324,
+      "step": 2745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.30469492574248,
+      "learning_rate": 1.9338450787466234e-05,
+      "loss": 0.7594,
+      "step": 2746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.01005121246038,
+      "learning_rate": 1.9337847467296783e-05,
+      "loss": 0.7594,
+      "step": 2747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.100149317739525,
+      "learning_rate": 1.9337243881564417e-05,
+      "loss": 0.7973,
+      "step": 2748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.149178918071595,
+      "learning_rate": 1.9336640030286293e-05,
+      "loss": 0.8795,
+      "step": 2749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.784110354669203,
+      "learning_rate": 1.933603591347959e-05,
+      "loss": 0.9595,
+      "step": 2750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.477578478672543,
+      "learning_rate": 1.9335431531161486e-05,
+      "loss": 0.7808,
+      "step": 2751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.147766612888061,
+      "learning_rate": 1.933482688334917e-05,
+      "loss": 0.7993,
+      "step": 2752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.406671432201627,
+      "learning_rate": 1.9334221970059837e-05,
+      "loss": 0.7119,
+      "step": 2753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.062424950640239,
+      "learning_rate": 1.9333616791310696e-05,
+      "loss": 0.7349,
+      "step": 2754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.54979717570123,
+      "learning_rate": 1.9333011347118953e-05,
+      "loss": 0.9561,
+      "step": 2755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.768112161300735,
+      "learning_rate": 1.9332405637501823e-05,
+      "loss": 0.8222,
+      "step": 2756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.354478960966798,
+      "learning_rate": 1.9331799662476537e-05,
+      "loss": 0.945,
+      "step": 2757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.542335764767387,
+      "learning_rate": 1.933119342206033e-05,
+      "loss": 0.8322,
+      "step": 2758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.29031479227513,
+      "learning_rate": 1.933058691627044e-05,
+      "loss": 0.8184,
+      "step": 2759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.46496870572795,
+      "learning_rate": 1.9329980145124115e-05,
+      "loss": 0.8366,
+      "step": 2760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.949423400994245,
+      "learning_rate": 1.9329373108638614e-05,
+      "loss": 0.7697,
+      "step": 2761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.251941269839598,
+      "learning_rate": 1.93287658068312e-05,
+      "loss": 0.8186,
+      "step": 2762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.620956735969866,
+      "learning_rate": 1.932815823971914e-05,
+      "loss": 0.7985,
+      "step": 2763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.906531339026152,
+      "learning_rate": 1.9327550407319717e-05,
+      "loss": 0.9988,
+      "step": 2764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.057307608259677,
+      "learning_rate": 1.932694230965022e-05,
+      "loss": 0.7492,
+      "step": 2765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.8766000479919,
+      "learning_rate": 1.9326333946727938e-05,
+      "loss": 0.8733,
+      "step": 2766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.366252670778806,
+      "learning_rate": 1.932572531857017e-05,
+      "loss": 0.8315,
+      "step": 2767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.97704605480904,
+      "learning_rate": 1.9325116425194235e-05,
+      "loss": 0.873,
+      "step": 2768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.93085267014662,
+      "learning_rate": 1.9324507266617444e-05,
+      "loss": 0.864,
+      "step": 2769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 25.14933070516083,
+      "learning_rate": 1.932389784285712e-05,
+      "loss": 0.9263,
+      "step": 2770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.647679178912497,
+      "learning_rate": 1.9323288153930595e-05,
+      "loss": 0.7297,
+      "step": 2771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.797743218961239,
+      "learning_rate": 1.932267819985521e-05,
+      "loss": 0.8237,
+      "step": 2772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.2180337131879,
+      "learning_rate": 1.932206798064831e-05,
+      "loss": 0.8558,
+      "step": 2773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.404121098252574,
+      "learning_rate": 1.932145749632725e-05,
+      "loss": 0.8191,
+      "step": 2774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.558964219507182,
+      "learning_rate": 1.932084674690939e-05,
+      "loss": 0.796,
+      "step": 2775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 25.654297185751886,
+      "learning_rate": 1.9320235732412104e-05,
+      "loss": 0.9271,
+      "step": 2776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.460692308822745,
+      "learning_rate": 1.9319624452852765e-05,
+      "loss": 0.7587,
+      "step": 2777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 25.55434798502312,
+      "learning_rate": 1.931901290824876e-05,
+      "loss": 0.8581,
+      "step": 2778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.45949459984029,
+      "learning_rate": 1.9318401098617475e-05,
+      "loss": 0.782,
+      "step": 2779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.010583702854948,
+      "learning_rate": 1.9317789023976314e-05,
+      "loss": 0.7699,
+      "step": 2780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.76329166300339,
+      "learning_rate": 1.9317176684342685e-05,
+      "loss": 0.8075,
+      "step": 2781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.381196447646076,
+      "learning_rate": 1.9316564079734005e-05,
+      "loss": 0.813,
+      "step": 2782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.620988882648282,
+      "learning_rate": 1.931595121016769e-05,
+      "loss": 0.8484,
+      "step": 2783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.305390716618515,
+      "learning_rate": 1.9315338075661172e-05,
+      "loss": 0.8336,
+      "step": 2784
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.44467630605795,
+      "learning_rate": 1.931472467623189e-05,
+      "loss": 0.772,
+      "step": 2785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.793004956683315,
+      "learning_rate": 1.9314111011897285e-05,
+      "loss": 0.9619,
+      "step": 2786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.540675940737646,
+      "learning_rate": 1.9313497082674813e-05,
+      "loss": 0.9043,
+      "step": 2787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.164699910261472,
+      "learning_rate": 1.931288288858193e-05,
+      "loss": 0.8337,
+      "step": 2788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.454483928343135,
+      "learning_rate": 1.9312268429636108e-05,
+      "loss": 0.6702,
+      "step": 2789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.001847195258666,
+      "learning_rate": 1.9311653705854817e-05,
+      "loss": 0.7726,
+      "step": 2790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 29.434434865819178,
+      "learning_rate": 1.9311038717255542e-05,
+      "loss": 1.0484,
+      "step": 2791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.90567106486018,
+      "learning_rate": 1.9310423463855774e-05,
+      "loss": 0.9265,
+      "step": 2792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.145564394716416,
+      "learning_rate": 1.930980794567301e-05,
+      "loss": 0.8952,
+      "step": 2793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 42.86813095108527,
+      "learning_rate": 1.9309192162724756e-05,
+      "loss": 0.9466,
+      "step": 2794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.99347205769062,
+      "learning_rate": 1.9308576115028515e-05,
+      "loss": 0.7671,
+      "step": 2795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 27.096805186308767,
+      "learning_rate": 1.9307959802601824e-05,
+      "loss": 0.856,
+      "step": 2796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.20832649231591,
+      "learning_rate": 1.9307343225462195e-05,
+      "loss": 0.7815,
+      "step": 2797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.836475105773456,
+      "learning_rate": 1.930672638362717e-05,
+      "loss": 0.8059,
+      "step": 2798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.827430523571735,
+      "learning_rate": 1.9306109277114292e-05,
+      "loss": 0.8531,
+      "step": 2799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.487774256797177,
+      "learning_rate": 1.930549190594111e-05,
+      "loss": 0.7303,
+      "step": 2800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.2695687591317,
+      "learning_rate": 1.930487427012518e-05,
+      "loss": 0.9313,
+      "step": 2801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.644877286452623,
+      "learning_rate": 1.930425636968407e-05,
+      "loss": 0.8546,
+      "step": 2802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 29.70245533147644,
+      "learning_rate": 1.9303638204635354e-05,
+      "loss": 0.8352,
+      "step": 2803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.042241174530986,
+      "learning_rate": 1.930301977499661e-05,
+      "loss": 0.8288,
+      "step": 2804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.242139636016038,
+      "learning_rate": 1.930240108078542e-05,
+      "loss": 0.8485,
+      "step": 2805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.388418784413442,
+      "learning_rate": 1.9301782122019392e-05,
+      "loss": 0.7612,
+      "step": 2806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.451465471103637,
+      "learning_rate": 1.9301162898716116e-05,
+      "loss": 0.8578,
+      "step": 2807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.405939431400558,
+      "learning_rate": 1.9300543410893213e-05,
+      "loss": 0.8356,
+      "step": 2808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 27.165890540818857,
+      "learning_rate": 1.9299923658568294e-05,
+      "loss": 0.7887,
+      "step": 2809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.407376016101875,
+      "learning_rate": 1.9299303641758986e-05,
+      "loss": 0.7502,
+      "step": 2810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.465288114829953,
+      "learning_rate": 1.9298683360482923e-05,
+      "loss": 0.8396,
+      "step": 2811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 34.92772760867393,
+      "learning_rate": 1.9298062814757746e-05,
+      "loss": 0.9241,
+      "step": 2812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.657996869052813,
+      "learning_rate": 1.92974420046011e-05,
+      "loss": 0.7611,
+      "step": 2813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.418309645230106,
+      "learning_rate": 1.9296820930030642e-05,
+      "loss": 0.7252,
+      "step": 2814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.719098678739265,
+      "learning_rate": 1.9296199591064036e-05,
+      "loss": 0.7518,
+      "step": 2815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.900043727099296,
+      "learning_rate": 1.9295577987718953e-05,
+      "loss": 0.8918,
+      "step": 2816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.353443178660008,
+      "learning_rate": 1.929495612001307e-05,
+      "loss": 0.8403,
+      "step": 2817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 29.80150160354507,
+      "learning_rate": 1.9294333987964067e-05,
+      "loss": 0.8056,
+      "step": 2818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.495019609607217,
+      "learning_rate": 1.9293711591589645e-05,
+      "loss": 0.8668,
+      "step": 2819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.93091527891047,
+      "learning_rate": 1.9293088930907505e-05,
+      "loss": 0.7623,
+      "step": 2820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.685629673804485,
+      "learning_rate": 1.9292466005935352e-05,
+      "loss": 0.7573,
+      "step": 2821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.229591706205877,
+      "learning_rate": 1.9291842816690898e-05,
+      "loss": 0.7855,
+      "step": 2822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.20478784208699,
+      "learning_rate": 1.9291219363191873e-05,
+      "loss": 0.8322,
+      "step": 2823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.574667659646149,
+      "learning_rate": 1.9290595645456003e-05,
+      "loss": 0.7588,
+      "step": 2824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.720982019705515,
+      "learning_rate": 1.9289971663501027e-05,
+      "loss": 0.8363,
+      "step": 2825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.736009845229447,
+      "learning_rate": 1.9289347417344694e-05,
+      "loss": 0.768,
+      "step": 2826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.70867224969513,
+      "learning_rate": 1.928872290700475e-05,
+      "loss": 0.8619,
+      "step": 2827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.94005813864722,
+      "learning_rate": 1.9288098132498966e-05,
+      "loss": 0.817,
+      "step": 2828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.729340832316538,
+      "learning_rate": 1.9287473093845102e-05,
+      "loss": 0.7662,
+      "step": 2829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.33096850540217,
+      "learning_rate": 1.9286847791060937e-05,
+      "loss": 0.8533,
+      "step": 2830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.731406791945428,
+      "learning_rate": 1.928622222416425e-05,
+      "loss": 0.7649,
+      "step": 2831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.22571892330583,
+      "learning_rate": 1.928559639317284e-05,
+      "loss": 0.8109,
+      "step": 2832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.07722226960523,
+      "learning_rate": 1.9284970298104495e-05,
+      "loss": 0.7884,
+      "step": 2833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.24392876091876,
+      "learning_rate": 1.928434393897703e-05,
+      "loss": 0.8451,
+      "step": 2834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.055815044775574,
+      "learning_rate": 1.9283717315808255e-05,
+      "loss": 0.8385,
+      "step": 2835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.294156363632272,
+      "learning_rate": 1.928309042861599e-05,
+      "loss": 0.7455,
+      "step": 2836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.476296005177545,
+      "learning_rate": 1.9282463277418062e-05,
+      "loss": 0.911,
+      "step": 2837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.64769102971277,
+      "learning_rate": 1.928183586223231e-05,
+      "loss": 0.8264,
+      "step": 2838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.59383549390099,
+      "learning_rate": 1.9281208183076576e-05,
+      "loss": 0.8206,
+      "step": 2839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.910436317962915,
+      "learning_rate": 1.9280580239968708e-05,
+      "loss": 0.8734,
+      "step": 2840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.43049350806382,
+      "learning_rate": 1.927995203292657e-05,
+      "loss": 0.88,
+      "step": 2841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.987091362221824,
+      "learning_rate": 1.9279323561968024e-05,
+      "loss": 0.7627,
+      "step": 2842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.835625789244995,
+      "learning_rate": 1.927869482711094e-05,
+      "loss": 0.7323,
+      "step": 2843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.991106596922414,
+      "learning_rate": 1.927806582837321e-05,
+      "loss": 0.8016,
+      "step": 2844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.10333686844645,
+      "learning_rate": 1.927743656577271e-05,
+      "loss": 0.9001,
+      "step": 2845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 28.852537134557487,
+      "learning_rate": 1.9276807039327344e-05,
+      "loss": 0.9389,
+      "step": 2846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.031735257963781,
+      "learning_rate": 1.9276177249055012e-05,
+      "loss": 0.7254,
+      "step": 2847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.636012338566934,
+      "learning_rate": 1.9275547194973626e-05,
+      "loss": 0.879,
+      "step": 2848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.620915853524366,
+      "learning_rate": 1.9274916877101104e-05,
+      "loss": 0.843,
+      "step": 2849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.696749001023305,
+      "learning_rate": 1.9274286295455373e-05,
+      "loss": 0.6976,
+      "step": 2850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.36699915098846,
+      "learning_rate": 1.927365545005436e-05,
+      "loss": 0.8912,
+      "step": 2851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 30.587851371664442,
+      "learning_rate": 1.9273024340916015e-05,
+      "loss": 0.9167,
+      "step": 2852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.504500437316278,
+      "learning_rate": 1.9272392968058282e-05,
+      "loss": 0.7861,
+      "step": 2853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 32.04225498861467,
+      "learning_rate": 1.9271761331499114e-05,
+      "loss": 0.9072,
+      "step": 2854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.428991956182465,
+      "learning_rate": 1.927112943125648e-05,
+      "loss": 0.8342,
+      "step": 2855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 30.74141216754505,
+      "learning_rate": 1.9270497267348348e-05,
+      "loss": 0.8617,
+      "step": 2856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.899881867271972,
+      "learning_rate": 1.9269864839792697e-05,
+      "loss": 0.8082,
+      "step": 2857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.883514953630044,
+      "learning_rate": 1.9269232148607515e-05,
+      "loss": 0.8591,
+      "step": 2858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.003423591312604,
+      "learning_rate": 1.926859919381079e-05,
+      "loss": 0.7978,
+      "step": 2859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.679820444582184,
+      "learning_rate": 1.926796597542053e-05,
+      "loss": 0.9388,
+      "step": 2860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.481095958348472,
+      "learning_rate": 1.9267332493454732e-05,
+      "loss": 0.7668,
+      "step": 2861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.287048395966394,
+      "learning_rate": 1.9266698747931425e-05,
+      "loss": 0.7206,
+      "step": 2862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.923573275715178,
+      "learning_rate": 1.9266064738868625e-05,
+      "loss": 0.847,
+      "step": 2863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.292903726689474,
+      "learning_rate": 1.9265430466284362e-05,
+      "loss": 0.871,
+      "step": 2864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 32.18323858162131,
+      "learning_rate": 1.9264795930196677e-05,
+      "loss": 0.9068,
+      "step": 2865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.809343413020226,
+      "learning_rate": 1.9264161130623618e-05,
+      "loss": 0.9016,
+      "step": 2866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.926370773381045,
+      "learning_rate": 1.9263526067583235e-05,
+      "loss": 0.8996,
+      "step": 2867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.92374671214326,
+      "learning_rate": 1.926289074109359e-05,
+      "loss": 0.7603,
+      "step": 2868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.096887386183415,
+      "learning_rate": 1.9262255151172752e-05,
+      "loss": 0.8825,
+      "step": 2869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.555201177905374,
+      "learning_rate": 1.9261619297838794e-05,
+      "loss": 0.8921,
+      "step": 2870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.373893600938462,
+      "learning_rate": 1.92609831811098e-05,
+      "loss": 0.8113,
+      "step": 2871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.639356349355918,
+      "learning_rate": 1.9260346801003864e-05,
+      "loss": 0.8037,
+      "step": 2872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.585625110995945,
+      "learning_rate": 1.925971015753908e-05,
+      "loss": 0.8173,
+      "step": 2873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.404036637200182,
+      "learning_rate": 1.925907325073356e-05,
+      "loss": 0.8761,
+      "step": 2874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.06670253819081,
+      "learning_rate": 1.925843608060541e-05,
+      "loss": 0.938,
+      "step": 2875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.330411233051962,
+      "learning_rate": 1.925779864717275e-05,
+      "loss": 0.7903,
+      "step": 2876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.969001946964816,
+      "learning_rate": 1.925716095045372e-05,
+      "loss": 0.8028,
+      "step": 2877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.030716748432223,
+      "learning_rate": 1.9256522990466445e-05,
+      "loss": 0.8031,
+      "step": 2878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.625710126905126,
+      "learning_rate": 1.925588476722907e-05,
+      "loss": 0.8059,
+      "step": 2879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 36.493411691443036,
+      "learning_rate": 1.9255246280759747e-05,
+      "loss": 0.802,
+      "step": 2880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.679559905608317,
+      "learning_rate": 1.9254607531076633e-05,
+      "loss": 0.735,
+      "step": 2881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.48163898319547,
+      "learning_rate": 1.9253968518197896e-05,
+      "loss": 0.8374,
+      "step": 2882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.901193054065537,
+      "learning_rate": 1.925332924214171e-05,
+      "loss": 0.8883,
+      "step": 2883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.677948101766702,
+      "learning_rate": 1.925268970292625e-05,
+      "loss": 0.8174,
+      "step": 2884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.600996014098943,
+      "learning_rate": 1.9252049900569707e-05,
+      "loss": 0.8718,
+      "step": 2885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.30245772040695,
+      "learning_rate": 1.925140983509028e-05,
+      "loss": 0.7605,
+      "step": 2886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.85453456772631,
+      "learning_rate": 1.9250769506506164e-05,
+      "loss": 0.8221,
+      "step": 2887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.577647359604168,
+      "learning_rate": 1.925012891483558e-05,
+      "loss": 0.7792,
+      "step": 2888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.794540209930005,
+      "learning_rate": 1.924948806009674e-05,
+      "loss": 0.7336,
+      "step": 2889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.356493493907163,
+      "learning_rate": 1.9248846942307867e-05,
+      "loss": 0.7764,
+      "step": 2890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.396885541591361,
+      "learning_rate": 1.92482055614872e-05,
+      "loss": 0.8077,
+      "step": 2891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 29.39138071912793,
+      "learning_rate": 1.9247563917652978e-05,
+      "loss": 0.74,
+      "step": 2892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.058876910531712,
+      "learning_rate": 1.9246922010823445e-05,
+      "loss": 0.7889,
+      "step": 2893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.35058201205367,
+      "learning_rate": 1.924627984101686e-05,
+      "loss": 0.8713,
+      "step": 2894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.303365962775334,
+      "learning_rate": 1.924563740825149e-05,
+      "loss": 0.7908,
+      "step": 2895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.218575881541433,
+      "learning_rate": 1.9244994712545596e-05,
+      "loss": 0.7346,
+      "step": 2896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.45784247951282,
+      "learning_rate": 1.924435175391746e-05,
+      "loss": 0.8048,
+      "step": 2897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.769205526806456,
+      "learning_rate": 1.924370853238537e-05,
+      "loss": 0.7424,
+      "step": 2898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.916617140223753,
+      "learning_rate": 1.9243065047967614e-05,
+      "loss": 0.7788,
+      "step": 2899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.888614085555783,
+      "learning_rate": 1.92424213006825e-05,
+      "loss": 0.8761,
+      "step": 2900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.87728224239912,
+      "learning_rate": 1.924177729054833e-05,
+      "loss": 0.894,
+      "step": 2901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 38.50258540399339,
+      "learning_rate": 1.9241133017583416e-05,
+      "loss": 1.0238,
+      "step": 2902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.942786794876184,
+      "learning_rate": 1.9240488481806086e-05,
+      "loss": 0.772,
+      "step": 2903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.247008721428116,
+      "learning_rate": 1.923984368323467e-05,
+      "loss": 0.7697,
+      "step": 2904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.128511498190324,
+      "learning_rate": 1.9239198621887505e-05,
+      "loss": 0.8602,
+      "step": 2905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.063971694778907,
+      "learning_rate": 1.9238553297782937e-05,
+      "loss": 0.7644,
+      "step": 2906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.009838367951684,
+      "learning_rate": 1.9237907710939317e-05,
+      "loss": 0.7811,
+      "step": 2907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.240008564197918,
+      "learning_rate": 1.9237261861375004e-05,
+      "loss": 0.9248,
+      "step": 2908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.23992104588431,
+      "learning_rate": 1.923661574910837e-05,
+      "loss": 0.7512,
+      "step": 2909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.939374063872858,
+      "learning_rate": 1.9235969374157786e-05,
+      "loss": 0.9421,
+      "step": 2910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.181644877669257,
+      "learning_rate": 1.9235322736541635e-05,
+      "loss": 0.8945,
+      "step": 2911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.533710042537482,
+      "learning_rate": 1.9234675836278308e-05,
+      "loss": 0.846,
+      "step": 2912
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.960547458816901,
+      "learning_rate": 1.9234028673386205e-05,
+      "loss": 0.7696,
+      "step": 2913
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.396502906181674,
+      "learning_rate": 1.9233381247883724e-05,
+      "loss": 0.8041,
+      "step": 2914
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.387321258067743,
+      "learning_rate": 1.9232733559789286e-05,
+      "loss": 0.796,
+      "step": 2915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.071487851572833,
+      "learning_rate": 1.9232085609121305e-05,
+      "loss": 0.8013,
+      "step": 2916
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.002491058816016,
+      "learning_rate": 1.9231437395898207e-05,
+      "loss": 0.7197,
+      "step": 2917
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.68117358451436,
+      "learning_rate": 1.9230788920138433e-05,
+      "loss": 0.8017,
+      "step": 2918
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.04612653430889,
+      "learning_rate": 1.923014018186042e-05,
+      "loss": 0.7759,
+      "step": 2919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.44459704258811,
+      "learning_rate": 1.922949118108262e-05,
+      "loss": 0.8843,
+      "step": 2920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 22.341203785582618,
+      "learning_rate": 1.9228841917823492e-05,
+      "loss": 0.8406,
+      "step": 2921
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.879511315024136,
+      "learning_rate": 1.92281923921015e-05,
+      "loss": 0.8005,
+      "step": 2922
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.180477080078573,
+      "learning_rate": 1.922754260393511e-05,
+      "loss": 0.7988,
+      "step": 2923
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.55693439175142,
+      "learning_rate": 1.9226892553342808e-05,
+      "loss": 0.7678,
+      "step": 2924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.63735886531153,
+      "learning_rate": 1.9226242240343082e-05,
+      "loss": 0.8642,
+      "step": 2925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.689179542921874,
+      "learning_rate": 1.9225591664954423e-05,
+      "loss": 0.7621,
+      "step": 2926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.62090388367294,
+      "learning_rate": 1.922494082719533e-05,
+      "loss": 0.8511,
+      "step": 2927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.44901606835908,
+      "learning_rate": 1.922428972708432e-05,
+      "loss": 0.8152,
+      "step": 2928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 273.49283357951964,
+      "learning_rate": 1.9223638364639902e-05,
+      "loss": 0.8089,
+      "step": 2929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.336963973047034,
+      "learning_rate": 1.9222986739880607e-05,
+      "loss": 0.8745,
+      "step": 2930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.416894450523952,
+      "learning_rate": 1.9222334852824966e-05,
+      "loss": 0.7592,
+      "step": 2931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 30.967808432945866,
+      "learning_rate": 1.922168270349152e-05,
+      "loss": 0.8561,
+      "step": 2932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.312385426811293,
+      "learning_rate": 1.9221030291898802e-05,
+      "loss": 0.8278,
+      "step": 2933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.97397118798427,
+      "learning_rate": 1.9220377618065383e-05,
+      "loss": 0.7038,
+      "step": 2934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.4901065990765,
+      "learning_rate": 1.921972468200982e-05,
+      "loss": 0.7889,
+      "step": 2935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 22.625603199526246,
+      "learning_rate": 1.9219071483750678e-05,
+      "loss": 0.8768,
+      "step": 2936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.20967454333816,
+      "learning_rate": 1.9218418023306536e-05,
+      "loss": 0.8274,
+      "step": 2937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 24.141432391247864,
+      "learning_rate": 1.9217764300695977e-05,
+      "loss": 0.7748,
+      "step": 2938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.316808523052693,
+      "learning_rate": 1.9217110315937597e-05,
+      "loss": 0.835,
+      "step": 2939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 30.469695912863475,
+      "learning_rate": 1.921645606904999e-05,
+      "loss": 0.891,
+      "step": 2940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.706821970594095,
+      "learning_rate": 1.9215801560051764e-05,
+      "loss": 0.8276,
+      "step": 2941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.769118284834967,
+      "learning_rate": 1.921514678896153e-05,
+      "loss": 0.7584,
+      "step": 2942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.253685453527048,
+      "learning_rate": 1.9214491755797916e-05,
+      "loss": 0.7512,
+      "step": 2943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.465678317509454,
+      "learning_rate": 1.9213836460579546e-05,
+      "loss": 0.8682,
+      "step": 2944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.11941303362085,
+      "learning_rate": 1.9213180903325056e-05,
+      "loss": 0.8247,
+      "step": 2945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 24.692209704748254,
+      "learning_rate": 1.921252508405309e-05,
+      "loss": 0.8497,
+      "step": 2946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.01600024794213,
+      "learning_rate": 1.92118690027823e-05,
+      "loss": 0.8728,
+      "step": 2947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.891205517207375,
+      "learning_rate": 1.9211212659531345e-05,
+      "loss": 0.809,
+      "step": 2948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.101412431874632,
+      "learning_rate": 1.9210556054318886e-05,
+      "loss": 0.7563,
+      "step": 2949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.680172858223028,
+      "learning_rate": 1.9209899187163606e-05,
+      "loss": 0.8506,
+      "step": 2950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.491085455744027,
+      "learning_rate": 1.920924205808418e-05,
+      "loss": 0.7501,
+      "step": 2951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.866830131057615,
+      "learning_rate": 1.92085846670993e-05,
+      "loss": 0.7544,
+      "step": 2952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.84464375166633,
+      "learning_rate": 1.9207927014227653e-05,
+      "loss": 0.8491,
+      "step": 2953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.44809408698163,
+      "learning_rate": 1.9207269099487953e-05,
+      "loss": 0.8276,
+      "step": 2954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.07766197027408,
+      "learning_rate": 1.9206610922898904e-05,
+      "loss": 0.7609,
+      "step": 2955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.54914848905122,
+      "learning_rate": 1.9205952484479225e-05,
+      "loss": 0.8494,
+      "step": 2956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.0218185979425,
+      "learning_rate": 1.920529378424764e-05,
+      "loss": 0.8645,
+      "step": 2957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.94251281096728,
+      "learning_rate": 1.920463482222289e-05,
+      "loss": 0.8819,
+      "step": 2958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.004192332047143,
+      "learning_rate": 1.920397559842371e-05,
+      "loss": 0.8254,
+      "step": 2959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.4700718577476,
+      "learning_rate": 1.9203316112868844e-05,
+      "loss": 0.7932,
+      "step": 2960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.28652544928423,
+      "learning_rate": 1.9202656365577056e-05,
+      "loss": 0.7781,
+      "step": 2961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 40.843648733817595,
+      "learning_rate": 1.9201996356567104e-05,
+      "loss": 0.9355,
+      "step": 2962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.72787100895518,
+      "learning_rate": 1.9201336085857757e-05,
+      "loss": 0.8533,
+      "step": 2963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.437005000967584,
+      "learning_rate": 1.9200675553467793e-05,
+      "loss": 0.7738,
+      "step": 2964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.514193885590164,
+      "learning_rate": 1.9200014759416002e-05,
+      "loss": 0.7954,
+      "step": 2965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.800568750003228,
+      "learning_rate": 1.919935370372117e-05,
+      "loss": 0.8311,
+      "step": 2966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.624515791542912,
+      "learning_rate": 1.91986923864021e-05,
+      "loss": 0.7354,
+      "step": 2967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.092666715845724,
+      "learning_rate": 1.91980308074776e-05,
+      "loss": 0.909,
+      "step": 2968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.732371659651548,
+      "learning_rate": 1.9197368966966486e-05,
+      "loss": 0.8271,
+      "step": 2969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 24.69417819706931,
+      "learning_rate": 1.919670686488758e-05,
+      "loss": 0.7414,
+      "step": 2970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.122450704110175,
+      "learning_rate": 1.9196044501259706e-05,
+      "loss": 0.8145,
+      "step": 2971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.592437767195474,
+      "learning_rate": 1.919538187610171e-05,
+      "loss": 0.7751,
+      "step": 2972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.147962863759687,
+      "learning_rate": 1.9194718989432434e-05,
+      "loss": 0.7556,
+      "step": 2973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.081973025903448,
+      "learning_rate": 1.9194055841270724e-05,
+      "loss": 0.7244,
+      "step": 2974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 22.151120993107924,
+      "learning_rate": 1.9193392431635447e-05,
+      "loss": 0.7764,
+      "step": 2975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.013099177799887,
+      "learning_rate": 1.9192728760545466e-05,
+      "loss": 0.8906,
+      "step": 2976
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.82556209101294,
+      "learning_rate": 1.9192064828019657e-05,
+      "loss": 0.8022,
+      "step": 2977
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 29.16419849113684,
+      "learning_rate": 1.91914006340769e-05,
+      "loss": 0.8661,
+      "step": 2978
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.058282397280657,
+      "learning_rate": 1.919073617873609e-05,
+      "loss": 0.7733,
+      "step": 2979
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.393799984046787,
+      "learning_rate": 1.9190071462016115e-05,
+      "loss": 0.7316,
+      "step": 2980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.690095128816516,
+      "learning_rate": 1.918940648393589e-05,
+      "loss": 0.8779,
+      "step": 2981
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.530499667083717,
+      "learning_rate": 1.9188741244514312e-05,
+      "loss": 0.8052,
+      "step": 2982
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.38641859368205,
+      "learning_rate": 1.9188075743770312e-05,
+      "loss": 0.8937,
+      "step": 2983
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.43112931390468,
+      "learning_rate": 1.9187409981722814e-05,
+      "loss": 0.801,
+      "step": 2984
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.866776106651184,
+      "learning_rate": 1.918674395839075e-05,
+      "loss": 0.7824,
+      "step": 2985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.97578799817641,
+      "learning_rate": 1.9186077673793062e-05,
+      "loss": 0.8566,
+      "step": 2986
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.05341934632158,
+      "learning_rate": 1.9185411127948695e-05,
+      "loss": 0.9179,
+      "step": 2987
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.067672739124912,
+      "learning_rate": 1.9184744320876612e-05,
+      "loss": 0.8418,
+      "step": 2988
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.47557472848811,
+      "learning_rate": 1.9184077252595777e-05,
+      "loss": 0.8147,
+      "step": 2989
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.80401669538672,
+      "learning_rate": 1.918340992312515e-05,
+      "loss": 0.8732,
+      "step": 2990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.328644617154742,
+      "learning_rate": 1.9182742332483722e-05,
+      "loss": 0.715,
+      "step": 2991
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.63828946756317,
+      "learning_rate": 1.9182074480690472e-05,
+      "loss": 0.7943,
+      "step": 2992
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.833982534164743,
+      "learning_rate": 1.9181406367764398e-05,
+      "loss": 0.7129,
+      "step": 2993
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.7322433730387,
+      "learning_rate": 1.9180737993724494e-05,
+      "loss": 0.8075,
+      "step": 2994
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.652519000273282,
+      "learning_rate": 1.9180069358589773e-05,
+      "loss": 0.7774,
+      "step": 2995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.711949560619978,
+      "learning_rate": 1.9179400462379248e-05,
+      "loss": 0.8935,
+      "step": 2996
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.270920943284427,
+      "learning_rate": 1.9178731305111952e-05,
+      "loss": 0.7492,
+      "step": 2997
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.175697308021366,
+      "learning_rate": 1.91780618868069e-05,
+      "loss": 0.8474,
+      "step": 2998
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 24.968055430767055,
+      "learning_rate": 1.9177392207483138e-05,
+      "loss": 0.7235,
+      "step": 2999
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.887994508659474,
+      "learning_rate": 1.9176722267159706e-05,
+      "loss": 0.8417,
+      "step": 3000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.83000504108188,
+      "learning_rate": 1.9176052065855666e-05,
+      "loss": 0.8184,
+      "step": 3001
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.695328402940113,
+      "learning_rate": 1.917538160359007e-05,
+      "loss": 0.8408,
+      "step": 3002
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.982270747971615,
+      "learning_rate": 1.9174710880381995e-05,
+      "loss": 0.8644,
+      "step": 3003
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.912944427613844,
+      "learning_rate": 1.9174039896250505e-05,
+      "loss": 0.7888,
+      "step": 3004
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.032150266921676,
+      "learning_rate": 1.9173368651214686e-05,
+      "loss": 0.7972,
+      "step": 3005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.035552993951598,
+      "learning_rate": 1.917269714529363e-05,
+      "loss": 0.7865,
+      "step": 3006
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.93616797064195,
+      "learning_rate": 1.9172025378506434e-05,
+      "loss": 0.7995,
+      "step": 3007
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.41459353378958,
+      "learning_rate": 1.9171353350872203e-05,
+      "loss": 0.767,
+      "step": 3008
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.700018681023064,
+      "learning_rate": 1.9170681062410042e-05,
+      "loss": 0.7218,
+      "step": 3009
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 31.0410510740856,
+      "learning_rate": 1.917000851313908e-05,
+      "loss": 0.8386,
+      "step": 3010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.574986941661635,
+      "learning_rate": 1.9169335703078443e-05,
+      "loss": 0.757,
+      "step": 3011
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.38841900934235,
+      "learning_rate": 1.916866263224726e-05,
+      "loss": 0.7719,
+      "step": 3012
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.931303081752528,
+      "learning_rate": 1.9167989300664675e-05,
+      "loss": 0.8328,
+      "step": 3013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.136696520403664,
+      "learning_rate": 1.9167315708349835e-05,
+      "loss": 0.8111,
+      "step": 3014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.187525535622676,
+      "learning_rate": 1.9166641855321902e-05,
+      "loss": 0.7314,
+      "step": 3015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.280160694669522,
+      "learning_rate": 1.9165967741600038e-05,
+      "loss": 0.7884,
+      "step": 3016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.690755728793388,
+      "learning_rate": 1.916529336720341e-05,
+      "loss": 0.74,
+      "step": 3017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.09338758509844,
+      "learning_rate": 1.9164618732151202e-05,
+      "loss": 0.835,
+      "step": 3018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 48.7374454853749,
+      "learning_rate": 1.91639438364626e-05,
+      "loss": 0.8324,
+      "step": 3019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 27.207459095194455,
+      "learning_rate": 1.9163268680156793e-05,
+      "loss": 0.879,
+      "step": 3020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 24.994638515091175,
+      "learning_rate": 1.9162593263252988e-05,
+      "loss": 0.7789,
+      "step": 3021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.173834867637698,
+      "learning_rate": 1.916191758577039e-05,
+      "loss": 0.8177,
+      "step": 3022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 78.70199739861276,
+      "learning_rate": 1.9161241647728213e-05,
+      "loss": 1.0178,
+      "step": 3023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.1262693230769,
+      "learning_rate": 1.9160565449145687e-05,
+      "loss": 0.787,
+      "step": 3024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.807728325083367,
+      "learning_rate": 1.9159888990042033e-05,
+      "loss": 0.708,
+      "step": 3025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.274863449463217,
+      "learning_rate": 1.91592122704365e-05,
+      "loss": 0.6912,
+      "step": 3026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.52421657493265,
+      "learning_rate": 1.915853529034832e-05,
+      "loss": 0.8269,
+      "step": 3027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.22484300063159,
+      "learning_rate": 1.915785804979676e-05,
+      "loss": 0.7822,
+      "step": 3028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 29.935498456404957,
+      "learning_rate": 1.915718054880107e-05,
+      "loss": 0.9521,
+      "step": 3029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.4065199819226,
+      "learning_rate": 1.9156502787380527e-05,
+      "loss": 0.7704,
+      "step": 3030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.14438303372769,
+      "learning_rate": 1.91558247655544e-05,
+      "loss": 0.8861,
+      "step": 3031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 188.43611629193143,
+      "learning_rate": 1.9155146483341972e-05,
+      "loss": 0.9285,
+      "step": 3032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.070695464394742,
+      "learning_rate": 1.9154467940762534e-05,
+      "loss": 0.7855,
+      "step": 3033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.187993590781332,
+      "learning_rate": 1.9153789137835384e-05,
+      "loss": 0.7894,
+      "step": 3034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 68.40038764914176,
+      "learning_rate": 1.9153110074579823e-05,
+      "loss": 0.866,
+      "step": 3035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.83281251750027,
+      "learning_rate": 1.915243075101517e-05,
+      "loss": 0.8398,
+      "step": 3036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.808878613981356,
+      "learning_rate": 1.9151751167160734e-05,
+      "loss": 0.7947,
+      "step": 3037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.707452924922112,
+      "learning_rate": 1.9151071323035856e-05,
+      "loss": 0.7885,
+      "step": 3038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.65563795249573,
+      "learning_rate": 1.9150391218659855e-05,
+      "loss": 0.8874,
+      "step": 3039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 32.72912747263712,
+      "learning_rate": 1.9149710854052087e-05,
+      "loss": 0.7335,
+      "step": 3040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.214797116661902,
+      "learning_rate": 1.914903022923189e-05,
+      "loss": 0.8206,
+      "step": 3041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.518177596767742,
+      "learning_rate": 1.914834934421863e-05,
+      "loss": 0.8378,
+      "step": 3042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.781896890045363,
+      "learning_rate": 1.9147668199031664e-05,
+      "loss": 0.8055,
+      "step": 3043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 27.430209166027087,
+      "learning_rate": 1.9146986793690363e-05,
+      "loss": 0.9628,
+      "step": 3044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.027555733919122,
+      "learning_rate": 1.914630512821411e-05,
+      "loss": 0.8761,
+      "step": 3045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.445006547376877,
+      "learning_rate": 1.9145623202622293e-05,
+      "loss": 0.8882,
+      "step": 3046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.351071810860862,
+      "learning_rate": 1.91449410169343e-05,
+      "loss": 0.7529,
+      "step": 3047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.19633968461515,
+      "learning_rate": 1.9144258571169533e-05,
+      "loss": 0.8132,
+      "step": 3048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.80901606358736,
+      "learning_rate": 1.9143575865347405e-05,
+      "loss": 0.9359,
+      "step": 3049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.474570051261475,
+      "learning_rate": 1.9142892899487323e-05,
+      "loss": 0.8576,
+      "step": 3050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.161613733349938,
+      "learning_rate": 1.9142209673608717e-05,
+      "loss": 0.879,
+      "step": 3051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 25.155900803505816,
+      "learning_rate": 1.914152618773102e-05,
+      "loss": 0.8732,
+      "step": 3052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.691004289524503,
+      "learning_rate": 1.914084244187366e-05,
+      "loss": 0.8542,
+      "step": 3053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.733439208562597,
+      "learning_rate": 1.914015843605609e-05,
+      "loss": 0.7865,
+      "step": 3054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 30.483586455117734,
+      "learning_rate": 1.9139474170297764e-05,
+      "loss": 0.8413,
+      "step": 3055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.977388508824916,
+      "learning_rate": 1.9138789644618138e-05,
+      "loss": 0.8357,
+      "step": 3056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.508870975541917,
+      "learning_rate": 1.9138104859036678e-05,
+      "loss": 0.8364,
+      "step": 3057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.79289521148358,
+      "learning_rate": 1.913741981357286e-05,
+      "loss": 0.9342,
+      "step": 3058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.60082752919683,
+      "learning_rate": 1.913673450824617e-05,
+      "loss": 0.7614,
+      "step": 3059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.314914554878236,
+      "learning_rate": 1.9136048943076098e-05,
+      "loss": 0.7982,
+      "step": 3060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.21496649621919,
+      "learning_rate": 1.9135363118082138e-05,
+      "loss": 0.7698,
+      "step": 3061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.68564685432869,
+      "learning_rate": 1.913467703328379e-05,
+      "loss": 0.813,
+      "step": 3062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.238265147154276,
+      "learning_rate": 1.9133990688700578e-05,
+      "loss": 0.7063,
+      "step": 3063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.684032162134578,
+      "learning_rate": 1.9133304084352006e-05,
+      "loss": 0.794,
+      "step": 3064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.306271986791536,
+      "learning_rate": 1.9132617220257613e-05,
+      "loss": 0.8492,
+      "step": 3065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.816758462426602,
+      "learning_rate": 1.913193009643693e-05,
+      "loss": 0.8228,
+      "step": 3066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.44329454946185,
+      "learning_rate": 1.9131242712909497e-05,
+      "loss": 0.913,
+      "step": 3067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.711856183312651,
+      "learning_rate": 1.913055506969486e-05,
+      "loss": 0.8675,
+      "step": 3068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.10911783078787,
+      "learning_rate": 1.9129867166812583e-05,
+      "loss": 0.7506,
+      "step": 3069
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.39252835287759,
+      "learning_rate": 1.9129179004282218e-05,
+      "loss": 0.7542,
+      "step": 3070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.396309603383,
+      "learning_rate": 1.9128490582123346e-05,
+      "loss": 0.8345,
+      "step": 3071
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.322095694118957,
+      "learning_rate": 1.912780190035554e-05,
+      "loss": 0.7505,
+      "step": 3072
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.454810197446502,
+      "learning_rate": 1.9127112958998392e-05,
+      "loss": 0.8345,
+      "step": 3073
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 105.19515645780156,
+      "learning_rate": 1.9126423758071486e-05,
+      "loss": 0.9433,
+      "step": 3074
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.251784971534057,
+      "learning_rate": 1.912573429759443e-05,
+      "loss": 0.863,
+      "step": 3075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.192685568633546,
+      "learning_rate": 1.9125044577586828e-05,
+      "loss": 0.784,
+      "step": 3076
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.968912127543344,
+      "learning_rate": 1.9124354598068294e-05,
+      "loss": 0.8748,
+      "step": 3077
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 26.428298764865872,
+      "learning_rate": 1.912366435905846e-05,
+      "loss": 0.9027,
+      "step": 3078
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.129201116824394,
+      "learning_rate": 1.9122973860576942e-05,
+      "loss": 0.9924,
+      "step": 3079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.870318116325166,
+      "learning_rate": 1.9122283102643383e-05,
+      "loss": 0.7567,
+      "step": 3080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.323786018100822,
+      "learning_rate": 1.9121592085277432e-05,
+      "loss": 0.7717,
+      "step": 3081
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 25.112334402216227,
+      "learning_rate": 1.912090080849874e-05,
+      "loss": 0.8616,
+      "step": 3082
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 25.934243678587745,
+      "learning_rate": 1.9120209272326962e-05,
+      "loss": 0.7796,
+      "step": 3083
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.82672576550508,
+      "learning_rate": 1.9119517476781766e-05,
+      "loss": 0.8885,
+      "step": 3084
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.10484160412152,
+      "learning_rate": 1.911882542188283e-05,
+      "loss": 0.8709,
+      "step": 3085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.074270100007478,
+      "learning_rate": 1.9118133107649833e-05,
+      "loss": 0.756,
+      "step": 3086
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 31.769694056903443,
+      "learning_rate": 1.9117440534102466e-05,
+      "loss": 0.9146,
+      "step": 3087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 31.294511954370197,
+      "learning_rate": 1.911674770126042e-05,
+      "loss": 0.8921,
+      "step": 3088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.86619813370443,
+      "learning_rate": 1.9116054609143408e-05,
+      "loss": 0.8159,
+      "step": 3089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.140514074510172,
+      "learning_rate": 1.9115361257771133e-05,
+      "loss": 0.7716,
+      "step": 3090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.32158532854001,
+      "learning_rate": 1.911466764716331e-05,
+      "loss": 0.9614,
+      "step": 3091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.20811950273685,
+      "learning_rate": 1.911397377733968e-05,
+      "loss": 0.7836,
+      "step": 3092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.772607778353215,
+      "learning_rate": 1.9113279648319964e-05,
+      "loss": 0.7588,
+      "step": 3093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 34.73969914570977,
+      "learning_rate": 1.9112585260123906e-05,
+      "loss": 0.9183,
+      "step": 3094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.482908049519,
+      "learning_rate": 1.9111890612771252e-05,
+      "loss": 0.886,
+      "step": 3095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.53036064214237,
+      "learning_rate": 1.911119570628176e-05,
+      "loss": 0.7775,
+      "step": 3096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.3446514356776,
+      "learning_rate": 1.9110500540675194e-05,
+      "loss": 0.9258,
+      "step": 3097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 26.49974695458246,
+      "learning_rate": 1.910980511597132e-05,
+      "loss": 0.8614,
+      "step": 3098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 27.22658414251488,
+      "learning_rate": 1.910910943218992e-05,
+      "loss": 0.8753,
+      "step": 3099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.928771941239514,
+      "learning_rate": 1.9108413489350775e-05,
+      "loss": 0.8703,
+      "step": 3100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.631568721769806,
+      "learning_rate": 1.910771728747368e-05,
+      "loss": 0.9159,
+      "step": 3101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.876356522433195,
+      "learning_rate": 1.910702082657843e-05,
+      "loss": 0.7536,
+      "step": 3102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.274547291080765,
+      "learning_rate": 1.910632410668484e-05,
+      "loss": 0.6968,
+      "step": 3103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.671944836418826,
+      "learning_rate": 1.9105627127812717e-05,
+      "loss": 0.8034,
+      "step": 3104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.764355759594338,
+      "learning_rate": 1.910492988998188e-05,
+      "loss": 0.8499,
+      "step": 3105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 29.226975306267484,
+      "learning_rate": 1.910423239321217e-05,
+      "loss": 0.8224,
+      "step": 3106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.579159736720895,
+      "learning_rate": 1.9103534637523414e-05,
+      "loss": 0.7808,
+      "step": 3107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.72727288419119,
+      "learning_rate": 1.910283662293546e-05,
+      "loss": 0.8869,
+      "step": 3108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.856564437993306,
+      "learning_rate": 1.9102138349468154e-05,
+      "loss": 0.8029,
+      "step": 3109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.953049028249062,
+      "learning_rate": 1.910143981714136e-05,
+      "loss": 0.8347,
+      "step": 3110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.9489918808129,
+      "learning_rate": 1.9100741025974943e-05,
+      "loss": 0.7513,
+      "step": 3111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 27.66708381117652,
+      "learning_rate": 1.9100041975988776e-05,
+      "loss": 0.8121,
+      "step": 3112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.065126588513714,
+      "learning_rate": 1.9099342667202733e-05,
+      "loss": 0.754,
+      "step": 3113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.67327893729973,
+      "learning_rate": 1.9098643099636714e-05,
+      "loss": 0.7429,
+      "step": 3114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.743642416760856,
+      "learning_rate": 1.9097943273310603e-05,
+      "loss": 0.8474,
+      "step": 3115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.69908849475767,
+      "learning_rate": 1.909724318824431e-05,
+      "loss": 0.8815,
+      "step": 3116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 25.1918363766617,
+      "learning_rate": 1.909654284445774e-05,
+      "loss": 0.8509,
+      "step": 3117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 106.24448763193921,
+      "learning_rate": 1.9095842241970817e-05,
+      "loss": 0.8482,
+      "step": 3118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.94313310015852,
+      "learning_rate": 1.9095141380803457e-05,
+      "loss": 0.7856,
+      "step": 3119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.2955491385221,
+      "learning_rate": 1.90944402609756e-05,
+      "loss": 0.7584,
+      "step": 3120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 27.399234865636398,
+      "learning_rate": 1.909373888250718e-05,
+      "loss": 0.8746,
+      "step": 3121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.065127674726245,
+      "learning_rate": 1.9093037245418147e-05,
+      "loss": 0.8033,
+      "step": 3122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.415516942768967,
+      "learning_rate": 1.909233534972845e-05,
+      "loss": 0.8502,
+      "step": 3123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.13339042803202,
+      "learning_rate": 1.9091633195458062e-05,
+      "loss": 0.82,
+      "step": 3124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.54999807055041,
+      "learning_rate": 1.9090930782626943e-05,
+      "loss": 0.7479,
+      "step": 3125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.961036975028698,
+      "learning_rate": 1.9090228111255066e-05,
+      "loss": 0.8619,
+      "step": 3126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.472341947734442,
+      "learning_rate": 1.908952518136242e-05,
+      "loss": 0.8502,
+      "step": 3127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.592292588739333,
+      "learning_rate": 1.9088821992969e-05,
+      "loss": 0.7517,
+      "step": 3128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.21307371670834,
+      "learning_rate": 1.9088118546094793e-05,
+      "loss": 0.8045,
+      "step": 3129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.981963037825658,
+      "learning_rate": 1.9087414840759812e-05,
+      "loss": 0.8468,
+      "step": 3130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.076005426268907,
+      "learning_rate": 1.9086710876984075e-05,
+      "loss": 0.7823,
+      "step": 3131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.660020628441846,
+      "learning_rate": 1.9086006654787588e-05,
+      "loss": 0.7982,
+      "step": 3132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.416622407082215,
+      "learning_rate": 1.908530217419039e-05,
+      "loss": 0.8989,
+      "step": 3133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.823122805150962,
+      "learning_rate": 1.9084597435212517e-05,
+      "loss": 0.8361,
+      "step": 3134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.627198711174408,
+      "learning_rate": 1.9083892437874004e-05,
+      "loss": 0.8226,
+      "step": 3135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.515210547695123,
+      "learning_rate": 1.90831871821949e-05,
+      "loss": 0.7261,
+      "step": 3136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 27.63891188206161,
+      "learning_rate": 1.9082481668195273e-05,
+      "loss": 0.8063,
+      "step": 3137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.223533788113098,
+      "learning_rate": 1.9081775895895177e-05,
+      "loss": 0.7253,
+      "step": 3138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.209520861701765,
+      "learning_rate": 1.9081069865314688e-05,
+      "loss": 0.7902,
+      "step": 3139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 25.65344581020031,
+      "learning_rate": 1.9080363576473883e-05,
+      "loss": 0.7909,
+      "step": 3140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.616322906054116,
+      "learning_rate": 1.907965702939285e-05,
+      "loss": 0.7744,
+      "step": 3141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.884314015271608,
+      "learning_rate": 1.9078950224091685e-05,
+      "loss": 0.8179,
+      "step": 3142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.443519655416107,
+      "learning_rate": 1.907824316059048e-05,
+      "loss": 0.8592,
+      "step": 3143
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.7428535376474,
+      "learning_rate": 1.9077535838909356e-05,
+      "loss": 0.7896,
+      "step": 3144
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.010459291984223,
+      "learning_rate": 1.9076828259068422e-05,
+      "loss": 0.7939,
+      "step": 3145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.17897469664532,
+      "learning_rate": 1.9076120421087806e-05,
+      "loss": 0.7815,
+      "step": 3146
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.02230568097062,
+      "learning_rate": 1.907541232498763e-05,
+      "loss": 0.9493,
+      "step": 3147
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 23.369242738149158,
+      "learning_rate": 1.9074703970788032e-05,
+      "loss": 0.9246,
+      "step": 3148
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.075400649309046,
+      "learning_rate": 1.907399535850917e-05,
+      "loss": 0.8102,
+      "step": 3149
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.134030874954295,
+      "learning_rate": 1.9073286488171184e-05,
+      "loss": 0.7831,
+      "step": 3150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.20100989319069,
+      "learning_rate": 1.9072577359794237e-05,
+      "loss": 0.797,
+      "step": 3151
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.31817177208678,
+      "learning_rate": 1.9071867973398502e-05,
+      "loss": 0.8151,
+      "step": 3152
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.3416210736141725,
+      "learning_rate": 1.9071158329004144e-05,
+      "loss": 0.8087,
+      "step": 3153
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.61621300103439,
+      "learning_rate": 1.907044842663135e-05,
+      "loss": 0.7402,
+      "step": 3154
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.41108041951863,
+      "learning_rate": 1.906973826630031e-05,
+      "loss": 0.865,
+      "step": 3155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.024465899899187,
+      "learning_rate": 1.906902784803122e-05,
+      "loss": 0.8306,
+      "step": 3156
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 26.99403006507065,
+      "learning_rate": 1.9068317171844284e-05,
+      "loss": 0.9317,
+      "step": 3157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.96428759991952,
+      "learning_rate": 1.9067606237759712e-05,
+      "loss": 0.7189,
+      "step": 3158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.963645590690497,
+      "learning_rate": 1.906689504579772e-05,
+      "loss": 0.8662,
+      "step": 3159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.218922496925103,
+      "learning_rate": 1.906618359597854e-05,
+      "loss": 0.9281,
+      "step": 3160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.79946724003632,
+      "learning_rate": 1.90654718883224e-05,
+      "loss": 0.6847,
+      "step": 3161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.17157078272671,
+      "learning_rate": 1.9064759922849544e-05,
+      "loss": 0.8478,
+      "step": 3162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.78562059796633,
+      "learning_rate": 1.906404769958022e-05,
+      "loss": 0.6997,
+      "step": 3163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.497370269004637,
+      "learning_rate": 1.9063335218534677e-05,
+      "loss": 0.7554,
+      "step": 3164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.649693669473898,
+      "learning_rate": 1.906262247973319e-05,
+      "loss": 0.7674,
+      "step": 3165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.00603353579781,
+      "learning_rate": 1.9061909483196012e-05,
+      "loss": 0.8077,
+      "step": 3166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.96696829319825,
+      "learning_rate": 1.9061196228943436e-05,
+      "loss": 0.9424,
+      "step": 3167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.666303651373918,
+      "learning_rate": 1.906048271699574e-05,
+      "loss": 0.7462,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.56346788169085,
+      "learning_rate": 1.905976894737321e-05,
+      "loss": 0.8058,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.983457982262538,
+      "learning_rate": 1.9059054920096155e-05,
+      "loss": 0.8306,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 22.155661529398557,
+      "learning_rate": 1.9058340635184877e-05,
+      "loss": 0.8405,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.57549470322433,
+      "learning_rate": 1.9057626092659692e-05,
+      "loss": 0.7334,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.827567746896081,
+      "learning_rate": 1.9056911292540915e-05,
+      "loss": 0.7195,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.972028228600916,
+      "learning_rate": 1.905619623484888e-05,
+      "loss": 0.7214,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.866056987188042,
+      "learning_rate": 1.9055480919603924e-05,
+      "loss": 0.7584,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 26.330229548889065,
+      "learning_rate": 1.905476534682639e-05,
+      "loss": 0.8473,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 31.325558457872724,
+      "learning_rate": 1.905404951653662e-05,
+      "loss": 0.8722,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.211834858164256,
+      "learning_rate": 1.9053333428754986e-05,
+      "loss": 0.7664,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.083965391892713,
+      "learning_rate": 1.9052617083501837e-05,
+      "loss": 0.85,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.158817002806252,
+      "learning_rate": 1.905190048079756e-05,
+      "loss": 0.8052,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.372182607235231,
+      "learning_rate": 1.9051183620662526e-05,
+      "loss": 0.7049,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 22.96415398086274,
+      "learning_rate": 1.9050466503117127e-05,
+      "loss": 0.8397,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.250823988311357,
+      "learning_rate": 1.9049749128181752e-05,
+      "loss": 0.9214,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.628738158483294,
+      "learning_rate": 1.904903149587681e-05,
+      "loss": 0.8246,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.580946174331254,
+      "learning_rate": 1.90483136062227e-05,
+      "loss": 0.7727,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.16508975837801,
+      "learning_rate": 1.9047595459239848e-05,
+      "loss": 0.7698,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.30438426938258,
+      "learning_rate": 1.904687705494867e-05,
+      "loss": 0.7347,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.562595660232425,
+      "learning_rate": 1.9046158393369608e-05,
+      "loss": 0.9778,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.90905310314675,
+      "learning_rate": 1.9045439474523086e-05,
+      "loss": 0.8563,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.126788763602624,
+      "learning_rate": 1.904472029842956e-05,
+      "loss": 0.7797,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 26.624222759758187,
+      "learning_rate": 1.904400086510948e-05,
+      "loss": 0.8578,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 24.69015382433746,
+      "learning_rate": 1.9043281174583305e-05,
+      "loss": 0.7512,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 28.265565583233567,
+      "learning_rate": 1.9042561226871506e-05,
+      "loss": 0.8479,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.815323731396376,
+      "learning_rate": 1.9041841021994552e-05,
+      "loss": 0.7862,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.924109704560887,
+      "learning_rate": 1.9041120559972927e-05,
+      "loss": 0.7745,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.877079630596655,
+      "learning_rate": 1.904039984082713e-05,
+      "loss": 0.8712,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.77250016992269,
+      "learning_rate": 1.9039678864577642e-05,
+      "loss": 0.8153,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 27.76538514853176,
+      "learning_rate": 1.903895763124498e-05,
+      "loss": 0.8957,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.117283871505574,
+      "learning_rate": 1.9038236140849648e-05,
+      "loss": 0.8031,
+      "step": 3199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.047662480005094,
+      "learning_rate": 1.9037514393412167e-05,
+      "loss": 0.8145,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.283046792237744,
+      "learning_rate": 1.9036792388953066e-05,
+      "loss": 0.7254,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.19781966004275,
+      "learning_rate": 1.9036070127492875e-05,
+      "loss": 0.76,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.17193392088283,
+      "learning_rate": 1.903534760905213e-05,
+      "loss": 0.8017,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.10453323656041,
+      "learning_rate": 1.9034624833651393e-05,
+      "loss": 0.7997,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.030486963000833,
+      "learning_rate": 1.9033901801311207e-05,
+      "loss": 0.8689,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.46829393479594,
+      "learning_rate": 1.903317851205214e-05,
+      "loss": 0.839,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.5378164201416,
+      "learning_rate": 1.903245496589476e-05,
+      "loss": 0.7806,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 22.730279853850114,
+      "learning_rate": 1.9031731162859642e-05,
+      "loss": 0.8162,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.84158372661201,
+      "learning_rate": 1.9031007102967375e-05,
+      "loss": 0.7447,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 26.09868887228873,
+      "learning_rate": 1.903028278623855e-05,
+      "loss": 0.888,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.788062582253868,
+      "learning_rate": 1.9029558212693764e-05,
+      "loss": 0.7907,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.156128065789513,
+      "learning_rate": 1.9028833382353624e-05,
+      "loss": 0.8118,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.067312722915801,
+      "learning_rate": 1.9028108295238745e-05,
+      "loss": 0.7429,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.730143143699177,
+      "learning_rate": 1.902738295136975e-05,
+      "loss": 0.7797,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.737672631233288,
+      "learning_rate": 1.902665735076726e-05,
+      "loss": 0.7292,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.47696348750332,
+      "learning_rate": 1.9025931493451917e-05,
+      "loss": 0.8176,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.45868765326416,
+      "learning_rate": 1.9025205379444362e-05,
+      "loss": 0.7827,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.305542756219992,
+      "learning_rate": 1.9024479008765246e-05,
+      "loss": 0.8556,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 22.36812396126054,
+      "learning_rate": 1.9023752381435227e-05,
+      "loss": 0.7753,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.08174138604477,
+      "learning_rate": 1.9023025497474968e-05,
+      "loss": 0.7337,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 42.66182033372968,
+      "learning_rate": 1.9022298356905146e-05,
+      "loss": 0.9066,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.446188973439455,
+      "learning_rate": 1.902157095974643e-05,
+      "loss": 0.8542,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 45.87349920130624,
+      "learning_rate": 1.902084330601952e-05,
+      "loss": 0.854,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.09499898530948,
+      "learning_rate": 1.9020115395745098e-05,
+      "loss": 0.7287,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.85777698342505,
+      "learning_rate": 1.9019387228943872e-05,
+      "loss": 0.8088,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.911027894602718,
+      "learning_rate": 1.9018658805636553e-05,
+      "loss": 0.8078,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.314574331483886,
+      "learning_rate": 1.901793012584385e-05,
+      "loss": 0.7852,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.346102203471863,
+      "learning_rate": 1.901720118958649e-05,
+      "loss": 0.8695,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 27.236645349492782,
+      "learning_rate": 1.9016471996885202e-05,
+      "loss": 0.839,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.432241548763045,
+      "learning_rate": 1.9015742547760726e-05,
+      "loss": 0.9157,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 24.47201569275253,
+      "learning_rate": 1.9015012842233807e-05,
+      "loss": 0.8392,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.995745037065667,
+      "learning_rate": 1.9014282880325194e-05,
+      "loss": 0.7897,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.31086148008573,
+      "learning_rate": 1.9013552662055652e-05,
+      "loss": 0.859,
+      "step": 3233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.7200266199151,
+      "learning_rate": 1.9012822187445944e-05,
+      "loss": 0.7593,
+      "step": 3234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.441460770125662,
+      "learning_rate": 1.9012091456516843e-05,
+      "loss": 0.8184,
+      "step": 3235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 23.760640379030786,
+      "learning_rate": 1.9011360469289138e-05,
+      "loss": 0.9675,
+      "step": 3236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.91864224073,
+      "learning_rate": 1.901062922578361e-05,
+      "loss": 0.7626,
+      "step": 3237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.97133714710073,
+      "learning_rate": 1.9009897726021058e-05,
+      "loss": 0.8719,
+      "step": 3238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.490950684369828,
+      "learning_rate": 1.9009165970022282e-05,
+      "loss": 0.7341,
+      "step": 3239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.52146375768583,
+      "learning_rate": 1.9008433957808102e-05,
+      "loss": 0.7873,
+      "step": 3240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.896604934104056,
+      "learning_rate": 1.9007701689399323e-05,
+      "loss": 0.7226,
+      "step": 3241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.72922314743194,
+      "learning_rate": 1.9006969164816784e-05,
+      "loss": 0.7948,
+      "step": 3242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.929676465637428,
+      "learning_rate": 1.9006236384081306e-05,
+      "loss": 0.7807,
+      "step": 3243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.201355750938077,
+      "learning_rate": 1.9005503347213738e-05,
+      "loss": 0.8165,
+      "step": 3244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 29.4712549006632,
+      "learning_rate": 1.900477005423492e-05,
+      "loss": 0.852,
+      "step": 3245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.618645060894476,
+      "learning_rate": 1.9004036505165708e-05,
+      "loss": 1.0108,
+      "step": 3246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.086511241966033,
+      "learning_rate": 1.9003302700026968e-05,
+      "loss": 0.6943,
+      "step": 3247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.071567673811153,
+      "learning_rate": 1.9002568638839566e-05,
+      "loss": 0.8185,
+      "step": 3248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 33.88515531423689,
+      "learning_rate": 1.9001834321624378e-05,
+      "loss": 0.9174,
+      "step": 3249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.64538851824194,
+      "learning_rate": 1.9001099748402287e-05,
+      "loss": 0.7773,
+      "step": 3250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.382574772326507,
+      "learning_rate": 1.9000364919194185e-05,
+      "loss": 0.7932,
+      "step": 3251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 36.10558050811683,
+      "learning_rate": 1.899962983402097e-05,
+      "loss": 0.9106,
+      "step": 3252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.984378653824976,
+      "learning_rate": 1.8998894492903545e-05,
+      "loss": 0.8735,
+      "step": 3253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.14826342527058,
+      "learning_rate": 1.8998158895862827e-05,
+      "loss": 0.7821,
+      "step": 3254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.4967278527968,
+      "learning_rate": 1.8997423042919734e-05,
+      "loss": 0.7565,
+      "step": 3255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.466214145418185,
+      "learning_rate": 1.899668693409519e-05,
+      "loss": 0.7963,
+      "step": 3256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 26.11605946292533,
+      "learning_rate": 1.8995950569410136e-05,
+      "loss": 0.9111,
+      "step": 3257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.69279396543573,
+      "learning_rate": 1.8995213948885508e-05,
+      "loss": 0.7317,
+      "step": 3258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.875436113216862,
+      "learning_rate": 1.899447707254226e-05,
+      "loss": 0.7685,
+      "step": 3259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.43809216700467,
+      "learning_rate": 1.8993739940401342e-05,
+      "loss": 0.7964,
+      "step": 3260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 47.01185875219237,
+      "learning_rate": 1.8993002552483726e-05,
+      "loss": 0.8537,
+      "step": 3261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.80303413004833,
+      "learning_rate": 1.8992264908810373e-05,
+      "loss": 0.7608,
+      "step": 3262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.31721858616252,
+      "learning_rate": 1.899152700940227e-05,
+      "loss": 0.7321,
+      "step": 3263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.495838070581346,
+      "learning_rate": 1.8990788854280397e-05,
+      "loss": 0.7526,
+      "step": 3264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.489194545562345,
+      "learning_rate": 1.899005044346575e-05,
+      "loss": 0.7448,
+      "step": 3265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.4744470208579,
+      "learning_rate": 1.8989311776979326e-05,
+      "loss": 0.8306,
+      "step": 3266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.526798336522805,
+      "learning_rate": 1.8988572854842133e-05,
+      "loss": 0.7805,
+      "step": 3267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.05069596271059,
+      "learning_rate": 1.898783367707519e-05,
+      "loss": 0.8492,
+      "step": 3268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.82385227201902,
+      "learning_rate": 1.898709424369951e-05,
+      "loss": 0.7442,
+      "step": 3269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.30131590781488,
+      "learning_rate": 1.898635455473613e-05,
+      "loss": 0.8237,
+      "step": 3270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.42282232759359,
+      "learning_rate": 1.898561461020608e-05,
+      "loss": 0.8917,
+      "step": 3271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.499431660115283,
+      "learning_rate": 1.8984874410130413e-05,
+      "loss": 0.7379,
+      "step": 3272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.93915737921503,
+      "learning_rate": 1.898413395453017e-05,
+      "loss": 0.8256,
+      "step": 3273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.358397824315356,
+      "learning_rate": 1.8983393243426413e-05,
+      "loss": 0.8956,
+      "step": 3274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.31304296526907,
+      "learning_rate": 1.8982652276840205e-05,
+      "loss": 0.726,
+      "step": 3275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.093361744468556,
+      "learning_rate": 1.8981911054792625e-05,
+      "loss": 0.8032,
+      "step": 3276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 36.918961722001136,
+      "learning_rate": 1.8981169577304747e-05,
+      "loss": 0.7639,
+      "step": 3277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.241479004072122,
+      "learning_rate": 1.898042784439766e-05,
+      "loss": 0.8295,
+      "step": 3278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.829385569119307,
+      "learning_rate": 1.897968585609246e-05,
+      "loss": 0.8646,
+      "step": 3279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 25.805204674449193,
+      "learning_rate": 1.8978943612410245e-05,
+      "loss": 0.8207,
+      "step": 3280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.210242630062517,
+      "learning_rate": 1.8978201113372128e-05,
+      "loss": 0.7854,
+      "step": 3281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.48862852748904,
+      "learning_rate": 1.8977458358999222e-05,
+      "loss": 0.7114,
+      "step": 3282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.770164608552005,
+      "learning_rate": 1.8976715349312652e-05,
+      "loss": 0.8092,
+      "step": 3283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.67252407677198,
+      "learning_rate": 1.897597208433355e-05,
+      "loss": 0.7627,
+      "step": 3284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.654441019912475,
+      "learning_rate": 1.8975228564083052e-05,
+      "loss": 0.8817,
+      "step": 3285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.432568941648473,
+      "learning_rate": 1.8974484788582303e-05,
+      "loss": 0.8926,
+      "step": 3286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.773174942927838,
+      "learning_rate": 1.897374075785246e-05,
+      "loss": 0.7816,
+      "step": 3287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.417549328863489,
+      "learning_rate": 1.8972996471914674e-05,
+      "loss": 0.8842,
+      "step": 3288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.497277064876052,
+      "learning_rate": 1.8972251930790124e-05,
+      "loss": 0.7477,
+      "step": 3289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.686390993504194,
+      "learning_rate": 1.8971507134499974e-05,
+      "loss": 0.7553,
+      "step": 3290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 33.51949868035587,
+      "learning_rate": 1.8970762083065408e-05,
+      "loss": 0.981,
+      "step": 3291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.377263269082434,
+      "learning_rate": 1.897001677650762e-05,
+      "loss": 0.8271,
+      "step": 3292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.536138019612448,
+      "learning_rate": 1.8969271214847795e-05,
+      "loss": 0.7148,
+      "step": 3293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.923094759355415,
+      "learning_rate": 1.896852539810715e-05,
+      "loss": 0.8304,
+      "step": 3294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.894773826650535,
+      "learning_rate": 1.8967779326306884e-05,
+      "loss": 0.9322,
+      "step": 3295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.304494596696205,
+      "learning_rate": 1.8967032999468226e-05,
+      "loss": 0.7831,
+      "step": 3296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.043791583152023,
+      "learning_rate": 1.8966286417612393e-05,
+      "loss": 0.8117,
+      "step": 3297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.07606661292513,
+      "learning_rate": 1.8965539580760615e-05,
+      "loss": 0.8522,
+      "step": 3298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.131179547515075,
+      "learning_rate": 1.8964792488934143e-05,
+      "loss": 0.8286,
+      "step": 3299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.86586056611989,
+      "learning_rate": 1.8964045142154212e-05,
+      "loss": 0.8386,
+      "step": 3300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.713210022562002,
+      "learning_rate": 1.896329754044208e-05,
+      "loss": 0.896,
+      "step": 3301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.974547955524066,
+      "learning_rate": 1.8962549683819013e-05,
+      "loss": 0.8399,
+      "step": 3302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.41597401849312,
+      "learning_rate": 1.8961801572306276e-05,
+      "loss": 0.8076,
+      "step": 3303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.818217938092527,
+      "learning_rate": 1.8961053205925143e-05,
+      "loss": 0.7319,
+      "step": 3304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.541694103242634,
+      "learning_rate": 1.89603045846969e-05,
+      "loss": 0.7611,
+      "step": 3305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.173817859470883,
+      "learning_rate": 1.8959555708642835e-05,
+      "loss": 0.8252,
+      "step": 3306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.150288053493515,
+      "learning_rate": 1.895880657778425e-05,
+      "loss": 0.7268,
+      "step": 3307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.08003356690973,
+      "learning_rate": 1.895805719214244e-05,
+      "loss": 0.7844,
+      "step": 3308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.322439721785628,
+      "learning_rate": 1.8957307551738727e-05,
+      "loss": 0.926,
+      "step": 3309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.02751792682329,
+      "learning_rate": 1.895655765659443e-05,
+      "loss": 0.7907,
+      "step": 3310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 33.30861495607743,
+      "learning_rate": 1.8955807506730872e-05,
+      "loss": 0.804,
+      "step": 3311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.57224672039511,
+      "learning_rate": 1.8955057102169385e-05,
+      "loss": 0.7755,
+      "step": 3312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.243166927249245,
+      "learning_rate": 1.8954306442931315e-05,
+      "loss": 0.8541,
+      "step": 3313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.078610801744933,
+      "learning_rate": 1.8953555529038006e-05,
+      "loss": 0.8728,
+      "step": 3314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.274344722894911,
+      "learning_rate": 1.8952804360510816e-05,
+      "loss": 0.8128,
+      "step": 3315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.425515175977296,
+      "learning_rate": 1.8952052937371105e-05,
+      "loss": 0.8284,
+      "step": 3316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 32.74106691373,
+      "learning_rate": 1.8951301259640252e-05,
+      "loss": 0.8772,
+      "step": 3317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.996978524597257,
+      "learning_rate": 1.895054932733962e-05,
+      "loss": 0.8976,
+      "step": 3318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 30.3681125249725,
+      "learning_rate": 1.8949797140490607e-05,
+      "loss": 0.9461,
+      "step": 3319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.100203911570139,
+      "learning_rate": 1.8949044699114592e-05,
+      "loss": 0.7677,
+      "step": 3320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.378046988326368,
+      "learning_rate": 1.8948292003232987e-05,
+      "loss": 0.9179,
+      "step": 3321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 22.65943203810927,
+      "learning_rate": 1.894753905286719e-05,
+      "loss": 0.8169,
+      "step": 3322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.6908656602609,
+      "learning_rate": 1.8946785848038614e-05,
+      "loss": 0.9243,
+      "step": 3323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.642141462670683,
+      "learning_rate": 1.8946032388768687e-05,
+      "loss": 0.8762,
+      "step": 3324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.503034261356934,
+      "learning_rate": 1.8945278675078828e-05,
+      "loss": 0.8297,
+      "step": 3325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.890176946029367,
+      "learning_rate": 1.894452470699048e-05,
+      "loss": 0.817,
+      "step": 3326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.339149383852504,
+      "learning_rate": 1.894377048452508e-05,
+      "loss": 0.9539,
+      "step": 3327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.530667503206878,
+      "learning_rate": 1.8943016007704078e-05,
+      "loss": 0.8668,
+      "step": 3328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.460202050300555,
+      "learning_rate": 1.8942261276548932e-05,
+      "loss": 0.8424,
+      "step": 3329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.204140872949036,
+      "learning_rate": 1.894150629108111e-05,
+      "loss": 0.8521,
+      "step": 3330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.107550123649787,
+      "learning_rate": 1.8940751051322075e-05,
+      "loss": 0.7128,
+      "step": 3331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.54912264364053,
+      "learning_rate": 1.8939995557293315e-05,
+      "loss": 0.7438,
+      "step": 3332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.289986236868678,
+      "learning_rate": 1.8939239809016306e-05,
+      "loss": 0.7825,
+      "step": 3333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.173740732385943,
+      "learning_rate": 1.893848380651255e-05,
+      "loss": 0.8543,
+      "step": 3334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.182607373474102,
+      "learning_rate": 1.893772754980354e-05,
+      "loss": 0.7827,
+      "step": 3335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.840961295812292,
+      "learning_rate": 1.8936971038910792e-05,
+      "loss": 0.8808,
+      "step": 3336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.335262489328013,
+      "learning_rate": 1.8936214273855813e-05,
+      "loss": 0.8171,
+      "step": 3337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.641496398805156,
+      "learning_rate": 1.8935457254660128e-05,
+      "loss": 0.8496,
+      "step": 3338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.289102779439688,
+      "learning_rate": 1.8934699981345265e-05,
+      "loss": 0.806,
+      "step": 3339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.84246254797321,
+      "learning_rate": 1.893394245393276e-05,
+      "loss": 0.8426,
+      "step": 3340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.598108331705184,
+      "learning_rate": 1.893318467244416e-05,
+      "loss": 0.8253,
+      "step": 3341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.54094295632851,
+      "learning_rate": 1.893242663690101e-05,
+      "loss": 0.8221,
+      "step": 3342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.41202666202375,
+      "learning_rate": 1.8931668347324877e-05,
+      "loss": 0.887,
+      "step": 3343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.909508859381846,
+      "learning_rate": 1.893090980373732e-05,
+      "loss": 0.7678,
+      "step": 3344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.05664375927195,
+      "learning_rate": 1.8930151006159908e-05,
+      "loss": 0.7798,
+      "step": 3345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.522019220476068,
+      "learning_rate": 1.892939195461423e-05,
+      "loss": 0.7704,
+      "step": 3346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.00418295928479,
+      "learning_rate": 1.8928632649121867e-05,
+      "loss": 0.8342,
+      "step": 3347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 30.24032633658489,
+      "learning_rate": 1.8927873089704416e-05,
+      "loss": 0.8104,
+      "step": 3348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 26.553372156383084,
+      "learning_rate": 1.8927113276383475e-05,
+      "loss": 0.8248,
+      "step": 3349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.53241335052198,
+      "learning_rate": 1.8926353209180655e-05,
+      "loss": 0.826,
+      "step": 3350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.55322187838542,
+      "learning_rate": 1.892559288811757e-05,
+      "loss": 0.8026,
+      "step": 3351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 26.519803613153854,
+      "learning_rate": 1.892483231321585e-05,
+      "loss": 0.9084,
+      "step": 3352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.11096696335605,
+      "learning_rate": 1.8924071484497114e-05,
+      "loss": 0.8003,
+      "step": 3353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.335539141926084,
+      "learning_rate": 1.892331040198301e-05,
+      "loss": 0.8166,
+      "step": 3354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.240287257457208,
+      "learning_rate": 1.8922549065695172e-05,
+      "loss": 0.7634,
+      "step": 3355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.75685767574243,
+      "learning_rate": 1.8921787475655265e-05,
+      "loss": 0.9598,
+      "step": 3356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.757156519389437,
+      "learning_rate": 1.8921025631884938e-05,
+      "loss": 0.8711,
+      "step": 3357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.124454988209056,
+      "learning_rate": 1.8920263534405858e-05,
+      "loss": 0.7593,
+      "step": 3358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.969145722440434,
+      "learning_rate": 1.8919501183239705e-05,
+      "loss": 0.8531,
+      "step": 3359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.07152944592918,
+      "learning_rate": 1.8918738578408157e-05,
+      "loss": 0.7959,
+      "step": 3360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.900326921630372,
+      "learning_rate": 1.8917975719932898e-05,
+      "loss": 0.7134,
+      "step": 3361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.994220511945322,
+      "learning_rate": 1.891721260783563e-05,
+      "loss": 0.7959,
+      "step": 3362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.25241374432694,
+      "learning_rate": 1.8916449242138047e-05,
+      "loss": 0.8172,
+      "step": 3363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.73083024684305,
+      "learning_rate": 1.8915685622861864e-05,
+      "loss": 0.8052,
+      "step": 3364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.64199132380459,
+      "learning_rate": 1.8914921750028804e-05,
+      "loss": 0.8448,
+      "step": 3365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.664915251452088,
+      "learning_rate": 1.891415762366058e-05,
+      "loss": 0.7772,
+      "step": 3366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 22.610154367840074,
+      "learning_rate": 1.8913393243778927e-05,
+      "loss": 0.7887,
+      "step": 3367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.32555711475076,
+      "learning_rate": 1.8912628610405585e-05,
+      "loss": 0.8175,
+      "step": 3368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.45858352462135,
+      "learning_rate": 1.89118637235623e-05,
+      "loss": 0.7907,
+      "step": 3369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.87556641526553,
+      "learning_rate": 1.891109858327082e-05,
+      "loss": 0.9231,
+      "step": 3370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.770753878588453,
+      "learning_rate": 1.8910333189552913e-05,
+      "loss": 0.6939,
+      "step": 3371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.903194134884835,
+      "learning_rate": 1.8909567542430344e-05,
+      "loss": 0.7623,
+      "step": 3372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.033742888282342,
+      "learning_rate": 1.890880164192488e-05,
+      "loss": 0.7945,
+      "step": 3373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.502029831964597,
+      "learning_rate": 1.8908035488058316e-05,
+      "loss": 0.7376,
+      "step": 3374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.609909045149866,
+      "learning_rate": 1.8907269080852432e-05,
+      "loss": 0.8078,
+      "step": 3375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.357426894513708,
+      "learning_rate": 1.8906502420329022e-05,
+      "loss": 0.7522,
+      "step": 3376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.214632161433357,
+      "learning_rate": 1.8905735506509897e-05,
+      "loss": 0.8444,
+      "step": 3377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.704193312769636,
+      "learning_rate": 1.890496833941686e-05,
+      "loss": 0.8611,
+      "step": 3378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 28.767409683180908,
+      "learning_rate": 1.890420091907174e-05,
+      "loss": 0.8217,
+      "step": 3379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.99141770017904,
+      "learning_rate": 1.8903433245496348e-05,
+      "loss": 0.736,
+      "step": 3380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 24.66708461176765,
+      "learning_rate": 1.8902665318712527e-05,
+      "loss": 0.8313,
+      "step": 3381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.466920263218196,
+      "learning_rate": 1.8901897138742107e-05,
+      "loss": 0.8133,
+      "step": 3382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.826265505340814,
+      "learning_rate": 1.890112870560694e-05,
+      "loss": 0.8638,
+      "step": 3383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.936165660345143,
+      "learning_rate": 1.8900360019328885e-05,
+      "loss": 0.9087,
+      "step": 3384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.720476132451292,
+      "learning_rate": 1.8899591079929796e-05,
+      "loss": 0.7638,
+      "step": 3385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 32.02774678123373,
+      "learning_rate": 1.8898821887431543e-05,
+      "loss": 0.8345,
+      "step": 3386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.123609201016286,
+      "learning_rate": 1.8898052441855997e-05,
+      "loss": 0.7579,
+      "step": 3387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.596070170489416,
+      "learning_rate": 1.8897282743225048e-05,
+      "loss": 0.7905,
+      "step": 3388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.662250099283213,
+      "learning_rate": 1.8896512791560584e-05,
+      "loss": 0.761,
+      "step": 3389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 27.909954965665126,
+      "learning_rate": 1.8895742586884502e-05,
+      "loss": 0.8593,
+      "step": 3390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.3871593925754,
+      "learning_rate": 1.88949721292187e-05,
+      "loss": 0.6885,
+      "step": 3391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.06501881648598,
+      "learning_rate": 1.8894201418585094e-05,
+      "loss": 0.9004,
+      "step": 3392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.926497168169508,
+      "learning_rate": 1.8893430455005604e-05,
+      "loss": 0.8322,
+      "step": 3393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.99428544627486,
+      "learning_rate": 1.8892659238502154e-05,
+      "loss": 0.8215,
+      "step": 3394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.148237302133445,
+      "learning_rate": 1.889188776909668e-05,
+      "loss": 0.7444,
+      "step": 3395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.47423651160239,
+      "learning_rate": 1.8891116046811116e-05,
+      "loss": 0.8691,
+      "step": 3396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.444399667196294,
+      "learning_rate": 1.8890344071667415e-05,
+      "loss": 0.8061,
+      "step": 3397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.23073772972432,
+      "learning_rate": 1.888957184368753e-05,
+      "loss": 0.7917,
+      "step": 3398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.70512289607133,
+      "learning_rate": 1.888879936289342e-05,
+      "loss": 0.8484,
+      "step": 3399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.84407015806765,
+      "learning_rate": 1.8888026629307056e-05,
+      "loss": 0.815,
+      "step": 3400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.48505899279683,
+      "learning_rate": 1.8887253642950414e-05,
+      "loss": 0.6795,
+      "step": 3401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.848641319132316,
+      "learning_rate": 1.8886480403845477e-05,
+      "loss": 0.8493,
+      "step": 3402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.291422213017542,
+      "learning_rate": 1.888570691201424e-05,
+      "loss": 0.7337,
+      "step": 3403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.811800274651663,
+      "learning_rate": 1.8884933167478688e-05,
+      "loss": 0.8441,
+      "step": 3404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.537586278207844,
+      "learning_rate": 1.888415917026084e-05,
+      "loss": 0.7338,
+      "step": 3405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.878420723399753,
+      "learning_rate": 1.8883384920382703e-05,
+      "loss": 0.8015,
+      "step": 3406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.533743463410543,
+      "learning_rate": 1.8882610417866295e-05,
+      "loss": 0.7818,
+      "step": 3407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 27.510857839030827,
+      "learning_rate": 1.888183566273364e-05,
+      "loss": 0.8128,
+      "step": 3408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.53714177437603,
+      "learning_rate": 1.888106065500678e-05,
+      "loss": 0.7775,
+      "step": 3409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.42868004840664,
+      "learning_rate": 1.8880285394707743e-05,
+      "loss": 0.7933,
+      "step": 3410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.942344776774117,
+      "learning_rate": 1.8879509881858587e-05,
+      "loss": 0.8449,
+      "step": 3411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.97399866407785,
+      "learning_rate": 1.8878734116481364e-05,
+      "loss": 0.7983,
+      "step": 3412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.976028293476137,
+      "learning_rate": 1.887795809859814e-05,
+      "loss": 0.8435,
+      "step": 3413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.749414522413606,
+      "learning_rate": 1.8877181828230978e-05,
+      "loss": 0.803,
+      "step": 3414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.371180048696683,
+      "learning_rate": 1.887640530540196e-05,
+      "loss": 0.7834,
+      "step": 3415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.418450622973566,
+      "learning_rate": 1.8875628530133164e-05,
+      "loss": 0.8457,
+      "step": 3416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.53891280145962,
+      "learning_rate": 1.887485150244669e-05,
+      "loss": 0.9259,
+      "step": 3417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.1538160886726,
+      "learning_rate": 1.8874074222364626e-05,
+      "loss": 0.7541,
+      "step": 3418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.298196457700744,
+      "learning_rate": 1.8873296689909084e-05,
+      "loss": 0.8898,
+      "step": 3419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.409744392510575,
+      "learning_rate": 1.8872518905102177e-05,
+      "loss": 0.8425,
+      "step": 3420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.777615546988404,
+      "learning_rate": 1.887174086796602e-05,
+      "loss": 0.8014,
+      "step": 3421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.753797718660437,
+      "learning_rate": 1.8870962578522746e-05,
+      "loss": 0.7904,
+      "step": 3422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.093606963144772,
+      "learning_rate": 1.887018403679448e-05,
+      "loss": 0.8219,
+      "step": 3423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.201060743220935,
+      "learning_rate": 1.8869405242803373e-05,
+      "loss": 0.7504,
+      "step": 3424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 31.126851586435176,
+      "learning_rate": 1.886862619657157e-05,
+      "loss": 0.9006,
+      "step": 3425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.103589940691286,
+      "learning_rate": 1.8867846898121223e-05,
+      "loss": 0.9261,
+      "step": 3426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.898382804457848,
+      "learning_rate": 1.8867067347474498e-05,
+      "loss": 0.8594,
+      "step": 3427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.037837845812941,
+      "learning_rate": 1.8866287544653566e-05,
+      "loss": 0.7686,
+      "step": 3428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 24.176376983851878,
+      "learning_rate": 1.8865507489680602e-05,
+      "loss": 0.9257,
+      "step": 3429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.022398017057679,
+      "learning_rate": 1.8864727182577792e-05,
+      "loss": 0.8068,
+      "step": 3430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.161244457387706,
+      "learning_rate": 1.8863946623367324e-05,
+      "loss": 0.7708,
+      "step": 3431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.072629002916553,
+      "learning_rate": 1.88631658120714e-05,
+      "loss": 0.7199,
+      "step": 3432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.8485785428595,
+      "learning_rate": 1.886238474871223e-05,
+      "loss": 0.8088,
+      "step": 3433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.682636580650488,
+      "learning_rate": 1.8861603433312017e-05,
+      "loss": 0.8052,
+      "step": 3434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.554400508601866,
+      "learning_rate": 1.8860821865892988e-05,
+      "loss": 0.7221,
+      "step": 3435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.396140113928965,
+      "learning_rate": 1.8860040046477367e-05,
+      "loss": 0.7557,
+      "step": 3436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.612653212013303,
+      "learning_rate": 1.8859257975087395e-05,
+      "loss": 0.7959,
+      "step": 3437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.94524016854678,
+      "learning_rate": 1.8858475651745304e-05,
+      "loss": 0.8156,
+      "step": 3438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.27030400103325,
+      "learning_rate": 1.8857693076473348e-05,
+      "loss": 0.7762,
+      "step": 3439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.113184566678713,
+      "learning_rate": 1.8856910249293783e-05,
+      "loss": 0.7738,
+      "step": 3440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 21.12987140278447,
+      "learning_rate": 1.8856127170228873e-05,
+      "loss": 0.8317,
+      "step": 3441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.703873130546501,
+      "learning_rate": 1.8855343839300885e-05,
+      "loss": 0.7512,
+      "step": 3442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 21.85220304606607,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.7905,
+      "step": 3443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 20.6116934174334,
+      "learning_rate": 1.88537764219448e-05,
+      "loss": 0.7489,
+      "step": 3444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.650546684540593,
+      "learning_rate": 1.885299233556128e-05,
+      "loss": 0.7357,
+      "step": 3445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.8245985911209,
+      "learning_rate": 1.8852207997403835e-05,
+      "loss": 0.7486,
+      "step": 3446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.30396428647669,
+      "learning_rate": 1.8851423407494774e-05,
+      "loss": 0.7502,
+      "step": 3447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.147550420567963,
+      "learning_rate": 1.8850638565856406e-05,
+      "loss": 0.7028,
+      "step": 3448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 25.946758679084766,
+      "learning_rate": 1.8849853472511057e-05,
+      "loss": 0.7767,
+      "step": 3449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 20.138954583112035,
+      "learning_rate": 1.8849068127481055e-05,
+      "loss": 0.7856,
+      "step": 3450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.999988545539324,
+      "learning_rate": 1.884828253078873e-05,
+      "loss": 0.808,
+      "step": 3451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.264519551217536,
+      "learning_rate": 1.8847496682456422e-05,
+      "loss": 0.8131,
+      "step": 3452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.128320698075523,
+      "learning_rate": 1.884671058250649e-05,
+      "loss": 0.7629,
+      "step": 3453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.566536444575174,
+      "learning_rate": 1.8845924230961278e-05,
+      "loss": 0.8165,
+      "step": 3454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.766168226097186,
+      "learning_rate": 1.884513762784316e-05,
+      "loss": 0.8274,
+      "step": 3455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 28.15575892515542,
+      "learning_rate": 1.8844350773174502e-05,
+      "loss": 0.7508,
+      "step": 3456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.40047403957846,
+      "learning_rate": 1.884356366697768e-05,
+      "loss": 0.8564,
+      "step": 3457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.091814960256134,
+      "learning_rate": 1.8842776309275083e-05,
+      "loss": 0.7324,
+      "step": 3458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.41725716508165,
+      "learning_rate": 1.8841988700089096e-05,
+      "loss": 0.8842,
+      "step": 3459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 20.600226610318046,
+      "learning_rate": 1.8841200839442128e-05,
+      "loss": 0.8848,
+      "step": 3460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.44865655447884,
+      "learning_rate": 1.8840412727356576e-05,
+      "loss": 0.776,
+      "step": 3461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 27.0947566516384,
+      "learning_rate": 1.883962436385486e-05,
+      "loss": 0.7059,
+      "step": 3462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 34.218234190689394,
+      "learning_rate": 1.88388357489594e-05,
+      "loss": 0.8564,
+      "step": 3463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.261275124625653,
+      "learning_rate": 1.883804688269262e-05,
+      "loss": 0.8148,
+      "step": 3464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.006673130392,
+      "learning_rate": 1.8837257765076956e-05,
+      "loss": 0.8582,
+      "step": 3465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 21.97831617132807,
+      "learning_rate": 1.883646839613485e-05,
+      "loss": 0.8575,
+      "step": 3466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.531856205878626,
+      "learning_rate": 1.8835678775888752e-05,
+      "loss": 0.8249,
+      "step": 3467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.242212969833199,
+      "learning_rate": 1.883488890436112e-05,
+      "loss": 0.844,
+      "step": 3468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.218738505526197,
+      "learning_rate": 1.8834098781574415e-05,
+      "loss": 0.785,
+      "step": 3469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.179441163258888,
+      "learning_rate": 1.883330840755111e-05,
+      "loss": 0.7026,
+      "step": 3470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.914691652235785,
+      "learning_rate": 1.883251778231368e-05,
+      "loss": 0.837,
+      "step": 3471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.636151214867326,
+      "learning_rate": 1.883172690588461e-05,
+      "loss": 0.7955,
+      "step": 3472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 26.70029326860049,
+      "learning_rate": 1.8830935778286393e-05,
+      "loss": 0.782,
+      "step": 3473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.354191565379452,
+      "learning_rate": 1.8830144399541533e-05,
+      "loss": 0.7538,
+      "step": 3474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 25.22030474964758,
+      "learning_rate": 1.8829352769672525e-05,
+      "loss": 0.8834,
+      "step": 3475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 25.52685199868065,
+      "learning_rate": 1.8828560888701895e-05,
+      "loss": 0.7973,
+      "step": 3476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 26.280431588977667,
+      "learning_rate": 1.8827768756652155e-05,
+      "loss": 0.8578,
+      "step": 3477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.79368563561777,
+      "learning_rate": 1.8826976373545838e-05,
+      "loss": 0.9458,
+      "step": 3478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.586325938225523,
+      "learning_rate": 1.8826183739405475e-05,
+      "loss": 0.7706,
+      "step": 3479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.495325419543924,
+      "learning_rate": 1.8825390854253605e-05,
+      "loss": 0.8013,
+      "step": 3480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.841185596979305,
+      "learning_rate": 1.882459771811279e-05,
+      "loss": 0.7905,
+      "step": 3481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 27.790002438057645,
+      "learning_rate": 1.8823804331005573e-05,
+      "loss": 0.9765,
+      "step": 3482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.265754893403322,
+      "learning_rate": 1.8823010692954523e-05,
+      "loss": 0.723,
+      "step": 3483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.58633743732954,
+      "learning_rate": 1.8822216803982214e-05,
+      "loss": 0.7811,
+      "step": 3484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.22621006744884,
+      "learning_rate": 1.8821422664111218e-05,
+      "loss": 0.9039,
+      "step": 3485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.987284765591708,
+      "learning_rate": 1.882062827336412e-05,
+      "loss": 0.787,
+      "step": 3486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.46565408548851,
+      "learning_rate": 1.8819833631763514e-05,
+      "loss": 0.683,
+      "step": 3487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 20.111958333714732,
+      "learning_rate": 1.8819038739332e-05,
+      "loss": 0.8128,
+      "step": 3488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 21.848316055120737,
+      "learning_rate": 1.8818243596092182e-05,
+      "loss": 0.8297,
+      "step": 3489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.662014347126476,
+      "learning_rate": 1.8817448202066677e-05,
+      "loss": 0.88,
+      "step": 3490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.752111273733133,
+      "learning_rate": 1.88166525572781e-05,
+      "loss": 0.7116,
+      "step": 3491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.515634826229782,
+      "learning_rate": 1.8815856661749085e-05,
+      "loss": 0.7744,
+      "step": 3492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.680512253111425,
+      "learning_rate": 1.8815060515502262e-05,
+      "loss": 0.7396,
+      "step": 3493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.066291974934833,
+      "learning_rate": 1.8814264118560274e-05,
+      "loss": 0.736,
+      "step": 3494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.829860335486249,
+      "learning_rate": 1.8813467470945774e-05,
+      "loss": 0.761,
+      "step": 3495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.75692336622185,
+      "learning_rate": 1.8812670572681413e-05,
+      "loss": 0.7411,
+      "step": 3496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.85813735818634,
+      "learning_rate": 1.8811873423789853e-05,
+      "loss": 0.7823,
+      "step": 3497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.627602908166725,
+      "learning_rate": 1.8811076024293774e-05,
+      "loss": 0.8384,
+      "step": 3498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 33.77782872172686,
+      "learning_rate": 1.881027837421584e-05,
+      "loss": 0.819,
+      "step": 3499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 33.12887411315028,
+      "learning_rate": 1.880948047357875e-05,
+      "loss": 0.7636,
+      "step": 3500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.36707713713431,
+      "learning_rate": 1.8808682322405184e-05,
+      "loss": 0.7357,
+      "step": 3501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.492143628995628,
+      "learning_rate": 1.8807883920717844e-05,
+      "loss": 0.865,
+      "step": 3502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.756720148068348,
+      "learning_rate": 1.8807085268539443e-05,
+      "loss": 0.77,
+      "step": 3503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.94826734523092,
+      "learning_rate": 1.8806286365892685e-05,
+      "loss": 0.8267,
+      "step": 3504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.55035426808554,
+      "learning_rate": 1.8805487212800297e-05,
+      "loss": 0.7403,
+      "step": 3505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.823150509122005,
+      "learning_rate": 1.8804687809285003e-05,
+      "loss": 0.8168,
+      "step": 3506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.015405160582734,
+      "learning_rate": 1.8803888155369538e-05,
+      "loss": 0.7991,
+      "step": 3507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.309295285606257,
+      "learning_rate": 1.8803088251076642e-05,
+      "loss": 0.8021,
+      "step": 3508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.422907803898124,
+      "learning_rate": 1.880228809642907e-05,
+      "loss": 0.7982,
+      "step": 3509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.902390401289402,
+      "learning_rate": 1.880148769144957e-05,
+      "loss": 0.808,
+      "step": 3510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.874141597630716,
+      "learning_rate": 1.880068703616091e-05,
+      "loss": 0.7047,
+      "step": 3511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.998691193772636,
+      "learning_rate": 1.8799886130585858e-05,
+      "loss": 0.7697,
+      "step": 3512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.20763225325,
+      "learning_rate": 1.8799084974747193e-05,
+      "loss": 0.8053,
+      "step": 3513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.560378502900267,
+      "learning_rate": 1.8798283568667698e-05,
+      "loss": 0.7729,
+      "step": 3514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 27.044188167107126,
+      "learning_rate": 1.8797481912370167e-05,
+      "loss": 0.7679,
+      "step": 3515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.8551137071639,
+      "learning_rate": 1.8796680005877397e-05,
+      "loss": 0.8702,
+      "step": 3516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.63964457223816,
+      "learning_rate": 1.879587784921219e-05,
+      "loss": 0.794,
+      "step": 3517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.036531375690243,
+      "learning_rate": 1.8795075442397367e-05,
+      "loss": 0.7565,
+      "step": 3518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.444531661007105,
+      "learning_rate": 1.879427278545574e-05,
+      "loss": 0.7602,
+      "step": 3519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.818201867740747,
+      "learning_rate": 1.8793469878410142e-05,
+      "loss": 0.7963,
+      "step": 3520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.78712336711334,
+      "learning_rate": 1.8792666721283406e-05,
+      "loss": 0.8398,
+      "step": 3521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.980181112045726,
+      "learning_rate": 1.8791863314098368e-05,
+      "loss": 0.8076,
+      "step": 3522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.180916078675132,
+      "learning_rate": 1.879105965687788e-05,
+      "loss": 0.8217,
+      "step": 3523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.73869338943173,
+      "learning_rate": 1.87902557496448e-05,
+      "loss": 0.7327,
+      "step": 3524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.342797615777204,
+      "learning_rate": 1.878945159242199e-05,
+      "loss": 0.7908,
+      "step": 3525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.183876213546846,
+      "learning_rate": 1.878864718523232e-05,
+      "loss": 0.7959,
+      "step": 3526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.645566801084128,
+      "learning_rate": 1.8787842528098657e-05,
+      "loss": 0.7731,
+      "step": 3527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.787845821507243,
+      "learning_rate": 1.87870376210439e-05,
+      "loss": 0.757,
+      "step": 3528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.189468052211144,
+      "learning_rate": 1.878623246409093e-05,
+      "loss": 0.8377,
+      "step": 3529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.148868128724928,
+      "learning_rate": 1.878542705726265e-05,
+      "loss": 0.7908,
+      "step": 3530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.809226666395197,
+      "learning_rate": 1.8784621400581966e-05,
+      "loss": 0.8775,
+      "step": 3531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.48633325163417,
+      "learning_rate": 1.8783815494071786e-05,
+      "loss": 0.7944,
+      "step": 3532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.74523896647049,
+      "learning_rate": 1.8783009337755032e-05,
+      "loss": 0.8008,
+      "step": 3533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.606981366940992,
+      "learning_rate": 1.878220293165463e-05,
+      "loss": 0.8824,
+      "step": 3534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.213001332379317,
+      "learning_rate": 1.8781396275793516e-05,
+      "loss": 0.7521,
+      "step": 3535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.088980811287946,
+      "learning_rate": 1.8780589370194626e-05,
+      "loss": 0.8775,
+      "step": 3536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 28.292222384602326,
+      "learning_rate": 1.8779782214880912e-05,
+      "loss": 0.8227,
+      "step": 3537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.035381930205684,
+      "learning_rate": 1.8778974809875326e-05,
+      "loss": 0.8039,
+      "step": 3538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.780408127763714,
+      "learning_rate": 1.8778167155200836e-05,
+      "loss": 0.7841,
+      "step": 3539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.79215376886102,
+      "learning_rate": 1.8777359250880404e-05,
+      "loss": 0.8581,
+      "step": 3540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 26.351232227480317,
+      "learning_rate": 1.8776551096937014e-05,
+      "loss": 0.9017,
+      "step": 3541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.091008297288706,
+      "learning_rate": 1.8775742693393638e-05,
+      "loss": 0.8004,
+      "step": 3542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.097193069179173,
+      "learning_rate": 1.8774934040273278e-05,
+      "loss": 0.8364,
+      "step": 3543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.74878090252714,
+      "learning_rate": 1.8774125137598926e-05,
+      "loss": 0.7653,
+      "step": 3544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.47398130237187,
+      "learning_rate": 1.877331598539359e-05,
+      "loss": 0.8068,
+      "step": 3545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 28.430191692316722,
+      "learning_rate": 1.8772506583680273e-05,
+      "loss": 0.9048,
+      "step": 3546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.8082207946181,
+      "learning_rate": 1.8771696932482008e-05,
+      "loss": 0.8238,
+      "step": 3547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.417594390523444,
+      "learning_rate": 1.8770887031821813e-05,
+      "loss": 0.903,
+      "step": 3548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.11493896140036,
+      "learning_rate": 1.8770076881722717e-05,
+      "loss": 0.675,
+      "step": 3549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.486352414912705,
+      "learning_rate": 1.8769266482207766e-05,
+      "loss": 0.8944,
+      "step": 3550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.881971707791365,
+      "learning_rate": 1.8768455833300007e-05,
+      "loss": 0.8369,
+      "step": 3551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 26.010122825104496,
+      "learning_rate": 1.8767644935022496e-05,
+      "loss": 0.8127,
+      "step": 3552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.726223891806299,
+      "learning_rate": 1.876683378739829e-05,
+      "loss": 0.6649,
+      "step": 3553
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 33.54011793628526,
+      "learning_rate": 1.876602239045046e-05,
+      "loss": 0.9012,
+      "step": 3554
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.61798279990791,
+      "learning_rate": 1.876521074420208e-05,
+      "loss": 0.7523,
+      "step": 3555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.164041866659867,
+      "learning_rate": 1.8764398848676232e-05,
+      "loss": 0.856,
+      "step": 3556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.812691038345378,
+      "learning_rate": 1.876358670389601e-05,
+      "loss": 0.8713,
+      "step": 3557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.691827570897786,
+      "learning_rate": 1.8762774309884508e-05,
+      "loss": 0.7943,
+      "step": 3558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.22391859148357,
+      "learning_rate": 1.876196166666483e-05,
+      "loss": 0.7599,
+      "step": 3559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 22.280253101801907,
+      "learning_rate": 1.876114877426009e-05,
+      "loss": 0.7013,
+      "step": 3560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.241165103207855,
+      "learning_rate": 1.8760335632693404e-05,
+      "loss": 0.7498,
+      "step": 3561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.492375995276703,
+      "learning_rate": 1.8759522241987896e-05,
+      "loss": 0.8936,
+      "step": 3562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.047379567577668,
+      "learning_rate": 1.87587086021667e-05,
+      "loss": 0.7544,
+      "step": 3563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.261741974576328,
+      "learning_rate": 1.8757894713252956e-05,
+      "loss": 0.7377,
+      "step": 3564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.04700849047223,
+      "learning_rate": 1.8757080575269808e-05,
+      "loss": 0.7349,
+      "step": 3565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.582964809993896,
+      "learning_rate": 1.8756266188240413e-05,
+      "loss": 0.8271,
+      "step": 3566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.764698830116274,
+      "learning_rate": 1.875545155218793e-05,
+      "loss": 0.7742,
+      "step": 3567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.89655761220839,
+      "learning_rate": 1.8754636667135523e-05,
+      "loss": 0.8057,
+      "step": 3568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.626023173192689,
+      "learning_rate": 1.8753821533106373e-05,
+      "loss": 0.8656,
+      "step": 3569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.137613945736003,
+      "learning_rate": 1.8753006150123662e-05,
+      "loss": 0.7823,
+      "step": 3570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.26057195828669,
+      "learning_rate": 1.8752190518210572e-05,
+      "loss": 0.8447,
+      "step": 3571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.026301947296844,
+      "learning_rate": 1.8751374637390304e-05,
+      "loss": 0.8068,
+      "step": 3572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.285851063941212,
+      "learning_rate": 1.8750558507686065e-05,
+      "loss": 0.8071,
+      "step": 3573
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.57400438624219,
+      "learning_rate": 1.8749742129121058e-05,
+      "loss": 0.8453,
+      "step": 3574
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.30709197570784,
+      "learning_rate": 1.87489255017185e-05,
+      "loss": 0.7422,
+      "step": 3575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.731462238209005,
+      "learning_rate": 1.8748108625501623e-05,
+      "loss": 0.9153,
+      "step": 3576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 25.09779022359157,
+      "learning_rate": 1.8747291500493653e-05,
+      "loss": 0.7667,
+      "step": 3577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.817954822661456,
+      "learning_rate": 1.874647412671783e-05,
+      "loss": 0.809,
+      "step": 3578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 23.494904352463458,
+      "learning_rate": 1.87456565041974e-05,
+      "loss": 0.863,
+      "step": 3579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.312299238826743,
+      "learning_rate": 1.8744838632955617e-05,
+      "loss": 0.8786,
+      "step": 3580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 29.869929817642944,
+      "learning_rate": 1.8744020513015733e-05,
+      "loss": 0.915,
+      "step": 3581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.34152127146327,
+      "learning_rate": 1.8743202144401027e-05,
+      "loss": 0.7927,
+      "step": 3582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.800781097231496,
+      "learning_rate": 1.874238352713476e-05,
+      "loss": 0.8538,
+      "step": 3583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.98585337606707,
+      "learning_rate": 1.8741564661240223e-05,
+      "loss": 0.7454,
+      "step": 3584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.135990223814957,
+      "learning_rate": 1.87407455467407e-05,
+      "loss": 0.9235,
+      "step": 3585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.53984675067897,
+      "learning_rate": 1.8739926183659492e-05,
+      "loss": 0.8527,
+      "step": 3586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.051889720266622,
+      "learning_rate": 1.873910657201989e-05,
+      "loss": 0.7914,
+      "step": 3587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.075225415406303,
+      "learning_rate": 1.873828671184521e-05,
+      "loss": 0.6688,
+      "step": 3588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.689917363989505,
+      "learning_rate": 1.873746660315877e-05,
+      "loss": 0.6685,
+      "step": 3589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.073484476381477,
+      "learning_rate": 1.873664624598389e-05,
+      "loss": 0.8412,
+      "step": 3590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.233145370390293,
+      "learning_rate": 1.8735825640343904e-05,
+      "loss": 0.7767,
+      "step": 3591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.73616713990554,
+      "learning_rate": 1.8735004786262142e-05,
+      "loss": 0.7929,
+      "step": 3592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.424976564636232,
+      "learning_rate": 1.8734183683761957e-05,
+      "loss": 0.7167,
+      "step": 3593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.365261378791256,
+      "learning_rate": 1.8733362332866698e-05,
+      "loss": 0.7045,
+      "step": 3594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.050867687142146,
+      "learning_rate": 1.873254073359972e-05,
+      "loss": 0.7009,
+      "step": 3595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.590000056221823,
+      "learning_rate": 1.8731718885984396e-05,
+      "loss": 0.7316,
+      "step": 3596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.923169066313335,
+      "learning_rate": 1.8730896790044092e-05,
+      "loss": 0.8225,
+      "step": 3597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.301342875710873,
+      "learning_rate": 1.873007444580219e-05,
+      "loss": 0.6974,
+      "step": 3598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.55983581164699,
+      "learning_rate": 1.872925185328208e-05,
+      "loss": 0.8101,
+      "step": 3599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.681606536335355,
+      "learning_rate": 1.872842901250715e-05,
+      "loss": 0.8205,
+      "step": 3600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.115853660671004,
+      "learning_rate": 1.872760592350081e-05,
+      "loss": 0.7441,
+      "step": 3601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.91763958736906,
+      "learning_rate": 1.872678258628646e-05,
+      "loss": 0.7799,
+      "step": 3602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 30.306542040101633,
+      "learning_rate": 1.872595900088752e-05,
+      "loss": 0.8051,
+      "step": 3603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 27.23274052060126,
+      "learning_rate": 1.872513516732741e-05,
+      "loss": 0.847,
+      "step": 3604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.02108160886153,
+      "learning_rate": 1.872431108562956e-05,
+      "loss": 0.8093,
+      "step": 3605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 26.913484732152558,
+      "learning_rate": 1.8723486755817405e-05,
+      "loss": 0.7594,
+      "step": 3606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.98816288542478,
+      "learning_rate": 1.872266217791439e-05,
+      "loss": 0.8622,
+      "step": 3607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.278605511314908,
+      "learning_rate": 1.8721837351943964e-05,
+      "loss": 0.7945,
+      "step": 3608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.05202768757929,
+      "learning_rate": 1.8721012277929588e-05,
+      "loss": 0.7563,
+      "step": 3609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.60643878756732,
+      "learning_rate": 1.8720186955894722e-05,
+      "loss": 0.7403,
+      "step": 3610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 23.26520798311354,
+      "learning_rate": 1.8719361385862843e-05,
+      "loss": 0.6803,
+      "step": 3611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.350552483449457,
+      "learning_rate": 1.8718535567857426e-05,
+      "loss": 0.8566,
+      "step": 3612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.610469631250647,
+      "learning_rate": 1.8717709501901956e-05,
+      "loss": 0.8075,
+      "step": 3613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 22.323171734891655,
+      "learning_rate": 1.8716883188019932e-05,
+      "loss": 0.733,
+      "step": 3614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 25.128167239825995,
+      "learning_rate": 1.8716056626234848e-05,
+      "loss": 0.9516,
+      "step": 3615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.73150825360642,
+      "learning_rate": 1.871522981657021e-05,
+      "loss": 0.8248,
+      "step": 3616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 22.36607078445375,
+      "learning_rate": 1.8714402759049535e-05,
+      "loss": 0.8285,
+      "step": 3617
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.57147295931924,
+      "learning_rate": 1.8713575453696344e-05,
+      "loss": 0.9086,
+      "step": 3618
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.40856968082394,
+      "learning_rate": 1.8712747900534164e-05,
+      "loss": 0.7795,
+      "step": 3619
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.859809590037942,
+      "learning_rate": 1.871192009958653e-05,
+      "loss": 0.7441,
+      "step": 3620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.53025016819523,
+      "learning_rate": 1.8711092050876982e-05,
+      "loss": 0.9138,
+      "step": 3621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 34.76080812152451,
+      "learning_rate": 1.8710263754429076e-05,
+      "loss": 0.8394,
+      "step": 3622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.259602599968307,
+      "learning_rate": 1.870943521026636e-05,
+      "loss": 0.7928,
+      "step": 3623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.41764770275707,
+      "learning_rate": 1.8708606418412408e-05,
+      "loss": 0.7217,
+      "step": 3624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.228803090629167,
+      "learning_rate": 1.8707777378890778e-05,
+      "loss": 0.7196,
+      "step": 3625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.779308315302089,
+      "learning_rate": 1.870694809172506e-05,
+      "loss": 0.6867,
+      "step": 3626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.212870471543894,
+      "learning_rate": 1.8706118556938826e-05,
+      "loss": 0.8893,
+      "step": 3627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.288949869540392,
+      "learning_rate": 1.870528877455567e-05,
+      "loss": 0.7834,
+      "step": 3628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.652253198347413,
+      "learning_rate": 1.8704458744599202e-05,
+      "loss": 0.7959,
+      "step": 3629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 24.136165796846026,
+      "learning_rate": 1.8703628467093017e-05,
+      "loss": 0.7148,
+      "step": 3630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.19935266429516,
+      "learning_rate": 1.8702797942060725e-05,
+      "loss": 0.7861,
+      "step": 3631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.491594517855711,
+      "learning_rate": 1.8701967169525958e-05,
+      "loss": 0.7346,
+      "step": 3632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.613146952174763,
+      "learning_rate": 1.8701136149512328e-05,
+      "loss": 0.8124,
+      "step": 3633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.47404466251857,
+      "learning_rate": 1.870030488204348e-05,
+      "loss": 0.7995,
+      "step": 3634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.611119590517344,
+      "learning_rate": 1.869947336714305e-05,
+      "loss": 0.7885,
+      "step": 3635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.579567387238896,
+      "learning_rate": 1.8698641604834687e-05,
+      "loss": 0.7388,
+      "step": 3636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.468340198010637,
+      "learning_rate": 1.8697809595142042e-05,
+      "loss": 0.8654,
+      "step": 3637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.634382934161103,
+      "learning_rate": 1.869697733808878e-05,
+      "loss": 0.7932,
+      "step": 3638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.187628459485262,
+      "learning_rate": 1.8696144833698574e-05,
+      "loss": 0.645,
+      "step": 3639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.130600509361752,
+      "learning_rate": 1.8695312081995096e-05,
+      "loss": 0.8204,
+      "step": 3640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.831058635649146,
+      "learning_rate": 1.8694479083002027e-05,
+      "loss": 0.8136,
+      "step": 3641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.409522530523994,
+      "learning_rate": 1.8693645836743062e-05,
+      "loss": 0.8237,
+      "step": 3642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.235013782217933,
+      "learning_rate": 1.8692812343241892e-05,
+      "loss": 0.7363,
+      "step": 3643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.526979839661074,
+      "learning_rate": 1.8691978602522226e-05,
+      "loss": 0.8215,
+      "step": 3644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.583651183965696,
+      "learning_rate": 1.869114461460777e-05,
+      "loss": 0.7235,
+      "step": 3645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.26246913640462,
+      "learning_rate": 1.8690310379522247e-05,
+      "loss": 0.7706,
+      "step": 3646
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.183451956300864,
+      "learning_rate": 1.8689475897289383e-05,
+      "loss": 0.8115,
+      "step": 3647
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.555517784457376,
+      "learning_rate": 1.8688641167932906e-05,
+      "loss": 0.7744,
+      "step": 3648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.314369426079622,
+      "learning_rate": 1.8687806191476555e-05,
+      "loss": 0.8448,
+      "step": 3649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 57.21910759185438,
+      "learning_rate": 1.8686970967944082e-05,
+      "loss": 0.911,
+      "step": 3650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.292246053565233,
+      "learning_rate": 1.8686135497359233e-05,
+      "loss": 0.7652,
+      "step": 3651
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.760178358034118,
+      "learning_rate": 1.868529977974577e-05,
+      "loss": 0.8735,
+      "step": 3652
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.26116852854488,
+      "learning_rate": 1.868446381512747e-05,
+      "loss": 0.763,
+      "step": 3653
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.453039699809953,
+      "learning_rate": 1.868362760352809e-05,
+      "loss": 0.818,
+      "step": 3654
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 26.40508058486648,
+      "learning_rate": 1.8682791144971426e-05,
+      "loss": 0.7489,
+      "step": 3655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.359591276708297,
+      "learning_rate": 1.8681954439481262e-05,
+      "loss": 0.7248,
+      "step": 3656
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.872460952596,
+      "learning_rate": 1.868111748708139e-05,
+      "loss": 0.7021,
+      "step": 3657
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.487588763402098,
+      "learning_rate": 1.8680280287795612e-05,
+      "loss": 0.7719,
+      "step": 3658
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 24.57308009136288,
+      "learning_rate": 1.8679442841647747e-05,
+      "loss": 0.7908,
+      "step": 3659
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.86155825162964,
+      "learning_rate": 1.8678605148661602e-05,
+      "loss": 0.7919,
+      "step": 3660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.30456815592562,
+      "learning_rate": 1.8677767208861e-05,
+      "loss": 0.8657,
+      "step": 3661
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.672394869080815,
+      "learning_rate": 1.867692902226978e-05,
+      "loss": 0.7507,
+      "step": 3662
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.984745160195164,
+      "learning_rate": 1.8676090588911774e-05,
+      "loss": 0.8663,
+      "step": 3663
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.687202469983326,
+      "learning_rate": 1.8675251908810824e-05,
+      "loss": 0.8457,
+      "step": 3664
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.454990390052906,
+      "learning_rate": 1.867441298199079e-05,
+      "loss": 0.8896,
+      "step": 3665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.650155715851447,
+      "learning_rate": 1.867357380847552e-05,
+      "loss": 0.8179,
+      "step": 3666
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.549687040618135,
+      "learning_rate": 1.8672734388288883e-05,
+      "loss": 0.8302,
+      "step": 3667
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 36.85061458123028,
+      "learning_rate": 1.867189472145476e-05,
+      "loss": 0.8611,
+      "step": 3668
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.57156035807646,
+      "learning_rate": 1.867105480799702e-05,
+      "loss": 0.8497,
+      "step": 3669
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.301634964109933,
+      "learning_rate": 1.867021464793955e-05,
+      "loss": 0.9199,
+      "step": 3670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.41043453638851,
+      "learning_rate": 1.8669374241306256e-05,
+      "loss": 0.8815,
+      "step": 3671
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.0655999081036,
+      "learning_rate": 1.8668533588121026e-05,
+      "loss": 0.6696,
+      "step": 3672
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.456877733810106,
+      "learning_rate": 1.866769268840777e-05,
+      "loss": 0.7239,
+      "step": 3673
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.17869941548231,
+      "learning_rate": 1.8666851542190406e-05,
+      "loss": 0.8522,
+      "step": 3674
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.974652400007232,
+      "learning_rate": 1.8666010149492853e-05,
+      "loss": 0.8909,
+      "step": 3675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.414487590279492,
+      "learning_rate": 1.8665168510339043e-05,
+      "loss": 0.8369,
+      "step": 3676
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.600487558833734,
+      "learning_rate": 1.866432662475291e-05,
+      "loss": 0.7624,
+      "step": 3677
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.683341203338003,
+      "learning_rate": 1.8663484492758393e-05,
+      "loss": 0.7371,
+      "step": 3678
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.098718588319226,
+      "learning_rate": 1.8662642114379444e-05,
+      "loss": 0.8288,
+      "step": 3679
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.80406669918209,
+      "learning_rate": 1.8661799489640025e-05,
+      "loss": 0.8039,
+      "step": 3680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.536797920397884,
+      "learning_rate": 1.8660956618564095e-05,
+      "loss": 0.8121,
+      "step": 3681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.463780093074181,
+      "learning_rate": 1.8660113501175622e-05,
+      "loss": 0.8281,
+      "step": 3682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.134779266009994,
+      "learning_rate": 1.8659270137498588e-05,
+      "loss": 0.8206,
+      "step": 3683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.628277560276816,
+      "learning_rate": 1.865842652755698e-05,
+      "loss": 0.7461,
+      "step": 3684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.614731937008038,
+      "learning_rate": 1.8657582671374787e-05,
+      "loss": 0.6644,
+      "step": 3685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.265353233388268,
+      "learning_rate": 1.8656738568976003e-05,
+      "loss": 0.8624,
+      "step": 3686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 40.75501088477457,
+      "learning_rate": 1.865589422038464e-05,
+      "loss": 0.9502,
+      "step": 3687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.772324339522127,
+      "learning_rate": 1.865504962562471e-05,
+      "loss": 0.8153,
+      "step": 3688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 33.304020535159474,
+      "learning_rate": 1.865420478472023e-05,
+      "loss": 0.8258,
+      "step": 3689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.301013521861792,
+      "learning_rate": 1.865335969769523e-05,
+      "loss": 0.792,
+      "step": 3690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.082822833479007,
+      "learning_rate": 1.865251436457374e-05,
+      "loss": 0.7324,
+      "step": 3691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.465329748446127,
+      "learning_rate": 1.8651668785379806e-05,
+      "loss": 0.8932,
+      "step": 3692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.513643689980302,
+      "learning_rate": 1.865082296013747e-05,
+      "loss": 0.7207,
+      "step": 3693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.25021015911653,
+      "learning_rate": 1.864997688887079e-05,
+      "loss": 0.8304,
+      "step": 3694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.24651118111633,
+      "learning_rate": 1.864913057160383e-05,
+      "loss": 0.7584,
+      "step": 3695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.393855193902642,
+      "learning_rate": 1.8648284008360654e-05,
+      "loss": 0.7448,
+      "step": 3696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.497330229931306,
+      "learning_rate": 1.864743719916534e-05,
+      "loss": 0.7827,
+      "step": 3697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.0206140060649,
+      "learning_rate": 1.8646590144041972e-05,
+      "loss": 0.7568,
+      "step": 3698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.920893952006146,
+      "learning_rate": 1.8645742843014635e-05,
+      "loss": 0.7157,
+      "step": 3699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.773234533872827,
+      "learning_rate": 1.8644895296107432e-05,
+      "loss": 0.8918,
+      "step": 3700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.331823485254711,
+      "learning_rate": 1.8644047503344465e-05,
+      "loss": 0.7402,
+      "step": 3701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.397191009440277,
+      "learning_rate": 1.8643199464749843e-05,
+      "loss": 0.8009,
+      "step": 3702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.073646718228165,
+      "learning_rate": 1.8642351180347682e-05,
+      "loss": 0.7898,
+      "step": 3703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 22.889933136258644,
+      "learning_rate": 1.864150265016211e-05,
+      "loss": 0.8683,
+      "step": 3704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 22.081720816621452,
+      "learning_rate": 1.8640653874217257e-05,
+      "loss": 0.7606,
+      "step": 3705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.892132729651188,
+      "learning_rate": 1.8639804852537264e-05,
+      "loss": 0.749,
+      "step": 3706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 22.552952171787748,
+      "learning_rate": 1.8638955585146277e-05,
+      "loss": 0.8249,
+      "step": 3707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.715828096238656,
+      "learning_rate": 1.8638106072068443e-05,
+      "loss": 0.8625,
+      "step": 3708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.97202987995221,
+      "learning_rate": 1.863725631332793e-05,
+      "loss": 0.7885,
+      "step": 3709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.576101891571074,
+      "learning_rate": 1.8636406308948894e-05,
+      "loss": 0.716,
+      "step": 3710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.46503447903622,
+      "learning_rate": 1.863555605895552e-05,
+      "loss": 0.8587,
+      "step": 3711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.752132537564197,
+      "learning_rate": 1.863470556337198e-05,
+      "loss": 0.7506,
+      "step": 3712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.947742279101046,
+      "learning_rate": 1.8633854822222467e-05,
+      "loss": 0.8411,
+      "step": 3713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.89918343127105,
+      "learning_rate": 1.8633003835531172e-05,
+      "loss": 0.8768,
+      "step": 3714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 21.52785553286638,
+      "learning_rate": 1.86321526033223e-05,
+      "loss": 0.8796,
+      "step": 3715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.747557255120682,
+      "learning_rate": 1.8631301125620056e-05,
+      "loss": 0.887,
+      "step": 3716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.558054906778917,
+      "learning_rate": 1.8630449402448655e-05,
+      "loss": 0.8198,
+      "step": 3717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.103131864803347,
+      "learning_rate": 1.8629597433832326e-05,
+      "loss": 0.7212,
+      "step": 3718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.514996598774914,
+      "learning_rate": 1.8628745219795292e-05,
+      "loss": 0.8049,
+      "step": 3719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.416992104377439,
+      "learning_rate": 1.862789276036179e-05,
+      "loss": 0.8523,
+      "step": 3720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.69594792705124,
+      "learning_rate": 1.862704005555607e-05,
+      "loss": 0.7291,
+      "step": 3721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.649249725546298,
+      "learning_rate": 1.8626187105402373e-05,
+      "loss": 0.8157,
+      "step": 3722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.471889943738677,
+      "learning_rate": 1.8625333909924964e-05,
+      "loss": 0.8382,
+      "step": 3723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.930556376606772,
+      "learning_rate": 1.8624480469148103e-05,
+      "loss": 0.7533,
+      "step": 3724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.18798221443737,
+      "learning_rate": 1.862362678309606e-05,
+      "loss": 0.7067,
+      "step": 3725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 33.96096729249176,
+      "learning_rate": 1.862277285179312e-05,
+      "loss": 0.7739,
+      "step": 3726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 23.539033528287007,
+      "learning_rate": 1.8621918675263564e-05,
+      "loss": 0.8901,
+      "step": 3727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.943651439129837,
+      "learning_rate": 1.862106425353168e-05,
+      "loss": 0.8123,
+      "step": 3728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.068993937394005,
+      "learning_rate": 1.8620209586621778e-05,
+      "loss": 0.6734,
+      "step": 3729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.885760329232188,
+      "learning_rate": 1.8619354674558155e-05,
+      "loss": 0.7728,
+      "step": 3730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.392613563181637,
+      "learning_rate": 1.8618499517365127e-05,
+      "loss": 0.8404,
+      "step": 3731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.644081427299254,
+      "learning_rate": 1.8617644115067013e-05,
+      "loss": 0.6939,
+      "step": 3732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 26.357701271049024,
+      "learning_rate": 1.8616788467688146e-05,
+      "loss": 0.7877,
+      "step": 3733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.70186861458413,
+      "learning_rate": 1.8615932575252855e-05,
+      "loss": 0.8398,
+      "step": 3734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.065060152623122,
+      "learning_rate": 1.8615076437785474e-05,
+      "loss": 0.8109,
+      "step": 3735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.992245761414257,
+      "learning_rate": 1.8614220055310367e-05,
+      "loss": 0.7348,
+      "step": 3736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.539017242837538,
+      "learning_rate": 1.8613363427851875e-05,
+      "loss": 0.77,
+      "step": 3737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.89321930033038,
+      "learning_rate": 1.8612506555434372e-05,
+      "loss": 0.748,
+      "step": 3738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.259032747631917,
+      "learning_rate": 1.861164943808222e-05,
+      "loss": 0.7775,
+      "step": 3739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 22.41070148109956,
+      "learning_rate": 1.861079207581979e-05,
+      "loss": 0.8158,
+      "step": 3740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.338031472144,
+      "learning_rate": 1.8609934468671474e-05,
+      "loss": 0.848,
+      "step": 3741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.376155894349836,
+      "learning_rate": 1.8609076616661658e-05,
+      "loss": 0.8179,
+      "step": 3742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.957199003573866,
+      "learning_rate": 1.860821851981474e-05,
+      "loss": 0.8948,
+      "step": 3743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.75913222270166,
+      "learning_rate": 1.8607360178155123e-05,
+      "loss": 0.7616,
+      "step": 3744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.465591898348475,
+      "learning_rate": 1.8606501591707217e-05,
+      "loss": 0.8039,
+      "step": 3745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.581807346691082,
+      "learning_rate": 1.860564276049544e-05,
+      "loss": 0.7629,
+      "step": 3746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.018448676515995,
+      "learning_rate": 1.860478368454422e-05,
+      "loss": 0.8146,
+      "step": 3747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.529173518563663,
+      "learning_rate": 1.860392436387798e-05,
+      "loss": 0.875,
+      "step": 3748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.22668664928939,
+      "learning_rate": 1.8603064798521168e-05,
+      "loss": 0.8169,
+      "step": 3749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.648865502914557,
+      "learning_rate": 1.8602204988498224e-05,
+      "loss": 0.6834,
+      "step": 3750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 24.235708710342703,
+      "learning_rate": 1.8601344933833605e-05,
+      "loss": 0.787,
+      "step": 3751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.519268344095945,
+      "learning_rate": 1.8600484634551767e-05,
+      "loss": 0.8487,
+      "step": 3752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.483473850314812,
+      "learning_rate": 1.8599624090677174e-05,
+      "loss": 0.8656,
+      "step": 3753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.485350888528718,
+      "learning_rate": 1.8598763302234306e-05,
+      "loss": 0.8484,
+      "step": 3754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.958964163665327,
+      "learning_rate": 1.859790226924764e-05,
+      "loss": 0.7221,
+      "step": 3755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 24.43862765361714,
+      "learning_rate": 1.8597040991741663e-05,
+      "loss": 0.8133,
+      "step": 3756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.426216796530715,
+      "learning_rate": 1.8596179469740863e-05,
+      "loss": 0.8019,
+      "step": 3757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.811719792830015,
+      "learning_rate": 1.8595317703269756e-05,
+      "loss": 0.8006,
+      "step": 3758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.485282200702745,
+      "learning_rate": 1.859445569235284e-05,
+      "loss": 0.7719,
+      "step": 3759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.848862857046242,
+      "learning_rate": 1.8593593437014627e-05,
+      "loss": 0.9183,
+      "step": 3760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.01885737200636,
+      "learning_rate": 1.8592730937279647e-05,
+      "loss": 0.8939,
+      "step": 3761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.799788601583767,
+      "learning_rate": 1.8591868193172423e-05,
+      "loss": 0.7175,
+      "step": 3762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 23.225276019629412,
+      "learning_rate": 1.8591005204717498e-05,
+      "loss": 0.7856,
+      "step": 3763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.592885455927046,
+      "learning_rate": 1.859014197193941e-05,
+      "loss": 0.7464,
+      "step": 3764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 23.39088362396005,
+      "learning_rate": 1.8589278494862707e-05,
+      "loss": 0.8755,
+      "step": 3765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.106150836220248,
+      "learning_rate": 1.858841477351195e-05,
+      "loss": 0.7518,
+      "step": 3766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.785924575441783,
+      "learning_rate": 1.85875508079117e-05,
+      "loss": 0.8277,
+      "step": 3767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.84390877051048,
+      "learning_rate": 1.858668659808653e-05,
+      "loss": 0.8027,
+      "step": 3768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.61427228954143,
+      "learning_rate": 1.8585822144061013e-05,
+      "loss": 0.8436,
+      "step": 3769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.665475394438053,
+      "learning_rate": 1.858495744585974e-05,
+      "loss": 0.6795,
+      "step": 3770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.333938397397688,
+      "learning_rate": 1.8584092503507296e-05,
+      "loss": 0.7503,
+      "step": 3771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.391209292576,
+      "learning_rate": 1.8583227317028287e-05,
+      "loss": 0.7334,
+      "step": 3772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.32468939906611,
+      "learning_rate": 1.8582361886447315e-05,
+      "loss": 0.84,
+      "step": 3773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.89209096820718,
+      "learning_rate": 1.8581496211788985e-05,
+      "loss": 0.722,
+      "step": 3774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.059009598705948,
+      "learning_rate": 1.8580630293077925e-05,
+      "loss": 0.7838,
+      "step": 3775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.446796155028977,
+      "learning_rate": 1.857976413033876e-05,
+      "loss": 0.7523,
+      "step": 3776
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.836283488509583,
+      "learning_rate": 1.8578897723596124e-05,
+      "loss": 0.818,
+      "step": 3777
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 24.57292858200032,
+      "learning_rate": 1.8578031072874655e-05,
+      "loss": 0.6849,
+      "step": 3778
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 28.16611248179845,
+      "learning_rate": 1.8577164178198995e-05,
+      "loss": 0.8842,
+      "step": 3779
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.157711751917876,
+      "learning_rate": 1.857629703959381e-05,
+      "loss": 0.7568,
+      "step": 3780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.410373774042984,
+      "learning_rate": 1.8575429657083753e-05,
+      "loss": 0.7815,
+      "step": 3781
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.342904453684604,
+      "learning_rate": 1.8574562030693492e-05,
+      "loss": 0.7891,
+      "step": 3782
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.830380088738504,
+      "learning_rate": 1.85736941604477e-05,
+      "loss": 0.7813,
+      "step": 3783
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 28.31186738581136,
+      "learning_rate": 1.8572826046371063e-05,
+      "loss": 0.8201,
+      "step": 3784
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.20612771751097,
+      "learning_rate": 1.8571957688488268e-05,
+      "loss": 0.7892,
+      "step": 3785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.547951486546346,
+      "learning_rate": 1.8571089086824014e-05,
+      "loss": 0.9203,
+      "step": 3786
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.98118268868699,
+      "learning_rate": 1.8570220241402996e-05,
+      "loss": 0.8133,
+      "step": 3787
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.630320797576523,
+      "learning_rate": 1.8569351152249932e-05,
+      "loss": 0.7927,
+      "step": 3788
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 19.285110558572953,
+      "learning_rate": 1.856848181938953e-05,
+      "loss": 0.8186,
+      "step": 3789
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.457573311308336,
+      "learning_rate": 1.856761224284652e-05,
+      "loss": 0.783,
+      "step": 3790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.97183920820715,
+      "learning_rate": 1.856674242264563e-05,
+      "loss": 0.7642,
+      "step": 3791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.592564884134067,
+      "learning_rate": 1.8565872358811594e-05,
+      "loss": 0.8134,
+      "step": 3792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.40412498552598,
+      "learning_rate": 1.856500205136916e-05,
+      "loss": 0.7384,
+      "step": 3793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.685917406879609,
+      "learning_rate": 1.8564131500343082e-05,
+      "loss": 0.8745,
+      "step": 3794
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.667348752559906,
+      "learning_rate": 1.8563260705758114e-05,
+      "loss": 0.7676,
+      "step": 3795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 26.836344030929713,
+      "learning_rate": 1.8562389667639016e-05,
+      "loss": 0.803,
+      "step": 3796
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.011330895596231,
+      "learning_rate": 1.8561518386010567e-05,
+      "loss": 0.7189,
+      "step": 3797
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.048253176286806,
+      "learning_rate": 1.856064686089754e-05,
+      "loss": 0.7731,
+      "step": 3798
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.730891031388188,
+      "learning_rate": 1.855977509232473e-05,
+      "loss": 0.7178,
+      "step": 3799
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.19745705653203,
+      "learning_rate": 1.8558903080316922e-05,
+      "loss": 0.9754,
+      "step": 3800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.072025045541388,
+      "learning_rate": 1.8558030824898915e-05,
+      "loss": 0.8115,
+      "step": 3801
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.687996764460728,
+      "learning_rate": 1.8557158326095515e-05,
+      "loss": 0.7821,
+      "step": 3802
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 27.291128992773302,
+      "learning_rate": 1.8556285583931547e-05,
+      "loss": 0.8319,
+      "step": 3803
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 31.95447375148788,
+      "learning_rate": 1.8555412598431814e-05,
+      "loss": 0.6964,
+      "step": 3804
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.714737086269803,
+      "learning_rate": 1.8554539369621156e-05,
+      "loss": 0.9218,
+      "step": 3805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.00982659509593,
+      "learning_rate": 1.85536658975244e-05,
+      "loss": 0.8843,
+      "step": 3806
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.160791135197275,
+      "learning_rate": 1.8552792182166393e-05,
+      "loss": 0.7222,
+      "step": 3807
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.08386377841724,
+      "learning_rate": 1.8551918223571975e-05,
+      "loss": 0.7301,
+      "step": 3808
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.394686121050924,
+      "learning_rate": 1.855104402176601e-05,
+      "loss": 0.7419,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.767436149750548,
+      "learning_rate": 1.8550169576773353e-05,
+      "loss": 0.7551,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.201831240165006,
+      "learning_rate": 1.8549294888618872e-05,
+      "loss": 0.7064,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 21.679244004258322,
+      "learning_rate": 1.8548419957327454e-05,
+      "loss": 0.7568,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 43.70672313633746,
+      "learning_rate": 1.8547544782923963e-05,
+      "loss": 0.809,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.798978130979503,
+      "learning_rate": 1.8546669365433305e-05,
+      "loss": 0.7657,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.444162700736577,
+      "learning_rate": 1.8545793704880367e-05,
+      "loss": 0.8721,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 26.541801941792407,
+      "learning_rate": 1.854491780129006e-05,
+      "loss": 0.6517,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.685965498701293,
+      "learning_rate": 1.8544041654687282e-05,
+      "loss": 0.7482,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.180410440072112,
+      "learning_rate": 1.8543165265096962e-05,
+      "loss": 0.793,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.74174721447432,
+      "learning_rate": 1.854228863254402e-05,
+      "loss": 0.7929,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.68191396786097,
+      "learning_rate": 1.854141175705338e-05,
+      "loss": 0.755,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.026995382528284,
+      "learning_rate": 1.8540534638649994e-05,
+      "loss": 0.8245,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.431292069860227,
+      "learning_rate": 1.8539657277358798e-05,
+      "loss": 0.7911,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.52814984093385,
+      "learning_rate": 1.8538779673204743e-05,
+      "loss": 0.8422,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.641874603503572,
+      "learning_rate": 1.8537901826212786e-05,
+      "loss": 0.7668,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.763300835906904,
+      "learning_rate": 1.8537023736407898e-05,
+      "loss": 0.7732,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.46968069224584,
+      "learning_rate": 1.853614540381505e-05,
+      "loss": 0.8077,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.594141751630684,
+      "learning_rate": 1.853526682845922e-05,
+      "loss": 0.8224,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.16988856553982,
+      "learning_rate": 1.8534388010365397e-05,
+      "loss": 0.839,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 64.85110000732195,
+      "learning_rate": 1.8533508949558564e-05,
+      "loss": 0.84,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.32852115215774,
+      "learning_rate": 1.8532629646063737e-05,
+      "loss": 0.8396,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.524113179146838,
+      "learning_rate": 1.853175009990591e-05,
+      "loss": 0.7346,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.23100621134276,
+      "learning_rate": 1.85308703111101e-05,
+      "loss": 0.7434,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.980683415202517,
+      "learning_rate": 1.852999027970133e-05,
+      "loss": 0.7483,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 30.459384500970334,
+      "learning_rate": 1.8529110005704627e-05,
+      "loss": 0.8256,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.785462742635673,
+      "learning_rate": 1.8528229489145023e-05,
+      "loss": 0.7698,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.966389276290089,
+      "learning_rate": 1.8527348730047563e-05,
+      "loss": 0.7426,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 25.438100229558053,
+      "learning_rate": 1.8526467728437292e-05,
+      "loss": 0.7875,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 30.353504072658048,
+      "learning_rate": 1.852558648433927e-05,
+      "loss": 0.762,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 24.14142603363089,
+      "learning_rate": 1.852470499777855e-05,
+      "loss": 0.919,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 21.39316143660563,
+      "learning_rate": 1.8523823268780214e-05,
+      "loss": 0.7851,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.969173797708438,
+      "learning_rate": 1.8522941297369324e-05,
+      "loss": 0.7629,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.20639261315958,
+      "learning_rate": 1.8522059083570972e-05,
+      "loss": 0.7311,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 29.990470734251467,
+      "learning_rate": 1.8521176627410246e-05,
+      "loss": 0.7724,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.021110580348523,
+      "learning_rate": 1.852029392891224e-05,
+      "loss": 0.8001,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.79630895066451,
+      "learning_rate": 1.8519410988102058e-05,
+      "loss": 0.7395,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.08821192734763,
+      "learning_rate": 1.8518527805004812e-05,
+      "loss": 0.7995,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.586739605506358,
+      "learning_rate": 1.851764437964562e-05,
+      "loss": 0.7433,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.41879667817107,
+      "learning_rate": 1.8516760712049603e-05,
+      "loss": 0.7605,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.800195746349994,
+      "learning_rate": 1.8515876802241894e-05,
+      "loss": 0.6754,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.735721712568434,
+      "learning_rate": 1.8514992650247627e-05,
+      "loss": 0.8175,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 25.940824373759135,
+      "learning_rate": 1.8514108256091953e-05,
+      "loss": 0.8489,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.821477867824576,
+      "learning_rate": 1.8513223619800022e-05,
+      "loss": 0.7689,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.62874136775155,
+      "learning_rate": 1.851233874139699e-05,
+      "loss": 0.8623,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.45067523042239,
+      "learning_rate": 1.8511453620908027e-05,
+      "loss": 0.7721,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.27249941593221,
+      "learning_rate": 1.8510568258358298e-05,
+      "loss": 0.7848,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.267452031925085,
+      "learning_rate": 1.850968265377299e-05,
+      "loss": 0.7377,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.04300267727044,
+      "learning_rate": 1.850879680717728e-05,
+      "loss": 0.8161,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.973969469370363,
+      "learning_rate": 1.850791071859637e-05,
+      "loss": 0.7655,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.764558773012604,
+      "learning_rate": 1.8507024388055456e-05,
+      "loss": 0.8755,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.076852204468675,
+      "learning_rate": 1.8506137815579744e-05,
+      "loss": 0.7986,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.58434569656616,
+      "learning_rate": 1.8505251001194445e-05,
+      "loss": 0.8154,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.40001003903859,
+      "learning_rate": 1.850436394492479e-05,
+      "loss": 0.7052,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.656556720356061,
+      "learning_rate": 1.8503476646795996e-05,
+      "loss": 0.6883,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.207146983455665,
+      "learning_rate": 1.85025891068333e-05,
+      "loss": 0.8264,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.353768880110001,
+      "learning_rate": 1.8501701325061943e-05,
+      "loss": 0.7402,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.248055625298326,
+      "learning_rate": 1.8500813301507174e-05,
+      "loss": 0.7408,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.465835975885266,
+      "learning_rate": 1.8499925036194247e-05,
+      "loss": 0.8808,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.006840873991008,
+      "learning_rate": 1.849903652914843e-05,
+      "loss": 0.785,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.900283448110795,
+      "learning_rate": 1.849814778039498e-05,
+      "loss": 0.8314,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 21.997378637458333,
+      "learning_rate": 1.849725878995918e-05,
+      "loss": 0.8069,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.670353523418115,
+      "learning_rate": 1.849636955786631e-05,
+      "loss": 0.7975,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.908022661833616,
+      "learning_rate": 1.8495480084141662e-05,
+      "loss": 0.8225,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 26.254595375864266,
+      "learning_rate": 1.849459036881053e-05,
+      "loss": 0.8323,
+      "step": 3873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 29.21850568801818,
+      "learning_rate": 1.8493700411898213e-05,
+      "loss": 0.8218,
+      "step": 3874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.41556248323299,
+      "learning_rate": 1.8492810213430027e-05,
+      "loss": 0.6852,
+      "step": 3875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 22.42568565445289,
+      "learning_rate": 1.849191977343129e-05,
+      "loss": 0.8137,
+      "step": 3876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.13855530497266,
+      "learning_rate": 1.849102909192732e-05,
+      "loss": 0.8085,
+      "step": 3877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.071614659543368,
+      "learning_rate": 1.849013816894345e-05,
+      "loss": 0.8206,
+      "step": 3878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.880815030354963,
+      "learning_rate": 1.8489247004505017e-05,
+      "loss": 0.8588,
+      "step": 3879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.932620809459976,
+      "learning_rate": 1.8488355598637362e-05,
+      "loss": 0.8097,
+      "step": 3880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.821513297094455,
+      "learning_rate": 1.8487463951365842e-05,
+      "loss": 0.6766,
+      "step": 3881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 25.381460687751073,
+      "learning_rate": 1.8486572062715814e-05,
+      "loss": 0.7676,
+      "step": 3882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.694321796201884,
+      "learning_rate": 1.848567993271264e-05,
+      "loss": 0.851,
+      "step": 3883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 23.91939712986684,
+      "learning_rate": 1.848478756138169e-05,
+      "loss": 0.7627,
+      "step": 3884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 27.575489992472782,
+      "learning_rate": 1.8483894948748348e-05,
+      "loss": 0.9076,
+      "step": 3885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.200503250736677,
+      "learning_rate": 1.8483002094837998e-05,
+      "loss": 0.7739,
+      "step": 3886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.171025317348672,
+      "learning_rate": 1.848210899967603e-05,
+      "loss": 0.7689,
+      "step": 3887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.639976547742176,
+      "learning_rate": 1.848121566328784e-05,
+      "loss": 0.7419,
+      "step": 3888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 25.321154784917912,
+      "learning_rate": 1.848032208569884e-05,
+      "loss": 0.8569,
+      "step": 3889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.313079087732586,
+      "learning_rate": 1.8479428266934442e-05,
+      "loss": 0.8096,
+      "step": 3890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.55765342883998,
+      "learning_rate": 1.8478534207020066e-05,
+      "loss": 0.7158,
+      "step": 3891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.29365973096387,
+      "learning_rate": 1.8477639905981135e-05,
+      "loss": 0.8127,
+      "step": 3892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 27.826071288049587,
+      "learning_rate": 1.8476745363843082e-05,
+      "loss": 0.7439,
+      "step": 3893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.669471236448274,
+      "learning_rate": 1.847585058063135e-05,
+      "loss": 0.8184,
+      "step": 3894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 27.851623443601042,
+      "learning_rate": 1.847495555637139e-05,
+      "loss": 0.7586,
+      "step": 3895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.347098516052323,
+      "learning_rate": 1.8474060291088646e-05,
+      "loss": 0.7504,
+      "step": 3896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.727692649101545,
+      "learning_rate": 1.8473164784808592e-05,
+      "loss": 0.8458,
+      "step": 3897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.36514862589803,
+      "learning_rate": 1.8472269037556683e-05,
+      "loss": 0.7923,
+      "step": 3898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 30.892618550586896,
+      "learning_rate": 1.84713730493584e-05,
+      "loss": 0.7549,
+      "step": 3899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.007721708458217,
+      "learning_rate": 1.847047682023922e-05,
+      "loss": 0.7426,
+      "step": 3900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.439207536485934,
+      "learning_rate": 1.8469580350224636e-05,
+      "loss": 0.7434,
+      "step": 3901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 30.340700094476045,
+      "learning_rate": 1.8468683639340145e-05,
+      "loss": 0.8639,
+      "step": 3902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.80755338434516,
+      "learning_rate": 1.846778668761124e-05,
+      "loss": 0.9562,
+      "step": 3903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.444557099233771,
+      "learning_rate": 1.846688949506344e-05,
+      "loss": 0.773,
+      "step": 3904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.873548635833972,
+      "learning_rate": 1.8465992061722255e-05,
+      "loss": 0.7355,
+      "step": 3905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.731886725897116,
+      "learning_rate": 1.8465094387613204e-05,
+      "loss": 0.8077,
+      "step": 3906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.668701540451185,
+      "learning_rate": 1.8464196472761825e-05,
+      "loss": 0.9629,
+      "step": 3907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.96158451594049,
+      "learning_rate": 1.8463298317193647e-05,
+      "loss": 0.8042,
+      "step": 3908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 22.66219790157908,
+      "learning_rate": 1.846239992093422e-05,
+      "loss": 0.7923,
+      "step": 3909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.93281560075687,
+      "learning_rate": 1.8461501284009085e-05,
+      "loss": 0.7495,
+      "step": 3910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.829182751826917,
+      "learning_rate": 1.8460602406443802e-05,
+      "loss": 0.7514,
+      "step": 3911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.564662925341402,
+      "learning_rate": 1.845970328826394e-05,
+      "loss": 0.81,
+      "step": 3912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.968870285685993,
+      "learning_rate": 1.8458803929495062e-05,
+      "loss": 0.8361,
+      "step": 3913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 27.28492253503119,
+      "learning_rate": 1.8457904330162752e-05,
+      "loss": 0.8537,
+      "step": 3914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.33909896703304,
+      "learning_rate": 1.8457004490292588e-05,
+      "loss": 0.7779,
+      "step": 3915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.2744878738717,
+      "learning_rate": 1.8456104409910165e-05,
+      "loss": 0.8338,
+      "step": 3916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 30.352666432201893,
+      "learning_rate": 1.845520408904108e-05,
+      "loss": 0.9,
+      "step": 3917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 26.65282672610522,
+      "learning_rate": 1.8454303527710934e-05,
+      "loss": 0.7512,
+      "step": 3918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.784527142814259,
+      "learning_rate": 1.8453402725945343e-05,
+      "loss": 0.7653,
+      "step": 3919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.254694785066972,
+      "learning_rate": 1.8452501683769924e-05,
+      "loss": 0.7724,
+      "step": 3920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.85627894130136,
+      "learning_rate": 1.8451600401210303e-05,
+      "loss": 0.7332,
+      "step": 3921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 26.047388374930797,
+      "learning_rate": 1.8450698878292106e-05,
+      "loss": 1.0138,
+      "step": 3922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.418785834027496,
+      "learning_rate": 1.844979711504098e-05,
+      "loss": 0.8606,
+      "step": 3923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 21.816517696932664,
+      "learning_rate": 1.8448895111482565e-05,
+      "loss": 0.8204,
+      "step": 3924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.214806980589778,
+      "learning_rate": 1.8447992867642518e-05,
+      "loss": 0.8153,
+      "step": 3925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.205753759461043,
+      "learning_rate": 1.8447090383546494e-05,
+      "loss": 0.689,
+      "step": 3926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.492803046007635,
+      "learning_rate": 1.8446187659220163e-05,
+      "loss": 0.698,
+      "step": 3927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.529158688303893,
+      "learning_rate": 1.8445284694689193e-05,
+      "loss": 0.8182,
+      "step": 3928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.343326230616558,
+      "learning_rate": 1.8444381489979267e-05,
+      "loss": 0.7947,
+      "step": 3929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.393176317521146,
+      "learning_rate": 1.8443478045116072e-05,
+      "loss": 0.7321,
+      "step": 3930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 22.799329624887452,
+      "learning_rate": 1.84425743601253e-05,
+      "loss": 0.7459,
+      "step": 3931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 23.987049691732953,
+      "learning_rate": 1.8441670435032654e-05,
+      "loss": 0.7579,
+      "step": 3932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.089735649065783,
+      "learning_rate": 1.8440766269863838e-05,
+      "loss": 0.7613,
+      "step": 3933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.310592131543064,
+      "learning_rate": 1.8439861864644567e-05,
+      "loss": 0.8116,
+      "step": 3934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 29.16116392434744,
+      "learning_rate": 1.843895721940056e-05,
+      "loss": 0.8648,
+      "step": 3935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 28.05762151270861,
+      "learning_rate": 1.843805233415755e-05,
+      "loss": 0.8286,
+      "step": 3936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.51066513673727,
+      "learning_rate": 1.8437147208941265e-05,
+      "loss": 0.8016,
+      "step": 3937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 28.732777431316652,
+      "learning_rate": 1.843624184377745e-05,
+      "loss": 0.7766,
+      "step": 3938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.46432995267984,
+      "learning_rate": 1.8435336238691853e-05,
+      "loss": 0.7877,
+      "step": 3939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 26.58450417602139,
+      "learning_rate": 1.8434430393710224e-05,
+      "loss": 0.8457,
+      "step": 3940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 22.97657099194777,
+      "learning_rate": 1.8433524308858334e-05,
+      "loss": 0.8222,
+      "step": 3941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 22.80400193802145,
+      "learning_rate": 1.843261798416194e-05,
+      "loss": 0.7666,
+      "step": 3942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.739879411047784,
+      "learning_rate": 1.8431711419646826e-05,
+      "loss": 0.7505,
+      "step": 3943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.493484900760645,
+      "learning_rate": 1.8430804615338774e-05,
+      "loss": 0.704,
+      "step": 3944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.62165709232176,
+      "learning_rate": 1.842989757126357e-05,
+      "loss": 0.8114,
+      "step": 3945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.9087229641836,
+      "learning_rate": 1.8428990287447006e-05,
+      "loss": 0.7421,
+      "step": 3946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.44888273325592,
+      "learning_rate": 1.8428082763914893e-05,
+      "loss": 0.7429,
+      "step": 3947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 27.407702641282988,
+      "learning_rate": 1.842717500069304e-05,
+      "loss": 0.7829,
+      "step": 3948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.696606699158387,
+      "learning_rate": 1.8426266997807252e-05,
+      "loss": 0.766,
+      "step": 3949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.182272009347923,
+      "learning_rate": 1.8425358755283365e-05,
+      "loss": 0.7823,
+      "step": 3950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.51357873522703,
+      "learning_rate": 1.84244502731472e-05,
+      "loss": 0.8084,
+      "step": 3951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.2355762588216,
+      "learning_rate": 1.84235415514246e-05,
+      "loss": 0.7781,
+      "step": 3952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.900462960225724,
+      "learning_rate": 1.8422632590141407e-05,
+      "loss": 0.7362,
+      "step": 3953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.374814147232982,
+      "learning_rate": 1.8421723389323468e-05,
+      "loss": 0.6842,
+      "step": 3954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.95554232832515,
+      "learning_rate": 1.8420813948996643e-05,
+      "loss": 0.7623,
+      "step": 3955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 26.167381137093408,
+      "learning_rate": 1.8419904269186797e-05,
+      "loss": 0.8981,
+      "step": 3956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 23.079421702367732,
+      "learning_rate": 1.8418994349919798e-05,
+      "loss": 0.9015,
+      "step": 3957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.286503494870637,
+      "learning_rate": 1.8418084191221522e-05,
+      "loss": 0.9776,
+      "step": 3958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.228907204678052,
+      "learning_rate": 1.8417173793117857e-05,
+      "loss": 0.7629,
+      "step": 3959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.949775742036135,
+      "learning_rate": 1.8416263155634695e-05,
+      "loss": 0.7898,
+      "step": 3960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 22.86129622296271,
+      "learning_rate": 1.8415352278797932e-05,
+      "loss": 0.8817,
+      "step": 3961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.527071654136954,
+      "learning_rate": 1.8414441162633473e-05,
+      "loss": 0.7815,
+      "step": 3962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.621064669147938,
+      "learning_rate": 1.8413529807167228e-05,
+      "loss": 0.7699,
+      "step": 3963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.06265569895257,
+      "learning_rate": 1.8412618212425116e-05,
+      "loss": 0.7837,
+      "step": 3964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 24.55817505382642,
+      "learning_rate": 1.841170637843307e-05,
+      "loss": 0.7721,
+      "step": 3965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.239676548274712,
+      "learning_rate": 1.8410794305217006e-05,
+      "loss": 0.7911,
+      "step": 3966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 22.03393289111414,
+      "learning_rate": 1.8409881992802882e-05,
+      "loss": 0.8064,
+      "step": 3967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 23.318176207464848,
+      "learning_rate": 1.8408969441216628e-05,
+      "loss": 0.8338,
+      "step": 3968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.12564241846807,
+      "learning_rate": 1.8408056650484204e-05,
+      "loss": 0.7723,
+      "step": 3969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.174953355677708,
+      "learning_rate": 1.8407143620631564e-05,
+      "loss": 0.7423,
+      "step": 3970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.970926900663347,
+      "learning_rate": 1.840623035168468e-05,
+      "loss": 0.7913,
+      "step": 3971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.744496493201806,
+      "learning_rate": 1.8405316843669524e-05,
+      "loss": 0.6713,
+      "step": 3972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 27.633855168561823,
+      "learning_rate": 1.840440309661207e-05,
+      "loss": 0.7697,
+      "step": 3973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.12806122126247,
+      "learning_rate": 1.8403489110538312e-05,
+      "loss": 0.7226,
+      "step": 3974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 29.946253078307834,
+      "learning_rate": 1.8402574885474238e-05,
+      "loss": 0.8742,
+      "step": 3975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 25.150682981912503,
+      "learning_rate": 1.840166042144585e-05,
+      "loss": 0.7331,
+      "step": 3976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.814120574315359,
+      "learning_rate": 1.8400745718479155e-05,
+      "loss": 0.8319,
+      "step": 3977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.955471699729365,
+      "learning_rate": 1.8399830776600164e-05,
+      "loss": 0.7965,
+      "step": 3978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.239781488610827,
+      "learning_rate": 1.83989155958349e-05,
+      "loss": 0.7555,
+      "step": 3979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 23.13886507426497,
+      "learning_rate": 1.839800017620939e-05,
+      "loss": 0.8598,
+      "step": 3980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.824127672720582,
+      "learning_rate": 1.8397084517749668e-05,
+      "loss": 0.8674,
+      "step": 3981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 24.81647318274528,
+      "learning_rate": 1.8396168620481772e-05,
+      "loss": 0.8192,
+      "step": 3982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.2916935686875,
+      "learning_rate": 1.8395252484431758e-05,
+      "loss": 0.8028,
+      "step": 3983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.95360975443012,
+      "learning_rate": 1.839433610962567e-05,
+      "loss": 0.8024,
+      "step": 3984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.174789954633832,
+      "learning_rate": 1.8393419496089577e-05,
+      "loss": 0.7694,
+      "step": 3985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 23.955312698652683,
+      "learning_rate": 1.839250264384954e-05,
+      "loss": 0.8709,
+      "step": 3986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 25.585178347674756,
+      "learning_rate": 1.839158555293164e-05,
+      "loss": 0.7702,
+      "step": 3987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.922154810201178,
+      "learning_rate": 1.839066822336195e-05,
+      "loss": 0.7046,
+      "step": 3988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.116644496070613,
+      "learning_rate": 1.8389750655166572e-05,
+      "loss": 0.7483,
+      "step": 3989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 31.005086219098988,
+      "learning_rate": 1.8388832848371588e-05,
+      "loss": 0.8069,
+      "step": 3990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.008058943035326,
+      "learning_rate": 1.8387914803003106e-05,
+      "loss": 0.7936,
+      "step": 3991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.647187100838746,
+      "learning_rate": 1.8386996519087236e-05,
+      "loss": 0.7547,
+      "step": 3992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.94987403505995,
+      "learning_rate": 1.8386077996650086e-05,
+      "loss": 0.7883,
+      "step": 3993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.975370734778773,
+      "learning_rate": 1.838515923571779e-05,
+      "loss": 0.8114,
+      "step": 3994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.02200142409825,
+      "learning_rate": 1.838424023631647e-05,
+      "loss": 0.8497,
+      "step": 3995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.750925009943543,
+      "learning_rate": 1.838332099847226e-05,
+      "loss": 0.6907,
+      "step": 3996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 27.149989611490202,
+      "learning_rate": 1.8382401522211304e-05,
+      "loss": 0.8797,
+      "step": 3997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 22.249574071512278,
+      "learning_rate": 1.838148180755975e-05,
+      "loss": 0.71,
+      "step": 3998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 25.632264021819047,
+      "learning_rate": 1.8380561854543763e-05,
+      "loss": 0.838,
+      "step": 3999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.405313161566008,
+      "learning_rate": 1.8379641663189493e-05,
+      "loss": 0.8205,
+      "step": 4000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.319739297686727,
+      "learning_rate": 1.837872123352312e-05,
+      "loss": 0.7407,
+      "step": 4001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.747272550442679,
+      "learning_rate": 1.837780056557082e-05,
+      "loss": 0.7633,
+      "step": 4002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.668417350866296,
+      "learning_rate": 1.8376879659358767e-05,
+      "loss": 0.7911,
+      "step": 4003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.531101403262607,
+      "learning_rate": 1.8375958514913158e-05,
+      "loss": 0.7659,
+      "step": 4004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.707772077855429,
+      "learning_rate": 1.8375037132260187e-05,
+      "loss": 0.7484,
+      "step": 4005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.826776165706086,
+      "learning_rate": 1.837411551142606e-05,
+      "loss": 0.8314,
+      "step": 4006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.591839304191762,
+      "learning_rate": 1.837319365243699e-05,
+      "loss": 0.7947,
+      "step": 4007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.507425071948727,
+      "learning_rate": 1.8372271555319185e-05,
+      "loss": 0.7932,
+      "step": 4008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 27.13151214310836,
+      "learning_rate": 1.837134922009888e-05,
+      "loss": 0.802,
+      "step": 4009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.195559047098417,
+      "learning_rate": 1.8370426646802298e-05,
+      "loss": 0.7496,
+      "step": 4010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 26.029585745047147,
+      "learning_rate": 1.8369503835455676e-05,
+      "loss": 0.9137,
+      "step": 4011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.172161446902468,
+      "learning_rate": 1.8368580786085265e-05,
+      "loss": 0.7937,
+      "step": 4012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.169397638376193,
+      "learning_rate": 1.836765749871731e-05,
+      "loss": 0.8181,
+      "step": 4013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.25379491607554,
+      "learning_rate": 1.8366733973378074e-05,
+      "loss": 0.8097,
+      "step": 4014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.523381642616236,
+      "learning_rate": 1.8365810210093814e-05,
+      "loss": 0.7363,
+      "step": 4015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.965855560785187,
+      "learning_rate": 1.836488620889081e-05,
+      "loss": 0.8354,
+      "step": 4016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.242756028629827,
+      "learning_rate": 1.836396196979533e-05,
+      "loss": 0.8586,
+      "step": 4017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 27.550900611148332,
+      "learning_rate": 1.8363037492833667e-05,
+      "loss": 0.9081,
+      "step": 4018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 27.709086276462795,
+      "learning_rate": 1.836211277803211e-05,
+      "loss": 0.7596,
+      "step": 4019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.05138982784889,
+      "learning_rate": 1.836118782541696e-05,
+      "loss": 0.7134,
+      "step": 4020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 44.184207096855985,
+      "learning_rate": 1.8360262635014513e-05,
+      "loss": 0.8022,
+      "step": 4021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.15848588444529,
+      "learning_rate": 1.8359337206851094e-05,
+      "loss": 0.7862,
+      "step": 4022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 32.793212961511045,
+      "learning_rate": 1.8358411540953008e-05,
+      "loss": 0.8714,
+      "step": 4023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 26.86531251475134,
+      "learning_rate": 1.835748563734659e-05,
+      "loss": 0.841,
+      "step": 4024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.677800799251358,
+      "learning_rate": 1.8356559496058172e-05,
+      "loss": 0.7518,
+      "step": 4025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.5701404761941,
+      "learning_rate": 1.835563311711409e-05,
+      "loss": 0.8105,
+      "step": 4026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.806335132849288,
+      "learning_rate": 1.835470650054069e-05,
+      "loss": 0.7603,
+      "step": 4027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.27433458205608,
+      "learning_rate": 1.8353779646364323e-05,
+      "loss": 0.822,
+      "step": 4028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.796653723875387,
+      "learning_rate": 1.835285255461135e-05,
+      "loss": 0.9318,
+      "step": 4029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 23.65776073187368,
+      "learning_rate": 1.8351925225308136e-05,
+      "loss": 0.8439,
+      "step": 4030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.733131728483166,
+      "learning_rate": 1.8350997658481054e-05,
+      "loss": 0.7087,
+      "step": 4031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.27642042085711,
+      "learning_rate": 1.835006985415649e-05,
+      "loss": 0.7264,
+      "step": 4032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.554432419541385,
+      "learning_rate": 1.8349141812360815e-05,
+      "loss": 0.8506,
+      "step": 4033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.956954534626647,
+      "learning_rate": 1.8348213533120437e-05,
+      "loss": 0.7942,
+      "step": 4034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.273529697752117,
+      "learning_rate": 1.8347285016461745e-05,
+      "loss": 0.8884,
+      "step": 4035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 23.249950737670268,
+      "learning_rate": 1.8346356262411157e-05,
+      "loss": 0.7631,
+      "step": 4036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.02906163759453,
+      "learning_rate": 1.8345427270995072e-05,
+      "loss": 0.7787,
+      "step": 4037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.05649738565577,
+      "learning_rate": 1.8344498042239922e-05,
+      "loss": 0.7353,
+      "step": 4038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.38100596628698,
+      "learning_rate": 1.8343568576172127e-05,
+      "loss": 0.7436,
+      "step": 4039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.546504893730834,
+      "learning_rate": 1.834263887281812e-05,
+      "loss": 0.82,
+      "step": 4040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.93560101889014,
+      "learning_rate": 1.834170893220435e-05,
+      "loss": 0.6939,
+      "step": 4041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.58899066922581,
+      "learning_rate": 1.8340778754357252e-05,
+      "loss": 0.7755,
+      "step": 4042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 25.40402393981906,
+      "learning_rate": 1.833984833930329e-05,
+      "loss": 0.8215,
+      "step": 4043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.259975279949824,
+      "learning_rate": 1.8338917687068916e-05,
+      "loss": 0.7213,
+      "step": 4044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.97638652598689,
+      "learning_rate": 1.8337986797680604e-05,
+      "loss": 0.8617,
+      "step": 4045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.826331171398678,
+      "learning_rate": 1.8337055671164826e-05,
+      "loss": 0.7763,
+      "step": 4046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.21938759892234,
+      "learning_rate": 1.833612430754806e-05,
+      "loss": 0.7074,
+      "step": 4047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.975197932986642,
+      "learning_rate": 1.8335192706856792e-05,
+      "loss": 0.7524,
+      "step": 4048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.644701512975637,
+      "learning_rate": 1.8334260869117524e-05,
+      "loss": 0.8038,
+      "step": 4049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.40626696966864,
+      "learning_rate": 1.833332879435675e-05,
+      "loss": 0.7759,
+      "step": 4050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 29.939665329817146,
+      "learning_rate": 1.833239648260098e-05,
+      "loss": 0.7455,
+      "step": 4051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.847727211801143,
+      "learning_rate": 1.833146393387673e-05,
+      "loss": 0.7344,
+      "step": 4052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.72879469205248,
+      "learning_rate": 1.8330531148210516e-05,
+      "loss": 0.8269,
+      "step": 4053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.034760398827235,
+      "learning_rate": 1.8329598125628873e-05,
+      "loss": 0.79,
+      "step": 4054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.130774426903695,
+      "learning_rate": 1.8328664866158332e-05,
+      "loss": 0.7332,
+      "step": 4055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.734666720286576,
+      "learning_rate": 1.8327731369825432e-05,
+      "loss": 0.788,
+      "step": 4056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.344273597094165,
+      "learning_rate": 1.8326797636656723e-05,
+      "loss": 0.8264,
+      "step": 4057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.630659119575697,
+      "learning_rate": 1.8325863666678764e-05,
+      "loss": 0.7591,
+      "step": 4058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.55776862643885,
+      "learning_rate": 1.832492945991811e-05,
+      "loss": 0.7203,
+      "step": 4059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.70146592654616,
+      "learning_rate": 1.8323995016401332e-05,
+      "loss": 0.7348,
+      "step": 4060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.597404789158826,
+      "learning_rate": 1.8323060336155008e-05,
+      "loss": 0.7367,
+      "step": 4061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.41308691752615,
+      "learning_rate": 1.8322125419205713e-05,
+      "loss": 0.7802,
+      "step": 4062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 23.890983435397562,
+      "learning_rate": 1.832119026558004e-05,
+      "loss": 0.8569,
+      "step": 4063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.789768444191667,
+      "learning_rate": 1.8320254875304585e-05,
+      "loss": 0.7193,
+      "step": 4064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.398659155709403,
+      "learning_rate": 1.831931924840595e-05,
+      "loss": 0.6454,
+      "step": 4065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.721580559436095,
+      "learning_rate": 1.831838338491074e-05,
+      "loss": 0.7313,
+      "step": 4066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.189410233777284,
+      "learning_rate": 1.8317447284845573e-05,
+      "loss": 0.7123,
+      "step": 4067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 25.44081514018268,
+      "learning_rate": 1.831651094823707e-05,
+      "loss": 0.7311,
+      "step": 4068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.73241113633123,
+      "learning_rate": 1.831557437511186e-05,
+      "loss": 0.8133,
+      "step": 4069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.433668434747723,
+      "learning_rate": 1.8314637565496584e-05,
+      "loss": 0.6534,
+      "step": 4070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 30.62790181026401,
+      "learning_rate": 1.8313700519417876e-05,
+      "loss": 0.8467,
+      "step": 4071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.14620854273327,
+      "learning_rate": 1.8312763236902387e-05,
+      "loss": 0.8438,
+      "step": 4072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.66952688602762,
+      "learning_rate": 1.831182571797678e-05,
+      "loss": 0.7178,
+      "step": 4073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.692903620687968,
+      "learning_rate": 1.8310887962667706e-05,
+      "loss": 0.7031,
+      "step": 4074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.533743314309838,
+      "learning_rate": 1.830994997100184e-05,
+      "loss": 0.884,
+      "step": 4075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.27041583951733,
+      "learning_rate": 1.8309011743005863e-05,
+      "loss": 0.7584,
+      "step": 4076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 86.65536448497078,
+      "learning_rate": 1.8308073278706448e-05,
+      "loss": 0.7623,
+      "step": 4077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.39951506978944,
+      "learning_rate": 1.8307134578130293e-05,
+      "loss": 0.7753,
+      "step": 4078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.932615527768895,
+      "learning_rate": 1.8306195641304088e-05,
+      "loss": 0.8052,
+      "step": 4079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.343993609897169,
+      "learning_rate": 1.8305256468254537e-05,
+      "loss": 0.8088,
+      "step": 4080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.93329382866461,
+      "learning_rate": 1.830431705900835e-05,
+      "loss": 0.7737,
+      "step": 4081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.29276556553434,
+      "learning_rate": 1.8303377413592248e-05,
+      "loss": 0.6871,
+      "step": 4082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.800445241228395,
+      "learning_rate": 1.8302437532032945e-05,
+      "loss": 0.7648,
+      "step": 4083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 21.153387648377237,
+      "learning_rate": 1.830149741435718e-05,
+      "loss": 0.8411,
+      "step": 4084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.861353873533375,
+      "learning_rate": 1.830055706059168e-05,
+      "loss": 0.8067,
+      "step": 4085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.48304642508233,
+      "learning_rate": 1.8299616470763196e-05,
+      "loss": 0.7107,
+      "step": 4086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.179646688753063,
+      "learning_rate": 1.829867564489847e-05,
+      "loss": 0.9648,
+      "step": 4087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 28.52022475833779,
+      "learning_rate": 1.8297734583024273e-05,
+      "loss": 0.808,
+      "step": 4088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.980345340979714,
+      "learning_rate": 1.8296793285167348e-05,
+      "loss": 0.6603,
+      "step": 4089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 24.50393409247012,
+      "learning_rate": 1.8295851751354482e-05,
+      "loss": 0.7467,
+      "step": 4090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.699693493333474,
+      "learning_rate": 1.8294909981612443e-05,
+      "loss": 0.7744,
+      "step": 4091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 24.123405910362294,
+      "learning_rate": 1.8293967975968016e-05,
+      "loss": 0.7457,
+      "step": 4092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.135385663796493,
+      "learning_rate": 1.8293025734447992e-05,
+      "loss": 0.7948,
+      "step": 4093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.12608952194189,
+      "learning_rate": 1.8292083257079166e-05,
+      "loss": 0.7302,
+      "step": 4094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.64810996948637,
+      "learning_rate": 1.8291140543888347e-05,
+      "loss": 0.8145,
+      "step": 4095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.120497737278335,
+      "learning_rate": 1.8290197594902337e-05,
+      "loss": 0.7315,
+      "step": 4096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.403996646144535,
+      "learning_rate": 1.8289254410147958e-05,
+      "loss": 0.8329,
+      "step": 4097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.843392320640191,
+      "learning_rate": 1.8288310989652035e-05,
+      "loss": 0.7139,
+      "step": 4098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.479391158151,
+      "learning_rate": 1.828736733344139e-05,
+      "loss": 0.7458,
+      "step": 4099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.982364150750014,
+      "learning_rate": 1.8286423441542874e-05,
+      "loss": 0.7826,
+      "step": 4100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.865509935831454,
+      "learning_rate": 1.828547931398332e-05,
+      "loss": 0.7012,
+      "step": 4101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.914275430842952,
+      "learning_rate": 1.8284534950789578e-05,
+      "loss": 0.7364,
+      "step": 4102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.00571356771551,
+      "learning_rate": 1.828359035198851e-05,
+      "loss": 0.7736,
+      "step": 4103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 29.542125641956837,
+      "learning_rate": 1.8282645517606977e-05,
+      "loss": 0.9103,
+      "step": 4104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.097036599178933,
+      "learning_rate": 1.8281700447671852e-05,
+      "loss": 0.8486,
+      "step": 4105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.684310183173224,
+      "learning_rate": 1.828075514221001e-05,
+      "loss": 0.6891,
+      "step": 4106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.84088828107749,
+      "learning_rate": 1.8279809601248342e-05,
+      "loss": 0.831,
+      "step": 4107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.74038264025481,
+      "learning_rate": 1.8278863824813726e-05,
+      "loss": 0.8437,
+      "step": 4108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.528611614866424,
+      "learning_rate": 1.8277917812933068e-05,
+      "loss": 0.7473,
+      "step": 4109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 28.163180080034113,
+      "learning_rate": 1.827697156563327e-05,
+      "loss": 0.8711,
+      "step": 4110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.354580146850937,
+      "learning_rate": 1.827602508294124e-05,
+      "loss": 0.7358,
+      "step": 4111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.790304353132047,
+      "learning_rate": 1.82750783648839e-05,
+      "loss": 0.7273,
+      "step": 4112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 33.003127082226484,
+      "learning_rate": 1.8274131411488172e-05,
+      "loss": 0.8199,
+      "step": 4113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.171991094486625,
+      "learning_rate": 1.8273184222780984e-05,
+      "loss": 0.794,
+      "step": 4114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.696733095273618,
+      "learning_rate": 1.827223679878928e-05,
+      "loss": 0.7188,
+      "step": 4115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.425840492548925,
+      "learning_rate": 1.8271289139539997e-05,
+      "loss": 0.8428,
+      "step": 4116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 26.09084472929248,
+      "learning_rate": 1.8270341245060095e-05,
+      "loss": 0.822,
+      "step": 4117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 21.6793608611648,
+      "learning_rate": 1.8269393115376522e-05,
+      "loss": 0.8358,
+      "step": 4118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 36.155094100226705,
+      "learning_rate": 1.8268444750516246e-05,
+      "loss": 0.7757,
+      "step": 4119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.56333594133736,
+      "learning_rate": 1.8267496150506242e-05,
+      "loss": 0.706,
+      "step": 4120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 25.576181748058023,
+      "learning_rate": 1.826654731537348e-05,
+      "loss": 0.8794,
+      "step": 4121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 27.442736306317713,
+      "learning_rate": 1.826559824514495e-05,
+      "loss": 0.8858,
+      "step": 4122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 28.96593825882559,
+      "learning_rate": 1.8264648939847642e-05,
+      "loss": 0.9125,
+      "step": 4123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.547110085785135,
+      "learning_rate": 1.826369939950855e-05,
+      "loss": 0.7612,
+      "step": 4124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.927598558929994,
+      "learning_rate": 1.8262749624154684e-05,
+      "loss": 0.8409,
+      "step": 4125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 33.57686183261153,
+      "learning_rate": 1.8261799613813053e-05,
+      "loss": 0.7878,
+      "step": 4126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.26701776238318,
+      "learning_rate": 1.826084936851067e-05,
+      "loss": 0.6831,
+      "step": 4127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.320788959766865,
+      "learning_rate": 1.8259898888274564e-05,
+      "loss": 0.7782,
+      "step": 4128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.862091160834396,
+      "learning_rate": 1.8258948173131766e-05,
+      "loss": 0.7341,
+      "step": 4129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.021932432436662,
+      "learning_rate": 1.8257997223109317e-05,
+      "loss": 0.7314,
+      "step": 4130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.693422991939816,
+      "learning_rate": 1.825704603823425e-05,
+      "loss": 0.7429,
+      "step": 4131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.957814172696747,
+      "learning_rate": 1.8256094618533633e-05,
+      "loss": 0.8141,
+      "step": 4132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.163449491609454,
+      "learning_rate": 1.825514296403451e-05,
+      "loss": 0.7616,
+      "step": 4133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.848775350195158,
+      "learning_rate": 1.825419107476395e-05,
+      "loss": 0.8326,
+      "step": 4134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.569001354101406,
+      "learning_rate": 1.8253238950749025e-05,
+      "loss": 0.8025,
+      "step": 4135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.710723629341924,
+      "learning_rate": 1.8252286592016812e-05,
+      "loss": 0.8281,
+      "step": 4136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.947511865867735,
+      "learning_rate": 1.8251333998594395e-05,
+      "loss": 0.7433,
+      "step": 4137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.028457690058588,
+      "learning_rate": 1.825038117050887e-05,
+      "loss": 0.7615,
+      "step": 4138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.60533198398627,
+      "learning_rate": 1.824942810778733e-05,
+      "loss": 0.7839,
+      "step": 4139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.996097847880318,
+      "learning_rate": 1.8248474810456874e-05,
+      "loss": 0.8439,
+      "step": 4140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.433477798959455,
+      "learning_rate": 1.8247521278544625e-05,
+      "loss": 0.7769,
+      "step": 4141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.530227256158849,
+      "learning_rate": 1.8246567512077694e-05,
+      "loss": 0.7159,
+      "step": 4142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.970803228492773,
+      "learning_rate": 1.824561351108321e-05,
+      "loss": 0.7677,
+      "step": 4143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.85053833675316,
+      "learning_rate": 1.8244659275588298e-05,
+      "loss": 0.8731,
+      "step": 4144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 24.81336237324034,
+      "learning_rate": 1.82437048056201e-05,
+      "loss": 0.7852,
+      "step": 4145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.01315384405582,
+      "learning_rate": 1.8242750101205757e-05,
+      "loss": 0.7627,
+      "step": 4146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.354121519100417,
+      "learning_rate": 1.8241795162372426e-05,
+      "loss": 0.8619,
+      "step": 4147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.814955547995996,
+      "learning_rate": 1.8240839989147264e-05,
+      "loss": 0.6823,
+      "step": 4148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.764312652754464,
+      "learning_rate": 1.823988458155743e-05,
+      "loss": 0.7452,
+      "step": 4149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.218329197629386,
+      "learning_rate": 1.8238928939630097e-05,
+      "loss": 0.7756,
+      "step": 4150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 24.214087248146637,
+      "learning_rate": 1.823797306339245e-05,
+      "loss": 0.7437,
+      "step": 4151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 27.496878017594884,
+      "learning_rate": 1.8237016952871664e-05,
+      "loss": 0.7958,
+      "step": 4152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.105387664446674,
+      "learning_rate": 1.8236060608094938e-05,
+      "loss": 0.7972,
+      "step": 4153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.260814013339807,
+      "learning_rate": 1.8235104029089465e-05,
+      "loss": 0.8071,
+      "step": 4154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.197953402440604,
+      "learning_rate": 1.823414721588245e-05,
+      "loss": 0.7347,
+      "step": 4155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 24.597249801764658,
+      "learning_rate": 1.8233190168501104e-05,
+      "loss": 0.7891,
+      "step": 4156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.397370830305267,
+      "learning_rate": 1.8232232886972643e-05,
+      "loss": 0.7062,
+      "step": 4157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.84046343709538,
+      "learning_rate": 1.8231275371324303e-05,
+      "loss": 0.696,
+      "step": 4158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.091233622436835,
+      "learning_rate": 1.82303176215833e-05,
+      "loss": 0.7187,
+      "step": 4159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.506800333979673,
+      "learning_rate": 1.8229359637776883e-05,
+      "loss": 0.746,
+      "step": 4160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.212309330351744,
+      "learning_rate": 1.8228401419932287e-05,
+      "loss": 0.8117,
+      "step": 4161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.157183395940265,
+      "learning_rate": 1.822744296807677e-05,
+      "loss": 0.7365,
+      "step": 4162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 31.04302911307338,
+      "learning_rate": 1.8226484282237587e-05,
+      "loss": 0.768,
+      "step": 4163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.208916504686204,
+      "learning_rate": 1.822552536244201e-05,
+      "loss": 0.6897,
+      "step": 4164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 28.746442931343466,
+      "learning_rate": 1.8224566208717296e-05,
+      "loss": 0.7417,
+      "step": 4165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.255061158054005,
+      "learning_rate": 1.8223606821090733e-05,
+      "loss": 0.7198,
+      "step": 4166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.535106175537841,
+      "learning_rate": 1.82226471995896e-05,
+      "loss": 0.7751,
+      "step": 4167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.6580760161256,
+      "learning_rate": 1.8221687344241193e-05,
+      "loss": 0.7885,
+      "step": 4168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.306401910524993,
+      "learning_rate": 1.822072725507281e-05,
+      "loss": 0.7807,
+      "step": 4169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.240129507863395,
+      "learning_rate": 1.8219766932111747e-05,
+      "loss": 0.7656,
+      "step": 4170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.08386420108185,
+      "learning_rate": 1.8218806375385326e-05,
+      "loss": 0.8176,
+      "step": 4171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.282687719982253,
+      "learning_rate": 1.8217845584920858e-05,
+      "loss": 0.7178,
+      "step": 4172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.343011743673248,
+      "learning_rate": 1.821688456074567e-05,
+      "loss": 0.757,
+      "step": 4173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.277970841041398,
+      "learning_rate": 1.8215923302887085e-05,
+      "loss": 0.8011,
+      "step": 4174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 70.69089067862052,
+      "learning_rate": 1.8214961811372455e-05,
+      "loss": 0.8367,
+      "step": 4175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.88569046519337,
+      "learning_rate": 1.8214000086229115e-05,
+      "loss": 0.7145,
+      "step": 4176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.862646877911239,
+      "learning_rate": 1.8213038127484415e-05,
+      "loss": 0.724,
+      "step": 4177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.664314450660363,
+      "learning_rate": 1.821207593516572e-05,
+      "loss": 0.8284,
+      "step": 4178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 21.66913914893567,
+      "learning_rate": 1.8211113509300384e-05,
+      "loss": 0.9162,
+      "step": 4179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.477847513495092,
+      "learning_rate": 1.8210150849915787e-05,
+      "loss": 0.7381,
+      "step": 4180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 34.222411037368175,
+      "learning_rate": 1.82091879570393e-05,
+      "loss": 0.7785,
+      "step": 4181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 24.01902531121858,
+      "learning_rate": 1.8208224830698312e-05,
+      "loss": 0.8747,
+      "step": 4182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.111924042495076,
+      "learning_rate": 1.820726147092021e-05,
+      "loss": 0.849,
+      "step": 4183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.247972078479735,
+      "learning_rate": 1.8206297877732393e-05,
+      "loss": 0.6994,
+      "step": 4184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.675256512218837,
+      "learning_rate": 1.8205334051162264e-05,
+      "loss": 0.7962,
+      "step": 4185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.884641425402634,
+      "learning_rate": 1.8204369991237237e-05,
+      "loss": 0.8158,
+      "step": 4186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 21.349592604424227,
+      "learning_rate": 1.8203405697984726e-05,
+      "loss": 0.7835,
+      "step": 4187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.53768262207262,
+      "learning_rate": 1.8202441171432156e-05,
+      "loss": 0.8492,
+      "step": 4188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.984575034237253,
+      "learning_rate": 1.8201476411606956e-05,
+      "loss": 0.7321,
+      "step": 4189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.785201187460796,
+      "learning_rate": 1.8200511418536563e-05,
+      "loss": 0.6348,
+      "step": 4190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.782846553680805,
+      "learning_rate": 1.819954619224843e-05,
+      "loss": 0.9021,
+      "step": 4191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 25.22696632069129,
+      "learning_rate": 1.8198580732769992e-05,
+      "loss": 0.8692,
+      "step": 4192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.90325030911037,
+      "learning_rate": 1.819761504012872e-05,
+      "loss": 0.8521,
+      "step": 4193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.209610022450809,
+      "learning_rate": 1.8196649114352065e-05,
+      "loss": 0.8209,
+      "step": 4194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.128908647560987,
+      "learning_rate": 1.819568295546751e-05,
+      "loss": 0.7317,
+      "step": 4195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.094954425899743,
+      "learning_rate": 1.819471656350252e-05,
+      "loss": 0.7932,
+      "step": 4196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 22.65058012686775,
+      "learning_rate": 1.819374993848459e-05,
+      "loss": 0.8494,
+      "step": 4197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 24.688260305744706,
+      "learning_rate": 1.8192783080441204e-05,
+      "loss": 0.8413,
+      "step": 4198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.12246524496961,
+      "learning_rate": 1.819181598939986e-05,
+      "loss": 0.7947,
+      "step": 4199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.72511579142144,
+      "learning_rate": 1.819084866538806e-05,
+      "loss": 0.8739,
+      "step": 4200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.70355082401617,
+      "learning_rate": 1.8189881108433317e-05,
+      "loss": 0.8834,
+      "step": 4201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.189093626553777,
+      "learning_rate": 1.8188913318563145e-05,
+      "loss": 0.8128,
+      "step": 4202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 22.59124555399027,
+      "learning_rate": 1.8187945295805066e-05,
+      "loss": 0.6995,
+      "step": 4203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.390925392819383,
+      "learning_rate": 1.8186977040186615e-05,
+      "loss": 0.8919,
+      "step": 4204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.713384502269177,
+      "learning_rate": 1.8186008551735327e-05,
+      "loss": 0.7108,
+      "step": 4205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.107845542862567,
+      "learning_rate": 1.8185039830478745e-05,
+      "loss": 0.8084,
+      "step": 4206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.85657301601761,
+      "learning_rate": 1.8184070876444417e-05,
+      "loss": 0.7317,
+      "step": 4207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 24.990322371369285,
+      "learning_rate": 1.8183101689659902e-05,
+      "loss": 0.7218,
+      "step": 4208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.208851136761886,
+      "learning_rate": 1.818213227015276e-05,
+      "loss": 0.7399,
+      "step": 4209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.142991538272394,
+      "learning_rate": 1.8181162617950563e-05,
+      "loss": 0.76,
+      "step": 4210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 22.846959313289375,
+      "learning_rate": 1.8180192733080887e-05,
+      "loss": 0.7868,
+      "step": 4211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.90326480343949,
+      "learning_rate": 1.817922261557132e-05,
+      "loss": 0.7735,
+      "step": 4212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.493681543379306,
+      "learning_rate": 1.817825226544944e-05,
+      "loss": 0.8084,
+      "step": 4213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.45435698397495,
+      "learning_rate": 1.817728168274285e-05,
+      "loss": 0.8268,
+      "step": 4214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 22.28494940819731,
+      "learning_rate": 1.8176310867479155e-05,
+      "loss": 0.7959,
+      "step": 4215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.652858458670133,
+      "learning_rate": 1.8175339819685965e-05,
+      "loss": 0.6771,
+      "step": 4216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.71053667378346,
+      "learning_rate": 1.8174368539390888e-05,
+      "loss": 0.7457,
+      "step": 4217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.2796520327813,
+      "learning_rate": 1.8173397026621554e-05,
+      "loss": 0.7368,
+      "step": 4218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.584496892201871,
+      "learning_rate": 1.817242528140559e-05,
+      "loss": 0.7424,
+      "step": 4219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.700335026067325,
+      "learning_rate": 1.8171453303770628e-05,
+      "loss": 0.8336,
+      "step": 4220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.60266369823785,
+      "learning_rate": 1.8170481093744317e-05,
+      "loss": 0.7677,
+      "step": 4221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.796280385952652,
+      "learning_rate": 1.81695086513543e-05,
+      "loss": 0.825,
+      "step": 4222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.339328609160805,
+      "learning_rate": 1.8168535976628242e-05,
+      "loss": 0.8076,
+      "step": 4223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.664036607996508,
+      "learning_rate": 1.8167563069593796e-05,
+      "loss": 0.8796,
+      "step": 4224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.685501119311006,
+      "learning_rate": 1.8166589930278637e-05,
+      "loss": 0.8069,
+      "step": 4225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.0908378763667,
+      "learning_rate": 1.816561655871043e-05,
+      "loss": 0.731,
+      "step": 4226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 33.870150718719664,
+      "learning_rate": 1.816464295491687e-05,
+      "loss": 0.6768,
+      "step": 4227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.729729567852427,
+      "learning_rate": 1.816366911892564e-05,
+      "loss": 0.7056,
+      "step": 4228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 25.702781940478168,
+      "learning_rate": 1.8162695050764437e-05,
+      "loss": 0.8006,
+      "step": 4229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 46.41708567786615,
+      "learning_rate": 1.8161720750460963e-05,
+      "loss": 0.771,
+      "step": 4230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 19.36104449712154,
+      "learning_rate": 1.8160746218042923e-05,
+      "loss": 0.7014,
+      "step": 4231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.573114581859631,
+      "learning_rate": 1.8159771453538034e-05,
+      "loss": 0.6727,
+      "step": 4232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.327036339677111,
+      "learning_rate": 1.8158796456974023e-05,
+      "loss": 0.7674,
+      "step": 4233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.592549386565253,
+      "learning_rate": 1.8157821228378607e-05,
+      "loss": 0.8118,
+      "step": 4234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.06741184879512,
+      "learning_rate": 1.8156845767779532e-05,
+      "loss": 0.8578,
+      "step": 4235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 19.853984133347314,
+      "learning_rate": 1.815587007520453e-05,
+      "loss": 0.7859,
+      "step": 4236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.56419704343956,
+      "learning_rate": 1.815489415068136e-05,
+      "loss": 0.7317,
+      "step": 4237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.464204674312242,
+      "learning_rate": 1.8153917994237765e-05,
+      "loss": 0.7163,
+      "step": 4238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 25.546874955670187,
+      "learning_rate": 1.8152941605901518e-05,
+      "loss": 0.8373,
+      "step": 4239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.286881393131456,
+      "learning_rate": 1.8151964985700375e-05,
+      "loss": 0.7966,
+      "step": 4240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.69725180837408,
+      "learning_rate": 1.815098813366212e-05,
+      "loss": 0.6484,
+      "step": 4241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.485323425701363,
+      "learning_rate": 1.8150011049814534e-05,
+      "loss": 0.7915,
+      "step": 4242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.325577534439056,
+      "learning_rate": 1.8149033734185396e-05,
+      "loss": 0.7345,
+      "step": 4243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.35399091586787,
+      "learning_rate": 1.814805618680251e-05,
+      "loss": 0.8457,
+      "step": 4244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 26.326481407887844,
+      "learning_rate": 1.814707840769367e-05,
+      "loss": 0.796,
+      "step": 4245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 19.757716482222545,
+      "learning_rate": 1.8146100396886685e-05,
+      "loss": 0.7758,
+      "step": 4246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.672847433984025,
+      "learning_rate": 1.814512215440937e-05,
+      "loss": 0.7704,
+      "step": 4247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 19.76956626890496,
+      "learning_rate": 1.8144143680289547e-05,
+      "loss": 0.777,
+      "step": 4248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.811800221589184,
+      "learning_rate": 1.8143164974555042e-05,
+      "loss": 0.724,
+      "step": 4249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 19.37250055988275,
+      "learning_rate": 1.8142186037233687e-05,
+      "loss": 0.8216,
+      "step": 4250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.467740504008308,
+      "learning_rate": 1.8141206868353327e-05,
+      "loss": 0.7109,
+      "step": 4251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.09127753365462,
+      "learning_rate": 1.8140227467941803e-05,
+      "loss": 0.6957,
+      "step": 4252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.512332261362477,
+      "learning_rate": 1.8139247836026977e-05,
+      "loss": 0.8107,
+      "step": 4253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.053714789069293,
+      "learning_rate": 1.8138267972636697e-05,
+      "loss": 0.8666,
+      "step": 4254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 27.58978856186665,
+      "learning_rate": 1.8137287877798842e-05,
+      "loss": 0.8388,
+      "step": 4255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.86409343623787,
+      "learning_rate": 1.8136307551541276e-05,
+      "loss": 0.8195,
+      "step": 4256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.1160706365079,
+      "learning_rate": 1.8135326993891886e-05,
+      "loss": 0.8557,
+      "step": 4257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.49182803457618,
+      "learning_rate": 1.8134346204878553e-05,
+      "loss": 0.8223,
+      "step": 4258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.701361360959423,
+      "learning_rate": 1.8133365184529176e-05,
+      "loss": 0.7515,
+      "step": 4259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 31.154165733893155,
+      "learning_rate": 1.8132383932871645e-05,
+      "loss": 0.6681,
+      "step": 4260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.96153690202711,
+      "learning_rate": 1.8131402449933875e-05,
+      "loss": 0.7608,
+      "step": 4261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.93937119749214,
+      "learning_rate": 1.813042073574378e-05,
+      "loss": 0.7109,
+      "step": 4262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.42305734839047,
+      "learning_rate": 1.8129438790329272e-05,
+      "loss": 0.7843,
+      "step": 4263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.325273479441474,
+      "learning_rate": 1.8128456613718278e-05,
+      "loss": 0.7861,
+      "step": 4264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.82262801814742,
+      "learning_rate": 1.812747420593874e-05,
+      "loss": 0.7791,
+      "step": 4265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.475188254677505,
+      "learning_rate": 1.8126491567018582e-05,
+      "loss": 0.7828,
+      "step": 4266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.533285579654216,
+      "learning_rate": 1.8125508696985763e-05,
+      "loss": 0.8698,
+      "step": 4267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.120126518310595,
+      "learning_rate": 1.8124525595868228e-05,
+      "loss": 0.7757,
+      "step": 4268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.919112486785913,
+      "learning_rate": 1.8123542263693935e-05,
+      "loss": 0.7934,
+      "step": 4269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.94620908456108,
+      "learning_rate": 1.8122558700490857e-05,
+      "loss": 0.7079,
+      "step": 4270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.697023743339116,
+      "learning_rate": 1.812157490628696e-05,
+      "loss": 0.7158,
+      "step": 4271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.91522053125438,
+      "learning_rate": 1.812059088111022e-05,
+      "loss": 0.7142,
+      "step": 4272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.362811639425157,
+      "learning_rate": 1.8119606624988632e-05,
+      "loss": 0.7401,
+      "step": 4273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.109241423859642,
+      "learning_rate": 1.8118622137950174e-05,
+      "loss": 0.7763,
+      "step": 4274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 27.86274346394573,
+      "learning_rate": 1.8117637420022858e-05,
+      "loss": 0.8347,
+      "step": 4275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.750907907969598,
+      "learning_rate": 1.8116652471234682e-05,
+      "loss": 0.7227,
+      "step": 4276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.535649071841243,
+      "learning_rate": 1.8115667291613653e-05,
+      "loss": 0.7447,
+      "step": 4277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.85389388916283,
+      "learning_rate": 1.8114681881187795e-05,
+      "loss": 0.7143,
+      "step": 4278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.848976007142422,
+      "learning_rate": 1.8113696239985134e-05,
+      "loss": 0.8332,
+      "step": 4279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.079137332828445,
+      "learning_rate": 1.81127103680337e-05,
+      "loss": 0.7348,
+      "step": 4280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.437528922415328,
+      "learning_rate": 1.8111724265361525e-05,
+      "loss": 0.6703,
+      "step": 4281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.77181728523707,
+      "learning_rate": 1.8110737931996658e-05,
+      "loss": 0.7071,
+      "step": 4282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.94980380782242,
+      "learning_rate": 1.8109751367967148e-05,
+      "loss": 0.7469,
+      "step": 4283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.21179685497273,
+      "learning_rate": 1.8108764573301054e-05,
+      "loss": 0.7592,
+      "step": 4284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 30.20020004284759,
+      "learning_rate": 1.810777754802644e-05,
+      "loss": 0.7037,
+      "step": 4285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 22.464562474179527,
+      "learning_rate": 1.8106790292171378e-05,
+      "loss": 0.7459,
+      "step": 4286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.493058091409495,
+      "learning_rate": 1.8105802805763938e-05,
+      "loss": 0.7197,
+      "step": 4287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.784912404683094,
+      "learning_rate": 1.810481508883221e-05,
+      "loss": 0.7689,
+      "step": 4288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.012711676814728,
+      "learning_rate": 1.810382714140428e-05,
+      "loss": 0.7586,
+      "step": 4289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.877625776276066,
+      "learning_rate": 1.8102838963508247e-05,
+      "loss": 0.7816,
+      "step": 4290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.141985244497537,
+      "learning_rate": 1.8101850555172214e-05,
+      "loss": 0.9162,
+      "step": 4291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.660256589872496,
+      "learning_rate": 1.8100861916424293e-05,
+      "loss": 0.8101,
+      "step": 4292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.79914055128638,
+      "learning_rate": 1.8099873047292596e-05,
+      "loss": 0.8228,
+      "step": 4293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 24.090390228412115,
+      "learning_rate": 1.8098883947805247e-05,
+      "loss": 0.8054,
+      "step": 4294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 29.20721872618258,
+      "learning_rate": 1.8097894617990378e-05,
+      "loss": 0.7194,
+      "step": 4295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.476324899080826,
+      "learning_rate": 1.809690505787612e-05,
+      "loss": 0.8397,
+      "step": 4296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.93750872429027,
+      "learning_rate": 1.8095915267490625e-05,
+      "loss": 0.7262,
+      "step": 4297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.124958189143264,
+      "learning_rate": 1.809492524686203e-05,
+      "loss": 0.7357,
+      "step": 4298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.13115606275349,
+      "learning_rate": 1.80939349960185e-05,
+      "loss": 0.8305,
+      "step": 4299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 23.035539915949922,
+      "learning_rate": 1.8092944514988193e-05,
+      "loss": 0.9096,
+      "step": 4300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.872842539960725,
+      "learning_rate": 1.8091953803799275e-05,
+      "loss": 0.6916,
+      "step": 4301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.77223099773228,
+      "learning_rate": 1.8090962862479927e-05,
+      "loss": 0.8311,
+      "step": 4302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.082046418904415,
+      "learning_rate": 1.8089971691058327e-05,
+      "loss": 0.8602,
+      "step": 4303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 29.242166431510118,
+      "learning_rate": 1.8088980289562666e-05,
+      "loss": 0.8083,
+      "step": 4304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.3028884639645,
+      "learning_rate": 1.8087988658021137e-05,
+      "loss": 0.7272,
+      "step": 4305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.733338198346054,
+      "learning_rate": 1.8086996796461943e-05,
+      "loss": 0.7799,
+      "step": 4306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.619174167735295,
+      "learning_rate": 1.808600470491329e-05,
+      "loss": 0.8411,
+      "step": 4307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.88761445214009,
+      "learning_rate": 1.8085012383403392e-05,
+      "loss": 0.7286,
+      "step": 4308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.059224807944098,
+      "learning_rate": 1.8084019831960474e-05,
+      "loss": 0.7332,
+      "step": 4309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.722668229425524,
+      "learning_rate": 1.808302705061276e-05,
+      "loss": 0.6538,
+      "step": 4310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.36886703664029,
+      "learning_rate": 1.8082034039388486e-05,
+      "loss": 0.7814,
+      "step": 4311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.47505662217455,
+      "learning_rate": 1.808104079831589e-05,
+      "loss": 0.8421,
+      "step": 4312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.327617185398385,
+      "learning_rate": 1.808004732742322e-05,
+      "loss": 0.7285,
+      "step": 4313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.1307651272322,
+      "learning_rate": 1.8079053626738732e-05,
+      "loss": 0.7086,
+      "step": 4314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 43.19829755649961,
+      "learning_rate": 1.8078059696290685e-05,
+      "loss": 0.8283,
+      "step": 4315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.407922130239847,
+      "learning_rate": 1.8077065536107347e-05,
+      "loss": 0.8116,
+      "step": 4316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.62592827042257,
+      "learning_rate": 1.8076071146216988e-05,
+      "loss": 0.8064,
+      "step": 4317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.946234029702975,
+      "learning_rate": 1.8075076526647892e-05,
+      "loss": 0.7978,
+      "step": 4318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.55398774569627,
+      "learning_rate": 1.807408167742834e-05,
+      "loss": 0.8183,
+      "step": 4319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.310680777675756,
+      "learning_rate": 1.8073086598586634e-05,
+      "loss": 0.8057,
+      "step": 4320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 22.56848686112926,
+      "learning_rate": 1.8072091290151063e-05,
+      "loss": 0.8163,
+      "step": 4321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 28.680198467444463,
+      "learning_rate": 1.807109575214994e-05,
+      "loss": 0.7474,
+      "step": 4322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.888657405865755,
+      "learning_rate": 1.8070099984611575e-05,
+      "loss": 0.6191,
+      "step": 4323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.62726442014608,
+      "learning_rate": 1.8069103987564286e-05,
+      "loss": 0.741,
+      "step": 4324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.28482620434681,
+      "learning_rate": 1.8068107761036402e-05,
+      "loss": 0.8525,
+      "step": 4325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.900999161629514,
+      "learning_rate": 1.806711130505625e-05,
+      "loss": 0.8357,
+      "step": 4326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.085255612111194,
+      "learning_rate": 1.806611461965217e-05,
+      "loss": 0.7387,
+      "step": 4327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.54601720172465,
+      "learning_rate": 1.8065117704852518e-05,
+      "loss": 0.8159,
+      "step": 4328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.934420265192585,
+      "learning_rate": 1.8064120560685626e-05,
+      "loss": 0.813,
+      "step": 4329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.78721142610371,
+      "learning_rate": 1.806312318717987e-05,
+      "loss": 0.8022,
+      "step": 4330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.46167464757203,
+      "learning_rate": 1.8062125584363605e-05,
+      "loss": 0.8014,
+      "step": 4331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.9684230010178,
+      "learning_rate": 1.8061127752265203e-05,
+      "loss": 0.7071,
+      "step": 4332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.1168461778214,
+      "learning_rate": 1.8060129690913045e-05,
+      "loss": 0.7522,
+      "step": 4333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.285889726700216,
+      "learning_rate": 1.8059131400335512e-05,
+      "loss": 0.8363,
+      "step": 4334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.347939714102884,
+      "learning_rate": 1.8058132880560998e-05,
+      "loss": 0.7959,
+      "step": 4335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.419360362360546,
+      "learning_rate": 1.80571341316179e-05,
+      "loss": 0.7799,
+      "step": 4336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.75484865077361,
+      "learning_rate": 1.8056135153534616e-05,
+      "loss": 0.7569,
+      "step": 4337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.53483295500399,
+      "learning_rate": 1.8055135946339563e-05,
+      "loss": 0.8076,
+      "step": 4338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.414970112993563,
+      "learning_rate": 1.8054136510061156e-05,
+      "loss": 0.7859,
+      "step": 4339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.435482671334562,
+      "learning_rate": 1.8053136844727822e-05,
+      "loss": 0.7089,
+      "step": 4340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.503406182091712,
+      "learning_rate": 1.8052136950367983e-05,
+      "loss": 0.7825,
+      "step": 4341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.00745055602217,
+      "learning_rate": 1.8051136827010077e-05,
+      "loss": 0.8883,
+      "step": 4342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.391674330898983,
+      "learning_rate": 1.8050136474682553e-05,
+      "loss": 0.7553,
+      "step": 4343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.554786039391917,
+      "learning_rate": 1.8049135893413858e-05,
+      "loss": 0.812,
+      "step": 4344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.314777588105425,
+      "learning_rate": 1.8048135083232446e-05,
+      "loss": 0.6781,
+      "step": 4345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.949065210061367,
+      "learning_rate": 1.804713404416678e-05,
+      "loss": 0.8333,
+      "step": 4346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.591601797200667,
+      "learning_rate": 1.8046132776245327e-05,
+      "loss": 0.6883,
+      "step": 4347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.932887640126669,
+      "learning_rate": 1.804513127949657e-05,
+      "loss": 0.8157,
+      "step": 4348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.288021167190276,
+      "learning_rate": 1.8044129553948978e-05,
+      "loss": 0.8966,
+      "step": 4349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.91643776899268,
+      "learning_rate": 1.8043127599631053e-05,
+      "loss": 0.7158,
+      "step": 4350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.067793257401863,
+      "learning_rate": 1.8042125416571284e-05,
+      "loss": 0.7801,
+      "step": 4351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.39998942130946,
+      "learning_rate": 1.8041123004798173e-05,
+      "loss": 0.7697,
+      "step": 4352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.100686927107667,
+      "learning_rate": 1.8040120364340223e-05,
+      "loss": 0.8217,
+      "step": 4353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 28.562866033299674,
+      "learning_rate": 1.8039117495225956e-05,
+      "loss": 0.9237,
+      "step": 4354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.732006388989692,
+      "learning_rate": 1.803811439748389e-05,
+      "loss": 0.7574,
+      "step": 4355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.841406363579924,
+      "learning_rate": 1.8037111071142552e-05,
+      "loss": 0.8425,
+      "step": 4356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.001851837765855,
+      "learning_rate": 1.8036107516230478e-05,
+      "loss": 0.7416,
+      "step": 4357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 27.63443504792,
+      "learning_rate": 1.8035103732776206e-05,
+      "loss": 0.8568,
+      "step": 4358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 40.96006777616897,
+      "learning_rate": 1.803409972080829e-05,
+      "loss": 0.7273,
+      "step": 4359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.101674452924147,
+      "learning_rate": 1.803309548035527e-05,
+      "loss": 0.7821,
+      "step": 4360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.56604787182817,
+      "learning_rate": 1.8032091011445714e-05,
+      "loss": 0.7877,
+      "step": 4361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 26.465669799806857,
+      "learning_rate": 1.803108631410819e-05,
+      "loss": 0.7313,
+      "step": 4362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1204724421504864,
+      "learning_rate": 1.803008138837127e-05,
+      "loss": 0.676,
+      "step": 4363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 40.72773430428415,
+      "learning_rate": 1.8029076234263534e-05,
+      "loss": 0.7996,
+      "step": 4364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.66358343026831,
+      "learning_rate": 1.8028070851813566e-05,
+      "loss": 0.7804,
+      "step": 4365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.6108139596374,
+      "learning_rate": 1.802706524104996e-05,
+      "loss": 0.8305,
+      "step": 4366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.040803518821154,
+      "learning_rate": 1.8026059402001313e-05,
+      "loss": 0.6919,
+      "step": 4367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.762513375167515,
+      "learning_rate": 1.8025053334696232e-05,
+      "loss": 0.7765,
+      "step": 4368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.129159322551544,
+      "learning_rate": 1.8024047039163325e-05,
+      "loss": 0.7246,
+      "step": 4369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.06298176332034,
+      "learning_rate": 1.802304051543122e-05,
+      "loss": 0.7454,
+      "step": 4370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.823940025223486,
+      "learning_rate": 1.8022033763528534e-05,
+      "loss": 0.8304,
+      "step": 4371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.53250667161487,
+      "learning_rate": 1.80210267834839e-05,
+      "loss": 0.707,
+      "step": 4372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.50934033879026,
+      "learning_rate": 1.802001957532596e-05,
+      "loss": 0.7716,
+      "step": 4373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.948848468000165,
+      "learning_rate": 1.801901213908335e-05,
+      "loss": 0.7268,
+      "step": 4374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.420163334666416,
+      "learning_rate": 1.801800447478473e-05,
+      "loss": 0.896,
+      "step": 4375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 26.897851962525642,
+      "learning_rate": 1.8016996582458746e-05,
+      "loss": 0.7131,
+      "step": 4376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.634898020198941,
+      "learning_rate": 1.8015988462134073e-05,
+      "loss": 0.7771,
+      "step": 4377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.821859129615284,
+      "learning_rate": 1.801498011383938e-05,
+      "loss": 0.7458,
+      "step": 4378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.127979127194294,
+      "learning_rate": 1.8013971537603336e-05,
+      "loss": 0.7777,
+      "step": 4379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.924722870477922,
+      "learning_rate": 1.8012962733454636e-05,
+      "loss": 0.7691,
+      "step": 4380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.171522156088209,
+      "learning_rate": 1.8011953701421957e-05,
+      "loss": 0.7542,
+      "step": 4381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.07723572297356,
+      "learning_rate": 1.8010944441534004e-05,
+      "loss": 0.7337,
+      "step": 4382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.69890807634116,
+      "learning_rate": 1.800993495381948e-05,
+      "loss": 0.7926,
+      "step": 4383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.605652896121875,
+      "learning_rate": 1.800892523830709e-05,
+      "loss": 0.8293,
+      "step": 4384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 26.276327549891764,
+      "learning_rate": 1.8007915295025552e-05,
+      "loss": 0.7653,
+      "step": 4385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.608032089735506,
+      "learning_rate": 1.8006905124003583e-05,
+      "loss": 0.8014,
+      "step": 4386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.580945152395373,
+      "learning_rate": 1.8005894725269918e-05,
+      "loss": 0.8588,
+      "step": 4387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.8424159468008,
+      "learning_rate": 1.8004884098853296e-05,
+      "loss": 0.7619,
+      "step": 4388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.82480906982031,
+      "learning_rate": 1.8003873244782447e-05,
+      "loss": 0.7729,
+      "step": 4389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.69975581915073,
+      "learning_rate": 1.8002862163086127e-05,
+      "loss": 0.7613,
+      "step": 4390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.51018684676257,
+      "learning_rate": 1.8001850853793084e-05,
+      "loss": 0.7006,
+      "step": 4391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.82898322524861,
+      "learning_rate": 1.8000839316932094e-05,
+      "loss": 0.7644,
+      "step": 4392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 20.407322404147806,
+      "learning_rate": 1.7999827552531905e-05,
+      "loss": 0.7579,
+      "step": 4393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.723444178959944,
+      "learning_rate": 1.7998815560621305e-05,
+      "loss": 0.8213,
+      "step": 4394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.889634416972665,
+      "learning_rate": 1.7997803341229067e-05,
+      "loss": 0.8172,
+      "step": 4395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.723829215174922,
+      "learning_rate": 1.7996790894383982e-05,
+      "loss": 0.6714,
+      "step": 4396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.59433678373567,
+      "learning_rate": 1.799577822011484e-05,
+      "loss": 0.7723,
+      "step": 4397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.131878716031913,
+      "learning_rate": 1.7994765318450446e-05,
+      "loss": 0.7368,
+      "step": 4398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.187226373704544,
+      "learning_rate": 1.7993752189419602e-05,
+      "loss": 0.7389,
+      "step": 4399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.6787192713573,
+      "learning_rate": 1.7992738833051122e-05,
+      "loss": 0.7607,
+      "step": 4400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 25.246798452805926,
+      "learning_rate": 1.7991725249373822e-05,
+      "loss": 0.8209,
+      "step": 4401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.80020465075268,
+      "learning_rate": 1.7990711438416536e-05,
+      "loss": 0.7652,
+      "step": 4402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.88989684970984,
+      "learning_rate": 1.798969740020809e-05,
+      "loss": 0.7749,
+      "step": 4403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 26.274208877065714,
+      "learning_rate": 1.798868313477732e-05,
+      "loss": 0.6947,
+      "step": 4404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.882254386309226,
+      "learning_rate": 1.7987668642153075e-05,
+      "loss": 0.7086,
+      "step": 4405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.090227360052566,
+      "learning_rate": 1.798665392236421e-05,
+      "loss": 0.6781,
+      "step": 4406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.378504637878486,
+      "learning_rate": 1.7985638975439578e-05,
+      "loss": 0.7496,
+      "step": 4407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 26.369406659588574,
+      "learning_rate": 1.7984623801408046e-05,
+      "loss": 0.7258,
+      "step": 4408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.660206540360726,
+      "learning_rate": 1.798360840029848e-05,
+      "loss": 0.7535,
+      "step": 4409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.986224646399426,
+      "learning_rate": 1.7982592772139766e-05,
+      "loss": 0.7267,
+      "step": 4410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.791240883508046,
+      "learning_rate": 1.7981576916960784e-05,
+      "loss": 0.8889,
+      "step": 4411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.433677329768178,
+      "learning_rate": 1.798056083479042e-05,
+      "loss": 0.732,
+      "step": 4412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.085880260665142,
+      "learning_rate": 1.7979544525657578e-05,
+      "loss": 0.7393,
+      "step": 4413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.73487400201541,
+      "learning_rate": 1.797852798959115e-05,
+      "loss": 0.7295,
+      "step": 4414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 20.464276245659956,
+      "learning_rate": 1.7977511226620062e-05,
+      "loss": 0.8915,
+      "step": 4415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.289352732438633,
+      "learning_rate": 1.7976494236773217e-05,
+      "loss": 0.8119,
+      "step": 4416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.103478341841534,
+      "learning_rate": 1.7975477020079544e-05,
+      "loss": 0.7636,
+      "step": 4417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.893535765235537,
+      "learning_rate": 1.7974459576567968e-05,
+      "loss": 0.7475,
+      "step": 4418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.040377117318236,
+      "learning_rate": 1.797344190626743e-05,
+      "loss": 0.8772,
+      "step": 4419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.792159452154106,
+      "learning_rate": 1.7972424009206865e-05,
+      "loss": 0.7556,
+      "step": 4420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.696463105651482,
+      "learning_rate": 1.7971405885415228e-05,
+      "loss": 0.7266,
+      "step": 4421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.240044561619914,
+      "learning_rate": 1.7970387534921467e-05,
+      "loss": 0.7683,
+      "step": 4422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 23.60420954626879,
+      "learning_rate": 1.796936895775455e-05,
+      "loss": 0.7696,
+      "step": 4423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.518673024693154,
+      "learning_rate": 1.7968350153943443e-05,
+      "loss": 0.8025,
+      "step": 4424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.974492961960626,
+      "learning_rate": 1.7967331123517114e-05,
+      "loss": 0.6361,
+      "step": 4425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.23256368530721,
+      "learning_rate": 1.7966311866504552e-05,
+      "loss": 0.7467,
+      "step": 4426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.19721673234662,
+      "learning_rate": 1.796529238293474e-05,
+      "loss": 0.8107,
+      "step": 4427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.03278740716948,
+      "learning_rate": 1.7964272672836675e-05,
+      "loss": 0.8674,
+      "step": 4428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.983059411990997,
+      "learning_rate": 1.7963252736239347e-05,
+      "loss": 0.7714,
+      "step": 4429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.646595055233774,
+      "learning_rate": 1.7962232573171772e-05,
+      "loss": 0.8475,
+      "step": 4430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.200130679363628,
+      "learning_rate": 1.7961212183662967e-05,
+      "loss": 0.7553,
+      "step": 4431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.025344381774524,
+      "learning_rate": 1.796019156774194e-05,
+      "loss": 0.8178,
+      "step": 4432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 247.01706348116528,
+      "learning_rate": 1.795917072543772e-05,
+      "loss": 0.7495,
+      "step": 4433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.127014079637586,
+      "learning_rate": 1.7958149656779343e-05,
+      "loss": 0.7655,
+      "step": 4434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.928762044211297,
+      "learning_rate": 1.7957128361795844e-05,
+      "loss": 0.7268,
+      "step": 4435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.02890202621121,
+      "learning_rate": 1.7956106840516268e-05,
+      "loss": 0.8222,
+      "step": 4436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.901316062245872,
+      "learning_rate": 1.7955085092969668e-05,
+      "loss": 0.7505,
+      "step": 4437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.681103566629638,
+      "learning_rate": 1.7954063119185105e-05,
+      "loss": 0.7627,
+      "step": 4438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 20.414639107509593,
+      "learning_rate": 1.7953040919191635e-05,
+      "loss": 0.8111,
+      "step": 4439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.012859023733897,
+      "learning_rate": 1.7952018493018335e-05,
+      "loss": 0.7608,
+      "step": 4440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.005937312750195,
+      "learning_rate": 1.7950995840694282e-05,
+      "loss": 0.8028,
+      "step": 4441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.745034940992863,
+      "learning_rate": 1.794997296224856e-05,
+      "loss": 0.6908,
+      "step": 4442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.67019884439328,
+      "learning_rate": 1.7948949857710256e-05,
+      "loss": 0.8683,
+      "step": 4443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 28.939592474780678,
+      "learning_rate": 1.7947926527108467e-05,
+      "loss": 0.8034,
+      "step": 4444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 25.855332846946425,
+      "learning_rate": 1.7946902970472297e-05,
+      "loss": 0.7717,
+      "step": 4445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.151155115022215,
+      "learning_rate": 1.7945879187830857e-05,
+      "loss": 0.7681,
+      "step": 4446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.3831178068288,
+      "learning_rate": 1.794485517921326e-05,
+      "loss": 0.7679,
+      "step": 4447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.126139486420012,
+      "learning_rate": 1.794383094464863e-05,
+      "loss": 0.7383,
+      "step": 4448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.3920308420861,
+      "learning_rate": 1.7942806484166096e-05,
+      "loss": 0.7641,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.317655831876813,
+      "learning_rate": 1.794178179779479e-05,
+      "loss": 0.8207,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.12111928353769,
+      "learning_rate": 1.7940756885563855e-05,
+      "loss": 0.6349,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.89606140772132,
+      "learning_rate": 1.793973174750244e-05,
+      "loss": 0.8081,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.8642243536,
+      "learning_rate": 1.79387063836397e-05,
+      "loss": 0.6497,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.988074198969503,
+      "learning_rate": 1.7937680794004793e-05,
+      "loss": 0.8457,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.446891520665506,
+      "learning_rate": 1.793665497862689e-05,
+      "loss": 0.7774,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.243367416713927,
+      "learning_rate": 1.793562893753516e-05,
+      "loss": 0.8724,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.285394610783442,
+      "learning_rate": 1.7934602670758786e-05,
+      "loss": 0.7699,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.909584013113722,
+      "learning_rate": 1.7933576178326952e-05,
+      "loss": 0.7427,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 24.405535239329332,
+      "learning_rate": 1.7932549460268856e-05,
+      "loss": 0.8373,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.428939132891044,
+      "learning_rate": 1.7931522516613688e-05,
+      "loss": 0.6933,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.68097958812621,
+      "learning_rate": 1.7930495347390665e-05,
+      "loss": 0.8012,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.95167454486104,
+      "learning_rate": 1.792946795262899e-05,
+      "loss": 0.756,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 25.440243898720556,
+      "learning_rate": 1.7928440332357885e-05,
+      "loss": 0.8529,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.31729846039767,
+      "learning_rate": 1.7927412486606578e-05,
+      "loss": 0.7239,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.547319519934067,
+      "learning_rate": 1.7926384415404298e-05,
+      "loss": 0.7182,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.749350629917217,
+      "learning_rate": 1.792535611878028e-05,
+      "loss": 0.8117,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.079585312386996,
+      "learning_rate": 1.792432759676377e-05,
+      "loss": 0.6649,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 26.710050192703306,
+      "learning_rate": 1.792329884938402e-05,
+      "loss": 0.785,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.633466685949253,
+      "learning_rate": 1.7922269876670282e-05,
+      "loss": 0.8311,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.615603397575217,
+      "learning_rate": 1.7921240678651823e-05,
+      "loss": 0.7499,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.947341529955526,
+      "learning_rate": 1.7920211255357918e-05,
+      "loss": 0.7379,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.11531134437023,
+      "learning_rate": 1.7919181606817836e-05,
+      "loss": 0.821,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 27.957349265596267,
+      "learning_rate": 1.791815173306086e-05,
+      "loss": 0.8164,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.651135891057883,
+      "learning_rate": 1.7917121634116284e-05,
+      "loss": 0.8202,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.631555885950647,
+      "learning_rate": 1.7916091310013396e-05,
+      "loss": 0.8001,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.496797906155138,
+      "learning_rate": 1.7915060760781505e-05,
+      "loss": 0.7961,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.339943354218075,
+      "learning_rate": 1.7914029986449916e-05,
+      "loss": 0.7516,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.98117911121771,
+      "learning_rate": 1.791299898704794e-05,
+      "loss": 0.7873,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.449011366411074,
+      "learning_rate": 1.7911967762604905e-05,
+      "loss": 0.7652,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.8835194482637,
+      "learning_rate": 1.7910936313150137e-05,
+      "loss": 0.8287,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.320333878908084,
+      "learning_rate": 1.7909904638712963e-05,
+      "loss": 0.7859,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.95907926110049,
+      "learning_rate": 1.790887273932273e-05,
+      "loss": 0.8135,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.83241715398812,
+      "learning_rate": 1.7907840615008787e-05,
+      "loss": 0.7302,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.654744967972178,
+      "learning_rate": 1.790680826580048e-05,
+      "loss": 0.7286,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.764342193790647,
+      "learning_rate": 1.7905775691727168e-05,
+      "loss": 0.8639,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.18982506913134,
+      "learning_rate": 1.7904742892818225e-05,
+      "loss": 0.8222,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.311592257830757,
+      "learning_rate": 1.7903709869103018e-05,
+      "loss": 0.7469,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.281143091571263,
+      "learning_rate": 1.7902676620610925e-05,
+      "loss": 0.7178,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.40254109882613,
+      "learning_rate": 1.7901643147371328e-05,
+      "loss": 0.8199,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.87214955697475,
+      "learning_rate": 1.7900609449413626e-05,
+      "loss": 0.8005,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.232942132534024,
+      "learning_rate": 1.7899575526767214e-05,
+      "loss": 0.7777,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.33288877982655,
+      "learning_rate": 1.7898541379461494e-05,
+      "loss": 0.7689,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.66610174799004,
+      "learning_rate": 1.7897507007525873e-05,
+      "loss": 0.7062,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.251050478095443,
+      "learning_rate": 1.789647241098978e-05,
+      "loss": 0.8165,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.139680554668924,
+      "learning_rate": 1.7895437589882627e-05,
+      "loss": 0.814,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.387941302325746,
+      "learning_rate": 1.7894402544233846e-05,
+      "loss": 0.8096,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.50296622152407,
+      "learning_rate": 1.789336727407288e-05,
+      "loss": 0.7492,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.119100092542315,
+      "learning_rate": 1.7892331779429164e-05,
+      "loss": 0.6993,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.749492042491251,
+      "learning_rate": 1.789129606033215e-05,
+      "loss": 0.7714,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 28.973994530305404,
+      "learning_rate": 1.789026011681129e-05,
+      "loss": 0.8197,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 28.70944916662741,
+      "learning_rate": 1.7889223948896046e-05,
+      "loss": 0.7195,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.64568218703805,
+      "learning_rate": 1.788818755661589e-05,
+      "loss": 0.7309,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.431507958326039,
+      "learning_rate": 1.7887150940000296e-05,
+      "loss": 0.722,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.803628546058574,
+      "learning_rate": 1.788611409907874e-05,
+      "loss": 0.7555,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.174416813826769,
+      "learning_rate": 1.7885077033880714e-05,
+      "loss": 0.6677,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.519255028776655,
+      "learning_rate": 1.788403974443571e-05,
+      "loss": 0.7111,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.945335141197912,
+      "learning_rate": 1.788300223077323e-05,
+      "loss": 0.7666,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 30.26109322944965,
+      "learning_rate": 1.7881964492922774e-05,
+      "loss": 0.7937,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.18380947710191,
+      "learning_rate": 1.7880926530913863e-05,
+      "loss": 0.7532,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.311576581211785,
+      "learning_rate": 1.7879888344776008e-05,
+      "loss": 0.7534,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.425904345629691,
+      "learning_rate": 1.787884993453874e-05,
+      "loss": 0.7623,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.30198697712689,
+      "learning_rate": 1.7877811300231587e-05,
+      "loss": 0.7699,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.377571393475165,
+      "learning_rate": 1.7876772441884093e-05,
+      "loss": 0.7981,
+      "step": 4513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.58466978839017,
+      "learning_rate": 1.7875733359525793e-05,
+      "loss": 0.8134,
+      "step": 4514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.9714839462326,
+      "learning_rate": 1.7874694053186246e-05,
+      "loss": 0.8238,
+      "step": 4515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.728531718417514,
+      "learning_rate": 1.7873654522895006e-05,
+      "loss": 0.8276,
+      "step": 4516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.510589664470007,
+      "learning_rate": 1.7872614768681638e-05,
+      "loss": 0.8108,
+      "step": 4517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.24816512107083,
+      "learning_rate": 1.7871574790575713e-05,
+      "loss": 0.8559,
+      "step": 4518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 24.814463942464236,
+      "learning_rate": 1.7870534588606804e-05,
+      "loss": 0.8228,
+      "step": 4519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.126995740101478,
+      "learning_rate": 1.7869494162804492e-05,
+      "loss": 0.7478,
+      "step": 4520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.225116719811254,
+      "learning_rate": 1.7868453513198375e-05,
+      "loss": 0.7609,
+      "step": 4521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.06792835085101,
+      "learning_rate": 1.786741263981804e-05,
+      "loss": 0.724,
+      "step": 4522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 25.883283735753,
+      "learning_rate": 1.7866371542693094e-05,
+      "loss": 0.7266,
+      "step": 4523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.20164873034085,
+      "learning_rate": 1.786533022185314e-05,
+      "loss": 0.7889,
+      "step": 4524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 30.722225829817987,
+      "learning_rate": 1.78642886773278e-05,
+      "loss": 0.8972,
+      "step": 4525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.33453424711936,
+      "learning_rate": 1.7863246909146688e-05,
+      "loss": 0.7599,
+      "step": 4526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.690056751794366,
+      "learning_rate": 1.786220491733943e-05,
+      "loss": 0.7537,
+      "step": 4527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.930520901810457,
+      "learning_rate": 1.7861162701935672e-05,
+      "loss": 0.7096,
+      "step": 4528
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.71866787237424,
+      "learning_rate": 1.7860120262965038e-05,
+      "loss": 0.7578,
+      "step": 4529
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 27.19743173764247,
+      "learning_rate": 1.7859077600457184e-05,
+      "loss": 0.7474,
+      "step": 4530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.763969372970255,
+      "learning_rate": 1.7858034714441764e-05,
+      "loss": 0.7404,
+      "step": 4531
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.289749176430142,
+      "learning_rate": 1.7856991604948428e-05,
+      "loss": 0.689,
+      "step": 4532
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.31819391225716,
+      "learning_rate": 1.7855948272006848e-05,
+      "loss": 0.7654,
+      "step": 4533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.21917595537159,
+      "learning_rate": 1.7854904715646697e-05,
+      "loss": 0.7669,
+      "step": 4534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.18341888851737,
+      "learning_rate": 1.785386093589765e-05,
+      "loss": 0.7457,
+      "step": 4535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.213289459037025,
+      "learning_rate": 1.7852816932789392e-05,
+      "loss": 0.85,
+      "step": 4536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.041481135578346,
+      "learning_rate": 1.7851772706351614e-05,
+      "loss": 0.7068,
+      "step": 4537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.176504845582247,
+      "learning_rate": 1.785072825661401e-05,
+      "loss": 0.7969,
+      "step": 4538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.287263358936663,
+      "learning_rate": 1.7849683583606293e-05,
+      "loss": 0.7074,
+      "step": 4539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.276838393556602,
+      "learning_rate": 1.7848638687358164e-05,
+      "loss": 0.6571,
+      "step": 4540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.05278429987945,
+      "learning_rate": 1.7847593567899344e-05,
+      "loss": 0.7659,
+      "step": 4541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.298943066125636,
+      "learning_rate": 1.7846548225259553e-05,
+      "loss": 0.7017,
+      "step": 4542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.871929997530419,
+      "learning_rate": 1.7845502659468522e-05,
+      "loss": 0.7217,
+      "step": 4543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.840309628514953,
+      "learning_rate": 1.784445687055598e-05,
+      "loss": 0.7104,
+      "step": 4544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.54040551532769,
+      "learning_rate": 1.7843410858551674e-05,
+      "loss": 0.6895,
+      "step": 4545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.885929705359707,
+      "learning_rate": 1.7842364623485356e-05,
+      "loss": 0.8651,
+      "step": 4546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.742020265499644,
+      "learning_rate": 1.784131816538677e-05,
+      "loss": 0.8096,
+      "step": 4547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.32213060918708,
+      "learning_rate": 1.7840271484285687e-05,
+      "loss": 0.7455,
+      "step": 4548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.294218046096653,
+      "learning_rate": 1.783922458021187e-05,
+      "loss": 0.7091,
+      "step": 4549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.886128865663519,
+      "learning_rate": 1.783817745319509e-05,
+      "loss": 0.693,
+      "step": 4550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.304773134884957,
+      "learning_rate": 1.783713010326513e-05,
+      "loss": 0.7035,
+      "step": 4551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.510328769850396,
+      "learning_rate": 1.7836082530451772e-05,
+      "loss": 0.8074,
+      "step": 4552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.01065139370198,
+      "learning_rate": 1.783503473478481e-05,
+      "loss": 0.6972,
+      "step": 4553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.235472748525677,
+      "learning_rate": 1.7833986716294046e-05,
+      "loss": 0.7201,
+      "step": 4554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.144588477717663,
+      "learning_rate": 1.783293847500928e-05,
+      "loss": 0.7781,
+      "step": 4555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.747317467243905,
+      "learning_rate": 1.783189001096033e-05,
+      "loss": 0.6988,
+      "step": 4556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.169701209631764,
+      "learning_rate": 1.7830841324177006e-05,
+      "loss": 0.7843,
+      "step": 4557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.938530348425555,
+      "learning_rate": 1.7829792414689135e-05,
+      "loss": 0.7218,
+      "step": 4558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.487101280375038,
+      "learning_rate": 1.782874328252655e-05,
+      "loss": 0.7515,
+      "step": 4559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 27.69630725295481,
+      "learning_rate": 1.7827693927719086e-05,
+      "loss": 0.8335,
+      "step": 4560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.951567465727559,
+      "learning_rate": 1.7826644350296588e-05,
+      "loss": 0.7452,
+      "step": 4561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.347824798761902,
+      "learning_rate": 1.7825594550288898e-05,
+      "loss": 0.7723,
+      "step": 4562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.25889149156369,
+      "learning_rate": 1.782454452772588e-05,
+      "loss": 0.719,
+      "step": 4563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.542686348007297,
+      "learning_rate": 1.782349428263739e-05,
+      "loss": 0.7425,
+      "step": 4564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.523572289632572,
+      "learning_rate": 1.7822443815053305e-05,
+      "loss": 0.848,
+      "step": 4565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.67062176082445,
+      "learning_rate": 1.782139312500349e-05,
+      "loss": 0.7283,
+      "step": 4566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.117571666016065,
+      "learning_rate": 1.7820342212517826e-05,
+      "loss": 0.714,
+      "step": 4567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.19939445039066,
+      "learning_rate": 1.7819291077626212e-05,
+      "loss": 0.8413,
+      "step": 4568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.50637241374402,
+      "learning_rate": 1.781823972035853e-05,
+      "loss": 0.8397,
+      "step": 4569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.80469806867194,
+      "learning_rate": 1.7817188140744682e-05,
+      "loss": 0.7116,
+      "step": 4570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.981495182970207,
+      "learning_rate": 1.7816136338814577e-05,
+      "loss": 0.7207,
+      "step": 4571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.554049720698364,
+      "learning_rate": 1.781508431459813e-05,
+      "loss": 0.7284,
+      "step": 4572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.198146139689523,
+      "learning_rate": 1.7814032068125254e-05,
+      "loss": 0.8335,
+      "step": 4573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.120963710368788,
+      "learning_rate": 1.781297959942588e-05,
+      "loss": 0.7382,
+      "step": 4574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.32251608311107,
+      "learning_rate": 1.7811926908529934e-05,
+      "loss": 0.7331,
+      "step": 4575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.51043225540429,
+      "learning_rate": 1.781087399546736e-05,
+      "loss": 0.7078,
+      "step": 4576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.47989384276147,
+      "learning_rate": 1.7809820860268096e-05,
+      "loss": 0.7913,
+      "step": 4577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.634376014463157,
+      "learning_rate": 1.7808767502962094e-05,
+      "loss": 0.72,
+      "step": 4578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 22.849754926480593,
+      "learning_rate": 1.7807713923579313e-05,
+      "loss": 0.6631,
+      "step": 4579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.006859729488482,
+      "learning_rate": 1.7806660122149714e-05,
+      "loss": 0.7598,
+      "step": 4580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.412321651336214,
+      "learning_rate": 1.7805606098703273e-05,
+      "loss": 0.699,
+      "step": 4581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.106188192033642,
+      "learning_rate": 1.7804551853269957e-05,
+      "loss": 0.7818,
+      "step": 4582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.032803466648353,
+      "learning_rate": 1.7803497385879752e-05,
+      "loss": 0.8391,
+      "step": 4583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.95640167772439,
+      "learning_rate": 1.7802442696562646e-05,
+      "loss": 0.6838,
+      "step": 4584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.11034547457215,
+      "learning_rate": 1.7801387785348637e-05,
+      "loss": 0.7524,
+      "step": 4585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.478175168142961,
+      "learning_rate": 1.780033265226772e-05,
+      "loss": 0.6551,
+      "step": 4586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.932257244091351,
+      "learning_rate": 1.7799277297349908e-05,
+      "loss": 0.6978,
+      "step": 4587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.86384159850776,
+      "learning_rate": 1.7798221720625205e-05,
+      "loss": 0.6796,
+      "step": 4588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.21155762223803,
+      "learning_rate": 1.779716592212365e-05,
+      "loss": 0.7782,
+      "step": 4589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.086023444183674,
+      "learning_rate": 1.7796109901875246e-05,
+      "loss": 0.7912,
+      "step": 4590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.905189667765715,
+      "learning_rate": 1.7795053659910044e-05,
+      "loss": 0.822,
+      "step": 4591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.280428154633523,
+      "learning_rate": 1.7793997196258074e-05,
+      "loss": 0.8209,
+      "step": 4592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.618783952315601,
+      "learning_rate": 1.7792940510949383e-05,
+      "loss": 0.7516,
+      "step": 4593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 26.049443477568158,
+      "learning_rate": 1.7791883604014023e-05,
+      "loss": 0.7825,
+      "step": 4594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.4910808696523,
+      "learning_rate": 1.7790826475482047e-05,
+      "loss": 0.812,
+      "step": 4595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.17971809472022,
+      "learning_rate": 1.7789769125383528e-05,
+      "loss": 0.7576,
+      "step": 4596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.355798234337353,
+      "learning_rate": 1.7788711553748533e-05,
+      "loss": 0.7372,
+      "step": 4597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.416064380501787,
+      "learning_rate": 1.7787653760607134e-05,
+      "loss": 0.7403,
+      "step": 4598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 27.892054013099575,
+      "learning_rate": 1.778659574598942e-05,
+      "loss": 0.7828,
+      "step": 4599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.44548862474979,
+      "learning_rate": 1.7785537509925478e-05,
+      "loss": 0.721,
+      "step": 4600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.884567290745816,
+      "learning_rate": 1.77844790524454e-05,
+      "loss": 0.7116,
+      "step": 4601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.758239248976416,
+      "learning_rate": 1.77834203735793e-05,
+      "loss": 0.7893,
+      "step": 4602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 28.94109036209768,
+      "learning_rate": 1.778236147335727e-05,
+      "loss": 0.6847,
+      "step": 4603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.537867110602596,
+      "learning_rate": 1.7781302351809437e-05,
+      "loss": 0.691,
+      "step": 4604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.825930645070017,
+      "learning_rate": 1.7780243008965915e-05,
+      "loss": 0.652,
+      "step": 4605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.45764941718774,
+      "learning_rate": 1.7779183444856833e-05,
+      "loss": 0.7485,
+      "step": 4606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 26.59329265824312,
+      "learning_rate": 1.7778123659512326e-05,
+      "loss": 0.7516,
+      "step": 4607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.52337750132507,
+      "learning_rate": 1.777706365296253e-05,
+      "loss": 0.6939,
+      "step": 4608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.87398120162674,
+      "learning_rate": 1.7776003425237592e-05,
+      "loss": 0.6667,
+      "step": 4609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.049088492545417,
+      "learning_rate": 1.7774942976367668e-05,
+      "loss": 0.7756,
+      "step": 4610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.44443025623951,
+      "learning_rate": 1.7773882306382913e-05,
+      "loss": 0.6874,
+      "step": 4611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.064901833986912,
+      "learning_rate": 1.7772821415313493e-05,
+      "loss": 0.7053,
+      "step": 4612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.36597587180189,
+      "learning_rate": 1.777176030318958e-05,
+      "loss": 0.7054,
+      "step": 4613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.365403532037313,
+      "learning_rate": 1.777069897004135e-05,
+      "loss": 0.7461,
+      "step": 4614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.22883278371235,
+      "learning_rate": 1.7769637415898982e-05,
+      "loss": 0.6973,
+      "step": 4615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.706301176988791,
+      "learning_rate": 1.776857564079268e-05,
+      "loss": 0.7002,
+      "step": 4616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 29.914397830832083,
+      "learning_rate": 1.7767513644752624e-05,
+      "loss": 0.8495,
+      "step": 4617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.378578762289482,
+      "learning_rate": 1.7766451427809026e-05,
+      "loss": 0.6789,
+      "step": 4618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.225273640111205,
+      "learning_rate": 1.7765388989992093e-05,
+      "loss": 0.7316,
+      "step": 4619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 23.14733087145608,
+      "learning_rate": 1.776432633133204e-05,
+      "loss": 0.7645,
+      "step": 4620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.36780120500211,
+      "learning_rate": 1.776326345185909e-05,
+      "loss": 0.7196,
+      "step": 4621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.967696973710073,
+      "learning_rate": 1.7762200351603465e-05,
+      "loss": 0.7228,
+      "step": 4622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.57145546333111,
+      "learning_rate": 1.7761137030595406e-05,
+      "loss": 0.7221,
+      "step": 4623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.224495556378056,
+      "learning_rate": 1.776007348886515e-05,
+      "loss": 0.7173,
+      "step": 4624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.778919065555696,
+      "learning_rate": 1.775900972644294e-05,
+      "loss": 0.7658,
+      "step": 4625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.968809065971032,
+      "learning_rate": 1.7757945743359033e-05,
+      "loss": 0.7949,
+      "step": 4626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.726221838120367,
+      "learning_rate": 1.775688153964369e-05,
+      "loss": 0.8367,
+      "step": 4627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.838458730080607,
+      "learning_rate": 1.775581711532717e-05,
+      "loss": 0.7264,
+      "step": 4628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.315085156088408,
+      "learning_rate": 1.7754752470439755e-05,
+      "loss": 0.752,
+      "step": 4629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.852962364319751,
+      "learning_rate": 1.7753687605011707e-05,
+      "loss": 0.763,
+      "step": 4630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.063412276120008,
+      "learning_rate": 1.7752622519073327e-05,
+      "loss": 0.7749,
+      "step": 4631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.92250672784313,
+      "learning_rate": 1.7751557212654893e-05,
+      "loss": 0.7456,
+      "step": 4632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.54124495642337,
+      "learning_rate": 1.7750491685786708e-05,
+      "loss": 0.8077,
+      "step": 4633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.289543972708875,
+      "learning_rate": 1.7749425938499073e-05,
+      "loss": 0.7602,
+      "step": 4634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.640675451080558,
+      "learning_rate": 1.77483599708223e-05,
+      "loss": 0.7679,
+      "step": 4635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.525962541647345,
+      "learning_rate": 1.7747293782786697e-05,
+      "loss": 0.8328,
+      "step": 4636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.983633664461529,
+      "learning_rate": 1.7746227374422597e-05,
+      "loss": 0.7545,
+      "step": 4637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 29.179535389699073,
+      "learning_rate": 1.7745160745760317e-05,
+      "loss": 0.7919,
+      "step": 4638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 21.79949564065988,
+      "learning_rate": 1.7744093896830196e-05,
+      "loss": 0.7646,
+      "step": 4639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 26.00539878640754,
+      "learning_rate": 1.7743026827662578e-05,
+      "loss": 0.7665,
+      "step": 4640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.839287454604097,
+      "learning_rate": 1.7741959538287807e-05,
+      "loss": 0.7677,
+      "step": 4641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.868581926904834,
+      "learning_rate": 1.7740892028736233e-05,
+      "loss": 0.7512,
+      "step": 4642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.953774704601337,
+      "learning_rate": 1.7739824299038217e-05,
+      "loss": 0.8183,
+      "step": 4643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.993655319477101,
+      "learning_rate": 1.773875634922413e-05,
+      "loss": 0.7837,
+      "step": 4644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.672320268022368,
+      "learning_rate": 1.7737688179324334e-05,
+      "loss": 0.789,
+      "step": 4645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.89504836425524,
+      "learning_rate": 1.773661978936922e-05,
+      "loss": 0.7179,
+      "step": 4646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.98533202926206,
+      "learning_rate": 1.7735551179389157e-05,
+      "loss": 0.7097,
+      "step": 4647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.47458291286288,
+      "learning_rate": 1.7734482349414547e-05,
+      "loss": 0.7666,
+      "step": 4648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.43875464033775,
+      "learning_rate": 1.7733413299475787e-05,
+      "loss": 0.8031,
+      "step": 4649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.87916699853392,
+      "learning_rate": 1.7732344029603273e-05,
+      "loss": 0.7293,
+      "step": 4650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.626097062298985,
+      "learning_rate": 1.773127453982742e-05,
+      "loss": 0.9176,
+      "step": 4651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.283764035368232,
+      "learning_rate": 1.7730204830178638e-05,
+      "loss": 0.7812,
+      "step": 4652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.78481579672387,
+      "learning_rate": 1.7729134900687354e-05,
+      "loss": 0.7556,
+      "step": 4653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 25.059231830123245,
+      "learning_rate": 1.7728064751383997e-05,
+      "loss": 0.784,
+      "step": 4654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 22.44174358356745,
+      "learning_rate": 1.7726994382299e-05,
+      "loss": 0.758,
+      "step": 4655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.80834578322753,
+      "learning_rate": 1.7725923793462798e-05,
+      "loss": 0.8446,
+      "step": 4656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.215338581736098,
+      "learning_rate": 1.7724852984905844e-05,
+      "loss": 0.6924,
+      "step": 4657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.856078966059064,
+      "learning_rate": 1.7723781956658593e-05,
+      "loss": 0.7276,
+      "step": 4658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.507102864187885,
+      "learning_rate": 1.77227107087515e-05,
+      "loss": 0.7842,
+      "step": 4659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.76527188781026,
+      "learning_rate": 1.7721639241215034e-05,
+      "loss": 0.7244,
+      "step": 4660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.368469322252448,
+      "learning_rate": 1.772056755407966e-05,
+      "loss": 0.7384,
+      "step": 4661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 22.35696192981688,
+      "learning_rate": 1.771949564737587e-05,
+      "loss": 0.769,
+      "step": 4662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.98192081147761,
+      "learning_rate": 1.771842352113413e-05,
+      "loss": 0.7962,
+      "step": 4663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.801141597518985,
+      "learning_rate": 1.7717351175384948e-05,
+      "loss": 0.7734,
+      "step": 4664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 28.0424575461503,
+      "learning_rate": 1.7716278610158806e-05,
+      "loss": 0.8066,
+      "step": 4665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.600696584361431,
+      "learning_rate": 1.7715205825486218e-05,
+      "loss": 0.7815,
+      "step": 4666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.76127266345328,
+      "learning_rate": 1.771413282139769e-05,
+      "loss": 0.7115,
+      "step": 4667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.362943337772816,
+      "learning_rate": 1.7713059597923735e-05,
+      "loss": 0.7848,
+      "step": 4668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.439576638543528,
+      "learning_rate": 1.771198615509488e-05,
+      "loss": 0.756,
+      "step": 4669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.119958743836833,
+      "learning_rate": 1.771091249294165e-05,
+      "loss": 0.6665,
+      "step": 4670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.20309282801442,
+      "learning_rate": 1.7709838611494577e-05,
+      "loss": 0.7591,
+      "step": 4671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.794715676459475,
+      "learning_rate": 1.7708764510784205e-05,
+      "loss": 0.725,
+      "step": 4672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.402110985595833,
+      "learning_rate": 1.770769019084108e-05,
+      "loss": 0.6992,
+      "step": 4673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 33.52548618675277,
+      "learning_rate": 1.7706615651695754e-05,
+      "loss": 0.8105,
+      "step": 4674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.938989561018383,
+      "learning_rate": 1.7705540893378787e-05,
+      "loss": 0.7831,
+      "step": 4675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.43314064498119,
+      "learning_rate": 1.7704465915920745e-05,
+      "loss": 0.815,
+      "step": 4676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 20.685658191031653,
+      "learning_rate": 1.77033907193522e-05,
+      "loss": 0.7443,
+      "step": 4677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.049858843156683,
+      "learning_rate": 1.770231530370373e-05,
+      "loss": 0.7456,
+      "step": 4678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.708947774163132,
+      "learning_rate": 1.7701239669005916e-05,
+      "loss": 0.8426,
+      "step": 4679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.484571068946565,
+      "learning_rate": 1.7700163815289353e-05,
+      "loss": 0.724,
+      "step": 4680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.628196660609422,
+      "learning_rate": 1.7699087742584636e-05,
+      "loss": 0.7379,
+      "step": 4681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 26.81608904812798,
+      "learning_rate": 1.769801145092237e-05,
+      "loss": 0.7381,
+      "step": 4682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.802430812466564,
+      "learning_rate": 1.7696934940333155e-05,
+      "loss": 0.8898,
+      "step": 4683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 27.12962424616862,
+      "learning_rate": 1.7695858210847615e-05,
+      "loss": 0.8014,
+      "step": 4684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.1015214166352,
+      "learning_rate": 1.769478126249637e-05,
+      "loss": 0.7168,
+      "step": 4685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 20.520628887616834,
+      "learning_rate": 1.769370409531005e-05,
+      "loss": 0.7749,
+      "step": 4686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 20.73359294572111,
+      "learning_rate": 1.7692626709319283e-05,
+      "loss": 0.7404,
+      "step": 4687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 26.08046606022863,
+      "learning_rate": 1.769154910455471e-05,
+      "loss": 0.7131,
+      "step": 4688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.21366844840311,
+      "learning_rate": 1.7690471281046982e-05,
+      "loss": 0.7684,
+      "step": 4689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.744214996514318,
+      "learning_rate": 1.768939323882675e-05,
+      "loss": 0.7194,
+      "step": 4690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.950315766400433,
+      "learning_rate": 1.7688314977924674e-05,
+      "loss": 0.7763,
+      "step": 4691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.612398469106186,
+      "learning_rate": 1.768723649837142e-05,
+      "loss": 0.744,
+      "step": 4692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.773618786659869,
+      "learning_rate": 1.7686157800197652e-05,
+      "loss": 0.7515,
+      "step": 4693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.05816044164395,
+      "learning_rate": 1.7685078883434054e-05,
+      "loss": 0.744,
+      "step": 4694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.406283108288324,
+      "learning_rate": 1.7683999748111306e-05,
+      "loss": 0.831,
+      "step": 4695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.617075782901715,
+      "learning_rate": 1.7682920394260102e-05,
+      "loss": 0.7737,
+      "step": 4696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.916444636448617,
+      "learning_rate": 1.768184082191114e-05,
+      "loss": 0.7374,
+      "step": 4697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.094554564256278,
+      "learning_rate": 1.7680761031095113e-05,
+      "loss": 0.6707,
+      "step": 4698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.64583492298938,
+      "learning_rate": 1.7679681021842738e-05,
+      "loss": 0.7114,
+      "step": 4699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.201698725408875,
+      "learning_rate": 1.767860079418473e-05,
+      "loss": 0.7649,
+      "step": 4700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 27.219623840705353,
+      "learning_rate": 1.7677520348151805e-05,
+      "loss": 0.798,
+      "step": 4701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.068731173647688,
+      "learning_rate": 1.767643968377469e-05,
+      "loss": 0.7095,
+      "step": 4702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.347959692297604,
+      "learning_rate": 1.7675358801084122e-05,
+      "loss": 0.6746,
+      "step": 4703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 20.983950511220705,
+      "learning_rate": 1.767427770011084e-05,
+      "loss": 0.7586,
+      "step": 4704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 22.607776450884366,
+      "learning_rate": 1.767319638088559e-05,
+      "loss": 0.8404,
+      "step": 4705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.93052476896135,
+      "learning_rate": 1.7672114843439127e-05,
+      "loss": 0.7683,
+      "step": 4706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 25.206016053666044,
+      "learning_rate": 1.76710330878022e-05,
+      "loss": 0.8398,
+      "step": 4707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.813816645625376,
+      "learning_rate": 1.7669951114005584e-05,
+      "loss": 0.7192,
+      "step": 4708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.644285375681033,
+      "learning_rate": 1.766886892208004e-05,
+      "loss": 0.7189,
+      "step": 4709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.0204235644049,
+      "learning_rate": 1.7667786512056352e-05,
+      "loss": 0.8104,
+      "step": 4710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.506676078951678,
+      "learning_rate": 1.7666703883965305e-05,
+      "loss": 0.7505,
+      "step": 4711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.526104483491093,
+      "learning_rate": 1.766562103783768e-05,
+      "loss": 0.8487,
+      "step": 4712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.166371296326044,
+      "learning_rate": 1.766453797370428e-05,
+      "loss": 0.8902,
+      "step": 4713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.098966153656235,
+      "learning_rate": 1.76634546915959e-05,
+      "loss": 0.8063,
+      "step": 4714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.266027946684616,
+      "learning_rate": 1.7662371191543355e-05,
+      "loss": 0.7278,
+      "step": 4715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 28.156781562832112,
+      "learning_rate": 1.7661287473577457e-05,
+      "loss": 0.7446,
+      "step": 4716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.59327000389072,
+      "learning_rate": 1.766020353772902e-05,
+      "loss": 0.8598,
+      "step": 4717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.87572390552177,
+      "learning_rate": 1.765911938402888e-05,
+      "loss": 0.7998,
+      "step": 4718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 23.873782269145302,
+      "learning_rate": 1.7658035012507863e-05,
+      "loss": 0.757,
+      "step": 4719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.234898673833182,
+      "learning_rate": 1.7656950423196807e-05,
+      "loss": 0.7828,
+      "step": 4720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.95689105247914,
+      "learning_rate": 1.7655865616126564e-05,
+      "loss": 0.7172,
+      "step": 4721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.574414710725613,
+      "learning_rate": 1.7654780591327983e-05,
+      "loss": 0.7864,
+      "step": 4722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.966824203599673,
+      "learning_rate": 1.7653695348831915e-05,
+      "loss": 0.7277,
+      "step": 4723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.66688520859272,
+      "learning_rate": 1.7652609888669234e-05,
+      "loss": 0.6598,
+      "step": 4724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.200763538571433,
+      "learning_rate": 1.76515242108708e-05,
+      "loss": 0.8736,
+      "step": 4725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.356492148095814,
+      "learning_rate": 1.7650438315467494e-05,
+      "loss": 0.7293,
+      "step": 4726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.214323658311407,
+      "learning_rate": 1.7649352202490198e-05,
+      "loss": 0.7564,
+      "step": 4727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.56004466670747,
+      "learning_rate": 1.7648265871969803e-05,
+      "loss": 0.768,
+      "step": 4728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.702671486249205,
+      "learning_rate": 1.76471793239372e-05,
+      "loss": 0.7815,
+      "step": 4729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.945495801921826,
+      "learning_rate": 1.7646092558423288e-05,
+      "loss": 0.8258,
+      "step": 4730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.69115327136434,
+      "learning_rate": 1.7645005575458977e-05,
+      "loss": 0.7168,
+      "step": 4731
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0180134369763225,
+      "learning_rate": 1.764391837507518e-05,
+      "loss": 0.6225,
+      "step": 4732
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.709364330858422,
+      "learning_rate": 1.7642830957302815e-05,
+      "loss": 0.7384,
+      "step": 4733
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.751418677749186,
+      "learning_rate": 1.7641743322172812e-05,
+      "loss": 0.7155,
+      "step": 4734
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 29.490568879027517,
+      "learning_rate": 1.7640655469716096e-05,
+      "loss": 0.8194,
+      "step": 4735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.936220906422324,
+      "learning_rate": 1.7639567399963607e-05,
+      "loss": 0.7671,
+      "step": 4736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.913759602721246,
+      "learning_rate": 1.7638479112946294e-05,
+      "loss": 0.7645,
+      "step": 4737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.82486069972475,
+      "learning_rate": 1.7637390608695102e-05,
+      "loss": 0.8512,
+      "step": 4738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.809928657989634,
+      "learning_rate": 1.7636301887240987e-05,
+      "loss": 0.7897,
+      "step": 4739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.471959093668424,
+      "learning_rate": 1.7635212948614915e-05,
+      "loss": 0.6799,
+      "step": 4740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.685886680214303,
+      "learning_rate": 1.7634123792847854e-05,
+      "loss": 0.7134,
+      "step": 4741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.1684793871101,
+      "learning_rate": 1.7633034419970775e-05,
+      "loss": 0.796,
+      "step": 4742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.809886761536184,
+      "learning_rate": 1.7631944830014663e-05,
+      "loss": 0.7088,
+      "step": 4743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.96983445565587,
+      "learning_rate": 1.763085502301051e-05,
+      "loss": 0.6873,
+      "step": 4744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 23.76149004516207,
+      "learning_rate": 1.76297649989893e-05,
+      "loss": 0.7967,
+      "step": 4745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.463363868330706,
+      "learning_rate": 1.7628674757982037e-05,
+      "loss": 0.8489,
+      "step": 4746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.672467960747674,
+      "learning_rate": 1.7627584300019727e-05,
+      "loss": 0.6577,
+      "step": 4747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 25.878883422841493,
+      "learning_rate": 1.7626493625133377e-05,
+      "loss": 0.7587,
+      "step": 4748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.673150511219983,
+      "learning_rate": 1.7625402733354015e-05,
+      "loss": 0.8432,
+      "step": 4749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.784277057915933,
+      "learning_rate": 1.7624311624712657e-05,
+      "loss": 0.6969,
+      "step": 4750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.754912266190452,
+      "learning_rate": 1.762322029924034e-05,
+      "loss": 0.8593,
+      "step": 4751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.59320934205328,
+      "learning_rate": 1.7622128756968095e-05,
+      "loss": 0.9065,
+      "step": 4752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.267190841368144,
+      "learning_rate": 1.762103699792697e-05,
+      "loss": 0.6885,
+      "step": 4753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.573975020037427,
+      "learning_rate": 1.7619945022148008e-05,
+      "loss": 0.7677,
+      "step": 4754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.137598308973239,
+      "learning_rate": 1.7618852829662264e-05,
+      "loss": 0.7774,
+      "step": 4755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.130612327390836,
+      "learning_rate": 1.7617760420500806e-05,
+      "loss": 0.7532,
+      "step": 4756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.903173049405172,
+      "learning_rate": 1.7616667794694697e-05,
+      "loss": 0.7776,
+      "step": 4757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.739616063788986,
+      "learning_rate": 1.7615574952275012e-05,
+      "loss": 0.8543,
+      "step": 4758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.746722267982145,
+      "learning_rate": 1.761448189327283e-05,
+      "loss": 0.7372,
+      "step": 4759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.923825014880418,
+      "learning_rate": 1.761338861771924e-05,
+      "loss": 0.7715,
+      "step": 4760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.34899404771331,
+      "learning_rate": 1.761229512564533e-05,
+      "loss": 0.8064,
+      "step": 4761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.528797054775271,
+      "learning_rate": 1.7611201417082196e-05,
+      "loss": 0.7003,
+      "step": 4762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.192845139928203,
+      "learning_rate": 1.761010749206095e-05,
+      "loss": 0.8327,
+      "step": 4763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.279638844594935,
+      "learning_rate": 1.7609013350612696e-05,
+      "loss": 0.6791,
+      "step": 4764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.375197317112875,
+      "learning_rate": 1.7607918992768554e-05,
+      "loss": 0.8303,
+      "step": 4765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 30.204422041251554,
+      "learning_rate": 1.7606824418559648e-05,
+      "loss": 0.8284,
+      "step": 4766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.738304781049953,
+      "learning_rate": 1.7605729628017108e-05,
+      "loss": 0.7289,
+      "step": 4767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 23.424944146577246,
+      "learning_rate": 1.760463462117206e-05,
+      "loss": 0.7679,
+      "step": 4768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.6465918798675,
+      "learning_rate": 1.7603539398055658e-05,
+      "loss": 0.7444,
+      "step": 4769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 25.41213395370621,
+      "learning_rate": 1.760244395869904e-05,
+      "loss": 0.7664,
+      "step": 4770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.23403088172201,
+      "learning_rate": 1.7601348303133364e-05,
+      "loss": 0.8096,
+      "step": 4771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.55227978095691,
+      "learning_rate": 1.760025243138979e-05,
+      "loss": 0.767,
+      "step": 4772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 25.0073222339007,
+      "learning_rate": 1.7599156343499482e-05,
+      "loss": 0.8258,
+      "step": 4773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.34440598750834,
+      "learning_rate": 1.7598060039493613e-05,
+      "loss": 0.6841,
+      "step": 4774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.792987881152875,
+      "learning_rate": 1.759696351940336e-05,
+      "loss": 0.8004,
+      "step": 4775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.556558176506321,
+      "learning_rate": 1.759586678325991e-05,
+      "loss": 0.7607,
+      "step": 4776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.958002349925415,
+      "learning_rate": 1.7594769831094452e-05,
+      "loss": 0.7425,
+      "step": 4777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.902998572006434,
+      "learning_rate": 1.759367266293818e-05,
+      "loss": 0.7837,
+      "step": 4778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.127361967484298,
+      "learning_rate": 1.7592575278822304e-05,
+      "loss": 0.7381,
+      "step": 4779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.620887369463702,
+      "learning_rate": 1.7591477678778027e-05,
+      "loss": 0.8289,
+      "step": 4780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.46171538137161,
+      "learning_rate": 1.7590379862836565e-05,
+      "loss": 0.7354,
+      "step": 4781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.731126361314573,
+      "learning_rate": 1.758928183102914e-05,
+      "loss": 0.7713,
+      "step": 4782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.345787646147702,
+      "learning_rate": 1.7588183583386982e-05,
+      "loss": 0.7797,
+      "step": 4783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.611504364860586,
+      "learning_rate": 1.7587085119941318e-05,
+      "loss": 0.7315,
+      "step": 4784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 24.601739066305928,
+      "learning_rate": 1.758598644072339e-05,
+      "loss": 0.678,
+      "step": 4785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.040475408622893,
+      "learning_rate": 1.7584887545764452e-05,
+      "loss": 0.6183,
+      "step": 4786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.977348198296745,
+      "learning_rate": 1.7583788435095746e-05,
+      "loss": 0.7622,
+      "step": 4787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.658938106779566,
+      "learning_rate": 1.758268910874853e-05,
+      "loss": 0.7202,
+      "step": 4788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.002904135147123,
+      "learning_rate": 1.7581589566754076e-05,
+      "loss": 0.6931,
+      "step": 4789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.676365673898395,
+      "learning_rate": 1.7580489809143648e-05,
+      "loss": 0.6883,
+      "step": 4790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.98718536532281,
+      "learning_rate": 1.7579389835948525e-05,
+      "loss": 0.7581,
+      "step": 4791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.523750962171704,
+      "learning_rate": 1.757828964719999e-05,
+      "loss": 0.8242,
+      "step": 4792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.675163396059236,
+      "learning_rate": 1.7577189242929325e-05,
+      "loss": 0.6967,
+      "step": 4793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.879273057364701,
+      "learning_rate": 1.7576088623167838e-05,
+      "loss": 0.7332,
+      "step": 4794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.176967468374464,
+      "learning_rate": 1.7574987787946817e-05,
+      "loss": 0.87,
+      "step": 4795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.13362650544997,
+      "learning_rate": 1.7573886737297575e-05,
+      "loss": 0.7196,
+      "step": 4796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.67366215048563,
+      "learning_rate": 1.757278547125143e-05,
+      "loss": 0.8146,
+      "step": 4797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 22.81321570893647,
+      "learning_rate": 1.7571683989839693e-05,
+      "loss": 0.7774,
+      "step": 4798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.753856486805756,
+      "learning_rate": 1.757058229309369e-05,
+      "loss": 0.7166,
+      "step": 4799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 23.24128792472426,
+      "learning_rate": 1.7569480381044758e-05,
+      "loss": 0.7229,
+      "step": 4800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.485896989409332,
+      "learning_rate": 1.756837825372423e-05,
+      "loss": 0.7756,
+      "step": 4801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.792856612255914,
+      "learning_rate": 1.7567275911163454e-05,
+      "loss": 0.7584,
+      "step": 4802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.3441967373095,
+      "learning_rate": 1.7566173353393778e-05,
+      "loss": 0.7485,
+      "step": 4803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.569963981785115,
+      "learning_rate": 1.7565070580446556e-05,
+      "loss": 0.6857,
+      "step": 4804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.88349512548097,
+      "learning_rate": 1.7563967592353152e-05,
+      "loss": 0.852,
+      "step": 4805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.22333099633842,
+      "learning_rate": 1.7562864389144936e-05,
+      "loss": 0.7922,
+      "step": 4806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 22.25885159312121,
+      "learning_rate": 1.756176097085328e-05,
+      "loss": 0.7922,
+      "step": 4807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.06409000577669,
+      "learning_rate": 1.7560657337509567e-05,
+      "loss": 0.826,
+      "step": 4808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.984912386875923,
+      "learning_rate": 1.755955348914518e-05,
+      "loss": 0.6804,
+      "step": 4809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 29.425783566771443,
+      "learning_rate": 1.7558449425791515e-05,
+      "loss": 0.7631,
+      "step": 4810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.734042760078,
+      "learning_rate": 1.7557345147479968e-05,
+      "loss": 0.7474,
+      "step": 4811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 33.94893108040185,
+      "learning_rate": 1.755624065424195e-05,
+      "loss": 0.7507,
+      "step": 4812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.545682669575285,
+      "learning_rate": 1.7555135946108866e-05,
+      "loss": 0.6758,
+      "step": 4813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 26.479266283301598,
+      "learning_rate": 1.7554031023112134e-05,
+      "loss": 0.8218,
+      "step": 4814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.218782299196178,
+      "learning_rate": 1.755292588528318e-05,
+      "loss": 0.7695,
+      "step": 4815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 24.94502200905209,
+      "learning_rate": 1.755182053265343e-05,
+      "loss": 0.7807,
+      "step": 4816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.34075820996803,
+      "learning_rate": 1.7550714965254325e-05,
+      "loss": 0.7112,
+      "step": 4817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.989275127157576,
+      "learning_rate": 1.7549609183117305e-05,
+      "loss": 0.7459,
+      "step": 4818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.15197623845428,
+      "learning_rate": 1.7548503186273812e-05,
+      "loss": 0.7195,
+      "step": 4819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 27.064862393196414,
+      "learning_rate": 1.7547396974755307e-05,
+      "loss": 0.7602,
+      "step": 4820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.261950742693383,
+      "learning_rate": 1.7546290548593242e-05,
+      "loss": 0.7356,
+      "step": 4821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 31.706864058009234,
+      "learning_rate": 1.7545183907819094e-05,
+      "loss": 0.7768,
+      "step": 4822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.51565134792058,
+      "learning_rate": 1.7544077052464327e-05,
+      "loss": 0.7456,
+      "step": 4823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.91997191231309,
+      "learning_rate": 1.7542969982560424e-05,
+      "loss": 0.7127,
+      "step": 4824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.831934449967568,
+      "learning_rate": 1.754186269813886e-05,
+      "loss": 0.6598,
+      "step": 4825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.17699316988014,
+      "learning_rate": 1.7540755199231145e-05,
+      "loss": 0.7775,
+      "step": 4826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.772209228806588,
+      "learning_rate": 1.7539647485868753e-05,
+      "loss": 0.6638,
+      "step": 4827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.126662285225219,
+      "learning_rate": 1.75385395580832e-05,
+      "loss": 0.8396,
+      "step": 4828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 23.78153301652085,
+      "learning_rate": 1.7537431415905995e-05,
+      "loss": 0.7789,
+      "step": 4829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.539708284450187,
+      "learning_rate": 1.7536323059368644e-05,
+      "loss": 0.8224,
+      "step": 4830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.345940725830468,
+      "learning_rate": 1.7535214488502677e-05,
+      "loss": 0.6338,
+      "step": 4831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 22.021739634453333,
+      "learning_rate": 1.753410570333962e-05,
+      "loss": 0.7438,
+      "step": 4832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.167628832709756,
+      "learning_rate": 1.7532996703911002e-05,
+      "loss": 0.7499,
+      "step": 4833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.612009739025503,
+      "learning_rate": 1.7531887490248364e-05,
+      "loss": 0.7419,
+      "step": 4834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.86547834964394,
+      "learning_rate": 1.7530778062383253e-05,
+      "loss": 0.7748,
+      "step": 4835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.43928118037929,
+      "learning_rate": 1.752966842034722e-05,
+      "loss": 0.7467,
+      "step": 4836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.462477024758247,
+      "learning_rate": 1.752855856417182e-05,
+      "loss": 0.7031,
+      "step": 4837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.15741089797868,
+      "learning_rate": 1.752744849388862e-05,
+      "loss": 0.8256,
+      "step": 4838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.898801636149011,
+      "learning_rate": 1.7526338209529184e-05,
+      "loss": 0.7411,
+      "step": 4839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.635275707134667,
+      "learning_rate": 1.7525227711125098e-05,
+      "loss": 0.7647,
+      "step": 4840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.956492013688138,
+      "learning_rate": 1.7524116998707933e-05,
+      "loss": 0.6666,
+      "step": 4841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.724817775681023,
+      "learning_rate": 1.7523006072309286e-05,
+      "loss": 0.7866,
+      "step": 4842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 30.70384077252155,
+      "learning_rate": 1.7521894931960742e-05,
+      "loss": 0.8051,
+      "step": 4843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.734744386778676,
+      "learning_rate": 1.7520783577693912e-05,
+      "loss": 0.7635,
+      "step": 4844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.164976594243466,
+      "learning_rate": 1.7519672009540394e-05,
+      "loss": 0.6252,
+      "step": 4845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 25.772837018721546,
+      "learning_rate": 1.7518560227531806e-05,
+      "loss": 0.7612,
+      "step": 4846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.238284836671102,
+      "learning_rate": 1.7517448231699758e-05,
+      "loss": 0.6688,
+      "step": 4847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 29.834792932520127,
+      "learning_rate": 1.7516336022075883e-05,
+      "loss": 0.7549,
+      "step": 4848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.412196228438255,
+      "learning_rate": 1.751522359869181e-05,
+      "loss": 0.7305,
+      "step": 4849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.79203448662454,
+      "learning_rate": 1.751411096157917e-05,
+      "loss": 0.7108,
+      "step": 4850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.71893000538938,
+      "learning_rate": 1.7512998110769613e-05,
+      "loss": 0.782,
+      "step": 4851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.70798477915841,
+      "learning_rate": 1.7511885046294782e-05,
+      "loss": 0.7218,
+      "step": 4852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.088597666822793,
+      "learning_rate": 1.751077176818634e-05,
+      "loss": 0.7287,
+      "step": 4853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.38114358112564,
+      "learning_rate": 1.7509658276475936e-05,
+      "loss": 0.7659,
+      "step": 4854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 22.80805134834718,
+      "learning_rate": 1.750854457119525e-05,
+      "loss": 0.8347,
+      "step": 4855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.59051237553145,
+      "learning_rate": 1.7507430652375943e-05,
+      "loss": 0.7803,
+      "step": 4856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.52014780433687,
+      "learning_rate": 1.7506316520049704e-05,
+      "loss": 0.6656,
+      "step": 4857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.592379901597482,
+      "learning_rate": 1.750520217424821e-05,
+      "loss": 0.7752,
+      "step": 4858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.259305076713584,
+      "learning_rate": 1.750408761500316e-05,
+      "loss": 0.7609,
+      "step": 4859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.62497678217625,
+      "learning_rate": 1.7502972842346248e-05,
+      "loss": 0.738,
+      "step": 4860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.26723370305961,
+      "learning_rate": 1.7501857856309176e-05,
+      "loss": 0.7739,
+      "step": 4861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 33.64519298872203,
+      "learning_rate": 1.7500742656923658e-05,
+      "loss": 0.7796,
+      "step": 4862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.616099507779385,
+      "learning_rate": 1.7499627244221403e-05,
+      "loss": 0.7832,
+      "step": 4863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.383847361173087,
+      "learning_rate": 1.7498511618234134e-05,
+      "loss": 0.7442,
+      "step": 4864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.867646029776123,
+      "learning_rate": 1.7497395778993585e-05,
+      "loss": 0.6778,
+      "step": 4865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 24.667337178461022,
+      "learning_rate": 1.749627972653149e-05,
+      "loss": 0.7646,
+      "step": 4866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.585988483431796,
+      "learning_rate": 1.749516346087958e-05,
+      "loss": 0.6406,
+      "step": 4867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.86307876541342,
+      "learning_rate": 1.7494046982069604e-05,
+      "loss": 0.7236,
+      "step": 4868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.67129283852544,
+      "learning_rate": 1.749293029013332e-05,
+      "loss": 0.7443,
+      "step": 4869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.84446944078611,
+      "learning_rate": 1.749181338510248e-05,
+      "loss": 0.7368,
+      "step": 4870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.127623775734374,
+      "learning_rate": 1.7490696267008845e-05,
+      "loss": 0.6603,
+      "step": 4871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.399211668374484,
+      "learning_rate": 1.74895789358842e-05,
+      "loss": 0.6618,
+      "step": 4872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 25.01011514832804,
+      "learning_rate": 1.7488461391760304e-05,
+      "loss": 0.8076,
+      "step": 4873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.397100702130807,
+      "learning_rate": 1.748734363466895e-05,
+      "loss": 0.6843,
+      "step": 4874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.45345986152622,
+      "learning_rate": 1.748622566464192e-05,
+      "loss": 0.7117,
+      "step": 4875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.40187664018827,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.6721,
+      "step": 4876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 33.539235193912916,
+      "learning_rate": 1.7483989085908027e-05,
+      "loss": 0.8621,
+      "step": 4877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.342273169264676,
+      "learning_rate": 1.748287047726477e-05,
+      "loss": 0.7188,
+      "step": 4878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 25.199967304182035,
+      "learning_rate": 1.7481751655813056e-05,
+      "loss": 0.7636,
+      "step": 4879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.462494589111566,
+      "learning_rate": 1.74806326215847e-05,
+      "loss": 0.8107,
+      "step": 4880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.340582240645666,
+      "learning_rate": 1.747951337461153e-05,
+      "loss": 0.5871,
+      "step": 4881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 29.550130129219415,
+      "learning_rate": 1.747839391492537e-05,
+      "loss": 0.7369,
+      "step": 4882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.411405450608832,
+      "learning_rate": 1.7477274242558064e-05,
+      "loss": 0.7911,
+      "step": 4883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.689339783440087,
+      "learning_rate": 1.7476154357541455e-05,
+      "loss": 0.8549,
+      "step": 4884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.88998864887052,
+      "learning_rate": 1.747503425990739e-05,
+      "loss": 0.8061,
+      "step": 4885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.532968591767025,
+      "learning_rate": 1.747391394968772e-05,
+      "loss": 0.7671,
+      "step": 4886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.844411140575536,
+      "learning_rate": 1.7472793426914313e-05,
+      "loss": 0.7971,
+      "step": 4887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.986351136859465,
+      "learning_rate": 1.7471672691619028e-05,
+      "loss": 0.704,
+      "step": 4888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.54876558019364,
+      "learning_rate": 1.7470551743833747e-05,
+      "loss": 0.7295,
+      "step": 4889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.1104018557556,
+      "learning_rate": 1.746943058359034e-05,
+      "loss": 0.8514,
+      "step": 4890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.648603297160975,
+      "learning_rate": 1.74683092109207e-05,
+      "loss": 0.7439,
+      "step": 4891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.10021493246341,
+      "learning_rate": 1.7467187625856713e-05,
+      "loss": 0.7427,
+      "step": 4892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.892852968974157,
+      "learning_rate": 1.7466065828430277e-05,
+      "loss": 0.7124,
+      "step": 4893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.42721033852216,
+      "learning_rate": 1.7464943818673302e-05,
+      "loss": 0.7044,
+      "step": 4894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 22.740172822091314,
+      "learning_rate": 1.7463821596617686e-05,
+      "loss": 0.8082,
+      "step": 4895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.424625993600092,
+      "learning_rate": 1.7462699162295353e-05,
+      "loss": 0.7107,
+      "step": 4896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.28094236211586,
+      "learning_rate": 1.746157651573822e-05,
+      "loss": 0.7964,
+      "step": 4897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.424930058330958,
+      "learning_rate": 1.7460453656978217e-05,
+      "loss": 0.7084,
+      "step": 4898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.99073109747475,
+      "learning_rate": 1.7459330586047273e-05,
+      "loss": 0.6947,
+      "step": 4899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.261263635849495,
+      "learning_rate": 1.7458207302977333e-05,
+      "loss": 0.7776,
+      "step": 4900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 21.060236984020634,
+      "learning_rate": 1.7457083807800342e-05,
+      "loss": 0.8081,
+      "step": 4901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.611709005554005,
+      "learning_rate": 1.745596010054825e-05,
+      "loss": 0.8036,
+      "step": 4902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.718040050439416,
+      "learning_rate": 1.745483618125301e-05,
+      "loss": 0.7284,
+      "step": 4903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 32.20314612982518,
+      "learning_rate": 1.745371204994659e-05,
+      "loss": 0.8614,
+      "step": 4904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.952299269384426,
+      "learning_rate": 1.7452587706660967e-05,
+      "loss": 0.7266,
+      "step": 4905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.38758085808322,
+      "learning_rate": 1.7451463151428104e-05,
+      "loss": 0.8472,
+      "step": 4906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.177508221646757,
+      "learning_rate": 1.7450338384279985e-05,
+      "loss": 0.7725,
+      "step": 4907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.94373766611307,
+      "learning_rate": 1.7449213405248607e-05,
+      "loss": 0.6757,
+      "step": 4908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.264865436291009,
+      "learning_rate": 1.7448088214365953e-05,
+      "loss": 0.7085,
+      "step": 4909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.981295953661448,
+      "learning_rate": 1.744696281166403e-05,
+      "loss": 0.7812,
+      "step": 4910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.35133221025492,
+      "learning_rate": 1.7445837197174836e-05,
+      "loss": 0.665,
+      "step": 4911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.685280684676613,
+      "learning_rate": 1.744471137093039e-05,
+      "loss": 0.7484,
+      "step": 4912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.87073515685023,
+      "learning_rate": 1.7443585332962715e-05,
+      "loss": 0.6615,
+      "step": 4913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.15167632578306,
+      "learning_rate": 1.744245908330382e-05,
+      "loss": 0.7008,
+      "step": 4914
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.078062662109442,
+      "learning_rate": 1.7441332621985742e-05,
+      "loss": 0.7807,
+      "step": 4915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.922468637256276,
+      "learning_rate": 1.744020594904052e-05,
+      "loss": 0.7061,
+      "step": 4916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 25.984085288414054,
+      "learning_rate": 1.7439079064500193e-05,
+      "loss": 0.7583,
+      "step": 4917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 375.8586389997877,
+      "learning_rate": 1.7437951968396808e-05,
+      "loss": 0.7949,
+      "step": 4918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.152516417959635,
+      "learning_rate": 1.743682466076242e-05,
+      "loss": 0.7221,
+      "step": 4919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.097579075819894,
+      "learning_rate": 1.7435697141629087e-05,
+      "loss": 0.7473,
+      "step": 4920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.952375938958086,
+      "learning_rate": 1.7434569411028883e-05,
+      "loss": 0.8087,
+      "step": 4921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.23459097673864,
+      "learning_rate": 1.743344146899387e-05,
+      "loss": 0.7221,
+      "step": 4922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 21.46926016388806,
+      "learning_rate": 1.7432313315556132e-05,
+      "loss": 0.7627,
+      "step": 4923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.914724127061033,
+      "learning_rate": 1.743118495074775e-05,
+      "loss": 0.6511,
+      "step": 4924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.53497233552534,
+      "learning_rate": 1.7430056374600813e-05,
+      "loss": 0.6766,
+      "step": 4925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 29.61839743366324,
+      "learning_rate": 1.7428927587147422e-05,
+      "loss": 0.7791,
+      "step": 4926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.54962927979832,
+      "learning_rate": 1.7427798588419674e-05,
+      "loss": 0.7249,
+      "step": 4927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.21038336777368,
+      "learning_rate": 1.7426669378449685e-05,
+      "loss": 0.7438,
+      "step": 4928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.168165871737898,
+      "learning_rate": 1.7425539957269557e-05,
+      "loss": 0.6793,
+      "step": 4929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.76502212412781,
+      "learning_rate": 1.742441032491142e-05,
+      "loss": 0.8016,
+      "step": 4930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 30.406634951680513,
+      "learning_rate": 1.7423280481407393e-05,
+      "loss": 0.8101,
+      "step": 4931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.897400229519402,
+      "learning_rate": 1.7422150426789613e-05,
+      "loss": 0.7459,
+      "step": 4932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.69640883254658,
+      "learning_rate": 1.7421020161090216e-05,
+      "loss": 0.6963,
+      "step": 4933
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.31423682794454,
+      "learning_rate": 1.741988968434135e-05,
+      "loss": 0.7501,
+      "step": 4934
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.355686037896728,
+      "learning_rate": 1.741875899657516e-05,
+      "loss": 0.6719,
+      "step": 4935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.52960706373916,
+      "learning_rate": 1.7417628097823802e-05,
+      "loss": 0.7426,
+      "step": 4936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 24.38273804178698,
+      "learning_rate": 1.741649698811944e-05,
+      "loss": 0.715,
+      "step": 4937
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.799724867602897,
+      "learning_rate": 1.7415365667494248e-05,
+      "loss": 0.7854,
+      "step": 4938
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.92200338982838,
+      "learning_rate": 1.741423413598039e-05,
+      "loss": 0.84,
+      "step": 4939
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.316727055810073,
+      "learning_rate": 1.741310239361005e-05,
+      "loss": 0.8222,
+      "step": 4940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.01069974595615,
+      "learning_rate": 1.7411970440415418e-05,
+      "loss": 0.7209,
+      "step": 4941
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.595927858845673,
+      "learning_rate": 1.7410838276428677e-05,
+      "loss": 0.7543,
+      "step": 4942
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.564628310031143,
+      "learning_rate": 1.7409705901682033e-05,
+      "loss": 0.804,
+      "step": 4943
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.485886607309236,
+      "learning_rate": 1.740857331620769e-05,
+      "loss": 0.6732,
+      "step": 4944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.846433901650936,
+      "learning_rate": 1.740744052003785e-05,
+      "loss": 0.7495,
+      "step": 4945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.38669637237412,
+      "learning_rate": 1.7406307513204742e-05,
+      "loss": 0.7389,
+      "step": 4946
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.912706189107379,
+      "learning_rate": 1.7405174295740577e-05,
+      "loss": 0.6628,
+      "step": 4947
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.065252702231586,
+      "learning_rate": 1.7404040867677587e-05,
+      "loss": 0.6699,
+      "step": 4948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.96734736895513,
+      "learning_rate": 1.740290722904801e-05,
+      "loss": 0.7858,
+      "step": 4949
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.967239962029407,
+      "learning_rate": 1.7401773379884078e-05,
+      "loss": 0.7964,
+      "step": 4950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.241067718393687,
+      "learning_rate": 1.7400639320218042e-05,
+      "loss": 0.7308,
+      "step": 4951
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.299745099140768,
+      "learning_rate": 1.739950505008215e-05,
+      "loss": 0.6156,
+      "step": 4952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.602985599130843,
+      "learning_rate": 1.7398370569508667e-05,
+      "loss": 0.7483,
+      "step": 4953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.337192921021662,
+      "learning_rate": 1.739723587852985e-05,
+      "loss": 0.7972,
+      "step": 4954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.357646844570873,
+      "learning_rate": 1.7396100977177975e-05,
+      "loss": 0.7109,
+      "step": 4955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.871261652590274,
+      "learning_rate": 1.7394965865485312e-05,
+      "loss": 0.7369,
+      "step": 4956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.456156405465862,
+      "learning_rate": 1.7393830543484147e-05,
+      "loss": 0.7759,
+      "step": 4957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 25.044864504838507,
+      "learning_rate": 1.7392695011206768e-05,
+      "loss": 0.7819,
+      "step": 4958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.588033656940288,
+      "learning_rate": 1.7391559268685464e-05,
+      "loss": 0.7667,
+      "step": 4959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 21.153757569746126,
+      "learning_rate": 1.739042331595254e-05,
+      "loss": 0.8124,
+      "step": 4960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.839408350346112,
+      "learning_rate": 1.73892871530403e-05,
+      "loss": 0.692,
+      "step": 4961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 26.906280886403078,
+      "learning_rate": 1.7388150779981057e-05,
+      "loss": 0.7894,
+      "step": 4962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.98817814852142,
+      "learning_rate": 1.738701419680713e-05,
+      "loss": 0.7946,
+      "step": 4963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.921376115637958,
+      "learning_rate": 1.7385877403550836e-05,
+      "loss": 0.7439,
+      "step": 4964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.48194463783497,
+      "learning_rate": 1.738474040024451e-05,
+      "loss": 0.7844,
+      "step": 4965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 25.02973950099229,
+      "learning_rate": 1.738360318692049e-05,
+      "loss": 0.7565,
+      "step": 4966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.903436595742544,
+      "learning_rate": 1.738246576361111e-05,
+      "loss": 0.6621,
+      "step": 4967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.457844870285887,
+      "learning_rate": 1.7381328130348727e-05,
+      "loss": 0.6961,
+      "step": 4968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 25.22048976456105,
+      "learning_rate": 1.7380190287165686e-05,
+      "loss": 0.7874,
+      "step": 4969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 22.782271585661586,
+      "learning_rate": 1.7379052234094353e-05,
+      "loss": 0.7587,
+      "step": 4970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.653046819361595,
+      "learning_rate": 1.737791397116709e-05,
+      "loss": 0.7627,
+      "step": 4971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 26.451389024828025,
+      "learning_rate": 1.737677549841627e-05,
+      "loss": 0.8129,
+      "step": 4972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.51047892184048,
+      "learning_rate": 1.7375636815874273e-05,
+      "loss": 0.7612,
+      "step": 4973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.190522751359087,
+      "learning_rate": 1.7374497923573473e-05,
+      "loss": 0.6705,
+      "step": 4974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 25.356071201725552,
+      "learning_rate": 1.7373358821546272e-05,
+      "loss": 0.7768,
+      "step": 4975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 22.612871549515546,
+      "learning_rate": 1.7372219509825056e-05,
+      "loss": 0.6925,
+      "step": 4976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.09954965700338,
+      "learning_rate": 1.737107998844223e-05,
+      "loss": 0.7263,
+      "step": 4977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.89917224533569,
+      "learning_rate": 1.7369940257430203e-05,
+      "loss": 0.6812,
+      "step": 4978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.32157037706258,
+      "learning_rate": 1.7368800316821387e-05,
+      "loss": 0.7067,
+      "step": 4979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.21420758448821,
+      "learning_rate": 1.7367660166648197e-05,
+      "loss": 0.7596,
+      "step": 4980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.927504539990547,
+      "learning_rate": 1.7366519806943067e-05,
+      "loss": 0.7706,
+      "step": 4981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.745854640258234,
+      "learning_rate": 1.736537923773842e-05,
+      "loss": 0.7822,
+      "step": 4982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.82011657499333,
+      "learning_rate": 1.736423845906669e-05,
+      "loss": 0.7312,
+      "step": 4983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 40.469379937486714,
+      "learning_rate": 1.7363097470960336e-05,
+      "loss": 0.7486,
+      "step": 4984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.289888085272274,
+      "learning_rate": 1.736195627345179e-05,
+      "loss": 0.7448,
+      "step": 4985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.01380473487693,
+      "learning_rate": 1.736081486657352e-05,
+      "loss": 0.7457,
+      "step": 4986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.8077806962746,
+      "learning_rate": 1.7359673250357977e-05,
+      "loss": 0.687,
+      "step": 4987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 25.211705960943725,
+      "learning_rate": 1.7358531424837633e-05,
+      "loss": 0.7321,
+      "step": 4988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.987147723469857,
+      "learning_rate": 1.7357389390044965e-05,
+      "loss": 0.7628,
+      "step": 4989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 26.934819811715208,
+      "learning_rate": 1.735624714601244e-05,
+      "loss": 0.7881,
+      "step": 4990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.968630449477843,
+      "learning_rate": 1.735510469277255e-05,
+      "loss": 0.7534,
+      "step": 4991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.594671351864825,
+      "learning_rate": 1.7353962030357786e-05,
+      "loss": 0.8809,
+      "step": 4992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.502093776252696,
+      "learning_rate": 1.7352819158800646e-05,
+      "loss": 0.9089,
+      "step": 4993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.798676835568042,
+      "learning_rate": 1.735167607813363e-05,
+      "loss": 0.7955,
+      "step": 4994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.52557333945106,
+      "learning_rate": 1.7350532788389248e-05,
+      "loss": 0.7633,
+      "step": 4995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.960497202377574,
+      "learning_rate": 1.734938928960001e-05,
+      "loss": 0.7301,
+      "step": 4996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.114733823159321,
+      "learning_rate": 1.7348245581798438e-05,
+      "loss": 0.6224,
+      "step": 4997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.798089532073249,
+      "learning_rate": 1.7347101665017064e-05,
+      "loss": 0.6808,
+      "step": 4998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.943326265187848,
+      "learning_rate": 1.7345957539288417e-05,
+      "loss": 0.6658,
+      "step": 4999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 35.25788677916932,
+      "learning_rate": 1.7344813204645033e-05,
+      "loss": 0.8039,
+      "step": 5000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.701814065027627,
+      "learning_rate": 1.734366866111946e-05,
+      "loss": 0.7383,
+      "step": 5001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 24.47408580973461,
+      "learning_rate": 1.7342523908744246e-05,
+      "loss": 0.6607,
+      "step": 5002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 46.325402325204784,
+      "learning_rate": 1.7341378947551946e-05,
+      "loss": 0.7624,
+      "step": 5003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.05457513775683,
+      "learning_rate": 1.7340233777575125e-05,
+      "loss": 0.7722,
+      "step": 5004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.92265491821371,
+      "learning_rate": 1.733908839884635e-05,
+      "loss": 0.7098,
+      "step": 5005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.640089091452882,
+      "learning_rate": 1.733794281139819e-05,
+      "loss": 0.6812,
+      "step": 5006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.545110240138342,
+      "learning_rate": 1.733679701526323e-05,
+      "loss": 0.7906,
+      "step": 5007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 24.377396656467127,
+      "learning_rate": 1.7335651010474057e-05,
+      "loss": 0.7975,
+      "step": 5008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.033133173062968,
+      "learning_rate": 1.7334504797063257e-05,
+      "loss": 0.8084,
+      "step": 5009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.281571613547705,
+      "learning_rate": 1.733335837506344e-05,
+      "loss": 0.7777,
+      "step": 5010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.245447107638785,
+      "learning_rate": 1.733221174450719e-05,
+      "loss": 0.782,
+      "step": 5011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.19563933656489,
+      "learning_rate": 1.7331064905427133e-05,
+      "loss": 0.7788,
+      "step": 5012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.869574781846504,
+      "learning_rate": 1.7329917857855875e-05,
+      "loss": 0.696,
+      "step": 5013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.764099954869234,
+      "learning_rate": 1.7328770601826047e-05,
+      "loss": 0.673,
+      "step": 5014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.066046134154801,
+      "learning_rate": 1.7327623137370265e-05,
+      "loss": 0.8146,
+      "step": 5015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 25.078479193328253,
+      "learning_rate": 1.732647546452117e-05,
+      "loss": 0.7619,
+      "step": 5016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.420310542099985,
+      "learning_rate": 1.7325327583311396e-05,
+      "loss": 0.609,
+      "step": 5017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.662384773600046,
+      "learning_rate": 1.7324179493773593e-05,
+      "loss": 0.8095,
+      "step": 5018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.057423188069045,
+      "learning_rate": 1.732303119594041e-05,
+      "loss": 0.7724,
+      "step": 5019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.762989985979699,
+      "learning_rate": 1.73218826898445e-05,
+      "loss": 0.704,
+      "step": 5020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.115908829384493,
+      "learning_rate": 1.7320733975518533e-05,
+      "loss": 0.6469,
+      "step": 5021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.836261484816877,
+      "learning_rate": 1.7319585052995177e-05,
+      "loss": 0.7122,
+      "step": 5022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.730244977642954,
+      "learning_rate": 1.7318435922307097e-05,
+      "loss": 0.6995,
+      "step": 5023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.261123366428865,
+      "learning_rate": 1.7317286583486983e-05,
+      "loss": 0.8975,
+      "step": 5024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.467894593355588,
+      "learning_rate": 1.7316137036567523e-05,
+      "loss": 0.6568,
+      "step": 5025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.87838022659654,
+      "learning_rate": 1.7314987281581403e-05,
+      "loss": 0.7126,
+      "step": 5026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 26.72086776237663,
+      "learning_rate": 1.731383731856132e-05,
+      "loss": 0.8774,
+      "step": 5027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.906667107057004,
+      "learning_rate": 1.731268714753999e-05,
+      "loss": 0.7681,
+      "step": 5028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.15878858084662,
+      "learning_rate": 1.7311536768550107e-05,
+      "loss": 0.7387,
+      "step": 5029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.56685865958884,
+      "learning_rate": 1.73103861816244e-05,
+      "loss": 0.9333,
+      "step": 5030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.591399114734493,
+      "learning_rate": 1.7309235386795585e-05,
+      "loss": 0.7584,
+      "step": 5031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.610174713067547,
+      "learning_rate": 1.7308084384096395e-05,
+      "loss": 0.7689,
+      "step": 5032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.458683640351946,
+      "learning_rate": 1.7306933173559554e-05,
+      "loss": 0.8298,
+      "step": 5033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.872977671443623,
+      "learning_rate": 1.7305781755217812e-05,
+      "loss": 0.6979,
+      "step": 5034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.247439795895005,
+      "learning_rate": 1.730463012910391e-05,
+      "loss": 0.7918,
+      "step": 5035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 28.85241462860156,
+      "learning_rate": 1.73034782952506e-05,
+      "loss": 0.8116,
+      "step": 5036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.957824468826395,
+      "learning_rate": 1.7302326253690643e-05,
+      "loss": 0.7945,
+      "step": 5037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.111495393334224,
+      "learning_rate": 1.7301174004456794e-05,
+      "loss": 0.6709,
+      "step": 5038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.74365873854793,
+      "learning_rate": 1.730002154758183e-05,
+      "loss": 0.7245,
+      "step": 5039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.507523721912314,
+      "learning_rate": 1.729886888309852e-05,
+      "loss": 0.7429,
+      "step": 5040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.073937723373685,
+      "learning_rate": 1.7297716011039654e-05,
+      "loss": 0.8302,
+      "step": 5041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.857892900153706,
+      "learning_rate": 1.729656293143801e-05,
+      "loss": 0.7171,
+      "step": 5042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.253516744410266,
+      "learning_rate": 1.7295409644326387e-05,
+      "loss": 0.6826,
+      "step": 5043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.94960649562677,
+      "learning_rate": 1.7294256149737577e-05,
+      "loss": 0.7902,
+      "step": 5044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.635953946089323,
+      "learning_rate": 1.7293102447704395e-05,
+      "loss": 0.6542,
+      "step": 5045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.098848532733925,
+      "learning_rate": 1.7291948538259644e-05,
+      "loss": 0.687,
+      "step": 5046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.597190720623495,
+      "learning_rate": 1.7290794421436138e-05,
+      "loss": 0.7154,
+      "step": 5047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.319660975909798,
+      "learning_rate": 1.7289640097266712e-05,
+      "loss": 0.739,
+      "step": 5048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.850288444146614,
+      "learning_rate": 1.728848556578418e-05,
+      "loss": 0.8241,
+      "step": 5049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.412644354745783,
+      "learning_rate": 1.7287330827021382e-05,
+      "loss": 0.7195,
+      "step": 5050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.979657389637712,
+      "learning_rate": 1.7286175881011158e-05,
+      "loss": 0.7461,
+      "step": 5051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.853701563777296,
+      "learning_rate": 1.7285020727786354e-05,
+      "loss": 0.7505,
+      "step": 5052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 25.178964584961466,
+      "learning_rate": 1.7283865367379826e-05,
+      "loss": 0.7019,
+      "step": 5053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.59595817869395,
+      "learning_rate": 1.7282709799824428e-05,
+      "loss": 0.7261,
+      "step": 5054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 25.532895160090938,
+      "learning_rate": 1.7281554025153018e-05,
+      "loss": 0.9205,
+      "step": 5055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.24995895276614,
+      "learning_rate": 1.7280398043398478e-05,
+      "loss": 0.8256,
+      "step": 5056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.58273686706776,
+      "learning_rate": 1.727924185459367e-05,
+      "loss": 0.8316,
+      "step": 5057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 26.184429837137692,
+      "learning_rate": 1.7278085458771485e-05,
+      "loss": 0.7576,
+      "step": 5058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 26.901705563019643,
+      "learning_rate": 1.7276928855964805e-05,
+      "loss": 0.685,
+      "step": 5059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 23.299143703119697,
+      "learning_rate": 1.727577204620653e-05,
+      "loss": 0.716,
+      "step": 5060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 42.928552792021215,
+      "learning_rate": 1.727461502952955e-05,
+      "loss": 0.8264,
+      "step": 5061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.698791213334815,
+      "learning_rate": 1.7273457805966773e-05,
+      "loss": 0.8002,
+      "step": 5062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.83940608694808,
+      "learning_rate": 1.7272300375551116e-05,
+      "loss": 0.6599,
+      "step": 5063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.194257719646686,
+      "learning_rate": 1.7271142738315485e-05,
+      "loss": 0.6953,
+      "step": 5064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 30.302458308170536,
+      "learning_rate": 1.7269984894292814e-05,
+      "loss": 0.8831,
+      "step": 5065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.28154251088473,
+      "learning_rate": 1.726882684351602e-05,
+      "loss": 0.7863,
+      "step": 5066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.52729123853482,
+      "learning_rate": 1.7267668586018044e-05,
+      "loss": 0.729,
+      "step": 5067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.744382800151502,
+      "learning_rate": 1.7266510121831824e-05,
+      "loss": 0.7486,
+      "step": 5068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.712542562381573,
+      "learning_rate": 1.726535145099031e-05,
+      "loss": 0.6335,
+      "step": 5069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.561469044247666,
+      "learning_rate": 1.726419257352645e-05,
+      "loss": 0.6944,
+      "step": 5070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.059782782473214,
+      "learning_rate": 1.7263033489473202e-05,
+      "loss": 0.7326,
+      "step": 5071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 28.439633716454974,
+      "learning_rate": 1.7261874198863533e-05,
+      "loss": 0.8421,
+      "step": 5072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.85165307626801,
+      "learning_rate": 1.7260714701730405e-05,
+      "loss": 0.6602,
+      "step": 5073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.32739867765844,
+      "learning_rate": 1.72595549981068e-05,
+      "loss": 0.7712,
+      "step": 5074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.7142273975174,
+      "learning_rate": 1.72583950880257e-05,
+      "loss": 0.763,
+      "step": 5075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.549305123802387,
+      "learning_rate": 1.7257234971520086e-05,
+      "loss": 0.7109,
+      "step": 5076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.870855238142497,
+      "learning_rate": 1.7256074648622958e-05,
+      "loss": 0.7423,
+      "step": 5077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.662565530964063,
+      "learning_rate": 1.725491411936731e-05,
+      "loss": 0.7172,
+      "step": 5078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.344096756724776,
+      "learning_rate": 1.7253753383786148e-05,
+      "loss": 0.771,
+      "step": 5079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.20295165341256,
+      "learning_rate": 1.7252592441912487e-05,
+      "loss": 0.7137,
+      "step": 5080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.099638587167057,
+      "learning_rate": 1.7251431293779334e-05,
+      "loss": 0.7776,
+      "step": 5081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 35.31761512252748,
+      "learning_rate": 1.725026993941972e-05,
+      "loss": 0.7658,
+      "step": 5082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.967257706907937,
+      "learning_rate": 1.724910837886667e-05,
+      "loss": 0.7261,
+      "step": 5083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 29.378218776803575,
+      "learning_rate": 1.7247946612153215e-05,
+      "loss": 0.8825,
+      "step": 5084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.34233699259649,
+      "learning_rate": 1.7246784639312403e-05,
+      "loss": 0.805,
+      "step": 5085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.742365402292652,
+      "learning_rate": 1.7245622460377274e-05,
+      "loss": 0.7195,
+      "step": 5086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.991661550158632,
+      "learning_rate": 1.724446007538088e-05,
+      "loss": 0.658,
+      "step": 5087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.917805067842313,
+      "learning_rate": 1.724329748435628e-05,
+      "loss": 0.7155,
+      "step": 5088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.351676272709078,
+      "learning_rate": 1.7242134687336535e-05,
+      "loss": 0.701,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.300366654504995,
+      "learning_rate": 1.7240971684354717e-05,
+      "loss": 0.6904,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 24.543106932966747,
+      "learning_rate": 1.72398084754439e-05,
+      "loss": 0.8452,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.219439792659678,
+      "learning_rate": 1.7238645060637165e-05,
+      "loss": 0.7456,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 31.200695625694124,
+      "learning_rate": 1.7237481439967602e-05,
+      "loss": 0.8066,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 31.05674417462991,
+      "learning_rate": 1.7236317613468294e-05,
+      "loss": 0.7463,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.318735471840476,
+      "learning_rate": 1.7235153581172353e-05,
+      "loss": 0.7005,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.241663489024187,
+      "learning_rate": 1.7233989343112872e-05,
+      "loss": 0.7386,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.114435285438898,
+      "learning_rate": 1.723282489932297e-05,
+      "loss": 0.6815,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.384308548542357,
+      "learning_rate": 1.7231660249835756e-05,
+      "loss": 0.6392,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.817426268006322,
+      "learning_rate": 1.7230495394684353e-05,
+      "loss": 0.7207,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.575994480720233,
+      "learning_rate": 1.7229330333901895e-05,
+      "loss": 0.7451,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.362047778565351,
+      "learning_rate": 1.722816506752151e-05,
+      "loss": 0.6762,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 27.31812986382629,
+      "learning_rate": 1.722699959557634e-05,
+      "loss": 0.6989,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.037783050881803,
+      "learning_rate": 1.7225833918099527e-05,
+      "loss": 0.8387,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 20.149734461983485,
+      "learning_rate": 1.7224668035124224e-05,
+      "loss": 0.7087,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 20.376804377223998,
+      "learning_rate": 1.722350194668359e-05,
+      "loss": 0.8158,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.28636350223484,
+      "learning_rate": 1.7222335652810788e-05,
+      "loss": 0.7634,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.50021746565481,
+      "learning_rate": 1.7221169153538978e-05,
+      "loss": 0.7332,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.11290429002052,
+      "learning_rate": 1.7220002448901346e-05,
+      "loss": 0.6832,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.245410589283114,
+      "learning_rate": 1.721883553893107e-05,
+      "loss": 0.7127,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.672782294523607,
+      "learning_rate": 1.7217668423661332e-05,
+      "loss": 0.6993,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.904317338145937,
+      "learning_rate": 1.7216501103125326e-05,
+      "loss": 0.6789,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 26.703068857597138,
+      "learning_rate": 1.721533357735625e-05,
+      "loss": 0.7675,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.342364338243668,
+      "learning_rate": 1.721416584638731e-05,
+      "loss": 0.6744,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.11973672871839,
+      "learning_rate": 1.721299791025171e-05,
+      "loss": 0.7176,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 25.50128034049107,
+      "learning_rate": 1.7211829768982672e-05,
+      "loss": 0.8299,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.21175173677941,
+      "learning_rate": 1.7210661422613412e-05,
+      "loss": 0.8219,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 22.4106500739761,
+      "learning_rate": 1.720949287117716e-05,
+      "loss": 0.806,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 38.73741170044382,
+      "learning_rate": 1.720832411470715e-05,
+      "loss": 0.8398,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.31007118908587,
+      "learning_rate": 1.720715515323662e-05,
+      "loss": 0.7324,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.226876249847756,
+      "learning_rate": 1.7205985986798808e-05,
+      "loss": 0.8309,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.664302245705011,
+      "learning_rate": 1.7204816615426972e-05,
+      "loss": 0.7168,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.232590517474655,
+      "learning_rate": 1.7203647039154367e-05,
+      "loss": 0.6735,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.68743278871682,
+      "learning_rate": 1.7202477258014256e-05,
+      "loss": 0.7258,
+      "step": 5123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 25.935429112939584,
+      "learning_rate": 1.72013072720399e-05,
+      "loss": 0.7146,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.39228771974679,
+      "learning_rate": 1.7200137081264584e-05,
+      "loss": 0.7438,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.290735597346607,
+      "learning_rate": 1.719896668572158e-05,
+      "loss": 0.805,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 21.979866063634727,
+      "learning_rate": 1.7197796085444173e-05,
+      "loss": 0.7777,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 21.221921175017695,
+      "learning_rate": 1.7196625280465656e-05,
+      "loss": 0.6502,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 38.09581345984922,
+      "learning_rate": 1.7195454270819326e-05,
+      "loss": 0.7592,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.941661498438886,
+      "learning_rate": 1.7194283056538486e-05,
+      "loss": 0.8059,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.06014691975666,
+      "learning_rate": 1.7193111637656445e-05,
+      "loss": 0.697,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 43.032591067507944,
+      "learning_rate": 1.7191940014206518e-05,
+      "loss": 0.7742,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.028730004207219,
+      "learning_rate": 1.719076818622202e-05,
+      "loss": 0.7147,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.75989389284088,
+      "learning_rate": 1.7189596153736285e-05,
+      "loss": 0.6818,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 22.58132415862778,
+      "learning_rate": 1.7188423916782637e-05,
+      "loss": 0.7869,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.37672344494824,
+      "learning_rate": 1.7187251475394423e-05,
+      "loss": 0.6452,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 32.285660965811566,
+      "learning_rate": 1.7186078829604978e-05,
+      "loss": 0.7822,
+      "step": 5137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.285971889756457,
+      "learning_rate": 1.7184905979447655e-05,
+      "loss": 0.685,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.100395532166885,
+      "learning_rate": 1.7183732924955808e-05,
+      "loss": 0.7418,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.721195950690603,
+      "learning_rate": 1.7182559666162802e-05,
+      "loss": 0.737,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.197672171376754,
+      "learning_rate": 1.7181386203102e-05,
+      "loss": 0.6763,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 21.89372631094041,
+      "learning_rate": 1.7180212535806776e-05,
+      "loss": 0.7672,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.410474215830988,
+      "learning_rate": 1.717903866431051e-05,
+      "loss": 0.7046,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 24.094264874727255,
+      "learning_rate": 1.717786458864658e-05,
+      "loss": 0.7272,
+      "step": 5144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.63603840510089,
+      "learning_rate": 1.717669030884838e-05,
+      "loss": 0.7827,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.835625627140793,
+      "learning_rate": 1.7175515824949304e-05,
+      "loss": 0.721,
+      "step": 5146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 20.555538623113,
+      "learning_rate": 1.7174341136982763e-05,
+      "loss": 0.707,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.694979288158706,
+      "learning_rate": 1.717316624498215e-05,
+      "loss": 0.7533,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.30406713104073,
+      "learning_rate": 1.7171991148980888e-05,
+      "loss": 0.7635,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.469966176659288,
+      "learning_rate": 1.717081584901239e-05,
+      "loss": 0.6965,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.156253639248597,
+      "learning_rate": 1.7169640345110088e-05,
+      "loss": 0.7239,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.503101974870242,
+      "learning_rate": 1.7168464637307408e-05,
+      "loss": 0.7429,
+      "step": 5152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.495866157500753,
+      "learning_rate": 1.7167288725637786e-05,
+      "loss": 0.7084,
+      "step": 5153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.819476779676247,
+      "learning_rate": 1.7166112610134665e-05,
+      "loss": 0.7504,
+      "step": 5154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.253693590638479,
+      "learning_rate": 1.7164936290831494e-05,
+      "loss": 0.7233,
+      "step": 5155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.691760306642331,
+      "learning_rate": 1.7163759767761727e-05,
+      "loss": 0.6635,
+      "step": 5156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.058269663818134,
+      "learning_rate": 1.716258304095882e-05,
+      "loss": 0.7584,
+      "step": 5157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.380375913891744,
+      "learning_rate": 1.716140611045625e-05,
+      "loss": 0.7411,
+      "step": 5158
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.897040454233423,
+      "learning_rate": 1.716022897628747e-05,
+      "loss": 0.6835,
+      "step": 5159
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.92237858185876,
+      "learning_rate": 1.7159051638485967e-05,
+      "loss": 0.7479,
+      "step": 5160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.31088891612536,
+      "learning_rate": 1.715787409708523e-05,
+      "loss": 0.8402,
+      "step": 5161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.15829211025891,
+      "learning_rate": 1.7156696352118735e-05,
+      "loss": 0.762,
+      "step": 5162
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.895683898265137,
+      "learning_rate": 1.7155518403619983e-05,
+      "loss": 0.6637,
+      "step": 5163
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.200923417126887,
+      "learning_rate": 1.7154340251622476e-05,
+      "loss": 0.7266,
+      "step": 5164
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.680402059006028,
+      "learning_rate": 1.7153161896159717e-05,
+      "loss": 0.8016,
+      "step": 5165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.499018163583965,
+      "learning_rate": 1.7151983337265217e-05,
+      "loss": 0.6742,
+      "step": 5166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.139926349637031,
+      "learning_rate": 1.7150804574972496e-05,
+      "loss": 0.6492,
+      "step": 5167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.32334262778569,
+      "learning_rate": 1.7149625609315076e-05,
+      "loss": 0.7421,
+      "step": 5168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.758706509086169,
+      "learning_rate": 1.7148446440326485e-05,
+      "loss": 0.7625,
+      "step": 5169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.378582217554925,
+      "learning_rate": 1.714726706804026e-05,
+      "loss": 0.7941,
+      "step": 5170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.14861364873735,
+      "learning_rate": 1.714608749248994e-05,
+      "loss": 0.7745,
+      "step": 5171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.184340133993782,
+      "learning_rate": 1.714490771370907e-05,
+      "loss": 0.6667,
+      "step": 5172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 27.235216965958795,
+      "learning_rate": 1.7143727731731204e-05,
+      "loss": 0.8347,
+      "step": 5173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.78384076910924,
+      "learning_rate": 1.7142547546589903e-05,
+      "loss": 0.7187,
+      "step": 5174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.327430603182503,
+      "learning_rate": 1.714136715831873e-05,
+      "loss": 0.7614,
+      "step": 5175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.341016486852183,
+      "learning_rate": 1.7140186566951248e-05,
+      "loss": 0.6902,
+      "step": 5176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.309475983495677,
+      "learning_rate": 1.7139005772521038e-05,
+      "loss": 0.7205,
+      "step": 5177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.225050607200238,
+      "learning_rate": 1.7137824775061682e-05,
+      "loss": 0.735,
+      "step": 5178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.36372698497285,
+      "learning_rate": 1.7136643574606764e-05,
+      "loss": 0.7876,
+      "step": 5179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.597623428331413,
+      "learning_rate": 1.7135462171189877e-05,
+      "loss": 0.6934,
+      "step": 5180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.694557004463306,
+      "learning_rate": 1.713428056484462e-05,
+      "loss": 0.6687,
+      "step": 5181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.611076331549317,
+      "learning_rate": 1.7133098755604595e-05,
+      "loss": 0.7197,
+      "step": 5182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.26785181360096,
+      "learning_rate": 1.713191674350342e-05,
+      "loss": 0.7426,
+      "step": 5183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.46083528019795,
+      "learning_rate": 1.71307345285747e-05,
+      "loss": 0.691,
+      "step": 5184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.027515995777403,
+      "learning_rate": 1.7129552110852064e-05,
+      "loss": 0.7468,
+      "step": 5185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.067577070744028,
+      "learning_rate": 1.7128369490369134e-05,
+      "loss": 0.6679,
+      "step": 5186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.996270422140327,
+      "learning_rate": 1.7127186667159547e-05,
+      "loss": 0.7136,
+      "step": 5187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 28.687060999991512,
+      "learning_rate": 1.712600364125694e-05,
+      "loss": 0.7427,
+      "step": 5188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.87675124948807,
+      "learning_rate": 1.712482041269496e-05,
+      "loss": 0.7439,
+      "step": 5189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.023500437888337,
+      "learning_rate": 1.7123636981507256e-05,
+      "loss": 0.7542,
+      "step": 5190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.226060276080307,
+      "learning_rate": 1.712245334772748e-05,
+      "loss": 0.7316,
+      "step": 5191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.60619267513184,
+      "learning_rate": 1.7121269511389298e-05,
+      "loss": 0.784,
+      "step": 5192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.594555092016382,
+      "learning_rate": 1.7120085472526377e-05,
+      "loss": 0.6827,
+      "step": 5193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.276446784427318,
+      "learning_rate": 1.711890123117239e-05,
+      "loss": 0.7894,
+      "step": 5194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.86603492865205,
+      "learning_rate": 1.711771678736102e-05,
+      "loss": 0.7585,
+      "step": 5195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.78643156013516,
+      "learning_rate": 1.7116532141125947e-05,
+      "loss": 0.7483,
+      "step": 5196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.21745251915064,
+      "learning_rate": 1.711534729250086e-05,
+      "loss": 0.7554,
+      "step": 5197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.214312822676618,
+      "learning_rate": 1.711416224151946e-05,
+      "loss": 0.6873,
+      "step": 5198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.00368890249207,
+      "learning_rate": 1.7112976988215445e-05,
+      "loss": 0.6451,
+      "step": 5199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 29.31716589810737,
+      "learning_rate": 1.711179153262253e-05,
+      "loss": 0.7897,
+      "step": 5200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 24.645334688695545,
+      "learning_rate": 1.711060587477442e-05,
+      "loss": 0.8364,
+      "step": 5201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.248071867907846,
+      "learning_rate": 1.710942001470484e-05,
+      "loss": 0.6988,
+      "step": 5202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.189003249359537,
+      "learning_rate": 1.7108233952447516e-05,
+      "loss": 0.7757,
+      "step": 5203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.825042063029482,
+      "learning_rate": 1.7107047688036175e-05,
+      "loss": 0.6276,
+      "step": 5204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.303921080303425,
+      "learning_rate": 1.7105861221504555e-05,
+      "loss": 0.7765,
+      "step": 5205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.345665834899343,
+      "learning_rate": 1.71046745528864e-05,
+      "loss": 0.7833,
+      "step": 5206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.304084153708335,
+      "learning_rate": 1.7103487682215454e-05,
+      "loss": 0.8053,
+      "step": 5207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.49489460334743,
+      "learning_rate": 1.710230060952548e-05,
+      "loss": 0.7643,
+      "step": 5208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.111518871150807,
+      "learning_rate": 1.7101113334850224e-05,
+      "loss": 0.7566,
+      "step": 5209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.556357640680986,
+      "learning_rate": 1.709992585822346e-05,
+      "loss": 0.8432,
+      "step": 5210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.020274380374257,
+      "learning_rate": 1.709873817967896e-05,
+      "loss": 0.7846,
+      "step": 5211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.61036679706035,
+      "learning_rate": 1.70975502992505e-05,
+      "loss": 0.7227,
+      "step": 5212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.708249171187916,
+      "learning_rate": 1.7096362216971857e-05,
+      "loss": 0.7751,
+      "step": 5213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.708748120169744,
+      "learning_rate": 1.7095173932876827e-05,
+      "loss": 0.7299,
+      "step": 5214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.909691180597793,
+      "learning_rate": 1.7093985446999203e-05,
+      "loss": 0.6862,
+      "step": 5215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.467779997791684,
+      "learning_rate": 1.7092796759372778e-05,
+      "loss": 0.7572,
+      "step": 5216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.466025192414246,
+      "learning_rate": 1.7091607870031362e-05,
+      "loss": 0.8308,
+      "step": 5217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.435405656064976,
+      "learning_rate": 1.7090418779008772e-05,
+      "loss": 0.6645,
+      "step": 5218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.10645770933609,
+      "learning_rate": 1.7089229486338815e-05,
+      "loss": 0.6955,
+      "step": 5219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.041225599807925,
+      "learning_rate": 1.708803999205532e-05,
+      "loss": 0.7755,
+      "step": 5220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 24.964851359696137,
+      "learning_rate": 1.7086850296192118e-05,
+      "loss": 0.7442,
+      "step": 5221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.45873126706079,
+      "learning_rate": 1.7085660398783032e-05,
+      "loss": 0.7241,
+      "step": 5222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 25.149234360924705,
+      "learning_rate": 1.7084470299861915e-05,
+      "loss": 0.6826,
+      "step": 5223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 26.213680701389535,
+      "learning_rate": 1.7083279999462603e-05,
+      "loss": 0.8048,
+      "step": 5224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.257926227654465,
+      "learning_rate": 1.7082089497618955e-05,
+      "loss": 0.7918,
+      "step": 5225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.216825118832926,
+      "learning_rate": 1.708089879436482e-05,
+      "loss": 0.7294,
+      "step": 5226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.91550840486508,
+      "learning_rate": 1.707970788973407e-05,
+      "loss": 0.6509,
+      "step": 5227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.22206902174581,
+      "learning_rate": 1.7078516783760568e-05,
+      "loss": 0.7002,
+      "step": 5228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.471909689941164,
+      "learning_rate": 1.707732547647819e-05,
+      "loss": 0.7251,
+      "step": 5229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 24.20088935978388,
+      "learning_rate": 1.7076133967920813e-05,
+      "loss": 0.7751,
+      "step": 5230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 30.134268155096944,
+      "learning_rate": 1.7074942258122326e-05,
+      "loss": 0.7393,
+      "step": 5231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 30.24635652893284,
+      "learning_rate": 1.707375034711662e-05,
+      "loss": 0.7391,
+      "step": 5232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.511066867966253,
+      "learning_rate": 1.7072558234937597e-05,
+      "loss": 0.7057,
+      "step": 5233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.00314918847954,
+      "learning_rate": 1.707136592161915e-05,
+      "loss": 0.7347,
+      "step": 5234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.868116686264358,
+      "learning_rate": 1.7070173407195193e-05,
+      "loss": 0.8165,
+      "step": 5235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.786545019757952,
+      "learning_rate": 1.7068980691699638e-05,
+      "loss": 0.7533,
+      "step": 5236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.024889323655994,
+      "learning_rate": 1.7067787775166414e-05,
+      "loss": 0.7822,
+      "step": 5237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 26.497529064221553,
+      "learning_rate": 1.7066594657629435e-05,
+      "loss": 0.6936,
+      "step": 5238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.644817973230241,
+      "learning_rate": 1.706540133912264e-05,
+      "loss": 0.7498,
+      "step": 5239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.682022740892513,
+      "learning_rate": 1.7064207819679964e-05,
+      "loss": 0.8008,
+      "step": 5240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.142779933096879,
+      "learning_rate": 1.7063014099335353e-05,
+      "loss": 0.7062,
+      "step": 5241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 29.532680526276256,
+      "learning_rate": 1.7061820178122746e-05,
+      "loss": 0.8107,
+      "step": 5242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.24548046361106,
+      "learning_rate": 1.7060626056076107e-05,
+      "loss": 0.7246,
+      "step": 5243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.7868676935813,
+      "learning_rate": 1.705943173322939e-05,
+      "loss": 0.8513,
+      "step": 5244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.52759771671025,
+      "learning_rate": 1.705823720961657e-05,
+      "loss": 0.7322,
+      "step": 5245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.44824281848892,
+      "learning_rate": 1.705704248527161e-05,
+      "loss": 0.7418,
+      "step": 5246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.02752852306113,
+      "learning_rate": 1.7055847560228495e-05,
+      "loss": 0.7536,
+      "step": 5247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.9162490112364,
+      "learning_rate": 1.7054652434521197e-05,
+      "loss": 0.6988,
+      "step": 5248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.729013259289736,
+      "learning_rate": 1.7053457108183713e-05,
+      "loss": 0.7439,
+      "step": 5249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.012753659661463,
+      "learning_rate": 1.7052261581250034e-05,
+      "loss": 0.7762,
+      "step": 5250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.4150008472449,
+      "learning_rate": 1.705106585375416e-05,
+      "loss": 0.7424,
+      "step": 5251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 28.2915625090407,
+      "learning_rate": 1.70498699257301e-05,
+      "loss": 0.7448,
+      "step": 5252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.18176275901338,
+      "learning_rate": 1.7048673797211863e-05,
+      "loss": 0.7461,
+      "step": 5253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.212802045895916,
+      "learning_rate": 1.7047477468233467e-05,
+      "loss": 0.8229,
+      "step": 5254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.20314254348072,
+      "learning_rate": 1.7046280938828932e-05,
+      "loss": 0.7776,
+      "step": 5255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.645115960251603,
+      "learning_rate": 1.704508420903229e-05,
+      "loss": 0.6577,
+      "step": 5256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 25.789337101271453,
+      "learning_rate": 1.7043887278877577e-05,
+      "loss": 0.6788,
+      "step": 5257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.30974265603834,
+      "learning_rate": 1.7042690148398825e-05,
+      "loss": 0.8397,
+      "step": 5258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.83332377320303,
+      "learning_rate": 1.704149281763009e-05,
+      "loss": 0.7066,
+      "step": 5259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 25.308804916958696,
+      "learning_rate": 1.7040295286605415e-05,
+      "loss": 0.7674,
+      "step": 5260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 26.81848377757063,
+      "learning_rate": 1.7039097555358862e-05,
+      "loss": 0.8434,
+      "step": 5261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.03986832890927,
+      "learning_rate": 1.7037899623924495e-05,
+      "loss": 0.7236,
+      "step": 5262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.090543859637418,
+      "learning_rate": 1.7036701492336378e-05,
+      "loss": 0.7969,
+      "step": 5263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.970197314733156,
+      "learning_rate": 1.7035503160628585e-05,
+      "loss": 0.7631,
+      "step": 5264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.222801017904905,
+      "learning_rate": 1.7034304628835198e-05,
+      "loss": 0.6801,
+      "step": 5265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.482936078396236,
+      "learning_rate": 1.70331058969903e-05,
+      "loss": 0.7057,
+      "step": 5266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.027679157388516,
+      "learning_rate": 1.703190696512799e-05,
+      "loss": 0.6521,
+      "step": 5267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.23955269185322,
+      "learning_rate": 1.703070783328236e-05,
+      "loss": 0.8073,
+      "step": 5268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.96725247722746,
+      "learning_rate": 1.702950850148751e-05,
+      "loss": 0.6897,
+      "step": 5269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.440301662020264,
+      "learning_rate": 1.7028308969777548e-05,
+      "loss": 0.7676,
+      "step": 5270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.081220980339747,
+      "learning_rate": 1.7027109238186593e-05,
+      "loss": 0.753,
+      "step": 5271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.28191990490078,
+      "learning_rate": 1.7025909306748757e-05,
+      "loss": 0.7668,
+      "step": 5272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.205375073967923,
+      "learning_rate": 1.7024709175498174e-05,
+      "loss": 0.6975,
+      "step": 5273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.929077265993786,
+      "learning_rate": 1.7023508844468974e-05,
+      "loss": 0.6462,
+      "step": 5274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.905851414711375,
+      "learning_rate": 1.702230831369529e-05,
+      "loss": 0.8903,
+      "step": 5275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.44763775331927,
+      "learning_rate": 1.7021107583211266e-05,
+      "loss": 0.7497,
+      "step": 5276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.32928537279355,
+      "learning_rate": 1.7019906653051045e-05,
+      "loss": 0.7195,
+      "step": 5277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.683651035194888,
+      "learning_rate": 1.701870552324879e-05,
+      "loss": 0.7452,
+      "step": 5278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.664000125118573,
+      "learning_rate": 1.7017504193838654e-05,
+      "loss": 0.7115,
+      "step": 5279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.395220431268605,
+      "learning_rate": 1.7016302664854802e-05,
+      "loss": 0.728,
+      "step": 5280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.27474301998892,
+      "learning_rate": 1.701510093633141e-05,
+      "loss": 0.6826,
+      "step": 5281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 20.15930575118074,
+      "learning_rate": 1.701389900830265e-05,
+      "loss": 0.8046,
+      "step": 5282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.83775068066402,
+      "learning_rate": 1.70126968808027e-05,
+      "loss": 0.7946,
+      "step": 5283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.934965622871642,
+      "learning_rate": 1.7011494553865762e-05,
+      "loss": 0.7219,
+      "step": 5284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.094198983144597,
+      "learning_rate": 1.7010292027526016e-05,
+      "loss": 0.6905,
+      "step": 5285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.386529025228718,
+      "learning_rate": 1.7009089301817666e-05,
+      "loss": 0.7556,
+      "step": 5286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.50589419609408,
+      "learning_rate": 1.7007886376774917e-05,
+      "loss": 0.7015,
+      "step": 5287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.505807325138306,
+      "learning_rate": 1.7006683252431977e-05,
+      "loss": 0.6908,
+      "step": 5288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.468875830523434,
+      "learning_rate": 1.7005479928823066e-05,
+      "loss": 0.7265,
+      "step": 5289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.890223480283264,
+      "learning_rate": 1.7004276405982404e-05,
+      "loss": 0.7221,
+      "step": 5290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.138049348636994,
+      "learning_rate": 1.7003072683944214e-05,
+      "loss": 0.7783,
+      "step": 5291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.008168949672765,
+      "learning_rate": 1.700186876274274e-05,
+      "loss": 0.803,
+      "step": 5292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.046225805401793,
+      "learning_rate": 1.700066464241221e-05,
+      "loss": 0.8066,
+      "step": 5293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.41782958252252,
+      "learning_rate": 1.6999460322986876e-05,
+      "loss": 0.7059,
+      "step": 5294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.223645780850129,
+      "learning_rate": 1.6998255804500983e-05,
+      "loss": 0.7338,
+      "step": 5295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.688801233732843,
+      "learning_rate": 1.699705108698879e-05,
+      "loss": 0.7164,
+      "step": 5296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.211439051488393,
+      "learning_rate": 1.6995846170484554e-05,
+      "loss": 0.7625,
+      "step": 5297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.793240609090326,
+      "learning_rate": 1.699464105502255e-05,
+      "loss": 0.7275,
+      "step": 5298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.8135925809831,
+      "learning_rate": 1.6993435740637042e-05,
+      "loss": 0.8877,
+      "step": 5299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.932537832555084,
+      "learning_rate": 1.6992230227362317e-05,
+      "loss": 0.6649,
+      "step": 5300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.850596382446989,
+      "learning_rate": 1.6991024515232652e-05,
+      "loss": 0.7352,
+      "step": 5301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.527752671497687,
+      "learning_rate": 1.698981860428234e-05,
+      "loss": 0.7362,
+      "step": 5302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.234730511952906,
+      "learning_rate": 1.6988612494545673e-05,
+      "loss": 0.6077,
+      "step": 5303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.151526856633746,
+      "learning_rate": 1.698740618605696e-05,
+      "loss": 0.7386,
+      "step": 5304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.161854040613125,
+      "learning_rate": 1.69861996788505e-05,
+      "loss": 0.6559,
+      "step": 5305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.160182836669758,
+      "learning_rate": 1.6984992972960606e-05,
+      "loss": 0.756,
+      "step": 5306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.711337702091864,
+      "learning_rate": 1.69837860684216e-05,
+      "loss": 0.6666,
+      "step": 5307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.52688328378112,
+      "learning_rate": 1.6982578965267805e-05,
+      "loss": 0.7041,
+      "step": 5308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.04005928769861,
+      "learning_rate": 1.6981371663533543e-05,
+      "loss": 0.7199,
+      "step": 5309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.322963190580786,
+      "learning_rate": 1.698016416325316e-05,
+      "loss": 0.7607,
+      "step": 5310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.216787203310748,
+      "learning_rate": 1.6978956464460993e-05,
+      "loss": 0.7545,
+      "step": 5311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.328090675901993,
+      "learning_rate": 1.6977748567191384e-05,
+      "loss": 0.729,
+      "step": 5312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.564781120714,
+      "learning_rate": 1.697654047147869e-05,
+      "loss": 0.7159,
+      "step": 5313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 22.83253972749377,
+      "learning_rate": 1.6975332177357258e-05,
+      "loss": 0.8002,
+      "step": 5314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 26.683272427810657,
+      "learning_rate": 1.697412368486147e-05,
+      "loss": 0.7561,
+      "step": 5315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 26.439257556016987,
+      "learning_rate": 1.697291499402568e-05,
+      "loss": 0.7028,
+      "step": 5316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.944124655796955,
+      "learning_rate": 1.6971706104884263e-05,
+      "loss": 0.7289,
+      "step": 5317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.98941214379324,
+      "learning_rate": 1.6970497017471608e-05,
+      "loss": 0.694,
+      "step": 5318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.85206005266668,
+      "learning_rate": 1.6969287731822092e-05,
+      "loss": 0.7589,
+      "step": 5319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.109176435557835,
+      "learning_rate": 1.6968078247970116e-05,
+      "loss": 0.8027,
+      "step": 5320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.40351651197082,
+      "learning_rate": 1.6966868565950062e-05,
+      "loss": 0.696,
+      "step": 5321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 33.89123415982081,
+      "learning_rate": 1.6965658685796347e-05,
+      "loss": 0.8422,
+      "step": 5322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.86977291992447,
+      "learning_rate": 1.6964448607543375e-05,
+      "loss": 0.7781,
+      "step": 5323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.757045640563643,
+      "learning_rate": 1.696323833122555e-05,
+      "loss": 0.7318,
+      "step": 5324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.32630022111386,
+      "learning_rate": 1.696202785687731e-05,
+      "loss": 0.7198,
+      "step": 5325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.799283365704994,
+      "learning_rate": 1.6960817184533068e-05,
+      "loss": 0.7747,
+      "step": 5326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 20.77355076994614,
+      "learning_rate": 1.6959606314227254e-05,
+      "loss": 0.7008,
+      "step": 5327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.142022414391043,
+      "learning_rate": 1.695839524599431e-05,
+      "loss": 0.8104,
+      "step": 5328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.127038314271974,
+      "learning_rate": 1.6957183979868673e-05,
+      "loss": 0.7259,
+      "step": 5329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.896891192166652,
+      "learning_rate": 1.6955972515884798e-05,
+      "loss": 0.6573,
+      "step": 5330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 27.293947471736764,
+      "learning_rate": 1.695476085407713e-05,
+      "loss": 0.804,
+      "step": 5331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.47729253411391,
+      "learning_rate": 1.695354899448013e-05,
+      "loss": 0.8239,
+      "step": 5332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 29.055762164147385,
+      "learning_rate": 1.695233693712827e-05,
+      "loss": 0.7221,
+      "step": 5333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.08650277415116,
+      "learning_rate": 1.695112468205601e-05,
+      "loss": 0.7172,
+      "step": 5334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.250135975018978,
+      "learning_rate": 1.6949912229297834e-05,
+      "loss": 0.6906,
+      "step": 5335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.04451229679336,
+      "learning_rate": 1.6948699578888218e-05,
+      "loss": 0.6787,
+      "step": 5336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.778587713069367,
+      "learning_rate": 1.6947486730861652e-05,
+      "loss": 0.5909,
+      "step": 5337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.90332788855892,
+      "learning_rate": 1.694627368525263e-05,
+      "loss": 0.6412,
+      "step": 5338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.719465036471096,
+      "learning_rate": 1.6945060442095642e-05,
+      "loss": 0.7676,
+      "step": 5339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.381134694038188,
+      "learning_rate": 1.69438470014252e-05,
+      "loss": 0.7599,
+      "step": 5340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 22.198136504812283,
+      "learning_rate": 1.6942633363275812e-05,
+      "loss": 0.7734,
+      "step": 5341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.840605630082592,
+      "learning_rate": 1.694141952768199e-05,
+      "loss": 0.6985,
+      "step": 5342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.279210167727102,
+      "learning_rate": 1.694020549467826e-05,
+      "loss": 0.7811,
+      "step": 5343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.04089971795267,
+      "learning_rate": 1.6938991264299145e-05,
+      "loss": 0.8073,
+      "step": 5344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.756741233425233,
+      "learning_rate": 1.693777683657918e-05,
+      "loss": 0.7204,
+      "step": 5345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.743478395737206,
+      "learning_rate": 1.6936562211552895e-05,
+      "loss": 0.7173,
+      "step": 5346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.310044583768175,
+      "learning_rate": 1.693534738925484e-05,
+      "loss": 0.7546,
+      "step": 5347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.8291652193765,
+      "learning_rate": 1.6934132369719563e-05,
+      "loss": 0.8646,
+      "step": 5348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.072921132344042,
+      "learning_rate": 1.693291715298162e-05,
+      "loss": 0.6123,
+      "step": 5349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.40706843792547,
+      "learning_rate": 1.6931701739075568e-05,
+      "loss": 0.7233,
+      "step": 5350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.430716859603752,
+      "learning_rate": 1.6930486128035974e-05,
+      "loss": 0.7303,
+      "step": 5351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.321403549064605,
+      "learning_rate": 1.6929270319897407e-05,
+      "loss": 0.796,
+      "step": 5352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.42729013989812,
+      "learning_rate": 1.6928054314694443e-05,
+      "loss": 0.8944,
+      "step": 5353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.109657402043945,
+      "learning_rate": 1.6926838112461668e-05,
+      "loss": 0.6913,
+      "step": 5354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.642529293682223,
+      "learning_rate": 1.692562171323367e-05,
+      "loss": 0.7497,
+      "step": 5355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.971605644649218,
+      "learning_rate": 1.6924405117045046e-05,
+      "loss": 0.7153,
+      "step": 5356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.764910216620862,
+      "learning_rate": 1.692318832393039e-05,
+      "loss": 0.7474,
+      "step": 5357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.43158527696729,
+      "learning_rate": 1.6921971333924304e-05,
+      "loss": 0.7534,
+      "step": 5358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.385573703015968,
+      "learning_rate": 1.6920754147061406e-05,
+      "loss": 0.6754,
+      "step": 5359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.13515113838597,
+      "learning_rate": 1.691953676337631e-05,
+      "loss": 0.7512,
+      "step": 5360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.899136083319725,
+      "learning_rate": 1.691831918290363e-05,
+      "loss": 0.8366,
+      "step": 5361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.289847191325563,
+      "learning_rate": 1.6917101405678004e-05,
+      "loss": 0.7825,
+      "step": 5362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.41337792309947,
+      "learning_rate": 1.691588343173406e-05,
+      "loss": 0.7244,
+      "step": 5363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.716262148435145,
+      "learning_rate": 1.6914665261106434e-05,
+      "loss": 0.8078,
+      "step": 5364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.652928943341834,
+      "learning_rate": 1.691344689382978e-05,
+      "loss": 0.7657,
+      "step": 5365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.2795215187537,
+      "learning_rate": 1.6912228329938734e-05,
+      "loss": 0.7703,
+      "step": 5366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.675518267624597,
+      "learning_rate": 1.691100956946796e-05,
+      "loss": 0.6667,
+      "step": 5367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.976858003617421,
+      "learning_rate": 1.6909790612452114e-05,
+      "loss": 0.6166,
+      "step": 5368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.270833606640224,
+      "learning_rate": 1.690857145892587e-05,
+      "loss": 0.6824,
+      "step": 5369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.864918181674533,
+      "learning_rate": 1.6907352108923892e-05,
+      "loss": 0.7064,
+      "step": 5370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.027192003091926,
+      "learning_rate": 1.690613256248086e-05,
+      "loss": 0.7339,
+      "step": 5371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.325723336958,
+      "learning_rate": 1.690491281963146e-05,
+      "loss": 0.7569,
+      "step": 5372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.778871219588932,
+      "learning_rate": 1.690369288041038e-05,
+      "loss": 0.6858,
+      "step": 5373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.28704337037985,
+      "learning_rate": 1.690247274485231e-05,
+      "loss": 0.7345,
+      "step": 5374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.14197846774523,
+      "learning_rate": 1.690125241299195e-05,
+      "loss": 0.8074,
+      "step": 5375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.866728701584494,
+      "learning_rate": 1.6900031884864013e-05,
+      "loss": 0.7064,
+      "step": 5376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.400844693638945,
+      "learning_rate": 1.6898811160503204e-05,
+      "loss": 0.7522,
+      "step": 5377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.037670620170804,
+      "learning_rate": 1.6897590239944242e-05,
+      "loss": 0.8515,
+      "step": 5378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 34.703549224891866,
+      "learning_rate": 1.6896369123221852e-05,
+      "loss": 0.7814,
+      "step": 5379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.171722151099456,
+      "learning_rate": 1.6895147810370755e-05,
+      "loss": 0.6885,
+      "step": 5380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.14283453523767,
+      "learning_rate": 1.6893926301425685e-05,
+      "loss": 0.7007,
+      "step": 5381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.788615669917164,
+      "learning_rate": 1.6892704596421386e-05,
+      "loss": 0.7271,
+      "step": 5382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.446338854851334,
+      "learning_rate": 1.6891482695392603e-05,
+      "loss": 0.7586,
+      "step": 5383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.167327045008195,
+      "learning_rate": 1.6890260598374083e-05,
+      "loss": 0.6385,
+      "step": 5384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.075636376512431,
+      "learning_rate": 1.688903830540058e-05,
+      "loss": 0.5901,
+      "step": 5385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.96259297581637,
+      "learning_rate": 1.6887815816506858e-05,
+      "loss": 0.7269,
+      "step": 5386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.875292372935663,
+      "learning_rate": 1.6886593131727687e-05,
+      "loss": 0.6637,
+      "step": 5387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.162009929500407,
+      "learning_rate": 1.688537025109783e-05,
+      "loss": 0.6942,
+      "step": 5388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.860024641923033,
+      "learning_rate": 1.6884147174652077e-05,
+      "loss": 0.8074,
+      "step": 5389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.332970647263213,
+      "learning_rate": 1.68829239024252e-05,
+      "loss": 0.6906,
+      "step": 5390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.801020061311059,
+      "learning_rate": 1.6881700434451996e-05,
+      "loss": 0.6766,
+      "step": 5391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 30.397282887052942,
+      "learning_rate": 1.6880476770767256e-05,
+      "loss": 0.7628,
+      "step": 5392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.59440561482592,
+      "learning_rate": 1.6879252911405782e-05,
+      "loss": 0.7597,
+      "step": 5393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.572461786664956,
+      "learning_rate": 1.6878028856402382e-05,
+      "loss": 0.5419,
+      "step": 5394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.872457975420073,
+      "learning_rate": 1.6876804605791864e-05,
+      "loss": 0.7453,
+      "step": 5395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.52819200151108,
+      "learning_rate": 1.6875580159609044e-05,
+      "loss": 0.65,
+      "step": 5396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.497709574665823,
+      "learning_rate": 1.6874355517888747e-05,
+      "loss": 0.7285,
+      "step": 5397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.48034715582227,
+      "learning_rate": 1.6873130680665798e-05,
+      "loss": 0.7966,
+      "step": 5398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.77543020116055,
+      "learning_rate": 1.6871905647975038e-05,
+      "loss": 0.6778,
+      "step": 5399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.77171730148631,
+      "learning_rate": 1.6870680419851297e-05,
+      "loss": 0.7279,
+      "step": 5400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.7211224584472,
+      "learning_rate": 1.686945499632943e-05,
+      "loss": 0.8393,
+      "step": 5401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.29289405996517,
+      "learning_rate": 1.6868229377444276e-05,
+      "loss": 0.6865,
+      "step": 5402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.51783646692002,
+      "learning_rate": 1.6867003563230697e-05,
+      "loss": 0.7408,
+      "step": 5403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.786608894086974,
+      "learning_rate": 1.6865777553723558e-05,
+      "loss": 0.676,
+      "step": 5404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.13278356432301,
+      "learning_rate": 1.6864551348957717e-05,
+      "loss": 0.723,
+      "step": 5405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.032227648738216,
+      "learning_rate": 1.686332494896805e-05,
+      "loss": 0.7232,
+      "step": 5406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.030171626454523,
+      "learning_rate": 1.686209835378944e-05,
+      "loss": 0.8027,
+      "step": 5407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.975415536603201,
+      "learning_rate": 1.6860871563456766e-05,
+      "loss": 0.7442,
+      "step": 5408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.163654531435537,
+      "learning_rate": 1.685964457800492e-05,
+      "loss": 0.7449,
+      "step": 5409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.836815275514981,
+      "learning_rate": 1.6858417397468792e-05,
+      "loss": 0.6989,
+      "step": 5410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.020592201147846,
+      "learning_rate": 1.6857190021883287e-05,
+      "loss": 0.7022,
+      "step": 5411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.64287639089137,
+      "learning_rate": 1.685596245128331e-05,
+      "loss": 0.6749,
+      "step": 5412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.36718262545153,
+      "learning_rate": 1.685473468570377e-05,
+      "loss": 0.746,
+      "step": 5413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 147.37386729526654,
+      "learning_rate": 1.6853506725179584e-05,
+      "loss": 0.6308,
+      "step": 5414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.43013961996076,
+      "learning_rate": 1.6852278569745678e-05,
+      "loss": 0.8726,
+      "step": 5415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.542861201709016,
+      "learning_rate": 1.685105021943698e-05,
+      "loss": 0.6851,
+      "step": 5416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.034092360070094,
+      "learning_rate": 1.6849821674288418e-05,
+      "loss": 0.6692,
+      "step": 5417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.483809558016567,
+      "learning_rate": 1.6848592934334934e-05,
+      "loss": 0.6107,
+      "step": 5418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.327963325773197,
+      "learning_rate": 1.6847363999611475e-05,
+      "loss": 0.8643,
+      "step": 5419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.43470713696328,
+      "learning_rate": 1.6846134870152987e-05,
+      "loss": 0.6726,
+      "step": 5420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 25.91558776474682,
+      "learning_rate": 1.684490554599443e-05,
+      "loss": 0.7237,
+      "step": 5421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.068149151594326,
+      "learning_rate": 1.6843676027170764e-05,
+      "loss": 0.7284,
+      "step": 5422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 22.815319879861377,
+      "learning_rate": 1.6842446313716957e-05,
+      "loss": 0.7744,
+      "step": 5423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.963717502821574,
+      "learning_rate": 1.6841216405667976e-05,
+      "loss": 0.7257,
+      "step": 5424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.63696057715217,
+      "learning_rate": 1.6839986303058803e-05,
+      "loss": 0.7225,
+      "step": 5425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.247224644113391,
+      "learning_rate": 1.6838756005924425e-05,
+      "loss": 0.7258,
+      "step": 5426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.773143098158165,
+      "learning_rate": 1.6837525514299823e-05,
+      "loss": 0.7017,
+      "step": 5427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.448104473049431,
+      "learning_rate": 1.6836294828219997e-05,
+      "loss": 0.6968,
+      "step": 5428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.374370569026592,
+      "learning_rate": 1.6835063947719943e-05,
+      "loss": 0.7632,
+      "step": 5429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.36995378960079,
+      "learning_rate": 1.683383287283467e-05,
+      "loss": 0.7148,
+      "step": 5430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.512365207142757,
+      "learning_rate": 1.683260160359919e-05,
+      "loss": 0.7483,
+      "step": 5431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 47.606828254324384,
+      "learning_rate": 1.6831370140048513e-05,
+      "loss": 0.7077,
+      "step": 5432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.440612262962913,
+      "learning_rate": 1.6830138482217667e-05,
+      "loss": 0.6902,
+      "step": 5433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 27.612076012484934,
+      "learning_rate": 1.6828906630141678e-05,
+      "loss": 0.7381,
+      "step": 5434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.960302204889638,
+      "learning_rate": 1.682767458385558e-05,
+      "loss": 0.7493,
+      "step": 5435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.36174247061062,
+      "learning_rate": 1.682644234339441e-05,
+      "loss": 0.777,
+      "step": 5436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.07842362981971,
+      "learning_rate": 1.6825209908793217e-05,
+      "loss": 0.6903,
+      "step": 5437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.78099443880849,
+      "learning_rate": 1.682397728008704e-05,
+      "loss": 0.7493,
+      "step": 5438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.07761205111984,
+      "learning_rate": 1.6822744457310948e-05,
+      "loss": 0.8222,
+      "step": 5439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.066881692623568,
+      "learning_rate": 1.682151144049999e-05,
+      "loss": 0.7605,
+      "step": 5440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.300542475118444,
+      "learning_rate": 1.682027822968924e-05,
+      "loss": 0.7117,
+      "step": 5441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.28297065403295,
+      "learning_rate": 1.6819044824913762e-05,
+      "loss": 0.7171,
+      "step": 5442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 26.16915410203333,
+      "learning_rate": 1.6817811226208643e-05,
+      "loss": 0.7454,
+      "step": 5443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.004041628462364,
+      "learning_rate": 1.681657743360896e-05,
+      "loss": 0.6326,
+      "step": 5444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 25.7243775616582,
+      "learning_rate": 1.6815343447149803e-05,
+      "loss": 0.6992,
+      "step": 5445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.069653091375866,
+      "learning_rate": 1.681410926686626e-05,
+      "loss": 0.7899,
+      "step": 5446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.86751397237882,
+      "learning_rate": 1.6812874892793443e-05,
+      "loss": 0.6546,
+      "step": 5447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.301640517827305,
+      "learning_rate": 1.6811640324966446e-05,
+      "loss": 0.7471,
+      "step": 5448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 26.96025782314002,
+      "learning_rate": 1.681040556342038e-05,
+      "loss": 0.7041,
+      "step": 5449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.10642878801702,
+      "learning_rate": 1.680917060819037e-05,
+      "loss": 0.7166,
+      "step": 5450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.825284947407024,
+      "learning_rate": 1.6807935459311528e-05,
+      "loss": 0.7401,
+      "step": 5451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.84335733979094,
+      "learning_rate": 1.6806700116818982e-05,
+      "loss": 0.7844,
+      "step": 5452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 28.750581994592896,
+      "learning_rate": 1.680546458074787e-05,
+      "loss": 0.6945,
+      "step": 5453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.368784116409262,
+      "learning_rate": 1.6804228851133326e-05,
+      "loss": 0.7021,
+      "step": 5454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.474745071948833,
+      "learning_rate": 1.6802992928010496e-05,
+      "loss": 0.7248,
+      "step": 5455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.29534970388788,
+      "learning_rate": 1.680175681141452e-05,
+      "loss": 0.7332,
+      "step": 5456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.158171505233792,
+      "learning_rate": 1.6800520501380564e-05,
+      "loss": 0.7238,
+      "step": 5457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.72776420089135,
+      "learning_rate": 1.679928399794378e-05,
+      "loss": 0.7926,
+      "step": 5458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.651488879121537,
+      "learning_rate": 1.6798047301139338e-05,
+      "loss": 0.7782,
+      "step": 5459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.378991828550618,
+      "learning_rate": 1.679681041100241e-05,
+      "loss": 0.692,
+      "step": 5460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.95584336348539,
+      "learning_rate": 1.6795573327568168e-05,
+      "loss": 0.7196,
+      "step": 5461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.752709687151274,
+      "learning_rate": 1.6794336050871797e-05,
+      "loss": 0.6768,
+      "step": 5462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.039901966332902,
+      "learning_rate": 1.6793098580948482e-05,
+      "loss": 0.7626,
+      "step": 5463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.38153448188531,
+      "learning_rate": 1.6791860917833417e-05,
+      "loss": 0.7007,
+      "step": 5464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 22.905846184974084,
+      "learning_rate": 1.67906230615618e-05,
+      "loss": 0.7494,
+      "step": 5465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.117005847887409,
+      "learning_rate": 1.6789385012168836e-05,
+      "loss": 0.7091,
+      "step": 5466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.86673968499415,
+      "learning_rate": 1.6788146769689734e-05,
+      "loss": 0.7533,
+      "step": 5467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.723153092673385,
+      "learning_rate": 1.678690833415971e-05,
+      "loss": 0.7846,
+      "step": 5468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.071843660843896,
+      "learning_rate": 1.678566970561398e-05,
+      "loss": 0.8239,
+      "step": 5469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.769982936477426,
+      "learning_rate": 1.678443088408778e-05,
+      "loss": 0.7651,
+      "step": 5470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.14182436378412,
+      "learning_rate": 1.6783191869616327e-05,
+      "loss": 0.678,
+      "step": 5471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.497020667499044,
+      "learning_rate": 1.678195266223487e-05,
+      "loss": 0.6751,
+      "step": 5472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 28.61302556831773,
+      "learning_rate": 1.6780713261978646e-05,
+      "loss": 0.742,
+      "step": 5473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.251215526847925,
+      "learning_rate": 1.67794736688829e-05,
+      "loss": 0.7207,
+      "step": 5474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.657820768118793,
+      "learning_rate": 1.6778233882982894e-05,
+      "loss": 0.7102,
+      "step": 5475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.817946007768445,
+      "learning_rate": 1.6776993904313875e-05,
+      "loss": 0.6995,
+      "step": 5476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.36170447777714,
+      "learning_rate": 1.677575373291112e-05,
+      "loss": 0.8064,
+      "step": 5477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.105219809241028,
+      "learning_rate": 1.6774513368809887e-05,
+      "loss": 0.8047,
+      "step": 5478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.222777176517035,
+      "learning_rate": 1.677327281204546e-05,
+      "loss": 0.6031,
+      "step": 5479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.26237220675424,
+      "learning_rate": 1.6772032062653115e-05,
+      "loss": 0.723,
+      "step": 5480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.956443900727262,
+      "learning_rate": 1.677079112066814e-05,
+      "loss": 0.6891,
+      "step": 5481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 27.18991411715872,
+      "learning_rate": 1.6769549986125827e-05,
+      "loss": 0.7149,
+      "step": 5482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.601994523287907,
+      "learning_rate": 1.6768308659061474e-05,
+      "loss": 0.7284,
+      "step": 5483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 23.101024903307696,
+      "learning_rate": 1.6767067139510383e-05,
+      "loss": 0.8209,
+      "step": 5484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.222146911217841,
+      "learning_rate": 1.6765825427507855e-05,
+      "loss": 0.7532,
+      "step": 5485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 23.919897616914454,
+      "learning_rate": 1.6764583523089214e-05,
+      "loss": 0.7917,
+      "step": 5486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 24.025062797275268,
+      "learning_rate": 1.6763341426289773e-05,
+      "loss": 0.67,
+      "step": 5487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.989780899025297,
+      "learning_rate": 1.676209913714486e-05,
+      "loss": 0.7046,
+      "step": 5488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.214736099628603,
+      "learning_rate": 1.6760856655689804e-05,
+      "loss": 0.6675,
+      "step": 5489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.572673796557577,
+      "learning_rate": 1.675961398195994e-05,
+      "loss": 0.7588,
+      "step": 5490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.24116876277585,
+      "learning_rate": 1.6758371115990607e-05,
+      "loss": 0.7645,
+      "step": 5491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.527954935404674,
+      "learning_rate": 1.6757128057817154e-05,
+      "loss": 0.6699,
+      "step": 5492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.71155999089906,
+      "learning_rate": 1.675588480747493e-05,
+      "loss": 0.6759,
+      "step": 5493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.509173275191113,
+      "learning_rate": 1.6754641364999297e-05,
+      "loss": 0.7605,
+      "step": 5494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.03881733773878,
+      "learning_rate": 1.675339773042561e-05,
+      "loss": 0.755,
+      "step": 5495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.29853177962315,
+      "learning_rate": 1.6752153903789247e-05,
+      "loss": 0.7541,
+      "step": 5496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.967974491386407,
+      "learning_rate": 1.6750909885125575e-05,
+      "loss": 0.7504,
+      "step": 5497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 25.734535862408897,
+      "learning_rate": 1.6749665674469975e-05,
+      "loss": 0.7387,
+      "step": 5498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.494666704626571,
+      "learning_rate": 1.6748421271857835e-05,
+      "loss": 0.7133,
+      "step": 5499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.040793883338267,
+      "learning_rate": 1.6747176677324535e-05,
+      "loss": 0.6771,
+      "step": 5500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.44159683520143,
+      "learning_rate": 1.674593189090548e-05,
+      "loss": 0.6774,
+      "step": 5501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.33537642841377,
+      "learning_rate": 1.674468691263607e-05,
+      "loss": 0.737,
+      "step": 5502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.05528572404925,
+      "learning_rate": 1.674344174255171e-05,
+      "loss": 0.7308,
+      "step": 5503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.338982081574864,
+      "learning_rate": 1.674219638068781e-05,
+      "loss": 0.6879,
+      "step": 5504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.994028790795618,
+      "learning_rate": 1.6740950827079786e-05,
+      "loss": 0.7206,
+      "step": 5505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.307530655447184,
+      "learning_rate": 1.673970508176307e-05,
+      "loss": 0.6481,
+      "step": 5506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.46851334128871,
+      "learning_rate": 1.673845914477308e-05,
+      "loss": 0.7108,
+      "step": 5507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.1834380711694,
+      "learning_rate": 1.6737213016145256e-05,
+      "loss": 0.6724,
+      "step": 5508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.63379417909364,
+      "learning_rate": 1.6735966695915036e-05,
+      "loss": 0.6387,
+      "step": 5509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.719482008337145,
+      "learning_rate": 1.673472018411786e-05,
+      "loss": 0.7228,
+      "step": 5510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.721704262125755,
+      "learning_rate": 1.6733473480789183e-05,
+      "loss": 0.7622,
+      "step": 5511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.955363955890217,
+      "learning_rate": 1.673222658596446e-05,
+      "loss": 0.7596,
+      "step": 5512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.99476577021787,
+      "learning_rate": 1.673097949967915e-05,
+      "loss": 0.8085,
+      "step": 5513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.60209281426596,
+      "learning_rate": 1.6729732221968722e-05,
+      "loss": 0.7025,
+      "step": 5514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.20081843509955,
+      "learning_rate": 1.6728484752868644e-05,
+      "loss": 0.6539,
+      "step": 5515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.372695521366598,
+      "learning_rate": 1.6727237092414397e-05,
+      "loss": 0.7045,
+      "step": 5516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.56112279789204,
+      "learning_rate": 1.6725989240641466e-05,
+      "loss": 0.7136,
+      "step": 5517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.575310276352525,
+      "learning_rate": 1.672474119758533e-05,
+      "loss": 0.7474,
+      "step": 5518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.956381313236854,
+      "learning_rate": 1.6723492963281492e-05,
+      "loss": 0.6584,
+      "step": 5519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.492598974159456,
+      "learning_rate": 1.6722244537765444e-05,
+      "loss": 0.7113,
+      "step": 5520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 24.99055338242651,
+      "learning_rate": 1.6720995921072698e-05,
+      "loss": 0.6908,
+      "step": 5521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.617679999352113,
+      "learning_rate": 1.6719747113238754e-05,
+      "loss": 0.6961,
+      "step": 5522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 21.938635807575874,
+      "learning_rate": 1.6718498114299138e-05,
+      "loss": 0.7129,
+      "step": 5523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.346676757710142,
+      "learning_rate": 1.6717248924289363e-05,
+      "loss": 0.7396,
+      "step": 5524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.889399115692704,
+      "learning_rate": 1.671599954324496e-05,
+      "loss": 0.8214,
+      "step": 5525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.567382454027097,
+      "learning_rate": 1.6714749971201457e-05,
+      "loss": 0.7353,
+      "step": 5526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.615304135524894,
+      "learning_rate": 1.6713500208194395e-05,
+      "loss": 0.7089,
+      "step": 5527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.31624027848153,
+      "learning_rate": 1.6712250254259313e-05,
+      "loss": 0.6713,
+      "step": 5528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.378537556372276,
+      "learning_rate": 1.6711000109431757e-05,
+      "loss": 0.7627,
+      "step": 5529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.144232575649506,
+      "learning_rate": 1.6709749773747292e-05,
+      "loss": 0.6812,
+      "step": 5530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.75782443736315,
+      "learning_rate": 1.670849924724146e-05,
+      "loss": 0.7546,
+      "step": 5531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.827908124889206,
+      "learning_rate": 1.670724852994984e-05,
+      "loss": 0.7323,
+      "step": 5532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.193446520468456,
+      "learning_rate": 1.6705997621907993e-05,
+      "loss": 0.7449,
+      "step": 5533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 21.531535550896763,
+      "learning_rate": 1.6704746523151498e-05,
+      "loss": 0.7342,
+      "step": 5534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.871039653514718,
+      "learning_rate": 1.670349523371593e-05,
+      "loss": 0.8125,
+      "step": 5535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.494552064670486,
+      "learning_rate": 1.670224375363688e-05,
+      "loss": 0.7303,
+      "step": 5536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.702676063156925,
+      "learning_rate": 1.670099208294994e-05,
+      "loss": 0.7046,
+      "step": 5537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.632325735255055,
+      "learning_rate": 1.6699740221690706e-05,
+      "loss": 0.8022,
+      "step": 5538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.587098073427606,
+      "learning_rate": 1.669848816989478e-05,
+      "loss": 0.7318,
+      "step": 5539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.253821634348277,
+      "learning_rate": 1.6697235927597763e-05,
+      "loss": 0.7121,
+      "step": 5540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 23.327291501224117,
+      "learning_rate": 1.6695983494835283e-05,
+      "loss": 0.7967,
+      "step": 5541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.954009283318214,
+      "learning_rate": 1.669473087164294e-05,
+      "loss": 0.7708,
+      "step": 5542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.513382041747036,
+      "learning_rate": 1.6693478058056375e-05,
+      "loss": 0.7594,
+      "step": 5543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.237630234348744,
+      "learning_rate": 1.6692225054111207e-05,
+      "loss": 0.6948,
+      "step": 5544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.505037142989893,
+      "learning_rate": 1.669097185984307e-05,
+      "loss": 0.696,
+      "step": 5545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.21335438607501,
+      "learning_rate": 1.668971847528761e-05,
+      "loss": 0.6997,
+      "step": 5546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.882450275568763,
+      "learning_rate": 1.668846490048047e-05,
+      "loss": 0.677,
+      "step": 5547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.483848613025124,
+      "learning_rate": 1.6687211135457304e-05,
+      "loss": 0.8614,
+      "step": 5548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.915116410356028,
+      "learning_rate": 1.668595718025376e-05,
+      "loss": 0.7381,
+      "step": 5549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.84288955139238,
+      "learning_rate": 1.6684703034905507e-05,
+      "loss": 0.7433,
+      "step": 5550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.270744241049968,
+      "learning_rate": 1.668344869944821e-05,
+      "loss": 0.6813,
+      "step": 5551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.405522495247816,
+      "learning_rate": 1.6682194173917543e-05,
+      "loss": 0.7991,
+      "step": 5552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.065625722874085,
+      "learning_rate": 1.6680939458349184e-05,
+      "loss": 0.6658,
+      "step": 5553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.149458870886434,
+      "learning_rate": 1.667968455277881e-05,
+      "loss": 0.8056,
+      "step": 5554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.546887985921607,
+      "learning_rate": 1.667842945724212e-05,
+      "loss": 0.675,
+      "step": 5555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.620954666543469,
+      "learning_rate": 1.6677174171774798e-05,
+      "loss": 0.6429,
+      "step": 5556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.3793106010298,
+      "learning_rate": 1.6675918696412552e-05,
+      "loss": 0.6852,
+      "step": 5557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.533755819574274,
+      "learning_rate": 1.6674663031191084e-05,
+      "loss": 0.7247,
+      "step": 5558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.292760146952514,
+      "learning_rate": 1.66734071761461e-05,
+      "loss": 0.7152,
+      "step": 5559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 23.396429482018345,
+      "learning_rate": 1.6672151131313324e-05,
+      "loss": 0.7848,
+      "step": 5560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.14182856454198,
+      "learning_rate": 1.6670894896728472e-05,
+      "loss": 0.7138,
+      "step": 5561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.90061012471951,
+      "learning_rate": 1.666963847242727e-05,
+      "loss": 0.7846,
+      "step": 5562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.584828852010293,
+      "learning_rate": 1.666838185844545e-05,
+      "loss": 0.7686,
+      "step": 5563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.569462244899825,
+      "learning_rate": 1.6667125054818756e-05,
+      "loss": 0.6719,
+      "step": 5564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 26.307846550925944,
+      "learning_rate": 1.6665868061582922e-05,
+      "loss": 0.7473,
+      "step": 5565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 31.89063735243966,
+      "learning_rate": 1.66646108787737e-05,
+      "loss": 0.6967,
+      "step": 5566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.311779746417574,
+      "learning_rate": 1.6663353506426845e-05,
+      "loss": 0.7495,
+      "step": 5567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 23.69390050301414,
+      "learning_rate": 1.6662095944578112e-05,
+      "loss": 0.6886,
+      "step": 5568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.175396334712502,
+      "learning_rate": 1.6660838193263268e-05,
+      "loss": 0.744,
+      "step": 5569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.753154560751634,
+      "learning_rate": 1.6659580252518083e-05,
+      "loss": 0.7076,
+      "step": 5570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.184822257793588,
+      "learning_rate": 1.665832212237833e-05,
+      "loss": 0.7598,
+      "step": 5571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.138235499359693,
+      "learning_rate": 1.6657063802879795e-05,
+      "loss": 0.6107,
+      "step": 5572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.665448903465165,
+      "learning_rate": 1.6655805294058256e-05,
+      "loss": 0.8026,
+      "step": 5573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 26.5927127239132,
+      "learning_rate": 1.665454659594951e-05,
+      "loss": 0.7968,
+      "step": 5574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.468444241270436,
+      "learning_rate": 1.665328770858935e-05,
+      "loss": 0.7041,
+      "step": 5575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.703204099191392,
+      "learning_rate": 1.665202863201358e-05,
+      "loss": 0.7069,
+      "step": 5576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.347136781098744,
+      "learning_rate": 1.6650769366258007e-05,
+      "loss": 0.6781,
+      "step": 5577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.703343744901467,
+      "learning_rate": 1.6649509911358444e-05,
+      "loss": 0.6943,
+      "step": 5578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.322211212314883,
+      "learning_rate": 1.6648250267350708e-05,
+      "loss": 0.8158,
+      "step": 5579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.781358037455213,
+      "learning_rate": 1.6646990434270625e-05,
+      "loss": 0.686,
+      "step": 5580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.481525669403204,
+      "learning_rate": 1.664573041215402e-05,
+      "loss": 0.7547,
+      "step": 5581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.18280275060542,
+      "learning_rate": 1.6644470201036732e-05,
+      "loss": 0.7663,
+      "step": 5582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.40391433849428,
+      "learning_rate": 1.6643209800954597e-05,
+      "loss": 0.7589,
+      "step": 5583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.155087411586884,
+      "learning_rate": 1.664194921194346e-05,
+      "loss": 0.6845,
+      "step": 5584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.26437062954492,
+      "learning_rate": 1.664068843403918e-05,
+      "loss": 0.8006,
+      "step": 5585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.616893719083563,
+      "learning_rate": 1.6639427467277597e-05,
+      "loss": 0.7166,
+      "step": 5586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 28.234724825838313,
+      "learning_rate": 1.663816631169459e-05,
+      "loss": 0.7036,
+      "step": 5587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.511589595465793,
+      "learning_rate": 1.663690496732601e-05,
+      "loss": 0.7027,
+      "step": 5588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.159613335204206,
+      "learning_rate": 1.663564343420773e-05,
+      "loss": 0.7976,
+      "step": 5589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.108872972837347,
+      "learning_rate": 1.6634381712375643e-05,
+      "loss": 0.7092,
+      "step": 5590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.131143095134703,
+      "learning_rate": 1.6633119801865617e-05,
+      "loss": 0.7527,
+      "step": 5591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.755097063373377,
+      "learning_rate": 1.6631857702713544e-05,
+      "loss": 0.7212,
+      "step": 5592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.457837596779466,
+      "learning_rate": 1.6630595414955314e-05,
+      "loss": 0.6246,
+      "step": 5593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.411398728124018,
+      "learning_rate": 1.6629332938626836e-05,
+      "loss": 0.6982,
+      "step": 5594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.84759840415231,
+      "learning_rate": 1.6628070273764002e-05,
+      "loss": 0.7234,
+      "step": 5595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.660024459391256,
+      "learning_rate": 1.6626807420402727e-05,
+      "loss": 0.7209,
+      "step": 5596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.33829364136699,
+      "learning_rate": 1.662554437857893e-05,
+      "loss": 0.8041,
+      "step": 5597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.02438790774433,
+      "learning_rate": 1.662428114832852e-05,
+      "loss": 0.7279,
+      "step": 5598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.76716925463418,
+      "learning_rate": 1.6623017729687432e-05,
+      "loss": 0.7577,
+      "step": 5599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.27211329754222,
+      "learning_rate": 1.6621754122691596e-05,
+      "loss": 0.6983,
+      "step": 5600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.032776306599608,
+      "learning_rate": 1.6620490327376947e-05,
+      "loss": 0.721,
+      "step": 5601
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.787681220196095,
+      "learning_rate": 1.6619226343779424e-05,
+      "loss": 0.6814,
+      "step": 5602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.348860665079567,
+      "learning_rate": 1.6617962171934975e-05,
+      "loss": 0.7318,
+      "step": 5603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.521396659286662,
+      "learning_rate": 1.6616697811879553e-05,
+      "loss": 0.7363,
+      "step": 5604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.860291335163293,
+      "learning_rate": 1.661543326364912e-05,
+      "loss": 0.7472,
+      "step": 5605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.75152451747405,
+      "learning_rate": 1.661416852727963e-05,
+      "loss": 0.7367,
+      "step": 5606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.039202729775827,
+      "learning_rate": 1.661290360280706e-05,
+      "loss": 0.674,
+      "step": 5607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.370202808262032,
+      "learning_rate": 1.6611638490267375e-05,
+      "loss": 0.6773,
+      "step": 5608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.7575965664809,
+      "learning_rate": 1.6610373189696565e-05,
+      "loss": 0.7473,
+      "step": 5609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.7891086912162,
+      "learning_rate": 1.6609107701130603e-05,
+      "loss": 0.6954,
+      "step": 5610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.139049392867005,
+      "learning_rate": 1.660784202460549e-05,
+      "loss": 0.7718,
+      "step": 5611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.288890370907275,
+      "learning_rate": 1.6606576160157212e-05,
+      "loss": 0.7744,
+      "step": 5612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 24.609885958256594,
+      "learning_rate": 1.6605310107821774e-05,
+      "loss": 0.7104,
+      "step": 5613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.668146188800378,
+      "learning_rate": 1.660404386763518e-05,
+      "loss": 0.7293,
+      "step": 5614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.494055250018512,
+      "learning_rate": 1.660277743963344e-05,
+      "loss": 0.6627,
+      "step": 5615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.059095645956646,
+      "learning_rate": 1.6601510823852574e-05,
+      "loss": 0.7107,
+      "step": 5616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.99842022057446,
+      "learning_rate": 1.66002440203286e-05,
+      "loss": 0.7479,
+      "step": 5617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.643079812755305,
+      "learning_rate": 1.659897702909755e-05,
+      "loss": 0.8548,
+      "step": 5618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 25.324072378295437,
+      "learning_rate": 1.659770985019545e-05,
+      "loss": 0.7261,
+      "step": 5619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.60092128892326,
+      "learning_rate": 1.6596442483658345e-05,
+      "loss": 0.792,
+      "step": 5620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.09456805373649,
+      "learning_rate": 1.6595174929522273e-05,
+      "loss": 0.7742,
+      "step": 5621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.991880673151805,
+      "learning_rate": 1.6593907187823284e-05,
+      "loss": 0.7054,
+      "step": 5622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.437698529741215,
+      "learning_rate": 1.6592639258597432e-05,
+      "loss": 0.7062,
+      "step": 5623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.745114927499325,
+      "learning_rate": 1.6591371141880776e-05,
+      "loss": 0.7459,
+      "step": 5624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.622390470927279,
+      "learning_rate": 1.6590102837709382e-05,
+      "loss": 0.7225,
+      "step": 5625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.870437478685915,
+      "learning_rate": 1.6588834346119315e-05,
+      "loss": 0.7657,
+      "step": 5626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.268626183483757,
+      "learning_rate": 1.6587565667146657e-05,
+      "loss": 0.887,
+      "step": 5627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.831310196369493,
+      "learning_rate": 1.6586296800827483e-05,
+      "loss": 0.7503,
+      "step": 5628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.596742024491475,
+      "learning_rate": 1.6585027747197882e-05,
+      "loss": 0.76,
+      "step": 5629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.465565701814533,
+      "learning_rate": 1.6583758506293945e-05,
+      "loss": 0.6691,
+      "step": 5630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.34487366382551,
+      "learning_rate": 1.6582489078151765e-05,
+      "loss": 0.7564,
+      "step": 5631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.745352710718397,
+      "learning_rate": 1.6581219462807444e-05,
+      "loss": 0.6095,
+      "step": 5632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.98266594176781,
+      "learning_rate": 1.6579949660297098e-05,
+      "loss": 0.8123,
+      "step": 5633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.608125391575303,
+      "learning_rate": 1.657867967065683e-05,
+      "loss": 0.6883,
+      "step": 5634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.35142454758878,
+      "learning_rate": 1.657740949392276e-05,
+      "loss": 0.7156,
+      "step": 5635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.73731976555211,
+      "learning_rate": 1.657613913013101e-05,
+      "loss": 0.7523,
+      "step": 5636
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.179669520627378,
+      "learning_rate": 1.657486857931771e-05,
+      "loss": 0.72,
+      "step": 5637
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.431933500835687,
+      "learning_rate": 1.6573597841518995e-05,
+      "loss": 0.6616,
+      "step": 5638
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.440175977244216,
+      "learning_rate": 1.6572326916771008e-05,
+      "loss": 0.7945,
+      "step": 5639
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.693828836488256,
+      "learning_rate": 1.657105580510988e-05,
+      "loss": 0.7288,
+      "step": 5640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.244733091183958,
+      "learning_rate": 1.6569784506571772e-05,
+      "loss": 0.7198,
+      "step": 5641
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.140556573491953,
+      "learning_rate": 1.6568513021192837e-05,
+      "loss": 0.7317,
+      "step": 5642
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.960201168688616,
+      "learning_rate": 1.6567241349009235e-05,
+      "loss": 0.6911,
+      "step": 5643
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.151401328821926,
+      "learning_rate": 1.6565969490057128e-05,
+      "loss": 0.6568,
+      "step": 5644
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 27.846305203315232,
+      "learning_rate": 1.656469744437269e-05,
+      "loss": 0.7271,
+      "step": 5645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.376494627860403,
+      "learning_rate": 1.65634252119921e-05,
+      "loss": 0.6899,
+      "step": 5646
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.58101857633776,
+      "learning_rate": 1.6562152792951534e-05,
+      "loss": 0.8136,
+      "step": 5647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.968154964645576,
+      "learning_rate": 1.656088018728718e-05,
+      "loss": 0.6698,
+      "step": 5648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.635239206441062,
+      "learning_rate": 1.6559607395035234e-05,
+      "loss": 0.6785,
+      "step": 5649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.85055862458894,
+      "learning_rate": 1.655833441623189e-05,
+      "loss": 0.7595,
+      "step": 5650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.452070321654034,
+      "learning_rate": 1.6557061250913352e-05,
+      "loss": 0.7528,
+      "step": 5651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.448101144352076,
+      "learning_rate": 1.6555787899115827e-05,
+      "loss": 0.7606,
+      "step": 5652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.298406981692136,
+      "learning_rate": 1.6554514360875528e-05,
+      "loss": 0.7146,
+      "step": 5653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.357977843134766,
+      "learning_rate": 1.655324063622868e-05,
+      "loss": 0.6706,
+      "step": 5654
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.054341309148466,
+      "learning_rate": 1.6551966725211497e-05,
+      "loss": 0.7434,
+      "step": 5655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.89844506270747,
+      "learning_rate": 1.6550692627860213e-05,
+      "loss": 0.6801,
+      "step": 5656
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.865825831171021,
+      "learning_rate": 1.6549418344211066e-05,
+      "loss": 0.679,
+      "step": 5657
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.409735980795922,
+      "learning_rate": 1.6548143874300292e-05,
+      "loss": 0.6125,
+      "step": 5658
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 24.94414499133065,
+      "learning_rate": 1.654686921816413e-05,
+      "loss": 0.6911,
+      "step": 5659
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.445974171891383,
+      "learning_rate": 1.6545594375838846e-05,
+      "loss": 0.6618,
+      "step": 5660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.57349780480745,
+      "learning_rate": 1.6544319347360685e-05,
+      "loss": 0.8199,
+      "step": 5661
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 27.230779203626952,
+      "learning_rate": 1.6543044132765907e-05,
+      "loss": 0.6828,
+      "step": 5662
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.855952965347512,
+      "learning_rate": 1.6541768732090784e-05,
+      "loss": 0.7227,
+      "step": 5663
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.783346947768226,
+      "learning_rate": 1.6540493145371582e-05,
+      "loss": 0.6645,
+      "step": 5664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.76466457619321,
+      "learning_rate": 1.6539217372644585e-05,
+      "loss": 0.6301,
+      "step": 5665
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.164926600786643,
+      "learning_rate": 1.653794141394607e-05,
+      "loss": 0.7843,
+      "step": 5666
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.844293454140125,
+      "learning_rate": 1.6536665269312324e-05,
+      "loss": 0.6663,
+      "step": 5667
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.45526192123226,
+      "learning_rate": 1.6535388938779644e-05,
+      "loss": 0.7432,
+      "step": 5668
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.74445819496282,
+      "learning_rate": 1.6534112422384324e-05,
+      "loss": 0.7317,
+      "step": 5669
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.787744545486513,
+      "learning_rate": 1.653283572016267e-05,
+      "loss": 0.6547,
+      "step": 5670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.759511184284943,
+      "learning_rate": 1.653155883215099e-05,
+      "loss": 0.7599,
+      "step": 5671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.623840955587834,
+      "learning_rate": 1.6530281758385597e-05,
+      "loss": 0.7579,
+      "step": 5672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.249405641458004,
+      "learning_rate": 1.652900449890281e-05,
+      "loss": 0.7172,
+      "step": 5673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.338067834282093,
+      "learning_rate": 1.6527727053738957e-05,
+      "loss": 0.6848,
+      "step": 5674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.435249072962097,
+      "learning_rate": 1.6526449422930363e-05,
+      "loss": 0.6931,
+      "step": 5675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.02453084021368,
+      "learning_rate": 1.6525171606513368e-05,
+      "loss": 0.7353,
+      "step": 5676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.410508837247384,
+      "learning_rate": 1.6523893604524304e-05,
+      "loss": 0.6824,
+      "step": 5677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.91434361640482,
+      "learning_rate": 1.652261541699953e-05,
+      "loss": 0.6474,
+      "step": 5678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.18647938135595,
+      "learning_rate": 1.6521337043975384e-05,
+      "loss": 0.6542,
+      "step": 5679
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.790304567604757,
+      "learning_rate": 1.652005848548823e-05,
+      "loss": 0.7275,
+      "step": 5680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.48159908862092,
+      "learning_rate": 1.6518779741574422e-05,
+      "loss": 0.7223,
+      "step": 5681
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.67648162039196,
+      "learning_rate": 1.6517500812270335e-05,
+      "loss": 0.7433,
+      "step": 5682
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.224208182662455,
+      "learning_rate": 1.6516221697612338e-05,
+      "loss": 0.6883,
+      "step": 5683
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 28.35615649915233,
+      "learning_rate": 1.6514942397636807e-05,
+      "loss": 0.8017,
+      "step": 5684
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.18832427651014,
+      "learning_rate": 1.6513662912380123e-05,
+      "loss": 0.793,
+      "step": 5685
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.798340100303275,
+      "learning_rate": 1.651238324187868e-05,
+      "loss": 0.6869,
+      "step": 5686
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 23.429544259524203,
+      "learning_rate": 1.6511103386168867e-05,
+      "loss": 0.76,
+      "step": 5687
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.18241841323595,
+      "learning_rate": 1.650982334528708e-05,
+      "loss": 0.6465,
+      "step": 5688
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.43709596600368,
+      "learning_rate": 1.6508543119269727e-05,
+      "loss": 0.7059,
+      "step": 5689
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.585352701860142,
+      "learning_rate": 1.6507262708153215e-05,
+      "loss": 0.7288,
+      "step": 5690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.692208607133892,
+      "learning_rate": 1.6505982111973956e-05,
+      "loss": 0.7269,
+      "step": 5691
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.412397001311842,
+      "learning_rate": 1.6504701330768372e-05,
+      "loss": 0.6766,
+      "step": 5692
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.154341760434438,
+      "learning_rate": 1.650342036457289e-05,
+      "loss": 0.7598,
+      "step": 5693
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 24.8203046995012,
+      "learning_rate": 1.6502139213423936e-05,
+      "loss": 0.6902,
+      "step": 5694
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.61772130210199,
+      "learning_rate": 1.650085787735795e-05,
+      "loss": 0.7521,
+      "step": 5695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.034694769455974,
+      "learning_rate": 1.649957635641136e-05,
+      "loss": 0.7226,
+      "step": 5696
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.32790611567556,
+      "learning_rate": 1.6498294650620628e-05,
+      "loss": 0.6905,
+      "step": 5697
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.756171411085806,
+      "learning_rate": 1.6497012760022197e-05,
+      "loss": 0.6105,
+      "step": 5698
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.9530410349578,
+      "learning_rate": 1.649573068465252e-05,
+      "loss": 0.8234,
+      "step": 5699
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.5295224375633,
+      "learning_rate": 1.6494448424548065e-05,
+      "loss": 0.7711,
+      "step": 5700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.882921491167963,
+      "learning_rate": 1.6493165979745294e-05,
+      "loss": 0.7079,
+      "step": 5701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.860371266449944,
+      "learning_rate": 1.649188335028068e-05,
+      "loss": 0.7249,
+      "step": 5702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.544225030604608,
+      "learning_rate": 1.6490600536190705e-05,
+      "loss": 0.8041,
+      "step": 5703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.9381631775993,
+      "learning_rate": 1.6489317537511846e-05,
+      "loss": 0.6826,
+      "step": 5704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.407314707999674,
+      "learning_rate": 1.6488034354280592e-05,
+      "loss": 0.7103,
+      "step": 5705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.182579225513994,
+      "learning_rate": 1.648675098653344e-05,
+      "loss": 0.762,
+      "step": 5706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.006120731722394,
+      "learning_rate": 1.648546743430688e-05,
+      "loss": 0.7462,
+      "step": 5707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.798014200121603,
+      "learning_rate": 1.648418369763742e-05,
+      "loss": 0.7125,
+      "step": 5708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.193339055305193,
+      "learning_rate": 1.648289977656157e-05,
+      "loss": 0.6597,
+      "step": 5709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.733403721864674,
+      "learning_rate": 1.6481615671115845e-05,
+      "loss": 0.7473,
+      "step": 5710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.467838994361564,
+      "learning_rate": 1.6480331381336757e-05,
+      "loss": 0.6704,
+      "step": 5711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.564381874933485,
+      "learning_rate": 1.647904690726084e-05,
+      "loss": 0.661,
+      "step": 5712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.289721502515405,
+      "learning_rate": 1.6477762248924616e-05,
+      "loss": 0.6799,
+      "step": 5713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.554585598715992,
+      "learning_rate": 1.6476477406364623e-05,
+      "loss": 0.7488,
+      "step": 5714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.731416276488957,
+      "learning_rate": 1.6475192379617405e-05,
+      "loss": 0.7656,
+      "step": 5715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.411782964186976,
+      "learning_rate": 1.64739071687195e-05,
+      "loss": 0.7907,
+      "step": 5716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.706146514036433,
+      "learning_rate": 1.6472621773707463e-05,
+      "loss": 0.7054,
+      "step": 5717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.554123431838466,
+      "learning_rate": 1.647133619461785e-05,
+      "loss": 0.717,
+      "step": 5718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.95087189918306,
+      "learning_rate": 1.647005043148722e-05,
+      "loss": 0.762,
+      "step": 5719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.028863548489635,
+      "learning_rate": 1.646876448435214e-05,
+      "loss": 0.7717,
+      "step": 5720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.460620633641483,
+      "learning_rate": 1.6467478353249186e-05,
+      "loss": 0.7567,
+      "step": 5721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.78109727286138,
+      "learning_rate": 1.6466192038214928e-05,
+      "loss": 0.684,
+      "step": 5722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.78884535322842,
+      "learning_rate": 1.646490553928595e-05,
+      "loss": 0.6817,
+      "step": 5723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.732138404346333,
+      "learning_rate": 1.6463618856498844e-05,
+      "loss": 0.7988,
+      "step": 5724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 24.295008041801893,
+      "learning_rate": 1.64623319898902e-05,
+      "loss": 0.7215,
+      "step": 5725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.610719198450784,
+      "learning_rate": 1.6461044939496606e-05,
+      "loss": 0.6587,
+      "step": 5726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.817251600648486,
+      "learning_rate": 1.645975770535468e-05,
+      "loss": 0.6557,
+      "step": 5727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.32195410909108,
+      "learning_rate": 1.6458470287501024e-05,
+      "loss": 0.7095,
+      "step": 5728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.27110069371097,
+      "learning_rate": 1.6457182685972248e-05,
+      "loss": 0.6062,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 25.095901566257552,
+      "learning_rate": 1.6455894900804974e-05,
+      "loss": 0.6872,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.916391614298977,
+      "learning_rate": 1.6454606932035827e-05,
+      "loss": 0.7936,
+      "step": 5731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.674545463712919,
+      "learning_rate": 1.6453318779701434e-05,
+      "loss": 0.6512,
+      "step": 5732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.96112751890239,
+      "learning_rate": 1.645203044383843e-05,
+      "loss": 0.7633,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.18483719736039,
+      "learning_rate": 1.645074192448346e-05,
+      "loss": 0.7853,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.53176577382229,
+      "learning_rate": 1.6449453221673154e-05,
+      "loss": 0.6666,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.473599699417615,
+      "learning_rate": 1.6448164335444178e-05,
+      "loss": 0.6325,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.893537718634708,
+      "learning_rate": 1.6446875265833178e-05,
+      "loss": 0.6931,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.35430756545539,
+      "learning_rate": 1.6445586012876814e-05,
+      "loss": 0.7357,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.71951126857608,
+      "learning_rate": 1.644429657661176e-05,
+      "loss": 0.7224,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.413789780042524,
+      "learning_rate": 1.6443006957074675e-05,
+      "loss": 0.8399,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.443157903238202,
+      "learning_rate": 1.6441717154302246e-05,
+      "loss": 0.7918,
+      "step": 5741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.37276849941754,
+      "learning_rate": 1.644042716833115e-05,
+      "loss": 0.6584,
+      "step": 5742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 23.25333712134639,
+      "learning_rate": 1.6439136999198068e-05,
+      "loss": 0.7796,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 23.405334323599018,
+      "learning_rate": 1.64378466469397e-05,
+      "loss": 0.8338,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.330854093823742,
+      "learning_rate": 1.643655611159274e-05,
+      "loss": 0.7011,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.47278854667994,
+      "learning_rate": 1.6435265393193886e-05,
+      "loss": 0.6948,
+      "step": 5746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.217504019519719,
+      "learning_rate": 1.6433974491779853e-05,
+      "loss": 0.7489,
+      "step": 5747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 21.288441049447194,
+      "learning_rate": 1.6432683407387346e-05,
+      "loss": 0.7664,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.15909214309748,
+      "learning_rate": 1.6431392140053088e-05,
+      "loss": 0.7481,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.147548078110866,
+      "learning_rate": 1.6430100689813797e-05,
+      "loss": 0.8429,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.291706603134074,
+      "learning_rate": 1.642880905670621e-05,
+      "loss": 0.7116,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.039495611518436,
+      "learning_rate": 1.6427517240767046e-05,
+      "loss": 0.6976,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.935437611792857,
+      "learning_rate": 1.6426225242033055e-05,
+      "loss": 0.7023,
+      "step": 5753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.821594706584087,
+      "learning_rate": 1.6424933060540978e-05,
+      "loss": 0.7242,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.602660503572892,
+      "learning_rate": 1.6423640696327564e-05,
+      "loss": 0.7304,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.888132968633195,
+      "learning_rate": 1.6422348149429566e-05,
+      "loss": 0.694,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.49301044136683,
+      "learning_rate": 1.6421055419883744e-05,
+      "loss": 0.6861,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.656531414497566,
+      "learning_rate": 1.641976250772686e-05,
+      "loss": 0.7489,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 26.013145738750463,
+      "learning_rate": 1.6418469412995685e-05,
+      "loss": 0.7409,
+      "step": 5759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.819613326906847,
+      "learning_rate": 1.6417176135726998e-05,
+      "loss": 0.7258,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.179174931135993,
+      "learning_rate": 1.641588267595757e-05,
+      "loss": 0.6894,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.704633209495057,
+      "learning_rate": 1.6414589033724197e-05,
+      "loss": 0.673,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.870078511105422,
+      "learning_rate": 1.6413295209063663e-05,
+      "loss": 0.7022,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.8587057441996,
+      "learning_rate": 1.6412001202012768e-05,
+      "loss": 0.744,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.374960559996676,
+      "learning_rate": 1.6410707012608303e-05,
+      "loss": 0.6695,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.30811657895733,
+      "learning_rate": 1.640941264088709e-05,
+      "loss": 0.7162,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.24656049153652,
+      "learning_rate": 1.6408118086885925e-05,
+      "loss": 0.7177,
+      "step": 5767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.595495521942496,
+      "learning_rate": 1.6406823350641628e-05,
+      "loss": 0.7463,
+      "step": 5768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.459095175127864,
+      "learning_rate": 1.6405528432191027e-05,
+      "loss": 0.7578,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.044689857474776,
+      "learning_rate": 1.6404233331570944e-05,
+      "loss": 0.6874,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.205197236482149,
+      "learning_rate": 1.6402938048818208e-05,
+      "loss": 0.6507,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.028392994915404,
+      "learning_rate": 1.640164258396966e-05,
+      "loss": 0.7461,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 27.12965333100808,
+      "learning_rate": 1.6400346937062147e-05,
+      "loss": 0.7316,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.985086341372234,
+      "learning_rate": 1.6399051108132507e-05,
+      "loss": 0.6895,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.008121465288855,
+      "learning_rate": 1.6397755097217598e-05,
+      "loss": 0.7771,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 21.732444575985188,
+      "learning_rate": 1.6396458904354275e-05,
+      "loss": 0.7354,
+      "step": 5776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 25.80816470182595,
+      "learning_rate": 1.6395162529579406e-05,
+      "loss": 0.7311,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.955277632507821,
+      "learning_rate": 1.6393865972929853e-05,
+      "loss": 0.7192,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.729889325866502,
+      "learning_rate": 1.6392569234442492e-05,
+      "loss": 0.7409,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.554252143990125,
+      "learning_rate": 1.63912723141542e-05,
+      "loss": 0.6934,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.4984187532941,
+      "learning_rate": 1.6389975212101867e-05,
+      "loss": 0.7349,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.212569607342665,
+      "learning_rate": 1.6388677928322376e-05,
+      "loss": 0.6987,
+      "step": 5782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 23.265658918500762,
+      "learning_rate": 1.638738046285262e-05,
+      "loss": 0.6853,
+      "step": 5783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.651439859095806,
+      "learning_rate": 1.63860828157295e-05,
+      "loss": 0.7496,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.696158599730714,
+      "learning_rate": 1.6384784986989918e-05,
+      "loss": 0.7441,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.1359620780321,
+      "learning_rate": 1.638348697667079e-05,
+      "loss": 0.7198,
+      "step": 5786
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.99321591202826,
+      "learning_rate": 1.6382188784809027e-05,
+      "loss": 0.7325,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.02321955580521,
+      "learning_rate": 1.638089041144155e-05,
+      "loss": 0.7557,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.966457452291284,
+      "learning_rate": 1.6379591856605274e-05,
+      "loss": 0.7063,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.163418201252806,
+      "learning_rate": 1.6378293120337144e-05,
+      "loss": 0.7001,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.953390495273855,
+      "learning_rate": 1.637699420267409e-05,
+      "loss": 0.7787,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.369755507357898,
+      "learning_rate": 1.6375695103653047e-05,
+      "loss": 0.7026,
+      "step": 5792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.90137034189558,
+      "learning_rate": 1.637439582331097e-05,
+      "loss": 0.5927,
+      "step": 5793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.616159394648477,
+      "learning_rate": 1.63730963616848e-05,
+      "loss": 0.6738,
+      "step": 5794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.90839275544625,
+      "learning_rate": 1.6371796718811496e-05,
+      "loss": 0.6792,
+      "step": 5795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 25.755384194776852,
+      "learning_rate": 1.637049689472803e-05,
+      "loss": 0.8311,
+      "step": 5796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 26.98814318289358,
+      "learning_rate": 1.636919688947135e-05,
+      "loss": 0.8479,
+      "step": 5797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 28.09579774842622,
+      "learning_rate": 1.636789670307844e-05,
+      "loss": 0.7434,
+      "step": 5798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.489971238392123,
+      "learning_rate": 1.636659633558627e-05,
+      "loss": 0.7615,
+      "step": 5799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.174140411815452,
+      "learning_rate": 1.6365295787031827e-05,
+      "loss": 0.6884,
+      "step": 5800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.195328238283686,
+      "learning_rate": 1.6363995057452098e-05,
+      "loss": 0.7759,
+      "step": 5801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 22.505019659154755,
+      "learning_rate": 1.6362694146884067e-05,
+      "loss": 0.7392,
+      "step": 5802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.42531078884606,
+      "learning_rate": 1.636139305536474e-05,
+      "loss": 0.6837,
+      "step": 5803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.879879409041795,
+      "learning_rate": 1.6360091782931115e-05,
+      "loss": 0.7237,
+      "step": 5804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.782277755517892,
+      "learning_rate": 1.63587903296202e-05,
+      "loss": 0.6456,
+      "step": 5805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.95610790913143,
+      "learning_rate": 1.635748869546901e-05,
+      "loss": 0.6609,
+      "step": 5806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.08536098097721,
+      "learning_rate": 1.6356186880514556e-05,
+      "loss": 0.6788,
+      "step": 5807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.278555583952873,
+      "learning_rate": 1.635488488479387e-05,
+      "loss": 0.7367,
+      "step": 5808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 28.759383400674686,
+      "learning_rate": 1.635358270834397e-05,
+      "loss": 0.6626,
+      "step": 5809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.078319271053218,
+      "learning_rate": 1.6352280351201898e-05,
+      "loss": 0.7372,
+      "step": 5810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.144959247533215,
+      "learning_rate": 1.635097781340469e-05,
+      "loss": 0.6672,
+      "step": 5811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.14920839463731,
+      "learning_rate": 1.6349675094989387e-05,
+      "loss": 0.7133,
+      "step": 5812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.656466561101208,
+      "learning_rate": 1.6348372195993037e-05,
+      "loss": 0.7127,
+      "step": 5813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.666831581568328,
+      "learning_rate": 1.63470691164527e-05,
+      "loss": 0.712,
+      "step": 5814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.696959494882545,
+      "learning_rate": 1.6345765856405424e-05,
+      "loss": 0.6774,
+      "step": 5815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.490884041950114,
+      "learning_rate": 1.6344462415888283e-05,
+      "loss": 0.7901,
+      "step": 5816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.732721353459777,
+      "learning_rate": 1.6343158794938342e-05,
+      "loss": 0.6554,
+      "step": 5817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 25.560658768655045,
+      "learning_rate": 1.6341854993592674e-05,
+      "loss": 0.6604,
+      "step": 5818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 24.328422611942173,
+      "learning_rate": 1.634055101188836e-05,
+      "loss": 0.7234,
+      "step": 5819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.392988013327987,
+      "learning_rate": 1.6339246849862488e-05,
+      "loss": 0.8028,
+      "step": 5820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.42406229215249,
+      "learning_rate": 1.633794250755214e-05,
+      "loss": 0.7765,
+      "step": 5821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.514115378247727,
+      "learning_rate": 1.6336637984994418e-05,
+      "loss": 0.7536,
+      "step": 5822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.812951527727318,
+      "learning_rate": 1.6335333282226414e-05,
+      "loss": 0.6994,
+      "step": 5823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.910757446580629,
+      "learning_rate": 1.6334028399285243e-05,
+      "loss": 0.6646,
+      "step": 5824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.882888151803336,
+      "learning_rate": 1.6332723336208007e-05,
+      "loss": 0.7778,
+      "step": 5825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.67503314209466,
+      "learning_rate": 1.633141809303182e-05,
+      "loss": 0.7605,
+      "step": 5826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.940135701907863,
+      "learning_rate": 1.633011266979381e-05,
+      "loss": 0.6965,
+      "step": 5827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.114629175374471,
+      "learning_rate": 1.6328807066531102e-05,
+      "loss": 0.7662,
+      "step": 5828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.789468579230917,
+      "learning_rate": 1.6327501283280816e-05,
+      "loss": 0.7287,
+      "step": 5829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 24.233231541498746,
+      "learning_rate": 1.6326195320080103e-05,
+      "loss": 0.6999,
+      "step": 5830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 29.19205879100767,
+      "learning_rate": 1.632488917696609e-05,
+      "loss": 0.6944,
+      "step": 5831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.753578174610723,
+      "learning_rate": 1.632358285397593e-05,
+      "loss": 0.7462,
+      "step": 5832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.644718473915397,
+      "learning_rate": 1.6322276351146774e-05,
+      "loss": 0.7603,
+      "step": 5833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.903770592158505,
+      "learning_rate": 1.6320969668515773e-05,
+      "loss": 0.8326,
+      "step": 5834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 22.879420381005456,
+      "learning_rate": 1.63196628061201e-05,
+      "loss": 0.7076,
+      "step": 5835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.43826235528824,
+      "learning_rate": 1.6318355763996908e-05,
+      "loss": 0.7249,
+      "step": 5836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.35383209283138,
+      "learning_rate": 1.6317048542183375e-05,
+      "loss": 0.7659,
+      "step": 5837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.356027427162466,
+      "learning_rate": 1.6315741140716675e-05,
+      "loss": 0.7482,
+      "step": 5838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.405609994848234,
+      "learning_rate": 1.6314433559633992e-05,
+      "loss": 0.7291,
+      "step": 5839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.420958412422095,
+      "learning_rate": 1.6313125798972514e-05,
+      "loss": 0.7589,
+      "step": 5840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.491793530351103,
+      "learning_rate": 1.6311817858769434e-05,
+      "loss": 0.6358,
+      "step": 5841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.172076856540357,
+      "learning_rate": 1.6310509739061946e-05,
+      "loss": 0.7548,
+      "step": 5842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.173773223292894,
+      "learning_rate": 1.6309201439887248e-05,
+      "loss": 0.7527,
+      "step": 5843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.80129068219147,
+      "learning_rate": 1.6307892961282555e-05,
+      "loss": 0.6585,
+      "step": 5844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.375898180116764,
+      "learning_rate": 1.6306584303285077e-05,
+      "loss": 0.6452,
+      "step": 5845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.877060956532212,
+      "learning_rate": 1.6305275465932027e-05,
+      "loss": 0.833,
+      "step": 5846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.77071229112161,
+      "learning_rate": 1.6303966449260636e-05,
+      "loss": 0.6422,
+      "step": 5847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 25.207241713057428,
+      "learning_rate": 1.630265725330812e-05,
+      "loss": 0.6187,
+      "step": 5848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.220453171151078,
+      "learning_rate": 1.6301347878111726e-05,
+      "loss": 0.8335,
+      "step": 5849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.177799544773135,
+      "learning_rate": 1.630003832370868e-05,
+      "loss": 0.6804,
+      "step": 5850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.285428087688317,
+      "learning_rate": 1.629872859013623e-05,
+      "loss": 0.7265,
+      "step": 5851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.01305889244609,
+      "learning_rate": 1.6297418677431625e-05,
+      "loss": 0.6638,
+      "step": 5852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 34.01237027495073,
+      "learning_rate": 1.629610858563212e-05,
+      "loss": 0.6628,
+      "step": 5853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.18372891680082,
+      "learning_rate": 1.6294798314774963e-05,
+      "loss": 0.7069,
+      "step": 5854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.38925319017957,
+      "learning_rate": 1.6293487864897425e-05,
+      "loss": 0.748,
+      "step": 5855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.62885086421975,
+      "learning_rate": 1.6292177236036776e-05,
+      "loss": 0.6876,
+      "step": 5856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.8710086442742,
+      "learning_rate": 1.629086642823029e-05,
+      "loss": 0.6719,
+      "step": 5857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 31.722702970495018,
+      "learning_rate": 1.628955544151524e-05,
+      "loss": 0.7324,
+      "step": 5858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 29.46370384451972,
+      "learning_rate": 1.6288244275928912e-05,
+      "loss": 0.7323,
+      "step": 5859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.859204978125025,
+      "learning_rate": 1.6286932931508596e-05,
+      "loss": 0.7308,
+      "step": 5860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.1169048200672,
+      "learning_rate": 1.6285621408291584e-05,
+      "loss": 0.7692,
+      "step": 5861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.490830373486926,
+      "learning_rate": 1.6284309706315178e-05,
+      "loss": 0.7694,
+      "step": 5862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.378687726053442,
+      "learning_rate": 1.6282997825616676e-05,
+      "loss": 0.7079,
+      "step": 5863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.007806749159085,
+      "learning_rate": 1.62816857662334e-05,
+      "loss": 0.7497,
+      "step": 5864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.422473914666208,
+      "learning_rate": 1.6280373528202648e-05,
+      "loss": 0.7115,
+      "step": 5865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.863384409275017,
+      "learning_rate": 1.627906111156175e-05,
+      "loss": 0.7473,
+      "step": 5866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.712622773388253,
+      "learning_rate": 1.6277748516348025e-05,
+      "loss": 0.638,
+      "step": 5867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.5264682178003,
+      "learning_rate": 1.6276435742598807e-05,
+      "loss": 0.6744,
+      "step": 5868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.79368025227889,
+      "learning_rate": 1.6275122790351426e-05,
+      "loss": 0.6629,
+      "step": 5869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.75003657096813,
+      "learning_rate": 1.6273809659643226e-05,
+      "loss": 0.6624,
+      "step": 5870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.537043094562865,
+      "learning_rate": 1.6272496350511547e-05,
+      "loss": 0.7322,
+      "step": 5871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.29564125696257,
+      "learning_rate": 1.6271182862993743e-05,
+      "loss": 0.6634,
+      "step": 5872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.113132191854387,
+      "learning_rate": 1.6269869197127162e-05,
+      "loss": 0.8624,
+      "step": 5873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.490609787042935,
+      "learning_rate": 1.6268555352949175e-05,
+      "loss": 0.6903,
+      "step": 5874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.2939526951468,
+      "learning_rate": 1.626724133049714e-05,
+      "loss": 0.6174,
+      "step": 5875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.409671632112559,
+      "learning_rate": 1.6265927129808426e-05,
+      "loss": 0.627,
+      "step": 5876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 27.136673234973063,
+      "learning_rate": 1.6264612750920406e-05,
+      "loss": 0.7712,
+      "step": 5877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.666245259950959,
+      "learning_rate": 1.6263298193870468e-05,
+      "loss": 0.6477,
+      "step": 5878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.55641363525317,
+      "learning_rate": 1.6261983458695996e-05,
+      "loss": 0.7585,
+      "step": 5879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 25.48309068437332,
+      "learning_rate": 1.6260668545434375e-05,
+      "loss": 0.7301,
+      "step": 5880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 27.784376618553992,
+      "learning_rate": 1.6259353454123003e-05,
+      "loss": 0.7419,
+      "step": 5881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.801995266137457,
+      "learning_rate": 1.625803818479928e-05,
+      "loss": 0.6598,
+      "step": 5882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.573069531722815,
+      "learning_rate": 1.6256722737500613e-05,
+      "loss": 0.7259,
+      "step": 5883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.47241078321202,
+      "learning_rate": 1.625540711226441e-05,
+      "loss": 0.7121,
+      "step": 5884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.727059569719156,
+      "learning_rate": 1.625409130912809e-05,
+      "loss": 0.7054,
+      "step": 5885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 22.850694802267544,
+      "learning_rate": 1.6252775328129067e-05,
+      "loss": 0.82,
+      "step": 5886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.648720053714523,
+      "learning_rate": 1.6251459169304776e-05,
+      "loss": 0.6951,
+      "step": 5887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.58796360555267,
+      "learning_rate": 1.6250142832692643e-05,
+      "loss": 0.7245,
+      "step": 5888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 28.586115250369748,
+      "learning_rate": 1.6248826318330103e-05,
+      "loss": 0.8198,
+      "step": 5889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.75473040052826,
+      "learning_rate": 1.6247509626254597e-05,
+      "loss": 0.767,
+      "step": 5890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.847430644488675,
+      "learning_rate": 1.6246192756503572e-05,
+      "loss": 0.6209,
+      "step": 5891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 24.909309793357508,
+      "learning_rate": 1.624487570911448e-05,
+      "loss": 0.7406,
+      "step": 5892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.864794776139332,
+      "learning_rate": 1.6243558484124778e-05,
+      "loss": 0.7169,
+      "step": 5893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.195730899481068,
+      "learning_rate": 1.6242241081571923e-05,
+      "loss": 0.6453,
+      "step": 5894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.801300941874892,
+      "learning_rate": 1.6240923501493383e-05,
+      "loss": 0.7624,
+      "step": 5895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.3320670730396,
+      "learning_rate": 1.6239605743926632e-05,
+      "loss": 0.8266,
+      "step": 5896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.097127002272018,
+      "learning_rate": 1.6238287808909137e-05,
+      "loss": 0.7413,
+      "step": 5897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.67196590260174,
+      "learning_rate": 1.6236969696478393e-05,
+      "loss": 0.6988,
+      "step": 5898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.623937590361077,
+      "learning_rate": 1.6235651406671875e-05,
+      "loss": 0.6864,
+      "step": 5899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.586423718896402,
+      "learning_rate": 1.623433293952708e-05,
+      "loss": 0.6383,
+      "step": 5900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 27.757080110335508,
+      "learning_rate": 1.6233014295081505e-05,
+      "loss": 0.6652,
+      "step": 5901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.259816567253296,
+      "learning_rate": 1.623169547337265e-05,
+      "loss": 0.6892,
+      "step": 5902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 25.2420301400005,
+      "learning_rate": 1.6230376474438018e-05,
+      "loss": 0.6876,
+      "step": 5903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.592569898485612,
+      "learning_rate": 1.6229057298315123e-05,
+      "loss": 0.7102,
+      "step": 5904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.735847455884493,
+      "learning_rate": 1.6227737945041485e-05,
+      "loss": 0.5997,
+      "step": 5905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.085673896160745,
+      "learning_rate": 1.6226418414654625e-05,
+      "loss": 0.7951,
+      "step": 5906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.890752029777746,
+      "learning_rate": 1.6225098707192063e-05,
+      "loss": 0.7078,
+      "step": 5907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 27.410334507826146,
+      "learning_rate": 1.622377882269134e-05,
+      "loss": 0.7008,
+      "step": 5908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 22.162051671703466,
+      "learning_rate": 1.6222458761189984e-05,
+      "loss": 0.7213,
+      "step": 5909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.494290652244455,
+      "learning_rate": 1.622113852272554e-05,
+      "loss": 0.6921,
+      "step": 5910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 24.428519991445437,
+      "learning_rate": 1.621981810733556e-05,
+      "loss": 0.7837,
+      "step": 5911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.87060602279722,
+      "learning_rate": 1.6218497515057593e-05,
+      "loss": 0.7829,
+      "step": 5912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.086965935111138,
+      "learning_rate": 1.621717674592919e-05,
+      "loss": 0.7273,
+      "step": 5913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.247245796677582,
+      "learning_rate": 1.621585579998792e-05,
+      "loss": 0.7406,
+      "step": 5914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.437697422186208,
+      "learning_rate": 1.6214534677271344e-05,
+      "loss": 0.6834,
+      "step": 5915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.005778743206983,
+      "learning_rate": 1.6213213377817035e-05,
+      "loss": 0.6712,
+      "step": 5916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.28936195148279,
+      "learning_rate": 1.6211891901662576e-05,
+      "loss": 0.6866,
+      "step": 5917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.603544973130766,
+      "learning_rate": 1.621057024884555e-05,
+      "loss": 0.7248,
+      "step": 5918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.067000101702876,
+      "learning_rate": 1.620924841940353e-05,
+      "loss": 0.6844,
+      "step": 5919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.65909563224819,
+      "learning_rate": 1.6207926413374124e-05,
+      "loss": 0.7633,
+      "step": 5920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.061215646648854,
+      "learning_rate": 1.620660423079492e-05,
+      "loss": 0.7529,
+      "step": 5921
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.69768886484089,
+      "learning_rate": 1.620528187170352e-05,
+      "loss": 0.6891,
+      "step": 5922
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.060081408626154,
+      "learning_rate": 1.6203959336137538e-05,
+      "loss": 0.6552,
+      "step": 5923
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.476902658896494,
+      "learning_rate": 1.620263662413458e-05,
+      "loss": 0.6676,
+      "step": 5924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.676627481719384,
+      "learning_rate": 1.6201313735732265e-05,
+      "loss": 0.6213,
+      "step": 5925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.727930664639576,
+      "learning_rate": 1.619999067096821e-05,
+      "loss": 0.6276,
+      "step": 5926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.159879646287866,
+      "learning_rate": 1.6198667429880054e-05,
+      "loss": 0.745,
+      "step": 5927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.112949457625067,
+      "learning_rate": 1.619734401250542e-05,
+      "loss": 0.7489,
+      "step": 5928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.354114584357836,
+      "learning_rate": 1.6196020418881947e-05,
+      "loss": 0.691,
+      "step": 5929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.80471084889467,
+      "learning_rate": 1.6194696649047278e-05,
+      "loss": 0.7834,
+      "step": 5930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.076001775085388,
+      "learning_rate": 1.619337270303906e-05,
+      "loss": 0.6526,
+      "step": 5931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.21434072043704,
+      "learning_rate": 1.6192048580894946e-05,
+      "loss": 0.7788,
+      "step": 5932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.87706993176504,
+      "learning_rate": 1.6190724282652594e-05,
+      "loss": 0.7708,
+      "step": 5933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.404084460435197,
+      "learning_rate": 1.6189399808349663e-05,
+      "loss": 0.6706,
+      "step": 5934
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.152339116005287,
+      "learning_rate": 1.6188075158023825e-05,
+      "loss": 0.7311,
+      "step": 5935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.15266799359661,
+      "learning_rate": 1.6186750331712744e-05,
+      "loss": 0.7385,
+      "step": 5936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.42989187876736,
+      "learning_rate": 1.6185425329454106e-05,
+      "loss": 0.7341,
+      "step": 5937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.289202157347265,
+      "learning_rate": 1.618410015128559e-05,
+      "loss": 0.7335,
+      "step": 5938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.004547217529506,
+      "learning_rate": 1.6182774797244882e-05,
+      "loss": 0.6932,
+      "step": 5939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 38.503908652520025,
+      "learning_rate": 1.6181449267369678e-05,
+      "loss": 0.8746,
+      "step": 5940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.426658899734946,
+      "learning_rate": 1.6180123561697672e-05,
+      "loss": 0.6522,
+      "step": 5941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.043240801405812,
+      "learning_rate": 1.6178797680266566e-05,
+      "loss": 0.6749,
+      "step": 5942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.961162540880018,
+      "learning_rate": 1.6177471623114068e-05,
+      "loss": 0.7554,
+      "step": 5943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.96801580950909,
+      "learning_rate": 1.6176145390277893e-05,
+      "loss": 0.7957,
+      "step": 5944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.16415807833147,
+      "learning_rate": 1.6174818981795756e-05,
+      "loss": 0.6562,
+      "step": 5945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.53753529643634,
+      "learning_rate": 1.6173492397705376e-05,
+      "loss": 0.7054,
+      "step": 5946
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.335069969031636,
+      "learning_rate": 1.6172165638044486e-05,
+      "loss": 0.6904,
+      "step": 5947
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.89637617765377,
+      "learning_rate": 1.6170838702850814e-05,
+      "loss": 0.7049,
+      "step": 5948
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 27.271486570438373,
+      "learning_rate": 1.61695115921621e-05,
+      "loss": 0.7481,
+      "step": 5949
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.494015256206872,
+      "learning_rate": 1.6168184306016092e-05,
+      "loss": 0.7544,
+      "step": 5950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.90461561884123,
+      "learning_rate": 1.6166856844450522e-05,
+      "loss": 0.6851,
+      "step": 5951
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.308063031837321,
+      "learning_rate": 1.6165529207503156e-05,
+      "loss": 0.7029,
+      "step": 5952
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.858244233132446,
+      "learning_rate": 1.616420139521174e-05,
+      "loss": 0.7325,
+      "step": 5953
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.418704637893903,
+      "learning_rate": 1.6162873407614047e-05,
+      "loss": 0.7511,
+      "step": 5954
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.27601236079115,
+      "learning_rate": 1.616154524474784e-05,
+      "loss": 0.7231,
+      "step": 5955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.736412367587192,
+      "learning_rate": 1.616021690665089e-05,
+      "loss": 0.7177,
+      "step": 5956
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.583090055979834,
+      "learning_rate": 1.615888839336097e-05,
+      "loss": 0.6823,
+      "step": 5957
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 30.736905367607875,
+      "learning_rate": 1.6157559704915873e-05,
+      "loss": 0.7765,
+      "step": 5958
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.72996953313437,
+      "learning_rate": 1.6156230841353376e-05,
+      "loss": 0.6867,
+      "step": 5959
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.616841350675035,
+      "learning_rate": 1.6154901802711277e-05,
+      "loss": 0.6481,
+      "step": 5960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.959188555290131,
+      "learning_rate": 1.615357258902737e-05,
+      "loss": 0.6785,
+      "step": 5961
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.279277267440737,
+      "learning_rate": 1.6152243200339462e-05,
+      "loss": 0.7434,
+      "step": 5962
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.872918304102438,
+      "learning_rate": 1.6150913636685356e-05,
+      "loss": 0.8089,
+      "step": 5963
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.24059477336781,
+      "learning_rate": 1.614958389810286e-05,
+      "loss": 0.6465,
+      "step": 5964
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.286999985837308,
+      "learning_rate": 1.6148253984629798e-05,
+      "loss": 0.7133,
+      "step": 5965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.91779360560661,
+      "learning_rate": 1.614692389630399e-05,
+      "loss": 0.7208,
+      "step": 5966
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.756621761095253,
+      "learning_rate": 1.6145593633163258e-05,
+      "loss": 0.7476,
+      "step": 5967
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.215467946248461,
+      "learning_rate": 1.614426319524544e-05,
+      "loss": 0.7045,
+      "step": 5968
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.00972725245103,
+      "learning_rate": 1.6142932582588375e-05,
+      "loss": 0.7449,
+      "step": 5969
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.352999536228005,
+      "learning_rate": 1.61416017952299e-05,
+      "loss": 0.7178,
+      "step": 5970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.142891198991848,
+      "learning_rate": 1.614027083320786e-05,
+      "loss": 0.6726,
+      "step": 5971
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.701998655521543,
+      "learning_rate": 1.613893969656011e-05,
+      "loss": 0.805,
+      "step": 5972
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.263784887695612,
+      "learning_rate": 1.6137608385324505e-05,
+      "loss": 0.6631,
+      "step": 5973
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.414910505893396,
+      "learning_rate": 1.613627689953891e-05,
+      "loss": 0.6877,
+      "step": 5974
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.738871387239564,
+      "learning_rate": 1.613494523924119e-05,
+      "loss": 0.6301,
+      "step": 5975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.638921087368857,
+      "learning_rate": 1.6133613404469216e-05,
+      "loss": 0.7647,
+      "step": 5976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.135337843386694,
+      "learning_rate": 1.6132281395260863e-05,
+      "loss": 0.6678,
+      "step": 5977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 27.808605120678386,
+      "learning_rate": 1.6130949211654017e-05,
+      "loss": 0.6802,
+      "step": 5978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 25.32361928761944,
+      "learning_rate": 1.612961685368656e-05,
+      "loss": 0.6662,
+      "step": 5979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.059270969583373,
+      "learning_rate": 1.6128284321396385e-05,
+      "loss": 0.6913,
+      "step": 5980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.265600688677793,
+      "learning_rate": 1.6126951614821388e-05,
+      "loss": 0.8176,
+      "step": 5981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.253895606148106,
+      "learning_rate": 1.6125618733999476e-05,
+      "loss": 0.7087,
+      "step": 5982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 26.790769623312986,
+      "learning_rate": 1.6124285678968546e-05,
+      "loss": 0.6605,
+      "step": 5983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.50153300660963,
+      "learning_rate": 1.6122952449766516e-05,
+      "loss": 0.7991,
+      "step": 5984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.855735052856446,
+      "learning_rate": 1.6121619046431297e-05,
+      "loss": 0.7114,
+      "step": 5985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 27.134727186948577,
+      "learning_rate": 1.612028546900082e-05,
+      "loss": 0.8299,
+      "step": 5986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 22.983470155317562,
+      "learning_rate": 1.6118951717513e-05,
+      "loss": 0.6333,
+      "step": 5987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.225499314885921,
+      "learning_rate": 1.611761779200577e-05,
+      "loss": 0.6733,
+      "step": 5988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.23512549465239,
+      "learning_rate": 1.6116283692517075e-05,
+      "loss": 0.7622,
+      "step": 5989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.93436108514862,
+      "learning_rate": 1.6114949419084846e-05,
+      "loss": 0.6438,
+      "step": 5990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 26.437615919648117,
+      "learning_rate": 1.6113614971747032e-05,
+      "loss": 0.7415,
+      "step": 5991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.42633262967853,
+      "learning_rate": 1.6112280350541584e-05,
+      "loss": 0.7304,
+      "step": 5992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.67108212467494,
+      "learning_rate": 1.6110945555506462e-05,
+      "loss": 0.69,
+      "step": 5993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.339503193389636,
+      "learning_rate": 1.6109610586679624e-05,
+      "loss": 0.7079,
+      "step": 5994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 29.556020902744823,
+      "learning_rate": 1.610827544409903e-05,
+      "loss": 0.7038,
+      "step": 5995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.64112554880865,
+      "learning_rate": 1.6106940127802658e-05,
+      "loss": 0.7173,
+      "step": 5996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.624462066208018,
+      "learning_rate": 1.610560463782848e-05,
+      "loss": 0.7088,
+      "step": 5997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 31.290247628273566,
+      "learning_rate": 1.610426897421448e-05,
+      "loss": 0.802,
+      "step": 5998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.606065118792602,
+      "learning_rate": 1.610293313699864e-05,
+      "loss": 0.6306,
+      "step": 5999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 29.70868308336053,
+      "learning_rate": 1.610159712621895e-05,
+      "loss": 0.6919,
+      "step": 6000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.370156368943082,
+      "learning_rate": 1.610026094191341e-05,
+      "loss": 0.7404,
+      "step": 6001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.185485222007488,
+      "learning_rate": 1.6098924584120017e-05,
+      "loss": 0.7808,
+      "step": 6002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.173276347237808,
+      "learning_rate": 1.6097588052876774e-05,
+      "loss": 0.7459,
+      "step": 6003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.47823266964914,
+      "learning_rate": 1.60962513482217e-05,
+      "loss": 0.7217,
+      "step": 6004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.99578884853381,
+      "learning_rate": 1.6094914470192796e-05,
+      "loss": 0.656,
+      "step": 6005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.3157444486773,
+      "learning_rate": 1.6093577418828094e-05,
+      "loss": 0.6596,
+      "step": 6006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 26.44876873752978,
+      "learning_rate": 1.6092240194165618e-05,
+      "loss": 0.7488,
+      "step": 6007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.546222867233915,
+      "learning_rate": 1.6090902796243386e-05,
+      "loss": 0.698,
+      "step": 6008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 22.527975887467075,
+      "learning_rate": 1.608956522509945e-05,
+      "loss": 0.6801,
+      "step": 6009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 35.15430544752382,
+      "learning_rate": 1.6088227480771835e-05,
+      "loss": 0.7016,
+      "step": 6010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.079673579247963,
+      "learning_rate": 1.6086889563298594e-05,
+      "loss": 0.6352,
+      "step": 6011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.300598659846223,
+      "learning_rate": 1.6085551472717775e-05,
+      "loss": 0.6901,
+      "step": 6012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.018386023503165,
+      "learning_rate": 1.6084213209067433e-05,
+      "loss": 0.7082,
+      "step": 6013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.904186541309263,
+      "learning_rate": 1.6082874772385628e-05,
+      "loss": 0.6874,
+      "step": 6014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.714295879602227,
+      "learning_rate": 1.6081536162710418e-05,
+      "loss": 0.6371,
+      "step": 6015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.127558041170627,
+      "learning_rate": 1.6080197380079883e-05,
+      "loss": 0.7249,
+      "step": 6016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.548216729359247,
+      "learning_rate": 1.6078858424532088e-05,
+      "loss": 0.6625,
+      "step": 6017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.252166740678206,
+      "learning_rate": 1.6077519296105112e-05,
+      "loss": 0.7389,
+      "step": 6018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.756865178148285,
+      "learning_rate": 1.6076179994837045e-05,
+      "loss": 0.7765,
+      "step": 6019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.465302958969573,
+      "learning_rate": 1.6074840520765976e-05,
+      "loss": 0.7671,
+      "step": 6020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.76529910418772,
+      "learning_rate": 1.607350087393e-05,
+      "loss": 0.7215,
+      "step": 6021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 24.15083210407022,
+      "learning_rate": 1.60721610543672e-05,
+      "loss": 0.8208,
+      "step": 6022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.675696418363845,
+      "learning_rate": 1.6070821062115704e-05,
+      "loss": 0.7002,
+      "step": 6023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.67845007038022,
+      "learning_rate": 1.60694808972136e-05,
+      "loss": 0.7252,
+      "step": 6024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.678972298276326,
+      "learning_rate": 1.606814055969901e-05,
+      "loss": 0.5876,
+      "step": 6025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.47609256700064,
+      "learning_rate": 1.606680004961006e-05,
+      "loss": 0.6981,
+      "step": 6026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.30427774542063,
+      "learning_rate": 1.6065459366984856e-05,
+      "loss": 0.6276,
+      "step": 6027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.338619783642514,
+      "learning_rate": 1.6064118511861543e-05,
+      "loss": 0.7362,
+      "step": 6028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.555603238376158,
+      "learning_rate": 1.6062777484278245e-05,
+      "loss": 0.6852,
+      "step": 6029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.551533864113665,
+      "learning_rate": 1.60614362842731e-05,
+      "loss": 0.7836,
+      "step": 6030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.280541871102116,
+      "learning_rate": 1.6060094911884255e-05,
+      "loss": 0.6696,
+      "step": 6031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.955626101863103,
+      "learning_rate": 1.6058753367149854e-05,
+      "loss": 0.7,
+      "step": 6032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.42183603685802,
+      "learning_rate": 1.6057411650108053e-05,
+      "loss": 0.7145,
+      "step": 6033
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.09152443537757,
+      "learning_rate": 1.6056069760797006e-05,
+      "loss": 0.7615,
+      "step": 6034
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.943154608542795,
+      "learning_rate": 1.605472769925488e-05,
+      "loss": 0.701,
+      "step": 6035
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.687209125321623,
+      "learning_rate": 1.6053385465519838e-05,
+      "loss": 0.6862,
+      "step": 6036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.024486516594113,
+      "learning_rate": 1.6052043059630058e-05,
+      "loss": 0.699,
+      "step": 6037
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.422543449254004,
+      "learning_rate": 1.6050700481623712e-05,
+      "loss": 0.6818,
+      "step": 6038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 24.572686157436372,
+      "learning_rate": 1.604935773153898e-05,
+      "loss": 0.7413,
+      "step": 6039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.045914884516726,
+      "learning_rate": 1.6048014809414057e-05,
+      "loss": 0.7308,
+      "step": 6040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.39324619124694,
+      "learning_rate": 1.604667171528713e-05,
+      "loss": 0.8003,
+      "step": 6041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.848160861965429,
+      "learning_rate": 1.60453284491964e-05,
+      "loss": 0.6883,
+      "step": 6042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 24.661372682422005,
+      "learning_rate": 1.6043985011180062e-05,
+      "loss": 0.7748,
+      "step": 6043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.149216708158821,
+      "learning_rate": 1.6042641401276326e-05,
+      "loss": 0.666,
+      "step": 6044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.574026378062864,
+      "learning_rate": 1.6041297619523405e-05,
+      "loss": 0.685,
+      "step": 6045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.945768089026576,
+      "learning_rate": 1.603995366595951e-05,
+      "loss": 0.7174,
+      "step": 6046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.517366707254787,
+      "learning_rate": 1.603860954062287e-05,
+      "loss": 0.685,
+      "step": 6047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.701941173207373,
+      "learning_rate": 1.6037265243551704e-05,
+      "loss": 0.6415,
+      "step": 6048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 24.461719324391474,
+      "learning_rate": 1.6035920774784247e-05,
+      "loss": 0.7319,
+      "step": 6049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.784512617949902,
+      "learning_rate": 1.6034576134358736e-05,
+      "loss": 0.7536,
+      "step": 6050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 21.90841418621444,
+      "learning_rate": 1.6033231322313406e-05,
+      "loss": 0.7027,
+      "step": 6051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.118060865550813,
+      "learning_rate": 1.603188633868651e-05,
+      "loss": 0.7073,
+      "step": 6052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.89676469955705,
+      "learning_rate": 1.6030541183516292e-05,
+      "loss": 0.783,
+      "step": 6053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.16690837320704,
+      "learning_rate": 1.6029195856841015e-05,
+      "loss": 0.6625,
+      "step": 6054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.620054953732446,
+      "learning_rate": 1.602785035869893e-05,
+      "loss": 0.6994,
+      "step": 6055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 22.97839752050187,
+      "learning_rate": 1.602650468912831e-05,
+      "loss": 0.7622,
+      "step": 6056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.639519712070097,
+      "learning_rate": 1.602515884816742e-05,
+      "loss": 0.6513,
+      "step": 6057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.51940529456008,
+      "learning_rate": 1.602381283585454e-05,
+      "loss": 0.8177,
+      "step": 6058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.49837469842179,
+      "learning_rate": 1.6022466652227944e-05,
+      "loss": 0.7179,
+      "step": 6059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.829366897843116,
+      "learning_rate": 1.602112029732592e-05,
+      "loss": 0.7255,
+      "step": 6060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.660995152306217,
+      "learning_rate": 1.6019773771186756e-05,
+      "loss": 0.792,
+      "step": 6061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.198536408158073,
+      "learning_rate": 1.601842707384875e-05,
+      "loss": 0.7614,
+      "step": 6062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 21.285158852870904,
+      "learning_rate": 1.60170802053502e-05,
+      "loss": 0.7865,
+      "step": 6063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.801970257246602,
+      "learning_rate": 1.6015733165729406e-05,
+      "loss": 0.6287,
+      "step": 6064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.967032262083727,
+      "learning_rate": 1.601438595502468e-05,
+      "loss": 0.7677,
+      "step": 6065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.766403354804233,
+      "learning_rate": 1.6013038573274336e-05,
+      "loss": 0.719,
+      "step": 6066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.585718313963557,
+      "learning_rate": 1.6011691020516696e-05,
+      "loss": 0.7175,
+      "step": 6067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.55071192695079,
+      "learning_rate": 1.6010343296790074e-05,
+      "loss": 0.7438,
+      "step": 6068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.923081299536978,
+      "learning_rate": 1.6008995402132807e-05,
+      "loss": 0.7245,
+      "step": 6069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.24145848716582,
+      "learning_rate": 1.6007647336583227e-05,
+      "loss": 0.6264,
+      "step": 6070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.482393296613985,
+      "learning_rate": 1.600629910017967e-05,
+      "loss": 0.6931,
+      "step": 6071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.306325195697898,
+      "learning_rate": 1.600495069296048e-05,
+      "loss": 0.6905,
+      "step": 6072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.890905039259907,
+      "learning_rate": 1.6003602114964007e-05,
+      "loss": 0.6545,
+      "step": 6073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.400955227430988,
+      "learning_rate": 1.60022533662286e-05,
+      "loss": 0.7456,
+      "step": 6074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.88797996672957,
+      "learning_rate": 1.600090444679262e-05,
+      "loss": 0.7343,
+      "step": 6075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.141422330507574,
+      "learning_rate": 1.5999555356694426e-05,
+      "loss": 0.6491,
+      "step": 6076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.750025277764754,
+      "learning_rate": 1.599820609597239e-05,
+      "loss": 0.6739,
+      "step": 6077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.70318905442603,
+      "learning_rate": 1.5996856664664876e-05,
+      "loss": 0.6953,
+      "step": 6078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.838339184236734,
+      "learning_rate": 1.599550706281027e-05,
+      "loss": 0.8145,
+      "step": 6079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 43.52356748710453,
+      "learning_rate": 1.5994157290446953e-05,
+      "loss": 0.6234,
+      "step": 6080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.221769869464637,
+      "learning_rate": 1.5992807347613305e-05,
+      "loss": 0.7163,
+      "step": 6081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 22.25877150012426,
+      "learning_rate": 1.5991457234347725e-05,
+      "loss": 0.7037,
+      "step": 6082
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.356749499155963,
+      "learning_rate": 1.5990106950688605e-05,
+      "loss": 0.6282,
+      "step": 6083
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.581752400514258,
+      "learning_rate": 1.5988756496674345e-05,
+      "loss": 0.6403,
+      "step": 6084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.290055549234074,
+      "learning_rate": 1.5987405872343358e-05,
+      "loss": 0.7299,
+      "step": 6085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 35.39351909164021,
+      "learning_rate": 1.5986055077734046e-05,
+      "loss": 0.6297,
+      "step": 6086
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.03817085595391,
+      "learning_rate": 1.5984704112884834e-05,
+      "loss": 0.7216,
+      "step": 6087
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 157.62999318720173,
+      "learning_rate": 1.5983352977834132e-05,
+      "loss": 0.7822,
+      "step": 6088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.102644789452551,
+      "learning_rate": 1.5982001672620376e-05,
+      "loss": 0.6764,
+      "step": 6089
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.830539053822466,
+      "learning_rate": 1.5980650197281994e-05,
+      "loss": 0.784,
+      "step": 6090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.20774198726569,
+      "learning_rate": 1.597929855185741e-05,
+      "loss": 0.7037,
+      "step": 6091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.601887583958202,
+      "learning_rate": 1.597794673638508e-05,
+      "loss": 0.6495,
+      "step": 6092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.91086018416681,
+      "learning_rate": 1.597659475090344e-05,
+      "loss": 0.6718,
+      "step": 6093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.735718613066542,
+      "learning_rate": 1.5975242595450946e-05,
+      "loss": 0.6484,
+      "step": 6094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 21.485949560554143,
+      "learning_rate": 1.5973890270066042e-05,
+      "loss": 0.694,
+      "step": 6095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.755962155331988,
+      "learning_rate": 1.5972537774787194e-05,
+      "loss": 0.6725,
+      "step": 6096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 25.94634781643838,
+      "learning_rate": 1.597118510965287e-05,
+      "loss": 0.8582,
+      "step": 6097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.60803522865118,
+      "learning_rate": 1.5969832274701528e-05,
+      "loss": 0.803,
+      "step": 6098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 23.74992923318208,
+      "learning_rate": 1.5968479269971654e-05,
+      "loss": 0.8157,
+      "step": 6099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 23.218071783898644,
+      "learning_rate": 1.596712609550172e-05,
+      "loss": 0.736,
+      "step": 6100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.684420037970854,
+      "learning_rate": 1.5965772751330207e-05,
+      "loss": 0.6614,
+      "step": 6101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.545800636779923,
+      "learning_rate": 1.596441923749561e-05,
+      "loss": 0.6814,
+      "step": 6102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.094761465892764,
+      "learning_rate": 1.5963065554036418e-05,
+      "loss": 0.6432,
+      "step": 6103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 26.097896790077538,
+      "learning_rate": 1.596171170099113e-05,
+      "loss": 0.775,
+      "step": 6104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.60526115145929,
+      "learning_rate": 1.596035767839825e-05,
+      "loss": 0.7764,
+      "step": 6105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.457857204946226,
+      "learning_rate": 1.595900348629628e-05,
+      "loss": 0.7346,
+      "step": 6106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.03574651350869,
+      "learning_rate": 1.5957649124723743e-05,
+      "loss": 0.6761,
+      "step": 6107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.79131792890106,
+      "learning_rate": 1.5956294593719147e-05,
+      "loss": 0.6428,
+      "step": 6108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 26.21022323169579,
+      "learning_rate": 1.5954939893321016e-05,
+      "loss": 0.7119,
+      "step": 6109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.838029359524974,
+      "learning_rate": 1.595358502356788e-05,
+      "loss": 0.7915,
+      "step": 6110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.35842348443619,
+      "learning_rate": 1.5952229984498265e-05,
+      "loss": 0.7209,
+      "step": 6111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.590018111377717,
+      "learning_rate": 1.5950874776150715e-05,
+      "loss": 0.7623,
+      "step": 6112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.048860509236775,
+      "learning_rate": 1.5949519398563766e-05,
+      "loss": 0.767,
+      "step": 6113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.0661945402471,
+      "learning_rate": 1.594816385177597e-05,
+      "loss": 0.6822,
+      "step": 6114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.486607023168693,
+      "learning_rate": 1.5946808135825866e-05,
+      "loss": 0.6685,
+      "step": 6115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.553040729827732,
+      "learning_rate": 1.5945452250752022e-05,
+      "loss": 0.7848,
+      "step": 6116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 24.321822710506687,
+      "learning_rate": 1.5944096196592994e-05,
+      "loss": 0.7468,
+      "step": 6117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.95007752930136,
+      "learning_rate": 1.5942739973387345e-05,
+      "loss": 0.6842,
+      "step": 6118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.568679781289745,
+      "learning_rate": 1.594138358117365e-05,
+      "loss": 0.782,
+      "step": 6119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.57447788772685,
+      "learning_rate": 1.594002701999048e-05,
+      "loss": 0.745,
+      "step": 6120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.92677475008517,
+      "learning_rate": 1.5938670289876418e-05,
+      "loss": 0.7654,
+      "step": 6121
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.45918067710494,
+      "learning_rate": 1.5937313390870045e-05,
+      "loss": 0.666,
+      "step": 6122
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 24.467953735978377,
+      "learning_rate": 1.5935956323009954e-05,
+      "loss": 0.7535,
+      "step": 6123
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.270819109929526,
+      "learning_rate": 1.5934599086334733e-05,
+      "loss": 0.618,
+      "step": 6124
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.83316831777706,
+      "learning_rate": 1.593324168088299e-05,
+      "loss": 0.7432,
+      "step": 6125
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.483902816731982,
+      "learning_rate": 1.5931884106693326e-05,
+      "loss": 0.6655,
+      "step": 6126
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.763631881407356,
+      "learning_rate": 1.5930526363804344e-05,
+      "loss": 0.7791,
+      "step": 6127
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.520092686101282,
+      "learning_rate": 1.5929168452254664e-05,
+      "loss": 0.7591,
+      "step": 6128
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.464502903706038,
+      "learning_rate": 1.5927810372082903e-05,
+      "loss": 0.7043,
+      "step": 6129
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.67035408170661,
+      "learning_rate": 1.592645212332768e-05,
+      "loss": 0.674,
+      "step": 6130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 23.229569894486914,
+      "learning_rate": 1.5925093706027625e-05,
+      "loss": 0.6965,
+      "step": 6131
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 24.16192466042968,
+      "learning_rate": 1.5923735120221372e-05,
+      "loss": 0.7232,
+      "step": 6132
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.857956823309355,
+      "learning_rate": 1.5922376365947557e-05,
+      "loss": 0.6704,
+      "step": 6133
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.807482009906417,
+      "learning_rate": 1.5921017443244825e-05,
+      "loss": 0.7533,
+      "step": 6134
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.69490538977884,
+      "learning_rate": 1.5919658352151818e-05,
+      "loss": 0.6534,
+      "step": 6135
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.612192071799736,
+      "learning_rate": 1.5918299092707194e-05,
+      "loss": 0.707,
+      "step": 6136
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.23309693192679,
+      "learning_rate": 1.5916939664949602e-05,
+      "loss": 0.6635,
+      "step": 6137
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 28.64551615845766,
+      "learning_rate": 1.591558006891771e-05,
+      "loss": 0.7467,
+      "step": 6138
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 22.691285514904354,
+      "learning_rate": 1.5914220304650177e-05,
+      "loss": 0.693,
+      "step": 6139
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.408445064087257,
+      "learning_rate": 1.5912860372185683e-05,
+      "loss": 0.7431,
+      "step": 6140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.917755096572193,
+      "learning_rate": 1.5911500271562895e-05,
+      "loss": 0.726,
+      "step": 6141
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.26411704320793,
+      "learning_rate": 1.5910140002820503e-05,
+      "loss": 0.5939,
+      "step": 6142
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.180750087450953,
+      "learning_rate": 1.5908779565997182e-05,
+      "loss": 0.6788,
+      "step": 6143
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 28.832142271379315,
+      "learning_rate": 1.5907418961131628e-05,
+      "loss": 0.8077,
+      "step": 6144
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 23.37061146476504,
+      "learning_rate": 1.5906058188262534e-05,
+      "loss": 0.6322,
+      "step": 6145
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.14986540833794,
+      "learning_rate": 1.5904697247428602e-05,
+      "loss": 0.6985,
+      "step": 6146
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.1414003195211,
+      "learning_rate": 1.5903336138668532e-05,
+      "loss": 0.6955,
+      "step": 6147
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.395505968840943,
+      "learning_rate": 1.5901974862021036e-05,
+      "loss": 0.753,
+      "step": 6148
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.91349934992944,
+      "learning_rate": 1.5900613417524827e-05,
+      "loss": 0.744,
+      "step": 6149
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.357660189111426,
+      "learning_rate": 1.5899251805218625e-05,
+      "loss": 0.7316,
+      "step": 6150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 27.979583490308016,
+      "learning_rate": 1.5897890025141153e-05,
+      "loss": 0.8225,
+      "step": 6151
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.383404537791456,
+      "learning_rate": 1.589652807733114e-05,
+      "loss": 0.6453,
+      "step": 6152
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.601501822409327,
+      "learning_rate": 1.5895165961827315e-05,
+      "loss": 0.6967,
+      "step": 6153
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 26.27885179032878,
+      "learning_rate": 1.5893803678668424e-05,
+      "loss": 0.6976,
+      "step": 6154
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 29.475166394151895,
+      "learning_rate": 1.5892441227893197e-05,
+      "loss": 0.7154,
+      "step": 6155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.082518067099613,
+      "learning_rate": 1.589107860954039e-05,
+      "loss": 0.768,
+      "step": 6156
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.038815927273703,
+      "learning_rate": 1.588971582364876e-05,
+      "loss": 0.6498,
+      "step": 6157
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.421322803865195,
+      "learning_rate": 1.588835287025705e-05,
+      "loss": 0.7573,
+      "step": 6158
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.50476274161912,
+      "learning_rate": 1.588698974940403e-05,
+      "loss": 0.6594,
+      "step": 6159
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.396100140412013,
+      "learning_rate": 1.5885626461128467e-05,
+      "loss": 0.7322,
+      "step": 6160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.76339265350174,
+      "learning_rate": 1.588426300546913e-05,
+      "loss": 0.7459,
+      "step": 6161
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 28.619583614843886,
+      "learning_rate": 1.5882899382464795e-05,
+      "loss": 0.7357,
+      "step": 6162
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.08433901353271,
+      "learning_rate": 1.5881535592154245e-05,
+      "loss": 0.7771,
+      "step": 6163
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.57934639374916,
+      "learning_rate": 1.5880171634576262e-05,
+      "loss": 0.7524,
+      "step": 6164
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.70357114893932,
+      "learning_rate": 1.587880750976964e-05,
+      "loss": 0.6779,
+      "step": 6165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.966815701738245,
+      "learning_rate": 1.5877443217773166e-05,
+      "loss": 0.7679,
+      "step": 6166
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.02784629787974,
+      "learning_rate": 1.5876078758625647e-05,
+      "loss": 0.7229,
+      "step": 6167
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.905083016833325,
+      "learning_rate": 1.5874714132365887e-05,
+      "loss": 0.7336,
+      "step": 6168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.604266264866526,
+      "learning_rate": 1.5873349339032693e-05,
+      "loss": 0.6868,
+      "step": 6169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.742797501879371,
+      "learning_rate": 1.587198437866488e-05,
+      "loss": 0.6729,
+      "step": 6170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.6087345679823,
+      "learning_rate": 1.5870619251301267e-05,
+      "loss": 0.7424,
+      "step": 6171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.605849624038399,
+      "learning_rate": 1.5869253956980672e-05,
+      "loss": 0.6427,
+      "step": 6172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 26.16999020180721,
+      "learning_rate": 1.5867888495741934e-05,
+      "loss": 0.7078,
+      "step": 6173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.37528765949897,
+      "learning_rate": 1.5866522867623876e-05,
+      "loss": 0.7846,
+      "step": 6174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.227007441281597,
+      "learning_rate": 1.586515707266534e-05,
+      "loss": 0.6812,
+      "step": 6175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.90701736600108,
+      "learning_rate": 1.5863791110905172e-05,
+      "loss": 0.7431,
+      "step": 6176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.573309327177764,
+      "learning_rate": 1.586242498238221e-05,
+      "loss": 0.6916,
+      "step": 6177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 23.951134870883674,
+      "learning_rate": 1.5861058687135314e-05,
+      "loss": 0.8912,
+      "step": 6178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.686810916603525,
+      "learning_rate": 1.5859692225203335e-05,
+      "loss": 0.6475,
+      "step": 6179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.335740569056334,
+      "learning_rate": 1.585832559662514e-05,
+      "loss": 0.66,
+      "step": 6180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.311238841262115,
+      "learning_rate": 1.585695880143959e-05,
+      "loss": 0.7251,
+      "step": 6181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.062592071560996,
+      "learning_rate": 1.5855591839685556e-05,
+      "loss": 0.7371,
+      "step": 6182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.24589320544766,
+      "learning_rate": 1.585422471140192e-05,
+      "loss": 0.7716,
+      "step": 6183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.928098998528,
+      "learning_rate": 1.5852857416627552e-05,
+      "loss": 0.6516,
+      "step": 6184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 28.28717905668244,
+      "learning_rate": 1.585148995540135e-05,
+      "loss": 0.7295,
+      "step": 6185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.73335874813438,
+      "learning_rate": 1.585012232776219e-05,
+      "loss": 0.7134,
+      "step": 6186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.872737960476485,
+      "learning_rate": 1.584875453374898e-05,
+      "loss": 0.7236,
+      "step": 6187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.06385799630569,
+      "learning_rate": 1.5847386573400605e-05,
+      "loss": 0.6823,
+      "step": 6188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.418897220507894,
+      "learning_rate": 1.584601844675598e-05,
+      "loss": 0.8347,
+      "step": 6189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.48595862593136,
+      "learning_rate": 1.584465015385401e-05,
+      "loss": 0.6605,
+      "step": 6190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.049688708968837,
+      "learning_rate": 1.584328169473361e-05,
+      "loss": 0.702,
+      "step": 6191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.301021634979367,
+      "learning_rate": 1.58419130694337e-05,
+      "loss": 0.7398,
+      "step": 6192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.44046627465891,
+      "learning_rate": 1.5840544277993193e-05,
+      "loss": 0.7179,
+      "step": 6193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.966605734474095,
+      "learning_rate": 1.5839175320451027e-05,
+      "loss": 0.7269,
+      "step": 6194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.76288541822951,
+      "learning_rate": 1.583780619684613e-05,
+      "loss": 0.6416,
+      "step": 6195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 21.82144262567529,
+      "learning_rate": 1.5836436907217438e-05,
+      "loss": 0.7189,
+      "step": 6196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.499975888359641,
+      "learning_rate": 1.58350674516039e-05,
+      "loss": 0.6673,
+      "step": 6197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.512830787258151,
+      "learning_rate": 1.583369783004445e-05,
+      "loss": 0.6559,
+      "step": 6198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.764767172207975,
+      "learning_rate": 1.5832328042578047e-05,
+      "loss": 0.6729,
+      "step": 6199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.18189812003746,
+      "learning_rate": 1.5830958089243652e-05,
+      "loss": 0.7005,
+      "step": 6200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.926758816958422,
+      "learning_rate": 1.5829587970080217e-05,
+      "loss": 0.6892,
+      "step": 6201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.682554361445563,
+      "learning_rate": 1.582821768512671e-05,
+      "loss": 0.6898,
+      "step": 6202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.80032111620127,
+      "learning_rate": 1.5826847234422102e-05,
+      "loss": 0.5931,
+      "step": 6203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.578395719655997,
+      "learning_rate": 1.5825476618005363e-05,
+      "loss": 0.7904,
+      "step": 6204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.478891744221988,
+      "learning_rate": 1.582410583591548e-05,
+      "loss": 0.6769,
+      "step": 6205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.466471831445197,
+      "learning_rate": 1.5822734888191437e-05,
+      "loss": 0.6986,
+      "step": 6206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.877754720031678,
+      "learning_rate": 1.5821363774872218e-05,
+      "loss": 0.7492,
+      "step": 6207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.61935295009391,
+      "learning_rate": 1.5819992495996818e-05,
+      "loss": 0.6531,
+      "step": 6208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 26.650594045296724,
+      "learning_rate": 1.5818621051604234e-05,
+      "loss": 0.6713,
+      "step": 6209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.13050746091519,
+      "learning_rate": 1.5817249441733475e-05,
+      "loss": 0.6609,
+      "step": 6210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.896457561928887,
+      "learning_rate": 1.5815877666423543e-05,
+      "loss": 0.733,
+      "step": 6211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 23.620864088589446,
+      "learning_rate": 1.5814505725713454e-05,
+      "loss": 0.6689,
+      "step": 6212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.294414013500338,
+      "learning_rate": 1.581313361964222e-05,
+      "loss": 0.6283,
+      "step": 6213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.857289407117033,
+      "learning_rate": 1.5811761348248872e-05,
+      "loss": 0.6373,
+      "step": 6214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.443122265630272,
+      "learning_rate": 1.5810388911572424e-05,
+      "loss": 0.7406,
+      "step": 6215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.48483365011168,
+      "learning_rate": 1.5809016309651916e-05,
+      "loss": 0.7768,
+      "step": 6216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.811447672544205,
+      "learning_rate": 1.5807643542526387e-05,
+      "loss": 0.749,
+      "step": 6217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 24.078537042571483,
+      "learning_rate": 1.580627061023487e-05,
+      "loss": 0.6296,
+      "step": 6218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.313146821825555,
+      "learning_rate": 1.5804897512816416e-05,
+      "loss": 0.733,
+      "step": 6219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.227180493819073,
+      "learning_rate": 1.580352425031007e-05,
+      "loss": 0.6935,
+      "step": 6220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.230630329294836,
+      "learning_rate": 1.5802150822754888e-05,
+      "loss": 0.8322,
+      "step": 6221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.605318888456289,
+      "learning_rate": 1.580077723018993e-05,
+      "loss": 0.7045,
+      "step": 6222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.056639086029023,
+      "learning_rate": 1.5799403472654266e-05,
+      "loss": 0.6404,
+      "step": 6223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.59353403238485,
+      "learning_rate": 1.5798029550186957e-05,
+      "loss": 0.692,
+      "step": 6224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.638346383175973,
+      "learning_rate": 1.5796655462827077e-05,
+      "loss": 0.6574,
+      "step": 6225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.147086033456212,
+      "learning_rate": 1.5795281210613708e-05,
+      "loss": 0.6948,
+      "step": 6226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.524126123015039,
+      "learning_rate": 1.5793906793585935e-05,
+      "loss": 0.6924,
+      "step": 6227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.899402099788624,
+      "learning_rate": 1.5792532211782837e-05,
+      "loss": 0.6904,
+      "step": 6228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.832425310976404,
+      "learning_rate": 1.5791157465243517e-05,
+      "loss": 0.7059,
+      "step": 6229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.189572613082234,
+      "learning_rate": 1.5789782554007063e-05,
+      "loss": 0.7952,
+      "step": 6230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.967314395288053,
+      "learning_rate": 1.578840747811258e-05,
+      "loss": 0.7456,
+      "step": 6231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.490645421287475,
+      "learning_rate": 1.5787032237599173e-05,
+      "loss": 0.6799,
+      "step": 6232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.400697747167488,
+      "learning_rate": 1.5785656832505956e-05,
+      "loss": 0.8336,
+      "step": 6233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.58609244321543,
+      "learning_rate": 1.5784281262872046e-05,
+      "loss": 0.6934,
+      "step": 6234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.762061707343198,
+      "learning_rate": 1.5782905528736558e-05,
+      "loss": 0.7065,
+      "step": 6235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.433959511728663,
+      "learning_rate": 1.578152963013862e-05,
+      "loss": 0.6244,
+      "step": 6236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.306377118909536,
+      "learning_rate": 1.578015356711736e-05,
+      "loss": 0.6028,
+      "step": 6237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 21.675394127297274,
+      "learning_rate": 1.5778777339711914e-05,
+      "loss": 0.8232,
+      "step": 6238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.718009130944154,
+      "learning_rate": 1.577740094796142e-05,
+      "loss": 0.6942,
+      "step": 6239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 24.662029480421435,
+      "learning_rate": 1.5776024391905026e-05,
+      "loss": 0.6917,
+      "step": 6240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.679605801722104,
+      "learning_rate": 1.5774647671581878e-05,
+      "loss": 0.7454,
+      "step": 6241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.049254008571342,
+      "learning_rate": 1.5773270787031124e-05,
+      "loss": 0.7414,
+      "step": 6242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 24.852855324957886,
+      "learning_rate": 1.577189373829193e-05,
+      "loss": 0.6756,
+      "step": 6243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.62704146910007,
+      "learning_rate": 1.5770516525403453e-05,
+      "loss": 0.7319,
+      "step": 6244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.562153743543485,
+      "learning_rate": 1.576913914840486e-05,
+      "loss": 0.6966,
+      "step": 6245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.28206774969318,
+      "learning_rate": 1.5767761607335327e-05,
+      "loss": 0.7001,
+      "step": 6246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.321958583315887,
+      "learning_rate": 1.5766383902234026e-05,
+      "loss": 0.6744,
+      "step": 6247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.249578056813633,
+      "learning_rate": 1.5765006033140142e-05,
+      "loss": 0.6593,
+      "step": 6248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.780097770914747,
+      "learning_rate": 1.5763628000092858e-05,
+      "loss": 0.7383,
+      "step": 6249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 26.861128854522114,
+      "learning_rate": 1.5762249803131365e-05,
+      "loss": 0.7002,
+      "step": 6250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.856258944628756,
+      "learning_rate": 1.5760871442294856e-05,
+      "loss": 0.8652,
+      "step": 6251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.94426362208592,
+      "learning_rate": 1.5759492917622537e-05,
+      "loss": 0.6385,
+      "step": 6252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.115648218541228,
+      "learning_rate": 1.5758114229153606e-05,
+      "loss": 0.7313,
+      "step": 6253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.829432571142346,
+      "learning_rate": 1.575673537692728e-05,
+      "loss": 0.6776,
+      "step": 6254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 26.05287736126642,
+      "learning_rate": 1.575535636098276e-05,
+      "loss": 0.6779,
+      "step": 6255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.580785969844754,
+      "learning_rate": 1.5753977181359277e-05,
+      "loss": 0.7603,
+      "step": 6256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.562898043745616,
+      "learning_rate": 1.5752597838096046e-05,
+      "loss": 0.6555,
+      "step": 6257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 19.524721865101647,
+      "learning_rate": 1.57512183312323e-05,
+      "loss": 0.693,
+      "step": 6258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.6832497298869,
+      "learning_rate": 1.574983866080727e-05,
+      "loss": 0.7587,
+      "step": 6259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 19.216483982755733,
+      "learning_rate": 1.5748458826860185e-05,
+      "loss": 0.683,
+      "step": 6260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.548388677261059,
+      "learning_rate": 1.5747078829430302e-05,
+      "loss": 0.7515,
+      "step": 6261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.187351731257,
+      "learning_rate": 1.5745698668556856e-05,
+      "loss": 0.7154,
+      "step": 6262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.402076516673862,
+      "learning_rate": 1.5744318344279103e-05,
+      "loss": 0.6895,
+      "step": 6263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 25.176644906322444,
+      "learning_rate": 1.5742937856636294e-05,
+      "loss": 0.6538,
+      "step": 6264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.097045165379267,
+      "learning_rate": 1.5741557205667688e-05,
+      "loss": 0.697,
+      "step": 6265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.51688942833326,
+      "learning_rate": 1.574017639141256e-05,
+      "loss": 0.7147,
+      "step": 6266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.356930145712544,
+      "learning_rate": 1.5738795413910174e-05,
+      "loss": 0.674,
+      "step": 6267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.798673484233632,
+      "learning_rate": 1.57374142731998e-05,
+      "loss": 0.6925,
+      "step": 6268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.950738697464558,
+      "learning_rate": 1.573603296932072e-05,
+      "loss": 0.7904,
+      "step": 6269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.508249749744827,
+      "learning_rate": 1.5734651502312218e-05,
+      "loss": 0.7407,
+      "step": 6270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.141409384066787,
+      "learning_rate": 1.5733269872213583e-05,
+      "loss": 0.7864,
+      "step": 6271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.444272982668206,
+      "learning_rate": 1.5731888079064107e-05,
+      "loss": 0.6863,
+      "step": 6272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.12951265467563,
+      "learning_rate": 1.5730506122903086e-05,
+      "loss": 0.7873,
+      "step": 6273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.93230822650624,
+      "learning_rate": 1.5729124003769826e-05,
+      "loss": 0.7465,
+      "step": 6274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.134391908430306,
+      "learning_rate": 1.572774172170363e-05,
+      "loss": 0.7148,
+      "step": 6275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.246212959067847,
+      "learning_rate": 1.5726359276743808e-05,
+      "loss": 0.6541,
+      "step": 6276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.883252963354481,
+      "learning_rate": 1.5724976668929678e-05,
+      "loss": 0.6974,
+      "step": 6277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.41424468561854,
+      "learning_rate": 1.5723593898300562e-05,
+      "loss": 0.7017,
+      "step": 6278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.743387526486405,
+      "learning_rate": 1.572221096489578e-05,
+      "loss": 0.65,
+      "step": 6279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.916708543579952,
+      "learning_rate": 1.572082786875467e-05,
+      "loss": 0.6345,
+      "step": 6280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 26.015664969519534,
+      "learning_rate": 1.5719444609916564e-05,
+      "loss": 0.7147,
+      "step": 6281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 30.076012645710914,
+      "learning_rate": 1.5718061188420793e-05,
+      "loss": 0.7767,
+      "step": 6282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.62416221234385,
+      "learning_rate": 1.571667760430671e-05,
+      "loss": 0.7267,
+      "step": 6283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.473050718298492,
+      "learning_rate": 1.5715293857613662e-05,
+      "loss": 0.7076,
+      "step": 6284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.892231357457838,
+      "learning_rate": 1.5713909948380995e-05,
+      "loss": 0.7583,
+      "step": 6285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.017917791540224,
+      "learning_rate": 1.5712525876648076e-05,
+      "loss": 0.7583,
+      "step": 6286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.676341254821256,
+      "learning_rate": 1.5711141642454258e-05,
+      "loss": 0.7764,
+      "step": 6287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.796402382012346,
+      "learning_rate": 1.5709757245838918e-05,
+      "loss": 0.7369,
+      "step": 6288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.376078308044274,
+      "learning_rate": 1.570837268684142e-05,
+      "loss": 0.7264,
+      "step": 6289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 19.632087278955975,
+      "learning_rate": 1.5706987965501142e-05,
+      "loss": 0.7711,
+      "step": 6290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.923638613941863,
+      "learning_rate": 1.5705603081857463e-05,
+      "loss": 0.6989,
+      "step": 6291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.242577810420823,
+      "learning_rate": 1.5704218035949773e-05,
+      "loss": 0.7212,
+      "step": 6292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.279137905210119,
+      "learning_rate": 1.570283282781746e-05,
+      "loss": 0.7621,
+      "step": 6293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.450926928379797,
+      "learning_rate": 1.5701447457499914e-05,
+      "loss": 0.791,
+      "step": 6294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.837420604916066,
+      "learning_rate": 1.570006192503654e-05,
+      "loss": 0.77,
+      "step": 6295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.909350604393193,
+      "learning_rate": 1.5698676230466735e-05,
+      "loss": 0.7334,
+      "step": 6296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.886295373527055,
+      "learning_rate": 1.5697290373829913e-05,
+      "loss": 0.6799,
+      "step": 6297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.233097679083436,
+      "learning_rate": 1.5695904355165486e-05,
+      "loss": 0.7532,
+      "step": 6298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.108783529765777,
+      "learning_rate": 1.5694518174512873e-05,
+      "loss": 0.6611,
+      "step": 6299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.657668291364995,
+      "learning_rate": 1.5693131831911494e-05,
+      "loss": 0.7844,
+      "step": 6300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 25.34681562116457,
+      "learning_rate": 1.5691745327400776e-05,
+      "loss": 0.7776,
+      "step": 6301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 30.17676812592073,
+      "learning_rate": 1.569035866102015e-05,
+      "loss": 0.7234,
+      "step": 6302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.39102641685156,
+      "learning_rate": 1.568897183280905e-05,
+      "loss": 0.7063,
+      "step": 6303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.93110483175266,
+      "learning_rate": 1.5687584842806925e-05,
+      "loss": 0.6177,
+      "step": 6304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.526691742169396,
+      "learning_rate": 1.568619769105321e-05,
+      "loss": 0.6045,
+      "step": 6305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 22.57082940305604,
+      "learning_rate": 1.568481037758736e-05,
+      "loss": 0.7763,
+      "step": 6306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.24234523855676,
+      "learning_rate": 1.5683422902448827e-05,
+      "loss": 0.7059,
+      "step": 6307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.985835218432385,
+      "learning_rate": 1.5682035265677074e-05,
+      "loss": 0.6476,
+      "step": 6308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.889131048595345,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.7627,
+      "step": 6309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 27.590260461076955,
+      "learning_rate": 1.5679259507391755e-05,
+      "loss": 0.7267,
+      "step": 6310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.37743047520126,
+      "learning_rate": 1.5677871385957134e-05,
+      "loss": 0.731,
+      "step": 6311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.17826006522586,
+      "learning_rate": 1.567648310304717e-05,
+      "loss": 0.7181,
+      "step": 6312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.444323788353781,
+      "learning_rate": 1.567509465870135e-05,
+      "loss": 0.7175,
+      "step": 6313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.491751149568916,
+      "learning_rate": 1.567370605295915e-05,
+      "loss": 0.7675,
+      "step": 6314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.453460728283682,
+      "learning_rate": 1.5672317285860076e-05,
+      "loss": 0.6342,
+      "step": 6315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 22.80354555471989,
+      "learning_rate": 1.5670928357443617e-05,
+      "loss": 0.7829,
+      "step": 6316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.4134458670892,
+      "learning_rate": 1.566953926774927e-05,
+      "loss": 0.7413,
+      "step": 6317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.142592560890225,
+      "learning_rate": 1.5668150016816545e-05,
+      "loss": 0.7177,
+      "step": 6318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.270432504738718,
+      "learning_rate": 1.5666760604684947e-05,
+      "loss": 0.7444,
+      "step": 6319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.252765065559505,
+      "learning_rate": 1.5665371031393994e-05,
+      "loss": 0.8004,
+      "step": 6320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 22.988404591763434,
+      "learning_rate": 1.56639812969832e-05,
+      "loss": 0.7443,
+      "step": 6321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.969777669751174,
+      "learning_rate": 1.5662591401492096e-05,
+      "loss": 0.7221,
+      "step": 6322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.280438987023008,
+      "learning_rate": 1.5661201344960203e-05,
+      "loss": 0.7201,
+      "step": 6323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 37.57959628327843,
+      "learning_rate": 1.5659811127427053e-05,
+      "loss": 0.767,
+      "step": 6324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.437502706932296,
+      "learning_rate": 1.5658420748932187e-05,
+      "loss": 0.6743,
+      "step": 6325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 22.148492767628333,
+      "learning_rate": 1.5657030209515146e-05,
+      "loss": 0.7354,
+      "step": 6326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.79360772813145,
+      "learning_rate": 1.5655639509215476e-05,
+      "loss": 0.6183,
+      "step": 6327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.334730897115843,
+      "learning_rate": 1.565424864807273e-05,
+      "loss": 0.6637,
+      "step": 6328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 22.47819394584062,
+      "learning_rate": 1.565285762612645e-05,
+      "loss": 0.6701,
+      "step": 6329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.261798390201033,
+      "learning_rate": 1.5651466443416213e-05,
+      "loss": 0.6882,
+      "step": 6330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.828376330727263,
+      "learning_rate": 1.5650075099981573e-05,
+      "loss": 0.7075,
+      "step": 6331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.908258572557916,
+      "learning_rate": 1.564868359586211e-05,
+      "loss": 0.7339,
+      "step": 6332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.5335566250218,
+      "learning_rate": 1.564729193109738e-05,
+      "loss": 0.6724,
+      "step": 6333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.72041759943538,
+      "learning_rate": 1.5645900105726976e-05,
+      "loss": 0.6783,
+      "step": 6334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 25.438276170094106,
+      "learning_rate": 1.5644508119790477e-05,
+      "loss": 0.7547,
+      "step": 6335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.496643036512626,
+      "learning_rate": 1.5643115973327464e-05,
+      "loss": 0.7009,
+      "step": 6336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 26.56720763395691,
+      "learning_rate": 1.5641723666377536e-05,
+      "loss": 0.7125,
+      "step": 6337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.754529238248484,
+      "learning_rate": 1.564033119898029e-05,
+      "loss": 0.7256,
+      "step": 6338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.333374418697414,
+      "learning_rate": 1.5638938571175324e-05,
+      "loss": 0.6783,
+      "step": 6339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.92363618146433,
+      "learning_rate": 1.563754578300224e-05,
+      "loss": 0.6673,
+      "step": 6340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.209951178308156,
+      "learning_rate": 1.5636152834500654e-05,
+      "loss": 0.6956,
+      "step": 6341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.734447474584446,
+      "learning_rate": 1.5634759725710178e-05,
+      "loss": 0.6897,
+      "step": 6342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.960526927169756,
+      "learning_rate": 1.5633366456670433e-05,
+      "loss": 0.7769,
+      "step": 6343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.17234637542746,
+      "learning_rate": 1.563197302742104e-05,
+      "loss": 0.6268,
+      "step": 6344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.327554166181422,
+      "learning_rate": 1.5630579438001626e-05,
+      "loss": 0.671,
+      "step": 6345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.89072928584197,
+      "learning_rate": 1.562918568845183e-05,
+      "loss": 0.6727,
+      "step": 6346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.526235828507357,
+      "learning_rate": 1.562779177881129e-05,
+      "loss": 0.6639,
+      "step": 6347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.72861549057825,
+      "learning_rate": 1.5626397709119638e-05,
+      "loss": 0.6892,
+      "step": 6348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 22.15658985786569,
+      "learning_rate": 1.562500347941653e-05,
+      "loss": 0.6799,
+      "step": 6349
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.507176640634484,
+      "learning_rate": 1.5623609089741608e-05,
+      "loss": 0.7295,
+      "step": 6350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.77365799424598,
+      "learning_rate": 1.5622214540134536e-05,
+      "loss": 0.7676,
+      "step": 6351
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.633065273269677,
+      "learning_rate": 1.5620819830634975e-05,
+      "loss": 0.72,
+      "step": 6352
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.624250783461548,
+      "learning_rate": 1.5619424961282585e-05,
+      "loss": 0.6405,
+      "step": 6353
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.837879893131927,
+      "learning_rate": 1.5618029932117035e-05,
+      "loss": 0.672,
+      "step": 6354
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.2073531832548,
+      "learning_rate": 1.5616634743177996e-05,
+      "loss": 0.6685,
+      "step": 6355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.607158163602943,
+      "learning_rate": 1.5615239394505154e-05,
+      "loss": 0.7397,
+      "step": 6356
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.430247388166993,
+      "learning_rate": 1.5613843886138192e-05,
+      "loss": 0.6731,
+      "step": 6357
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.615028495457203,
+      "learning_rate": 1.561244821811679e-05,
+      "loss": 0.7111,
+      "step": 6358
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.608515768526523,
+      "learning_rate": 1.5611052390480646e-05,
+      "loss": 0.7206,
+      "step": 6359
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.936369934115836,
+      "learning_rate": 1.5609656403269452e-05,
+      "loss": 0.6903,
+      "step": 6360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.944315726390709,
+      "learning_rate": 1.560826025652291e-05,
+      "loss": 0.7447,
+      "step": 6361
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.259277886198085,
+      "learning_rate": 1.560686395028073e-05,
+      "loss": 0.6132,
+      "step": 6362
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.02463172180988,
+      "learning_rate": 1.560546748458262e-05,
+      "loss": 0.6949,
+      "step": 6363
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 30.59667730938663,
+      "learning_rate": 1.5604070859468292e-05,
+      "loss": 0.6782,
+      "step": 6364
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.042902305654117,
+      "learning_rate": 1.5602674074977467e-05,
+      "loss": 0.7063,
+      "step": 6365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.800659256294253,
+      "learning_rate": 1.560127713114987e-05,
+      "loss": 0.7687,
+      "step": 6366
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.89127829661326,
+      "learning_rate": 1.559988002802523e-05,
+      "loss": 0.6818,
+      "step": 6367
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.499096024221645,
+      "learning_rate": 1.5598482765643273e-05,
+      "loss": 0.6369,
+      "step": 6368
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.145687953391345,
+      "learning_rate": 1.5597085344043742e-05,
+      "loss": 0.6976,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 27.72505787980115,
+      "learning_rate": 1.5595687763266378e-05,
+      "loss": 0.7195,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.090530167227787,
+      "learning_rate": 1.5594290023350933e-05,
+      "loss": 0.6804,
+      "step": 6371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 23.81879930735099,
+      "learning_rate": 1.5592892124337145e-05,
+      "loss": 0.7284,
+      "step": 6372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.79588388477555,
+      "learning_rate": 1.559149406626478e-05,
+      "loss": 0.7053,
+      "step": 6373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.071680652907293,
+      "learning_rate": 1.5590095849173597e-05,
+      "loss": 0.6733,
+      "step": 6374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.068402627502268,
+      "learning_rate": 1.5588697473103355e-05,
+      "loss": 0.6621,
+      "step": 6375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 23.994208110113572,
+      "learning_rate": 1.5587298938093828e-05,
+      "loss": 0.6938,
+      "step": 6376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 27.89978263360765,
+      "learning_rate": 1.5585900244184785e-05,
+      "loss": 0.7703,
+      "step": 6377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 22.50729676979544,
+      "learning_rate": 1.5584501391416008e-05,
+      "loss": 0.6832,
+      "step": 6378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.379438339225747,
+      "learning_rate": 1.558310237982728e-05,
+      "loss": 0.6573,
+      "step": 6379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.45984504218609,
+      "learning_rate": 1.558170320945838e-05,
+      "loss": 0.6878,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.599490474879698,
+      "learning_rate": 1.5580303880349113e-05,
+      "loss": 0.6284,
+      "step": 6381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.223205712944264,
+      "learning_rate": 1.5578904392539268e-05,
+      "loss": 0.6712,
+      "step": 6382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.117554979449,
+      "learning_rate": 1.557750474606864e-05,
+      "loss": 0.7686,
+      "step": 6383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.909667976630182,
+      "learning_rate": 1.5576104940977045e-05,
+      "loss": 0.6952,
+      "step": 6384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 21.715535244308136,
+      "learning_rate": 1.5574704977304286e-05,
+      "loss": 0.7224,
+      "step": 6385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.4192327668256,
+      "learning_rate": 1.557330485509018e-05,
+      "loss": 0.6875,
+      "step": 6386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.048427537679707,
+      "learning_rate": 1.5571904574374543e-05,
+      "loss": 0.7214,
+      "step": 6387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.260240005029573,
+      "learning_rate": 1.55705041351972e-05,
+      "loss": 0.7188,
+      "step": 6388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.199730289557138,
+      "learning_rate": 1.5569103537597972e-05,
+      "loss": 0.7141,
+      "step": 6389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.794480954963547,
+      "learning_rate": 1.55677027816167e-05,
+      "loss": 0.736,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.265690492589275,
+      "learning_rate": 1.5566301867293223e-05,
+      "loss": 0.6804,
+      "step": 6391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.674947584000783,
+      "learning_rate": 1.5564900794667372e-05,
+      "loss": 0.6766,
+      "step": 6392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.604538746373462,
+      "learning_rate": 1.5563499563779002e-05,
+      "loss": 0.7749,
+      "step": 6393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.775019820371973,
+      "learning_rate": 1.5562098174667957e-05,
+      "loss": 0.6711,
+      "step": 6394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 27.382511781573122,
+      "learning_rate": 1.556069662737409e-05,
+      "loss": 0.8826,
+      "step": 6395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.75612464784674,
+      "learning_rate": 1.5559294921937272e-05,
+      "loss": 0.7504,
+      "step": 6396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.36544359471965,
+      "learning_rate": 1.555789305839735e-05,
+      "loss": 0.5571,
+      "step": 6397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.937426314464096,
+      "learning_rate": 1.5556491036794204e-05,
+      "loss": 0.6322,
+      "step": 6398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.087657573041733,
+      "learning_rate": 1.5555088857167703e-05,
+      "loss": 0.6066,
+      "step": 6399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.06324289781649,
+      "learning_rate": 1.5553686519557726e-05,
+      "loss": 0.7523,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.220169261789898,
+      "learning_rate": 1.5552284024004154e-05,
+      "loss": 0.5992,
+      "step": 6401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.19145050819094,
+      "learning_rate": 1.555088137054687e-05,
+      "loss": 0.6211,
+      "step": 6402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.2437545282828,
+      "learning_rate": 1.554947855922577e-05,
+      "loss": 0.7152,
+      "step": 6403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 27.16781968871853,
+      "learning_rate": 1.5548075590080745e-05,
+      "loss": 0.7184,
+      "step": 6404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 28.54908774558578,
+      "learning_rate": 1.5546672463151695e-05,
+      "loss": 0.6511,
+      "step": 6405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 62.210625402081355,
+      "learning_rate": 1.5545269178478523e-05,
+      "loss": 0.6291,
+      "step": 6406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.581450136064625,
+      "learning_rate": 1.554386573610114e-05,
+      "loss": 0.6526,
+      "step": 6407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.038554100261715,
+      "learning_rate": 1.554246213605946e-05,
+      "loss": 0.6116,
+      "step": 6408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.306176302260706,
+      "learning_rate": 1.55410583783934e-05,
+      "loss": 0.7377,
+      "step": 6409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 29.4897848532468,
+      "learning_rate": 1.5539654463142878e-05,
+      "loss": 0.6722,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 24.97828358986796,
+      "learning_rate": 1.5538250390347825e-05,
+      "loss": 0.7174,
+      "step": 6411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.532038375590826,
+      "learning_rate": 1.5536846160048172e-05,
+      "loss": 0.6915,
+      "step": 6412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 30.187267639221538,
+      "learning_rate": 1.553544177228385e-05,
+      "loss": 0.6716,
+      "step": 6413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.411902530744527,
+      "learning_rate": 1.5534037227094807e-05,
+      "loss": 0.6789,
+      "step": 6414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 21.86502560255622,
+      "learning_rate": 1.553263252452098e-05,
+      "loss": 0.7738,
+      "step": 6415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.6030213927747,
+      "learning_rate": 1.553122766460232e-05,
+      "loss": 0.6946,
+      "step": 6416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 22.28326880563784,
+      "learning_rate": 1.552982264737878e-05,
+      "loss": 0.7725,
+      "step": 6417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.524584318138132,
+      "learning_rate": 1.5528417472890324e-05,
+      "loss": 0.6993,
+      "step": 6418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.205418527364696,
+      "learning_rate": 1.5527012141176904e-05,
+      "loss": 0.6616,
+      "step": 6419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.703448832232999,
+      "learning_rate": 1.5525606652278493e-05,
+      "loss": 0.6215,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.331133644745545,
+      "learning_rate": 1.5524201006235063e-05,
+      "loss": 0.7806,
+      "step": 6421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.713294211999813,
+      "learning_rate": 1.5522795203086584e-05,
+      "loss": 0.7002,
+      "step": 6422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.876963129760735,
+      "learning_rate": 1.5521389242873044e-05,
+      "loss": 0.6968,
+      "step": 6423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.581924863610826,
+      "learning_rate": 1.551998312563442e-05,
+      "loss": 0.6997,
+      "step": 6424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.814924504690554,
+      "learning_rate": 1.551857685141071e-05,
+      "loss": 0.6691,
+      "step": 6425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.574550622803667,
+      "learning_rate": 1.5517170420241897e-05,
+      "loss": 0.7012,
+      "step": 6426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.427129710839353,
+      "learning_rate": 1.551576383216799e-05,
+      "loss": 0.6208,
+      "step": 6427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.742300797727502,
+      "learning_rate": 1.5514357087228985e-05,
+      "loss": 0.7013,
+      "step": 6428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.469801133051082,
+      "learning_rate": 1.551295018546489e-05,
+      "loss": 0.556,
+      "step": 6429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.812763396449327,
+      "learning_rate": 1.5511543126915713e-05,
+      "loss": 0.6248,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.462182426274687,
+      "learning_rate": 1.5510135911621474e-05,
+      "loss": 0.8001,
+      "step": 6431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.442260031825697,
+      "learning_rate": 1.5508728539622196e-05,
+      "loss": 0.6867,
+      "step": 6432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.667367790946475,
+      "learning_rate": 1.5507321010957903e-05,
+      "loss": 0.6411,
+      "step": 6433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 26.82405009618746,
+      "learning_rate": 1.550591332566862e-05,
+      "loss": 0.8276,
+      "step": 6434
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 28.317963098612843,
+      "learning_rate": 1.550450548379438e-05,
+      "loss": 0.678,
+      "step": 6435
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.657453142697225,
+      "learning_rate": 1.5503097485375228e-05,
+      "loss": 0.635,
+      "step": 6436
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.13644110396231,
+      "learning_rate": 1.5501689330451203e-05,
+      "loss": 0.6308,
+      "step": 6437
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.74563325137112,
+      "learning_rate": 1.5500281019062347e-05,
+      "loss": 0.6649,
+      "step": 6438
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.28041175908318,
+      "learning_rate": 1.5498872551248722e-05,
+      "loss": 0.7186,
+      "step": 6439
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.692231973433241,
+      "learning_rate": 1.5497463927050375e-05,
+      "loss": 0.6133,
+      "step": 6440
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 24.996811089384508,
+      "learning_rate": 1.5496055146507368e-05,
+      "loss": 0.6893,
+      "step": 6441
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.039620496579285,
+      "learning_rate": 1.5494646209659775e-05,
+      "loss": 0.6556,
+      "step": 6442
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.81479343267713,
+      "learning_rate": 1.549323711654765e-05,
+      "loss": 0.6677,
+      "step": 6443
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.438537720370615,
+      "learning_rate": 1.549182786721108e-05,
+      "loss": 0.5729,
+      "step": 6444
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.437771074477322,
+      "learning_rate": 1.5490418461690137e-05,
+      "loss": 0.6372,
+      "step": 6445
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.703700780947685,
+      "learning_rate": 1.5489008900024903e-05,
+      "loss": 0.6979,
+      "step": 6446
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.84046898071911,
+      "learning_rate": 1.5487599182255467e-05,
+      "loss": 0.7347,
+      "step": 6447
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 25.446927423152637,
+      "learning_rate": 1.5486189308421922e-05,
+      "loss": 0.7067,
+      "step": 6448
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.861818366852788,
+      "learning_rate": 1.5484779278564363e-05,
+      "loss": 0.739,
+      "step": 6449
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.445507919874448,
+      "learning_rate": 1.5483369092722888e-05,
+      "loss": 0.6521,
+      "step": 6450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.52276352149899,
+      "learning_rate": 1.5481958750937605e-05,
+      "loss": 0.7105,
+      "step": 6451
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.278302783551332,
+      "learning_rate": 1.548054825324862e-05,
+      "loss": 0.7706,
+      "step": 6452
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.19005566113112,
+      "learning_rate": 1.547913759969605e-05,
+      "loss": 0.6834,
+      "step": 6453
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.101977136359437,
+      "learning_rate": 1.5477726790320012e-05,
+      "loss": 0.6386,
+      "step": 6454
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.9440510386669,
+      "learning_rate": 1.547631582516063e-05,
+      "loss": 0.6406,
+      "step": 6455
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.728639640331524,
+      "learning_rate": 1.5474904704258027e-05,
+      "loss": 0.5915,
+      "step": 6456
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 24.502157493094778,
+      "learning_rate": 1.5473493427652337e-05,
+      "loss": 0.708,
+      "step": 6457
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.493871406810868,
+      "learning_rate": 1.54720819953837e-05,
+      "loss": 0.69,
+      "step": 6458
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.440185088053983,
+      "learning_rate": 1.547067040749225e-05,
+      "loss": 0.702,
+      "step": 6459
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.26024440494361,
+      "learning_rate": 1.546925866401813e-05,
+      "loss": 0.7452,
+      "step": 6460
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 30.164224661248184,
+      "learning_rate": 1.5467846765001496e-05,
+      "loss": 0.7826,
+      "step": 6461
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.484245985135512,
+      "learning_rate": 1.54664347104825e-05,
+      "loss": 0.6766,
+      "step": 6462
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.16069597947086,
+      "learning_rate": 1.5465022500501294e-05,
+      "loss": 0.6465,
+      "step": 6463
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.559388702718227,
+      "learning_rate": 1.5463610135098048e-05,
+      "loss": 0.7666,
+      "step": 6464
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.008984004839487,
+      "learning_rate": 1.5462197614312926e-05,
+      "loss": 0.7048,
+      "step": 6465
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.238208929654018,
+      "learning_rate": 1.54607849381861e-05,
+      "loss": 0.6362,
+      "step": 6466
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.94097853826782,
+      "learning_rate": 1.5459372106757742e-05,
+      "loss": 0.7061,
+      "step": 6467
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.899308479426832,
+      "learning_rate": 1.5457959120068036e-05,
+      "loss": 0.6635,
+      "step": 6468
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.943075280311621,
+      "learning_rate": 1.5456545978157168e-05,
+      "loss": 0.6428,
+      "step": 6469
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.13668959818432,
+      "learning_rate": 1.545513268106532e-05,
+      "loss": 0.6835,
+      "step": 6470
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 57.91829494776299,
+      "learning_rate": 1.5453719228832692e-05,
+      "loss": 0.6759,
+      "step": 6471
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.221148865546716,
+      "learning_rate": 1.5452305621499483e-05,
+      "loss": 0.7349,
+      "step": 6472
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.936886695518524,
+      "learning_rate": 1.5450891859105884e-05,
+      "loss": 0.7431,
+      "step": 6473
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.768308261848315,
+      "learning_rate": 1.5449477941692114e-05,
+      "loss": 0.6369,
+      "step": 6474
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.73339342343991,
+      "learning_rate": 1.544806386929838e-05,
+      "loss": 0.7103,
+      "step": 6475
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.01587585681051,
+      "learning_rate": 1.5446649641964895e-05,
+      "loss": 0.634,
+      "step": 6476
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.830770775546783,
+      "learning_rate": 1.544523525973188e-05,
+      "loss": 0.7456,
+      "step": 6477
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.09747496837828,
+      "learning_rate": 1.544382072263956e-05,
+      "loss": 0.6198,
+      "step": 6478
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.985267152490426,
+      "learning_rate": 1.5442406030728166e-05,
+      "loss": 0.6579,
+      "step": 6479
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.73792426638232,
+      "learning_rate": 1.5440991184037924e-05,
+      "loss": 0.6999,
+      "step": 6480
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.700077622907656,
+      "learning_rate": 1.5439576182609077e-05,
+      "loss": 0.7502,
+      "step": 6481
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 22.47298910821798,
+      "learning_rate": 1.5438161026481866e-05,
+      "loss": 0.6358,
+      "step": 6482
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 26.84028716228371,
+      "learning_rate": 1.5436745715696535e-05,
+      "loss": 0.744,
+      "step": 6483
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 31.05960894616586,
+      "learning_rate": 1.543533025029334e-05,
+      "loss": 0.7034,
+      "step": 6484
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.039469146826423,
+      "learning_rate": 1.5433914630312527e-05,
+      "loss": 0.6496,
+      "step": 6485
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.360632751056993,
+      "learning_rate": 1.5432498855794364e-05,
+      "loss": 0.6024,
+      "step": 6486
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.719325236245606,
+      "learning_rate": 1.543108292677911e-05,
+      "loss": 0.6599,
+      "step": 6487
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.95426578067958,
+      "learning_rate": 1.5429666843307035e-05,
+      "loss": 0.6399,
+      "step": 6488
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.729858326264344,
+      "learning_rate": 1.5428250605418414e-05,
+      "loss": 0.6643,
+      "step": 6489
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.274789936001337,
+      "learning_rate": 1.542683421315352e-05,
+      "loss": 0.6547,
+      "step": 6490
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.05461939242172,
+      "learning_rate": 1.5425417666552635e-05,
+      "loss": 0.6833,
+      "step": 6491
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 19.758131550316673,
+      "learning_rate": 1.5424000965656042e-05,
+      "loss": 0.707,
+      "step": 6492
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.878210758802147,
+      "learning_rate": 1.542258411050404e-05,
+      "loss": 0.6852,
+      "step": 6493
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.76383856789012,
+      "learning_rate": 1.5421167101136917e-05,
+      "loss": 0.6823,
+      "step": 6494
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 21.582764347105275,
+      "learning_rate": 1.5419749937594967e-05,
+      "loss": 0.7516,
+      "step": 6495
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.840335613120832,
+      "learning_rate": 1.5418332619918507e-05,
+      "loss": 0.6507,
+      "step": 6496
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.363139546935002,
+      "learning_rate": 1.541691514814783e-05,
+      "loss": 0.6675,
+      "step": 6497
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.466826583583153,
+      "learning_rate": 1.541549752232326e-05,
+      "loss": 0.7122,
+      "step": 6498
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.442843736145985,
+      "learning_rate": 1.541407974248511e-05,
+      "loss": 0.6983,
+      "step": 6499
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.599299483498584,
+      "learning_rate": 1.5412661808673694e-05,
+      "loss": 0.6219,
+      "step": 6500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.347073533096834,
+      "learning_rate": 1.5411243720929342e-05,
+      "loss": 0.6531,
+      "step": 6501
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.872325090621768,
+      "learning_rate": 1.5409825479292388e-05,
+      "loss": 0.6819,
+      "step": 6502
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.181120917833676,
+      "learning_rate": 1.5408407083803162e-05,
+      "loss": 0.6597,
+      "step": 6503
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.683661268486603,
+      "learning_rate": 1.5406988534502002e-05,
+      "loss": 0.6862,
+      "step": 6504
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 28.451001613869806,
+      "learning_rate": 1.5405569831429247e-05,
+      "loss": 0.6581,
+      "step": 6505
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 23.754498773432474,
+      "learning_rate": 1.5404150974625254e-05,
+      "loss": 0.7414,
+      "step": 6506
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.520233758083858,
+      "learning_rate": 1.5402731964130365e-05,
+      "loss": 0.6471,
+      "step": 6507
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.36371344976559,
+      "learning_rate": 1.5401312799984943e-05,
+      "loss": 0.5999,
+      "step": 6508
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.94754732621128,
+      "learning_rate": 1.539989348222934e-05,
+      "loss": 0.6978,
+      "step": 6509
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.382679134675595,
+      "learning_rate": 1.5398474010903927e-05,
+      "loss": 0.7457,
+      "step": 6510
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.69835778231014,
+      "learning_rate": 1.5397054386049072e-05,
+      "loss": 0.6914,
+      "step": 6511
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.07778185244271,
+      "learning_rate": 1.5395634607705145e-05,
+      "loss": 0.6611,
+      "step": 6512
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.33259452530617,
+      "learning_rate": 1.539421467591253e-05,
+      "loss": 0.7345,
+      "step": 6513
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.867619554428977,
+      "learning_rate": 1.5392794590711605e-05,
+      "loss": 0.6659,
+      "step": 6514
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.37208631571248,
+      "learning_rate": 1.5391374352142752e-05,
+      "loss": 0.6822,
+      "step": 6515
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.438450565047656,
+      "learning_rate": 1.538995396024637e-05,
+      "loss": 0.714,
+      "step": 6516
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 27.06842814846263,
+      "learning_rate": 1.5388533415062848e-05,
+      "loss": 0.5554,
+      "step": 6517
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 27.561295686305304,
+      "learning_rate": 1.5387112716632594e-05,
+      "loss": 0.6792,
+      "step": 6518
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.8032830643274,
+      "learning_rate": 1.5385691864995998e-05,
+      "loss": 0.6521,
+      "step": 6519
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.67270866080271,
+      "learning_rate": 1.5384270860193477e-05,
+      "loss": 0.6598,
+      "step": 6520
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.561964563900505,
+      "learning_rate": 1.5382849702265447e-05,
+      "loss": 0.6752,
+      "step": 6521
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 28.522186184979937,
+      "learning_rate": 1.538142839125232e-05,
+      "loss": 0.7348,
+      "step": 6522
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 24.088772860199423,
+      "learning_rate": 1.538000692719451e-05,
+      "loss": 0.7279,
+      "step": 6523
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.539333245538113,
+      "learning_rate": 1.5378585310132458e-05,
+      "loss": 0.7308,
+      "step": 6524
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.89188983239716,
+      "learning_rate": 1.5377163540106582e-05,
+      "loss": 0.6546,
+      "step": 6525
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.668147509891359,
+      "learning_rate": 1.5375741617157324e-05,
+      "loss": 0.6505,
+      "step": 6526
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.063535418279532,
+      "learning_rate": 1.5374319541325114e-05,
+      "loss": 0.7976,
+      "step": 6527
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.380752379373973,
+      "learning_rate": 1.5372897312650406e-05,
+      "loss": 0.7622,
+      "step": 6528
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 26.530187130588722,
+      "learning_rate": 1.5371474931173638e-05,
+      "loss": 0.7456,
+      "step": 6529
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.945712580293577,
+      "learning_rate": 1.5370052396935268e-05,
+      "loss": 0.7097,
+      "step": 6530
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.622305601523422,
+      "learning_rate": 1.5368629709975745e-05,
+      "loss": 0.7191,
+      "step": 6531
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.725128365690392,
+      "learning_rate": 1.5367206870335536e-05,
+      "loss": 0.6044,
+      "step": 6532
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.28518469812207,
+      "learning_rate": 1.5365783878055103e-05,
+      "loss": 0.6909,
+      "step": 6533
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 22.472772158677415,
+      "learning_rate": 1.5364360733174916e-05,
+      "loss": 0.6582,
+      "step": 6534
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.542231148209925,
+      "learning_rate": 1.536293743573545e-05,
+      "loss": 0.6751,
+      "step": 6535
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.47603248188668,
+      "learning_rate": 1.5361513985777175e-05,
+      "loss": 0.6323,
+      "step": 6536
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.942876716357123,
+      "learning_rate": 1.536009038334058e-05,
+      "loss": 0.6564,
+      "step": 6537
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.88624460991718,
+      "learning_rate": 1.5358666628466154e-05,
+      "loss": 0.6908,
+      "step": 6538
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.097432343778436,
+      "learning_rate": 1.535724272119438e-05,
+      "loss": 0.6468,
+      "step": 6539
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.533646284586915,
+      "learning_rate": 1.535581866156576e-05,
+      "loss": 0.6792,
+      "step": 6540
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.373033787711773,
+      "learning_rate": 1.535439444962079e-05,
+      "loss": 0.6745,
+      "step": 6541
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.30349281989993,
+      "learning_rate": 1.535297008539997e-05,
+      "loss": 0.6978,
+      "step": 6542
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.16978710001151,
+      "learning_rate": 1.5351545568943818e-05,
+      "loss": 0.7,
+      "step": 6543
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.654371946662845,
+      "learning_rate": 1.5350120900292833e-05,
+      "loss": 0.6959,
+      "step": 6544
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.19838647026938,
+      "learning_rate": 1.5348696079487547e-05,
+      "loss": 0.7448,
+      "step": 6545
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.045007251105677,
+      "learning_rate": 1.534727110656847e-05,
+      "loss": 0.6897,
+      "step": 6546
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 26.992515178849985,
+      "learning_rate": 1.534584598157613e-05,
+      "loss": 0.7007,
+      "step": 6547
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.780094752167326,
+      "learning_rate": 1.5344420704551058e-05,
+      "loss": 0.647,
+      "step": 6548
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.98574558852048,
+      "learning_rate": 1.534299527553379e-05,
+      "loss": 0.6448,
+      "step": 6549
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 22.06125811020938,
+      "learning_rate": 1.534156969456486e-05,
+      "loss": 0.683,
+      "step": 6550
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.182842330895035,
+      "learning_rate": 1.5340143961684813e-05,
+      "loss": 0.6714,
+      "step": 6551
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.8774928974933,
+      "learning_rate": 1.5338718076934195e-05,
+      "loss": 0.7379,
+      "step": 6552
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.287842793250592,
+      "learning_rate": 1.5337292040353555e-05,
+      "loss": 0.6541,
+      "step": 6553
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.490607593925503,
+      "learning_rate": 1.5335865851983456e-05,
+      "loss": 0.6495,
+      "step": 6554
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.359068395388427,
+      "learning_rate": 1.5334439511864453e-05,
+      "loss": 0.7372,
+      "step": 6555
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.46820770526866,
+      "learning_rate": 1.533301302003711e-05,
+      "loss": 0.6705,
+      "step": 6556
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.33734881215871,
+      "learning_rate": 1.5331586376541997e-05,
+      "loss": 0.7181,
+      "step": 6557
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 26.14072399439881,
+      "learning_rate": 1.5330159581419687e-05,
+      "loss": 0.8224,
+      "step": 6558
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 22.334981350149064,
+      "learning_rate": 1.532873263471075e-05,
+      "loss": 0.8401,
+      "step": 6559
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.26068576114692,
+      "learning_rate": 1.5327305536455786e-05,
+      "loss": 0.7529,
+      "step": 6560
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.45708392616138,
+      "learning_rate": 1.5325878286695362e-05,
+      "loss": 0.6644,
+      "step": 6561
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 19.169113866224432,
+      "learning_rate": 1.5324450885470078e-05,
+      "loss": 0.6645,
+      "step": 6562
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 26.22562730353031,
+      "learning_rate": 1.5323023332820517e-05,
+      "loss": 0.7402,
+      "step": 6563
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.552562716194432,
+      "learning_rate": 1.5321595628787297e-05,
+      "loss": 0.623,
+      "step": 6564
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.47646629420186,
+      "learning_rate": 1.5320167773411004e-05,
+      "loss": 0.6306,
+      "step": 6565
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.24444339180681,
+      "learning_rate": 1.5318739766732255e-05,
+      "loss": 0.6996,
+      "step": 6566
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.445902434090621,
+      "learning_rate": 1.5317311608791656e-05,
+      "loss": 0.5829,
+      "step": 6567
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.940875641665073,
+      "learning_rate": 1.5315883299629825e-05,
+      "loss": 0.6225,
+      "step": 6568
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.88560715845788,
+      "learning_rate": 1.531445483928738e-05,
+      "loss": 0.656,
+      "step": 6569
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.068383973428357,
+      "learning_rate": 1.531302622780496e-05,
+      "loss": 0.7427,
+      "step": 6570
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.72585627283115,
+      "learning_rate": 1.5311597465223173e-05,
+      "loss": 0.6016,
+      "step": 6571
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.24498861073896,
+      "learning_rate": 1.531016855158266e-05,
+      "loss": 0.6832,
+      "step": 6572
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.651745441165993,
+      "learning_rate": 1.5308739486924064e-05,
+      "loss": 0.7824,
+      "step": 6573
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 20.544739473709814,
+      "learning_rate": 1.530731027128802e-05,
+      "loss": 0.7702,
+      "step": 6574
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.67457569626726,
+      "learning_rate": 1.5305880904715177e-05,
+      "loss": 0.6796,
+      "step": 6575
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.914617544826108,
+      "learning_rate": 1.530445138724619e-05,
+      "loss": 0.6844,
+      "step": 6576
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.174804068189223,
+      "learning_rate": 1.53030217189217e-05,
+      "loss": 0.6634,
+      "step": 6577
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 24.172706655658924,
+      "learning_rate": 1.5301591899782376e-05,
+      "loss": 0.6934,
+      "step": 6578
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.762627190097165,
+      "learning_rate": 1.5300161929868886e-05,
+      "loss": 0.7323,
+      "step": 6579
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.243914724334267,
+      "learning_rate": 1.5298731809221886e-05,
+      "loss": 0.6541,
+      "step": 6580
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.242675904007758,
+      "learning_rate": 1.5297301537882055e-05,
+      "loss": 0.7242,
+      "step": 6581
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.2606390878493,
+      "learning_rate": 1.5295871115890066e-05,
+      "loss": 0.7013,
+      "step": 6582
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.768022064124573,
+      "learning_rate": 1.52944405432866e-05,
+      "loss": 0.7681,
+      "step": 6583
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 20.229685884432772,
+      "learning_rate": 1.5293009820112345e-05,
+      "loss": 0.7712,
+      "step": 6584
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.52838174973978,
+      "learning_rate": 1.5291578946407985e-05,
+      "loss": 0.7438,
+      "step": 6585
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 25.288551781476045,
+      "learning_rate": 1.5290147922214212e-05,
+      "loss": 0.736,
+      "step": 6586
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.289904916199976,
+      "learning_rate": 1.5288716747571735e-05,
+      "loss": 0.7038,
+      "step": 6587
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.69966445189918,
+      "learning_rate": 1.5287285422521238e-05,
+      "loss": 0.6714,
+      "step": 6588
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 24.50131665980954,
+      "learning_rate": 1.5285853947103446e-05,
+      "loss": 0.682,
+      "step": 6589
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.383778157224622,
+      "learning_rate": 1.5284422321359054e-05,
+      "loss": 0.6443,
+      "step": 6590
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.396775495388287,
+      "learning_rate": 1.5282990545328782e-05,
+      "loss": 0.6823,
+      "step": 6591
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.96607641078348,
+      "learning_rate": 1.5281558619053353e-05,
+      "loss": 0.7675,
+      "step": 6592
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.713466718998358,
+      "learning_rate": 1.5280126542573484e-05,
+      "loss": 0.7566,
+      "step": 6593
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.48877243942122,
+      "learning_rate": 1.5278694315929906e-05,
+      "loss": 0.816,
+      "step": 6594
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.04490309713365,
+      "learning_rate": 1.527726193916335e-05,
+      "loss": 0.6772,
+      "step": 6595
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.346228167348286,
+      "learning_rate": 1.5275829412314547e-05,
+      "loss": 0.6375,
+      "step": 6596
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.052906827864728,
+      "learning_rate": 1.5274396735424244e-05,
+      "loss": 0.664,
+      "step": 6597
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.91804025692924,
+      "learning_rate": 1.5272963908533184e-05,
+      "loss": 0.6322,
+      "step": 6598
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.815045410789473,
+      "learning_rate": 1.5271530931682116e-05,
+      "loss": 0.7188,
+      "step": 6599
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.794946515808697,
+      "learning_rate": 1.5270097804911794e-05,
+      "loss": 0.7057,
+      "step": 6600
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.807764284614507,
+      "learning_rate": 1.526866452826297e-05,
+      "loss": 0.7643,
+      "step": 6601
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.98111661753733,
+      "learning_rate": 1.526723110177641e-05,
+      "loss": 0.6432,
+      "step": 6602
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.439288304104851,
+      "learning_rate": 1.5265797525492878e-05,
+      "loss": 0.7841,
+      "step": 6603
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.356104020241695,
+      "learning_rate": 1.526436379945315e-05,
+      "loss": 0.6421,
+      "step": 6604
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.35263202201593,
+      "learning_rate": 1.5262929923697986e-05,
+      "loss": 0.739,
+      "step": 6605
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.834669111293513,
+      "learning_rate": 1.526149589826818e-05,
+      "loss": 0.6509,
+      "step": 6606
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.840777058710355,
+      "learning_rate": 1.5260061723204506e-05,
+      "loss": 0.7117,
+      "step": 6607
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.481618551021654,
+      "learning_rate": 1.5258627398547754e-05,
+      "loss": 0.6568,
+      "step": 6608
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.80696078053773,
+      "learning_rate": 1.5257192924338715e-05,
+      "loss": 0.6788,
+      "step": 6609
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.284141943746125,
+      "learning_rate": 1.5255758300618184e-05,
+      "loss": 0.7012,
+      "step": 6610
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.337081794919776,
+      "learning_rate": 1.5254323527426964e-05,
+      "loss": 0.6966,
+      "step": 6611
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.714745961228799,
+      "learning_rate": 1.5252888604805853e-05,
+      "loss": 0.586,
+      "step": 6612
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 25.742324016596896,
+      "learning_rate": 1.5251453532795663e-05,
+      "loss": 0.7441,
+      "step": 6613
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 26.83939254152738,
+      "learning_rate": 1.5250018311437212e-05,
+      "loss": 0.7659,
+      "step": 6614
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.680834414274619,
+      "learning_rate": 1.5248582940771306e-05,
+      "loss": 0.7014,
+      "step": 6615
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.952776197281985,
+      "learning_rate": 1.5247147420838774e-05,
+      "loss": 0.7046,
+      "step": 6616
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.594671322979067,
+      "learning_rate": 1.5245711751680438e-05,
+      "loss": 0.7383,
+      "step": 6617
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.595836471358915,
+      "learning_rate": 1.524427593333713e-05,
+      "loss": 0.6726,
+      "step": 6618
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.10038194654784,
+      "learning_rate": 1.524283996584968e-05,
+      "loss": 0.6717,
+      "step": 6619
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.71986501598634,
+      "learning_rate": 1.524140384925893e-05,
+      "loss": 0.681,
+      "step": 6620
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.48799162203246,
+      "learning_rate": 1.5239967583605719e-05,
+      "loss": 0.6372,
+      "step": 6621
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 20.083716389523538,
+      "learning_rate": 1.5238531168930894e-05,
+      "loss": 0.6297,
+      "step": 6622
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.48344413153892,
+      "learning_rate": 1.5237094605275311e-05,
+      "loss": 0.6848,
+      "step": 6623
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.151736963888386,
+      "learning_rate": 1.5235657892679818e-05,
+      "loss": 0.7363,
+      "step": 6624
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.90782770211155,
+      "learning_rate": 1.523422103118528e-05,
+      "loss": 0.6543,
+      "step": 6625
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 28.46399421910907,
+      "learning_rate": 1.5232784020832556e-05,
+      "loss": 0.6591,
+      "step": 6626
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 23.43643280905722,
+      "learning_rate": 1.5231346861662518e-05,
+      "loss": 0.7411,
+      "step": 6627
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.530727861752872,
+      "learning_rate": 1.5229909553716032e-05,
+      "loss": 0.6726,
+      "step": 6628
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.942418488788668,
+      "learning_rate": 1.5228472097033979e-05,
+      "loss": 0.7344,
+      "step": 6629
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.667152133851134,
+      "learning_rate": 1.5227034491657235e-05,
+      "loss": 0.6592,
+      "step": 6630
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 22.635822310648,
+      "learning_rate": 1.5225596737626695e-05,
+      "loss": 0.7934,
+      "step": 6631
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.207611311074004,
+      "learning_rate": 1.5224158834983234e-05,
+      "loss": 0.6699,
+      "step": 6632
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 30.850181217816736,
+      "learning_rate": 1.5222720783767755e-05,
+      "loss": 0.6371,
+      "step": 6633
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 22.081180126253468,
+      "learning_rate": 1.522128258402115e-05,
+      "loss": 0.6717,
+      "step": 6634
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.79146770108563,
+      "learning_rate": 1.5219844235784326e-05,
+      "loss": 0.6334,
+      "step": 6635
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.525451205576074,
+      "learning_rate": 1.5218405739098183e-05,
+      "loss": 0.6134,
+      "step": 6636
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.671374621780412,
+      "learning_rate": 1.5216967094003633e-05,
+      "loss": 0.6702,
+      "step": 6637
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.72162200427294,
+      "learning_rate": 1.5215528300541593e-05,
+      "loss": 0.6383,
+      "step": 6638
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.024548302186307,
+      "learning_rate": 1.5214089358752979e-05,
+      "loss": 0.7096,
+      "step": 6639
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.861294975777007,
+      "learning_rate": 1.5212650268678711e-05,
+      "loss": 0.7022,
+      "step": 6640
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.541978602912499,
+      "learning_rate": 1.5211211030359719e-05,
+      "loss": 0.6873,
+      "step": 6641
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.905069210285983,
+      "learning_rate": 1.5209771643836937e-05,
+      "loss": 0.7557,
+      "step": 6642
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.949176372184894,
+      "learning_rate": 1.5208332109151295e-05,
+      "loss": 0.7601,
+      "step": 6643
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.168979017672584,
+      "learning_rate": 1.5206892426343736e-05,
+      "loss": 0.6954,
+      "step": 6644
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.81713949689299,
+      "learning_rate": 1.5205452595455198e-05,
+      "loss": 0.6907,
+      "step": 6645
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.719257058760887,
+      "learning_rate": 1.5204012616526637e-05,
+      "loss": 0.6739,
+      "step": 6646
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.61770966904786,
+      "learning_rate": 1.5202572489598998e-05,
+      "loss": 0.713,
+      "step": 6647
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.00530202475908,
+      "learning_rate": 1.5201132214713249e-05,
+      "loss": 0.7099,
+      "step": 6648
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.75996954853697,
+      "learning_rate": 1.5199691791910333e-05,
+      "loss": 0.7586,
+      "step": 6649
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 21.906346931052607,
+      "learning_rate": 1.5198251221231232e-05,
+      "loss": 0.7403,
+      "step": 6650
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.89578159295,
+      "learning_rate": 1.51968105027169e-05,
+      "loss": 0.6511,
+      "step": 6651
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 25.635977239549483,
+      "learning_rate": 1.5195369636408322e-05,
+      "loss": 0.7877,
+      "step": 6652
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.404539385126125,
+      "learning_rate": 1.519392862234647e-05,
+      "loss": 0.5677,
+      "step": 6653
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.03915373203264,
+      "learning_rate": 1.5192487460572327e-05,
+      "loss": 0.7096,
+      "step": 6654
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.096133952151117,
+      "learning_rate": 1.5191046151126876e-05,
+      "loss": 0.7408,
+      "step": 6655
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.70834941359846,
+      "learning_rate": 1.5189604694051111e-05,
+      "loss": 0.6242,
+      "step": 6656
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.453598259372303,
+      "learning_rate": 1.5188163089386022e-05,
+      "loss": 0.6273,
+      "step": 6657
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.856765901077996,
+      "learning_rate": 1.5186721337172614e-05,
+      "loss": 0.6754,
+      "step": 6658
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.852426534897962,
+      "learning_rate": 1.5185279437451881e-05,
+      "loss": 0.7661,
+      "step": 6659
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.698674770647406,
+      "learning_rate": 1.5183837390264833e-05,
+      "loss": 0.6751,
+      "step": 6660
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 21.809745454791692,
+      "learning_rate": 1.5182395195652487e-05,
+      "loss": 0.6831,
+      "step": 6661
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.689005475298254,
+      "learning_rate": 1.518095285365585e-05,
+      "loss": 0.6706,
+      "step": 6662
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.43701011788531,
+      "learning_rate": 1.5179510364315945e-05,
+      "loss": 0.6119,
+      "step": 6663
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.094374598430834,
+      "learning_rate": 1.5178067727673797e-05,
+      "loss": 0.6518,
+      "step": 6664
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.6080674989379,
+      "learning_rate": 1.517662494377043e-05,
+      "loss": 0.7072,
+      "step": 6665
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 21.187186529997053,
+      "learning_rate": 1.5175182012646877e-05,
+      "loss": 0.6354,
+      "step": 6666
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.323068565102645,
+      "learning_rate": 1.5173738934344178e-05,
+      "loss": 0.59,
+      "step": 6667
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.293548632210413,
+      "learning_rate": 1.5172295708903368e-05,
+      "loss": 0.7842,
+      "step": 6668
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.640274083217776,
+      "learning_rate": 1.5170852336365494e-05,
+      "loss": 0.6901,
+      "step": 6669
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 28.419465718381375,
+      "learning_rate": 1.5169408816771605e-05,
+      "loss": 0.6439,
+      "step": 6670
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 24.974292597093342,
+      "learning_rate": 1.5167965150162753e-05,
+      "loss": 0.6539,
+      "step": 6671
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.167377202645568,
+      "learning_rate": 1.5166521336579996e-05,
+      "loss": 0.6445,
+      "step": 6672
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.12948753071187,
+      "learning_rate": 1.5165077376064394e-05,
+      "loss": 0.6076,
+      "step": 6673
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 25.008362679118434,
+      "learning_rate": 1.516363326865701e-05,
+      "loss": 0.7471,
+      "step": 6674
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.526367413997082,
+      "learning_rate": 1.5162189014398923e-05,
+      "loss": 0.7342,
+      "step": 6675
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 28.163684441387435,
+      "learning_rate": 1.5160744613331196e-05,
+      "loss": 0.6954,
+      "step": 6676
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 27.353512265996105,
+      "learning_rate": 1.5159300065494912e-05,
+      "loss": 0.8157,
+      "step": 6677
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.217737586967633,
+      "learning_rate": 1.515785537093115e-05,
+      "loss": 0.7461,
+      "step": 6678
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.31736374654659,
+      "learning_rate": 1.5156410529681e-05,
+      "loss": 0.73,
+      "step": 6679
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 28.388256885394128,
+      "learning_rate": 1.5154965541785554e-05,
+      "loss": 0.6695,
+      "step": 6680
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.050016300627473,
+      "learning_rate": 1.51535204072859e-05,
+      "loss": 0.6573,
+      "step": 6681
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.55720826704021,
+      "learning_rate": 1.515207512622314e-05,
+      "loss": 0.6634,
+      "step": 6682
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.29011207442339,
+      "learning_rate": 1.515062969863838e-05,
+      "loss": 0.712,
+      "step": 6683
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.737950136059283,
+      "learning_rate": 1.514918412457272e-05,
+      "loss": 0.7008,
+      "step": 6684
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.189511755562448,
+      "learning_rate": 1.5147738404067279e-05,
+      "loss": 0.6075,
+      "step": 6685
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.262714429181987,
+      "learning_rate": 1.5146292537163168e-05,
+      "loss": 0.6746,
+      "step": 6686
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.636403245099917,
+      "learning_rate": 1.5144846523901508e-05,
+      "loss": 0.7292,
+      "step": 6687
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.09836613456225,
+      "learning_rate": 1.5143400364323424e-05,
+      "loss": 0.6327,
+      "step": 6688
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.782418149998275,
+      "learning_rate": 1.5141954058470041e-05,
+      "loss": 0.6115,
+      "step": 6689
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 24.250980679970176,
+      "learning_rate": 1.5140507606382496e-05,
+      "loss": 0.7613,
+      "step": 6690
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.284831227840556,
+      "learning_rate": 1.5139061008101915e-05,
+      "loss": 0.6453,
+      "step": 6691
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 30.36856014956248,
+      "learning_rate": 1.5137614263669451e-05,
+      "loss": 0.7937,
+      "step": 6692
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.24527272262378,
+      "learning_rate": 1.5136167373126239e-05,
+      "loss": 0.7274,
+      "step": 6693
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.303322647975598,
+      "learning_rate": 1.5134720336513433e-05,
+      "loss": 0.6853,
+      "step": 6694
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.510953574668997,
+      "learning_rate": 1.5133273153872186e-05,
+      "loss": 0.6834,
+      "step": 6695
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.93447534633121,
+      "learning_rate": 1.5131825825243651e-05,
+      "loss": 0.6649,
+      "step": 6696
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.860722140994472,
+      "learning_rate": 1.5130378350668996e-05,
+      "loss": 0.724,
+      "step": 6697
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.81838713951837,
+      "learning_rate": 1.5128930730189377e-05,
+      "loss": 0.6582,
+      "step": 6698
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 19.702926904739805,
+      "learning_rate": 1.5127482963845968e-05,
+      "loss": 0.7186,
+      "step": 6699
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 29.46530149899022,
+      "learning_rate": 1.5126035051679946e-05,
+      "loss": 0.7247,
+      "step": 6700
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.876680611502888,
+      "learning_rate": 1.5124586993732483e-05,
+      "loss": 0.7157,
+      "step": 6701
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.079512483867564,
+      "learning_rate": 1.5123138790044767e-05,
+      "loss": 0.6618,
+      "step": 6702
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.463695534052768,
+      "learning_rate": 1.5121690440657977e-05,
+      "loss": 0.6404,
+      "step": 6703
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.689674616117216,
+      "learning_rate": 1.512024194561331e-05,
+      "loss": 0.7385,
+      "step": 6704
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.592732213822007,
+      "learning_rate": 1.5118793304951955e-05,
+      "loss": 0.6684,
+      "step": 6705
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.832450765138345,
+      "learning_rate": 1.5117344518715115e-05,
+      "loss": 0.6442,
+      "step": 6706
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 22.80305041043113,
+      "learning_rate": 1.511589558694399e-05,
+      "loss": 0.6963,
+      "step": 6707
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 19.56725129076971,
+      "learning_rate": 1.5114446509679783e-05,
+      "loss": 0.7202,
+      "step": 6708
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.534769864974145,
+      "learning_rate": 1.5112997286963715e-05,
+      "loss": 0.6659,
+      "step": 6709
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.26281817306663,
+      "learning_rate": 1.5111547918836994e-05,
+      "loss": 0.6833,
+      "step": 6710
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.506014359211708,
+      "learning_rate": 1.5110098405340838e-05,
+      "loss": 0.6883,
+      "step": 6711
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.400366255735799,
+      "learning_rate": 1.5108648746516475e-05,
+      "loss": 0.6027,
+      "step": 6712
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.308729778132559,
+      "learning_rate": 1.5107198942405128e-05,
+      "loss": 0.6443,
+      "step": 6713
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 22.632348640831825,
+      "learning_rate": 1.5105748993048032e-05,
+      "loss": 0.7034,
+      "step": 6714
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.452454173052796,
+      "learning_rate": 1.5104298898486423e-05,
+      "loss": 0.7327,
+      "step": 6715
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 24.4268050596143,
+      "learning_rate": 1.5102848658761535e-05,
+      "loss": 0.739,
+      "step": 6716
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.614295445019987,
+      "learning_rate": 1.5101398273914623e-05,
+      "loss": 0.6471,
+      "step": 6717
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 20.486895905852382,
+      "learning_rate": 1.509994774398692e-05,
+      "loss": 0.7768,
+      "step": 6718
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 22.19455573091933,
+      "learning_rate": 1.5098497069019693e-05,
+      "loss": 0.794,
+      "step": 6719
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.27445002489998,
+      "learning_rate": 1.5097046249054187e-05,
+      "loss": 0.7013,
+      "step": 6720
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.220557597984667,
+      "learning_rate": 1.5095595284131674e-05,
+      "loss": 0.7077,
+      "step": 6721
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.517174452293958,
+      "learning_rate": 1.5094144174293406e-05,
+      "loss": 0.6298,
+      "step": 6722
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.60737034958921,
+      "learning_rate": 1.509269291958066e-05,
+      "loss": 0.7088,
+      "step": 6723
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.646346818630947,
+      "learning_rate": 1.5091241520034705e-05,
+      "loss": 0.7184,
+      "step": 6724
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.632190975079098,
+      "learning_rate": 1.508978997569682e-05,
+      "loss": 0.6687,
+      "step": 6725
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.012759653616886,
+      "learning_rate": 1.5088338286608287e-05,
+      "loss": 0.7022,
+      "step": 6726
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.635143369369796,
+      "learning_rate": 1.508688645281039e-05,
+      "loss": 0.7386,
+      "step": 6727
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.433604095379572,
+      "learning_rate": 1.5085434474344416e-05,
+      "loss": 0.6502,
+      "step": 6728
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.862097299616963,
+      "learning_rate": 1.508398235125166e-05,
+      "loss": 0.6882,
+      "step": 6729
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 20.576157288620845,
+      "learning_rate": 1.5082530083573421e-05,
+      "loss": 0.7651,
+      "step": 6730
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 19.890640653962166,
+      "learning_rate": 1.5081077671350998e-05,
+      "loss": 0.7228,
+      "step": 6731
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.263088205884493,
+      "learning_rate": 1.5079625114625701e-05,
+      "loss": 0.6851,
+      "step": 6732
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.431106882462366,
+      "learning_rate": 1.5078172413438836e-05,
+      "loss": 0.6179,
+      "step": 6733
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.37782525191314,
+      "learning_rate": 1.507671956783172e-05,
+      "loss": 0.7089,
+      "step": 6734
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.936303321450204,
+      "learning_rate": 1.5075266577845668e-05,
+      "loss": 0.6976,
+      "step": 6735
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.63058520146432,
+      "learning_rate": 1.5073813443522005e-05,
+      "loss": 0.6687,
+      "step": 6736
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.870071077366125,
+      "learning_rate": 1.5072360164902052e-05,
+      "loss": 0.5894,
+      "step": 6737
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 26.251515197166363,
+      "learning_rate": 1.5070906742027145e-05,
+      "loss": 0.7214,
+      "step": 6738
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 32.10956064806358,
+      "learning_rate": 1.5069453174938618e-05,
+      "loss": 0.6366,
+      "step": 6739
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.493373581536158,
+      "learning_rate": 1.5067999463677807e-05,
+      "loss": 0.7083,
+      "step": 6740
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.7424982863581,
+      "learning_rate": 1.5066545608286056e-05,
+      "loss": 0.7417,
+      "step": 6741
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.184725924503475,
+      "learning_rate": 1.5065091608804712e-05,
+      "loss": 0.7666,
+      "step": 6742
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.12462423938081,
+      "learning_rate": 1.5063637465275127e-05,
+      "loss": 0.7064,
+      "step": 6743
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.929054207961638,
+      "learning_rate": 1.5062183177738652e-05,
+      "loss": 0.6309,
+      "step": 6744
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.467203612465383,
+      "learning_rate": 1.5060728746236648e-05,
+      "loss": 0.6572,
+      "step": 6745
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.7776663156548,
+      "learning_rate": 1.5059274170810483e-05,
+      "loss": 0.7138,
+      "step": 6746
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.73596376469182,
+      "learning_rate": 1.505781945150152e-05,
+      "loss": 0.6955,
+      "step": 6747
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.814254819269365,
+      "learning_rate": 1.5056364588351127e-05,
+      "loss": 0.6977,
+      "step": 6748
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.190434135687397,
+      "learning_rate": 1.5054909581400683e-05,
+      "loss": 0.6426,
+      "step": 6749
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 24.528145942382583,
+      "learning_rate": 1.505345443069157e-05,
+      "loss": 0.6864,
+      "step": 6750
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.28437851602784,
+      "learning_rate": 1.505199913626517e-05,
+      "loss": 0.6202,
+      "step": 6751
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.074780002212918,
+      "learning_rate": 1.5050543698162865e-05,
+      "loss": 0.6105,
+      "step": 6752
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.03098004109674,
+      "learning_rate": 1.5049088116426056e-05,
+      "loss": 0.64,
+      "step": 6753
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.2625640121314,
+      "learning_rate": 1.5047632391096132e-05,
+      "loss": 0.7421,
+      "step": 6754
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.196997694040093,
+      "learning_rate": 1.5046176522214497e-05,
+      "loss": 0.6098,
+      "step": 6755
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 21.544319961933265,
+      "learning_rate": 1.5044720509822553e-05,
+      "loss": 0.6651,
+      "step": 6756
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.107893537816315,
+      "learning_rate": 1.5043264353961711e-05,
+      "loss": 0.6649,
+      "step": 6757
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.688034368292268,
+      "learning_rate": 1.5041808054673379e-05,
+      "loss": 0.6494,
+      "step": 6758
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 27.510079136349486,
+      "learning_rate": 1.5040351611998975e-05,
+      "loss": 0.698,
+      "step": 6759
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 26.915014561283886,
+      "learning_rate": 1.5038895025979919e-05,
+      "loss": 0.7072,
+      "step": 6760
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.561983282825631,
+      "learning_rate": 1.503743829665764e-05,
+      "loss": 0.6981,
+      "step": 6761
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.339591041413275,
+      "learning_rate": 1.503598142407356e-05,
+      "loss": 0.5992,
+      "step": 6762
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.477260059333883,
+      "learning_rate": 1.5034524408269115e-05,
+      "loss": 0.6849,
+      "step": 6763
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 22.58687186312458,
+      "learning_rate": 1.503306724928574e-05,
+      "loss": 0.7011,
+      "step": 6764
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.31629383606017,
+      "learning_rate": 1.5031609947164876e-05,
+      "loss": 0.6922,
+      "step": 6765
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.86701573990697,
+      "learning_rate": 1.503015250194797e-05,
+      "loss": 0.6886,
+      "step": 6766
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.194127112590767,
+      "learning_rate": 1.5028694913676469e-05,
+      "loss": 0.6928,
+      "step": 6767
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 26.367849941550556,
+      "learning_rate": 1.5027237182391825e-05,
+      "loss": 0.7701,
+      "step": 6768
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.472990147525692,
+      "learning_rate": 1.5025779308135499e-05,
+      "loss": 0.792,
+      "step": 6769
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.092443508716261,
+      "learning_rate": 1.5024321290948948e-05,
+      "loss": 0.7054,
+      "step": 6770
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 20.347386482692883,
+      "learning_rate": 1.502286313087364e-05,
+      "loss": 0.6532,
+      "step": 6771
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.677010330138664,
+      "learning_rate": 1.5021404827951039e-05,
+      "loss": 0.6664,
+      "step": 6772
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.927849904431202,
+      "learning_rate": 1.5019946382222626e-05,
+      "loss": 0.6334,
+      "step": 6773
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.654999269114985,
+      "learning_rate": 1.5018487793729875e-05,
+      "loss": 0.6635,
+      "step": 6774
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 20.958678695046423,
+      "learning_rate": 1.5017029062514263e-05,
+      "loss": 0.6941,
+      "step": 6775
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.987532676438875,
+      "learning_rate": 1.5015570188617283e-05,
+      "loss": 0.7844,
+      "step": 6776
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 22.552739288482833,
+      "learning_rate": 1.5014111172080416e-05,
+      "loss": 0.604,
+      "step": 6777
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 25.16805239729693,
+      "learning_rate": 1.5012652012945169e-05,
+      "loss": 0.6479,
+      "step": 6778
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 32.77274945324569,
+      "learning_rate": 1.5011192711253021e-05,
+      "loss": 0.7088,
+      "step": 6779
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.730486535911156,
+      "learning_rate": 1.5009733267045492e-05,
+      "loss": 0.631,
+      "step": 6780
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.1626431667129,
+      "learning_rate": 1.5008273680364074e-05,
+      "loss": 0.7112,
+      "step": 6781
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.652303983635978,
+      "learning_rate": 1.5006813951250286e-05,
+      "loss": 0.7087,
+      "step": 6782
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 27.013477595308252,
+      "learning_rate": 1.5005354079745636e-05,
+      "loss": 0.8159,
+      "step": 6783
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.46465218051016,
+      "learning_rate": 1.5003894065891644e-05,
+      "loss": 0.7353,
+      "step": 6784
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 24.82916107560699,
+      "learning_rate": 1.5002433909729832e-05,
+      "loss": 0.6299,
+      "step": 6785
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.352828822555992,
+      "learning_rate": 1.5000973611301729e-05,
+      "loss": 0.8654,
+      "step": 6786
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 25.386211684563552,
+      "learning_rate": 1.4999513170648856e-05,
+      "loss": 0.6896,
+      "step": 6787
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 31.012217609550927,
+      "learning_rate": 1.4998052587812759e-05,
+      "loss": 0.6279,
+      "step": 6788
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 44.16336133377703,
+      "learning_rate": 1.4996591862834965e-05,
+      "loss": 0.756,
+      "step": 6789
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.678696744111086,
+      "learning_rate": 1.4995130995757029e-05,
+      "loss": 0.7094,
+      "step": 6790
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.709091299218962,
+      "learning_rate": 1.499366998662048e-05,
+      "loss": 0.6427,
+      "step": 6791
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.878654305948793,
+      "learning_rate": 1.4992208835466884e-05,
+      "loss": 0.725,
+      "step": 6792
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 32.343372325258606,
+      "learning_rate": 1.4990747542337787e-05,
+      "loss": 0.7414,
+      "step": 6793
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.317289096279897,
+      "learning_rate": 1.498928610727475e-05,
+      "loss": 0.6178,
+      "step": 6794
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.479834453502043,
+      "learning_rate": 1.4987824530319335e-05,
+      "loss": 0.7074,
+      "step": 6795
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 22.55327039355677,
+      "learning_rate": 1.4986362811513107e-05,
+      "loss": 0.6459,
+      "step": 6796
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 41.87650463475585,
+      "learning_rate": 1.4984900950897638e-05,
+      "loss": 0.7785,
+      "step": 6797
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.145241705308184,
+      "learning_rate": 1.4983438948514501e-05,
+      "loss": 0.7369,
+      "step": 6798
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 44.90480650013255,
+      "learning_rate": 1.4981976804405278e-05,
+      "loss": 0.76,
+      "step": 6799
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.24738458163355,
+      "learning_rate": 1.4980514518611546e-05,
+      "loss": 0.6368,
+      "step": 6800
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.602200210767613,
+      "learning_rate": 1.49790520911749e-05,
+      "loss": 0.7114,
+      "step": 6801
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 28.55989867119618,
+      "learning_rate": 1.497758952213692e-05,
+      "loss": 0.6931,
+      "step": 6802
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.88186985496182,
+      "learning_rate": 1.4976126811539207e-05,
+      "loss": 0.697,
+      "step": 6803
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 23.79902455808823,
+      "learning_rate": 1.4974663959423358e-05,
+      "loss": 0.6749,
+      "step": 6804
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.462700658944087,
+      "learning_rate": 1.497320096583098e-05,
+      "loss": 0.7276,
+      "step": 6805
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 20.465791524138826,
+      "learning_rate": 1.497173783080367e-05,
+      "loss": 0.7689,
+      "step": 6806
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.511313547679306,
+      "learning_rate": 1.497027455438305e-05,
+      "loss": 0.5923,
+      "step": 6807
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.123060524762046,
+      "learning_rate": 1.4968811136610724e-05,
+      "loss": 0.6836,
+      "step": 6808
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.695988198598952,
+      "learning_rate": 1.4967347577528319e-05,
+      "loss": 0.7611,
+      "step": 6809
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.851156695683098,
+      "learning_rate": 1.4965883877177454e-05,
+      "loss": 0.6637,
+      "step": 6810
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 27.71474204552171,
+      "learning_rate": 1.4964420035599756e-05,
+      "loss": 0.7271,
+      "step": 6811
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.816600197094832,
+      "learning_rate": 1.4962956052836858e-05,
+      "loss": 0.7371,
+      "step": 6812
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.377530578813502,
+      "learning_rate": 1.4961491928930394e-05,
+      "loss": 0.6727,
+      "step": 6813
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.31630213409514,
+      "learning_rate": 1.4960027663921996e-05,
+      "loss": 0.6912,
+      "step": 6814
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.73835271599576,
+      "learning_rate": 1.495856325785332e-05,
+      "loss": 0.7391,
+      "step": 6815
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 20.810605633427603,
+      "learning_rate": 1.4957098710766e-05,
+      "loss": 0.6262,
+      "step": 6816
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.876832552132404,
+      "learning_rate": 1.4955634022701695e-05,
+      "loss": 0.6713,
+      "step": 6817
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 24.96261159479596,
+      "learning_rate": 1.4954169193702058e-05,
+      "loss": 0.6871,
+      "step": 6818
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.196235310255712,
+      "learning_rate": 1.4952704223808745e-05,
+      "loss": 0.6898,
+      "step": 6819
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 29.31788666994457,
+      "learning_rate": 1.4951239113063425e-05,
+      "loss": 0.7327,
+      "step": 6820
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.302621169001725,
+      "learning_rate": 1.4949773861507754e-05,
+      "loss": 0.6682,
+      "step": 6821
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.85392665305955,
+      "learning_rate": 1.4948308469183419e-05,
+      "loss": 0.7063,
+      "step": 6822
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.082330922027204,
+      "learning_rate": 1.494684293613208e-05,
+      "loss": 0.666,
+      "step": 6823
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.04374310965196,
+      "learning_rate": 1.4945377262395425e-05,
+      "loss": 0.6601,
+      "step": 6824
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.389755900554096,
+      "learning_rate": 1.4943911448015129e-05,
+      "loss": 0.6056,
+      "step": 6825
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.100140596913102,
+      "learning_rate": 1.4942445493032887e-05,
+      "loss": 0.6693,
+      "step": 6826
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.478692802455363,
+      "learning_rate": 1.4940979397490386e-05,
+      "loss": 0.5297,
+      "step": 6827
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.680751222512853,
+      "learning_rate": 1.493951316142932e-05,
+      "loss": 0.5804,
+      "step": 6828
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.251738047120982,
+      "learning_rate": 1.493804678489139e-05,
+      "loss": 0.6372,
+      "step": 6829
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 28.469253560709763,
+      "learning_rate": 1.4936580267918299e-05,
+      "loss": 0.8205,
+      "step": 6830
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.202536620085596,
+      "learning_rate": 1.4935113610551751e-05,
+      "loss": 0.6718,
+      "step": 6831
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.015048666122139,
+      "learning_rate": 1.4933646812833461e-05,
+      "loss": 0.6796,
+      "step": 6832
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.696228174848482,
+      "learning_rate": 1.493217987480514e-05,
+      "loss": 0.7411,
+      "step": 6833
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.884923190534414,
+      "learning_rate": 1.4930712796508511e-05,
+      "loss": 0.6621,
+      "step": 6834
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 24.478785677170265,
+      "learning_rate": 1.4929245577985292e-05,
+      "loss": 0.7677,
+      "step": 6835
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.19751390149639,
+      "learning_rate": 1.4927778219277214e-05,
+      "loss": 0.6948,
+      "step": 6836
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.090736125151793,
+      "learning_rate": 1.4926310720426005e-05,
+      "loss": 0.6565,
+      "step": 6837
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.064623228708854,
+      "learning_rate": 1.4924843081473401e-05,
+      "loss": 0.7807,
+      "step": 6838
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 24.472623109604786,
+      "learning_rate": 1.492337530246114e-05,
+      "loss": 0.6843,
+      "step": 6839
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.504632696364414,
+      "learning_rate": 1.4921907383430964e-05,
+      "loss": 0.7186,
+      "step": 6840
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.362482820556192,
+      "learning_rate": 1.4920439324424627e-05,
+      "loss": 0.6502,
+      "step": 6841
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.316981648675004,
+      "learning_rate": 1.491897112548387e-05,
+      "loss": 0.6199,
+      "step": 6842
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.835712021789067,
+      "learning_rate": 1.4917502786650451e-05,
+      "loss": 0.7515,
+      "step": 6843
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.10075321578142,
+      "learning_rate": 1.4916034307966132e-05,
+      "loss": 0.7038,
+      "step": 6844
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.994877366130115,
+      "learning_rate": 1.4914565689472671e-05,
+      "loss": 0.7293,
+      "step": 6845
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.10547906115209,
+      "learning_rate": 1.4913096931211839e-05,
+      "loss": 0.6755,
+      "step": 6846
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.547108326881645,
+      "learning_rate": 1.4911628033225402e-05,
+      "loss": 0.7262,
+      "step": 6847
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.29024413453594,
+      "learning_rate": 1.4910158995555138e-05,
+      "loss": 0.6144,
+      "step": 6848
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.26053868031611,
+      "learning_rate": 1.4908689818242826e-05,
+      "loss": 0.7477,
+      "step": 6849
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.024619263793335,
+      "learning_rate": 1.4907220501330245e-05,
+      "loss": 0.6162,
+      "step": 6850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.205784198534877,
+      "learning_rate": 1.4905751044859185e-05,
+      "loss": 0.7409,
+      "step": 6851
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.162922534434607,
+      "learning_rate": 1.4904281448871431e-05,
+      "loss": 0.6259,
+      "step": 6852
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.590950670662059,
+      "learning_rate": 1.4902811713408784e-05,
+      "loss": 0.6334,
+      "step": 6853
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.42584840697611,
+      "learning_rate": 1.4901341838513044e-05,
+      "loss": 0.7229,
+      "step": 6854
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.918537038141352,
+      "learning_rate": 1.4899871824226004e-05,
+      "loss": 0.683,
+      "step": 6855
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.438816199764084,
+      "learning_rate": 1.4898401670589478e-05,
+      "loss": 0.6605,
+      "step": 6856
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.56973616255662,
+      "learning_rate": 1.4896931377645273e-05,
+      "loss": 0.6847,
+      "step": 6857
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.68732531901476,
+      "learning_rate": 1.4895460945435205e-05,
+      "loss": 0.6687,
+      "step": 6858
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.047093463976806,
+      "learning_rate": 1.489399037400109e-05,
+      "loss": 0.6741,
+      "step": 6859
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 24.990936668935863,
+      "learning_rate": 1.4892519663384755e-05,
+      "loss": 0.6475,
+      "step": 6860
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.909315085755992,
+      "learning_rate": 1.489104881362802e-05,
+      "loss": 0.6919,
+      "step": 6861
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.86212752560986,
+      "learning_rate": 1.4889577824772719e-05,
+      "loss": 0.6404,
+      "step": 6862
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.620987791412155,
+      "learning_rate": 1.4888106696860684e-05,
+      "loss": 0.6714,
+      "step": 6863
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.86707473784369,
+      "learning_rate": 1.4886635429933756e-05,
+      "loss": 0.6344,
+      "step": 6864
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.752094637922625,
+      "learning_rate": 1.4885164024033771e-05,
+      "loss": 0.7448,
+      "step": 6865
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.351449011788484,
+      "learning_rate": 1.4883692479202582e-05,
+      "loss": 0.6105,
+      "step": 6866
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 28.396978297255078,
+      "learning_rate": 1.4882220795482037e-05,
+      "loss": 0.7575,
+      "step": 6867
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.887766752180045,
+      "learning_rate": 1.4880748972913987e-05,
+      "loss": 0.7021,
+      "step": 6868
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.62241245670707,
+      "learning_rate": 1.4879277011540289e-05,
+      "loss": 0.698,
+      "step": 6869
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 27.94651450317409,
+      "learning_rate": 1.4877804911402809e-05,
+      "loss": 0.7187,
+      "step": 6870
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.232966576535382,
+      "learning_rate": 1.4876332672543412e-05,
+      "loss": 0.7409,
+      "step": 6871
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.66010890127837,
+      "learning_rate": 1.4874860295003965e-05,
+      "loss": 0.6974,
+      "step": 6872
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.804826238581002,
+      "learning_rate": 1.4873387778826344e-05,
+      "loss": 0.61,
+      "step": 6873
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.506525522659032,
+      "learning_rate": 1.4871915124052426e-05,
+      "loss": 0.5513,
+      "step": 6874
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 24.77126192268069,
+      "learning_rate": 1.4870442330724088e-05,
+      "loss": 0.6856,
+      "step": 6875
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.463418757623485,
+      "learning_rate": 1.4868969398883224e-05,
+      "loss": 0.6515,
+      "step": 6876
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 26.449551904890306,
+      "learning_rate": 1.4867496328571716e-05,
+      "loss": 0.6854,
+      "step": 6877
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.694815587233965,
+      "learning_rate": 1.4866023119831461e-05,
+      "loss": 0.7199,
+      "step": 6878
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.633412048002835,
+      "learning_rate": 1.4864549772704353e-05,
+      "loss": 0.7222,
+      "step": 6879
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.048754914532285,
+      "learning_rate": 1.4863076287232297e-05,
+      "loss": 0.6516,
+      "step": 6880
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.338989017140193,
+      "learning_rate": 1.4861602663457195e-05,
+      "loss": 0.702,
+      "step": 6881
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.753267535878702,
+      "learning_rate": 1.4860128901420958e-05,
+      "loss": 0.7323,
+      "step": 6882
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.328161696134313,
+      "learning_rate": 1.4858655001165498e-05,
+      "loss": 0.557,
+      "step": 6883
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.837909827298652,
+      "learning_rate": 1.4857180962732731e-05,
+      "loss": 0.7135,
+      "step": 6884
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.3281535883019,
+      "learning_rate": 1.4855706786164579e-05,
+      "loss": 0.7526,
+      "step": 6885
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 23.341651623031126,
+      "learning_rate": 1.4854232471502967e-05,
+      "loss": 0.6791,
+      "step": 6886
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.346725633794186,
+      "learning_rate": 1.4852758018789826e-05,
+      "loss": 0.6881,
+      "step": 6887
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.20036394845212,
+      "learning_rate": 1.4851283428067083e-05,
+      "loss": 0.7463,
+      "step": 6888
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.835894446335644,
+      "learning_rate": 1.4849808699376675e-05,
+      "loss": 0.6627,
+      "step": 6889
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.724042714175972,
+      "learning_rate": 1.4848333832760549e-05,
+      "loss": 0.6696,
+      "step": 6890
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.012475708687663,
+      "learning_rate": 1.4846858828260642e-05,
+      "loss": 0.6936,
+      "step": 6891
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.062443667954245,
+      "learning_rate": 1.4845383685918906e-05,
+      "loss": 0.7597,
+      "step": 6892
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.44012516103373,
+      "learning_rate": 1.4843908405777294e-05,
+      "loss": 0.7649,
+      "step": 6893
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.611921070592302,
+      "learning_rate": 1.4842432987877758e-05,
+      "loss": 0.716,
+      "step": 6894
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.432880388216796,
+      "learning_rate": 1.4840957432262265e-05,
+      "loss": 0.6406,
+      "step": 6895
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.203478788616906,
+      "learning_rate": 1.483948173897277e-05,
+      "loss": 0.7051,
+      "step": 6896
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.47160488268398,
+      "learning_rate": 1.4838005908051245e-05,
+      "loss": 0.6831,
+      "step": 6897
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.68274434016696,
+      "learning_rate": 1.4836529939539665e-05,
+      "loss": 0.6279,
+      "step": 6898
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 25.558568519814205,
+      "learning_rate": 1.4835053833480003e-05,
+      "loss": 0.6641,
+      "step": 6899
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.304881011551856,
+      "learning_rate": 1.4833577589914234e-05,
+      "loss": 0.6803,
+      "step": 6900
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.183987779783546,
+      "learning_rate": 1.4832101208884352e-05,
+      "loss": 0.6123,
+      "step": 6901
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.32523948960862,
+      "learning_rate": 1.4830624690432331e-05,
+      "loss": 0.6468,
+      "step": 6902
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 27.145288723905725,
+      "learning_rate": 1.4829148034600174e-05,
+      "loss": 0.6132,
+      "step": 6903
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 22.06759863853126,
+      "learning_rate": 1.4827671241429873e-05,
+      "loss": 0.6542,
+      "step": 6904
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.76076857723192,
+      "learning_rate": 1.4826194310963424e-05,
+      "loss": 0.7168,
+      "step": 6905
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.705553563992417,
+      "learning_rate": 1.4824717243242835e-05,
+      "loss": 0.6569,
+      "step": 6906
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.847961762781033,
+      "learning_rate": 1.4823240038310108e-05,
+      "loss": 0.6935,
+      "step": 6907
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.936835498837047,
+      "learning_rate": 1.4821762696207254e-05,
+      "loss": 0.5734,
+      "step": 6908
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.75644804650147,
+      "learning_rate": 1.4820285216976288e-05,
+      "loss": 0.7417,
+      "step": 6909
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.68685259125747,
+      "learning_rate": 1.4818807600659237e-05,
+      "loss": 0.6866,
+      "step": 6910
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 24.248879402485336,
+      "learning_rate": 1.4817329847298112e-05,
+      "loss": 0.6538,
+      "step": 6911
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.528503137326087,
+      "learning_rate": 1.4815851956934948e-05,
+      "loss": 0.6704,
+      "step": 6912
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.891502665721167,
+      "learning_rate": 1.4814373929611768e-05,
+      "loss": 0.7168,
+      "step": 6913
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.781617483807377,
+      "learning_rate": 1.4812895765370609e-05,
+      "loss": 0.7302,
+      "step": 6914
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.535920767769795,
+      "learning_rate": 1.4811417464253512e-05,
+      "loss": 0.6499,
+      "step": 6915
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 23.921355271128462,
+      "learning_rate": 1.4809939026302518e-05,
+      "loss": 0.6869,
+      "step": 6916
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.897900614003222,
+      "learning_rate": 1.4808460451559669e-05,
+      "loss": 0.6351,
+      "step": 6917
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.154011517809057,
+      "learning_rate": 1.4806981740067018e-05,
+      "loss": 0.588,
+      "step": 6918
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.81200540722553,
+      "learning_rate": 1.480550289186662e-05,
+      "loss": 0.644,
+      "step": 6919
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.645432092217323,
+      "learning_rate": 1.480402390700053e-05,
+      "loss": 0.7193,
+      "step": 6920
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.905216777564288,
+      "learning_rate": 1.4802544785510811e-05,
+      "loss": 0.7288,
+      "step": 6921
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.696237168603231,
+      "learning_rate": 1.4801065527439526e-05,
+      "loss": 0.5994,
+      "step": 6922
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.590824371024818,
+      "learning_rate": 1.4799586132828747e-05,
+      "loss": 0.6722,
+      "step": 6923
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 22.933085539106568,
+      "learning_rate": 1.4798106601720545e-05,
+      "loss": 0.6759,
+      "step": 6924
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.985911898940657,
+      "learning_rate": 1.4796626934157e-05,
+      "loss": 0.6826,
+      "step": 6925
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.291782400686994,
+      "learning_rate": 1.4795147130180192e-05,
+      "loss": 0.6581,
+      "step": 6926
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 23.520752775450195,
+      "learning_rate": 1.4793667189832201e-05,
+      "loss": 0.8795,
+      "step": 6927
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.867146188830034,
+      "learning_rate": 1.4792187113155121e-05,
+      "loss": 0.6364,
+      "step": 6928
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.228691108022982,
+      "learning_rate": 1.4790706900191043e-05,
+      "loss": 0.6169,
+      "step": 6929
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.889255036721124,
+      "learning_rate": 1.4789226550982062e-05,
+      "loss": 0.7175,
+      "step": 6930
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.9396156249684,
+      "learning_rate": 1.4787746065570282e-05,
+      "loss": 0.6739,
+      "step": 6931
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 23.950394583679003,
+      "learning_rate": 1.4786265443997805e-05,
+      "loss": 0.6497,
+      "step": 6932
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.86424577334854,
+      "learning_rate": 1.4784784686306737e-05,
+      "loss": 0.7775,
+      "step": 6933
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 25.279852644021776,
+      "learning_rate": 1.478330379253919e-05,
+      "loss": 0.7147,
+      "step": 6934
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.13845804282142,
+      "learning_rate": 1.4781822762737284e-05,
+      "loss": 0.6882,
+      "step": 6935
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.626746143671477,
+      "learning_rate": 1.4780341596943133e-05,
+      "loss": 0.6497,
+      "step": 6936
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.964296727852883,
+      "learning_rate": 1.4778860295198869e-05,
+      "loss": 0.644,
+      "step": 6937
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.62900263410205,
+      "learning_rate": 1.4777378857546606e-05,
+      "loss": 0.6984,
+      "step": 6938
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.706145554232766,
+      "learning_rate": 1.4775897284028491e-05,
+      "loss": 0.6706,
+      "step": 6939
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 23.973669696114257,
+      "learning_rate": 1.4774415574686645e-05,
+      "loss": 0.7294,
+      "step": 6940
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.739109649954834,
+      "learning_rate": 1.4772933729563215e-05,
+      "loss": 0.6569,
+      "step": 6941
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.057039554068108,
+      "learning_rate": 1.4771451748700344e-05,
+      "loss": 0.6426,
+      "step": 6942
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.33816397406472,
+      "learning_rate": 1.4769969632140173e-05,
+      "loss": 0.6911,
+      "step": 6943
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 24.74663158515733,
+      "learning_rate": 1.4768487379924857e-05,
+      "loss": 0.6781,
+      "step": 6944
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.80884483756154,
+      "learning_rate": 1.476700499209655e-05,
+      "loss": 0.7354,
+      "step": 6945
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.843543977182733,
+      "learning_rate": 1.4765522468697406e-05,
+      "loss": 0.7084,
+      "step": 6946
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.51182711738,
+      "learning_rate": 1.4764039809769598e-05,
+      "loss": 0.7018,
+      "step": 6947
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.914673627688813,
+      "learning_rate": 1.4762557015355278e-05,
+      "loss": 0.6176,
+      "step": 6948
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.854363101035872,
+      "learning_rate": 1.4761074085496625e-05,
+      "loss": 0.6629,
+      "step": 6949
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.88423216077385,
+      "learning_rate": 1.4759591020235812e-05,
+      "loss": 0.699,
+      "step": 6950
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 26.588928320100546,
+      "learning_rate": 1.4758107819615014e-05,
+      "loss": 0.6286,
+      "step": 6951
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.338798535155828,
+      "learning_rate": 1.4756624483676412e-05,
+      "loss": 0.6679,
+      "step": 6952
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.15223280847857,
+      "learning_rate": 1.475514101246219e-05,
+      "loss": 0.6892,
+      "step": 6953
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.254974697743698,
+      "learning_rate": 1.4753657406014545e-05,
+      "loss": 0.6922,
+      "step": 6954
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.536834734925447,
+      "learning_rate": 1.475217366437566e-05,
+      "loss": 0.6449,
+      "step": 6955
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.775998131666926,
+      "learning_rate": 1.4750689787587735e-05,
+      "loss": 0.6154,
+      "step": 6956
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.990829974346735,
+      "learning_rate": 1.4749205775692974e-05,
+      "loss": 0.7237,
+      "step": 6957
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.495954436179797,
+      "learning_rate": 1.4747721628733578e-05,
+      "loss": 0.6836,
+      "step": 6958
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.21748560181008,
+      "learning_rate": 1.4746237346751757e-05,
+      "loss": 0.6481,
+      "step": 6959
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.270806569916324,
+      "learning_rate": 1.4744752929789723e-05,
+      "loss": 0.5648,
+      "step": 6960
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.705429537995972,
+      "learning_rate": 1.474326837788969e-05,
+      "loss": 0.6576,
+      "step": 6961
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.46108868467707,
+      "learning_rate": 1.4741783691093879e-05,
+      "loss": 0.6488,
+      "step": 6962
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.672031985733904,
+      "learning_rate": 1.4740298869444512e-05,
+      "loss": 0.6764,
+      "step": 6963
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.330488535854496,
+      "learning_rate": 1.4738813912983823e-05,
+      "loss": 0.6726,
+      "step": 6964
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.997999848576214,
+      "learning_rate": 1.4737328821754034e-05,
+      "loss": 0.7138,
+      "step": 6965
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.873263037092144,
+      "learning_rate": 1.4735843595797388e-05,
+      "loss": 0.656,
+      "step": 6966
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.484086140844646,
+      "learning_rate": 1.4734358235156117e-05,
+      "loss": 0.6322,
+      "step": 6967
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.733840856449074,
+      "learning_rate": 1.4732872739872469e-05,
+      "loss": 0.6889,
+      "step": 6968
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.444808700051054,
+      "learning_rate": 1.4731387109988689e-05,
+      "loss": 0.7014,
+      "step": 6969
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.971813462003162,
+      "learning_rate": 1.4729901345547027e-05,
+      "loss": 0.6809,
+      "step": 6970
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.401088114612397,
+      "learning_rate": 1.4728415446589736e-05,
+      "loss": 0.6512,
+      "step": 6971
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.771706887609707,
+      "learning_rate": 1.4726929413159077e-05,
+      "loss": 0.6208,
+      "step": 6972
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 25.89304416188458,
+      "learning_rate": 1.472544324529731e-05,
+      "loss": 0.7429,
+      "step": 6973
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 22.627261270755298,
+      "learning_rate": 1.4723956943046705e-05,
+      "loss": 0.7454,
+      "step": 6974
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.86831949760473,
+      "learning_rate": 1.4722470506449525e-05,
+      "loss": 0.6715,
+      "step": 6975
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.250413323455156,
+      "learning_rate": 1.4720983935548048e-05,
+      "loss": 0.7234,
+      "step": 6976
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.19040646253769,
+      "learning_rate": 1.4719497230384548e-05,
+      "loss": 0.6732,
+      "step": 6977
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.156288867094405,
+      "learning_rate": 1.4718010391001306e-05,
+      "loss": 0.6864,
+      "step": 6978
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.736330097090796,
+      "learning_rate": 1.4716523417440613e-05,
+      "loss": 0.7096,
+      "step": 6979
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.31613064249538,
+      "learning_rate": 1.4715036309744749e-05,
+      "loss": 0.6621,
+      "step": 6980
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.727156947986735,
+      "learning_rate": 1.4713549067956011e-05,
+      "loss": 0.7025,
+      "step": 6981
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.937049214648837,
+      "learning_rate": 1.4712061692116697e-05,
+      "loss": 0.6454,
+      "step": 6982
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.215261549031933,
+      "learning_rate": 1.4710574182269103e-05,
+      "loss": 0.6641,
+      "step": 6983
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.26781923507785,
+      "learning_rate": 1.4709086538455537e-05,
+      "loss": 0.5997,
+      "step": 6984
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 28.266171831774347,
+      "learning_rate": 1.4707598760718305e-05,
+      "loss": 0.6695,
+      "step": 6985
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 25.94063095590854,
+      "learning_rate": 1.4706110849099715e-05,
+      "loss": 0.8163,
+      "step": 6986
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.96683123747085,
+      "learning_rate": 1.4704622803642089e-05,
+      "loss": 0.7017,
+      "step": 6987
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 25.847383674981895,
+      "learning_rate": 1.470313462438774e-05,
+      "loss": 0.6739,
+      "step": 6988
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.503805068844486,
+      "learning_rate": 1.4701646311378994e-05,
+      "loss": 0.7649,
+      "step": 6989
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.18007624496995,
+      "learning_rate": 1.4700157864658176e-05,
+      "loss": 0.6699,
+      "step": 6990
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.102198666801241,
+      "learning_rate": 1.4698669284267619e-05,
+      "loss": 0.613,
+      "step": 6991
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.507259196056836,
+      "learning_rate": 1.4697180570249655e-05,
+      "loss": 0.6672,
+      "step": 6992
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.20135285938748,
+      "learning_rate": 1.4695691722646626e-05,
+      "loss": 0.6565,
+      "step": 6993
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.399815078758454,
+      "learning_rate": 1.4694202741500871e-05,
+      "loss": 0.7691,
+      "step": 6994
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.254121472892848,
+      "learning_rate": 1.4692713626854735e-05,
+      "loss": 0.6645,
+      "step": 6995
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.544987441451005,
+      "learning_rate": 1.4691224378750569e-05,
+      "loss": 0.5736,
+      "step": 6996
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.133386995122116,
+      "learning_rate": 1.4689734997230724e-05,
+      "loss": 0.6855,
+      "step": 6997
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.569562101305245,
+      "learning_rate": 1.4688245482337565e-05,
+      "loss": 0.7389,
+      "step": 6998
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.67870523712877,
+      "learning_rate": 1.468675583411344e-05,
+      "loss": 0.7144,
+      "step": 6999
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.019001725156665,
+      "learning_rate": 1.4685266052600724e-05,
+      "loss": 0.6206,
+      "step": 7000
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.342001245159917,
+      "learning_rate": 1.468377613784178e-05,
+      "loss": 0.6379,
+      "step": 7001
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.443360173874197,
+      "learning_rate": 1.4682286089878985e-05,
+      "loss": 0.7224,
+      "step": 7002
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 23.846929963695523,
+      "learning_rate": 1.4680795908754713e-05,
+      "loss": 0.703,
+      "step": 7003
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 24.87026691275611,
+      "learning_rate": 1.4679305594511344e-05,
+      "loss": 0.7048,
+      "step": 7004
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.959445866404707,
+      "learning_rate": 1.4677815147191256e-05,
+      "loss": 0.7042,
+      "step": 7005
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.11895754842285,
+      "learning_rate": 1.4676324566836848e-05,
+      "loss": 0.7079,
+      "step": 7006
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.83222527330517,
+      "learning_rate": 1.4674833853490502e-05,
+      "loss": 0.7459,
+      "step": 7007
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.030453288455655,
+      "learning_rate": 1.4673343007194618e-05,
+      "loss": 0.6139,
+      "step": 7008
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.907334622091476,
+      "learning_rate": 1.4671852027991589e-05,
+      "loss": 0.6868,
+      "step": 7009
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 22.42533027097129,
+      "learning_rate": 1.4670360915923824e-05,
+      "loss": 0.6339,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.657633555780176,
+      "learning_rate": 1.4668869671033726e-05,
+      "loss": 0.6438,
+      "step": 7011
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.88928070322481,
+      "learning_rate": 1.4667378293363706e-05,
+      "loss": 0.7048,
+      "step": 7012
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.70095618885727,
+      "learning_rate": 1.4665886782956178e-05,
+      "loss": 0.61,
+      "step": 7013
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.453035742007238,
+      "learning_rate": 1.4664395139853558e-05,
+      "loss": 0.7035,
+      "step": 7014
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 26.709407210272033,
+      "learning_rate": 1.466290336409827e-05,
+      "loss": 0.6475,
+      "step": 7015
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 24.27223713843969,
+      "learning_rate": 1.4661411455732735e-05,
+      "loss": 0.7413,
+      "step": 7016
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.31140263843974,
+      "learning_rate": 1.4659919414799389e-05,
+      "loss": 0.6015,
+      "step": 7017
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.056088089260825,
+      "learning_rate": 1.465842724134066e-05,
+      "loss": 0.6879,
+      "step": 7018
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.38249931811291,
+      "learning_rate": 1.4656934935398983e-05,
+      "loss": 0.6906,
+      "step": 7019
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.207129877611006,
+      "learning_rate": 1.4655442497016802e-05,
+      "loss": 0.6343,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.646688022751317,
+      "learning_rate": 1.4653949926236562e-05,
+      "loss": 0.6299,
+      "step": 7021
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.525901557415224,
+      "learning_rate": 1.4652457223100704e-05,
+      "loss": 0.7107,
+      "step": 7022
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.2278939265862,
+      "learning_rate": 1.4650964387651692e-05,
+      "loss": 0.715,
+      "step": 7023
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.366730467816783,
+      "learning_rate": 1.4649471419931964e-05,
+      "loss": 0.6328,
+      "step": 7024
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.22701759975661,
+      "learning_rate": 1.4647978319983998e-05,
+      "loss": 0.6437,
+      "step": 7025
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 25.290032370833387,
+      "learning_rate": 1.4646485087850239e-05,
+      "loss": 0.7371,
+      "step": 7026
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.123669322687615,
+      "learning_rate": 1.4644991723573167e-05,
+      "loss": 0.7021,
+      "step": 7027
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.399931850758918,
+      "learning_rate": 1.4643498227195247e-05,
+      "loss": 0.7094,
+      "step": 7028
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.0638921235817,
+      "learning_rate": 1.4642004598758955e-05,
+      "loss": 0.6051,
+      "step": 7029
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.992809223377687,
+      "learning_rate": 1.4640510838306765e-05,
+      "loss": 0.6004,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.715305468849856,
+      "learning_rate": 1.4639016945881164e-05,
+      "loss": 0.7904,
+      "step": 7031
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 28.687614468349267,
+      "learning_rate": 1.4637522921524632e-05,
+      "loss": 0.6982,
+      "step": 7032
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.598117797439969,
+      "learning_rate": 1.4636028765279662e-05,
+      "loss": 0.5972,
+      "step": 7033
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.615879165286403,
+      "learning_rate": 1.4634534477188745e-05,
+      "loss": 0.6154,
+      "step": 7034
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.878544153149125,
+      "learning_rate": 1.4633040057294379e-05,
+      "loss": 0.7251,
+      "step": 7035
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.490215179963524,
+      "learning_rate": 1.4631545505639066e-05,
+      "loss": 0.638,
+      "step": 7036
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.998263878485687,
+      "learning_rate": 1.4630050822265306e-05,
+      "loss": 0.6917,
+      "step": 7037
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.251567147849286,
+      "learning_rate": 1.4628556007215612e-05,
+      "loss": 0.6591,
+      "step": 7038
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.090590785622393,
+      "learning_rate": 1.462706106053249e-05,
+      "loss": 0.6737,
+      "step": 7039
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.11842552024782,
+      "learning_rate": 1.4625565982258461e-05,
+      "loss": 0.7529,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.220165905724727,
+      "learning_rate": 1.4624070772436037e-05,
+      "loss": 0.6819,
+      "step": 7041
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 45.387024255695025,
+      "learning_rate": 1.4622575431107748e-05,
+      "loss": 0.7692,
+      "step": 7042
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.686625203750875,
+      "learning_rate": 1.4621079958316118e-05,
+      "loss": 0.6618,
+      "step": 7043
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.755306786830555,
+      "learning_rate": 1.4619584354103675e-05,
+      "loss": 0.6575,
+      "step": 7044
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 24.8441490561147,
+      "learning_rate": 1.4618088618512957e-05,
+      "loss": 0.7926,
+      "step": 7045
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.99545148704661,
+      "learning_rate": 1.4616592751586503e-05,
+      "loss": 0.6647,
+      "step": 7046
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.447742875621604,
+      "learning_rate": 1.4615096753366849e-05,
+      "loss": 0.715,
+      "step": 7047
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.451496408826745,
+      "learning_rate": 1.4613600623896542e-05,
+      "loss": 0.6422,
+      "step": 7048
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.301495811328188,
+      "learning_rate": 1.4612104363218132e-05,
+      "loss": 0.6394,
+      "step": 7049
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.492657231754652,
+      "learning_rate": 1.4610607971374175e-05,
+      "loss": 0.733,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.673587667932145,
+      "learning_rate": 1.460911144840722e-05,
+      "loss": 0.7125,
+      "step": 7051
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.09370370967656,
+      "learning_rate": 1.4607614794359836e-05,
+      "loss": 0.6866,
+      "step": 7052
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.41064252760542,
+      "learning_rate": 1.4606118009274582e-05,
+      "loss": 0.6593,
+      "step": 7053
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 24.503598446450553,
+      "learning_rate": 1.4604621093194024e-05,
+      "loss": 0.684,
+      "step": 7054
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.500413183706783,
+      "learning_rate": 1.4603124046160737e-05,
+      "loss": 0.7107,
+      "step": 7055
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.284920481997407,
+      "learning_rate": 1.4601626868217296e-05,
+      "loss": 0.6495,
+      "step": 7056
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.54281996045134,
+      "learning_rate": 1.4600129559406278e-05,
+      "loss": 0.7752,
+      "step": 7057
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.973435400765425,
+      "learning_rate": 1.4598632119770266e-05,
+      "loss": 0.7222,
+      "step": 7058
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 26.655250829102528,
+      "learning_rate": 1.4597134549351845e-05,
+      "loss": 0.6997,
+      "step": 7059
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 24.968223785852004,
+      "learning_rate": 1.459563684819361e-05,
+      "loss": 0.7215,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.77558051936856,
+      "learning_rate": 1.4594139016338149e-05,
+      "loss": 0.6302,
+      "step": 7061
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.65315032446035,
+      "learning_rate": 1.4592641053828063e-05,
+      "loss": 0.5942,
+      "step": 7062
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.807305845723146,
+      "learning_rate": 1.4591142960705951e-05,
+      "loss": 0.7034,
+      "step": 7063
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.018211518192775,
+      "learning_rate": 1.4589644737014418e-05,
+      "loss": 0.6423,
+      "step": 7064
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.051010367016783,
+      "learning_rate": 1.4588146382796075e-05,
+      "loss": 0.6319,
+      "step": 7065
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.951180637386592,
+      "learning_rate": 1.4586647898093534e-05,
+      "loss": 0.6474,
+      "step": 7066
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.7070851335649,
+      "learning_rate": 1.4585149282949409e-05,
+      "loss": 0.6891,
+      "step": 7067
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.384739329077902,
+      "learning_rate": 1.458365053740632e-05,
+      "loss": 0.6133,
+      "step": 7068
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.041241613600196,
+      "learning_rate": 1.4582151661506894e-05,
+      "loss": 0.5978,
+      "step": 7069
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.890526369782368,
+      "learning_rate": 1.458065265529375e-05,
+      "loss": 0.6614,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.048859252621583,
+      "learning_rate": 1.4579153518809527e-05,
+      "loss": 0.7275,
+      "step": 7071
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.468198100574117,
+      "learning_rate": 1.4577654252096856e-05,
+      "loss": 0.6838,
+      "step": 7072
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.25671740439634,
+      "learning_rate": 1.4576154855198377e-05,
+      "loss": 0.6564,
+      "step": 7073
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.427955231875117,
+      "learning_rate": 1.457465532815673e-05,
+      "loss": 0.7075,
+      "step": 7074
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 23.81805510441936,
+      "learning_rate": 1.4573155671014563e-05,
+      "loss": 0.729,
+      "step": 7075
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 20.555953099584958,
+      "learning_rate": 1.457165588381452e-05,
+      "loss": 0.7012,
+      "step": 7076
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.117698708104072,
+      "learning_rate": 1.4570155966599264e-05,
+      "loss": 0.6668,
+      "step": 7077
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.652709015139353,
+      "learning_rate": 1.456865591941144e-05,
+      "loss": 0.6553,
+      "step": 7078
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.639431213934756,
+      "learning_rate": 1.4567155742293717e-05,
+      "loss": 0.6589,
+      "step": 7079
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 28.755967781325303,
+      "learning_rate": 1.4565655435288756e-05,
+      "loss": 0.7819,
+      "step": 7080
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.085008097855225,
+      "learning_rate": 1.4564154998439225e-05,
+      "loss": 0.656,
+      "step": 7081
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.128384263745025,
+      "learning_rate": 1.4562654431787797e-05,
+      "loss": 0.6186,
+      "step": 7082
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.361402765257665,
+      "learning_rate": 1.4561153735377147e-05,
+      "loss": 0.6589,
+      "step": 7083
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.039083891368954,
+      "learning_rate": 1.455965290924995e-05,
+      "loss": 0.7627,
+      "step": 7084
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.872707496763248,
+      "learning_rate": 1.4558151953448891e-05,
+      "loss": 0.7493,
+      "step": 7085
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 20.48439826934913,
+      "learning_rate": 1.4556650868016661e-05,
+      "loss": 0.6464,
+      "step": 7086
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.80123666528361,
+      "learning_rate": 1.4555149652995942e-05,
+      "loss": 0.6105,
+      "step": 7087
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 20.48556201952367,
+      "learning_rate": 1.4553648308429433e-05,
+      "loss": 0.677,
+      "step": 7088
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.585760968648692,
+      "learning_rate": 1.455214683435983e-05,
+      "loss": 0.6422,
+      "step": 7089
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.374523158354936,
+      "learning_rate": 1.4550645230829837e-05,
+      "loss": 0.6592,
+      "step": 7090
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 23.031335417745673,
+      "learning_rate": 1.4549143497882152e-05,
+      "loss": 0.6826,
+      "step": 7091
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.136101080664933,
+      "learning_rate": 1.4547641635559488e-05,
+      "loss": 0.7095,
+      "step": 7092
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 29.673752854120632,
+      "learning_rate": 1.4546139643904553e-05,
+      "loss": 0.7743,
+      "step": 7093
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 28.004237428043275,
+      "learning_rate": 1.4544637522960072e-05,
+      "loss": 0.6392,
+      "step": 7094
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.403888674916924,
+      "learning_rate": 1.4543135272768753e-05,
+      "loss": 0.7561,
+      "step": 7095
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 22.2972724280328,
+      "learning_rate": 1.454163289337333e-05,
+      "loss": 0.7921,
+      "step": 7096
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.257958892697793,
+      "learning_rate": 1.454013038481652e-05,
+      "loss": 0.6995,
+      "step": 7097
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.417911662751203,
+      "learning_rate": 1.4538627747141058e-05,
+      "loss": 0.665,
+      "step": 7098
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 25.90885183627213,
+      "learning_rate": 1.4537124980389678e-05,
+      "loss": 0.7621,
+      "step": 7099
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.131756108475784,
+      "learning_rate": 1.4535622084605119e-05,
+      "loss": 0.6533,
+      "step": 7100
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 28.440700924110484,
+      "learning_rate": 1.453411905983012e-05,
+      "loss": 0.6031,
+      "step": 7101
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.446422980777664,
+      "learning_rate": 1.453261590610743e-05,
+      "loss": 0.6583,
+      "step": 7102
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.393097887192482,
+      "learning_rate": 1.453111262347979e-05,
+      "loss": 0.6708,
+      "step": 7103
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 25.235441142846927,
+      "learning_rate": 1.4529609211989962e-05,
+      "loss": 0.6485,
+      "step": 7104
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.911867624274608,
+      "learning_rate": 1.4528105671680697e-05,
+      "loss": 0.6828,
+      "step": 7105
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 25.534884267697496,
+      "learning_rate": 1.4526602002594755e-05,
+      "loss": 0.645,
+      "step": 7106
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.25805458823257,
+      "learning_rate": 1.45250982047749e-05,
+      "loss": 0.6715,
+      "step": 7107
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.48968898836983,
+      "learning_rate": 1.4523594278263899e-05,
+      "loss": 0.7343,
+      "step": 7108
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.322576181327083,
+      "learning_rate": 1.4522090223104523e-05,
+      "loss": 0.6218,
+      "step": 7109
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.337220602420146,
+      "learning_rate": 1.4520586039339545e-05,
+      "loss": 0.6086,
+      "step": 7110
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.37134331126699,
+      "learning_rate": 1.4519081727011747e-05,
+      "loss": 0.7534,
+      "step": 7111
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.238374376450317,
+      "learning_rate": 1.4517577286163906e-05,
+      "loss": 0.651,
+      "step": 7112
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.46718639999983,
+      "learning_rate": 1.4516072716838815e-05,
+      "loss": 0.6485,
+      "step": 7113
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 20.217345109327145,
+      "learning_rate": 1.4514568019079252e-05,
+      "loss": 0.8135,
+      "step": 7114
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.922151751322875,
+      "learning_rate": 1.4513063192928016e-05,
+      "loss": 0.6534,
+      "step": 7115
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.125813466611653,
+      "learning_rate": 1.4511558238427905e-05,
+      "loss": 0.6134,
+      "step": 7116
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.48041140043487,
+      "learning_rate": 1.4510053155621714e-05,
+      "loss": 0.6565,
+      "step": 7117
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 23.86290110105597,
+      "learning_rate": 1.4508547944552254e-05,
+      "loss": 0.6473,
+      "step": 7118
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 24.96522334530517,
+      "learning_rate": 1.4507042605262324e-05,
+      "loss": 0.6947,
+      "step": 7119
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.667638920649857,
+      "learning_rate": 1.450553713779474e-05,
+      "loss": 0.7727,
+      "step": 7120
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.191781196154576,
+      "learning_rate": 1.4504031542192319e-05,
+      "loss": 0.6742,
+      "step": 7121
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.86807658352979,
+      "learning_rate": 1.450252581849787e-05,
+      "loss": 0.6663,
+      "step": 7122
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 22.07577298831755,
+      "learning_rate": 1.4501019966754225e-05,
+      "loss": 0.6357,
+      "step": 7123
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 26.80208455819756,
+      "learning_rate": 1.4499513987004201e-05,
+      "loss": 0.7979,
+      "step": 7124
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.814780257309796,
+      "learning_rate": 1.4498007879290635e-05,
+      "loss": 0.7274,
+      "step": 7125
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.112242110632847,
+      "learning_rate": 1.4496501643656356e-05,
+      "loss": 0.6785,
+      "step": 7126
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.779309510510934,
+      "learning_rate": 1.4494995280144198e-05,
+      "loss": 0.5376,
+      "step": 7127
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.730295629502738,
+      "learning_rate": 1.4493488788797006e-05,
+      "loss": 0.6951,
+      "step": 7128
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.67425831866003,
+      "learning_rate": 1.449198216965762e-05,
+      "loss": 0.679,
+      "step": 7129
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 43.810620102977005,
+      "learning_rate": 1.449047542276889e-05,
+      "loss": 0.7325,
+      "step": 7130
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.433666699274976,
+      "learning_rate": 1.4488968548173662e-05,
+      "loss": 0.6103,
+      "step": 7131
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.24231067535535,
+      "learning_rate": 1.4487461545914801e-05,
+      "loss": 0.6429,
+      "step": 7132
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.28116405538859,
+      "learning_rate": 1.4485954416035152e-05,
+      "loss": 0.6813,
+      "step": 7133
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.072778834871823,
+      "learning_rate": 1.4484447158577588e-05,
+      "loss": 0.6368,
+      "step": 7134
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 25.47481399146278,
+      "learning_rate": 1.448293977358497e-05,
+      "loss": 0.6792,
+      "step": 7135
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.532519097950544,
+      "learning_rate": 1.4481432261100168e-05,
+      "loss": 0.6942,
+      "step": 7136
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.272208164300157,
+      "learning_rate": 1.4479924621166051e-05,
+      "loss": 0.6786,
+      "step": 7137
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.632029391016218,
+      "learning_rate": 1.4478416853825502e-05,
+      "loss": 0.6437,
+      "step": 7138
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.045931425652277,
+      "learning_rate": 1.4476908959121394e-05,
+      "loss": 0.7058,
+      "step": 7139
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.9929101364993,
+      "learning_rate": 1.4475400937096617e-05,
+      "loss": 0.628,
+      "step": 7140
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.128778170156263,
+      "learning_rate": 1.4473892787794053e-05,
+      "loss": 0.7265,
+      "step": 7141
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.684913205118107,
+      "learning_rate": 1.4472384511256597e-05,
+      "loss": 0.6601,
+      "step": 7142
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.101327858799248,
+      "learning_rate": 1.4470876107527141e-05,
+      "loss": 0.7012,
+      "step": 7143
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.549378449755283,
+      "learning_rate": 1.4469367576648582e-05,
+      "loss": 0.683,
+      "step": 7144
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.643840399672822,
+      "learning_rate": 1.4467858918663826e-05,
+      "loss": 0.652,
+      "step": 7145
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.099977081906715,
+      "learning_rate": 1.4466350133615776e-05,
+      "loss": 0.6943,
+      "step": 7146
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.526825632591546,
+      "learning_rate": 1.4464841221547339e-05,
+      "loss": 0.6407,
+      "step": 7147
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.440793436888754,
+      "learning_rate": 1.4463332182501431e-05,
+      "loss": 0.6954,
+      "step": 7148
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.649252492030804,
+      "learning_rate": 1.4461823016520967e-05,
+      "loss": 0.7201,
+      "step": 7149
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.52141929028158,
+      "learning_rate": 1.4460313723648866e-05,
+      "loss": 0.6271,
+      "step": 7150
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.022794139236988,
+      "learning_rate": 1.4458804303928053e-05,
+      "loss": 0.6861,
+      "step": 7151
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.31227642761135,
+      "learning_rate": 1.4457294757401452e-05,
+      "loss": 0.6074,
+      "step": 7152
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 30.454482830572918,
+      "learning_rate": 1.4455785084111997e-05,
+      "loss": 0.6284,
+      "step": 7153
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.638165825402233,
+      "learning_rate": 1.445427528410262e-05,
+      "loss": 0.688,
+      "step": 7154
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 24.754279296529734,
+      "learning_rate": 1.4452765357416261e-05,
+      "loss": 0.6578,
+      "step": 7155
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 30.252084088847063,
+      "learning_rate": 1.4451255304095857e-05,
+      "loss": 0.6368,
+      "step": 7156
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.34131684838313,
+      "learning_rate": 1.444974512418436e-05,
+      "loss": 0.6447,
+      "step": 7157
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.375665156126406,
+      "learning_rate": 1.444823481772471e-05,
+      "loss": 0.6622,
+      "step": 7158
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 25.45230349496516,
+      "learning_rate": 1.4446724384759869e-05,
+      "loss": 0.6687,
+      "step": 7159
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.27005710284545,
+      "learning_rate": 1.4445213825332784e-05,
+      "loss": 0.6446,
+      "step": 7160
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.183637495789984,
+      "learning_rate": 1.444370313948642e-05,
+      "loss": 0.5617,
+      "step": 7161
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.80311784112368,
+      "learning_rate": 1.4442192327263737e-05,
+      "loss": 0.5712,
+      "step": 7162
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.468068875926573,
+      "learning_rate": 1.4440681388707704e-05,
+      "loss": 0.6664,
+      "step": 7163
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.19284891834503,
+      "learning_rate": 1.4439170323861289e-05,
+      "loss": 0.7225,
+      "step": 7164
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 27.27110475766207,
+      "learning_rate": 1.443765913276747e-05,
+      "loss": 0.7459,
+      "step": 7165
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 23.593908107935167,
+      "learning_rate": 1.4436147815469217e-05,
+      "loss": 0.6545,
+      "step": 7166
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.296511303974057,
+      "learning_rate": 1.4434636372009516e-05,
+      "loss": 0.7165,
+      "step": 7167
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.056383111186385,
+      "learning_rate": 1.4433124802431353e-05,
+      "loss": 0.6576,
+      "step": 7168
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.743213363914144,
+      "learning_rate": 1.4431613106777712e-05,
+      "loss": 0.5647,
+      "step": 7169
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.870715013580075,
+      "learning_rate": 1.4430101285091587e-05,
+      "loss": 0.6001,
+      "step": 7170
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.314093600279405,
+      "learning_rate": 1.4428589337415975e-05,
+      "loss": 0.6606,
+      "step": 7171
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.44392831472725,
+      "learning_rate": 1.442707726379387e-05,
+      "loss": 0.6873,
+      "step": 7172
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.4437503236935,
+      "learning_rate": 1.4425565064268276e-05,
+      "loss": 0.6569,
+      "step": 7173
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 22.267867267873793,
+      "learning_rate": 1.4424052738882203e-05,
+      "loss": 0.6765,
+      "step": 7174
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.534561724241765,
+      "learning_rate": 1.442254028767866e-05,
+      "loss": 0.6394,
+      "step": 7175
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 24.281917991495554,
+      "learning_rate": 1.4421027710700655e-05,
+      "loss": 0.7031,
+      "step": 7176
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.474470189530354,
+      "learning_rate": 1.441951500799121e-05,
+      "loss": 0.7319,
+      "step": 7177
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.35123254428233,
+      "learning_rate": 1.4418002179593345e-05,
+      "loss": 0.64,
+      "step": 7178
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.71189170849506,
+      "learning_rate": 1.441648922555008e-05,
+      "loss": 0.6774,
+      "step": 7179
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.746596146425777,
+      "learning_rate": 1.4414976145904445e-05,
+      "loss": 0.5948,
+      "step": 7180
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.093170712107323,
+      "learning_rate": 1.441346294069947e-05,
+      "loss": 0.6454,
+      "step": 7181
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.990572226482872,
+      "learning_rate": 1.4411949609978196e-05,
+      "loss": 0.6853,
+      "step": 7182
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.34061084299404,
+      "learning_rate": 1.441043615378365e-05,
+      "loss": 0.6357,
+      "step": 7183
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.196777537404106,
+      "learning_rate": 1.4408922572158883e-05,
+      "loss": 0.5732,
+      "step": 7184
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.584729596707398,
+      "learning_rate": 1.4407408865146935e-05,
+      "loss": 0.6176,
+      "step": 7185
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.18308444134311,
+      "learning_rate": 1.4405895032790858e-05,
+      "loss": 0.6675,
+      "step": 7186
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.587270234188065,
+      "learning_rate": 1.4404381075133706e-05,
+      "loss": 0.6121,
+      "step": 7187
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.64533149878886,
+      "learning_rate": 1.4402866992218529e-05,
+      "loss": 0.6496,
+      "step": 7188
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 30.97312875118572,
+      "learning_rate": 1.4401352784088394e-05,
+      "loss": 0.7026,
+      "step": 7189
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 22.50748127962067,
+      "learning_rate": 1.4399838450786358e-05,
+      "loss": 0.6849,
+      "step": 7190
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.264482633260354,
+      "learning_rate": 1.439832399235549e-05,
+      "loss": 0.7039,
+      "step": 7191
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.890241899626659,
+      "learning_rate": 1.4396809408838862e-05,
+      "loss": 0.588,
+      "step": 7192
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 24.062036142584752,
+      "learning_rate": 1.4395294700279545e-05,
+      "loss": 0.5902,
+      "step": 7193
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 23.765563902378236,
+      "learning_rate": 1.439377986672062e-05,
+      "loss": 0.7383,
+      "step": 7194
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.300892123349808,
+      "learning_rate": 1.4392264908205165e-05,
+      "loss": 0.7094,
+      "step": 7195
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.3762345138238,
+      "learning_rate": 1.4390749824776264e-05,
+      "loss": 0.6302,
+      "step": 7196
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.207671099661752,
+      "learning_rate": 1.438923461647701e-05,
+      "loss": 0.7919,
+      "step": 7197
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.203450093611192,
+      "learning_rate": 1.4387719283350488e-05,
+      "loss": 0.663,
+      "step": 7198
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.132287068378385,
+      "learning_rate": 1.43862038254398e-05,
+      "loss": 0.5977,
+      "step": 7199
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.118131337038932,
+      "learning_rate": 1.4384688242788034e-05,
+      "loss": 0.727,
+      "step": 7200
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 17.18136887261186,
+      "learning_rate": 1.4383172535438306e-05,
+      "loss": 0.6412,
+      "step": 7201
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.190909346982185,
+      "learning_rate": 1.4381656703433713e-05,
+      "loss": 0.7353,
+      "step": 7202
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 21.15093742302803,
+      "learning_rate": 1.4380140746817366e-05,
+      "loss": 0.6698,
+      "step": 7203
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.195003159654558,
+      "learning_rate": 1.4378624665632379e-05,
+      "loss": 0.6264,
+      "step": 7204
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.480302149440988,
+      "learning_rate": 1.4377108459921867e-05,
+      "loss": 0.6304,
+      "step": 7205
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.97901000046173,
+      "learning_rate": 1.4375592129728952e-05,
+      "loss": 0.6901,
+      "step": 7206
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.494333391847798,
+      "learning_rate": 1.4374075675096757e-05,
+      "loss": 0.5668,
+      "step": 7207
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 20.394876210865057,
+      "learning_rate": 1.4372559096068406e-05,
+      "loss": 0.6654,
+      "step": 7208
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.987401928186944,
+      "learning_rate": 1.4371042392687036e-05,
+      "loss": 0.7057,
+      "step": 7209
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 24.61080116741353,
+      "learning_rate": 1.4369525564995774e-05,
+      "loss": 0.6659,
+      "step": 7210
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.792474335419126,
+      "learning_rate": 1.4368008613037763e-05,
+      "loss": 0.6541,
+      "step": 7211
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.998459234696274,
+      "learning_rate": 1.436649153685614e-05,
+      "loss": 0.6462,
+      "step": 7212
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 27.13593411714945,
+      "learning_rate": 1.4364974336494055e-05,
+      "loss": 0.6775,
+      "step": 7213
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.118755586011249,
+      "learning_rate": 1.436345701199465e-05,
+      "loss": 0.5968,
+      "step": 7214
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 21.300974833288958,
+      "learning_rate": 1.4361939563401082e-05,
+      "loss": 0.6587,
+      "step": 7215
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 22.01647242383209,
+      "learning_rate": 1.4360421990756506e-05,
+      "loss": 0.6577,
+      "step": 7216
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.977488267357753,
+      "learning_rate": 1.4358904294104076e-05,
+      "loss": 0.6365,
+      "step": 7217
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.596867110520396,
+      "learning_rate": 1.4357386473486961e-05,
+      "loss": 0.7022,
+      "step": 7218
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 26.604504721962428,
+      "learning_rate": 1.435586852894832e-05,
+      "loss": 0.7153,
+      "step": 7219
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 23.031523037796934,
+      "learning_rate": 1.435435046053133e-05,
+      "loss": 0.7289,
+      "step": 7220
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.795522085783244,
+      "learning_rate": 1.4352832268279158e-05,
+      "loss": 0.7591,
+      "step": 7221
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.248858315808214,
+      "learning_rate": 1.4351313952234985e-05,
+      "loss": 0.684,
+      "step": 7222
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.92443869022873,
+      "learning_rate": 1.4349795512441984e-05,
+      "loss": 0.6102,
+      "step": 7223
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.636637438034517,
+      "learning_rate": 1.434827694894335e-05,
+      "loss": 0.6419,
+      "step": 7224
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 27.59543016584774,
+      "learning_rate": 1.4346758261782256e-05,
+      "loss": 0.7201,
+      "step": 7225
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.881622650791597,
+      "learning_rate": 1.4345239451001905e-05,
+      "loss": 0.7057,
+      "step": 7226
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.200292531914716,
+      "learning_rate": 1.4343720516645482e-05,
+      "loss": 0.7032,
+      "step": 7227
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 20.024773365993283,
+      "learning_rate": 1.4342201458756193e-05,
+      "loss": 0.6261,
+      "step": 7228
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.231735378911848,
+      "learning_rate": 1.434068227737723e-05,
+      "loss": 0.6029,
+      "step": 7229
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.76573305409677,
+      "learning_rate": 1.4339162972551806e-05,
+      "loss": 0.5652,
+      "step": 7230
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.957174920035747,
+      "learning_rate": 1.4337643544323124e-05,
+      "loss": 0.6785,
+      "step": 7231
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.60017781909677,
+      "learning_rate": 1.4336123992734396e-05,
+      "loss": 0.6322,
+      "step": 7232
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.529308365836304,
+      "learning_rate": 1.433460431782884e-05,
+      "loss": 0.7221,
+      "step": 7233
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.9133651038106,
+      "learning_rate": 1.4333084519649671e-05,
+      "loss": 0.675,
+      "step": 7234
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.231114544916153,
+      "learning_rate": 1.4331564598240112e-05,
+      "loss": 0.7332,
+      "step": 7235
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.315115823194645,
+      "learning_rate": 1.4330044553643391e-05,
+      "loss": 0.7017,
+      "step": 7236
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 32.258094438063104,
+      "learning_rate": 1.4328524385902738e-05,
+      "loss": 0.6993,
+      "step": 7237
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 21.244949787649077,
+      "learning_rate": 1.432700409506138e-05,
+      "loss": 0.632,
+      "step": 7238
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.06964998762685,
+      "learning_rate": 1.4325483681162562e-05,
+      "loss": 0.6341,
+      "step": 7239
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.866102818706402,
+      "learning_rate": 1.4323963144249514e-05,
+      "loss": 0.6796,
+      "step": 7240
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.021833442279746,
+      "learning_rate": 1.4322442484365486e-05,
+      "loss": 0.6807,
+      "step": 7241
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.452541188727096,
+      "learning_rate": 1.432092170155372e-05,
+      "loss": 0.6833,
+      "step": 7242
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.42122978030149,
+      "learning_rate": 1.4319400795857472e-05,
+      "loss": 0.6333,
+      "step": 7243
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 21.716899455771948,
+      "learning_rate": 1.431787976731999e-05,
+      "loss": 0.6892,
+      "step": 7244
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.069079270938037,
+      "learning_rate": 1.4316358615984532e-05,
+      "loss": 0.6718,
+      "step": 7245
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.104257318416137,
+      "learning_rate": 1.4314837341894361e-05,
+      "loss": 0.7103,
+      "step": 7246
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.881999526249366,
+      "learning_rate": 1.431331594509274e-05,
+      "loss": 0.6932,
+      "step": 7247
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.72241974066135,
+      "learning_rate": 1.4311794425622937e-05,
+      "loss": 0.6597,
+      "step": 7248
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.534884557157698,
+      "learning_rate": 1.4310272783528223e-05,
+      "loss": 0.6009,
+      "step": 7249
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 25.739369021894525,
+      "learning_rate": 1.4308751018851872e-05,
+      "loss": 0.7261,
+      "step": 7250
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.952255815194379,
+      "learning_rate": 1.4307229131637163e-05,
+      "loss": 0.5514,
+      "step": 7251
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 26.721540954729356,
+      "learning_rate": 1.4305707121927377e-05,
+      "loss": 0.7217,
+      "step": 7252
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.57838382215206,
+      "learning_rate": 1.43041849897658e-05,
+      "loss": 0.6106,
+      "step": 7253
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 25.813005878857766,
+      "learning_rate": 1.4302662735195717e-05,
+      "loss": 0.7072,
+      "step": 7254
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.81141440722146,
+      "learning_rate": 1.4301140358260426e-05,
+      "loss": 0.5864,
+      "step": 7255
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 23.280836446020455,
+      "learning_rate": 1.4299617859003218e-05,
+      "loss": 0.6065,
+      "step": 7256
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 29.162200611807993,
+      "learning_rate": 1.4298095237467394e-05,
+      "loss": 0.7162,
+      "step": 7257
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 23.580302761666395,
+      "learning_rate": 1.4296572493696255e-05,
+      "loss": 0.7014,
+      "step": 7258
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.698055707703222,
+      "learning_rate": 1.4295049627733107e-05,
+      "loss": 0.7062,
+      "step": 7259
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 22.704559497119337,
+      "learning_rate": 1.4293526639621262e-05,
+      "loss": 0.7014,
+      "step": 7260
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.237680704039285,
+      "learning_rate": 1.4292003529404028e-05,
+      "loss": 0.6506,
+      "step": 7261
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 20.000050231751302,
+      "learning_rate": 1.4290480297124726e-05,
+      "loss": 0.7142,
+      "step": 7262
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.553125203381537,
+      "learning_rate": 1.4288956942826674e-05,
+      "loss": 0.6221,
+      "step": 7263
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.705421841070356,
+      "learning_rate": 1.4287433466553198e-05,
+      "loss": 0.6185,
+      "step": 7264
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.954092152866856,
+      "learning_rate": 1.4285909868347621e-05,
+      "loss": 0.6653,
+      "step": 7265
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 25.50243836086091,
+      "learning_rate": 1.4284386148253273e-05,
+      "loss": 0.6396,
+      "step": 7266
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.404806212376876,
+      "learning_rate": 1.4282862306313488e-05,
+      "loss": 0.6636,
+      "step": 7267
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 26.71300850298545,
+      "learning_rate": 1.4281338342571609e-05,
+      "loss": 0.7065,
+      "step": 7268
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 25.39621370050001,
+      "learning_rate": 1.4279814257070967e-05,
+      "loss": 0.6672,
+      "step": 7269
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 26.62580982619853,
+      "learning_rate": 1.4278290049854917e-05,
+      "loss": 0.7053,
+      "step": 7270
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.372850870615956,
+      "learning_rate": 1.4276765720966797e-05,
+      "loss": 0.7387,
+      "step": 7271
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 26.858369018973544,
+      "learning_rate": 1.4275241270449962e-05,
+      "loss": 0.6253,
+      "step": 7272
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.310326414065248,
+      "learning_rate": 1.4273716698347766e-05,
+      "loss": 0.6862,
+      "step": 7273
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.1063625762404,
+      "learning_rate": 1.4272192004703569e-05,
+      "loss": 0.5218,
+      "step": 7274
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.870113432159268,
+      "learning_rate": 1.4270667189560727e-05,
+      "loss": 0.5224,
+      "step": 7275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.723494026688144,
+      "learning_rate": 1.426914225296261e-05,
+      "loss": 0.6658,
+      "step": 7276
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.30637815410105,
+      "learning_rate": 1.4267617194952588e-05,
+      "loss": 0.6535,
+      "step": 7277
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 24.63036808573288,
+      "learning_rate": 1.4266092015574027e-05,
+      "loss": 0.707,
+      "step": 7278
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.63144918739665,
+      "learning_rate": 1.4264566714870301e-05,
+      "loss": 0.6415,
+      "step": 7279
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 27.568212890235827,
+      "learning_rate": 1.4263041292884795e-05,
+      "loss": 0.643,
+      "step": 7280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 27.427697458550004,
+      "learning_rate": 1.426151574966089e-05,
+      "loss": 0.7036,
+      "step": 7281
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 23.635511133263815,
+      "learning_rate": 1.425999008524197e-05,
+      "loss": 0.6852,
+      "step": 7282
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.44841963596834,
+      "learning_rate": 1.4258464299671425e-05,
+      "loss": 0.6534,
+      "step": 7283
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 23.55068954080406,
+      "learning_rate": 1.4256938392992641e-05,
+      "loss": 0.6783,
+      "step": 7284
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.273292658109778,
+      "learning_rate": 1.4255412365249027e-05,
+      "loss": 0.6484,
+      "step": 7285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.905888167398263,
+      "learning_rate": 1.4253886216483968e-05,
+      "loss": 0.7714,
+      "step": 7286
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 41.88415731093706,
+      "learning_rate": 1.4252359946740877e-05,
+      "loss": 0.7661,
+      "step": 7287
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.98892526540278,
+      "learning_rate": 1.4250833556063155e-05,
+      "loss": 0.6841,
+      "step": 7288
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.74233157436609,
+      "learning_rate": 1.4249307044494214e-05,
+      "loss": 0.7232,
+      "step": 7289
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 28.876401321252764,
+      "learning_rate": 1.4247780412077466e-05,
+      "loss": 0.8119,
+      "step": 7290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.898143076431495,
+      "learning_rate": 1.4246253658856329e-05,
+      "loss": 0.669,
+      "step": 7291
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.584322122119293,
+      "learning_rate": 1.4244726784874221e-05,
+      "loss": 0.609,
+      "step": 7292
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.693950548453664,
+      "learning_rate": 1.4243199790174566e-05,
+      "loss": 0.742,
+      "step": 7293
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.46314194949273,
+      "learning_rate": 1.4241672674800791e-05,
+      "loss": 0.6365,
+      "step": 7294
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.011452578360043,
+      "learning_rate": 1.4240145438796329e-05,
+      "loss": 0.8068,
+      "step": 7295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.842204373113004,
+      "learning_rate": 1.4238618082204609e-05,
+      "loss": 0.7043,
+      "step": 7296
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.18058377223884,
+      "learning_rate": 1.4237090605069072e-05,
+      "loss": 0.6929,
+      "step": 7297
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.439286518547199,
+      "learning_rate": 1.4235563007433153e-05,
+      "loss": 0.6226,
+      "step": 7298
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.97953463894396,
+      "learning_rate": 1.4234035289340304e-05,
+      "loss": 0.5827,
+      "step": 7299
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 25.155805540601076,
+      "learning_rate": 1.4232507450833966e-05,
+      "loss": 0.6884,
+      "step": 7300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.895742233843588,
+      "learning_rate": 1.4230979491957592e-05,
+      "loss": 0.7232,
+      "step": 7301
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.15858242606363,
+      "learning_rate": 1.4229451412754638e-05,
+      "loss": 0.6397,
+      "step": 7302
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 24.3522504691592,
+      "learning_rate": 1.4227923213268557e-05,
+      "loss": 0.7252,
+      "step": 7303
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 40.257306912484495,
+      "learning_rate": 1.4226394893542817e-05,
+      "loss": 0.7522,
+      "step": 7304
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.102134458095758,
+      "learning_rate": 1.4224866453620875e-05,
+      "loss": 0.6615,
+      "step": 7305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.660909969686745,
+      "learning_rate": 1.4223337893546206e-05,
+      "loss": 0.7609,
+      "step": 7306
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.064511993637204,
+      "learning_rate": 1.4221809213362275e-05,
+      "loss": 0.7111,
+      "step": 7307
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.592858773860158,
+      "learning_rate": 1.422028041311256e-05,
+      "loss": 0.6496,
+      "step": 7308
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.298560536797847,
+      "learning_rate": 1.421875149284054e-05,
+      "loss": 0.7658,
+      "step": 7309
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.27792632629873,
+      "learning_rate": 1.4217222452589694e-05,
+      "loss": 0.6584,
+      "step": 7310
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 29.62450832952857,
+      "learning_rate": 1.4215693292403508e-05,
+      "loss": 0.6047,
+      "step": 7311
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.945433246974968,
+      "learning_rate": 1.4214164012325475e-05,
+      "loss": 0.6657,
+      "step": 7312
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.511089778686312,
+      "learning_rate": 1.421263461239908e-05,
+      "loss": 0.6407,
+      "step": 7313
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.651154781658498,
+      "learning_rate": 1.421110509266782e-05,
+      "loss": 0.5592,
+      "step": 7314
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.131092057490608,
+      "learning_rate": 1.4209575453175195e-05,
+      "loss": 0.5954,
+      "step": 7315
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.539222297112726,
+      "learning_rate": 1.4208045693964707e-05,
+      "loss": 0.5758,
+      "step": 7316
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.98470227391721,
+      "learning_rate": 1.4206515815079862e-05,
+      "loss": 0.7989,
+      "step": 7317
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.938613006826344,
+      "learning_rate": 1.4204985816564167e-05,
+      "loss": 0.7022,
+      "step": 7318
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.651870943641747,
+      "learning_rate": 1.4203455698461135e-05,
+      "loss": 0.6421,
+      "step": 7319
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 27.187722911140124,
+      "learning_rate": 1.4201925460814282e-05,
+      "loss": 0.7116,
+      "step": 7320
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 24.278937345022634,
+      "learning_rate": 1.4200395103667126e-05,
+      "loss": 0.6897,
+      "step": 7321
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.002033966297134,
+      "learning_rate": 1.4198864627063194e-05,
+      "loss": 0.7879,
+      "step": 7322
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.94732913452051,
+      "learning_rate": 1.4197334031046004e-05,
+      "loss": 0.5438,
+      "step": 7323
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.17885160212382,
+      "learning_rate": 1.4195803315659092e-05,
+      "loss": 0.6195,
+      "step": 7324
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 24.56809037171448,
+      "learning_rate": 1.4194272480945987e-05,
+      "loss": 0.6291,
+      "step": 7325
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.22303539589026,
+      "learning_rate": 1.4192741526950226e-05,
+      "loss": 0.6927,
+      "step": 7326
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.75937794270382,
+      "learning_rate": 1.4191210453715351e-05,
+      "loss": 0.6513,
+      "step": 7327
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.299243211425,
+      "learning_rate": 1.4189679261284899e-05,
+      "loss": 0.699,
+      "step": 7328
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.46224169867028,
+      "learning_rate": 1.4188147949702425e-05,
+      "loss": 0.6191,
+      "step": 7329
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.030742795884645,
+      "learning_rate": 1.4186616519011467e-05,
+      "loss": 0.7157,
+      "step": 7330
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.007342884044387,
+      "learning_rate": 1.4185084969255589e-05,
+      "loss": 0.6542,
+      "step": 7331
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.038201794405612,
+      "learning_rate": 1.4183553300478339e-05,
+      "loss": 0.643,
+      "step": 7332
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.23321810154991,
+      "learning_rate": 1.4182021512723283e-05,
+      "loss": 0.6499,
+      "step": 7333
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.29804193438696,
+      "learning_rate": 1.418048960603398e-05,
+      "loss": 0.7239,
+      "step": 7334
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.06691928320673,
+      "learning_rate": 1.4178957580454e-05,
+      "loss": 0.7259,
+      "step": 7335
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.185520240060747,
+      "learning_rate": 1.417742543602691e-05,
+      "loss": 0.6354,
+      "step": 7336
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.189090704259083,
+      "learning_rate": 1.4175893172796285e-05,
+      "loss": 0.5948,
+      "step": 7337
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.50046685807299,
+      "learning_rate": 1.4174360790805699e-05,
+      "loss": 0.7353,
+      "step": 7338
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.518134735725692,
+      "learning_rate": 1.417282829009874e-05,
+      "loss": 0.5918,
+      "step": 7339
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.761593482433817,
+      "learning_rate": 1.4171295670718976e-05,
+      "loss": 0.6593,
+      "step": 7340
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.163263791284107,
+      "learning_rate": 1.4169762932710012e-05,
+      "loss": 0.6502,
+      "step": 7341
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.779312368198045,
+      "learning_rate": 1.4168230076115424e-05,
+      "loss": 0.5405,
+      "step": 7342
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.719819136330365,
+      "learning_rate": 1.4166697100978812e-05,
+      "loss": 0.6644,
+      "step": 7343
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.919184132740725,
+      "learning_rate": 1.4165164007343772e-05,
+      "loss": 0.634,
+      "step": 7344
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.712967475570874,
+      "learning_rate": 1.4163630795253904e-05,
+      "loss": 0.688,
+      "step": 7345
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.3363748803825,
+      "learning_rate": 1.4162097464752815e-05,
+      "loss": 0.6943,
+      "step": 7346
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.560537049527449,
+      "learning_rate": 1.4160564015884103e-05,
+      "loss": 0.5626,
+      "step": 7347
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.223194824677954,
+      "learning_rate": 1.415903044869139e-05,
+      "loss": 0.6359,
+      "step": 7348
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.69957061072878,
+      "learning_rate": 1.4157496763218278e-05,
+      "loss": 0.6065,
+      "step": 7349
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.588844511724094,
+      "learning_rate": 1.4155962959508396e-05,
+      "loss": 0.7005,
+      "step": 7350
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.55260639971917,
+      "learning_rate": 1.4154429037605358e-05,
+      "loss": 0.7367,
+      "step": 7351
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.867184974711531,
+      "learning_rate": 1.4152894997552786e-05,
+      "loss": 0.636,
+      "step": 7352
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.56582770342848,
+      "learning_rate": 1.415136083939431e-05,
+      "loss": 0.7422,
+      "step": 7353
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.27299115539783,
+      "learning_rate": 1.414982656317356e-05,
+      "loss": 0.7147,
+      "step": 7354
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.87205477638366,
+      "learning_rate": 1.414829216893417e-05,
+      "loss": 0.6581,
+      "step": 7355
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 25.777476602169482,
+      "learning_rate": 1.4146757656719781e-05,
+      "loss": 0.622,
+      "step": 7356
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.911269385805255,
+      "learning_rate": 1.4145223026574027e-05,
+      "loss": 0.6752,
+      "step": 7357
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 26.591560678767447,
+      "learning_rate": 1.4143688278540554e-05,
+      "loss": 0.6516,
+      "step": 7358
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 27.284983232587212,
+      "learning_rate": 1.4142153412663012e-05,
+      "loss": 0.7673,
+      "step": 7359
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 30.947000097337614,
+      "learning_rate": 1.414061842898505e-05,
+      "loss": 0.6534,
+      "step": 7360
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.162060374390233,
+      "learning_rate": 1.413908332755032e-05,
+      "loss": 0.7408,
+      "step": 7361
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.071875773411218,
+      "learning_rate": 1.4137548108402483e-05,
+      "loss": 0.607,
+      "step": 7362
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.91990479229881,
+      "learning_rate": 1.41360127715852e-05,
+      "loss": 0.6592,
+      "step": 7363
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.913176749864778,
+      "learning_rate": 1.4134477317142133e-05,
+      "loss": 0.7209,
+      "step": 7364
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 22.581044234735447,
+      "learning_rate": 1.4132941745116946e-05,
+      "loss": 0.6888,
+      "step": 7365
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 22.878128883946122,
+      "learning_rate": 1.4131406055553316e-05,
+      "loss": 0.7104,
+      "step": 7366
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.10784189075606,
+      "learning_rate": 1.4129870248494913e-05,
+      "loss": 0.6707,
+      "step": 7367
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.9504778036491,
+      "learning_rate": 1.4128334323985418e-05,
+      "loss": 0.7455,
+      "step": 7368
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.608233045903027,
+      "learning_rate": 1.412679828206851e-05,
+      "loss": 0.6429,
+      "step": 7369
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.300653763023575,
+      "learning_rate": 1.4125262122787872e-05,
+      "loss": 0.621,
+      "step": 7370
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.980337563510073,
+      "learning_rate": 1.4123725846187193e-05,
+      "loss": 0.6928,
+      "step": 7371
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.370839955994054,
+      "learning_rate": 1.412218945231016e-05,
+      "loss": 0.6911,
+      "step": 7372
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.750375623657416,
+      "learning_rate": 1.4120652941200477e-05,
+      "loss": 0.6599,
+      "step": 7373
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.144720114282386,
+      "learning_rate": 1.4119116312901828e-05,
+      "loss": 0.6486,
+      "step": 7374
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.228050529146017,
+      "learning_rate": 1.4117579567457927e-05,
+      "loss": 0.5787,
+      "step": 7375
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.31204141916774,
+      "learning_rate": 1.4116042704912465e-05,
+      "loss": 0.7116,
+      "step": 7376
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.973238864326664,
+      "learning_rate": 1.411450572530916e-05,
+      "loss": 0.686,
+      "step": 7377
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.533872217040669,
+      "learning_rate": 1.411296862869172e-05,
+      "loss": 0.6718,
+      "step": 7378
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.361131588851002,
+      "learning_rate": 1.4111431415103858e-05,
+      "loss": 0.716,
+      "step": 7379
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.509238595237658,
+      "learning_rate": 1.4109894084589291e-05,
+      "loss": 0.6428,
+      "step": 7380
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.586743325361361,
+      "learning_rate": 1.410835663719174e-05,
+      "loss": 0.6645,
+      "step": 7381
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.186221632749692,
+      "learning_rate": 1.410681907295493e-05,
+      "loss": 0.6604,
+      "step": 7382
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.939096133396044,
+      "learning_rate": 1.410528139192259e-05,
+      "loss": 0.7495,
+      "step": 7383
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 22.933165789733653,
+      "learning_rate": 1.4103743594138443e-05,
+      "loss": 0.6805,
+      "step": 7384
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.45622464524285,
+      "learning_rate": 1.4102205679646236e-05,
+      "loss": 0.7013,
+      "step": 7385
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.055066577056435,
+      "learning_rate": 1.4100667648489692e-05,
+      "loss": 0.6227,
+      "step": 7386
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.882454329212287,
+      "learning_rate": 1.4099129500712562e-05,
+      "loss": 0.5995,
+      "step": 7387
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.79465860119624,
+      "learning_rate": 1.4097591236358588e-05,
+      "loss": 0.7247,
+      "step": 7388
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.272643075951262,
+      "learning_rate": 1.4096052855471519e-05,
+      "loss": 0.6833,
+      "step": 7389
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.842453650218186,
+      "learning_rate": 1.4094514358095096e-05,
+      "loss": 0.6822,
+      "step": 7390
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.675001128036836,
+      "learning_rate": 1.4092975744273082e-05,
+      "loss": 0.5861,
+      "step": 7391
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.759574079585438,
+      "learning_rate": 1.4091437014049234e-05,
+      "loss": 0.6755,
+      "step": 7392
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.376489050078144,
+      "learning_rate": 1.408989816746731e-05,
+      "loss": 0.7123,
+      "step": 7393
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.247837955897612,
+      "learning_rate": 1.4088359204571076e-05,
+      "loss": 0.7154,
+      "step": 7394
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.001454248068562,
+      "learning_rate": 1.4086820125404299e-05,
+      "loss": 0.6979,
+      "step": 7395
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.785621788868074,
+      "learning_rate": 1.4085280930010745e-05,
+      "loss": 0.6524,
+      "step": 7396
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.620393958684955,
+      "learning_rate": 1.4083741618434192e-05,
+      "loss": 0.6626,
+      "step": 7397
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 24.826052385045198,
+      "learning_rate": 1.4082202190718417e-05,
+      "loss": 0.6586,
+      "step": 7398
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 22.273155315649465,
+      "learning_rate": 1.40806626469072e-05,
+      "loss": 0.6882,
+      "step": 7399
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.282204565720075,
+      "learning_rate": 1.4079122987044324e-05,
+      "loss": 0.6899,
+      "step": 7400
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.549876274179365,
+      "learning_rate": 1.4077583211173575e-05,
+      "loss": 0.717,
+      "step": 7401
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.33697810834557,
+      "learning_rate": 1.4076043319338748e-05,
+      "loss": 0.5725,
+      "step": 7402
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.315063533646104,
+      "learning_rate": 1.407450331158363e-05,
+      "loss": 0.7549,
+      "step": 7403
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.744429739733686,
+      "learning_rate": 1.4072963187952023e-05,
+      "loss": 0.5895,
+      "step": 7404
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.463991743229743,
+      "learning_rate": 1.4071422948487725e-05,
+      "loss": 0.6726,
+      "step": 7405
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.788281130249198,
+      "learning_rate": 1.406988259323454e-05,
+      "loss": 0.6566,
+      "step": 7406
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.20010528951628,
+      "learning_rate": 1.4068342122236275e-05,
+      "loss": 0.7284,
+      "step": 7407
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 28.13856353601223,
+      "learning_rate": 1.406680153553674e-05,
+      "loss": 0.7362,
+      "step": 7408
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.419744186157724,
+      "learning_rate": 1.4065260833179748e-05,
+      "loss": 0.6689,
+      "step": 7409
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 22.797216267413933,
+      "learning_rate": 1.4063720015209117e-05,
+      "loss": 0.6516,
+      "step": 7410
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.152084730649868,
+      "learning_rate": 1.4062179081668665e-05,
+      "loss": 0.6049,
+      "step": 7411
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.964435438638063,
+      "learning_rate": 1.4060638032602218e-05,
+      "loss": 0.6738,
+      "step": 7412
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.80899551607031,
+      "learning_rate": 1.4059096868053601e-05,
+      "loss": 0.6962,
+      "step": 7413
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.636531243613337,
+      "learning_rate": 1.405755558806664e-05,
+      "loss": 0.7195,
+      "step": 7414
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.075681713180828,
+      "learning_rate": 1.4056014192685175e-05,
+      "loss": 0.6948,
+      "step": 7415
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.800611271161323,
+      "learning_rate": 1.4054472681953035e-05,
+      "loss": 0.6461,
+      "step": 7416
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.98961262095591,
+      "learning_rate": 1.405293105591407e-05,
+      "loss": 0.6318,
+      "step": 7417
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.766827578629822,
+      "learning_rate": 1.4051389314612112e-05,
+      "loss": 0.6247,
+      "step": 7418
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.22760036980565,
+      "learning_rate": 1.4049847458091014e-05,
+      "loss": 0.6329,
+      "step": 7419
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.506367963971051,
+      "learning_rate": 1.404830548639462e-05,
+      "loss": 0.6172,
+      "step": 7420
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.971184970853464,
+      "learning_rate": 1.404676339956679e-05,
+      "loss": 0.686,
+      "step": 7421
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.00354369947476,
+      "learning_rate": 1.4045221197651375e-05,
+      "loss": 0.7335,
+      "step": 7422
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.898873788162456,
+      "learning_rate": 1.4043678880692235e-05,
+      "loss": 0.5801,
+      "step": 7423
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.175083200681893,
+      "learning_rate": 1.4042136448733235e-05,
+      "loss": 0.6682,
+      "step": 7424
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.586331558829869,
+      "learning_rate": 1.4040593901818239e-05,
+      "loss": 0.6266,
+      "step": 7425
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.76058851999795,
+      "learning_rate": 1.403905123999111e-05,
+      "loss": 0.6591,
+      "step": 7426
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.896263202572664,
+      "learning_rate": 1.4037508463295735e-05,
+      "loss": 0.5713,
+      "step": 7427
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.76547623133555,
+      "learning_rate": 1.4035965571775976e-05,
+      "loss": 0.6287,
+      "step": 7428
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.94958560604843,
+      "learning_rate": 1.403442256547572e-05,
+      "loss": 0.6983,
+      "step": 7429
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.673931826429664,
+      "learning_rate": 1.4032879444438843e-05,
+      "loss": 0.6669,
+      "step": 7430
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.755173992948183,
+      "learning_rate": 1.4031336208709236e-05,
+      "loss": 0.6923,
+      "step": 7431
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.853443592981684,
+      "learning_rate": 1.4029792858330783e-05,
+      "loss": 0.7087,
+      "step": 7432
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.04611829145561,
+      "learning_rate": 1.4028249393347382e-05,
+      "loss": 0.605,
+      "step": 7433
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.575987522186704,
+      "learning_rate": 1.4026705813802923e-05,
+      "loss": 0.6012,
+      "step": 7434
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.225078286878,
+      "learning_rate": 1.4025162119741304e-05,
+      "loss": 0.5948,
+      "step": 7435
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.87707332705314,
+      "learning_rate": 1.4023618311206432e-05,
+      "loss": 0.7036,
+      "step": 7436
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 25.970216250655813,
+      "learning_rate": 1.4022074388242208e-05,
+      "loss": 0.7093,
+      "step": 7437
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.026213796162736,
+      "learning_rate": 1.4020530350892542e-05,
+      "loss": 0.7042,
+      "step": 7438
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.980603976272196,
+      "learning_rate": 1.4018986199201345e-05,
+      "loss": 0.761,
+      "step": 7439
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 24.443586857918696,
+      "learning_rate": 1.4017441933212532e-05,
+      "loss": 0.6455,
+      "step": 7440
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.55953781980156,
+      "learning_rate": 1.4015897552970017e-05,
+      "loss": 0.6394,
+      "step": 7441
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.26280079205925,
+      "learning_rate": 1.401435305851773e-05,
+      "loss": 0.6492,
+      "step": 7442
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.57450492333809,
+      "learning_rate": 1.4012808449899584e-05,
+      "loss": 0.6636,
+      "step": 7443
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 25.879526449869036,
+      "learning_rate": 1.4011263727159521e-05,
+      "loss": 0.6698,
+      "step": 7444
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.09169694808246,
+      "learning_rate": 1.4009718890341457e-05,
+      "loss": 0.6293,
+      "step": 7445
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 28.297327064344987,
+      "learning_rate": 1.4008173939489338e-05,
+      "loss": 0.6776,
+      "step": 7446
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.47554753763949,
+      "learning_rate": 1.4006628874647094e-05,
+      "loss": 0.6542,
+      "step": 7447
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.084212500370837,
+      "learning_rate": 1.400508369585867e-05,
+      "loss": 0.6238,
+      "step": 7448
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 21.652779760758925,
+      "learning_rate": 1.400353840316801e-05,
+      "loss": 0.7362,
+      "step": 7449
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.954309514971452,
+      "learning_rate": 1.4001992996619056e-05,
+      "loss": 0.6846,
+      "step": 7450
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.143083624679395,
+      "learning_rate": 1.4000447476255765e-05,
+      "loss": 0.7219,
+      "step": 7451
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.73512113971914,
+      "learning_rate": 1.3998901842122088e-05,
+      "loss": 0.6412,
+      "step": 7452
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.6357970465466,
+      "learning_rate": 1.3997356094261977e-05,
+      "loss": 0.6264,
+      "step": 7453
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.022154026352275,
+      "learning_rate": 1.3995810232719405e-05,
+      "loss": 0.6172,
+      "step": 7454
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.373682419762957,
+      "learning_rate": 1.3994264257538324e-05,
+      "loss": 0.696,
+      "step": 7455
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 22.068361951330044,
+      "learning_rate": 1.3992718168762702e-05,
+      "loss": 0.5801,
+      "step": 7456
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.87333762075109,
+      "learning_rate": 1.3991171966436513e-05,
+      "loss": 0.7128,
+      "step": 7457
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.790901233762101,
+      "learning_rate": 1.3989625650603729e-05,
+      "loss": 0.5648,
+      "step": 7458
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.56375106814535,
+      "learning_rate": 1.3988079221308323e-05,
+      "loss": 0.619,
+      "step": 7459
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.73698580098692,
+      "learning_rate": 1.3986532678594277e-05,
+      "loss": 0.6288,
+      "step": 7460
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.035665876046952,
+      "learning_rate": 1.3984986022505578e-05,
+      "loss": 0.7057,
+      "step": 7461
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 17.590239793799036,
+      "learning_rate": 1.3983439253086201e-05,
+      "loss": 0.6575,
+      "step": 7462
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.885276763696012,
+      "learning_rate": 1.3981892370380146e-05,
+      "loss": 0.6823,
+      "step": 7463
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 22.377524375366338,
+      "learning_rate": 1.3980345374431401e-05,
+      "loss": 0.7452,
+      "step": 7464
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.54329056116013,
+      "learning_rate": 1.3978798265283962e-05,
+      "loss": 0.6429,
+      "step": 7465
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 25.903684229042685,
+      "learning_rate": 1.3977251042981828e-05,
+      "loss": 0.7628,
+      "step": 7466
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.999764541180905,
+      "learning_rate": 1.3975703707569e-05,
+      "loss": 0.6401,
+      "step": 7467
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.223170027261705,
+      "learning_rate": 1.3974156259089486e-05,
+      "loss": 0.7002,
+      "step": 7468
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.750197318135148,
+      "learning_rate": 1.3972608697587292e-05,
+      "loss": 0.6167,
+      "step": 7469
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.489287971411944,
+      "learning_rate": 1.3971061023106428e-05,
+      "loss": 0.6141,
+      "step": 7470
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.059324109406212,
+      "learning_rate": 1.3969513235690915e-05,
+      "loss": 0.6861,
+      "step": 7471
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.256802116335816,
+      "learning_rate": 1.3967965335384766e-05,
+      "loss": 0.6624,
+      "step": 7472
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.706032848147292,
+      "learning_rate": 1.3966417322232006e-05,
+      "loss": 0.6457,
+      "step": 7473
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 17.277762907093944,
+      "learning_rate": 1.3964869196276654e-05,
+      "loss": 0.659,
+      "step": 7474
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.595718087145244,
+      "learning_rate": 1.3963320957562743e-05,
+      "loss": 0.7015,
+      "step": 7475
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.606199690884935,
+      "learning_rate": 1.3961772606134303e-05,
+      "loss": 0.6491,
+      "step": 7476
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 17.200989268409902,
+      "learning_rate": 1.3960224142035368e-05,
+      "loss": 0.6213,
+      "step": 7477
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.51283058533687,
+      "learning_rate": 1.3958675565309974e-05,
+      "loss": 0.6469,
+      "step": 7478
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.902464307051382,
+      "learning_rate": 1.3957126876002164e-05,
+      "loss": 0.7062,
+      "step": 7479
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.246684168649182,
+      "learning_rate": 1.3955578074155978e-05,
+      "loss": 0.6201,
+      "step": 7480
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.684449043338914,
+      "learning_rate": 1.3954029159815468e-05,
+      "loss": 0.5924,
+      "step": 7481
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.218361384951006,
+      "learning_rate": 1.3952480133024682e-05,
+      "loss": 0.6665,
+      "step": 7482
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.73341601683904,
+      "learning_rate": 1.3950930993827671e-05,
+      "loss": 0.6646,
+      "step": 7483
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.65514275334491,
+      "learning_rate": 1.3949381742268496e-05,
+      "loss": 0.6029,
+      "step": 7484
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.227009257813034,
+      "learning_rate": 1.394783237839121e-05,
+      "loss": 0.5666,
+      "step": 7485
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.874543880928986,
+      "learning_rate": 1.3946282902239886e-05,
+      "loss": 0.6344,
+      "step": 7486
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.48261663396154,
+      "learning_rate": 1.3944733313858583e-05,
+      "loss": 0.7229,
+      "step": 7487
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.218143652509674,
+      "learning_rate": 1.3943183613291374e-05,
+      "loss": 0.73,
+      "step": 7488
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.645593324817906,
+      "learning_rate": 1.3941633800582325e-05,
+      "loss": 0.7254,
+      "step": 7489
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.409587059496285,
+      "learning_rate": 1.394008387577552e-05,
+      "loss": 0.7066,
+      "step": 7490
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.0758242745071,
+      "learning_rate": 1.3938533838915033e-05,
+      "loss": 0.6805,
+      "step": 7491
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.481029815308045,
+      "learning_rate": 1.3936983690044949e-05,
+      "loss": 0.6164,
+      "step": 7492
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.771087089235197,
+      "learning_rate": 1.3935433429209352e-05,
+      "loss": 0.6318,
+      "step": 7493
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.97081639388146,
+      "learning_rate": 1.3933883056452331e-05,
+      "loss": 0.7006,
+      "step": 7494
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.4972055204725,
+      "learning_rate": 1.3932332571817975e-05,
+      "loss": 0.6624,
+      "step": 7495
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.643876100745285,
+      "learning_rate": 1.3930781975350382e-05,
+      "loss": 0.6949,
+      "step": 7496
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.89933613779166,
+      "learning_rate": 1.3929231267093646e-05,
+      "loss": 0.7052,
+      "step": 7497
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.711042071980877,
+      "learning_rate": 1.3927680447091876e-05,
+      "loss": 0.811,
+      "step": 7498
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.793785704220694,
+      "learning_rate": 1.3926129515389168e-05,
+      "loss": 0.5723,
+      "step": 7499
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.612089721566733,
+      "learning_rate": 1.3924578472029637e-05,
+      "loss": 0.6871,
+      "step": 7500
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.881192241790526,
+      "learning_rate": 1.3923027317057388e-05,
+      "loss": 0.7151,
+      "step": 7501
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.59567292101353,
+      "learning_rate": 1.3921476050516538e-05,
+      "loss": 0.7365,
+      "step": 7502
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.47455473428712,
+      "learning_rate": 1.3919924672451201e-05,
+      "loss": 0.7011,
+      "step": 7503
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.764728500842555,
+      "learning_rate": 1.3918373182905501e-05,
+      "loss": 0.6463,
+      "step": 7504
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.85860007557551,
+      "learning_rate": 1.391682158192356e-05,
+      "loss": 0.7198,
+      "step": 7505
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.907522179228543,
+      "learning_rate": 1.3915269869549504e-05,
+      "loss": 0.6414,
+      "step": 7506
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.510044998498573,
+      "learning_rate": 1.3913718045827462e-05,
+      "loss": 0.6663,
+      "step": 7507
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.59371658033721,
+      "learning_rate": 1.391216611080157e-05,
+      "loss": 0.6849,
+      "step": 7508
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.287951147669038,
+      "learning_rate": 1.3910614064515964e-05,
+      "loss": 0.6739,
+      "step": 7509
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.478886626050734,
+      "learning_rate": 1.3909061907014781e-05,
+      "loss": 0.6944,
+      "step": 7510
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.072511998812043,
+      "learning_rate": 1.390750963834216e-05,
+      "loss": 0.5905,
+      "step": 7511
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.05425884238126,
+      "learning_rate": 1.3905957258542253e-05,
+      "loss": 0.7121,
+      "step": 7512
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.01832958738399,
+      "learning_rate": 1.390440476765921e-05,
+      "loss": 0.6673,
+      "step": 7513
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.41394322718299,
+      "learning_rate": 1.3902852165737172e-05,
+      "loss": 0.683,
+      "step": 7514
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.456126261052685,
+      "learning_rate": 1.390129945282031e-05,
+      "loss": 0.678,
+      "step": 7515
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.400980989672785,
+      "learning_rate": 1.3899746628952766e-05,
+      "loss": 0.7271,
+      "step": 7516
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.431509641391173,
+      "learning_rate": 1.3898193694178714e-05,
+      "loss": 0.7039,
+      "step": 7517
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.610124583257402,
+      "learning_rate": 1.3896640648542312e-05,
+      "loss": 0.6537,
+      "step": 7518
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 17.39909699504327,
+      "learning_rate": 1.389508749208773e-05,
+      "loss": 0.7293,
+      "step": 7519
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.21891755594496,
+      "learning_rate": 1.389353422485914e-05,
+      "loss": 0.6648,
+      "step": 7520
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.923362457703828,
+      "learning_rate": 1.3891980846900712e-05,
+      "loss": 0.7032,
+      "step": 7521
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.66710044209375,
+      "learning_rate": 1.3890427358256626e-05,
+      "loss": 0.6923,
+      "step": 7522
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.958698425630674,
+      "learning_rate": 1.388887375897106e-05,
+      "loss": 0.6708,
+      "step": 7523
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.38599063342712,
+      "learning_rate": 1.3887320049088202e-05,
+      "loss": 0.6418,
+      "step": 7524
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.83205121493252,
+      "learning_rate": 1.3885766228652235e-05,
+      "loss": 0.598,
+      "step": 7525
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.79574681968676,
+      "learning_rate": 1.3884212297707349e-05,
+      "loss": 0.6237,
+      "step": 7526
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.677812889038439,
+      "learning_rate": 1.3882658256297738e-05,
+      "loss": 0.7044,
+      "step": 7527
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.830279366624833,
+      "learning_rate": 1.3881104104467598e-05,
+      "loss": 0.6114,
+      "step": 7528
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.913860664342462,
+      "learning_rate": 1.3879549842261123e-05,
+      "loss": 0.6359,
+      "step": 7529
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.892387249923937,
+      "learning_rate": 1.3877995469722527e-05,
+      "loss": 0.6188,
+      "step": 7530
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.58478861911699,
+      "learning_rate": 1.3876440986896004e-05,
+      "loss": 0.6768,
+      "step": 7531
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.869863317629406,
+      "learning_rate": 1.387488639382577e-05,
+      "loss": 0.7038,
+      "step": 7532
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 30.157638454592195,
+      "learning_rate": 1.387333169055603e-05,
+      "loss": 0.6805,
+      "step": 7533
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.382775782984638,
+      "learning_rate": 1.3871776877131007e-05,
+      "loss": 0.7045,
+      "step": 7534
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.424394742648765,
+      "learning_rate": 1.3870221953594912e-05,
+      "loss": 0.6368,
+      "step": 7535
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.88825740236287,
+      "learning_rate": 1.3868666919991966e-05,
+      "loss": 0.6372,
+      "step": 7536
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.92365602822914,
+      "learning_rate": 1.38671117763664e-05,
+      "loss": 0.6643,
+      "step": 7537
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.729266756545332,
+      "learning_rate": 1.3865556522762438e-05,
+      "loss": 0.679,
+      "step": 7538
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.845997522974661,
+      "learning_rate": 1.3864001159224306e-05,
+      "loss": 0.6784,
+      "step": 7539
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 27.252948094053966,
+      "learning_rate": 1.3862445685796244e-05,
+      "loss": 0.6914,
+      "step": 7540
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 26.367243305288532,
+      "learning_rate": 1.3860890102522482e-05,
+      "loss": 0.7281,
+      "step": 7541
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.747856897861515,
+      "learning_rate": 1.385933440944727e-05,
+      "loss": 0.6155,
+      "step": 7542
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.461379410839402,
+      "learning_rate": 1.385777860661484e-05,
+      "loss": 0.6093,
+      "step": 7543
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.55451364925784,
+      "learning_rate": 1.3856222694069446e-05,
+      "loss": 0.6032,
+      "step": 7544
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 26.457505973072525,
+      "learning_rate": 1.3854666671855334e-05,
+      "loss": 0.624,
+      "step": 7545
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.36800395976475,
+      "learning_rate": 1.3853110540016759e-05,
+      "loss": 0.649,
+      "step": 7546
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.234854057709228,
+      "learning_rate": 1.3851554298597971e-05,
+      "loss": 0.7901,
+      "step": 7547
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.349697126593004,
+      "learning_rate": 1.384999794764323e-05,
+      "loss": 0.6993,
+      "step": 7548
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.34435634976441,
+      "learning_rate": 1.3848441487196804e-05,
+      "loss": 0.7236,
+      "step": 7549
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.937564966427612,
+      "learning_rate": 1.3846884917302952e-05,
+      "loss": 0.6597,
+      "step": 7550
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.575933954124604,
+      "learning_rate": 1.3845328238005942e-05,
+      "loss": 0.7012,
+      "step": 7551
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 28.011320616404007,
+      "learning_rate": 1.3843771449350047e-05,
+      "loss": 0.7467,
+      "step": 7552
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.13286322652592,
+      "learning_rate": 1.3842214551379542e-05,
+      "loss": 0.6014,
+      "step": 7553
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 25.82566865028165,
+      "learning_rate": 1.3840657544138702e-05,
+      "loss": 0.7456,
+      "step": 7554
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.244024722615029,
+      "learning_rate": 1.3839100427671808e-05,
+      "loss": 0.6363,
+      "step": 7555
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.811069034658267,
+      "learning_rate": 1.3837543202023143e-05,
+      "loss": 0.6291,
+      "step": 7556
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.160090923291442,
+      "learning_rate": 1.3835985867236995e-05,
+      "loss": 0.652,
+      "step": 7557
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.11189345926525,
+      "learning_rate": 1.383442842335765e-05,
+      "loss": 0.6006,
+      "step": 7558
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 26.496299290692853,
+      "learning_rate": 1.383287087042941e-05,
+      "loss": 0.7171,
+      "step": 7559
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.572744585304314,
+      "learning_rate": 1.383131320849656e-05,
+      "loss": 0.5934,
+      "step": 7560
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.945321114284212,
+      "learning_rate": 1.3829755437603405e-05,
+      "loss": 0.6502,
+      "step": 7561
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.953599670817265,
+      "learning_rate": 1.3828197557794245e-05,
+      "loss": 0.6227,
+      "step": 7562
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.30852525159876,
+      "learning_rate": 1.3826639569113384e-05,
+      "loss": 0.5657,
+      "step": 7563
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 27.66963524803588,
+      "learning_rate": 1.3825081471605137e-05,
+      "loss": 0.6587,
+      "step": 7564
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.762473263502336,
+      "learning_rate": 1.3823523265313808e-05,
+      "loss": 0.6205,
+      "step": 7565
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.477765238289495,
+      "learning_rate": 1.3821964950283713e-05,
+      "loss": 0.5983,
+      "step": 7566
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.924384916632853,
+      "learning_rate": 1.3820406526559169e-05,
+      "loss": 0.6099,
+      "step": 7567
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.781352330984465,
+      "learning_rate": 1.38188479941845e-05,
+      "loss": 0.6457,
+      "step": 7568
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 27.584528475708602,
+      "learning_rate": 1.381728935320403e-05,
+      "loss": 0.6183,
+      "step": 7569
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 25.564806360203946,
+      "learning_rate": 1.3815730603662083e-05,
+      "loss": 0.6289,
+      "step": 7570
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.089715600091765,
+      "learning_rate": 1.3814171745602991e-05,
+      "loss": 0.6958,
+      "step": 7571
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.6109904551384,
+      "learning_rate": 1.3812612779071084e-05,
+      "loss": 0.7949,
+      "step": 7572
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.26732133080964,
+      "learning_rate": 1.3811053704110697e-05,
+      "loss": 0.6563,
+      "step": 7573
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.909600846430628,
+      "learning_rate": 1.3809494520766179e-05,
+      "loss": 0.5822,
+      "step": 7574
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.463115556294614,
+      "learning_rate": 1.3807935229081859e-05,
+      "loss": 0.6293,
+      "step": 7575
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.67937550398115,
+      "learning_rate": 1.3806375829102092e-05,
+      "loss": 0.6278,
+      "step": 7576
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.35971127037589,
+      "learning_rate": 1.3804816320871221e-05,
+      "loss": 0.6333,
+      "step": 7577
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.773158144834433,
+      "learning_rate": 1.38032567044336e-05,
+      "loss": 0.6524,
+      "step": 7578
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.53549174216908,
+      "learning_rate": 1.3801696979833586e-05,
+      "loss": 0.7427,
+      "step": 7579
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.63057206910241,
+      "learning_rate": 1.3800137147115531e-05,
+      "loss": 0.7508,
+      "step": 7580
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.8101798712467,
+      "learning_rate": 1.37985772063238e-05,
+      "loss": 0.7527,
+      "step": 7581
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.001134149949873,
+      "learning_rate": 1.3797017157502754e-05,
+      "loss": 0.6924,
+      "step": 7582
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.23867863927879,
+      "learning_rate": 1.3795457000696762e-05,
+      "loss": 0.6913,
+      "step": 7583
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.772363631571118,
+      "learning_rate": 1.3793896735950195e-05,
+      "loss": 0.6638,
+      "step": 7584
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.612161568413985,
+      "learning_rate": 1.3792336363307419e-05,
+      "loss": 0.6242,
+      "step": 7585
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.39163862709829,
+      "learning_rate": 1.3790775882812818e-05,
+      "loss": 0.7153,
+      "step": 7586
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 28.381249904275546,
+      "learning_rate": 1.378921529451077e-05,
+      "loss": 0.7509,
+      "step": 7587
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.591754432341034,
+      "learning_rate": 1.3787654598445656e-05,
+      "loss": 0.777,
+      "step": 7588
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.40724545142196,
+      "learning_rate": 1.3786093794661858e-05,
+      "loss": 0.6377,
+      "step": 7589
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.488630621416277,
+      "learning_rate": 1.3784532883203769e-05,
+      "loss": 0.7281,
+      "step": 7590
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.80473266926859,
+      "learning_rate": 1.378297186411578e-05,
+      "loss": 0.6796,
+      "step": 7591
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.218804443536794,
+      "learning_rate": 1.378141073744228e-05,
+      "loss": 0.7338,
+      "step": 7592
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.18899216318968,
+      "learning_rate": 1.3779849503227675e-05,
+      "loss": 0.5817,
+      "step": 7593
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.51721928104307,
+      "learning_rate": 1.377828816151636e-05,
+      "loss": 0.7058,
+      "step": 7594
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.66716454915918,
+      "learning_rate": 1.377672671235274e-05,
+      "loss": 0.7502,
+      "step": 7595
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.583078490297684,
+      "learning_rate": 1.377516515578122e-05,
+      "loss": 0.6846,
+      "step": 7596
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 39.52117862531305,
+      "learning_rate": 1.3773603491846213e-05,
+      "loss": 0.6456,
+      "step": 7597
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 25.83714727863196,
+      "learning_rate": 1.3772041720592131e-05,
+      "loss": 0.719,
+      "step": 7598
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.26837292379662,
+      "learning_rate": 1.3770479842063387e-05,
+      "loss": 0.5968,
+      "step": 7599
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.494083477936199,
+      "learning_rate": 1.37689178563044e-05,
+      "loss": 0.6826,
+      "step": 7600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.190142649637103,
+      "learning_rate": 1.37673557633596e-05,
+      "loss": 0.7216,
+      "step": 7601
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 21.799027146975,
+      "learning_rate": 1.3765793563273402e-05,
+      "loss": 0.6388,
+      "step": 7602
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.595024924285916,
+      "learning_rate": 1.376423125609024e-05,
+      "loss": 0.6552,
+      "step": 7603
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.91833612434623,
+      "learning_rate": 1.376266884185454e-05,
+      "loss": 0.7241,
+      "step": 7604
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 33.587980181204244,
+      "learning_rate": 1.376110632061074e-05,
+      "loss": 0.6916,
+      "step": 7605
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 25.88662066425963,
+      "learning_rate": 1.3759543692403278e-05,
+      "loss": 0.6966,
+      "step": 7606
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.996353307665114,
+      "learning_rate": 1.3757980957276594e-05,
+      "loss": 0.7129,
+      "step": 7607
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.07633098406556,
+      "learning_rate": 1.3756418115275128e-05,
+      "loss": 0.6685,
+      "step": 7608
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.98864413724588,
+      "learning_rate": 1.3754855166443326e-05,
+      "loss": 0.6743,
+      "step": 7609
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.32496062053142,
+      "learning_rate": 1.375329211082564e-05,
+      "loss": 0.6902,
+      "step": 7610
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.280417487759227,
+      "learning_rate": 1.3751728948466526e-05,
+      "loss": 0.6503,
+      "step": 7611
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.466755148350478,
+      "learning_rate": 1.3750165679410431e-05,
+      "loss": 0.6718,
+      "step": 7612
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.280107986493373,
+      "learning_rate": 1.3748602303701821e-05,
+      "loss": 0.6518,
+      "step": 7613
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.770372852212237,
+      "learning_rate": 1.3747038821385155e-05,
+      "loss": 0.6022,
+      "step": 7614
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.899196220495579,
+      "learning_rate": 1.3745475232504895e-05,
+      "loss": 0.6409,
+      "step": 7615
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.657098477968363,
+      "learning_rate": 1.374391153710551e-05,
+      "loss": 0.6713,
+      "step": 7616
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.680950969218298,
+      "learning_rate": 1.374234773523147e-05,
+      "loss": 0.7287,
+      "step": 7617
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.114378093032236,
+      "learning_rate": 1.3740783826927252e-05,
+      "loss": 0.7616,
+      "step": 7618
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.38611065463507,
+      "learning_rate": 1.373921981223733e-05,
+      "loss": 0.7034,
+      "step": 7619
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.81119139213464,
+      "learning_rate": 1.3737655691206184e-05,
+      "loss": 0.6361,
+      "step": 7620
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 29.82213783823889,
+      "learning_rate": 1.3736091463878296e-05,
+      "loss": 0.6597,
+      "step": 7621
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 31.98501895615954,
+      "learning_rate": 1.3734527130298153e-05,
+      "loss": 0.7065,
+      "step": 7622
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.824982803353503,
+      "learning_rate": 1.373296269051024e-05,
+      "loss": 0.6366,
+      "step": 7623
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.33518289711661,
+      "learning_rate": 1.3731398144559056e-05,
+      "loss": 0.6591,
+      "step": 7624
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.55279041450296,
+      "learning_rate": 1.3729833492489091e-05,
+      "loss": 0.6656,
+      "step": 7625
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.840946987019695,
+      "learning_rate": 1.3728268734344842e-05,
+      "loss": 0.6558,
+      "step": 7626
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.403385587937322,
+      "learning_rate": 1.372670387017081e-05,
+      "loss": 0.7045,
+      "step": 7627
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.664214010145676,
+      "learning_rate": 1.3725138900011503e-05,
+      "loss": 0.6638,
+      "step": 7628
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.38587578731434,
+      "learning_rate": 1.372357382391142e-05,
+      "loss": 0.6434,
+      "step": 7629
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.366065858470517,
+      "learning_rate": 1.3722008641915081e-05,
+      "loss": 0.7008,
+      "step": 7630
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.743109412817446,
+      "learning_rate": 1.372044335406699e-05,
+      "loss": 0.6805,
+      "step": 7631
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 21.621152475237675,
+      "learning_rate": 1.3718877960411669e-05,
+      "loss": 0.6004,
+      "step": 7632
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.985923084486366,
+      "learning_rate": 1.3717312460993632e-05,
+      "loss": 0.6458,
+      "step": 7633
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.831499571936895,
+      "learning_rate": 1.3715746855857407e-05,
+      "loss": 0.5965,
+      "step": 7634
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.378227400774456,
+      "learning_rate": 1.3714181145047512e-05,
+      "loss": 0.682,
+      "step": 7635
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 21.8385788468174,
+      "learning_rate": 1.3712615328608476e-05,
+      "loss": 0.6703,
+      "step": 7636
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.541157472343233,
+      "learning_rate": 1.3711049406584836e-05,
+      "loss": 0.6786,
+      "step": 7637
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 28.222698862644222,
+      "learning_rate": 1.3709483379021118e-05,
+      "loss": 0.6697,
+      "step": 7638
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.572778544125555,
+      "learning_rate": 1.3707917245961864e-05,
+      "loss": 0.7053,
+      "step": 7639
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.73125718557368,
+      "learning_rate": 1.3706351007451616e-05,
+      "loss": 0.6334,
+      "step": 7640
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.35657112144822,
+      "learning_rate": 1.3704784663534909e-05,
+      "loss": 0.6475,
+      "step": 7641
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.2966139470998,
+      "learning_rate": 1.3703218214256295e-05,
+      "loss": 0.7055,
+      "step": 7642
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.124191457199956,
+      "learning_rate": 1.3701651659660322e-05,
+      "loss": 0.6314,
+      "step": 7643
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.737879939036816,
+      "learning_rate": 1.3700084999791539e-05,
+      "loss": 0.6366,
+      "step": 7644
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.617573194716968,
+      "learning_rate": 1.3698518234694507e-05,
+      "loss": 0.7464,
+      "step": 7645
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.01912822833064,
+      "learning_rate": 1.3696951364413776e-05,
+      "loss": 0.654,
+      "step": 7646
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 23.066088949309595,
+      "learning_rate": 1.3695384388993915e-05,
+      "loss": 0.7175,
+      "step": 7647
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.099934492948222,
+      "learning_rate": 1.369381730847948e-05,
+      "loss": 0.6197,
+      "step": 7648
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.964683048858276,
+      "learning_rate": 1.3692250122915042e-05,
+      "loss": 0.6001,
+      "step": 7649
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.877150163643428,
+      "learning_rate": 1.3690682832345172e-05,
+      "loss": 0.6811,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.667626102270013,
+      "learning_rate": 1.3689115436814442e-05,
+      "loss": 0.6262,
+      "step": 7651
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.954731895040663,
+      "learning_rate": 1.3687547936367424e-05,
+      "loss": 0.6491,
+      "step": 7652
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.787930592887133,
+      "learning_rate": 1.36859803310487e-05,
+      "loss": 0.7097,
+      "step": 7653
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.17398208354797,
+      "learning_rate": 1.3684412620902854e-05,
+      "loss": 0.7381,
+      "step": 7654
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.47236551599288,
+      "learning_rate": 1.3682844805974466e-05,
+      "loss": 0.7036,
+      "step": 7655
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.80344222823844,
+      "learning_rate": 1.3681276886308129e-05,
+      "loss": 0.6652,
+      "step": 7656
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.467410604034352,
+      "learning_rate": 1.3679708861948429e-05,
+      "loss": 0.6617,
+      "step": 7657
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.298707388184887,
+      "learning_rate": 1.3678140732939962e-05,
+      "loss": 0.6192,
+      "step": 7658
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.506839285419602,
+      "learning_rate": 1.3676572499327322e-05,
+      "loss": 0.677,
+      "step": 7659
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.5937809117265,
+      "learning_rate": 1.3675004161155116e-05,
+      "loss": 0.688,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.131816096625517,
+      "learning_rate": 1.3673435718467936e-05,
+      "loss": 0.6501,
+      "step": 7661
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.64486544268422,
+      "learning_rate": 1.3671867171310398e-05,
+      "loss": 0.6203,
+      "step": 7662
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.945936554089887,
+      "learning_rate": 1.36702985197271e-05,
+      "loss": 0.6201,
+      "step": 7663
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.82345189607954,
+      "learning_rate": 1.3668729763762664e-05,
+      "loss": 0.6751,
+      "step": 7664
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.04941698441705,
+      "learning_rate": 1.3667160903461699e-05,
+      "loss": 0.7807,
+      "step": 7665
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.25502799540227,
+      "learning_rate": 1.366559193886882e-05,
+      "loss": 0.6261,
+      "step": 7666
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 25.027603123063447,
+      "learning_rate": 1.3664022870028655e-05,
+      "loss": 0.7956,
+      "step": 7667
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.668726061547922,
+      "learning_rate": 1.366245369698582e-05,
+      "loss": 0.6413,
+      "step": 7668
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 30.615400659331648,
+      "learning_rate": 1.3660884419784944e-05,
+      "loss": 0.6724,
+      "step": 7669
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.61355365501805,
+      "learning_rate": 1.3659315038470657e-05,
+      "loss": 0.6764,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 25.317622178655764,
+      "learning_rate": 1.3657745553087591e-05,
+      "loss": 0.6982,
+      "step": 7671
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.344091877725926,
+      "learning_rate": 1.3656175963680385e-05,
+      "loss": 0.6288,
+      "step": 7672
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 26.068582226163738,
+      "learning_rate": 1.3654606270293666e-05,
+      "loss": 0.6353,
+      "step": 7673
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.689203383696714,
+      "learning_rate": 1.3653036472972088e-05,
+      "loss": 0.67,
+      "step": 7674
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.55984097125743,
+      "learning_rate": 1.365146657176029e-05,
+      "loss": 0.6526,
+      "step": 7675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.987938094205383,
+      "learning_rate": 1.3649896566702916e-05,
+      "loss": 0.6221,
+      "step": 7676
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.697086014618723,
+      "learning_rate": 1.364832645784462e-05,
+      "loss": 0.685,
+      "step": 7677
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.614157721106924,
+      "learning_rate": 1.3646756245230056e-05,
+      "loss": 0.7203,
+      "step": 7678
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 24.74984298431076,
+      "learning_rate": 1.3645185928903874e-05,
+      "loss": 0.5951,
+      "step": 7679
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.58654666237059,
+      "learning_rate": 1.3643615508910734e-05,
+      "loss": 0.5818,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.597103354334525,
+      "learning_rate": 1.3642044985295307e-05,
+      "loss": 0.6439,
+      "step": 7681
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.33460279471346,
+      "learning_rate": 1.3640474358102247e-05,
+      "loss": 0.7883,
+      "step": 7682
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.261608012690708,
+      "learning_rate": 1.3638903627376227e-05,
+      "loss": 0.66,
+      "step": 7683
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 20.19728792220861,
+      "learning_rate": 1.3637332793161918e-05,
+      "loss": 0.7208,
+      "step": 7684
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.352946889048283,
+      "learning_rate": 1.363576185550399e-05,
+      "loss": 0.6549,
+      "step": 7685
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.845646273305002,
+      "learning_rate": 1.3634190814447124e-05,
+      "loss": 0.6223,
+      "step": 7686
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.27922149754853,
+      "learning_rate": 1.3632619670035996e-05,
+      "loss": 0.7209,
+      "step": 7687
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 24.391918448658416,
+      "learning_rate": 1.3631048422315288e-05,
+      "loss": 0.6488,
+      "step": 7688
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.26403424560335,
+      "learning_rate": 1.3629477071329689e-05,
+      "loss": 0.5684,
+      "step": 7689
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.135181514972738,
+      "learning_rate": 1.3627905617123884e-05,
+      "loss": 0.5704,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.742921262836303,
+      "learning_rate": 1.3626334059742567e-05,
+      "loss": 0.5532,
+      "step": 7691
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 26.149796006398997,
+      "learning_rate": 1.362476239923043e-05,
+      "loss": 0.6653,
+      "step": 7692
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.8417530230639,
+      "learning_rate": 1.3623190635632171e-05,
+      "loss": 0.6207,
+      "step": 7693
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.328355889939534,
+      "learning_rate": 1.3621618768992488e-05,
+      "loss": 0.5995,
+      "step": 7694
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.363896644185626,
+      "learning_rate": 1.3620046799356088e-05,
+      "loss": 0.5512,
+      "step": 7695
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.73881009138261,
+      "learning_rate": 1.3618474726767675e-05,
+      "loss": 0.6583,
+      "step": 7696
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 24.47576820942793,
+      "learning_rate": 1.3616902551271955e-05,
+      "loss": 0.7518,
+      "step": 7697
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.461286948632727,
+      "learning_rate": 1.3615330272913643e-05,
+      "loss": 0.6328,
+      "step": 7698
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 30.251965504230686,
+      "learning_rate": 1.3613757891737453e-05,
+      "loss": 0.7129,
+      "step": 7699
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.70995665462603,
+      "learning_rate": 1.3612185407788104e-05,
+      "loss": 0.5815,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.64415549573565,
+      "learning_rate": 1.3610612821110315e-05,
+      "loss": 0.6377,
+      "step": 7701
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.733259576745162,
+      "learning_rate": 1.360904013174881e-05,
+      "loss": 0.6813,
+      "step": 7702
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.172275078371847,
+      "learning_rate": 1.360746733974831e-05,
+      "loss": 0.7182,
+      "step": 7703
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.055222941510106,
+      "learning_rate": 1.3605894445153556e-05,
+      "loss": 0.7182,
+      "step": 7704
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.476187080902122,
+      "learning_rate": 1.3604321448009268e-05,
+      "loss": 0.5663,
+      "step": 7705
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.040039513294065,
+      "learning_rate": 1.3602748348360191e-05,
+      "loss": 0.6797,
+      "step": 7706
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.51030181028526,
+      "learning_rate": 1.3601175146251056e-05,
+      "loss": 0.5196,
+      "step": 7707
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.308671933221458,
+      "learning_rate": 1.3599601841726608e-05,
+      "loss": 0.6526,
+      "step": 7708
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.208077474119694,
+      "learning_rate": 1.3598028434831584e-05,
+      "loss": 0.6512,
+      "step": 7709
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.929046116626616,
+      "learning_rate": 1.3596454925610742e-05,
+      "loss": 0.6529,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 25.957286642294974,
+      "learning_rate": 1.3594881314108823e-05,
+      "loss": 0.7335,
+      "step": 7711
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.292747626136237,
+      "learning_rate": 1.3593307600370585e-05,
+      "loss": 0.6245,
+      "step": 7712
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 26.95357431894152,
+      "learning_rate": 1.359173378444078e-05,
+      "loss": 0.6813,
+      "step": 7713
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.831349505024185,
+      "learning_rate": 1.3590159866364166e-05,
+      "loss": 0.6319,
+      "step": 7714
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.648316742756453,
+      "learning_rate": 1.3588585846185502e-05,
+      "loss": 0.7271,
+      "step": 7715
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.373605419492204,
+      "learning_rate": 1.3587011723949564e-05,
+      "loss": 0.5684,
+      "step": 7716
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 28.810962346352422,
+      "learning_rate": 1.3585437499701104e-05,
+      "loss": 0.7036,
+      "step": 7717
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 24.28804142187009,
+      "learning_rate": 1.3583863173484902e-05,
+      "loss": 0.6169,
+      "step": 7718
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.76797237766549,
+      "learning_rate": 1.3582288745345728e-05,
+      "loss": 0.6861,
+      "step": 7719
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.63998005565017,
+      "learning_rate": 1.3580714215328357e-05,
+      "loss": 0.5862,
+      "step": 7720
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 24.53303063818445,
+      "learning_rate": 1.357913958347757e-05,
+      "loss": 0.6792,
+      "step": 7721
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.826674323848138,
+      "learning_rate": 1.3577564849838146e-05,
+      "loss": 0.6953,
+      "step": 7722
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.313937702809529,
+      "learning_rate": 1.3575990014454869e-05,
+      "loss": 0.5925,
+      "step": 7723
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.865880868610812,
+      "learning_rate": 1.3574415077372527e-05,
+      "loss": 0.6402,
+      "step": 7724
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.010248769582482,
+      "learning_rate": 1.3572840038635913e-05,
+      "loss": 0.6418,
+      "step": 7725
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.122639224503757,
+      "learning_rate": 1.357126489828982e-05,
+      "loss": 0.6886,
+      "step": 7726
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.33183266466506,
+      "learning_rate": 1.356968965637904e-05,
+      "loss": 0.6792,
+      "step": 7727
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.824989547549205,
+      "learning_rate": 1.3568114312948375e-05,
+      "loss": 0.5982,
+      "step": 7728
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.848025552165538,
+      "learning_rate": 1.3566538868042627e-05,
+      "loss": 0.6607,
+      "step": 7729
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 24.38716803079882,
+      "learning_rate": 1.3564963321706597e-05,
+      "loss": 0.7229,
+      "step": 7730
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.13679198167889,
+      "learning_rate": 1.35633876739851e-05,
+      "loss": 0.5827,
+      "step": 7731
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.391859682018913,
+      "learning_rate": 1.3561811924922939e-05,
+      "loss": 0.6427,
+      "step": 7732
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 28.225278217525755,
+      "learning_rate": 1.3560236074564935e-05,
+      "loss": 0.6938,
+      "step": 7733
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.87479051137091,
+      "learning_rate": 1.3558660122955894e-05,
+      "loss": 0.5692,
+      "step": 7734
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.634308373254914,
+      "learning_rate": 1.3557084070140644e-05,
+      "loss": 0.6371,
+      "step": 7735
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.132658715521163,
+      "learning_rate": 1.3555507916164002e-05,
+      "loss": 0.6319,
+      "step": 7736
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 59.79078517262928,
+      "learning_rate": 1.3553931661070796e-05,
+      "loss": 0.7993,
+      "step": 7737
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.675090233356418,
+      "learning_rate": 1.3552355304905851e-05,
+      "loss": 0.6695,
+      "step": 7738
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.822836955615433,
+      "learning_rate": 1.3550778847714001e-05,
+      "loss": 0.6723,
+      "step": 7739
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.56423243472237,
+      "learning_rate": 1.3549202289540077e-05,
+      "loss": 0.6465,
+      "step": 7740
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.799417812501762,
+      "learning_rate": 1.3547625630428919e-05,
+      "loss": 0.6235,
+      "step": 7741
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.001394396734433,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.7434,
+      "step": 7742
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.618670841687017,
+      "learning_rate": 1.3544472009574243e-05,
+      "loss": 0.6331,
+      "step": 7743
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.760969114618792,
+      "learning_rate": 1.354289504792042e-05,
+      "loss": 0.6606,
+      "step": 7744
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.166953750820301,
+      "learning_rate": 1.3541317985508733e-05,
+      "loss": 0.7009,
+      "step": 7745
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.23684946609206,
+      "learning_rate": 1.3539740822384034e-05,
+      "loss": 0.6593,
+      "step": 7746
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.166807897073213,
+      "learning_rate": 1.3538163558591175e-05,
+      "loss": 0.6959,
+      "step": 7747
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.08195940191766,
+      "learning_rate": 1.3536586194175019e-05,
+      "loss": 0.7125,
+      "step": 7748
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.443448226338507,
+      "learning_rate": 1.3535008729180414e-05,
+      "loss": 0.7189,
+      "step": 7749
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.9016326289895,
+      "learning_rate": 1.3533431163652235e-05,
+      "loss": 0.7998,
+      "step": 7750
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.443605478528706,
+      "learning_rate": 1.3531853497635336e-05,
+      "loss": 0.6639,
+      "step": 7751
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.645828533533575,
+      "learning_rate": 1.3530275731174591e-05,
+      "loss": 0.694,
+      "step": 7752
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 27.10189880268002,
+      "learning_rate": 1.352869786431487e-05,
+      "loss": 0.6703,
+      "step": 7753
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.97391631528998,
+      "learning_rate": 1.3527119897101046e-05,
+      "loss": 0.6051,
+      "step": 7754
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.39083757071754,
+      "learning_rate": 1.3525541829577997e-05,
+      "loss": 0.6977,
+      "step": 7755
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.660695229985553,
+      "learning_rate": 1.3523963661790597e-05,
+      "loss": 0.566,
+      "step": 7756
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.514916987503454,
+      "learning_rate": 1.3522385393783735e-05,
+      "loss": 0.6181,
+      "step": 7757
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.44983075362788,
+      "learning_rate": 1.3520807025602291e-05,
+      "loss": 0.6142,
+      "step": 7758
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.112730421360908,
+      "learning_rate": 1.3519228557291156e-05,
+      "loss": 0.7191,
+      "step": 7759
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.87400981712163,
+      "learning_rate": 1.3517649988895221e-05,
+      "loss": 0.641,
+      "step": 7760
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.363589170268963,
+      "learning_rate": 1.3516071320459374e-05,
+      "loss": 0.6954,
+      "step": 7761
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.762579793334773,
+      "learning_rate": 1.3514492552028518e-05,
+      "loss": 0.7501,
+      "step": 7762
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.25704190722339,
+      "learning_rate": 1.351291368364755e-05,
+      "loss": 0.6863,
+      "step": 7763
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.79564088841133,
+      "learning_rate": 1.3511334715361371e-05,
+      "loss": 0.6296,
+      "step": 7764
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.96447351733195,
+      "learning_rate": 1.3509755647214888e-05,
+      "loss": 0.6559,
+      "step": 7765
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.17274124564975,
+      "learning_rate": 1.3508176479253003e-05,
+      "loss": 0.6416,
+      "step": 7766
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.323950186945373,
+      "learning_rate": 1.3506597211520638e-05,
+      "loss": 0.6761,
+      "step": 7767
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.553193918228306,
+      "learning_rate": 1.3505017844062693e-05,
+      "loss": 0.6436,
+      "step": 7768
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.341149612035274,
+      "learning_rate": 1.3503438376924096e-05,
+      "loss": 0.6985,
+      "step": 7769
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.187834744797613,
+      "learning_rate": 1.350185881014976e-05,
+      "loss": 0.7258,
+      "step": 7770
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.118010345781553,
+      "learning_rate": 1.3500279143784606e-05,
+      "loss": 0.5968,
+      "step": 7771
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.766183982076797,
+      "learning_rate": 1.3498699377873563e-05,
+      "loss": 0.6071,
+      "step": 7772
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.648475346017172,
+      "learning_rate": 1.3497119512461555e-05,
+      "loss": 0.5594,
+      "step": 7773
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.256902129731618,
+      "learning_rate": 1.3495539547593512e-05,
+      "loss": 0.6319,
+      "step": 7774
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.95187187555169,
+      "learning_rate": 1.3493959483314373e-05,
+      "loss": 0.73,
+      "step": 7775
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 26.797598910771534,
+      "learning_rate": 1.349237931966907e-05,
+      "loss": 0.7152,
+      "step": 7776
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.137483614216,
+      "learning_rate": 1.3490799056702542e-05,
+      "loss": 0.6849,
+      "step": 7777
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 30.073841916592357,
+      "learning_rate": 1.3489218694459727e-05,
+      "loss": 0.6337,
+      "step": 7778
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.22687657442029,
+      "learning_rate": 1.3487638232985575e-05,
+      "loss": 0.7377,
+      "step": 7779
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.285290256993726,
+      "learning_rate": 1.3486057672325035e-05,
+      "loss": 0.7236,
+      "step": 7780
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 39.5672733664932,
+      "learning_rate": 1.348447701252305e-05,
+      "loss": 0.8154,
+      "step": 7781
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 23.84282254868855,
+      "learning_rate": 1.348289625362458e-05,
+      "loss": 0.7073,
+      "step": 7782
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.170662522190558,
+      "learning_rate": 1.3481315395674577e-05,
+      "loss": 0.6956,
+      "step": 7783
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.927685363625908,
+      "learning_rate": 1.3479734438718002e-05,
+      "loss": 0.581,
+      "step": 7784
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.836569437461584,
+      "learning_rate": 1.3478153382799812e-05,
+      "loss": 0.6611,
+      "step": 7785
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.520857340134633,
+      "learning_rate": 1.3476572227964976e-05,
+      "loss": 0.6539,
+      "step": 7786
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.932599178936254,
+      "learning_rate": 1.347499097425846e-05,
+      "loss": 0.6455,
+      "step": 7787
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.924515073712305,
+      "learning_rate": 1.3473409621725233e-05,
+      "loss": 0.6104,
+      "step": 7788
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 22.843174330924167,
+      "learning_rate": 1.3471828170410268e-05,
+      "loss": 0.6356,
+      "step": 7789
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.22999083816796,
+      "learning_rate": 1.3470246620358543e-05,
+      "loss": 0.645,
+      "step": 7790
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.155008353193594,
+      "learning_rate": 1.3468664971615031e-05,
+      "loss": 0.6822,
+      "step": 7791
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 23.88314308827424,
+      "learning_rate": 1.3467083224224719e-05,
+      "loss": 0.692,
+      "step": 7792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.887869052967515,
+      "learning_rate": 1.3465501378232586e-05,
+      "loss": 0.7156,
+      "step": 7793
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.55371340810835,
+      "learning_rate": 1.3463919433683624e-05,
+      "loss": 0.7328,
+      "step": 7794
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.815173644195019,
+      "learning_rate": 1.3462337390622814e-05,
+      "loss": 0.6379,
+      "step": 7795
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.7295557383263,
+      "learning_rate": 1.3460755249095161e-05,
+      "loss": 0.6373,
+      "step": 7796
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.688595746460194,
+      "learning_rate": 1.345917300914565e-05,
+      "loss": 0.7018,
+      "step": 7797
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.029232767640135,
+      "learning_rate": 1.3457590670819283e-05,
+      "loss": 0.6712,
+      "step": 7798
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.215186818242362,
+      "learning_rate": 1.3456008234161057e-05,
+      "loss": 0.6219,
+      "step": 7799
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 28.221219862978312,
+      "learning_rate": 1.3454425699215982e-05,
+      "loss": 0.6038,
+      "step": 7800
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.461458556760373,
+      "learning_rate": 1.3452843066029058e-05,
+      "loss": 0.5516,
+      "step": 7801
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.928907133449648,
+      "learning_rate": 1.3451260334645299e-05,
+      "loss": 0.66,
+      "step": 7802
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.87503209949902,
+      "learning_rate": 1.3449677505109716e-05,
+      "loss": 0.5877,
+      "step": 7803
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 27.80302327163075,
+      "learning_rate": 1.3448094577467323e-05,
+      "loss": 0.6431,
+      "step": 7804
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 39.054457456823506,
+      "learning_rate": 1.3446511551763133e-05,
+      "loss": 0.6847,
+      "step": 7805
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.944321259074611,
+      "learning_rate": 1.3444928428042175e-05,
+      "loss": 0.724,
+      "step": 7806
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.451371262771637,
+      "learning_rate": 1.3443345206349466e-05,
+      "loss": 0.6378,
+      "step": 7807
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.279431429503372,
+      "learning_rate": 1.3441761886730033e-05,
+      "loss": 0.6591,
+      "step": 7808
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 30.09793328981426,
+      "learning_rate": 1.3440178469228906e-05,
+      "loss": 0.7033,
+      "step": 7809
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 25.74143649237283,
+      "learning_rate": 1.3438594953891113e-05,
+      "loss": 0.606,
+      "step": 7810
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 23.527783235374834,
+      "learning_rate": 1.3437011340761697e-05,
+      "loss": 0.7218,
+      "step": 7811
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.67283128367126,
+      "learning_rate": 1.3435427629885684e-05,
+      "loss": 0.6986,
+      "step": 7812
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.904130784686412,
+      "learning_rate": 1.343384382130812e-05,
+      "loss": 0.7566,
+      "step": 7813
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 21.393523155060652,
+      "learning_rate": 1.3432259915074048e-05,
+      "loss": 0.6636,
+      "step": 7814
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.681988834630545,
+      "learning_rate": 1.3430675911228512e-05,
+      "loss": 0.5694,
+      "step": 7815
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.390931557749273,
+      "learning_rate": 1.342909180981656e-05,
+      "loss": 0.6919,
+      "step": 7816
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.62703428212765,
+      "learning_rate": 1.3427507610883242e-05,
+      "loss": 0.5966,
+      "step": 7817
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.082857439457646,
+      "learning_rate": 1.342592331447361e-05,
+      "loss": 0.5937,
+      "step": 7818
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.16480701483057,
+      "learning_rate": 1.342433892063273e-05,
+      "loss": 0.6725,
+      "step": 7819
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.29070869203043,
+      "learning_rate": 1.3422754429405647e-05,
+      "loss": 0.6131,
+      "step": 7820
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.209781259413884,
+      "learning_rate": 1.3421169840837433e-05,
+      "loss": 0.6267,
+      "step": 7821
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.400291058307083,
+      "learning_rate": 1.341958515497315e-05,
+      "loss": 0.6059,
+      "step": 7822
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.10039043456538,
+      "learning_rate": 1.3418000371857866e-05,
+      "loss": 0.7005,
+      "step": 7823
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 40.55081024845028,
+      "learning_rate": 1.341641549153665e-05,
+      "loss": 0.7274,
+      "step": 7824
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.49873915770387,
+      "learning_rate": 1.3414830514054578e-05,
+      "loss": 0.6495,
+      "step": 7825
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.50741681027481,
+      "learning_rate": 1.3413245439456721e-05,
+      "loss": 0.5725,
+      "step": 7826
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.86678812325382,
+      "learning_rate": 1.3411660267788163e-05,
+      "loss": 0.6922,
+      "step": 7827
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.908302403716277,
+      "learning_rate": 1.341007499909398e-05,
+      "loss": 0.5855,
+      "step": 7828
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.360341223899788,
+      "learning_rate": 1.3408489633419261e-05,
+      "loss": 0.7458,
+      "step": 7829
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.839485972258483,
+      "learning_rate": 1.340690417080909e-05,
+      "loss": 0.6703,
+      "step": 7830
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.7909383248384,
+      "learning_rate": 1.3405318611308558e-05,
+      "loss": 0.5847,
+      "step": 7831
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.451240739427591,
+      "learning_rate": 1.3403732954962755e-05,
+      "loss": 0.6302,
+      "step": 7832
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.27548092829371,
+      "learning_rate": 1.3402147201816779e-05,
+      "loss": 0.6838,
+      "step": 7833
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.857950330962739,
+      "learning_rate": 1.3400561351915728e-05,
+      "loss": 0.6174,
+      "step": 7834
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.737197755700027,
+      "learning_rate": 1.33989754053047e-05,
+      "loss": 0.6696,
+      "step": 7835
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.101315306506343,
+      "learning_rate": 1.3397389362028802e-05,
+      "loss": 0.573,
+      "step": 7836
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.274495057910308,
+      "learning_rate": 1.3395803222133133e-05,
+      "loss": 0.741,
+      "step": 7837
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.494150524395614,
+      "learning_rate": 1.3394216985662813e-05,
+      "loss": 0.7125,
+      "step": 7838
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.088645319283465,
+      "learning_rate": 1.3392630652662943e-05,
+      "loss": 0.6642,
+      "step": 7839
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.428706659746524,
+      "learning_rate": 1.3391044223178641e-05,
+      "loss": 0.6716,
+      "step": 7840
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.56534645434312,
+      "learning_rate": 1.338945769725503e-05,
+      "loss": 0.5711,
+      "step": 7841
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.620159165927667,
+      "learning_rate": 1.3387871074937222e-05,
+      "loss": 0.7104,
+      "step": 7842
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.010273246035972,
+      "learning_rate": 1.3386284356270342e-05,
+      "loss": 0.6918,
+      "step": 7843
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 24.633946700862996,
+      "learning_rate": 1.3384697541299518e-05,
+      "loss": 0.6386,
+      "step": 7844
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.301930476555217,
+      "learning_rate": 1.3383110630069874e-05,
+      "loss": 0.6633,
+      "step": 7845
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.57836072690391,
+      "learning_rate": 1.3381523622626544e-05,
+      "loss": 0.7097,
+      "step": 7846
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.97680065682913,
+      "learning_rate": 1.3379936519014658e-05,
+      "loss": 0.6742,
+      "step": 7847
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.171551114944574,
+      "learning_rate": 1.3378349319279358e-05,
+      "loss": 0.6585,
+      "step": 7848
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 25.342515185104073,
+      "learning_rate": 1.3376762023465775e-05,
+      "loss": 0.7317,
+      "step": 7849
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.011294051955705,
+      "learning_rate": 1.337517463161906e-05,
+      "loss": 0.6369,
+      "step": 7850
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.294206516236073,
+      "learning_rate": 1.337358714378435e-05,
+      "loss": 0.6676,
+      "step": 7851
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.85916209683056,
+      "learning_rate": 1.3371999560006796e-05,
+      "loss": 0.6662,
+      "step": 7852
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.731471469049188,
+      "learning_rate": 1.3370411880331549e-05,
+      "loss": 0.6359,
+      "step": 7853
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.577569937946773,
+      "learning_rate": 1.3368824104803755e-05,
+      "loss": 0.5964,
+      "step": 7854
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.292603830930343,
+      "learning_rate": 1.3367236233468579e-05,
+      "loss": 0.6249,
+      "step": 7855
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.053976986132536,
+      "learning_rate": 1.3365648266371171e-05,
+      "loss": 0.6138,
+      "step": 7856
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.281809365988988,
+      "learning_rate": 1.3364060203556697e-05,
+      "loss": 0.6376,
+      "step": 7857
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.587373227775677,
+      "learning_rate": 1.3362472045070318e-05,
+      "loss": 0.607,
+      "step": 7858
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.037851861391571,
+      "learning_rate": 1.3360883790957202e-05,
+      "loss": 0.6241,
+      "step": 7859
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 24.69572943528869,
+      "learning_rate": 1.3359295441262516e-05,
+      "loss": 0.677,
+      "step": 7860
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 25.246291058144273,
+      "learning_rate": 1.3357706996031434e-05,
+      "loss": 0.7144,
+      "step": 7861
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 26.800505409439165,
+      "learning_rate": 1.3356118455309127e-05,
+      "loss": 0.6671,
+      "step": 7862
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.097743590918743,
+      "learning_rate": 1.3354529819140777e-05,
+      "loss": 0.6569,
+      "step": 7863
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.808594838449327,
+      "learning_rate": 1.3352941087571558e-05,
+      "loss": 0.6306,
+      "step": 7864
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.444113129241654,
+      "learning_rate": 1.3351352260646662e-05,
+      "loss": 0.6454,
+      "step": 7865
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.65124130678045,
+      "learning_rate": 1.3349763338411261e-05,
+      "loss": 0.6305,
+      "step": 7866
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.59297812853889,
+      "learning_rate": 1.3348174320910554e-05,
+      "loss": 0.5736,
+      "step": 7867
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.205577460427556,
+      "learning_rate": 1.3346585208189727e-05,
+      "loss": 0.5189,
+      "step": 7868
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.536255799702765,
+      "learning_rate": 1.3344996000293976e-05,
+      "loss": 0.6064,
+      "step": 7869
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.37655068788581,
+      "learning_rate": 1.3343406697268493e-05,
+      "loss": 0.7551,
+      "step": 7870
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 22.403108360591908,
+      "learning_rate": 1.334181729915848e-05,
+      "loss": 0.6887,
+      "step": 7871
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.3271059961191,
+      "learning_rate": 1.3340227806009138e-05,
+      "loss": 0.6999,
+      "step": 7872
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.470230556526593,
+      "learning_rate": 1.3338638217865673e-05,
+      "loss": 0.6047,
+      "step": 7873
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.523624682434654,
+      "learning_rate": 1.333704853477329e-05,
+      "loss": 0.6704,
+      "step": 7874
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.07322435052991,
+      "learning_rate": 1.3335458756777195e-05,
+      "loss": 0.6204,
+      "step": 7875
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.890217170444934,
+      "learning_rate": 1.3333868883922606e-05,
+      "loss": 0.6307,
+      "step": 7876
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 23.18885976989232,
+      "learning_rate": 1.3332278916254739e-05,
+      "loss": 0.6319,
+      "step": 7877
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 22.45499080585571,
+      "learning_rate": 1.3330688853818808e-05,
+      "loss": 0.6339,
+      "step": 7878
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.178983751781235,
+      "learning_rate": 1.332909869666003e-05,
+      "loss": 0.5528,
+      "step": 7879
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.828388373565346,
+      "learning_rate": 1.3327508444823637e-05,
+      "loss": 0.6013,
+      "step": 7880
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.98520209943347,
+      "learning_rate": 1.3325918098354848e-05,
+      "loss": 0.6541,
+      "step": 7881
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.243383701530203,
+      "learning_rate": 1.33243276572989e-05,
+      "loss": 0.6377,
+      "step": 7882
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.21215025543746,
+      "learning_rate": 1.3322737121701011e-05,
+      "loss": 0.6345,
+      "step": 7883
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.014958809743952,
+      "learning_rate": 1.3321146491606423e-05,
+      "loss": 0.6191,
+      "step": 7884
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.491245505540082,
+      "learning_rate": 1.3319555767060373e-05,
+      "loss": 0.6163,
+      "step": 7885
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.035706632181144,
+      "learning_rate": 1.33179649481081e-05,
+      "loss": 0.6931,
+      "step": 7886
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.461097830565055,
+      "learning_rate": 1.3316374034794844e-05,
+      "loss": 0.7712,
+      "step": 7887
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.399169550948923,
+      "learning_rate": 1.331478302716585e-05,
+      "loss": 0.6629,
+      "step": 7888
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.88543885916358,
+      "learning_rate": 1.3313191925266362e-05,
+      "loss": 0.721,
+      "step": 7889
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.65867786762948,
+      "learning_rate": 1.3311600729141641e-05,
+      "loss": 0.6391,
+      "step": 7890
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.215032964594407,
+      "learning_rate": 1.3310009438836925e-05,
+      "loss": 0.6456,
+      "step": 7891
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.163176239584654,
+      "learning_rate": 1.3308418054397485e-05,
+      "loss": 0.5854,
+      "step": 7892
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 22.397403010577648,
+      "learning_rate": 1.3306826575868565e-05,
+      "loss": 0.6801,
+      "step": 7893
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.50753423426254,
+      "learning_rate": 1.3305235003295434e-05,
+      "loss": 0.6,
+      "step": 7894
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.560138784772928,
+      "learning_rate": 1.330364333672335e-05,
+      "loss": 0.5809,
+      "step": 7895
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.55651368342618,
+      "learning_rate": 1.3302051576197586e-05,
+      "loss": 0.6925,
+      "step": 7896
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.831075461221058,
+      "learning_rate": 1.3300459721763405e-05,
+      "loss": 0.6789,
+      "step": 7897
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.710369055180365,
+      "learning_rate": 1.329886777346608e-05,
+      "loss": 0.6549,
+      "step": 7898
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.264327182730028,
+      "learning_rate": 1.3297275731350885e-05,
+      "loss": 0.6392,
+      "step": 7899
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.441937706898234,
+      "learning_rate": 1.3295683595463099e-05,
+      "loss": 0.6144,
+      "step": 7900
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.786126687444444,
+      "learning_rate": 1.3294091365847998e-05,
+      "loss": 0.6627,
+      "step": 7901
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 24.377213004717987,
+      "learning_rate": 1.3292499042550869e-05,
+      "loss": 0.7348,
+      "step": 7902
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 23.047238160521076,
+      "learning_rate": 1.3290906625616988e-05,
+      "loss": 0.6707,
+      "step": 7903
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 24.753436743419407,
+      "learning_rate": 1.3289314115091649e-05,
+      "loss": 0.6353,
+      "step": 7904
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.532643287509494,
+      "learning_rate": 1.3287721511020145e-05,
+      "loss": 0.6173,
+      "step": 7905
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.753650588851634,
+      "learning_rate": 1.3286128813447759e-05,
+      "loss": 0.6767,
+      "step": 7906
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 27.803578055323058,
+      "learning_rate": 1.3284536022419796e-05,
+      "loss": 0.7903,
+      "step": 7907
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.273160275093604,
+      "learning_rate": 1.3282943137981546e-05,
+      "loss": 0.6562,
+      "step": 7908
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.305561522661897,
+      "learning_rate": 1.3281350160178317e-05,
+      "loss": 0.6933,
+      "step": 7909
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.470618459293842,
+      "learning_rate": 1.3279757089055404e-05,
+      "loss": 0.6052,
+      "step": 7910
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.91177093019644,
+      "learning_rate": 1.3278163924658123e-05,
+      "loss": 0.622,
+      "step": 7911
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.094124083530875,
+      "learning_rate": 1.3276570667031777e-05,
+      "loss": 0.6626,
+      "step": 7912
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.997377877039073,
+      "learning_rate": 1.3274977316221678e-05,
+      "loss": 0.6484,
+      "step": 7913
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.02435372977003,
+      "learning_rate": 1.3273383872273137e-05,
+      "loss": 0.7073,
+      "step": 7914
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.765552197205572,
+      "learning_rate": 1.3271790335231476e-05,
+      "loss": 0.6007,
+      "step": 7915
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.841112129305039,
+      "learning_rate": 1.327019670514201e-05,
+      "loss": 0.6273,
+      "step": 7916
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.69000467959322,
+      "learning_rate": 1.3268602982050064e-05,
+      "loss": 0.611,
+      "step": 7917
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.28029239987682,
+      "learning_rate": 1.3267009166000961e-05,
+      "loss": 0.7759,
+      "step": 7918
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.22753663293937,
+      "learning_rate": 1.3265415257040031e-05,
+      "loss": 0.7013,
+      "step": 7919
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.891110487103646,
+      "learning_rate": 1.3263821255212598e-05,
+      "loss": 0.5992,
+      "step": 7920
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.740556329509829,
+      "learning_rate": 1.3262227160564001e-05,
+      "loss": 0.588,
+      "step": 7921
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.409217790425616,
+      "learning_rate": 1.3260632973139569e-05,
+      "loss": 0.5627,
+      "step": 7922
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.057686791152825,
+      "learning_rate": 1.3259038692984643e-05,
+      "loss": 0.7109,
+      "step": 7923
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 23.50835141204126,
+      "learning_rate": 1.3257444320144566e-05,
+      "loss": 0.6567,
+      "step": 7924
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.554456563616693,
+      "learning_rate": 1.3255849854664674e-05,
+      "loss": 0.6055,
+      "step": 7925
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.78730080790689,
+      "learning_rate": 1.325425529659032e-05,
+      "loss": 0.6669,
+      "step": 7926
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 26.201441033424327,
+      "learning_rate": 1.3252660645966846e-05,
+      "loss": 0.6913,
+      "step": 7927
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.291437704780622,
+      "learning_rate": 1.3251065902839608e-05,
+      "loss": 0.5889,
+      "step": 7928
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.25152501407792,
+      "learning_rate": 1.3249471067253957e-05,
+      "loss": 0.6122,
+      "step": 7929
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.182677067842906,
+      "learning_rate": 1.3247876139255249e-05,
+      "loss": 0.666,
+      "step": 7930
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 28.480784499600276,
+      "learning_rate": 1.3246281118888843e-05,
+      "loss": 0.6199,
+      "step": 7931
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.85401371600529,
+      "learning_rate": 1.3244686006200105e-05,
+      "loss": 0.7349,
+      "step": 7932
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.55603224050496,
+      "learning_rate": 1.324309080123439e-05,
+      "loss": 0.6342,
+      "step": 7933
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.490307000093008,
+      "learning_rate": 1.3241495504037074e-05,
+      "loss": 0.6602,
+      "step": 7934
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.061803130403938,
+      "learning_rate": 1.323990011465352e-05,
+      "loss": 0.6368,
+      "step": 7935
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.654416692014685,
+      "learning_rate": 1.3238304633129102e-05,
+      "loss": 0.6791,
+      "step": 7936
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.013629017532137,
+      "learning_rate": 1.3236709059509194e-05,
+      "loss": 0.6531,
+      "step": 7937
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 17.749109220651647,
+      "learning_rate": 1.3235113393839174e-05,
+      "loss": 0.6401,
+      "step": 7938
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.597698698166603,
+      "learning_rate": 1.3233517636164423e-05,
+      "loss": 0.5637,
+      "step": 7939
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.845483500689486,
+      "learning_rate": 1.3231921786530323e-05,
+      "loss": 0.6072,
+      "step": 7940
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 23.05340353228807,
+      "learning_rate": 1.3230325844982254e-05,
+      "loss": 0.6468,
+      "step": 7941
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.41042156977892,
+      "learning_rate": 1.3228729811565608e-05,
+      "loss": 0.6402,
+      "step": 7942
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.345713386359616,
+      "learning_rate": 1.3227133686325777e-05,
+      "loss": 0.6406,
+      "step": 7943
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 17.9661174778001,
+      "learning_rate": 1.3225537469308152e-05,
+      "loss": 0.8235,
+      "step": 7944
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.523520140653662,
+      "learning_rate": 1.3223941160558127e-05,
+      "loss": 0.5726,
+      "step": 7945
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.432950747673917,
+      "learning_rate": 1.3222344760121105e-05,
+      "loss": 0.6464,
+      "step": 7946
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.092475980503007,
+      "learning_rate": 1.3220748268042478e-05,
+      "loss": 0.7446,
+      "step": 7947
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.004634284806414,
+      "learning_rate": 1.3219151684367656e-05,
+      "loss": 0.6777,
+      "step": 7948
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.929152988664814,
+      "learning_rate": 1.3217555009142044e-05,
+      "loss": 0.602,
+      "step": 7949
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.8817185716788,
+      "learning_rate": 1.3215958242411048e-05,
+      "loss": 0.7217,
+      "step": 7950
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.319623139418203,
+      "learning_rate": 1.3214361384220084e-05,
+      "loss": 0.6298,
+      "step": 7951
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.567336582154155,
+      "learning_rate": 1.321276443461456e-05,
+      "loss": 0.7221,
+      "step": 7952
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 29.23238888410719,
+      "learning_rate": 1.3211167393639898e-05,
+      "loss": 0.6523,
+      "step": 7953
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.35537994463111,
+      "learning_rate": 1.320957026134151e-05,
+      "loss": 0.6673,
+      "step": 7954
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.716767447961214,
+      "learning_rate": 1.3207973037764822e-05,
+      "loss": 0.6886,
+      "step": 7955
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.993860901341591,
+      "learning_rate": 1.3206375722955261e-05,
+      "loss": 0.6252,
+      "step": 7956
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.21098975756675,
+      "learning_rate": 1.3204778316958247e-05,
+      "loss": 0.6027,
+      "step": 7957
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.75832085191696,
+      "learning_rate": 1.3203180819819213e-05,
+      "loss": 0.6426,
+      "step": 7958
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.815084283267451,
+      "learning_rate": 1.320158323158359e-05,
+      "loss": 0.6695,
+      "step": 7959
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.100789063992094,
+      "learning_rate": 1.319998555229681e-05,
+      "loss": 0.603,
+      "step": 7960
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 25.813543146220944,
+      "learning_rate": 1.3198387782004315e-05,
+      "loss": 0.7146,
+      "step": 7961
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.048277472308529,
+      "learning_rate": 1.3196789920751546e-05,
+      "loss": 0.6135,
+      "step": 7962
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 25.871516541486958,
+      "learning_rate": 1.3195191968583938e-05,
+      "loss": 0.689,
+      "step": 7963
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.541956828209123,
+      "learning_rate": 1.3193593925546941e-05,
+      "loss": 0.5964,
+      "step": 7964
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 24.38098700551929,
+      "learning_rate": 1.3191995791686002e-05,
+      "loss": 0.6473,
+      "step": 7965
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.753004758398816,
+      "learning_rate": 1.3190397567046569e-05,
+      "loss": 0.6561,
+      "step": 7966
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.73712064953473,
+      "learning_rate": 1.3188799251674092e-05,
+      "loss": 0.6794,
+      "step": 7967
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.066982157870687,
+      "learning_rate": 1.3187200845614035e-05,
+      "loss": 0.6959,
+      "step": 7968
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.87475163601495,
+      "learning_rate": 1.3185602348911846e-05,
+      "loss": 0.5211,
+      "step": 7969
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.02567790223042,
+      "learning_rate": 1.318400376161299e-05,
+      "loss": 0.5565,
+      "step": 7970
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.72316915540688,
+      "learning_rate": 1.3182405083762935e-05,
+      "loss": 0.6656,
+      "step": 7971
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.445719403217705,
+      "learning_rate": 1.3180806315407137e-05,
+      "loss": 0.6301,
+      "step": 7972
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.544344864032784,
+      "learning_rate": 1.3179207456591067e-05,
+      "loss": 0.6285,
+      "step": 7973
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.56743709737352,
+      "learning_rate": 1.31776085073602e-05,
+      "loss": 0.6398,
+      "step": 7974
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.49886792134699,
+      "learning_rate": 1.3176009467760005e-05,
+      "loss": 0.6281,
+      "step": 7975
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.215587319370943,
+      "learning_rate": 1.3174410337835956e-05,
+      "loss": 0.6139,
+      "step": 7976
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.98417351690066,
+      "learning_rate": 1.3172811117633536e-05,
+      "loss": 0.622,
+      "step": 7977
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.407426807415938,
+      "learning_rate": 1.3171211807198225e-05,
+      "loss": 0.6029,
+      "step": 7978
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.876736337735665,
+      "learning_rate": 1.3169612406575501e-05,
+      "loss": 0.5869,
+      "step": 7979
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.51903827206621,
+      "learning_rate": 1.316801291581086e-05,
+      "loss": 0.6657,
+      "step": 7980
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.993202123709066,
+      "learning_rate": 1.316641333494978e-05,
+      "loss": 0.622,
+      "step": 7981
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.914745675136928,
+      "learning_rate": 1.316481366403776e-05,
+      "loss": 0.6857,
+      "step": 7982
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.067066034141959,
+      "learning_rate": 1.316321390312029e-05,
+      "loss": 0.5794,
+      "step": 7983
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.279088552644822,
+      "learning_rate": 1.3161614052242869e-05,
+      "loss": 0.644,
+      "step": 7984
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.278151384964875,
+      "learning_rate": 1.3160014111450989e-05,
+      "loss": 0.6082,
+      "step": 7985
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 24.751250024738408,
+      "learning_rate": 1.3158414080790158e-05,
+      "loss": 0.6466,
+      "step": 7986
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 25.5915203357572,
+      "learning_rate": 1.315681396030588e-05,
+      "loss": 0.6963,
+      "step": 7987
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.11976135716292,
+      "learning_rate": 1.3155213750043658e-05,
+      "loss": 0.6684,
+      "step": 7988
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.56733587035152,
+      "learning_rate": 1.3153613450049006e-05,
+      "loss": 0.6152,
+      "step": 7989
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.830153845485786,
+      "learning_rate": 1.3152013060367428e-05,
+      "loss": 0.8028,
+      "step": 7990
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.26611231509897,
+      "learning_rate": 1.3150412581044446e-05,
+      "loss": 0.672,
+      "step": 7991
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.139062711391306,
+      "learning_rate": 1.314881201212557e-05,
+      "loss": 0.6382,
+      "step": 7992
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.39103672851601,
+      "learning_rate": 1.3147211353656327e-05,
+      "loss": 0.5999,
+      "step": 7993
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.83432799329184,
+      "learning_rate": 1.314561060568223e-05,
+      "loss": 0.6059,
+      "step": 7994
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.23246007104005,
+      "learning_rate": 1.3144009768248808e-05,
+      "loss": 0.655,
+      "step": 7995
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.06712140910257,
+      "learning_rate": 1.3142408841401586e-05,
+      "loss": 0.6001,
+      "step": 7996
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.775752023871869,
+      "learning_rate": 1.3140807825186098e-05,
+      "loss": 0.5959,
+      "step": 7997
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 25.43728491038459,
+      "learning_rate": 1.313920671964787e-05,
+      "loss": 0.6157,
+      "step": 7998
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.1862003240563,
+      "learning_rate": 1.3137605524832442e-05,
+      "loss": 0.7027,
+      "step": 7999
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.532901553101294,
+      "learning_rate": 1.3136004240785348e-05,
+      "loss": 0.6413,
+      "step": 8000
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.918692224404275,
+      "learning_rate": 1.3134402867552125e-05,
+      "loss": 0.6516,
+      "step": 8001
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 26.022916189083826,
+      "learning_rate": 1.3132801405178322e-05,
+      "loss": 0.7436,
+      "step": 8002
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 28.172128322898967,
+      "learning_rate": 1.3131199853709477e-05,
+      "loss": 0.6289,
+      "step": 8003
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 33.08401007639743,
+      "learning_rate": 1.3129598213191139e-05,
+      "loss": 0.6346,
+      "step": 8004
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.320825942732977,
+      "learning_rate": 1.3127996483668862e-05,
+      "loss": 0.6322,
+      "step": 8005
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.723695398040176,
+      "learning_rate": 1.3126394665188191e-05,
+      "loss": 0.5866,
+      "step": 8006
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.3880531225369,
+      "learning_rate": 1.3124792757794687e-05,
+      "loss": 0.6105,
+      "step": 8007
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.773889711396897,
+      "learning_rate": 1.3123190761533903e-05,
+      "loss": 0.6621,
+      "step": 8008
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.007287362743273,
+      "learning_rate": 1.3121588676451405e-05,
+      "loss": 0.6335,
+      "step": 8009
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.835750770914707,
+      "learning_rate": 1.3119986502592747e-05,
+      "loss": 0.8118,
+      "step": 8010
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.21119332996405,
+      "learning_rate": 1.3118384240003498e-05,
+      "loss": 0.6193,
+      "step": 8011
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.051947208073248,
+      "learning_rate": 1.311678188872923e-05,
+      "loss": 0.6799,
+      "step": 8012
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.9895501395088,
+      "learning_rate": 1.3115179448815503e-05,
+      "loss": 0.7574,
+      "step": 8013
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.752326133175949,
+      "learning_rate": 1.3113576920307896e-05,
+      "loss": 0.6685,
+      "step": 8014
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 29.550321243598614,
+      "learning_rate": 1.3111974303251984e-05,
+      "loss": 0.6871,
+      "step": 8015
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.699380170060376,
+      "learning_rate": 1.3110371597693344e-05,
+      "loss": 0.5708,
+      "step": 8016
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.613952519430622,
+      "learning_rate": 1.3108768803677554e-05,
+      "loss": 0.6425,
+      "step": 8017
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.555677728944183,
+      "learning_rate": 1.3107165921250197e-05,
+      "loss": 0.7518,
+      "step": 8018
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.73028546974145,
+      "learning_rate": 1.3105562950456857e-05,
+      "loss": 0.6634,
+      "step": 8019
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.786130834689537,
+      "learning_rate": 1.310395989134313e-05,
+      "loss": 0.702,
+      "step": 8020
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.777253248459047,
+      "learning_rate": 1.3102356743954594e-05,
+      "loss": 0.6694,
+      "step": 8021
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.333587093903443,
+      "learning_rate": 1.3100753508336852e-05,
+      "loss": 0.6967,
+      "step": 8022
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.969389769343948,
+      "learning_rate": 1.3099150184535487e-05,
+      "loss": 0.6445,
+      "step": 8023
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.006206591118794,
+      "learning_rate": 1.309754677259611e-05,
+      "loss": 0.6807,
+      "step": 8024
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.049078108190155,
+      "learning_rate": 1.3095943272564312e-05,
+      "loss": 0.618,
+      "step": 8025
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.950898794490673,
+      "learning_rate": 1.3094339684485698e-05,
+      "loss": 0.5541,
+      "step": 8026
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.673019773239924,
+      "learning_rate": 1.3092736008405875e-05,
+      "loss": 0.6066,
+      "step": 8027
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.261889639083932,
+      "learning_rate": 1.3091132244370449e-05,
+      "loss": 0.5696,
+      "step": 8028
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.930248221286238,
+      "learning_rate": 1.3089528392425029e-05,
+      "loss": 0.581,
+      "step": 8029
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.690711249989143,
+      "learning_rate": 1.3087924452615228e-05,
+      "loss": 0.7141,
+      "step": 8030
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.32094288704457,
+      "learning_rate": 1.3086320424986662e-05,
+      "loss": 0.6354,
+      "step": 8031
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.845228194962765,
+      "learning_rate": 1.308471630958495e-05,
+      "loss": 0.6104,
+      "step": 8032
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.53215942226686,
+      "learning_rate": 1.308311210645571e-05,
+      "loss": 0.7366,
+      "step": 8033
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.13644579044986,
+      "learning_rate": 1.3081507815644567e-05,
+      "loss": 0.7059,
+      "step": 8034
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.96785430537651,
+      "learning_rate": 1.3079903437197143e-05,
+      "loss": 0.6275,
+      "step": 8035
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.643482824029313,
+      "learning_rate": 1.3078298971159067e-05,
+      "loss": 0.6371,
+      "step": 8036
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.96079941038394,
+      "learning_rate": 1.307669441757597e-05,
+      "loss": 0.7108,
+      "step": 8037
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.606243149464927,
+      "learning_rate": 1.3075089776493481e-05,
+      "loss": 0.6766,
+      "step": 8038
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 22.976607181592534,
+      "learning_rate": 1.3073485047957242e-05,
+      "loss": 0.753,
+      "step": 8039
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.59863223795007,
+      "learning_rate": 1.3071880232012881e-05,
+      "loss": 0.7284,
+      "step": 8040
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 26.25509165928565,
+      "learning_rate": 1.3070275328706047e-05,
+      "loss": 0.6254,
+      "step": 8041
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.182875356608985,
+      "learning_rate": 1.306867033808238e-05,
+      "loss": 0.591,
+      "step": 8042
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.29877226011105,
+      "learning_rate": 1.306706526018752e-05,
+      "loss": 0.6691,
+      "step": 8043
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.16888109006233,
+      "learning_rate": 1.306546009506712e-05,
+      "loss": 0.6923,
+      "step": 8044
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.08747598980473,
+      "learning_rate": 1.3063854842766828e-05,
+      "loss": 0.6637,
+      "step": 8045
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 28.459688069690944,
+      "learning_rate": 1.3062249503332294e-05,
+      "loss": 0.7626,
+      "step": 8046
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.987700725973603,
+      "learning_rate": 1.3060644076809178e-05,
+      "loss": 0.689,
+      "step": 8047
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.207565630561497,
+      "learning_rate": 1.3059038563243134e-05,
+      "loss": 0.6706,
+      "step": 8048
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.32752337163448,
+      "learning_rate": 1.3057432962679823e-05,
+      "loss": 0.634,
+      "step": 8049
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.357077477434743,
+      "learning_rate": 1.3055827275164908e-05,
+      "loss": 0.6368,
+      "step": 8050
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.405643486569828,
+      "learning_rate": 1.305422150074405e-05,
+      "loss": 0.5992,
+      "step": 8051
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.089647997557105,
+      "learning_rate": 1.3052615639462923e-05,
+      "loss": 0.6943,
+      "step": 8052
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.605909577930653,
+      "learning_rate": 1.3051009691367191e-05,
+      "loss": 0.6964,
+      "step": 8053
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 24.936611174038248,
+      "learning_rate": 1.304940365650253e-05,
+      "loss": 0.6061,
+      "step": 8054
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.87232370907216,
+      "learning_rate": 1.3047797534914609e-05,
+      "loss": 0.6753,
+      "step": 8055
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.715961010743476,
+      "learning_rate": 1.3046191326649114e-05,
+      "loss": 0.7218,
+      "step": 8056
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.881813148601875,
+      "learning_rate": 1.3044585031751716e-05,
+      "loss": 0.6388,
+      "step": 8057
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.289576357327363,
+      "learning_rate": 1.3042978650268102e-05,
+      "loss": 0.6415,
+      "step": 8058
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.86277932209255,
+      "learning_rate": 1.3041372182243955e-05,
+      "loss": 0.6875,
+      "step": 8059
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.82818851325902,
+      "learning_rate": 1.3039765627724961e-05,
+      "loss": 0.6484,
+      "step": 8060
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.779006779809293,
+      "learning_rate": 1.3038158986756812e-05,
+      "loss": 0.6049,
+      "step": 8061
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.29816063334219,
+      "learning_rate": 1.30365522593852e-05,
+      "loss": 0.687,
+      "step": 8062
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.63552919362852,
+      "learning_rate": 1.3034945445655813e-05,
+      "loss": 0.5901,
+      "step": 8063
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.039137659256628,
+      "learning_rate": 1.303333854561436e-05,
+      "loss": 0.6291,
+      "step": 8064
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.518064989659592,
+      "learning_rate": 1.3031731559306528e-05,
+      "loss": 0.6836,
+      "step": 8065
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.162309839383838,
+      "learning_rate": 1.3030124486778029e-05,
+      "loss": 0.6957,
+      "step": 8066
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.673784059481592,
+      "learning_rate": 1.3028517328074557e-05,
+      "loss": 0.5744,
+      "step": 8067
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.798210789957974,
+      "learning_rate": 1.3026910083241825e-05,
+      "loss": 0.6016,
+      "step": 8068
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.60857404840023,
+      "learning_rate": 1.3025302752325544e-05,
+      "loss": 0.68,
+      "step": 8069
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.599724971756796,
+      "learning_rate": 1.302369533537142e-05,
+      "loss": 0.6848,
+      "step": 8070
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 29.297590207334487,
+      "learning_rate": 1.302208783242517e-05,
+      "loss": 0.7169,
+      "step": 8071
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.50109058549094,
+      "learning_rate": 1.302048024353251e-05,
+      "loss": 0.6008,
+      "step": 8072
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.230349797811378,
+      "learning_rate": 1.3018872568739158e-05,
+      "loss": 0.5577,
+      "step": 8073
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.740862755684619,
+      "learning_rate": 1.3017264808090834e-05,
+      "loss": 0.6639,
+      "step": 8074
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.590514992000543,
+      "learning_rate": 1.3015656961633267e-05,
+      "loss": 0.6889,
+      "step": 8075
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.703502957913768,
+      "learning_rate": 1.301404902941218e-05,
+      "loss": 0.6794,
+      "step": 8076
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.48843429438833,
+      "learning_rate": 1.3012441011473302e-05,
+      "loss": 0.6521,
+      "step": 8077
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 25.176779369281245,
+      "learning_rate": 1.3010832907862363e-05,
+      "loss": 0.6505,
+      "step": 8078
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.06897907515434,
+      "learning_rate": 1.30092247186251e-05,
+      "loss": 0.607,
+      "step": 8079
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 29.02435386207123,
+      "learning_rate": 1.300761644380724e-05,
+      "loss": 0.6733,
+      "step": 8080
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.404721088481462,
+      "learning_rate": 1.3006008083454535e-05,
+      "loss": 0.6356,
+      "step": 8081
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.623189423948606,
+      "learning_rate": 1.3004399637612712e-05,
+      "loss": 0.5975,
+      "step": 8082
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.212860929356678,
+      "learning_rate": 1.3002791106327528e-05,
+      "loss": 0.6029,
+      "step": 8083
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.77139714966941,
+      "learning_rate": 1.3001182489644715e-05,
+      "loss": 0.5814,
+      "step": 8084
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.927539945752825,
+      "learning_rate": 1.2999573787610031e-05,
+      "loss": 0.6126,
+      "step": 8085
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.795299804758624,
+      "learning_rate": 1.2997965000269223e-05,
+      "loss": 0.5802,
+      "step": 8086
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.780677363893393,
+      "learning_rate": 1.2996356127668044e-05,
+      "loss": 0.6255,
+      "step": 8087
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 22.077401849251853,
+      "learning_rate": 1.2994747169852249e-05,
+      "loss": 0.6261,
+      "step": 8088
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.559667540661923,
+      "learning_rate": 1.2993138126867597e-05,
+      "loss": 0.6299,
+      "step": 8089
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 28.457817103440483,
+      "learning_rate": 1.2991528998759844e-05,
+      "loss": 0.6094,
+      "step": 8090
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.339257345874962,
+      "learning_rate": 1.2989919785574764e-05,
+      "loss": 0.7234,
+      "step": 8091
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.6435415185879,
+      "learning_rate": 1.2988310487358107e-05,
+      "loss": 0.6359,
+      "step": 8092
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.987047190713486,
+      "learning_rate": 1.298670110415565e-05,
+      "loss": 0.6762,
+      "step": 8093
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.850866606065082,
+      "learning_rate": 1.2985091636013163e-05,
+      "loss": 0.6225,
+      "step": 8094
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.571227599151083,
+      "learning_rate": 1.2983482082976412e-05,
+      "loss": 0.6764,
+      "step": 8095
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.67571502732276,
+      "learning_rate": 1.2981872445091179e-05,
+      "loss": 0.7227,
+      "step": 8096
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 21.467369648786548,
+      "learning_rate": 1.2980262722403238e-05,
+      "loss": 0.6298,
+      "step": 8097
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.703392703334963,
+      "learning_rate": 1.297865291495837e-05,
+      "loss": 0.6477,
+      "step": 8098
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.284576594262212,
+      "learning_rate": 1.2977043022802351e-05,
+      "loss": 0.6747,
+      "step": 8099
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.193384234372612,
+      "learning_rate": 1.2975433045980975e-05,
+      "loss": 0.6105,
+      "step": 8100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.67586365123682,
+      "learning_rate": 1.2973822984540018e-05,
+      "loss": 0.6995,
+      "step": 8101
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.10651846798882,
+      "learning_rate": 1.2972212838525278e-05,
+      "loss": 0.6199,
+      "step": 8102
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.45786557803001,
+      "learning_rate": 1.2970602607982546e-05,
+      "loss": 0.7185,
+      "step": 8103
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.34438554224968,
+      "learning_rate": 1.2968992292957609e-05,
+      "loss": 0.5187,
+      "step": 8104
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.936576690697496,
+      "learning_rate": 1.296738189349627e-05,
+      "loss": 0.6408,
+      "step": 8105
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.64420570169021,
+      "learning_rate": 1.2965771409644328e-05,
+      "loss": 0.6539,
+      "step": 8106
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 23.30442986456231,
+      "learning_rate": 1.2964160841447577e-05,
+      "loss": 0.618,
+      "step": 8107
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 26.527710628823286,
+      "learning_rate": 1.2962550188951832e-05,
+      "loss": 0.6227,
+      "step": 8108
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 20.65186116964494,
+      "learning_rate": 1.2960939452202888e-05,
+      "loss": 0.6377,
+      "step": 8109
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.48742288029853,
+      "learning_rate": 1.2959328631246562e-05,
+      "loss": 0.6404,
+      "step": 8110
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.599566165277086,
+      "learning_rate": 1.2957717726128656e-05,
+      "loss": 0.6358,
+      "step": 8111
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.049314629175925,
+      "learning_rate": 1.295610673689499e-05,
+      "loss": 0.721,
+      "step": 8112
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.902807623466806,
+      "learning_rate": 1.295449566359138e-05,
+      "loss": 0.6448,
+      "step": 8113
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.78937737953778,
+      "learning_rate": 1.2952884506263641e-05,
+      "loss": 0.5574,
+      "step": 8114
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.79619761208542,
+      "learning_rate": 1.2951273264957591e-05,
+      "loss": 0.6552,
+      "step": 8115
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.187879162107276,
+      "learning_rate": 1.2949661939719059e-05,
+      "loss": 0.7062,
+      "step": 8116
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.66637270677793,
+      "learning_rate": 1.2948050530593863e-05,
+      "loss": 0.556,
+      "step": 8117
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.678856294276525,
+      "learning_rate": 1.2946439037627838e-05,
+      "loss": 0.6095,
+      "step": 8118
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.89260742317735,
+      "learning_rate": 1.2944827460866809e-05,
+      "loss": 0.6259,
+      "step": 8119
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.78214248384012,
+      "learning_rate": 1.294321580035661e-05,
+      "loss": 0.6526,
+      "step": 8120
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.834424581502493,
+      "learning_rate": 1.2941604056143078e-05,
+      "loss": 0.6312,
+      "step": 8121
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 24.361240146110788,
+      "learning_rate": 1.2939992228272043e-05,
+      "loss": 0.69,
+      "step": 8122
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.877071231404777,
+      "learning_rate": 1.2938380316789352e-05,
+      "loss": 0.5819,
+      "step": 8123
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.473555748512167,
+      "learning_rate": 1.293676832174084e-05,
+      "loss": 0.6196,
+      "step": 8124
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.73949488865913,
+      "learning_rate": 1.293515624317236e-05,
+      "loss": 0.6337,
+      "step": 8125
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.448364607954518,
+      "learning_rate": 1.293354408112975e-05,
+      "loss": 0.5691,
+      "step": 8126
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.694756018415156,
+      "learning_rate": 1.2931931835658864e-05,
+      "loss": 0.6521,
+      "step": 8127
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 38.066722819233796,
+      "learning_rate": 1.293031950680555e-05,
+      "loss": 0.737,
+      "step": 8128
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.798866402502114,
+      "learning_rate": 1.2928707094615665e-05,
+      "loss": 0.6981,
+      "step": 8129
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.41769476532043,
+      "learning_rate": 1.2927094599135064e-05,
+      "loss": 0.5406,
+      "step": 8130
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.64114958708201,
+      "learning_rate": 1.2925482020409604e-05,
+      "loss": 0.6444,
+      "step": 8131
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.039166211848157,
+      "learning_rate": 1.2923869358485147e-05,
+      "loss": 0.6988,
+      "step": 8132
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.34886971034679,
+      "learning_rate": 1.2922256613407553e-05,
+      "loss": 0.5802,
+      "step": 8133
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.549011014723588,
+      "learning_rate": 1.292064378522269e-05,
+      "loss": 0.6199,
+      "step": 8134
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 20.731477949708044,
+      "learning_rate": 1.2919030873976433e-05,
+      "loss": 0.6691,
+      "step": 8135
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.514022325330988,
+      "learning_rate": 1.291741787971464e-05,
+      "loss": 0.6005,
+      "step": 8136
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 27.042457301323097,
+      "learning_rate": 1.2915804802483192e-05,
+      "loss": 0.8196,
+      "step": 8137
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 30.26239914656705,
+      "learning_rate": 1.2914191642327958e-05,
+      "loss": 0.6548,
+      "step": 8138
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.61272050246207,
+      "learning_rate": 1.2912578399294822e-05,
+      "loss": 0.5848,
+      "step": 8139
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 23.194163024471106,
+      "learning_rate": 1.2910965073429657e-05,
+      "loss": 0.6923,
+      "step": 8140
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 28.114345680387576,
+      "learning_rate": 1.2909351664778352e-05,
+      "loss": 0.689,
+      "step": 8141
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.409550353111786,
+      "learning_rate": 1.2907738173386783e-05,
+      "loss": 0.6354,
+      "step": 8142
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.480181081951525,
+      "learning_rate": 1.2906124599300842e-05,
+      "loss": 0.6481,
+      "step": 8143
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.429377995836642,
+      "learning_rate": 1.2904510942566421e-05,
+      "loss": 0.6804,
+      "step": 8144
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.722834172160264,
+      "learning_rate": 1.2902897203229406e-05,
+      "loss": 0.5646,
+      "step": 8145
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.759729081395175,
+      "learning_rate": 1.2901283381335691e-05,
+      "loss": 0.6659,
+      "step": 8146
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.444978165930042,
+      "learning_rate": 1.2899669476931176e-05,
+      "loss": 0.5811,
+      "step": 8147
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.518306735423433,
+      "learning_rate": 1.2898055490061757e-05,
+      "loss": 0.6521,
+      "step": 8148
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.733359029327055,
+      "learning_rate": 1.2896441420773335e-05,
+      "loss": 0.6341,
+      "step": 8149
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.477306501553883,
+      "learning_rate": 1.2894827269111811e-05,
+      "loss": 0.6213,
+      "step": 8150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.630328221854942,
+      "learning_rate": 1.2893213035123096e-05,
+      "loss": 0.6439,
+      "step": 8151
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.360006763758854,
+      "learning_rate": 1.2891598718853096e-05,
+      "loss": 0.7555,
+      "step": 8152
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.679010002465652,
+      "learning_rate": 1.2889984320347715e-05,
+      "loss": 0.6289,
+      "step": 8153
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.306412535242888,
+      "learning_rate": 1.2888369839652875e-05,
+      "loss": 0.6526,
+      "step": 8154
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.080063890437046,
+      "learning_rate": 1.2886755276814484e-05,
+      "loss": 0.6543,
+      "step": 8155
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.375367553028894,
+      "learning_rate": 1.2885140631878463e-05,
+      "loss": 0.6382,
+      "step": 8156
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.28122932117928,
+      "learning_rate": 1.2883525904890728e-05,
+      "loss": 0.5818,
+      "step": 8157
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.21215980378753,
+      "learning_rate": 1.2881911095897206e-05,
+      "loss": 0.5374,
+      "step": 8158
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.70605472892297,
+      "learning_rate": 1.2880296204943816e-05,
+      "loss": 0.65,
+      "step": 8159
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.843607854466118,
+      "learning_rate": 1.287868123207649e-05,
+      "loss": 0.6771,
+      "step": 8160
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.619776545342837,
+      "learning_rate": 1.2877066177341148e-05,
+      "loss": 0.6139,
+      "step": 8161
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.732893863656638,
+      "learning_rate": 1.287545104078373e-05,
+      "loss": 0.5971,
+      "step": 8162
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 22.347713784442327,
+      "learning_rate": 1.2873835822450168e-05,
+      "loss": 0.7457,
+      "step": 8163
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.417072269309283,
+      "learning_rate": 1.2872220522386396e-05,
+      "loss": 0.6432,
+      "step": 8164
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.638863247439584,
+      "learning_rate": 1.2870605140638352e-05,
+      "loss": 0.596,
+      "step": 8165
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.743124205002776,
+      "learning_rate": 1.2868989677251976e-05,
+      "loss": 0.6887,
+      "step": 8166
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.636283051586897,
+      "learning_rate": 1.2867374132273214e-05,
+      "loss": 0.6665,
+      "step": 8167
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.321634925503275,
+      "learning_rate": 1.2865758505748006e-05,
+      "loss": 0.5936,
+      "step": 8168
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.366738158522285,
+      "learning_rate": 1.2864142797722305e-05,
+      "loss": 0.6217,
+      "step": 8169
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.724378762419924,
+      "learning_rate": 1.2862527008242056e-05,
+      "loss": 0.6125,
+      "step": 8170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.261670719380938,
+      "learning_rate": 1.2860911137353219e-05,
+      "loss": 0.6728,
+      "step": 8171
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.35691244855049,
+      "learning_rate": 1.2859295185101736e-05,
+      "loss": 0.6272,
+      "step": 8172
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.572375423177817,
+      "learning_rate": 1.2857679151533576e-05,
+      "loss": 0.7258,
+      "step": 8173
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.28698023090884,
+      "learning_rate": 1.2856063036694688e-05,
+      "loss": 0.6612,
+      "step": 8174
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.575975106319408,
+      "learning_rate": 1.2854446840631041e-05,
+      "loss": 0.6047,
+      "step": 8175
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.183492852947104,
+      "learning_rate": 1.2852830563388597e-05,
+      "loss": 0.5923,
+      "step": 8176
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.053828227886665,
+      "learning_rate": 1.2851214205013317e-05,
+      "loss": 0.6942,
+      "step": 8177
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.7969470888414,
+      "learning_rate": 1.2849597765551171e-05,
+      "loss": 0.6402,
+      "step": 8178
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.286436343268512,
+      "learning_rate": 1.2847981245048137e-05,
+      "loss": 0.6593,
+      "step": 8179
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.417870226809434,
+      "learning_rate": 1.284636464355018e-05,
+      "loss": 0.6639,
+      "step": 8180
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.517238998406224,
+      "learning_rate": 1.2844747961103279e-05,
+      "loss": 0.6272,
+      "step": 8181
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 27.43945814826806,
+      "learning_rate": 1.2843131197753408e-05,
+      "loss": 0.6415,
+      "step": 8182
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 26.910037510156272,
+      "learning_rate": 1.2841514353546549e-05,
+      "loss": 0.6448,
+      "step": 8183
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.863559745081396,
+      "learning_rate": 1.2839897428528687e-05,
+      "loss": 0.594,
+      "step": 8184
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.722278407000786,
+      "learning_rate": 1.2838280422745801e-05,
+      "loss": 0.5537,
+      "step": 8185
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.28998282903744,
+      "learning_rate": 1.283666333624388e-05,
+      "loss": 0.7031,
+      "step": 8186
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 24.678736125870845,
+      "learning_rate": 1.2835046169068912e-05,
+      "loss": 0.6506,
+      "step": 8187
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.86576640145763,
+      "learning_rate": 1.283342892126689e-05,
+      "loss": 0.6597,
+      "step": 8188
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.68417514707934,
+      "learning_rate": 1.283181159288381e-05,
+      "loss": 0.6305,
+      "step": 8189
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.088556117146169,
+      "learning_rate": 1.2830194183965663e-05,
+      "loss": 0.6077,
+      "step": 8190
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.664025697295775,
+      "learning_rate": 1.2828576694558447e-05,
+      "loss": 0.5667,
+      "step": 8191
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 24.491983022855344,
+      "learning_rate": 1.2826959124708167e-05,
+      "loss": 0.767,
+      "step": 8192
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.348674336485477,
+      "learning_rate": 1.2825341474460824e-05,
+      "loss": 0.6835,
+      "step": 8193
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 27.10513878849774,
+      "learning_rate": 1.2823723743862422e-05,
+      "loss": 0.7139,
+      "step": 8194
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.087777241478815,
+      "learning_rate": 1.2822105932958964e-05,
+      "loss": 0.5762,
+      "step": 8195
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 27.34988598560056,
+      "learning_rate": 1.282048804179647e-05,
+      "loss": 0.5978,
+      "step": 8196
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.475743284566807,
+      "learning_rate": 1.2818870070420942e-05,
+      "loss": 0.6952,
+      "step": 8197
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 24.90319452575244,
+      "learning_rate": 1.2817252018878401e-05,
+      "loss": 0.6428,
+      "step": 8198
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.821598848943367,
+      "learning_rate": 1.2815633887214857e-05,
+      "loss": 0.6362,
+      "step": 8199
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.267088470829696,
+      "learning_rate": 1.2814015675476336e-05,
+      "loss": 0.6174,
+      "step": 8200
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 32.68403077269346,
+      "learning_rate": 1.2812397383708854e-05,
+      "loss": 0.6853,
+      "step": 8201
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.39607547817,
+      "learning_rate": 1.2810779011958437e-05,
+      "loss": 0.6587,
+      "step": 8202
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.088205178262257,
+      "learning_rate": 1.2809160560271108e-05,
+      "loss": 0.6183,
+      "step": 8203
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.81991632310064,
+      "learning_rate": 1.2807542028692894e-05,
+      "loss": 0.7255,
+      "step": 8204
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.652369874199216,
+      "learning_rate": 1.2805923417269828e-05,
+      "loss": 0.6284,
+      "step": 8205
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.30337913744467,
+      "learning_rate": 1.2804304726047943e-05,
+      "loss": 0.61,
+      "step": 8206
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.251616359030656,
+      "learning_rate": 1.2802685955073271e-05,
+      "loss": 0.6537,
+      "step": 8207
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.35665278828155,
+      "learning_rate": 1.280106710439185e-05,
+      "loss": 0.645,
+      "step": 8208
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.34360182130856,
+      "learning_rate": 1.279944817404972e-05,
+      "loss": 0.5877,
+      "step": 8209
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.940694352410677,
+      "learning_rate": 1.279782916409292e-05,
+      "loss": 0.6063,
+      "step": 8210
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.1460525655355,
+      "learning_rate": 1.2796210074567499e-05,
+      "loss": 0.6644,
+      "step": 8211
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.391992489499236,
+      "learning_rate": 1.2794590905519494e-05,
+      "loss": 0.5939,
+      "step": 8212
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 27.558598731004682,
+      "learning_rate": 1.2792971656994963e-05,
+      "loss": 0.6897,
+      "step": 8213
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 21.736035336015192,
+      "learning_rate": 1.2791352329039948e-05,
+      "loss": 0.5729,
+      "step": 8214
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.35966969126824,
+      "learning_rate": 1.278973292170051e-05,
+      "loss": 0.6441,
+      "step": 8215
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 26.81296011334332,
+      "learning_rate": 1.2788113435022696e-05,
+      "loss": 0.7154,
+      "step": 8216
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.376780871287782,
+      "learning_rate": 1.2786493869052568e-05,
+      "loss": 0.5998,
+      "step": 8217
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 29.08429770774486,
+      "learning_rate": 1.2784874223836184e-05,
+      "loss": 0.6733,
+      "step": 8218
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.842001650702121,
+      "learning_rate": 1.2783254499419607e-05,
+      "loss": 0.5726,
+      "step": 8219
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.339599638625465,
+      "learning_rate": 1.2781634695848899e-05,
+      "loss": 0.6483,
+      "step": 8220
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 24.412345451094303,
+      "learning_rate": 1.278001481317013e-05,
+      "loss": 0.649,
+      "step": 8221
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.097347138905876,
+      "learning_rate": 1.2778394851429361e-05,
+      "loss": 0.5526,
+      "step": 8222
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.73414152023992,
+      "learning_rate": 1.2776774810672674e-05,
+      "loss": 0.632,
+      "step": 8223
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.325303280716902,
+      "learning_rate": 1.277515469094613e-05,
+      "loss": 0.5003,
+      "step": 8224
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.971576566258243,
+      "learning_rate": 1.2773534492295813e-05,
+      "loss": 0.5308,
+      "step": 8225
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.055650524859043,
+      "learning_rate": 1.2771914214767798e-05,
+      "loss": 0.6434,
+      "step": 8226
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.336058272079114,
+      "learning_rate": 1.2770293858408166e-05,
+      "loss": 0.6633,
+      "step": 8227
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.182235426854886,
+      "learning_rate": 1.2768673423262992e-05,
+      "loss": 0.7339,
+      "step": 8228
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 22.68572989959388,
+      "learning_rate": 1.2767052909378367e-05,
+      "loss": 0.7338,
+      "step": 8229
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.320254660896573,
+      "learning_rate": 1.2765432316800379e-05,
+      "loss": 0.626,
+      "step": 8230
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.825806270231375,
+      "learning_rate": 1.276381164557511e-05,
+      "loss": 0.6617,
+      "step": 8231
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 23.305797496759737,
+      "learning_rate": 1.2762190895748657e-05,
+      "loss": 0.6881,
+      "step": 8232
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.713697541779094,
+      "learning_rate": 1.2760570067367111e-05,
+      "loss": 0.6689,
+      "step": 8233
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.242289927193006,
+      "learning_rate": 1.2758949160476567e-05,
+      "loss": 0.6251,
+      "step": 8234
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.520502193029696,
+      "learning_rate": 1.2757328175123122e-05,
+      "loss": 0.5961,
+      "step": 8235
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 23.25639740864642,
+      "learning_rate": 1.2755707111352877e-05,
+      "loss": 0.7622,
+      "step": 8236
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.405445122730182,
+      "learning_rate": 1.2754085969211932e-05,
+      "loss": 0.6287,
+      "step": 8237
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.009959486758213,
+      "learning_rate": 1.2752464748746395e-05,
+      "loss": 0.6654,
+      "step": 8238
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 29.488844121426467,
+      "learning_rate": 1.275084345000237e-05,
+      "loss": 0.6218,
+      "step": 8239
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.191574689186826,
+      "learning_rate": 1.2749222073025968e-05,
+      "loss": 0.5641,
+      "step": 8240
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 22.50691927839262,
+      "learning_rate": 1.2747600617863296e-05,
+      "loss": 0.6369,
+      "step": 8241
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.675275036425337,
+      "learning_rate": 1.2745979084560474e-05,
+      "loss": 0.6113,
+      "step": 8242
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.42540606595315,
+      "learning_rate": 1.274435747316361e-05,
+      "loss": 0.7135,
+      "step": 8243
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.502848249262144,
+      "learning_rate": 1.2742735783718825e-05,
+      "loss": 0.6044,
+      "step": 8244
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.234961526703646,
+      "learning_rate": 1.2741114016272239e-05,
+      "loss": 0.581,
+      "step": 8245
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.207085752975964,
+      "learning_rate": 1.2739492170869973e-05,
+      "loss": 0.6448,
+      "step": 8246
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.858919183687561,
+      "learning_rate": 1.2737870247558152e-05,
+      "loss": 0.5923,
+      "step": 8247
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.561342703826355,
+      "learning_rate": 1.2736248246382906e-05,
+      "loss": 0.6666,
+      "step": 8248
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.343705813715111,
+      "learning_rate": 1.2734626167390356e-05,
+      "loss": 0.5232,
+      "step": 8249
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.694151070144503,
+      "learning_rate": 1.2733004010626639e-05,
+      "loss": 0.7189,
+      "step": 8250
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.76889183044537,
+      "learning_rate": 1.273138177613789e-05,
+      "loss": 0.673,
+      "step": 8251
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.191555895339366,
+      "learning_rate": 1.2729759463970238e-05,
+      "loss": 0.6596,
+      "step": 8252
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.413788901767116,
+      "learning_rate": 1.2728137074169824e-05,
+      "loss": 0.622,
+      "step": 8253
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.702806584416553,
+      "learning_rate": 1.2726514606782785e-05,
+      "loss": 0.7032,
+      "step": 8254
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.454562870184347,
+      "learning_rate": 1.2724892061855271e-05,
+      "loss": 0.6099,
+      "step": 8255
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.534468667838016,
+      "learning_rate": 1.2723269439433416e-05,
+      "loss": 0.5489,
+      "step": 8256
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.893511429360114,
+      "learning_rate": 1.2721646739563378e-05,
+      "loss": 0.6107,
+      "step": 8257
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 25.034105700567117,
+      "learning_rate": 1.2720023962291292e-05,
+      "loss": 0.697,
+      "step": 8258
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.347027819371986,
+      "learning_rate": 1.2718401107663318e-05,
+      "loss": 0.6945,
+      "step": 8259
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.475692892839113,
+      "learning_rate": 1.2716778175725606e-05,
+      "loss": 0.6346,
+      "step": 8260
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.219003174438452,
+      "learning_rate": 1.2715155166524313e-05,
+      "loss": 0.6401,
+      "step": 8261
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.103182003521006,
+      "learning_rate": 1.2713532080105595e-05,
+      "loss": 0.6422,
+      "step": 8262
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 22.474250747499287,
+      "learning_rate": 1.2711908916515613e-05,
+      "loss": 0.617,
+      "step": 8263
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.380625495399574,
+      "learning_rate": 1.2710285675800527e-05,
+      "loss": 0.5819,
+      "step": 8264
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.53238287748573,
+      "learning_rate": 1.2708662358006504e-05,
+      "loss": 0.5705,
+      "step": 8265
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 23.17753445404153,
+      "learning_rate": 1.2707038963179704e-05,
+      "loss": 0.717,
+      "step": 8266
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.869701037519455,
+      "learning_rate": 1.2705415491366303e-05,
+      "loss": 0.6855,
+      "step": 8267
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.1292536875374,
+      "learning_rate": 1.2703791942612462e-05,
+      "loss": 0.5989,
+      "step": 8268
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.419536762127363,
+      "learning_rate": 1.2702168316964365e-05,
+      "loss": 0.5862,
+      "step": 8269
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.78543399158514,
+      "learning_rate": 1.2700544614468181e-05,
+      "loss": 0.6635,
+      "step": 8270
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.072822686719643,
+      "learning_rate": 1.2698920835170088e-05,
+      "loss": 0.6921,
+      "step": 8271
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.085295042332895,
+      "learning_rate": 1.2697296979116265e-05,
+      "loss": 0.6539,
+      "step": 8272
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.33598617663096,
+      "learning_rate": 1.2695673046352893e-05,
+      "loss": 0.619,
+      "step": 8273
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.727146128274091,
+      "learning_rate": 1.2694049036926159e-05,
+      "loss": 0.5424,
+      "step": 8274
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.767142420036183,
+      "learning_rate": 1.2692424950882242e-05,
+      "loss": 0.6286,
+      "step": 8275
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.435146818243783,
+      "learning_rate": 1.2690800788267338e-05,
+      "loss": 0.5762,
+      "step": 8276
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.64200408620659,
+      "learning_rate": 1.2689176549127633e-05,
+      "loss": 0.7087,
+      "step": 8277
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.426749474636896,
+      "learning_rate": 1.2687552233509322e-05,
+      "loss": 0.6936,
+      "step": 8278
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.95753997236476,
+      "learning_rate": 1.2685927841458593e-05,
+      "loss": 0.6355,
+      "step": 8279
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.704809227276208,
+      "learning_rate": 1.2684303373021651e-05,
+      "loss": 0.6469,
+      "step": 8280
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.786845307106773,
+      "learning_rate": 1.268267882824469e-05,
+      "loss": 0.6742,
+      "step": 8281
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.523714175237412,
+      "learning_rate": 1.2681054207173915e-05,
+      "loss": 0.599,
+      "step": 8282
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.42877414153336,
+      "learning_rate": 1.2679429509855521e-05,
+      "loss": 0.659,
+      "step": 8283
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.899460323960357,
+      "learning_rate": 1.2677804736335726e-05,
+      "loss": 0.5583,
+      "step": 8284
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.53406389686106,
+      "learning_rate": 1.2676179886660725e-05,
+      "loss": 0.6189,
+      "step": 8285
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.087217950500186,
+      "learning_rate": 1.2674554960876737e-05,
+      "loss": 0.6769,
+      "step": 8286
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 24.029417775453116,
+      "learning_rate": 1.2672929959029969e-05,
+      "loss": 0.6117,
+      "step": 8287
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 24.828266838393592,
+      "learning_rate": 1.2671304881166638e-05,
+      "loss": 0.6575,
+      "step": 8288
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.012288101359115,
+      "learning_rate": 1.2669679727332957e-05,
+      "loss": 0.5936,
+      "step": 8289
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.185367081076276,
+      "learning_rate": 1.2668054497575147e-05,
+      "loss": 0.6373,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.471258515197395,
+      "learning_rate": 1.2666429191939428e-05,
+      "loss": 0.6036,
+      "step": 8291
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 32.78390001926139,
+      "learning_rate": 1.2664803810472025e-05,
+      "loss": 0.672,
+      "step": 8292
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 20.973422872550596,
+      "learning_rate": 1.2663178353219154e-05,
+      "loss": 0.6369,
+      "step": 8293
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.30912832403255,
+      "learning_rate": 1.2661552820227053e-05,
+      "loss": 0.6168,
+      "step": 8294
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 20.21009902999088,
+      "learning_rate": 1.2659927211541946e-05,
+      "loss": 0.6063,
+      "step": 8295
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.766984977642572,
+      "learning_rate": 1.2658301527210066e-05,
+      "loss": 0.6186,
+      "step": 8296
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.254903317822391,
+      "learning_rate": 1.2656675767277647e-05,
+      "loss": 0.5818,
+      "step": 8297
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 40.45768699840152,
+      "learning_rate": 1.2655049931790919e-05,
+      "loss": 0.6823,
+      "step": 8298
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.93386568161306,
+      "learning_rate": 1.2653424020796128e-05,
+      "loss": 0.5915,
+      "step": 8299
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 27.277270190650405,
+      "learning_rate": 1.2651798034339506e-05,
+      "loss": 0.6445,
+      "step": 8300
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.89493076950185,
+      "learning_rate": 1.2650171972467304e-05,
+      "loss": 0.6528,
+      "step": 8301
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 24.165653551312477,
+      "learning_rate": 1.2648545835225758e-05,
+      "loss": 0.6742,
+      "step": 8302
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.142076057643568,
+      "learning_rate": 1.2646919622661119e-05,
+      "loss": 0.6438,
+      "step": 8303
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.194847990832569,
+      "learning_rate": 1.2645293334819635e-05,
+      "loss": 0.5588,
+      "step": 8304
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.47180693523747,
+      "learning_rate": 1.2643666971747555e-05,
+      "loss": 0.6949,
+      "step": 8305
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.225884641457455,
+      "learning_rate": 1.2642040533491133e-05,
+      "loss": 0.6261,
+      "step": 8306
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 20.36221919133966,
+      "learning_rate": 1.2640414020096626e-05,
+      "loss": 0.5965,
+      "step": 8307
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.204110906216208,
+      "learning_rate": 1.2638787431610285e-05,
+      "loss": 0.6384,
+      "step": 8308
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.597296453442517,
+      "learning_rate": 1.2637160768078376e-05,
+      "loss": 0.6544,
+      "step": 8309
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 23.160980914446505,
+      "learning_rate": 1.2635534029547155e-05,
+      "loss": 0.6486,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.00884000496649,
+      "learning_rate": 1.263390721606289e-05,
+      "loss": 0.6161,
+      "step": 8311
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 23.374112095773256,
+      "learning_rate": 1.2632280327671842e-05,
+      "loss": 0.6349,
+      "step": 8312
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.90759129436889,
+      "learning_rate": 1.2630653364420285e-05,
+      "loss": 0.7162,
+      "step": 8313
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 28.444149062689792,
+      "learning_rate": 1.2629026326354481e-05,
+      "loss": 0.6268,
+      "step": 8314
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.3915209329876,
+      "learning_rate": 1.262739921352071e-05,
+      "loss": 0.6736,
+      "step": 8315
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.4954904953057,
+      "learning_rate": 1.2625772025965241e-05,
+      "loss": 0.6483,
+      "step": 8316
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.94987849039004,
+      "learning_rate": 1.262414476373435e-05,
+      "loss": 0.644,
+      "step": 8317
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.480263768799166,
+      "learning_rate": 1.2622517426874322e-05,
+      "loss": 0.6107,
+      "step": 8318
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.973122043753486,
+      "learning_rate": 1.2620890015431427e-05,
+      "loss": 0.5114,
+      "step": 8319
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.78530759285365,
+      "learning_rate": 1.2619262529451954e-05,
+      "loss": 0.6009,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.290259431183483,
+      "learning_rate": 1.2617634968982189e-05,
+      "loss": 0.6141,
+      "step": 8321
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.916252844856091,
+      "learning_rate": 1.2616007334068417e-05,
+      "loss": 0.6312,
+      "step": 8322
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.10366944305978,
+      "learning_rate": 1.2614379624756924e-05,
+      "loss": 0.6477,
+      "step": 8323
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.974929096587086,
+      "learning_rate": 1.2612751841094006e-05,
+      "loss": 0.6621,
+      "step": 8324
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 44.209934620285786,
+      "learning_rate": 1.2611123983125951e-05,
+      "loss": 0.6666,
+      "step": 8325
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.474935030700554,
+      "learning_rate": 1.2609496050899061e-05,
+      "loss": 0.6136,
+      "step": 8326
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 20.948634831066823,
+      "learning_rate": 1.2607868044459629e-05,
+      "loss": 0.5512,
+      "step": 8327
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.758153153662324,
+      "learning_rate": 1.2606239963853957e-05,
+      "loss": 0.645,
+      "step": 8328
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.338863175005827,
+      "learning_rate": 1.260461180912834e-05,
+      "loss": 0.666,
+      "step": 8329
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.871535718989445,
+      "learning_rate": 1.260298358032909e-05,
+      "loss": 0.6818,
+      "step": 8330
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.4711589734319,
+      "learning_rate": 1.260135527750251e-05,
+      "loss": 0.6062,
+      "step": 8331
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.597299414938952,
+      "learning_rate": 1.2599726900694909e-05,
+      "loss": 0.6556,
+      "step": 8332
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.01644488202975,
+      "learning_rate": 1.2598098449952594e-05,
+      "loss": 0.6689,
+      "step": 8333
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.478950747518574,
+      "learning_rate": 1.2596469925321877e-05,
+      "loss": 0.6645,
+      "step": 8334
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.898143156471157,
+      "learning_rate": 1.2594841326849075e-05,
+      "loss": 0.671,
+      "step": 8335
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.316397881603008,
+      "learning_rate": 1.2593212654580507e-05,
+      "loss": 0.7539,
+      "step": 8336
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.919082616649789,
+      "learning_rate": 1.2591583908562483e-05,
+      "loss": 0.5255,
+      "step": 8337
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 43.385797618939,
+      "learning_rate": 1.2589955088841332e-05,
+      "loss": 0.5732,
+      "step": 8338
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 24.524400495887633,
+      "learning_rate": 1.2588326195463373e-05,
+      "loss": 0.6069,
+      "step": 8339
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.70334773584294,
+      "learning_rate": 1.2586697228474931e-05,
+      "loss": 0.6646,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.06490269003772,
+      "learning_rate": 1.2585068187922333e-05,
+      "loss": 0.6353,
+      "step": 8341
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.61264025552755,
+      "learning_rate": 1.2583439073851907e-05,
+      "loss": 0.6275,
+      "step": 8342
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.231405004863914,
+      "learning_rate": 1.2581809886309989e-05,
+      "loss": 0.5737,
+      "step": 8343
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.215214434693078,
+      "learning_rate": 1.25801806253429e-05,
+      "loss": 0.5884,
+      "step": 8344
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.457123827986855,
+      "learning_rate": 1.2578551290996993e-05,
+      "loss": 0.6552,
+      "step": 8345
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.390290753966006,
+      "learning_rate": 1.2576921883318589e-05,
+      "loss": 0.6073,
+      "step": 8346
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.580823521516413,
+      "learning_rate": 1.2575292402354036e-05,
+      "loss": 0.6572,
+      "step": 8347
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.342939219638488,
+      "learning_rate": 1.2573662848149674e-05,
+      "loss": 0.6532,
+      "step": 8348
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.704701681590626,
+      "learning_rate": 1.2572033220751844e-05,
+      "loss": 0.6629,
+      "step": 8349
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.744592104506047,
+      "learning_rate": 1.2570403520206898e-05,
+      "loss": 0.6081,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.026847720474173,
+      "learning_rate": 1.2568773746561179e-05,
+      "loss": 0.5171,
+      "step": 8351
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.078173749479243,
+      "learning_rate": 1.2567143899861031e-05,
+      "loss": 0.6911,
+      "step": 8352
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.63746392616011,
+      "learning_rate": 1.2565513980152822e-05,
+      "loss": 0.6007,
+      "step": 8353
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.722548822687365,
+      "learning_rate": 1.2563883987482887e-05,
+      "loss": 0.5726,
+      "step": 8354
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.49407482065009,
+      "learning_rate": 1.2562253921897598e-05,
+      "loss": 0.6477,
+      "step": 8355
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.053773822740483,
+      "learning_rate": 1.2560623783443302e-05,
+      "loss": 0.6122,
+      "step": 8356
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.992982872981976,
+      "learning_rate": 1.2558993572166365e-05,
+      "loss": 0.6202,
+      "step": 8357
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.554575835820508,
+      "learning_rate": 1.255736328811315e-05,
+      "loss": 0.6036,
+      "step": 8358
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 28.007069121355418,
+      "learning_rate": 1.2555732931330015e-05,
+      "loss": 0.6098,
+      "step": 8359
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.900518265884386,
+      "learning_rate": 1.2554102501863333e-05,
+      "loss": 0.5865,
+      "step": 8360
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.19558522659384,
+      "learning_rate": 1.2552471999759469e-05,
+      "loss": 0.5392,
+      "step": 8361
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.110822537651774,
+      "learning_rate": 1.2550841425064795e-05,
+      "loss": 0.5385,
+      "step": 8362
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.938318039318656,
+      "learning_rate": 1.254921077782568e-05,
+      "loss": 0.5992,
+      "step": 8363
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.815408864791664,
+      "learning_rate": 1.2547580058088507e-05,
+      "loss": 0.7195,
+      "step": 8364
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.682037749183138,
+      "learning_rate": 1.2545949265899645e-05,
+      "loss": 0.6883,
+      "step": 8365
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.833900433071957,
+      "learning_rate": 1.2544318401305476e-05,
+      "loss": 0.7107,
+      "step": 8366
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.506565240216386,
+      "learning_rate": 1.2542687464352378e-05,
+      "loss": 0.6388,
+      "step": 8367
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.664891985924683,
+      "learning_rate": 1.2541056455086737e-05,
+      "loss": 0.6404,
+      "step": 8368
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.537311302131393,
+      "learning_rate": 1.2539425373554937e-05,
+      "loss": 0.6769,
+      "step": 8369
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.35579949645393,
+      "learning_rate": 1.2537794219803368e-05,
+      "loss": 0.6158,
+      "step": 8370
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.55089074986457,
+      "learning_rate": 1.2536162993878409e-05,
+      "loss": 0.5631,
+      "step": 8371
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.08641090051207,
+      "learning_rate": 1.2534531695826466e-05,
+      "loss": 0.66,
+      "step": 8372
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.393151781789907,
+      "learning_rate": 1.2532900325693917e-05,
+      "loss": 0.6303,
+      "step": 8373
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 24.21301103091029,
+      "learning_rate": 1.2531268883527168e-05,
+      "loss": 0.7126,
+      "step": 8374
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.535537246309605,
+      "learning_rate": 1.2529637369372614e-05,
+      "loss": 0.6424,
+      "step": 8375
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.814339562844467,
+      "learning_rate": 1.2528005783276652e-05,
+      "loss": 0.5745,
+      "step": 8376
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.077784393687985,
+      "learning_rate": 1.2526374125285683e-05,
+      "loss": 0.5842,
+      "step": 8377
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.764549310036568,
+      "learning_rate": 1.2524742395446112e-05,
+      "loss": 0.6442,
+      "step": 8378
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 24.32706219404166,
+      "learning_rate": 1.2523110593804344e-05,
+      "loss": 0.6665,
+      "step": 8379
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.587623230692884,
+      "learning_rate": 1.252147872040679e-05,
+      "loss": 0.6504,
+      "step": 8380
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.64738030869867,
+      "learning_rate": 1.251984677529985e-05,
+      "loss": 0.6724,
+      "step": 8381
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.66866432322756,
+      "learning_rate": 1.2518214758529947e-05,
+      "loss": 0.6555,
+      "step": 8382
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.135922908489277,
+      "learning_rate": 1.251658267014349e-05,
+      "loss": 0.5536,
+      "step": 8383
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.683570287004873,
+      "learning_rate": 1.2514950510186892e-05,
+      "loss": 0.5341,
+      "step": 8384
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.497442436918337,
+      "learning_rate": 1.2513318278706573e-05,
+      "loss": 0.6684,
+      "step": 8385
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.794121792194833,
+      "learning_rate": 1.2511685975748948e-05,
+      "loss": 0.6374,
+      "step": 8386
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 25.733633631743633,
+      "learning_rate": 1.251005360136045e-05,
+      "loss": 0.7394,
+      "step": 8387
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 46.83819900012865,
+      "learning_rate": 1.2508421155587492e-05,
+      "loss": 0.635,
+      "step": 8388
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.8407095466611,
+      "learning_rate": 1.2506788638476506e-05,
+      "loss": 0.5833,
+      "step": 8389
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.778555733521987,
+      "learning_rate": 1.2505156050073914e-05,
+      "loss": 0.5958,
+      "step": 8390
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 27.671684704432664,
+      "learning_rate": 1.2503523390426153e-05,
+      "loss": 0.7228,
+      "step": 8391
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.50419597507629,
+      "learning_rate": 1.2501890659579649e-05,
+      "loss": 0.6231,
+      "step": 8392
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 28.585759525417487,
+      "learning_rate": 1.250025785758084e-05,
+      "loss": 0.6347,
+      "step": 8393
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.109868363012563,
+      "learning_rate": 1.249862498447616e-05,
+      "loss": 0.5274,
+      "step": 8394
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.593890241020958,
+      "learning_rate": 1.2496992040312045e-05,
+      "loss": 0.655,
+      "step": 8395
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.28844610735789,
+      "learning_rate": 1.2495359025134939e-05,
+      "loss": 0.5972,
+      "step": 8396
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.615189672600541,
+      "learning_rate": 1.2493725938991283e-05,
+      "loss": 0.5638,
+      "step": 8397
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.034943350656782,
+      "learning_rate": 1.2492092781927517e-05,
+      "loss": 0.6228,
+      "step": 8398
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.604180662260138,
+      "learning_rate": 1.2490459553990095e-05,
+      "loss": 0.6158,
+      "step": 8399
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 25.844669714295254,
+      "learning_rate": 1.2488826255225455e-05,
+      "loss": 0.6275,
+      "step": 8400
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.912448924384213,
+      "learning_rate": 1.2487192885680053e-05,
+      "loss": 0.6135,
+      "step": 8401
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.16203081307279,
+      "learning_rate": 1.2485559445400343e-05,
+      "loss": 0.6634,
+      "step": 8402
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.671311987717258,
+      "learning_rate": 1.2483925934432774e-05,
+      "loss": 0.6482,
+      "step": 8403
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 25.97376114848387,
+      "learning_rate": 1.2482292352823806e-05,
+      "loss": 0.6776,
+      "step": 8404
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 25.836135582904237,
+      "learning_rate": 1.2480658700619891e-05,
+      "loss": 0.6545,
+      "step": 8405
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.747727832968042,
+      "learning_rate": 1.24790249778675e-05,
+      "loss": 0.6735,
+      "step": 8406
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.52613928250092,
+      "learning_rate": 1.2477391184613086e-05,
+      "loss": 0.6163,
+      "step": 8407
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.685962541820313,
+      "learning_rate": 1.2475757320903117e-05,
+      "loss": 0.7279,
+      "step": 8408
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.845276350411908,
+      "learning_rate": 1.2474123386784059e-05,
+      "loss": 0.6353,
+      "step": 8409
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.96359114795371,
+      "learning_rate": 1.2472489382302377e-05,
+      "loss": 0.648,
+      "step": 8410
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.820878426393136,
+      "learning_rate": 1.2470855307504544e-05,
+      "loss": 0.6637,
+      "step": 8411
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.835597795712292,
+      "learning_rate": 1.2469221162437033e-05,
+      "loss": 0.6047,
+      "step": 8412
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.10453886893205,
+      "learning_rate": 1.2467586947146313e-05,
+      "loss": 0.5672,
+      "step": 8413
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.020047285174645,
+      "learning_rate": 1.2465952661678866e-05,
+      "loss": 0.6774,
+      "step": 8414
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.203443262377164,
+      "learning_rate": 1.2464318306081167e-05,
+      "loss": 0.6614,
+      "step": 8415
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 25.966718314210883,
+      "learning_rate": 1.24626838803997e-05,
+      "loss": 0.626,
+      "step": 8416
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.429948645357744,
+      "learning_rate": 1.2461049384680938e-05,
+      "loss": 0.6385,
+      "step": 8417
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.410800452892463,
+      "learning_rate": 1.2459414818971376e-05,
+      "loss": 0.7324,
+      "step": 8418
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.28931554132851,
+      "learning_rate": 1.2457780183317496e-05,
+      "loss": 0.6756,
+      "step": 8419
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.80160122486069,
+      "learning_rate": 1.2456145477765782e-05,
+      "loss": 0.6449,
+      "step": 8420
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 26.596933002921876,
+      "learning_rate": 1.245451070236273e-05,
+      "loss": 0.7206,
+      "step": 8421
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.168266110261689,
+      "learning_rate": 1.2452875857154827e-05,
+      "loss": 0.6589,
+      "step": 8422
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.552074673880554,
+      "learning_rate": 1.2451240942188569e-05,
+      "loss": 0.7514,
+      "step": 8423
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.089566779995492,
+      "learning_rate": 1.2449605957510456e-05,
+      "loss": 0.5992,
+      "step": 8424
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.558456475792575,
+      "learning_rate": 1.244797090316698e-05,
+      "loss": 0.7114,
+      "step": 8425
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.750146685851046,
+      "learning_rate": 1.2446335779204647e-05,
+      "loss": 0.7159,
+      "step": 8426
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.94879042559237,
+      "learning_rate": 1.2444700585669952e-05,
+      "loss": 0.5895,
+      "step": 8427
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.606569309814386,
+      "learning_rate": 1.2443065322609402e-05,
+      "loss": 0.6628,
+      "step": 8428
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.28941968393588,
+      "learning_rate": 1.2441429990069507e-05,
+      "loss": 0.6623,
+      "step": 8429
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.82479284197508,
+      "learning_rate": 1.2439794588096767e-05,
+      "loss": 0.6832,
+      "step": 8430
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.99441871821392,
+      "learning_rate": 1.24381591167377e-05,
+      "loss": 0.6229,
+      "step": 8431
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.94365852748525,
+      "learning_rate": 1.243652357603881e-05,
+      "loss": 0.681,
+      "step": 8432
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.147005642907533,
+      "learning_rate": 1.243488796604662e-05,
+      "loss": 0.6554,
+      "step": 8433
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 28.43136384596959,
+      "learning_rate": 1.2433252286807635e-05,
+      "loss": 0.6552,
+      "step": 8434
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.172208024675033,
+      "learning_rate": 1.2431616538368383e-05,
+      "loss": 0.6674,
+      "step": 8435
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.567540897048898,
+      "learning_rate": 1.2429980720775376e-05,
+      "loss": 0.6981,
+      "step": 8436
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.623454990605445,
+      "learning_rate": 1.2428344834075142e-05,
+      "loss": 0.6922,
+      "step": 8437
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.0994253181439,
+      "learning_rate": 1.2426708878314197e-05,
+      "loss": 0.6782,
+      "step": 8438
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.673935467175266,
+      "learning_rate": 1.2425072853539076e-05,
+      "loss": 0.6562,
+      "step": 8439
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.8969765086303,
+      "learning_rate": 1.2423436759796296e-05,
+      "loss": 0.6404,
+      "step": 8440
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.58288717159023,
+      "learning_rate": 1.2421800597132399e-05,
+      "loss": 0.6153,
+      "step": 8441
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.26700704340193,
+      "learning_rate": 1.2420164365593903e-05,
+      "loss": 0.5948,
+      "step": 8442
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.903703649971675,
+      "learning_rate": 1.2418528065227354e-05,
+      "loss": 0.5366,
+      "step": 8443
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.758806024914684,
+      "learning_rate": 1.2416891696079276e-05,
+      "loss": 0.6088,
+      "step": 8444
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.369139554869104,
+      "learning_rate": 1.2415255258196215e-05,
+      "loss": 0.6122,
+      "step": 8445
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 24.64057546391706,
+      "learning_rate": 1.2413618751624708e-05,
+      "loss": 0.6863,
+      "step": 8446
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.249161263125092,
+      "learning_rate": 1.2411982176411294e-05,
+      "loss": 0.6737,
+      "step": 8447
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.616167868424817,
+      "learning_rate": 1.2410345532602518e-05,
+      "loss": 0.679,
+      "step": 8448
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.542266887724022,
+      "learning_rate": 1.2408708820244926e-05,
+      "loss": 0.6725,
+      "step": 8449
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.2780886429179,
+      "learning_rate": 1.2407072039385064e-05,
+      "loss": 0.6734,
+      "step": 8450
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 23.760429935599344,
+      "learning_rate": 1.2405435190069481e-05,
+      "loss": 0.6293,
+      "step": 8451
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.352998041416104,
+      "learning_rate": 1.2403798272344729e-05,
+      "loss": 0.7218,
+      "step": 8452
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.34735796232701,
+      "learning_rate": 1.240216128625736e-05,
+      "loss": 0.6762,
+      "step": 8453
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.848365222889896,
+      "learning_rate": 1.2400524231853929e-05,
+      "loss": 0.6442,
+      "step": 8454
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.527076288966455,
+      "learning_rate": 1.2398887109180992e-05,
+      "loss": 0.651,
+      "step": 8455
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.76830119459625,
+      "learning_rate": 1.2397249918285113e-05,
+      "loss": 0.5973,
+      "step": 8456
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.009791511833946,
+      "learning_rate": 1.2395612659212844e-05,
+      "loss": 0.6102,
+      "step": 8457
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.919847215104042,
+      "learning_rate": 1.2393975332010757e-05,
+      "loss": 0.6027,
+      "step": 8458
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.250803343229645,
+      "learning_rate": 1.2392337936725408e-05,
+      "loss": 0.6422,
+      "step": 8459
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.980697480658229,
+      "learning_rate": 1.239070047340337e-05,
+      "loss": 0.5356,
+      "step": 8460
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 36.12678984436896,
+      "learning_rate": 1.2389062942091209e-05,
+      "loss": 0.6401,
+      "step": 8461
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.517682775584007,
+      "learning_rate": 1.2387425342835492e-05,
+      "loss": 0.6006,
+      "step": 8462
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.33834159567959,
+      "learning_rate": 1.2385787675682799e-05,
+      "loss": 0.61,
+      "step": 8463
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.59121940021831,
+      "learning_rate": 1.2384149940679697e-05,
+      "loss": 0.7406,
+      "step": 8464
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.764711292442753,
+      "learning_rate": 1.2382512137872769e-05,
+      "loss": 0.6407,
+      "step": 8465
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.010763261304934,
+      "learning_rate": 1.2380874267308586e-05,
+      "loss": 0.656,
+      "step": 8466
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.710717990713249,
+      "learning_rate": 1.237923632903373e-05,
+      "loss": 0.6055,
+      "step": 8467
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 34.495774283715484,
+      "learning_rate": 1.2377598323094788e-05,
+      "loss": 0.6495,
+      "step": 8468
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.69431043267568,
+      "learning_rate": 1.2375960249538341e-05,
+      "loss": 0.6576,
+      "step": 8469
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.511810830278588,
+      "learning_rate": 1.2374322108410974e-05,
+      "loss": 0.6334,
+      "step": 8470
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.853325801537327,
+      "learning_rate": 1.2372683899759274e-05,
+      "loss": 0.6293,
+      "step": 8471
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.866740655037848,
+      "learning_rate": 1.2371045623629834e-05,
+      "loss": 0.5808,
+      "step": 8472
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.115381657602597,
+      "learning_rate": 1.2369407280069241e-05,
+      "loss": 0.6865,
+      "step": 8473
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.801582747227275,
+      "learning_rate": 1.2367768869124091e-05,
+      "loss": 0.6638,
+      "step": 8474
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.664340221147935,
+      "learning_rate": 1.2366130390840982e-05,
+      "loss": 0.6152,
+      "step": 8475
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.415242830972065,
+      "learning_rate": 1.2364491845266506e-05,
+      "loss": 0.5968,
+      "step": 8476
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.067682735023382,
+      "learning_rate": 1.2362853232447267e-05,
+      "loss": 0.5581,
+      "step": 8477
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.29645373881489,
+      "learning_rate": 1.2361214552429863e-05,
+      "loss": 0.6734,
+      "step": 8478
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.764378467558146,
+      "learning_rate": 1.23595758052609e-05,
+      "loss": 0.6377,
+      "step": 8479
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 26.29545478929641,
+      "learning_rate": 1.235793699098698e-05,
+      "loss": 0.5976,
+      "step": 8480
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.952746884278923,
+      "learning_rate": 1.2356298109654712e-05,
+      "loss": 0.6185,
+      "step": 8481
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.902903533041194,
+      "learning_rate": 1.2354659161310704e-05,
+      "loss": 0.6704,
+      "step": 8482
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.132129194364175,
+      "learning_rate": 1.2353020146001568e-05,
+      "loss": 0.648,
+      "step": 8483
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.529667552345707,
+      "learning_rate": 1.2351381063773913e-05,
+      "loss": 0.6833,
+      "step": 8484
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.40850758105997,
+      "learning_rate": 1.2349741914674361e-05,
+      "loss": 0.6454,
+      "step": 8485
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.17867539293414,
+      "learning_rate": 1.2348102698749518e-05,
+      "loss": 0.6472,
+      "step": 8486
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 21.911434192083497,
+      "learning_rate": 1.234646341604601e-05,
+      "loss": 0.6109,
+      "step": 8487
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 23.287101348233122,
+      "learning_rate": 1.2344824066610454e-05,
+      "loss": 0.6755,
+      "step": 8488
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.26716566563346,
+      "learning_rate": 1.2343184650489476e-05,
+      "loss": 0.6091,
+      "step": 8489
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.40588813351155,
+      "learning_rate": 1.2341545167729693e-05,
+      "loss": 0.5919,
+      "step": 8490
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 20.77361262600216,
+      "learning_rate": 1.2339905618377739e-05,
+      "loss": 0.6741,
+      "step": 8491
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 21.582979000219552,
+      "learning_rate": 1.2338266002480237e-05,
+      "loss": 0.5845,
+      "step": 8492
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.768318299826905,
+      "learning_rate": 1.2336626320083816e-05,
+      "loss": 0.6034,
+      "step": 8493
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.772503873865713,
+      "learning_rate": 1.233498657123511e-05,
+      "loss": 0.6968,
+      "step": 8494
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 25.65378679074082,
+      "learning_rate": 1.2333346755980753e-05,
+      "loss": 0.7412,
+      "step": 8495
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.7526211061879,
+      "learning_rate": 1.233170687436738e-05,
+      "loss": 0.6636,
+      "step": 8496
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.085844378766186,
+      "learning_rate": 1.2330066926441626e-05,
+      "loss": 0.6107,
+      "step": 8497
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 20.880085868092525,
+      "learning_rate": 1.232842691225013e-05,
+      "loss": 0.6404,
+      "step": 8498
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.821906419986114,
+      "learning_rate": 1.2326786831839536e-05,
+      "loss": 0.5547,
+      "step": 8499
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.788811926836328,
+      "learning_rate": 1.2325146685256489e-05,
+      "loss": 0.6134,
+      "step": 8500
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.759688831202755,
+      "learning_rate": 1.2323506472547626e-05,
+      "loss": 0.6144,
+      "step": 8501
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.276056399474264,
+      "learning_rate": 1.2321866193759602e-05,
+      "loss": 0.6455,
+      "step": 8502
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.869654232969632,
+      "learning_rate": 1.2320225848939059e-05,
+      "loss": 0.6269,
+      "step": 8503
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 22.588172582308815,
+      "learning_rate": 1.2318585438132654e-05,
+      "loss": 0.6206,
+      "step": 8504
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 23.0132422032763,
+      "learning_rate": 1.2316944961387028e-05,
+      "loss": 0.5531,
+      "step": 8505
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.95920531327305,
+      "learning_rate": 1.2315304418748848e-05,
+      "loss": 0.6361,
+      "step": 8506
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 21.701476311867843,
+      "learning_rate": 1.2313663810264762e-05,
+      "loss": 0.6137,
+      "step": 8507
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.3641054747069,
+      "learning_rate": 1.2312023135981434e-05,
+      "loss": 0.6001,
+      "step": 8508
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.87934186055937,
+      "learning_rate": 1.231038239594552e-05,
+      "loss": 0.616,
+      "step": 8509
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.62906589022556,
+      "learning_rate": 1.230874159020368e-05,
+      "loss": 0.6292,
+      "step": 8510
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.017662707481655,
+      "learning_rate": 1.2307100718802579e-05,
+      "loss": 0.5669,
+      "step": 8511
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.850245509893645,
+      "learning_rate": 1.2305459781788885e-05,
+      "loss": 0.6347,
+      "step": 8512
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.579043334833766,
+      "learning_rate": 1.2303818779209264e-05,
+      "loss": 0.6614,
+      "step": 8513
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.089340029140562,
+      "learning_rate": 1.2302177711110384e-05,
+      "loss": 0.6952,
+      "step": 8514
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 21.053311086757766,
+      "learning_rate": 1.2300536577538917e-05,
+      "loss": 0.6103,
+      "step": 8515
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.962015982455814,
+      "learning_rate": 1.2298895378541536e-05,
+      "loss": 0.5827,
+      "step": 8516
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.934177167292416,
+      "learning_rate": 1.2297254114164914e-05,
+      "loss": 0.6862,
+      "step": 8517
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 22.84563689781987,
+      "learning_rate": 1.2295612784455728e-05,
+      "loss": 0.6806,
+      "step": 8518
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.973864316809564,
+      "learning_rate": 1.2293971389460659e-05,
+      "loss": 0.6555,
+      "step": 8519
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.418891846811245,
+      "learning_rate": 1.2292329929226385e-05,
+      "loss": 0.6581,
+      "step": 8520
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.18932821156568,
+      "learning_rate": 1.229068840379959e-05,
+      "loss": 0.606,
+      "step": 8521
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.743693873701552,
+      "learning_rate": 1.2289046813226954e-05,
+      "loss": 0.6379,
+      "step": 8522
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.77122972883461,
+      "learning_rate": 1.2287405157555168e-05,
+      "loss": 0.6134,
+      "step": 8523
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 23.03307996425988,
+      "learning_rate": 1.2285763436830917e-05,
+      "loss": 0.675,
+      "step": 8524
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.659364245108755,
+      "learning_rate": 1.2284121651100891e-05,
+      "loss": 0.5764,
+      "step": 8525
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 26.19314833728055,
+      "learning_rate": 1.228247980041178e-05,
+      "loss": 0.6735,
+      "step": 8526
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.713701830037026,
+      "learning_rate": 1.2280837884810282e-05,
+      "loss": 0.591,
+      "step": 8527
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.15943430902438,
+      "learning_rate": 1.2279195904343084e-05,
+      "loss": 0.6002,
+      "step": 8528
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.048214151727812,
+      "learning_rate": 1.2277553859056894e-05,
+      "loss": 0.6848,
+      "step": 8529
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.690646851684782,
+      "learning_rate": 1.22759117489984e-05,
+      "loss": 0.5433,
+      "step": 8530
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.987558771553786,
+      "learning_rate": 1.227426957421431e-05,
+      "loss": 0.5566,
+      "step": 8531
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.66383854167598,
+      "learning_rate": 1.227262733475132e-05,
+      "loss": 0.6899,
+      "step": 8532
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.022869830178324,
+      "learning_rate": 1.2270985030656139e-05,
+      "loss": 0.5529,
+      "step": 8533
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.35945595900608,
+      "learning_rate": 1.2269342661975474e-05,
+      "loss": 0.6456,
+      "step": 8534
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.18422387360619,
+      "learning_rate": 1.2267700228756033e-05,
+      "loss": 0.6298,
+      "step": 8535
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.660415692858141,
+      "learning_rate": 1.2266057731044521e-05,
+      "loss": 0.6397,
+      "step": 8536
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.309528642527022,
+      "learning_rate": 1.2264415168887651e-05,
+      "loss": 0.5893,
+      "step": 8537
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 21.20277945083012,
+      "learning_rate": 1.226277254233214e-05,
+      "loss": 0.6605,
+      "step": 8538
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.896931020350566,
+      "learning_rate": 1.2261129851424703e-05,
+      "loss": 0.5987,
+      "step": 8539
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.659557426552695,
+      "learning_rate": 1.2259487096212055e-05,
+      "loss": 0.66,
+      "step": 8540
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.572156330212678,
+      "learning_rate": 1.2257844276740916e-05,
+      "loss": 0.6837,
+      "step": 8541
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.48955546162416,
+      "learning_rate": 1.2256201393058006e-05,
+      "loss": 0.6782,
+      "step": 8542
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.208081511215612,
+      "learning_rate": 1.2254558445210048e-05,
+      "loss": 0.5595,
+      "step": 8543
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.310728373998613,
+      "learning_rate": 1.2252915433243768e-05,
+      "loss": 0.5796,
+      "step": 8544
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.754076669892413,
+      "learning_rate": 1.225127235720589e-05,
+      "loss": 0.6264,
+      "step": 8545
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.674181956880906,
+      "learning_rate": 1.2249629217143143e-05,
+      "loss": 0.5549,
+      "step": 8546
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.92216467619158,
+      "learning_rate": 1.2247986013102258e-05,
+      "loss": 0.6989,
+      "step": 8547
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.064078355667892,
+      "learning_rate": 1.2246342745129964e-05,
+      "loss": 0.6911,
+      "step": 8548
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.170920627284424,
+      "learning_rate": 1.2244699413272998e-05,
+      "loss": 0.6604,
+      "step": 8549
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.94594602994399,
+      "learning_rate": 1.2243056017578095e-05,
+      "loss": 0.6326,
+      "step": 8550
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.776293552586147,
+      "learning_rate": 1.2241412558091988e-05,
+      "loss": 0.6862,
+      "step": 8551
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.479472675607774,
+      "learning_rate": 1.2239769034861423e-05,
+      "loss": 0.6175,
+      "step": 8552
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.63204416552825,
+      "learning_rate": 1.2238125447933134e-05,
+      "loss": 0.6443,
+      "step": 8553
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 20.82843613445841,
+      "learning_rate": 1.2236481797353865e-05,
+      "loss": 0.6338,
+      "step": 8554
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.37390498443152,
+      "learning_rate": 1.2234838083170362e-05,
+      "loss": 0.5719,
+      "step": 8555
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.549117574979292,
+      "learning_rate": 1.2233194305429375e-05,
+      "loss": 0.6037,
+      "step": 8556
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.57356127272504,
+      "learning_rate": 1.2231550464177646e-05,
+      "loss": 0.5215,
+      "step": 8557
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.570649246300786,
+      "learning_rate": 1.2229906559461925e-05,
+      "loss": 0.6638,
+      "step": 8558
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.945698632700093,
+      "learning_rate": 1.222826259132897e-05,
+      "loss": 0.6179,
+      "step": 8559
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 26.915847750115947,
+      "learning_rate": 1.2226618559825529e-05,
+      "loss": 0.6985,
+      "step": 8560
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.076808096148143,
+      "learning_rate": 1.222497446499836e-05,
+      "loss": 0.5829,
+      "step": 8561
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 25.728779161227575,
+      "learning_rate": 1.2223330306894215e-05,
+      "loss": 0.6485,
+      "step": 8562
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.566633584150182,
+      "learning_rate": 1.222168608555986e-05,
+      "loss": 0.6531,
+      "step": 8563
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.665807427988359,
+      "learning_rate": 1.222004180104205e-05,
+      "loss": 0.6951,
+      "step": 8564
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.09666819832233,
+      "learning_rate": 1.2218397453387551e-05,
+      "loss": 0.8099,
+      "step": 8565
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 22.251314463926256,
+      "learning_rate": 1.2216753042643128e-05,
+      "loss": 0.656,
+      "step": 8566
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.174028588709266,
+      "learning_rate": 1.2215108568855545e-05,
+      "loss": 0.5178,
+      "step": 8567
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.88891556827831,
+      "learning_rate": 1.2213464032071567e-05,
+      "loss": 0.5962,
+      "step": 8568
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.799382599683735,
+      "learning_rate": 1.221181943233797e-05,
+      "loss": 0.5768,
+      "step": 8569
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.91518573033513,
+      "learning_rate": 1.221017476970152e-05,
+      "loss": 0.6383,
+      "step": 8570
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.456165007982904,
+      "learning_rate": 1.2208530044208995e-05,
+      "loss": 0.6282,
+      "step": 8571
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.63572919042834,
+      "learning_rate": 1.2206885255907163e-05,
+      "loss": 0.6224,
+      "step": 8572
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 22.210550375297608,
+      "learning_rate": 1.2205240404842811e-05,
+      "loss": 0.5794,
+      "step": 8573
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 25.877061355759512,
+      "learning_rate": 1.2203595491062707e-05,
+      "loss": 0.6282,
+      "step": 8574
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.423044212995645,
+      "learning_rate": 1.2201950514613638e-05,
+      "loss": 0.6323,
+      "step": 8575
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.954159829670836,
+      "learning_rate": 1.2200305475542385e-05,
+      "loss": 0.6524,
+      "step": 8576
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.71081158954826,
+      "learning_rate": 1.2198660373895731e-05,
+      "loss": 0.5714,
+      "step": 8577
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.267509780902145,
+      "learning_rate": 1.2197015209720462e-05,
+      "loss": 0.5689,
+      "step": 8578
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 26.785402962476226,
+      "learning_rate": 1.2195369983063368e-05,
+      "loss": 0.6925,
+      "step": 8579
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.482483312382136,
+      "learning_rate": 1.2193724693971235e-05,
+      "loss": 0.6086,
+      "step": 8580
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 20.127896025743652,
+      "learning_rate": 1.2192079342490851e-05,
+      "loss": 0.6026,
+      "step": 8581
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.353481856921645,
+      "learning_rate": 1.2190433928669015e-05,
+      "loss": 0.5335,
+      "step": 8582
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.15072360140765,
+      "learning_rate": 1.218878845255252e-05,
+      "loss": 0.6624,
+      "step": 8583
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.802699797114263,
+      "learning_rate": 1.2187142914188161e-05,
+      "loss": 0.6546,
+      "step": 8584
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.262017076436383,
+      "learning_rate": 1.218549731362274e-05,
+      "loss": 0.6721,
+      "step": 8585
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.179441932199305,
+      "learning_rate": 1.218385165090305e-05,
+      "loss": 0.5582,
+      "step": 8586
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.147349544241326,
+      "learning_rate": 1.2182205926075899e-05,
+      "loss": 0.6305,
+      "step": 8587
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.79689105727175,
+      "learning_rate": 1.2180560139188088e-05,
+      "loss": 0.6199,
+      "step": 8588
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.124370497161896,
+      "learning_rate": 1.217891429028642e-05,
+      "loss": 0.5722,
+      "step": 8589
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.527256327436305,
+      "learning_rate": 1.2177268379417708e-05,
+      "loss": 0.6247,
+      "step": 8590
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.956928681494368,
+      "learning_rate": 1.2175622406628754e-05,
+      "loss": 0.549,
+      "step": 8591
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.97446677005685,
+      "learning_rate": 1.2173976371966372e-05,
+      "loss": 0.6974,
+      "step": 8592
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.493586167850015,
+      "learning_rate": 1.2172330275477374e-05,
+      "loss": 0.6115,
+      "step": 8593
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.941489858062244,
+      "learning_rate": 1.2170684117208573e-05,
+      "loss": 0.6264,
+      "step": 8594
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.27002217046163,
+      "learning_rate": 1.2169037897206787e-05,
+      "loss": 0.6026,
+      "step": 8595
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.086005392739107,
+      "learning_rate": 1.2167391615518831e-05,
+      "loss": 0.6347,
+      "step": 8596
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.84576977340459,
+      "learning_rate": 1.2165745272191524e-05,
+      "loss": 0.6643,
+      "step": 8597
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.689365290258255,
+      "learning_rate": 1.2164098867271694e-05,
+      "loss": 0.7244,
+      "step": 8598
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.649201631581684,
+      "learning_rate": 1.2162452400806151e-05,
+      "loss": 0.5708,
+      "step": 8599
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.300732506662012,
+      "learning_rate": 1.216080587284173e-05,
+      "loss": 0.6756,
+      "step": 8600
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.766556312517203,
+      "learning_rate": 1.2159159283425254e-05,
+      "loss": 0.6162,
+      "step": 8601
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 26.39327628356211,
+      "learning_rate": 1.2157512632603553e-05,
+      "loss": 0.6694,
+      "step": 8602
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 24.74339040488667,
+      "learning_rate": 1.215586592042345e-05,
+      "loss": 0.6752,
+      "step": 8603
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.849812490823286,
+      "learning_rate": 1.2154219146931786e-05,
+      "loss": 0.6391,
+      "step": 8604
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.199938153408523,
+      "learning_rate": 1.2152572312175388e-05,
+      "loss": 0.579,
+      "step": 8605
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.43259540466348,
+      "learning_rate": 1.2150925416201091e-05,
+      "loss": 0.6109,
+      "step": 8606
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.49210279563776,
+      "learning_rate": 1.2149278459055737e-05,
+      "loss": 0.6209,
+      "step": 8607
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 20.70374795012466,
+      "learning_rate": 1.2147631440786156e-05,
+      "loss": 0.6015,
+      "step": 8608
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.84303984924125,
+      "learning_rate": 1.2145984361439197e-05,
+      "loss": 0.6145,
+      "step": 8609
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.83683998891923,
+      "learning_rate": 1.2144337221061697e-05,
+      "loss": 0.6278,
+      "step": 8610
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.27728890070193,
+      "learning_rate": 1.2142690019700503e-05,
+      "loss": 0.6035,
+      "step": 8611
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.829886528611222,
+      "learning_rate": 1.2141042757402454e-05,
+      "loss": 0.59,
+      "step": 8612
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.899656288015105,
+      "learning_rate": 1.2139395434214406e-05,
+      "loss": 0.584,
+      "step": 8613
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.5394635526347,
+      "learning_rate": 1.2137748050183197e-05,
+      "loss": 0.6026,
+      "step": 8614
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.832506498603806,
+      "learning_rate": 1.2136100605355691e-05,
+      "loss": 0.618,
+      "step": 8615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 31.684148808281535,
+      "learning_rate": 1.213445309977873e-05,
+      "loss": 0.6659,
+      "step": 8616
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.11390394060753,
+      "learning_rate": 1.2132805533499172e-05,
+      "loss": 0.6353,
+      "step": 8617
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.570365225381163,
+      "learning_rate": 1.213115790656387e-05,
+      "loss": 0.6731,
+      "step": 8618
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.545611620505994,
+      "learning_rate": 1.2129510219019686e-05,
+      "loss": 0.6404,
+      "step": 8619
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.941143432735313,
+      "learning_rate": 1.2127862470913478e-05,
+      "loss": 0.5982,
+      "step": 8620
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.405501643083866,
+      "learning_rate": 1.2126214662292106e-05,
+      "loss": 0.6401,
+      "step": 8621
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 21.49025598247731,
+      "learning_rate": 1.2124566793202432e-05,
+      "loss": 0.6717,
+      "step": 8622
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.20007527781142,
+      "learning_rate": 1.2122918863691322e-05,
+      "loss": 0.6624,
+      "step": 8623
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.363786412918277,
+      "learning_rate": 1.2121270873805638e-05,
+      "loss": 0.5814,
+      "step": 8624
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.313352477778924,
+      "learning_rate": 1.2119622823592254e-05,
+      "loss": 0.6372,
+      "step": 8625
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.40802895622932,
+      "learning_rate": 1.2117974713098038e-05,
+      "loss": 0.6714,
+      "step": 8626
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.594557231218044,
+      "learning_rate": 1.2116326542369859e-05,
+      "loss": 0.5989,
+      "step": 8627
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.18178038960349,
+      "learning_rate": 1.211467831145459e-05,
+      "loss": 0.5986,
+      "step": 8628
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.942473229376064,
+      "learning_rate": 1.2113030020399107e-05,
+      "loss": 0.6953,
+      "step": 8629
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.832735425818143,
+      "learning_rate": 1.2111381669250288e-05,
+      "loss": 0.6188,
+      "step": 8630
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 26.321500638717836,
+      "learning_rate": 1.2109733258055007e-05,
+      "loss": 0.6906,
+      "step": 8631
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 21.573466512152592,
+      "learning_rate": 1.210808478686015e-05,
+      "loss": 0.6716,
+      "step": 8632
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.15335020999387,
+      "learning_rate": 1.210643625571259e-05,
+      "loss": 0.6121,
+      "step": 8633
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.133454025061795,
+      "learning_rate": 1.2104787664659221e-05,
+      "loss": 0.6633,
+      "step": 8634
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.280080881719982,
+      "learning_rate": 1.2103139013746919e-05,
+      "loss": 0.6343,
+      "step": 8635
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 27.531134442791593,
+      "learning_rate": 1.2101490303022571e-05,
+      "loss": 0.702,
+      "step": 8636
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.797471289585328,
+      "learning_rate": 1.2099841532533073e-05,
+      "loss": 0.6256,
+      "step": 8637
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.276726431224354,
+      "learning_rate": 1.2098192702325309e-05,
+      "loss": 0.6478,
+      "step": 8638
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.89552403801863,
+      "learning_rate": 1.209654381244617e-05,
+      "loss": 0.6902,
+      "step": 8639
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.958346612008572,
+      "learning_rate": 1.2094894862942552e-05,
+      "loss": 0.6106,
+      "step": 8640
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.678519720501198,
+      "learning_rate": 1.2093245853861349e-05,
+      "loss": 0.657,
+      "step": 8641
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.707797702739459,
+      "learning_rate": 1.2091596785249461e-05,
+      "loss": 0.6076,
+      "step": 8642
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 21.43371740228696,
+      "learning_rate": 1.208994765715378e-05,
+      "loss": 0.647,
+      "step": 8643
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 27.12832496025796,
+      "learning_rate": 1.2088298469621214e-05,
+      "loss": 0.6718,
+      "step": 8644
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 24.98985447894854,
+      "learning_rate": 1.208664922269866e-05,
+      "loss": 0.5796,
+      "step": 8645
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.767594889857637,
+      "learning_rate": 1.2084999916433022e-05,
+      "loss": 0.5757,
+      "step": 8646
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.86950795375119,
+      "learning_rate": 1.2083350550871206e-05,
+      "loss": 0.6389,
+      "step": 8647
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.182343188120832,
+      "learning_rate": 1.2081701126060121e-05,
+      "loss": 0.6121,
+      "step": 8648
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 25.74199996439996,
+      "learning_rate": 1.2080051642046674e-05,
+      "loss": 0.6565,
+      "step": 8649
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.432182822891118,
+      "learning_rate": 1.2078402098877771e-05,
+      "loss": 0.6103,
+      "step": 8650
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.90295544119717,
+      "learning_rate": 1.2076752496600333e-05,
+      "loss": 0.6326,
+      "step": 8651
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.870214430344216,
+      "learning_rate": 1.2075102835261264e-05,
+      "loss": 0.6406,
+      "step": 8652
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.562156110943068,
+      "learning_rate": 1.2073453114907485e-05,
+      "loss": 0.665,
+      "step": 8653
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.820579893890073,
+      "learning_rate": 1.2071803335585917e-05,
+      "loss": 0.6558,
+      "step": 8654
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 42.492516257039505,
+      "learning_rate": 1.207015349734347e-05,
+      "loss": 0.6626,
+      "step": 8655
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.752472217332123,
+      "learning_rate": 1.206850360022707e-05,
+      "loss": 0.7339,
+      "step": 8656
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.302521378499183,
+      "learning_rate": 1.2066853644283639e-05,
+      "loss": 0.6109,
+      "step": 8657
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.986431140602697,
+      "learning_rate": 1.2065203629560093e-05,
+      "loss": 0.5829,
+      "step": 8658
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.375647208678277,
+      "learning_rate": 1.2063553556103372e-05,
+      "loss": 0.5178,
+      "step": 8659
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.58411080173229,
+      "learning_rate": 1.206190342396039e-05,
+      "loss": 0.67,
+      "step": 8660
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.888372249392383,
+      "learning_rate": 1.2060253233178086e-05,
+      "loss": 0.5886,
+      "step": 8661
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.246519981283104,
+      "learning_rate": 1.2058602983803378e-05,
+      "loss": 0.6475,
+      "step": 8662
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.864418853513957,
+      "learning_rate": 1.2056952675883208e-05,
+      "loss": 0.6008,
+      "step": 8663
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.578250072884106,
+      "learning_rate": 1.2055302309464509e-05,
+      "loss": 0.5911,
+      "step": 8664
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.716578899167196,
+      "learning_rate": 1.2053651884594215e-05,
+      "loss": 0.6514,
+      "step": 8665
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.838624148899495,
+      "learning_rate": 1.2052001401319262e-05,
+      "loss": 0.6989,
+      "step": 8666
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.869598103963412,
+      "learning_rate": 1.205035085968659e-05,
+      "loss": 0.7058,
+      "step": 8667
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 25.229050145928856,
+      "learning_rate": 1.2048700259743136e-05,
+      "loss": 0.6901,
+      "step": 8668
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.2123019603452105,
+      "learning_rate": 1.2047049601535847e-05,
+      "loss": 0.602,
+      "step": 8669
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 21.594150492574343,
+      "learning_rate": 1.2045398885111665e-05,
+      "loss": 0.6425,
+      "step": 8670
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.441651733548678,
+      "learning_rate": 1.2043748110517536e-05,
+      "loss": 0.6666,
+      "step": 8671
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.520323122735798,
+      "learning_rate": 1.2042097277800406e-05,
+      "loss": 0.5862,
+      "step": 8672
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.330987991507707,
+      "learning_rate": 1.2040446387007222e-05,
+      "loss": 0.6194,
+      "step": 8673
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.146208037695654,
+      "learning_rate": 1.203879543818494e-05,
+      "loss": 0.6213,
+      "step": 8674
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.66988023487055,
+      "learning_rate": 1.2037144431380506e-05,
+      "loss": 0.5863,
+      "step": 8675
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.40602767817765,
+      "learning_rate": 1.2035493366640879e-05,
+      "loss": 0.5943,
+      "step": 8676
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 23.25615465959148,
+      "learning_rate": 1.2033842244013006e-05,
+      "loss": 0.5958,
+      "step": 8677
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.37008311490887,
+      "learning_rate": 1.2032191063543855e-05,
+      "loss": 0.6931,
+      "step": 8678
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.431357786931859,
+      "learning_rate": 1.2030539825280373e-05,
+      "loss": 0.5809,
+      "step": 8679
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.634978234535726,
+      "learning_rate": 1.202888852926953e-05,
+      "loss": 0.706,
+      "step": 8680
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.180166034206074,
+      "learning_rate": 1.2027237175558283e-05,
+      "loss": 0.6727,
+      "step": 8681
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.274704392230202,
+      "learning_rate": 1.2025585764193597e-05,
+      "loss": 0.6655,
+      "step": 8682
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.758746647604836,
+      "learning_rate": 1.2023934295222437e-05,
+      "loss": 0.589,
+      "step": 8683
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.147222840172123,
+      "learning_rate": 1.202228276869177e-05,
+      "loss": 0.7426,
+      "step": 8684
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.242841270216683,
+      "learning_rate": 1.202063118464856e-05,
+      "loss": 0.5979,
+      "step": 8685
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.265055188701282,
+      "learning_rate": 1.2018979543139788e-05,
+      "loss": 0.6719,
+      "step": 8686
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.970906259754297,
+      "learning_rate": 1.2017327844212414e-05,
+      "loss": 0.573,
+      "step": 8687
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.97843936655783,
+      "learning_rate": 1.2015676087913418e-05,
+      "loss": 0.6731,
+      "step": 8688
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 26.719685570471505,
+      "learning_rate": 1.2014024274289773e-05,
+      "loss": 0.7976,
+      "step": 8689
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.657509474436512,
+      "learning_rate": 1.2012372403388457e-05,
+      "loss": 0.6603,
+      "step": 8690
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.92491280555814,
+      "learning_rate": 1.2010720475256446e-05,
+      "loss": 0.6932,
+      "step": 8691
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.058209762427303,
+      "learning_rate": 1.200906848994072e-05,
+      "loss": 0.6104,
+      "step": 8692
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.012169296408299,
+      "learning_rate": 1.2007416447488263e-05,
+      "loss": 0.6158,
+      "step": 8693
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.49370475224746,
+      "learning_rate": 1.2005764347946053e-05,
+      "loss": 0.6646,
+      "step": 8694
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.485652090188943,
+      "learning_rate": 1.200411219136108e-05,
+      "loss": 0.5479,
+      "step": 8695
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.097223393070108,
+      "learning_rate": 1.2002459977780331e-05,
+      "loss": 0.6201,
+      "step": 8696
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.79586073218258,
+      "learning_rate": 1.200080770725079e-05,
+      "loss": 0.6067,
+      "step": 8697
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.52462496183499,
+      "learning_rate": 1.1999155379819449e-05,
+      "loss": 0.5849,
+      "step": 8698
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.27744664362888,
+      "learning_rate": 1.1997502995533299e-05,
+      "loss": 0.7032,
+      "step": 8699
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 23.678167306299112,
+      "learning_rate": 1.1995850554439332e-05,
+      "loss": 0.6332,
+      "step": 8700
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 27.32049155237977,
+      "learning_rate": 1.199419805658454e-05,
+      "loss": 0.7102,
+      "step": 8701
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.668013411662498,
+      "learning_rate": 1.1992545502015923e-05,
+      "loss": 0.6285,
+      "step": 8702
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.35178649391694,
+      "learning_rate": 1.1990892890780482e-05,
+      "loss": 0.6692,
+      "step": 8703
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.67873565852468,
+      "learning_rate": 1.1989240222925206e-05,
+      "loss": 0.6064,
+      "step": 8704
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.204817147330086,
+      "learning_rate": 1.1987587498497107e-05,
+      "loss": 0.6411,
+      "step": 8705
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.827820825019513,
+      "learning_rate": 1.1985934717543178e-05,
+      "loss": 0.6193,
+      "step": 8706
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.466607331382953,
+      "learning_rate": 1.1984281880110427e-05,
+      "loss": 0.664,
+      "step": 8707
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.920055425695603,
+      "learning_rate": 1.198262898624586e-05,
+      "loss": 0.6509,
+      "step": 8708
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.04460874009785,
+      "learning_rate": 1.1980976035996488e-05,
+      "loss": 0.5718,
+      "step": 8709
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.65760851930483,
+      "learning_rate": 1.1979323029409316e-05,
+      "loss": 0.6145,
+      "step": 8710
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.045625629775152,
+      "learning_rate": 1.1977669966531353e-05,
+      "loss": 0.616,
+      "step": 8711
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.667504673998017,
+      "learning_rate": 1.197601684740961e-05,
+      "loss": 0.6646,
+      "step": 8712
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.207071307030063,
+      "learning_rate": 1.197436367209111e-05,
+      "loss": 0.6228,
+      "step": 8713
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.933617073651327,
+      "learning_rate": 1.1972710440622858e-05,
+      "loss": 0.6474,
+      "step": 8714
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.28544016770247,
+      "learning_rate": 1.1971057153051878e-05,
+      "loss": 0.707,
+      "step": 8715
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.115497778856586,
+      "learning_rate": 1.1969403809425183e-05,
+      "loss": 0.702,
+      "step": 8716
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 24.303170521308342,
+      "learning_rate": 1.1967750409789796e-05,
+      "loss": 0.6456,
+      "step": 8717
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.41790776206668,
+      "learning_rate": 1.196609695419274e-05,
+      "loss": 0.6288,
+      "step": 8718
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.29997157684022,
+      "learning_rate": 1.1964443442681036e-05,
+      "loss": 0.667,
+      "step": 8719
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.997276983135546,
+      "learning_rate": 1.196278987530171e-05,
+      "loss": 0.5949,
+      "step": 8720
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.344192490004842,
+      "learning_rate": 1.1961136252101786e-05,
+      "loss": 0.5479,
+      "step": 8721
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.402597818887436,
+      "learning_rate": 1.1959482573128298e-05,
+      "loss": 0.659,
+      "step": 8722
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.250074908346086,
+      "learning_rate": 1.1957828838428269e-05,
+      "loss": 0.6589,
+      "step": 8723
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.78168894829748,
+      "learning_rate": 1.1956175048048734e-05,
+      "loss": 0.5787,
+      "step": 8724
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 26.64106913053749,
+      "learning_rate": 1.1954521202036726e-05,
+      "loss": 0.6823,
+      "step": 8725
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 25.159660220161182,
+      "learning_rate": 1.1952867300439276e-05,
+      "loss": 0.6809,
+      "step": 8726
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.948444492476462,
+      "learning_rate": 1.1951213343303425e-05,
+      "loss": 0.7293,
+      "step": 8727
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.259032591867555,
+      "learning_rate": 1.1949559330676209e-05,
+      "loss": 0.695,
+      "step": 8728
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.339335834764851,
+      "learning_rate": 1.194790526260466e-05,
+      "loss": 0.6094,
+      "step": 8729
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.742763378389107,
+      "learning_rate": 1.1946251139135831e-05,
+      "loss": 0.62,
+      "step": 8730
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.939018681191225,
+      "learning_rate": 1.1944596960316755e-05,
+      "loss": 0.5929,
+      "step": 8731
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.431779897274023,
+      "learning_rate": 1.1942942726194477e-05,
+      "loss": 0.6364,
+      "step": 8732
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.683718808267724,
+      "learning_rate": 1.194128843681605e-05,
+      "loss": 0.6156,
+      "step": 8733
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 24.38668062835554,
+      "learning_rate": 1.1939634092228511e-05,
+      "loss": 0.6481,
+      "step": 8734
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.017589097508985,
+      "learning_rate": 1.1937979692478915e-05,
+      "loss": 0.6234,
+      "step": 8735
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.01555311471506,
+      "learning_rate": 1.1936325237614312e-05,
+      "loss": 0.6355,
+      "step": 8736
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.86445135205235,
+      "learning_rate": 1.193467072768175e-05,
+      "loss": 0.5934,
+      "step": 8737
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.803704844700196,
+      "learning_rate": 1.1933016162728281e-05,
+      "loss": 0.5933,
+      "step": 8738
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.139628692497936,
+      "learning_rate": 1.1931361542800968e-05,
+      "loss": 0.5912,
+      "step": 8739
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.402821710042435,
+      "learning_rate": 1.192970686794686e-05,
+      "loss": 0.5911,
+      "step": 8740
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.761679372666222,
+      "learning_rate": 1.192805213821302e-05,
+      "loss": 0.6233,
+      "step": 8741
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.400814986295323,
+      "learning_rate": 1.1926397353646501e-05,
+      "loss": 0.602,
+      "step": 8742
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.277909358812714,
+      "learning_rate": 1.1924742514294371e-05,
+      "loss": 0.7104,
+      "step": 8743
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.739870507452032,
+      "learning_rate": 1.1923087620203688e-05,
+      "loss": 0.5877,
+      "step": 8744
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.39542170293932,
+      "learning_rate": 1.1921432671421523e-05,
+      "loss": 0.5963,
+      "step": 8745
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 23.54429346999875,
+      "learning_rate": 1.1919777667994932e-05,
+      "loss": 0.5965,
+      "step": 8746
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 27.29421663577674,
+      "learning_rate": 1.191812260997099e-05,
+      "loss": 0.6987,
+      "step": 8747
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.545700450931495,
+      "learning_rate": 1.1916467497396759e-05,
+      "loss": 0.6507,
+      "step": 8748
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.808383496813654,
+      "learning_rate": 1.1914812330319318e-05,
+      "loss": 0.6615,
+      "step": 8749
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.436963995758067,
+      "learning_rate": 1.1913157108785731e-05,
+      "loss": 0.7119,
+      "step": 8750
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.583041417119194,
+      "learning_rate": 1.1911501832843077e-05,
+      "loss": 0.6371,
+      "step": 8751
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.972580252186157,
+      "learning_rate": 1.1909846502538429e-05,
+      "loss": 0.6289,
+      "step": 8752
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 20.192975607421552,
+      "learning_rate": 1.1908191117918864e-05,
+      "loss": 0.6453,
+      "step": 8753
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.280189820668895,
+      "learning_rate": 1.190653567903146e-05,
+      "loss": 0.6793,
+      "step": 8754
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.86481619178849,
+      "learning_rate": 1.1904880185923295e-05,
+      "loss": 0.5888,
+      "step": 8755
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 19.206713653849533,
+      "learning_rate": 1.190322463864145e-05,
+      "loss": 0.5912,
+      "step": 8756
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.95514795879096,
+      "learning_rate": 1.1901569037233012e-05,
+      "loss": 0.6591,
+      "step": 8757
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.45718708322206,
+      "learning_rate": 1.1899913381745062e-05,
+      "loss": 0.6098,
+      "step": 8758
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.174229752905077,
+      "learning_rate": 1.189825767222469e-05,
+      "loss": 0.6077,
+      "step": 8759
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.924854490395116,
+      "learning_rate": 1.1896601908718979e-05,
+      "loss": 0.555,
+      "step": 8760
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.766709660096474,
+      "learning_rate": 1.1894946091275014e-05,
+      "loss": 0.6823,
+      "step": 8761
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.711140559234362,
+      "learning_rate": 1.1893290219939899e-05,
+      "loss": 0.6426,
+      "step": 8762
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.665283668070913,
+      "learning_rate": 1.1891634294760713e-05,
+      "loss": 0.6003,
+      "step": 8763
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.673144886994365,
+      "learning_rate": 1.1889978315784557e-05,
+      "loss": 0.6776,
+      "step": 8764
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.903327367053109,
+      "learning_rate": 1.1888322283058517e-05,
+      "loss": 0.5077,
+      "step": 8765
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 20.94289715722972,
+      "learning_rate": 1.1886666196629701e-05,
+      "loss": 0.6484,
+      "step": 8766
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.498087574032546,
+      "learning_rate": 1.1885010056545204e-05,
+      "loss": 0.673,
+      "step": 8767
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.40149669127506,
+      "learning_rate": 1.1883353862852121e-05,
+      "loss": 0.6498,
+      "step": 8768
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.29558471312626,
+      "learning_rate": 1.1881697615597554e-05,
+      "loss": 0.5583,
+      "step": 8769
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.698319888606086,
+      "learning_rate": 1.188004131482861e-05,
+      "loss": 0.6484,
+      "step": 8770
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.05487926657686,
+      "learning_rate": 1.187838496059239e-05,
+      "loss": 0.7046,
+      "step": 8771
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.347937588728534,
+      "learning_rate": 1.1876728552936e-05,
+      "loss": 0.6062,
+      "step": 8772
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.477226872362207,
+      "learning_rate": 1.1875072091906547e-05,
+      "loss": 0.5326,
+      "step": 8773
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.41050072044705,
+      "learning_rate": 1.1873415577551146e-05,
+      "loss": 0.6239,
+      "step": 8774
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 29.094114957911238,
+      "learning_rate": 1.1871759009916897e-05,
+      "loss": 0.7097,
+      "step": 8775
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.523528699479314,
+      "learning_rate": 1.1870102389050917e-05,
+      "loss": 0.6356,
+      "step": 8776
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.427232202669828,
+      "learning_rate": 1.186844571500032e-05,
+      "loss": 0.5878,
+      "step": 8777
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 19.328231193694823,
+      "learning_rate": 1.1866788987812219e-05,
+      "loss": 0.5914,
+      "step": 8778
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.47172187406481,
+      "learning_rate": 1.1865132207533731e-05,
+      "loss": 0.6255,
+      "step": 8779
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.157636511901286,
+      "learning_rate": 1.1863475374211974e-05,
+      "loss": 0.5833,
+      "step": 8780
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.718110224006761,
+      "learning_rate": 1.1861818487894072e-05,
+      "loss": 0.6306,
+      "step": 8781
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.783794854776737,
+      "learning_rate": 1.1860161548627137e-05,
+      "loss": 0.592,
+      "step": 8782
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.493314127303073,
+      "learning_rate": 1.1858504556458294e-05,
+      "loss": 0.6051,
+      "step": 8783
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.525089320586517,
+      "learning_rate": 1.1856847511434673e-05,
+      "loss": 0.575,
+      "step": 8784
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.28510709241336,
+      "learning_rate": 1.1855190413603392e-05,
+      "loss": 0.605,
+      "step": 8785
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 23.72661662166227,
+      "learning_rate": 1.1853533263011583e-05,
+      "loss": 0.6353,
+      "step": 8786
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.017655812338674,
+      "learning_rate": 1.185187605970637e-05,
+      "loss": 0.6292,
+      "step": 8787
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.77108743548212,
+      "learning_rate": 1.1850218803734886e-05,
+      "loss": 0.6616,
+      "step": 8788
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 19.064875782206286,
+      "learning_rate": 1.1848561495144263e-05,
+      "loss": 0.6098,
+      "step": 8789
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.372480403586309,
+      "learning_rate": 1.184690413398163e-05,
+      "loss": 0.6271,
+      "step": 8790
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 23.617858075109126,
+      "learning_rate": 1.1845246720294129e-05,
+      "loss": 0.5457,
+      "step": 8791
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.57863554500613,
+      "learning_rate": 1.1843589254128884e-05,
+      "loss": 0.6431,
+      "step": 8792
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.010458838038158,
+      "learning_rate": 1.1841931735533043e-05,
+      "loss": 0.6162,
+      "step": 8793
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 23.40619553481973,
+      "learning_rate": 1.184027416455374e-05,
+      "loss": 0.6197,
+      "step": 8794
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.926978034401095,
+      "learning_rate": 1.1838616541238115e-05,
+      "loss": 0.5773,
+      "step": 8795
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 24.16491994313919,
+      "learning_rate": 1.1836958865633315e-05,
+      "loss": 0.6239,
+      "step": 8796
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.247293061595126,
+      "learning_rate": 1.1835301137786476e-05,
+      "loss": 0.7009,
+      "step": 8797
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.60317748373623,
+      "learning_rate": 1.1833643357744747e-05,
+      "loss": 0.5932,
+      "step": 8798
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.256813926854264,
+      "learning_rate": 1.1831985525555274e-05,
+      "loss": 0.6461,
+      "step": 8799
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.181852749825417,
+      "learning_rate": 1.1830327641265202e-05,
+      "loss": 0.6034,
+      "step": 8800
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.984551365590939,
+      "learning_rate": 1.1828669704921685e-05,
+      "loss": 0.5431,
+      "step": 8801
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 27.292590424626432,
+      "learning_rate": 1.182701171657187e-05,
+      "loss": 0.6692,
+      "step": 8802
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.036929265693516,
+      "learning_rate": 1.1825353676262914e-05,
+      "loss": 0.615,
+      "step": 8803
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 20.25751258357393,
+      "learning_rate": 1.1823695584041963e-05,
+      "loss": 0.613,
+      "step": 8804
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 27.153551680277655,
+      "learning_rate": 1.1822037439956178e-05,
+      "loss": 0.6862,
+      "step": 8805
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.63115168470582,
+      "learning_rate": 1.1820379244052715e-05,
+      "loss": 0.653,
+      "step": 8806
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.23232320913089,
+      "learning_rate": 1.1818720996378729e-05,
+      "loss": 0.5638,
+      "step": 8807
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.930645730054778,
+      "learning_rate": 1.1817062696981384e-05,
+      "loss": 0.6471,
+      "step": 8808
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.33081934192666,
+      "learning_rate": 1.1815404345907837e-05,
+      "loss": 0.6343,
+      "step": 8809
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.683788137352334,
+      "learning_rate": 1.1813745943205254e-05,
+      "loss": 0.5866,
+      "step": 8810
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.684608961215451,
+      "learning_rate": 1.1812087488920798e-05,
+      "loss": 0.5788,
+      "step": 8811
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.180297513575756,
+      "learning_rate": 1.1810428983101632e-05,
+      "loss": 0.6191,
+      "step": 8812
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.243397929638935,
+      "learning_rate": 1.1808770425794927e-05,
+      "loss": 0.5692,
+      "step": 8813
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.72967744177661,
+      "learning_rate": 1.1807111817047846e-05,
+      "loss": 0.6357,
+      "step": 8814
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 20.66109854889212,
+      "learning_rate": 1.1805453156907562e-05,
+      "loss": 0.6164,
+      "step": 8815
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.436350847256517,
+      "learning_rate": 1.1803794445421251e-05,
+      "loss": 0.6708,
+      "step": 8816
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.339018356708436,
+      "learning_rate": 1.1802135682636076e-05,
+      "loss": 0.6642,
+      "step": 8817
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.560486225980682,
+      "learning_rate": 1.1800476868599222e-05,
+      "loss": 0.653,
+      "step": 8818
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.680373201413264,
+      "learning_rate": 1.1798818003357853e-05,
+      "loss": 0.558,
+      "step": 8819
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 22.154495874987695,
+      "learning_rate": 1.1797159086959156e-05,
+      "loss": 0.5709,
+      "step": 8820
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 20.7590174255286,
+      "learning_rate": 1.1795500119450305e-05,
+      "loss": 0.6297,
+      "step": 8821
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.793729425057617,
+      "learning_rate": 1.179384110087848e-05,
+      "loss": 0.602,
+      "step": 8822
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.675922302858655,
+      "learning_rate": 1.1792182031290867e-05,
+      "loss": 0.6355,
+      "step": 8823
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.56581067935375,
+      "learning_rate": 1.1790522910734638e-05,
+      "loss": 0.592,
+      "step": 8824
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.423739930853916,
+      "learning_rate": 1.1788863739256992e-05,
+      "loss": 0.6738,
+      "step": 8825
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.300026753623985,
+      "learning_rate": 1.1787204516905104e-05,
+      "loss": 0.5973,
+      "step": 8826
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.190096682563716,
+      "learning_rate": 1.1785545243726166e-05,
+      "loss": 0.5915,
+      "step": 8827
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.673453136647232,
+      "learning_rate": 1.1783885919767368e-05,
+      "loss": 0.6839,
+      "step": 8828
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.876111351146506,
+      "learning_rate": 1.1782226545075896e-05,
+      "loss": 0.6366,
+      "step": 8829
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.039812517661556,
+      "learning_rate": 1.1780567119698944e-05,
+      "loss": 0.6643,
+      "step": 8830
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 26.43417260015495,
+      "learning_rate": 1.1778907643683704e-05,
+      "loss": 0.5339,
+      "step": 8831
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.071657149789413,
+      "learning_rate": 1.1777248117077371e-05,
+      "loss": 0.59,
+      "step": 8832
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.47647137885765,
+      "learning_rate": 1.1775588539927142e-05,
+      "loss": 0.4877,
+      "step": 8833
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.44229360361136,
+      "learning_rate": 1.1773928912280213e-05,
+      "loss": 0.6178,
+      "step": 8834
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 27.87886765509306,
+      "learning_rate": 1.1772269234183786e-05,
+      "loss": 0.6472,
+      "step": 8835
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 30.0246233759278,
+      "learning_rate": 1.1770609505685056e-05,
+      "loss": 0.6759,
+      "step": 8836
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 22.69169194991702,
+      "learning_rate": 1.1768949726831228e-05,
+      "loss": 0.6553,
+      "step": 8837
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.54049871464825,
+      "learning_rate": 1.1767289897669505e-05,
+      "loss": 0.696,
+      "step": 8838
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 24.74826716608812,
+      "learning_rate": 1.1765630018247089e-05,
+      "loss": 0.6519,
+      "step": 8839
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.132074428380475,
+      "learning_rate": 1.1763970088611192e-05,
+      "loss": 0.5551,
+      "step": 8840
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.130480533561737,
+      "learning_rate": 1.1762310108809017e-05,
+      "loss": 0.6302,
+      "step": 8841
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.78296848692802,
+      "learning_rate": 1.176065007888777e-05,
+      "loss": 0.6442,
+      "step": 8842
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.249799158044773,
+      "learning_rate": 1.1758989998894667e-05,
+      "loss": 0.67,
+      "step": 8843
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.989843346260734,
+      "learning_rate": 1.1757329868876917e-05,
+      "loss": 0.5932,
+      "step": 8844
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.585443668142796,
+      "learning_rate": 1.1755669688881732e-05,
+      "loss": 0.6167,
+      "step": 8845
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.87361622114679,
+      "learning_rate": 1.175400945895633e-05,
+      "loss": 0.5859,
+      "step": 8846
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.710688289680192,
+      "learning_rate": 1.1752349179147926e-05,
+      "loss": 0.6055,
+      "step": 8847
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.391839388802753,
+      "learning_rate": 1.1750688849503735e-05,
+      "loss": 0.6189,
+      "step": 8848
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.569850741263114,
+      "learning_rate": 1.1749028470070975e-05,
+      "loss": 0.6183,
+      "step": 8849
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 29.473755080491895,
+      "learning_rate": 1.1747368040896875e-05,
+      "loss": 0.657,
+      "step": 8850
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.53333045561157,
+      "learning_rate": 1.1745707562028643e-05,
+      "loss": 0.5385,
+      "step": 8851
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 22.16195506698316,
+      "learning_rate": 1.1744047033513514e-05,
+      "loss": 0.6817,
+      "step": 8852
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.38801996331766,
+      "learning_rate": 1.1742386455398704e-05,
+      "loss": 0.5677,
+      "step": 8853
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.618734703035798,
+      "learning_rate": 1.1740725827731446e-05,
+      "loss": 0.6374,
+      "step": 8854
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.860257317008685,
+      "learning_rate": 1.1739065150558961e-05,
+      "loss": 0.606,
+      "step": 8855
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.30032536088489,
+      "learning_rate": 1.1737404423928482e-05,
+      "loss": 0.6202,
+      "step": 8856
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.39155890836942,
+      "learning_rate": 1.1735743647887237e-05,
+      "loss": 0.6448,
+      "step": 8857
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.356398879559249,
+      "learning_rate": 1.1734082822482457e-05,
+      "loss": 0.5771,
+      "step": 8858
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.78085065068071,
+      "learning_rate": 1.1732421947761377e-05,
+      "loss": 0.5897,
+      "step": 8859
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 37.65702955657859,
+      "learning_rate": 1.173076102377123e-05,
+      "loss": 0.714,
+      "step": 8860
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.294290677908684,
+      "learning_rate": 1.1729100050559252e-05,
+      "loss": 0.6297,
+      "step": 8861
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.82242889957776,
+      "learning_rate": 1.1727439028172682e-05,
+      "loss": 0.5688,
+      "step": 8862
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.457085996274138,
+      "learning_rate": 1.1725777956658752e-05,
+      "loss": 0.6314,
+      "step": 8863
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 24.38195712133182,
+      "learning_rate": 1.172411683606471e-05,
+      "loss": 0.7306,
+      "step": 8864
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.683424076152892,
+      "learning_rate": 1.1722455666437793e-05,
+      "loss": 0.643,
+      "step": 8865
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.81138611707172,
+      "learning_rate": 1.1720794447825245e-05,
+      "loss": 0.6803,
+      "step": 8866
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.127828186591692,
+      "learning_rate": 1.171913318027431e-05,
+      "loss": 0.6178,
+      "step": 8867
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 34.470922097308126,
+      "learning_rate": 1.1717471863832231e-05,
+      "loss": 0.683,
+      "step": 8868
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.81591433011889,
+      "learning_rate": 1.1715810498546259e-05,
+      "loss": 0.5453,
+      "step": 8869
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.681676905898172,
+      "learning_rate": 1.1714149084463638e-05,
+      "loss": 0.6379,
+      "step": 8870
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.621112405574397,
+      "learning_rate": 1.1712487621631621e-05,
+      "loss": 0.6087,
+      "step": 8871
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.987682156107702,
+      "learning_rate": 1.1710826110097457e-05,
+      "loss": 0.5933,
+      "step": 8872
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.562717508864173,
+      "learning_rate": 1.17091645499084e-05,
+      "loss": 0.5919,
+      "step": 8873
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.549489103415727,
+      "learning_rate": 1.1707502941111704e-05,
+      "loss": 0.5727,
+      "step": 8874
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.998997583450262,
+      "learning_rate": 1.1705841283754622e-05,
+      "loss": 0.5683,
+      "step": 8875
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.374022289081182,
+      "learning_rate": 1.170417957788441e-05,
+      "loss": 0.6948,
+      "step": 8876
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.439125885229366,
+      "learning_rate": 1.1702517823548332e-05,
+      "loss": 0.6025,
+      "step": 8877
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.614318817235965,
+      "learning_rate": 1.1700856020793639e-05,
+      "loss": 0.5508,
+      "step": 8878
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.54863954143879,
+      "learning_rate": 1.1699194169667598e-05,
+      "loss": 0.5586,
+      "step": 8879
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.1742295062101,
+      "learning_rate": 1.1697532270217466e-05,
+      "loss": 0.6709,
+      "step": 8880
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.145459674676427,
+      "learning_rate": 1.1695870322490512e-05,
+      "loss": 0.5692,
+      "step": 8881
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.531136037419943,
+      "learning_rate": 1.1694208326533997e-05,
+      "loss": 0.6319,
+      "step": 8882
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.172624563730935,
+      "learning_rate": 1.169254628239519e-05,
+      "loss": 0.5992,
+      "step": 8883
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.61402977301959,
+      "learning_rate": 1.1690884190121356e-05,
+      "loss": 0.6144,
+      "step": 8884
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.10082069893346,
+      "learning_rate": 1.1689222049759765e-05,
+      "loss": 0.6038,
+      "step": 8885
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.112785282709446,
+      "learning_rate": 1.1687559861357685e-05,
+      "loss": 0.6546,
+      "step": 8886
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.74565906233532,
+      "learning_rate": 1.1685897624962392e-05,
+      "loss": 0.5352,
+      "step": 8887
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.639195117563393,
+      "learning_rate": 1.1684235340621155e-05,
+      "loss": 0.6038,
+      "step": 8888
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.48490717796426,
+      "learning_rate": 1.1682573008381252e-05,
+      "loss": 0.6278,
+      "step": 8889
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.658684348570834,
+      "learning_rate": 1.1680910628289956e-05,
+      "loss": 0.6125,
+      "step": 8890
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.835226346700075,
+      "learning_rate": 1.1679248200394546e-05,
+      "loss": 0.7187,
+      "step": 8891
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.99453620761519,
+      "learning_rate": 1.1677585724742298e-05,
+      "loss": 0.6022,
+      "step": 8892
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 22.909605502560105,
+      "learning_rate": 1.1675923201380493e-05,
+      "loss": 0.622,
+      "step": 8893
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.935382409610092,
+      "learning_rate": 1.1674260630356415e-05,
+      "loss": 0.5615,
+      "step": 8894
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.553511302307438,
+      "learning_rate": 1.167259801171734e-05,
+      "loss": 0.627,
+      "step": 8895
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.566553538810293,
+      "learning_rate": 1.1670935345510561e-05,
+      "loss": 0.5479,
+      "step": 8896
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.748620073649047,
+      "learning_rate": 1.1669272631783354e-05,
+      "loss": 0.5502,
+      "step": 8897
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.17891408769377,
+      "learning_rate": 1.1667609870583012e-05,
+      "loss": 0.6716,
+      "step": 8898
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.30397120615585,
+      "learning_rate": 1.1665947061956821e-05,
+      "loss": 0.6013,
+      "step": 8899
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 44.10363793773203,
+      "learning_rate": 1.166428420595207e-05,
+      "loss": 0.6198,
+      "step": 8900
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.83430853108769,
+      "learning_rate": 1.166262130261605e-05,
+      "loss": 0.5914,
+      "step": 8901
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.042748495398728,
+      "learning_rate": 1.1660958351996051e-05,
+      "loss": 0.6349,
+      "step": 8902
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.83199374171938,
+      "learning_rate": 1.1659295354139368e-05,
+      "loss": 0.5785,
+      "step": 8903
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.144907496287269,
+      "learning_rate": 1.1657632309093301e-05,
+      "loss": 0.6123,
+      "step": 8904
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.796133662545763,
+      "learning_rate": 1.1655969216905134e-05,
+      "loss": 0.6385,
+      "step": 8905
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.98031227577902,
+      "learning_rate": 1.1654306077622176e-05,
+      "loss": 0.5588,
+      "step": 8906
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.62861061204906,
+      "learning_rate": 1.1652642891291717e-05,
+      "loss": 0.5689,
+      "step": 8907
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.20944149014515,
+      "learning_rate": 1.1650979657961063e-05,
+      "loss": 0.6959,
+      "step": 8908
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.67912927951192,
+      "learning_rate": 1.1649316377677513e-05,
+      "loss": 0.6151,
+      "step": 8909
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.476384363588675,
+      "learning_rate": 1.164765305048837e-05,
+      "loss": 0.5469,
+      "step": 8910
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 23.744040991437682,
+      "learning_rate": 1.1645989676440938e-05,
+      "loss": 0.6541,
+      "step": 8911
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 28.569482728186248,
+      "learning_rate": 1.1644326255582523e-05,
+      "loss": 0.6799,
+      "step": 8912
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 22.658019062560932,
+      "learning_rate": 1.164266278796043e-05,
+      "loss": 0.6158,
+      "step": 8913
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.828817107270327,
+      "learning_rate": 1.1640999273621969e-05,
+      "loss": 0.5203,
+      "step": 8914
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.081379738490785,
+      "learning_rate": 1.1639335712614451e-05,
+      "loss": 0.5795,
+      "step": 8915
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.46435336575244,
+      "learning_rate": 1.1637672104985182e-05,
+      "loss": 0.6023,
+      "step": 8916
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 24.036083264315902,
+      "learning_rate": 1.1636008450781475e-05,
+      "loss": 0.597,
+      "step": 8917
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.408367795334165,
+      "learning_rate": 1.1634344750050648e-05,
+      "loss": 0.6426,
+      "step": 8918
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.13800700053045,
+      "learning_rate": 1.163268100284001e-05,
+      "loss": 0.5833,
+      "step": 8919
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.642906726672628,
+      "learning_rate": 1.1631017209196878e-05,
+      "loss": 0.5988,
+      "step": 8920
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.048421595169614,
+      "learning_rate": 1.1629353369168574e-05,
+      "loss": 0.6942,
+      "step": 8921
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.33704772617256,
+      "learning_rate": 1.162768948280241e-05,
+      "loss": 0.5742,
+      "step": 8922
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.085674109185405,
+      "learning_rate": 1.1626025550145714e-05,
+      "loss": 0.5937,
+      "step": 8923
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.09362559480258,
+      "learning_rate": 1.16243615712458e-05,
+      "loss": 0.7127,
+      "step": 8924
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.298571636574305,
+      "learning_rate": 1.1622697546149992e-05,
+      "loss": 0.5892,
+      "step": 8925
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.749609536590906,
+      "learning_rate": 1.1621033474905617e-05,
+      "loss": 0.6755,
+      "step": 8926
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.381152615711747,
+      "learning_rate": 1.1619369357559998e-05,
+      "loss": 0.533,
+      "step": 8927
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.306365367498675,
+      "learning_rate": 1.161770519416046e-05,
+      "loss": 0.5943,
+      "step": 8928
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.434557268926508,
+      "learning_rate": 1.1616040984754333e-05,
+      "loss": 0.6459,
+      "step": 8929
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.75157432399176,
+      "learning_rate": 1.1614376729388946e-05,
+      "loss": 0.5812,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.785748338378092,
+      "learning_rate": 1.161271242811163e-05,
+      "loss": 0.5867,
+      "step": 8931
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.899215954112893,
+      "learning_rate": 1.1611048080969715e-05,
+      "loss": 0.5809,
+      "step": 8932
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.941007743290774,
+      "learning_rate": 1.1609383688010536e-05,
+      "loss": 0.5935,
+      "step": 8933
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 24.552889559950803,
+      "learning_rate": 1.1607719249281426e-05,
+      "loss": 0.6503,
+      "step": 8934
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.732238530327162,
+      "learning_rate": 1.1606054764829723e-05,
+      "loss": 0.668,
+      "step": 8935
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.736266633633413,
+      "learning_rate": 1.1604390234702758e-05,
+      "loss": 0.5471,
+      "step": 8936
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.658529238310255,
+      "learning_rate": 1.1602725658947873e-05,
+      "loss": 0.5252,
+      "step": 8937
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 37.67316527591635,
+      "learning_rate": 1.1601061037612413e-05,
+      "loss": 0.5912,
+      "step": 8938
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 20.72212938866781,
+      "learning_rate": 1.1599396370743707e-05,
+      "loss": 0.6876,
+      "step": 8939
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.025153776868871,
+      "learning_rate": 1.159773165838911e-05,
+      "loss": 0.5517,
+      "step": 8940
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.624871202166297,
+      "learning_rate": 1.1596066900595954e-05,
+      "loss": 0.6611,
+      "step": 8941
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.767632991852402,
+      "learning_rate": 1.1594402097411595e-05,
+      "loss": 0.6476,
+      "step": 8942
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 24.77659046484016,
+      "learning_rate": 1.159273724888337e-05,
+      "loss": 0.6358,
+      "step": 8943
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 23.103510608716086,
+      "learning_rate": 1.1591072355058629e-05,
+      "loss": 0.6641,
+      "step": 8944
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.655814361064117,
+      "learning_rate": 1.1589407415984721e-05,
+      "loss": 0.6319,
+      "step": 8945
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.430856029348964,
+      "learning_rate": 1.1587742431708997e-05,
+      "loss": 0.5775,
+      "step": 8946
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 20.641534294354134,
+      "learning_rate": 1.1586077402278804e-05,
+      "loss": 0.5636,
+      "step": 8947
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.833329482247713,
+      "learning_rate": 1.1584412327741501e-05,
+      "loss": 0.6544,
+      "step": 8948
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 31.28449221794735,
+      "learning_rate": 1.1582747208144435e-05,
+      "loss": 0.612,
+      "step": 8949
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.982731508595581,
+      "learning_rate": 1.1581082043534968e-05,
+      "loss": 0.6152,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.69315728320189,
+      "learning_rate": 1.157941683396045e-05,
+      "loss": 0.6447,
+      "step": 8951
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.134849385511526,
+      "learning_rate": 1.1577751579468242e-05,
+      "loss": 0.6076,
+      "step": 8952
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 25.03793340984905,
+      "learning_rate": 1.1576086280105702e-05,
+      "loss": 0.6063,
+      "step": 8953
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.249844101619445,
+      "learning_rate": 1.1574420935920192e-05,
+      "loss": 0.6639,
+      "step": 8954
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.4426587226952,
+      "learning_rate": 1.157275554695907e-05,
+      "loss": 0.6062,
+      "step": 8955
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.042862773607467,
+      "learning_rate": 1.1571090113269697e-05,
+      "loss": 0.6121,
+      "step": 8956
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.273812716384683,
+      "learning_rate": 1.1569424634899441e-05,
+      "loss": 0.6205,
+      "step": 8957
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.980340108535618,
+      "learning_rate": 1.156775911189567e-05,
+      "loss": 0.5797,
+      "step": 8958
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.760529708721084,
+      "learning_rate": 1.1566093544305747e-05,
+      "loss": 0.6132,
+      "step": 8959
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 20.046573704293593,
+      "learning_rate": 1.1564427932177037e-05,
+      "loss": 0.6122,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.435986591621763,
+      "learning_rate": 1.1562762275556916e-05,
+      "loss": 0.615,
+      "step": 8961
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 20.714940123573943,
+      "learning_rate": 1.1561096574492745e-05,
+      "loss": 0.5803,
+      "step": 8962
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.705839174480742,
+      "learning_rate": 1.1559430829031905e-05,
+      "loss": 0.5994,
+      "step": 8963
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.009704724260164,
+      "learning_rate": 1.155776503922176e-05,
+      "loss": 0.5652,
+      "step": 8964
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.796618555212422,
+      "learning_rate": 1.1556099205109694e-05,
+      "loss": 0.5592,
+      "step": 8965
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.947415085453642,
+      "learning_rate": 1.155443332674307e-05,
+      "loss": 0.5901,
+      "step": 8966
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.57799767968796,
+      "learning_rate": 1.1552767404169281e-05,
+      "loss": 0.5742,
+      "step": 8967
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.846545519231139,
+      "learning_rate": 1.1551101437435686e-05,
+      "loss": 0.5587,
+      "step": 8968
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 19.203446933385028,
+      "learning_rate": 1.1549435426589678e-05,
+      "loss": 0.6159,
+      "step": 8969
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.861264901860107,
+      "learning_rate": 1.1547769371678632e-05,
+      "loss": 0.6122,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 27.995382434851084,
+      "learning_rate": 1.1546103272749931e-05,
+      "loss": 0.5866,
+      "step": 8971
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.95391270082982,
+      "learning_rate": 1.1544437129850956e-05,
+      "loss": 0.515,
+      "step": 8972
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.54622634895558,
+      "learning_rate": 1.1542770943029096e-05,
+      "loss": 0.5953,
+      "step": 8973
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.626418754952542,
+      "learning_rate": 1.1541104712331728e-05,
+      "loss": 0.6332,
+      "step": 8974
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 20.772675761717576,
+      "learning_rate": 1.1539438437806245e-05,
+      "loss": 0.7193,
+      "step": 8975
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.878241759765308,
+      "learning_rate": 1.1537772119500035e-05,
+      "loss": 0.6371,
+      "step": 8976
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 19.42295789116523,
+      "learning_rate": 1.1536105757460485e-05,
+      "loss": 0.5887,
+      "step": 8977
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.367424480206616,
+      "learning_rate": 1.1534439351734986e-05,
+      "loss": 0.6164,
+      "step": 8978
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.34381620775164,
+      "learning_rate": 1.1532772902370929e-05,
+      "loss": 0.5721,
+      "step": 8979
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.17088806577735,
+      "learning_rate": 1.1531106409415706e-05,
+      "loss": 0.6079,
+      "step": 8980
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 25.690804804924554,
+      "learning_rate": 1.1529439872916712e-05,
+      "loss": 0.6134,
+      "step": 8981
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.90920049372513,
+      "learning_rate": 1.1527773292921347e-05,
+      "loss": 0.5996,
+      "step": 8982
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.57414107107544,
+      "learning_rate": 1.1526106669476999e-05,
+      "loss": 0.6256,
+      "step": 8983
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.237908356120657,
+      "learning_rate": 1.1524440002631073e-05,
+      "loss": 0.6122,
+      "step": 8984
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.84641441723242,
+      "learning_rate": 1.1522773292430963e-05,
+      "loss": 0.6135,
+      "step": 8985
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 22.795206548805538,
+      "learning_rate": 1.1521106538924071e-05,
+      "loss": 0.6405,
+      "step": 8986
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.33282839749868,
+      "learning_rate": 1.15194397421578e-05,
+      "loss": 0.5875,
+      "step": 8987
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.667473735969466,
+      "learning_rate": 1.1517772902179554e-05,
+      "loss": 0.6352,
+      "step": 8988
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.271874362929152,
+      "learning_rate": 1.1516106019036733e-05,
+      "loss": 0.5384,
+      "step": 8989
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.605000335146926,
+      "learning_rate": 1.1514439092776741e-05,
+      "loss": 0.6122,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.496008624354705,
+      "learning_rate": 1.1512772123446988e-05,
+      "loss": 0.5473,
+      "step": 8991
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.405462524102404,
+      "learning_rate": 1.1511105111094884e-05,
+      "loss": 0.6454,
+      "step": 8992
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.47929706894094,
+      "learning_rate": 1.150943805576783e-05,
+      "loss": 0.5362,
+      "step": 8993
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.107262922064121,
+      "learning_rate": 1.1507770957513245e-05,
+      "loss": 0.6412,
+      "step": 8994
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.85739110563627,
+      "learning_rate": 1.1506103816378533e-05,
+      "loss": 0.6541,
+      "step": 8995
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.063767656794987,
+      "learning_rate": 1.1504436632411112e-05,
+      "loss": 0.5636,
+      "step": 8996
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.551905877143591,
+      "learning_rate": 1.150276940565839e-05,
+      "loss": 0.5732,
+      "step": 8997
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.483792454592585,
+      "learning_rate": 1.1501102136167788e-05,
+      "loss": 0.6391,
+      "step": 8998
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.324170060319688,
+      "learning_rate": 1.1499434823986719e-05,
+      "loss": 0.6567,
+      "step": 8999
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.33599461930103,
+      "learning_rate": 1.1497767469162598e-05,
+      "loss": 0.5386,
+      "step": 9000
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.664917872369813,
+      "learning_rate": 1.1496100071742849e-05,
+      "loss": 0.5914,
+      "step": 9001
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 23.243480797696296,
+      "learning_rate": 1.1494432631774888e-05,
+      "loss": 0.5619,
+      "step": 9002
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.05223211151562,
+      "learning_rate": 1.1492765149306137e-05,
+      "loss": 0.5855,
+      "step": 9003
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.276544287648605,
+      "learning_rate": 1.149109762438402e-05,
+      "loss": 0.5184,
+      "step": 9004
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 23.335126113067542,
+      "learning_rate": 1.1489430057055959e-05,
+      "loss": 0.5605,
+      "step": 9005
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 35.99915808737943,
+      "learning_rate": 1.1487762447369375e-05,
+      "loss": 0.6487,
+      "step": 9006
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 25.421618981411598,
+      "learning_rate": 1.1486094795371703e-05,
+      "loss": 0.6516,
+      "step": 9007
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.423986668758584,
+      "learning_rate": 1.1484427101110359e-05,
+      "loss": 0.6055,
+      "step": 9008
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.111500551723328,
+      "learning_rate": 1.148275936463278e-05,
+      "loss": 0.6073,
+      "step": 9009
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.382029610480792,
+      "learning_rate": 1.148109158598639e-05,
+      "loss": 0.6652,
+      "step": 9010
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.41212379704429,
+      "learning_rate": 1.1479423765218625e-05,
+      "loss": 0.6017,
+      "step": 9011
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.243962914789135,
+      "learning_rate": 1.1477755902376908e-05,
+      "loss": 0.5556,
+      "step": 9012
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.75227774273833,
+      "learning_rate": 1.147608799750868e-05,
+      "loss": 0.6113,
+      "step": 9013
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.742528265328865,
+      "learning_rate": 1.1474420050661374e-05,
+      "loss": 0.6155,
+      "step": 9014
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.653549596306117,
+      "learning_rate": 1.1472752061882426e-05,
+      "loss": 0.5651,
+      "step": 9015
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.244776740031117,
+      "learning_rate": 1.1471084031219268e-05,
+      "loss": 0.5875,
+      "step": 9016
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.00375170692068,
+      "learning_rate": 1.146941595871934e-05,
+      "loss": 0.5834,
+      "step": 9017
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.79248906652147,
+      "learning_rate": 1.1467747844430082e-05,
+      "loss": 0.5934,
+      "step": 9018
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.18208711647645,
+      "learning_rate": 1.1466079688398937e-05,
+      "loss": 0.5653,
+      "step": 9019
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.989205817157451,
+      "learning_rate": 1.146441149067334e-05,
+      "loss": 0.6278,
+      "step": 9020
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.380178522762158,
+      "learning_rate": 1.1462743251300739e-05,
+      "loss": 0.7111,
+      "step": 9021
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.215860721788788,
+      "learning_rate": 1.1461074970328574e-05,
+      "loss": 0.5509,
+      "step": 9022
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.994802569519802,
+      "learning_rate": 1.1459406647804295e-05,
+      "loss": 0.6439,
+      "step": 9023
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.916998712009644,
+      "learning_rate": 1.1457738283775339e-05,
+      "loss": 0.5621,
+      "step": 9024
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.976651007592288,
+      "learning_rate": 1.1456069878289161e-05,
+      "loss": 0.6414,
+      "step": 9025
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.180718418326556,
+      "learning_rate": 1.1454401431393212e-05,
+      "loss": 0.5068,
+      "step": 9026
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 28.58602278530663,
+      "learning_rate": 1.145273294313493e-05,
+      "loss": 0.5983,
+      "step": 9027
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 26.57496777821415,
+      "learning_rate": 1.1451064413561776e-05,
+      "loss": 0.6932,
+      "step": 9028
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.214888109323557,
+      "learning_rate": 1.1449395842721201e-05,
+      "loss": 0.5989,
+      "step": 9029
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.505302495435384,
+      "learning_rate": 1.1447727230660654e-05,
+      "loss": 0.6125,
+      "step": 9030
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.371601697784598,
+      "learning_rate": 1.1446058577427592e-05,
+      "loss": 0.5455,
+      "step": 9031
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 31.773109916459667,
+      "learning_rate": 1.1444389883069471e-05,
+      "loss": 0.6116,
+      "step": 9032
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.058310238120455,
+      "learning_rate": 1.1442721147633744e-05,
+      "loss": 0.6278,
+      "step": 9033
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.038650264041706,
+      "learning_rate": 1.1441052371167876e-05,
+      "loss": 0.6426,
+      "step": 9034
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.568615878545533,
+      "learning_rate": 1.1439383553719315e-05,
+      "loss": 0.5955,
+      "step": 9035
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.927676643839913,
+      "learning_rate": 1.1437714695335534e-05,
+      "loss": 0.5605,
+      "step": 9036
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 24.555699644064862,
+      "learning_rate": 1.1436045796063983e-05,
+      "loss": 0.6343,
+      "step": 9037
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.433361387157586,
+      "learning_rate": 1.1434376855952136e-05,
+      "loss": 0.5932,
+      "step": 9038
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.801446360782016,
+      "learning_rate": 1.1432707875047445e-05,
+      "loss": 0.6008,
+      "step": 9039
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.820679051484227,
+      "learning_rate": 1.143103885339738e-05,
+      "loss": 0.6079,
+      "step": 9040
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.998013283692515,
+      "learning_rate": 1.1429369791049409e-05,
+      "loss": 0.6169,
+      "step": 9041
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.860633692269566,
+      "learning_rate": 1.1427700688050998e-05,
+      "loss": 0.5423,
+      "step": 9042
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.42527479279007,
+      "learning_rate": 1.1426031544449614e-05,
+      "loss": 0.5335,
+      "step": 9043
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.489213408677326,
+      "learning_rate": 1.1424362360292725e-05,
+      "loss": 0.5786,
+      "step": 9044
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.797013522957908,
+      "learning_rate": 1.1422693135627804e-05,
+      "loss": 0.5931,
+      "step": 9045
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.216643216115255,
+      "learning_rate": 1.1421023870502324e-05,
+      "loss": 0.5823,
+      "step": 9046
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.452861149415751,
+      "learning_rate": 1.1419354564963756e-05,
+      "loss": 0.6471,
+      "step": 9047
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.712113179015002,
+      "learning_rate": 1.1417685219059576e-05,
+      "loss": 0.6075,
+      "step": 9048
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.09751338432782,
+      "learning_rate": 1.1416015832837258e-05,
+      "loss": 0.601,
+      "step": 9049
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.08564511640922,
+      "learning_rate": 1.1414346406344272e-05,
+      "loss": 0.6791,
+      "step": 9050
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.037675992504557,
+      "learning_rate": 1.1412676939628108e-05,
+      "loss": 0.6227,
+      "step": 9051
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.526436426946656,
+      "learning_rate": 1.1411007432736236e-05,
+      "loss": 0.6669,
+      "step": 9052
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.49169064310133,
+      "learning_rate": 1.1409337885716141e-05,
+      "loss": 0.6462,
+      "step": 9053
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.349309727778657,
+      "learning_rate": 1.1407668298615296e-05,
+      "loss": 0.5469,
+      "step": 9054
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.755873802426294,
+      "learning_rate": 1.1405998671481191e-05,
+      "loss": 0.6872,
+      "step": 9055
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.277805061036986,
+      "learning_rate": 1.1404329004361306e-05,
+      "loss": 0.6122,
+      "step": 9056
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.948319764812553,
+      "learning_rate": 1.1402659297303124e-05,
+      "loss": 0.6444,
+      "step": 9057
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.131996583476063,
+      "learning_rate": 1.1400989550354133e-05,
+      "loss": 0.5869,
+      "step": 9058
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.669039271476105,
+      "learning_rate": 1.1399319763561821e-05,
+      "loss": 0.6279,
+      "step": 9059
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 24.808556724404184,
+      "learning_rate": 1.1397649936973672e-05,
+      "loss": 0.6098,
+      "step": 9060
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.779127564593985,
+      "learning_rate": 1.1395980070637175e-05,
+      "loss": 0.6201,
+      "step": 9061
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.949160058980471,
+      "learning_rate": 1.139431016459982e-05,
+      "loss": 0.6068,
+      "step": 9062
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.81452301011464,
+      "learning_rate": 1.1392640218909103e-05,
+      "loss": 0.5838,
+      "step": 9063
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.88530604372359,
+      "learning_rate": 1.1390970233612513e-05,
+      "loss": 0.5541,
+      "step": 9064
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.12488839846441,
+      "learning_rate": 1.138930020875754e-05,
+      "loss": 0.6342,
+      "step": 9065
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.314105152263146,
+      "learning_rate": 1.1387630144391685e-05,
+      "loss": 0.5846,
+      "step": 9066
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 20.729774355647102,
+      "learning_rate": 1.138596004056244e-05,
+      "loss": 0.6847,
+      "step": 9067
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.456214600661525,
+      "learning_rate": 1.1384289897317302e-05,
+      "loss": 0.5532,
+      "step": 9068
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.074690685609944,
+      "learning_rate": 1.1382619714703767e-05,
+      "loss": 0.5645,
+      "step": 9069
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.098515367923971,
+      "learning_rate": 1.1380949492769339e-05,
+      "loss": 0.5844,
+      "step": 9070
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.28447161701956,
+      "learning_rate": 1.1379279231561514e-05,
+      "loss": 0.6042,
+      "step": 9071
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.510734081140757,
+      "learning_rate": 1.1377608931127792e-05,
+      "loss": 0.6153,
+      "step": 9072
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.839027655715256,
+      "learning_rate": 1.137593859151568e-05,
+      "loss": 0.6751,
+      "step": 9073
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.811467209898392,
+      "learning_rate": 1.137426821277268e-05,
+      "loss": 0.68,
+      "step": 9074
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.72735095959442,
+      "learning_rate": 1.1372597794946298e-05,
+      "loss": 0.6713,
+      "step": 9075
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 20.814761549418616,
+      "learning_rate": 1.1370927338084035e-05,
+      "loss": 0.5791,
+      "step": 9076
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.49466500218632,
+      "learning_rate": 1.1369256842233399e-05,
+      "loss": 0.6257,
+      "step": 9077
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.443216470792667,
+      "learning_rate": 1.1367586307441902e-05,
+      "loss": 0.6048,
+      "step": 9078
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.067267870846962,
+      "learning_rate": 1.136591573375705e-05,
+      "loss": 0.6292,
+      "step": 9079
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.285803065725048,
+      "learning_rate": 1.1364245121226354e-05,
+      "loss": 0.6121,
+      "step": 9080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.894933058358427,
+      "learning_rate": 1.1362574469897322e-05,
+      "loss": 0.6188,
+      "step": 9081
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 29.614628034449026,
+      "learning_rate": 1.136090377981747e-05,
+      "loss": 0.6406,
+      "step": 9082
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.403628824033497,
+      "learning_rate": 1.1359233051034316e-05,
+      "loss": 0.6175,
+      "step": 9083
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 26.84342563830639,
+      "learning_rate": 1.1357562283595363e-05,
+      "loss": 0.6301,
+      "step": 9084
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 23.60368557270249,
+      "learning_rate": 1.1355891477548136e-05,
+      "loss": 0.6775,
+      "step": 9085
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 23.069978213576917,
+      "learning_rate": 1.1354220632940147e-05,
+      "loss": 0.6347,
+      "step": 9086
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.572896261741263,
+      "learning_rate": 1.1352549749818916e-05,
+      "loss": 0.6391,
+      "step": 9087
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.69603869095432,
+      "learning_rate": 1.135087882823196e-05,
+      "loss": 0.5302,
+      "step": 9088
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 25.01557777371735,
+      "learning_rate": 1.1349207868226801e-05,
+      "loss": 0.5949,
+      "step": 9089
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.342854894218238,
+      "learning_rate": 1.1347536869850959e-05,
+      "loss": 0.7235,
+      "step": 9090
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.25456287398254,
+      "learning_rate": 1.134586583315196e-05,
+      "loss": 0.6007,
+      "step": 9091
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.083450642548637,
+      "learning_rate": 1.134419475817732e-05,
+      "loss": 0.638,
+      "step": 9092
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.231451875127494,
+      "learning_rate": 1.1342523644974567e-05,
+      "loss": 0.632,
+      "step": 9093
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.93877235301933,
+      "learning_rate": 1.134085249359123e-05,
+      "loss": 0.6454,
+      "step": 9094
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.033638809775347,
+      "learning_rate": 1.1339181304074833e-05,
+      "loss": 0.647,
+      "step": 9095
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 29.423520161219532,
+      "learning_rate": 1.13375100764729e-05,
+      "loss": 0.6064,
+      "step": 9096
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.54665340897153,
+      "learning_rate": 1.1335838810832966e-05,
+      "loss": 0.7019,
+      "step": 9097
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.72381920385396,
+      "learning_rate": 1.1334167507202553e-05,
+      "loss": 0.5706,
+      "step": 9098
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 38.26470266899626,
+      "learning_rate": 1.1332496165629201e-05,
+      "loss": 0.6931,
+      "step": 9099
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.353207287815692,
+      "learning_rate": 1.1330824786160437e-05,
+      "loss": 0.6489,
+      "step": 9100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 20.343490737370498,
+      "learning_rate": 1.1329153368843793e-05,
+      "loss": 0.6839,
+      "step": 9101
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.03310480837816,
+      "learning_rate": 1.1327481913726806e-05,
+      "loss": 0.6535,
+      "step": 9102
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.648559477832315,
+      "learning_rate": 1.1325810420857009e-05,
+      "loss": 0.617,
+      "step": 9103
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.355797718121442,
+      "learning_rate": 1.1324138890281937e-05,
+      "loss": 0.6064,
+      "step": 9104
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.46330109342206,
+      "learning_rate": 1.1322467322049135e-05,
+      "loss": 0.5981,
+      "step": 9105
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.75941963592894,
+      "learning_rate": 1.1320795716206132e-05,
+      "loss": 0.7807,
+      "step": 9106
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.167837482896456,
+      "learning_rate": 1.1319124072800471e-05,
+      "loss": 0.5619,
+      "step": 9107
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.71152468366489,
+      "learning_rate": 1.1317452391879695e-05,
+      "loss": 0.5719,
+      "step": 9108
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.4066642659776,
+      "learning_rate": 1.1315780673491344e-05,
+      "loss": 0.6372,
+      "step": 9109
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.55319533344844,
+      "learning_rate": 1.131410891768296e-05,
+      "loss": 0.6287,
+      "step": 9110
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 26.391365222701964,
+      "learning_rate": 1.1312437124502086e-05,
+      "loss": 0.6055,
+      "step": 9111
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.06958664821942,
+      "learning_rate": 1.131076529399627e-05,
+      "loss": 0.5371,
+      "step": 9112
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.792279032666656,
+      "learning_rate": 1.1309093426213051e-05,
+      "loss": 0.6333,
+      "step": 9113
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 23.371635136812454,
+      "learning_rate": 1.1307421521199988e-05,
+      "loss": 0.6905,
+      "step": 9114
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.001999132647843,
+      "learning_rate": 1.1305749579004618e-05,
+      "loss": 0.5812,
+      "step": 9115
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.828004745260056,
+      "learning_rate": 1.1304077599674493e-05,
+      "loss": 0.6056,
+      "step": 9116
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.639540626435203,
+      "learning_rate": 1.1302405583257163e-05,
+      "loss": 0.6281,
+      "step": 9117
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.16487097871779,
+      "learning_rate": 1.1300733529800183e-05,
+      "loss": 0.5788,
+      "step": 9118
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 23.968955207577906,
+      "learning_rate": 1.1299061439351102e-05,
+      "loss": 0.6387,
+      "step": 9119
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.807406211577261,
+      "learning_rate": 1.1297389311957472e-05,
+      "loss": 0.5559,
+      "step": 9120
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.176759074569787,
+      "learning_rate": 1.1295717147666848e-05,
+      "loss": 0.6406,
+      "step": 9121
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.275687869233185,
+      "learning_rate": 1.129404494652679e-05,
+      "loss": 0.6397,
+      "step": 9122
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.290152658163905,
+      "learning_rate": 1.1292372708584847e-05,
+      "loss": 0.7365,
+      "step": 9123
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.600511221700568,
+      "learning_rate": 1.1290700433888584e-05,
+      "loss": 0.6836,
+      "step": 9124
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.36856989679862,
+      "learning_rate": 1.1289028122485553e-05,
+      "loss": 0.6109,
+      "step": 9125
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.96650136820152,
+      "learning_rate": 1.1287355774423316e-05,
+      "loss": 0.6098,
+      "step": 9126
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.440399361721234,
+      "learning_rate": 1.1285683389749434e-05,
+      "loss": 0.6564,
+      "step": 9127
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.37273353754732,
+      "learning_rate": 1.128401096851147e-05,
+      "loss": 0.6347,
+      "step": 9128
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.22226318455248,
+      "learning_rate": 1.1282338510756985e-05,
+      "loss": 0.6322,
+      "step": 9129
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.428904535895661,
+      "learning_rate": 1.1280666016533542e-05,
+      "loss": 0.5889,
+      "step": 9130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.379043486184198,
+      "learning_rate": 1.1278993485888704e-05,
+      "loss": 0.6049,
+      "step": 9131
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.240538401824047,
+      "learning_rate": 1.1277320918870043e-05,
+      "loss": 0.617,
+      "step": 9132
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.327043346213088,
+      "learning_rate": 1.1275648315525123e-05,
+      "loss": 0.6593,
+      "step": 9133
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.678445781150614,
+      "learning_rate": 1.127397567590151e-05,
+      "loss": 0.6154,
+      "step": 9134
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.86344744757711,
+      "learning_rate": 1.1272303000046774e-05,
+      "loss": 0.6913,
+      "step": 9135
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.504715011914662,
+      "learning_rate": 1.1270630288008484e-05,
+      "loss": 0.6695,
+      "step": 9136
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.092482862634203,
+      "learning_rate": 1.1268957539834213e-05,
+      "loss": 0.6672,
+      "step": 9137
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.48369333117928,
+      "learning_rate": 1.1267284755571528e-05,
+      "loss": 0.5981,
+      "step": 9138
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.308538836365852,
+      "learning_rate": 1.1265611935268013e-05,
+      "loss": 0.5965,
+      "step": 9139
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.073746797439478,
+      "learning_rate": 1.1263939078971229e-05,
+      "loss": 0.6243,
+      "step": 9140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.810150379953935,
+      "learning_rate": 1.1262266186728763e-05,
+      "loss": 0.6773,
+      "step": 9141
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.019139879984305,
+      "learning_rate": 1.1260593258588179e-05,
+      "loss": 0.5797,
+      "step": 9142
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.68618465276088,
+      "learning_rate": 1.1258920294597063e-05,
+      "loss": 0.5153,
+      "step": 9143
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.780031276852068,
+      "learning_rate": 1.1257247294802988e-05,
+      "loss": 0.5815,
+      "step": 9144
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.872738059414592,
+      "learning_rate": 1.125557425925354e-05,
+      "loss": 0.5487,
+      "step": 9145
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.979703139241888,
+      "learning_rate": 1.125390118799629e-05,
+      "loss": 0.6997,
+      "step": 9146
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.524404827507347,
+      "learning_rate": 1.1252228081078826e-05,
+      "loss": 0.5704,
+      "step": 9147
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.949117590326956,
+      "learning_rate": 1.1250554938548726e-05,
+      "loss": 0.6567,
+      "step": 9148
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.889068687313657,
+      "learning_rate": 1.1248881760453578e-05,
+      "loss": 0.6515,
+      "step": 9149
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.429801151464815,
+      "learning_rate": 1.124720854684096e-05,
+      "loss": 0.6001,
+      "step": 9150
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.819143058741489,
+      "learning_rate": 1.1245535297758463e-05,
+      "loss": 0.5381,
+      "step": 9151
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.710729237204703,
+      "learning_rate": 1.124386201325367e-05,
+      "loss": 0.6662,
+      "step": 9152
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.605071221177864,
+      "learning_rate": 1.124218869337417e-05,
+      "loss": 0.5742,
+      "step": 9153
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.094322992716332,
+      "learning_rate": 1.1240515338167548e-05,
+      "loss": 0.5979,
+      "step": 9154
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.685898919014859,
+      "learning_rate": 1.1238841947681399e-05,
+      "loss": 0.5981,
+      "step": 9155
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.8066106956894,
+      "learning_rate": 1.1237168521963307e-05,
+      "loss": 0.5302,
+      "step": 9156
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.51211495266274,
+      "learning_rate": 1.1235495061060864e-05,
+      "loss": 0.5423,
+      "step": 9157
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.017995690202573,
+      "learning_rate": 1.123382156502167e-05,
+      "loss": 0.5371,
+      "step": 9158
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.274334587612588,
+      "learning_rate": 1.1232148033893307e-05,
+      "loss": 0.4745,
+      "step": 9159
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.28087899645096,
+      "learning_rate": 1.1230474467723376e-05,
+      "loss": 0.594,
+      "step": 9160
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.606996366315286,
+      "learning_rate": 1.1228800866559472e-05,
+      "loss": 0.5989,
+      "step": 9161
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.246749941264525,
+      "learning_rate": 1.1227127230449191e-05,
+      "loss": 0.5773,
+      "step": 9162
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.688941519878224,
+      "learning_rate": 1.1225453559440128e-05,
+      "loss": 0.6094,
+      "step": 9163
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.601185173683334,
+      "learning_rate": 1.1223779853579884e-05,
+      "loss": 0.6012,
+      "step": 9164
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.220288273403998,
+      "learning_rate": 1.1222106112916053e-05,
+      "loss": 0.6064,
+      "step": 9165
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.02275981584391,
+      "learning_rate": 1.1220432337496244e-05,
+      "loss": 0.6672,
+      "step": 9166
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.058959748511075,
+      "learning_rate": 1.121875852736805e-05,
+      "loss": 0.5952,
+      "step": 9167
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.952123799971428,
+      "learning_rate": 1.1217084682579077e-05,
+      "loss": 0.6151,
+      "step": 9168
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.647652856467833,
+      "learning_rate": 1.1215410803176927e-05,
+      "loss": 0.5902,
+      "step": 9169
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.473547970343066,
+      "learning_rate": 1.1213736889209207e-05,
+      "loss": 0.6666,
+      "step": 9170
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 23.354751545277264,
+      "learning_rate": 1.1212062940723519e-05,
+      "loss": 0.5764,
+      "step": 9171
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 26.54377848980917,
+      "learning_rate": 1.1210388957767471e-05,
+      "loss": 0.5804,
+      "step": 9172
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.864543439766816,
+      "learning_rate": 1.1208714940388668e-05,
+      "loss": 0.6083,
+      "step": 9173
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.573375387447467,
+      "learning_rate": 1.1207040888634719e-05,
+      "loss": 0.5952,
+      "step": 9174
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.071871097091563,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.5872,
+      "step": 9175
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.281087903176285,
+      "learning_rate": 1.1203692682191819e-05,
+      "loss": 0.6939,
+      "step": 9176
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.202771754635325,
+      "learning_rate": 1.1202018527598091e-05,
+      "loss": 0.6457,
+      "step": 9177
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.67415285283405,
+      "learning_rate": 1.120034433881966e-05,
+      "loss": 0.59,
+      "step": 9178
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.601946449548517,
+      "learning_rate": 1.1198670115904136e-05,
+      "loss": 0.6297,
+      "step": 9179
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 30.470285551551125,
+      "learning_rate": 1.1196995858899138e-05,
+      "loss": 0.6937,
+      "step": 9180
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.413605388640091,
+      "learning_rate": 1.1195321567852275e-05,
+      "loss": 0.5675,
+      "step": 9181
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.069825486226854,
+      "learning_rate": 1.1193647242811165e-05,
+      "loss": 0.5391,
+      "step": 9182
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.292613018269247,
+      "learning_rate": 1.1191972883823433e-05,
+      "loss": 0.5359,
+      "step": 9183
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.862600239254476,
+      "learning_rate": 1.119029849093668e-05,
+      "loss": 0.647,
+      "step": 9184
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 25.713910930367874,
+      "learning_rate": 1.118862406419854e-05,
+      "loss": 0.5467,
+      "step": 9185
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 23.153794562571782,
+      "learning_rate": 1.1186949603656624e-05,
+      "loss": 0.7108,
+      "step": 9186
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.312513565935227,
+      "learning_rate": 1.1185275109358558e-05,
+      "loss": 0.5972,
+      "step": 9187
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.569038125764617,
+      "learning_rate": 1.118360058135196e-05,
+      "loss": 0.6932,
+      "step": 9188
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 29.725426942939055,
+      "learning_rate": 1.1181926019684454e-05,
+      "loss": 0.6175,
+      "step": 9189
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 15.614473932249652,
+      "learning_rate": 1.1180251424403666e-05,
+      "loss": 0.5703,
+      "step": 9190
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.54772774245434,
+      "learning_rate": 1.1178576795557214e-05,
+      "loss": 0.5395,
+      "step": 9191
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.772009570027226,
+      "learning_rate": 1.117690213319273e-05,
+      "loss": 0.7171,
+      "step": 9192
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.63290593937698,
+      "learning_rate": 1.1175227437357835e-05,
+      "loss": 0.6536,
+      "step": 9193
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.0844609992272,
+      "learning_rate": 1.1173552708100162e-05,
+      "loss": 0.5932,
+      "step": 9194
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 22.895576886320796,
+      "learning_rate": 1.1171877945467333e-05,
+      "loss": 0.6676,
+      "step": 9195
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 22.679667293790565,
+      "learning_rate": 1.1170203149506984e-05,
+      "loss": 0.5905,
+      "step": 9196
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.82699749983028,
+      "learning_rate": 1.1168528320266743e-05,
+      "loss": 0.5795,
+      "step": 9197
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.148849002528173,
+      "learning_rate": 1.1166853457794238e-05,
+      "loss": 0.5694,
+      "step": 9198
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.59946377445537,
+      "learning_rate": 1.1165178562137103e-05,
+      "loss": 0.6562,
+      "step": 9199
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 22.98983943847259,
+      "learning_rate": 1.1163503633342971e-05,
+      "loss": 0.7734,
+      "step": 9200
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.280237031384548,
+      "learning_rate": 1.1161828671459475e-05,
+      "loss": 0.615,
+      "step": 9201
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.66468511156852,
+      "learning_rate": 1.1160153676534256e-05,
+      "loss": 0.5803,
+      "step": 9202
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.658845253320656,
+      "learning_rate": 1.115847864861494e-05,
+      "loss": 0.6408,
+      "step": 9203
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.169203517702094,
+      "learning_rate": 1.1156803587749172e-05,
+      "loss": 0.5945,
+      "step": 9204
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.953103423871607,
+      "learning_rate": 1.1155128493984588e-05,
+      "loss": 0.6681,
+      "step": 9205
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.969191934066426,
+      "learning_rate": 1.1153453367368824e-05,
+      "loss": 0.6249,
+      "step": 9206
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 22.91817980087681,
+      "learning_rate": 1.115177820794952e-05,
+      "loss": 0.5894,
+      "step": 9207
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.48646017948929,
+      "learning_rate": 1.1150103015774318e-05,
+      "loss": 0.6155,
+      "step": 9208
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 23.897741915870306,
+      "learning_rate": 1.1148427790890857e-05,
+      "loss": 0.6374,
+      "step": 9209
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.232398398685255,
+      "learning_rate": 1.1146752533346786e-05,
+      "loss": 0.6724,
+      "step": 9210
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.359026797213556,
+      "learning_rate": 1.1145077243189739e-05,
+      "loss": 0.6039,
+      "step": 9211
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.296470177581037,
+      "learning_rate": 1.114340192046737e-05,
+      "loss": 0.575,
+      "step": 9212
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.605814467960386,
+      "learning_rate": 1.1141726565227313e-05,
+      "loss": 0.7205,
+      "step": 9213
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.400132068023424,
+      "learning_rate": 1.1140051177517224e-05,
+      "loss": 0.6781,
+      "step": 9214
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.485535025884852,
+      "learning_rate": 1.1138375757384747e-05,
+      "loss": 0.5879,
+      "step": 9215
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.18111534724461,
+      "learning_rate": 1.113670030487753e-05,
+      "loss": 0.6068,
+      "step": 9216
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 20.93906984991635,
+      "learning_rate": 1.1135024820043218e-05,
+      "loss": 0.5994,
+      "step": 9217
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.576505744951646,
+      "learning_rate": 1.1133349302929468e-05,
+      "loss": 0.6009,
+      "step": 9218
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 23.1509549950664,
+      "learning_rate": 1.1131673753583922e-05,
+      "loss": 0.5916,
+      "step": 9219
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.72266107176036,
+      "learning_rate": 1.112999817205424e-05,
+      "loss": 0.6482,
+      "step": 9220
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.915140540956552,
+      "learning_rate": 1.1128322558388071e-05,
+      "loss": 0.6096,
+      "step": 9221
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.446700158244614,
+      "learning_rate": 1.1126646912633068e-05,
+      "loss": 0.5952,
+      "step": 9222
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.73237618249963,
+      "learning_rate": 1.1124971234836886e-05,
+      "loss": 0.589,
+      "step": 9223
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.254016859650473,
+      "learning_rate": 1.1123295525047178e-05,
+      "loss": 0.6781,
+      "step": 9224
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.664382352297014,
+      "learning_rate": 1.1121619783311606e-05,
+      "loss": 0.5769,
+      "step": 9225
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.017923541978643,
+      "learning_rate": 1.111994400967782e-05,
+      "loss": 0.6411,
+      "step": 9226
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.551425547660248,
+      "learning_rate": 1.1118268204193485e-05,
+      "loss": 0.6588,
+      "step": 9227
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.759691531412997,
+      "learning_rate": 1.1116592366906254e-05,
+      "loss": 0.5924,
+      "step": 9228
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.396146381919781,
+      "learning_rate": 1.111491649786379e-05,
+      "loss": 0.5966,
+      "step": 9229
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.314983633208765,
+      "learning_rate": 1.1113240597113752e-05,
+      "loss": 0.6199,
+      "step": 9230
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.73033330998373,
+      "learning_rate": 1.1111564664703805e-05,
+      "loss": 0.7141,
+      "step": 9231
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.42948871421432,
+      "learning_rate": 1.1109888700681608e-05,
+      "loss": 0.5544,
+      "step": 9232
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.724222380238306,
+      "learning_rate": 1.1108212705094826e-05,
+      "loss": 0.6953,
+      "step": 9233
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.683155058442054,
+      "learning_rate": 1.1106536677991122e-05,
+      "loss": 0.6156,
+      "step": 9234
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.15696689943428,
+      "learning_rate": 1.1104860619418162e-05,
+      "loss": 0.5368,
+      "step": 9235
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.71613632600167,
+      "learning_rate": 1.1103184529423612e-05,
+      "loss": 0.6409,
+      "step": 9236
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.456197802359966,
+      "learning_rate": 1.1101508408055141e-05,
+      "loss": 0.6639,
+      "step": 9237
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 24.49740585848573,
+      "learning_rate": 1.1099832255360412e-05,
+      "loss": 0.5518,
+      "step": 9238
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.69709484610711,
+      "learning_rate": 1.1098156071387101e-05,
+      "loss": 0.6742,
+      "step": 9239
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.395890251304616,
+      "learning_rate": 1.1096479856182872e-05,
+      "loss": 0.5318,
+      "step": 9240
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.377089384249047,
+      "learning_rate": 1.1094803609795398e-05,
+      "loss": 0.5615,
+      "step": 9241
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 20.295350144498446,
+      "learning_rate": 1.109312733227235e-05,
+      "loss": 0.6399,
+      "step": 9242
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.574306593655788,
+      "learning_rate": 1.10914510236614e-05,
+      "loss": 0.6261,
+      "step": 9243
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.288961762324547,
+      "learning_rate": 1.1089774684010219e-05,
+      "loss": 0.6474,
+      "step": 9244
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.94701582591727,
+      "learning_rate": 1.1088098313366484e-05,
+      "loss": 0.5891,
+      "step": 9245
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.822434631643524,
+      "learning_rate": 1.108642191177787e-05,
+      "loss": 0.573,
+      "step": 9246
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 15.142014519928907,
+      "learning_rate": 1.1084745479292054e-05,
+      "loss": 0.5446,
+      "step": 9247
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.005720786212926,
+      "learning_rate": 1.108306901595671e-05,
+      "loss": 0.6057,
+      "step": 9248
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.07260389531164,
+      "learning_rate": 1.1081392521819518e-05,
+      "loss": 0.6217,
+      "step": 9249
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.374317807588838,
+      "learning_rate": 1.1079715996928156e-05,
+      "loss": 0.672,
+      "step": 9250
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.466625453375757,
+      "learning_rate": 1.1078039441330298e-05,
+      "loss": 0.6079,
+      "step": 9251
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.851873231007477,
+      "learning_rate": 1.1076362855073635e-05,
+      "loss": 0.598,
+      "step": 9252
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 28.064885948167905,
+      "learning_rate": 1.1074686238205837e-05,
+      "loss": 0.6714,
+      "step": 9253
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.867303519064384,
+      "learning_rate": 1.1073009590774595e-05,
+      "loss": 0.5728,
+      "step": 9254
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.335319741740042,
+      "learning_rate": 1.1071332912827584e-05,
+      "loss": 0.6313,
+      "step": 9255
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.180464268866153,
+      "learning_rate": 1.1069656204412496e-05,
+      "loss": 0.5479,
+      "step": 9256
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.80640386217679,
+      "learning_rate": 1.1067979465577005e-05,
+      "loss": 0.617,
+      "step": 9257
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.29597278537783,
+      "learning_rate": 1.106630269636881e-05,
+      "loss": 0.5618,
+      "step": 9258
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 22.1431439002648,
+      "learning_rate": 1.1064625896835585e-05,
+      "loss": 0.5572,
+      "step": 9259
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.42095537813848,
+      "learning_rate": 1.1062949067025024e-05,
+      "loss": 0.6221,
+      "step": 9260
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.78267029483533,
+      "learning_rate": 1.1061272206984815e-05,
+      "loss": 0.6412,
+      "step": 9261
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.15203782380583,
+      "learning_rate": 1.1059595316762644e-05,
+      "loss": 0.5633,
+      "step": 9262
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.537579408795157,
+      "learning_rate": 1.1057918396406197e-05,
+      "loss": 0.5938,
+      "step": 9263
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.498481395676407,
+      "learning_rate": 1.1056241445963177e-05,
+      "loss": 0.6043,
+      "step": 9264
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.583496270867375,
+      "learning_rate": 1.1054564465481263e-05,
+      "loss": 0.6441,
+      "step": 9265
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.38927262129966,
+      "learning_rate": 1.1052887455008157e-05,
+      "loss": 0.5766,
+      "step": 9266
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.729221358477853,
+      "learning_rate": 1.1051210414591544e-05,
+      "loss": 0.5274,
+      "step": 9267
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.39367556733507,
+      "learning_rate": 1.1049533344279122e-05,
+      "loss": 0.599,
+      "step": 9268
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 25.45838002144944,
+      "learning_rate": 1.1047856244118591e-05,
+      "loss": 0.6777,
+      "step": 9269
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.46941887856235,
+      "learning_rate": 1.1046179114157636e-05,
+      "loss": 0.6293,
+      "step": 9270
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.313819413982287,
+      "learning_rate": 1.1044501954443962e-05,
+      "loss": 0.5981,
+      "step": 9271
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.764063283786262,
+      "learning_rate": 1.104282476502526e-05,
+      "loss": 0.4817,
+      "step": 9272
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.95439232245066,
+      "learning_rate": 1.1041147545949233e-05,
+      "loss": 0.5305,
+      "step": 9273
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.06889737895641,
+      "learning_rate": 1.103947029726358e-05,
+      "loss": 0.6047,
+      "step": 9274
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.000177674173145,
+      "learning_rate": 1.1037793019015999e-05,
+      "loss": 0.5771,
+      "step": 9275
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.596997755010054,
+      "learning_rate": 1.1036115711254191e-05,
+      "loss": 0.5705,
+      "step": 9276
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.040014524156327,
+      "learning_rate": 1.103443837402586e-05,
+      "loss": 0.5791,
+      "step": 9277
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 28.342455143566227,
+      "learning_rate": 1.1032761007378708e-05,
+      "loss": 0.6149,
+      "step": 9278
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.76329390046126,
+      "learning_rate": 1.1031083611360433e-05,
+      "loss": 0.5815,
+      "step": 9279
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 21.4693686369783,
+      "learning_rate": 1.1029406186018745e-05,
+      "loss": 0.6036,
+      "step": 9280
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 22.6934949150342,
+      "learning_rate": 1.102772873140135e-05,
+      "loss": 0.6247,
+      "step": 9281
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 23.703464840066538,
+      "learning_rate": 1.1026051247555947e-05,
+      "loss": 0.5954,
+      "step": 9282
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.936623447971115,
+      "learning_rate": 1.1024373734530247e-05,
+      "loss": 0.6855,
+      "step": 9283
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.189429309421666,
+      "learning_rate": 1.102269619237196e-05,
+      "loss": 0.6611,
+      "step": 9284
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.165220812394546,
+      "learning_rate": 1.102101862112879e-05,
+      "loss": 0.6039,
+      "step": 9285
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 22.82262879304276,
+      "learning_rate": 1.1019341020848449e-05,
+      "loss": 0.6157,
+      "step": 9286
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.60149013056682,
+      "learning_rate": 1.1017663391578644e-05,
+      "loss": 0.6286,
+      "step": 9287
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.476861249404273,
+      "learning_rate": 1.1015985733367087e-05,
+      "loss": 0.6079,
+      "step": 9288
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 31.370785933627776,
+      "learning_rate": 1.1014308046261491e-05,
+      "loss": 0.6477,
+      "step": 9289
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.87850610610669,
+      "learning_rate": 1.1012630330309567e-05,
+      "loss": 0.5596,
+      "step": 9290
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.24132793096575,
+      "learning_rate": 1.1010952585559029e-05,
+      "loss": 0.6938,
+      "step": 9291
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.959268132691284,
+      "learning_rate": 1.1009274812057592e-05,
+      "loss": 0.5619,
+      "step": 9292
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 21.89586029783041,
+      "learning_rate": 1.1007597009852968e-05,
+      "loss": 0.6731,
+      "step": 9293
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 21.91133291751002,
+      "learning_rate": 1.1005919178992876e-05,
+      "loss": 0.524,
+      "step": 9294
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.36805678174434,
+      "learning_rate": 1.1004241319525028e-05,
+      "loss": 0.6244,
+      "step": 9295
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.889540809995655,
+      "learning_rate": 1.1002563431497151e-05,
+      "loss": 0.6113,
+      "step": 9296
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.685101366562206,
+      "learning_rate": 1.1000885514956949e-05,
+      "loss": 0.618,
+      "step": 9297
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.93623806920666,
+      "learning_rate": 1.0999207569952155e-05,
+      "loss": 0.7182,
+      "step": 9298
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.81101575060987,
+      "learning_rate": 1.0997529596530477e-05,
+      "loss": 0.6416,
+      "step": 9299
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.326877634694508,
+      "learning_rate": 1.0995851594739642e-05,
+      "loss": 0.6346,
+      "step": 9300
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.503003877467766,
+      "learning_rate": 1.099417356462737e-05,
+      "loss": 0.603,
+      "step": 9301
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.843762097267835,
+      "learning_rate": 1.0992495506241387e-05,
+      "loss": 0.645,
+      "step": 9302
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.308978150167626,
+      "learning_rate": 1.0990817419629409e-05,
+      "loss": 0.6139,
+      "step": 9303
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.311669976135995,
+      "learning_rate": 1.0989139304839165e-05,
+      "loss": 0.6507,
+      "step": 9304
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 24.02842702456249,
+      "learning_rate": 1.0987461161918375e-05,
+      "loss": 0.6218,
+      "step": 9305
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.82349988290152,
+      "learning_rate": 1.0985782990914771e-05,
+      "loss": 0.6529,
+      "step": 9306
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.726658065638848,
+      "learning_rate": 1.098410479187607e-05,
+      "loss": 0.6127,
+      "step": 9307
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.259433790246426,
+      "learning_rate": 1.0982426564850009e-05,
+      "loss": 0.5724,
+      "step": 9308
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.976708911408448,
+      "learning_rate": 1.098074830988431e-05,
+      "loss": 0.5547,
+      "step": 9309
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.170615903933054,
+      "learning_rate": 1.0979070027026703e-05,
+      "loss": 0.5575,
+      "step": 9310
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.485875205866122,
+      "learning_rate": 1.0977391716324916e-05,
+      "loss": 0.5526,
+      "step": 9311
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.027373340606335,
+      "learning_rate": 1.097571337782668e-05,
+      "loss": 0.5961,
+      "step": 9312
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.743595569193275,
+      "learning_rate": 1.0974035011579731e-05,
+      "loss": 0.6316,
+      "step": 9313
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.71870219454767,
+      "learning_rate": 1.097235661763179e-05,
+      "loss": 0.6095,
+      "step": 9314
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.657334626731686,
+      "learning_rate": 1.09706781960306e-05,
+      "loss": 0.584,
+      "step": 9315
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.001788818695566,
+      "learning_rate": 1.0968999746823888e-05,
+      "loss": 0.5891,
+      "step": 9316
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.441643764078986,
+      "learning_rate": 1.0967321270059391e-05,
+      "loss": 0.6439,
+      "step": 9317
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.053219437131116,
+      "learning_rate": 1.0965642765784841e-05,
+      "loss": 0.6044,
+      "step": 9318
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.90568478291575,
+      "learning_rate": 1.0963964234047977e-05,
+      "loss": 0.6336,
+      "step": 9319
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 79.01080314115897,
+      "learning_rate": 1.0962285674896534e-05,
+      "loss": 0.7097,
+      "step": 9320
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.373823724694624,
+      "learning_rate": 1.096060708837825e-05,
+      "loss": 0.6142,
+      "step": 9321
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.400886317917156,
+      "learning_rate": 1.0958928474540857e-05,
+      "loss": 0.6652,
+      "step": 9322
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.61111164975765,
+      "learning_rate": 1.0957249833432108e-05,
+      "loss": 0.6009,
+      "step": 9323
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 29.695094435206546,
+      "learning_rate": 1.0955571165099728e-05,
+      "loss": 0.6241,
+      "step": 9324
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.494431626007302,
+      "learning_rate": 1.0953892469591464e-05,
+      "loss": 0.5911,
+      "step": 9325
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 27.481600748807796,
+      "learning_rate": 1.0952213746955056e-05,
+      "loss": 0.6171,
+      "step": 9326
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 25.74845349832007,
+      "learning_rate": 1.0950534997238246e-05,
+      "loss": 0.5717,
+      "step": 9327
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.79694738805393,
+      "learning_rate": 1.0948856220488779e-05,
+      "loss": 0.652,
+      "step": 9328
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.7749297189614,
+      "learning_rate": 1.0947177416754396e-05,
+      "loss": 0.6805,
+      "step": 9329
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.015635277052905,
+      "learning_rate": 1.0945498586082838e-05,
+      "loss": 0.5753,
+      "step": 9330
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.101593173226394,
+      "learning_rate": 1.0943819728521854e-05,
+      "loss": 0.5702,
+      "step": 9331
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.889657821696332,
+      "learning_rate": 1.0942140844119194e-05,
+      "loss": 0.6042,
+      "step": 9332
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.450217602308165,
+      "learning_rate": 1.0940461932922593e-05,
+      "loss": 0.5951,
+      "step": 9333
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.44554068663033,
+      "learning_rate": 1.0938782994979808e-05,
+      "loss": 0.6493,
+      "step": 9334
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.847142159759,
+      "learning_rate": 1.0937104030338584e-05,
+      "loss": 0.5713,
+      "step": 9335
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.85219721732854,
+      "learning_rate": 1.0935425039046668e-05,
+      "loss": 0.5854,
+      "step": 9336
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.14891029592234,
+      "learning_rate": 1.0933746021151813e-05,
+      "loss": 0.5597,
+      "step": 9337
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.617218879894757,
+      "learning_rate": 1.0932066976701767e-05,
+      "loss": 0.5958,
+      "step": 9338
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.19051853896591,
+      "learning_rate": 1.0930387905744278e-05,
+      "loss": 0.6437,
+      "step": 9339
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 27.79981532287448,
+      "learning_rate": 1.0928708808327107e-05,
+      "loss": 0.7169,
+      "step": 9340
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.750440615270596,
+      "learning_rate": 1.0927029684497995e-05,
+      "loss": 0.5916,
+      "step": 9341
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.773517717452904,
+      "learning_rate": 1.0925350534304706e-05,
+      "loss": 0.596,
+      "step": 9342
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.23432686737092,
+      "learning_rate": 1.0923671357794986e-05,
+      "loss": 0.6124,
+      "step": 9343
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.46536881756696,
+      "learning_rate": 1.0921992155016594e-05,
+      "loss": 0.6399,
+      "step": 9344
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.835094566774238,
+      "learning_rate": 1.0920312926017285e-05,
+      "loss": 0.5395,
+      "step": 9345
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.85248532785278,
+      "learning_rate": 1.0918633670844813e-05,
+      "loss": 0.6304,
+      "step": 9346
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.692073893187192,
+      "learning_rate": 1.0916954389546939e-05,
+      "loss": 0.5905,
+      "step": 9347
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.087433340637027,
+      "learning_rate": 1.0915275082171419e-05,
+      "loss": 0.5698,
+      "step": 9348
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.552102892874942,
+      "learning_rate": 1.0913595748766009e-05,
+      "loss": 0.5512,
+      "step": 9349
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 24.961806478880785,
+      "learning_rate": 1.0911916389378473e-05,
+      "loss": 0.5684,
+      "step": 9350
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 25.915952094357614,
+      "learning_rate": 1.0910237004056563e-05,
+      "loss": 0.6457,
+      "step": 9351
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.217591032081387,
+      "learning_rate": 1.0908557592848048e-05,
+      "loss": 0.6163,
+      "step": 9352
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 23.803704479040587,
+      "learning_rate": 1.0906878155800686e-05,
+      "loss": 0.7503,
+      "step": 9353
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.070667471784837,
+      "learning_rate": 1.0905198692962241e-05,
+      "loss": 0.6108,
+      "step": 9354
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.253869194386121,
+      "learning_rate": 1.0903519204380475e-05,
+      "loss": 0.605,
+      "step": 9355
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.43850178502428,
+      "learning_rate": 1.090183969010315e-05,
+      "loss": 0.5053,
+      "step": 9356
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.94831951734522,
+      "learning_rate": 1.0900160150178034e-05,
+      "loss": 0.6282,
+      "step": 9357
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.79610114732136,
+      "learning_rate": 1.0898480584652887e-05,
+      "loss": 0.6296,
+      "step": 9358
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.24718245874271,
+      "learning_rate": 1.0896800993575482e-05,
+      "loss": 0.6255,
+      "step": 9359
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.759223427063468,
+      "learning_rate": 1.0895121376993576e-05,
+      "loss": 0.5951,
+      "step": 9360
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 27.871080497733203,
+      "learning_rate": 1.0893441734954944e-05,
+      "loss": 0.5584,
+      "step": 9361
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.92422331578606,
+      "learning_rate": 1.0891762067507353e-05,
+      "loss": 0.6447,
+      "step": 9362
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.245075625056977,
+      "learning_rate": 1.089008237469857e-05,
+      "loss": 0.5667,
+      "step": 9363
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.180661107784466,
+      "learning_rate": 1.0888402656576364e-05,
+      "loss": 0.5551,
+      "step": 9364
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.099482853924005,
+      "learning_rate": 1.0886722913188508e-05,
+      "loss": 0.6097,
+      "step": 9365
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.237800511740094,
+      "learning_rate": 1.0885043144582765e-05,
+      "loss": 0.578,
+      "step": 9366
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 24.35403595069551,
+      "learning_rate": 1.0883363350806922e-05,
+      "loss": 0.5912,
+      "step": 9367
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.135482664141634,
+      "learning_rate": 1.0881683531908733e-05,
+      "loss": 0.6131,
+      "step": 9368
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.470869481661918,
+      "learning_rate": 1.0880003687935988e-05,
+      "loss": 0.6347,
+      "step": 9369
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.97363563802261,
+      "learning_rate": 1.0878323818936448e-05,
+      "loss": 0.5712,
+      "step": 9370
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.5706278316438,
+      "learning_rate": 1.0876643924957892e-05,
+      "loss": 0.6928,
+      "step": 9371
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.891631202386465,
+      "learning_rate": 1.0874964006048098e-05,
+      "loss": 0.6189,
+      "step": 9372
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.001800649486196,
+      "learning_rate": 1.087328406225484e-05,
+      "loss": 0.5915,
+      "step": 9373
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 24.739046890599845,
+      "learning_rate": 1.0871604093625894e-05,
+      "loss": 0.5962,
+      "step": 9374
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.372756715849441,
+      "learning_rate": 1.0869924100209032e-05,
+      "loss": 0.5949,
+      "step": 9375
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.676144707632373,
+      "learning_rate": 1.0868244082052047e-05,
+      "loss": 0.6773,
+      "step": 9376
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.00720326423703,
+      "learning_rate": 1.0866564039202701e-05,
+      "loss": 0.7093,
+      "step": 9377
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 24.900872300899252,
+      "learning_rate": 1.0864883971708782e-05,
+      "loss": 0.5787,
+      "step": 9378
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.036677754152258,
+      "learning_rate": 1.0863203879618069e-05,
+      "loss": 0.5634,
+      "step": 9379
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.653135864055255,
+      "learning_rate": 1.0861523762978343e-05,
+      "loss": 0.6148,
+      "step": 9380
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.19453208812757,
+      "learning_rate": 1.0859843621837386e-05,
+      "loss": 0.6133,
+      "step": 9381
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.224947890261403,
+      "learning_rate": 1.085816345624298e-05,
+      "loss": 0.5717,
+      "step": 9382
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.338802059912311,
+      "learning_rate": 1.0856483266242903e-05,
+      "loss": 0.4999,
+      "step": 9383
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.246108394990104,
+      "learning_rate": 1.085480305188495e-05,
+      "loss": 0.5871,
+      "step": 9384
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.480610727703029,
+      "learning_rate": 1.0853122813216891e-05,
+      "loss": 0.5697,
+      "step": 9385
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.552867714016873,
+      "learning_rate": 1.0851442550286524e-05,
+      "loss": 0.5627,
+      "step": 9386
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.694715413826017,
+      "learning_rate": 1.0849762263141625e-05,
+      "loss": 0.6263,
+      "step": 9387
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.15524028019837,
+      "learning_rate": 1.0848081951829985e-05,
+      "loss": 0.5954,
+      "step": 9388
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.591500600347356,
+      "learning_rate": 1.0846401616399392e-05,
+      "loss": 0.6272,
+      "step": 9389
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 24.016059650664904,
+      "learning_rate": 1.0844721256897633e-05,
+      "loss": 0.6562,
+      "step": 9390
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.775444815102258,
+      "learning_rate": 1.0843040873372493e-05,
+      "loss": 0.6982,
+      "step": 9391
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.60872089662641,
+      "learning_rate": 1.0841360465871765e-05,
+      "loss": 0.5756,
+      "step": 9392
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 20.712522712256504,
+      "learning_rate": 1.0839680034443234e-05,
+      "loss": 0.5979,
+      "step": 9393
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.86911710516593,
+      "learning_rate": 1.0837999579134699e-05,
+      "loss": 0.6265,
+      "step": 9394
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.528542272546858,
+      "learning_rate": 1.0836319099993944e-05,
+      "loss": 0.6186,
+      "step": 9395
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 20.391496309029126,
+      "learning_rate": 1.0834638597068763e-05,
+      "loss": 0.5755,
+      "step": 9396
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.280628048508916,
+      "learning_rate": 1.0832958070406949e-05,
+      "loss": 0.548,
+      "step": 9397
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.738299791033334,
+      "learning_rate": 1.0831277520056296e-05,
+      "loss": 0.6854,
+      "step": 9398
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.076698314995113,
+      "learning_rate": 1.0829596946064595e-05,
+      "loss": 0.6017,
+      "step": 9399
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.164490710427323,
+      "learning_rate": 1.0827916348479641e-05,
+      "loss": 0.627,
+      "step": 9400
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.47180817728054,
+      "learning_rate": 1.0826235727349235e-05,
+      "loss": 0.6082,
+      "step": 9401
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.532146106264022,
+      "learning_rate": 1.0824555082721161e-05,
+      "loss": 0.622,
+      "step": 9402
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.87756875146567,
+      "learning_rate": 1.0822874414643229e-05,
+      "loss": 0.6658,
+      "step": 9403
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.87691566185493,
+      "learning_rate": 1.0821193723163228e-05,
+      "loss": 0.6399,
+      "step": 9404
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 317.9635921230517,
+      "learning_rate": 1.0819513008328957e-05,
+      "loss": 0.592,
+      "step": 9405
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.982118411228477,
+      "learning_rate": 1.0817832270188217e-05,
+      "loss": 0.5933,
+      "step": 9406
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 45.05185597761114,
+      "learning_rate": 1.0816151508788804e-05,
+      "loss": 0.6157,
+      "step": 9407
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.376543944123615,
+      "learning_rate": 1.081447072417852e-05,
+      "loss": 0.5731,
+      "step": 9408
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.80013261806668,
+      "learning_rate": 1.0812789916405167e-05,
+      "loss": 0.6376,
+      "step": 9409
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 24.585975999077736,
+      "learning_rate": 1.081110908551654e-05,
+      "loss": 0.5928,
+      "step": 9410
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.844602445573592,
+      "learning_rate": 1.0809428231560451e-05,
+      "loss": 0.6208,
+      "step": 9411
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.132977277481936,
+      "learning_rate": 1.0807747354584692e-05,
+      "loss": 0.5376,
+      "step": 9412
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.942128926670657,
+      "learning_rate": 1.0806066454637074e-05,
+      "loss": 0.6012,
+      "step": 9413
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.963368986364024,
+      "learning_rate": 1.0804385531765394e-05,
+      "loss": 0.6952,
+      "step": 9414
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.20483350455384,
+      "learning_rate": 1.0802704586017463e-05,
+      "loss": 0.5912,
+      "step": 9415
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.641716544747602,
+      "learning_rate": 1.0801023617441082e-05,
+      "loss": 0.5807,
+      "step": 9416
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.490538248553587,
+      "learning_rate": 1.0799342626084057e-05,
+      "loss": 0.632,
+      "step": 9417
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.217425944648927,
+      "learning_rate": 1.0797661611994196e-05,
+      "loss": 0.5509,
+      "step": 9418
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.932601520218793,
+      "learning_rate": 1.0795980575219305e-05,
+      "loss": 0.6256,
+      "step": 9419
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.35474050889783,
+      "learning_rate": 1.0794299515807191e-05,
+      "loss": 0.6549,
+      "step": 9420
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.936730895849756,
+      "learning_rate": 1.0792618433805666e-05,
+      "loss": 0.6483,
+      "step": 9421
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.44849353224946,
+      "learning_rate": 1.0790937329262537e-05,
+      "loss": 0.6675,
+      "step": 9422
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.453121421799935,
+      "learning_rate": 1.0789256202225611e-05,
+      "loss": 0.6725,
+      "step": 9423
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.045690951106884,
+      "learning_rate": 1.07875750527427e-05,
+      "loss": 0.6397,
+      "step": 9424
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 25.129528171380745,
+      "learning_rate": 1.0785893880861616e-05,
+      "loss": 0.6001,
+      "step": 9425
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.170946033509956,
+      "learning_rate": 1.0784212686630171e-05,
+      "loss": 0.5264,
+      "step": 9426
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.591933308486732,
+      "learning_rate": 1.0782531470096171e-05,
+      "loss": 0.6915,
+      "step": 9427
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.98558754210368,
+      "learning_rate": 1.078085023130744e-05,
+      "loss": 0.6217,
+      "step": 9428
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.663204068755721,
+      "learning_rate": 1.0779168970311782e-05,
+      "loss": 0.5677,
+      "step": 9429
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.071986070294194,
+      "learning_rate": 1.0777487687157018e-05,
+      "loss": 0.5633,
+      "step": 9430
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.05711221520912,
+      "learning_rate": 1.0775806381890951e-05,
+      "loss": 0.6033,
+      "step": 9431
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.609692545754523,
+      "learning_rate": 1.0774125054561411e-05,
+      "loss": 0.6702,
+      "step": 9432
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.828614111617625,
+      "learning_rate": 1.0772443705216206e-05,
+      "loss": 0.6436,
+      "step": 9433
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.221197034597598,
+      "learning_rate": 1.0770762333903154e-05,
+      "loss": 0.5892,
+      "step": 9434
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.655164095824736,
+      "learning_rate": 1.0769080940670072e-05,
+      "loss": 0.6246,
+      "step": 9435
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.871111394935483,
+      "learning_rate": 1.0767399525564777e-05,
+      "loss": 0.5908,
+      "step": 9436
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.680634806307495,
+      "learning_rate": 1.0765718088635086e-05,
+      "loss": 0.6012,
+      "step": 9437
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.00025038638523,
+      "learning_rate": 1.0764036629928827e-05,
+      "loss": 0.57,
+      "step": 9438
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.24499639660738,
+      "learning_rate": 1.0762355149493808e-05,
+      "loss": 0.5749,
+      "step": 9439
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 25.77828202936811,
+      "learning_rate": 1.0760673647377855e-05,
+      "loss": 0.6088,
+      "step": 9440
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 23.354121597287953,
+      "learning_rate": 1.0758992123628792e-05,
+      "loss": 0.7406,
+      "step": 9441
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 41.03289745928189,
+      "learning_rate": 1.0757310578294433e-05,
+      "loss": 0.7082,
+      "step": 9442
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.45371777845439,
+      "learning_rate": 1.0755629011422605e-05,
+      "loss": 0.5759,
+      "step": 9443
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.571716547566247,
+      "learning_rate": 1.0753947423061131e-05,
+      "loss": 0.6467,
+      "step": 9444
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.299547865296514,
+      "learning_rate": 1.0752265813257833e-05,
+      "loss": 0.6574,
+      "step": 9445
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.86579688039223,
+      "learning_rate": 1.0750584182060535e-05,
+      "loss": 0.6207,
+      "step": 9446
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.80949575604331,
+      "learning_rate": 1.0748902529517065e-05,
+      "loss": 0.5757,
+      "step": 9447
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.01128324591708,
+      "learning_rate": 1.074722085567524e-05,
+      "loss": 0.5821,
+      "step": 9448
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.561519825384572,
+      "learning_rate": 1.0745539160582894e-05,
+      "loss": 0.5665,
+      "step": 9449
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.751996929248858,
+      "learning_rate": 1.074385744428785e-05,
+      "loss": 0.5297,
+      "step": 9450
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.69825900169725,
+      "learning_rate": 1.0742175706837936e-05,
+      "loss": 0.6072,
+      "step": 9451
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.325386852186696,
+      "learning_rate": 1.0740493948280976e-05,
+      "loss": 0.6216,
+      "step": 9452
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 30.758691158442254,
+      "learning_rate": 1.0738812168664804e-05,
+      "loss": 0.7611,
+      "step": 9453
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.43667945880691,
+      "learning_rate": 1.0737130368037245e-05,
+      "loss": 0.5615,
+      "step": 9454
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.917657728250465,
+      "learning_rate": 1.0735448546446132e-05,
+      "loss": 0.5749,
+      "step": 9455
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.572068139474975,
+      "learning_rate": 1.073376670393929e-05,
+      "loss": 0.6511,
+      "step": 9456
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.302822306266066,
+      "learning_rate": 1.0732084840564556e-05,
+      "loss": 0.6853,
+      "step": 9457
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.795132239927618,
+      "learning_rate": 1.0730402956369753e-05,
+      "loss": 0.5679,
+      "step": 9458
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.018679810606356,
+      "learning_rate": 1.0728721051402718e-05,
+      "loss": 0.568,
+      "step": 9459
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.406772044107694,
+      "learning_rate": 1.0727039125711283e-05,
+      "loss": 0.5909,
+      "step": 9460
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 25.010215222578704,
+      "learning_rate": 1.0725357179343282e-05,
+      "loss": 0.5392,
+      "step": 9461
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.532561240117714,
+      "learning_rate": 1.0723675212346546e-05,
+      "loss": 0.6206,
+      "step": 9462
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.360520181037266,
+      "learning_rate": 1.0721993224768908e-05,
+      "loss": 0.5557,
+      "step": 9463
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 22.81751436097232,
+      "learning_rate": 1.0720311216658205e-05,
+      "loss": 0.6083,
+      "step": 9464
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.101448116207504,
+      "learning_rate": 1.0718629188062275e-05,
+      "loss": 0.5691,
+      "step": 9465
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.293912554498625,
+      "learning_rate": 1.0716947139028953e-05,
+      "loss": 0.563,
+      "step": 9466
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 22.40208876114644,
+      "learning_rate": 1.071526506960607e-05,
+      "loss": 0.5423,
+      "step": 9467
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.219565201540604,
+      "learning_rate": 1.071358297984147e-05,
+      "loss": 0.6123,
+      "step": 9468
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 22.08994336835152,
+      "learning_rate": 1.0711900869782983e-05,
+      "loss": 0.6473,
+      "step": 9469
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.957967786928343,
+      "learning_rate": 1.0710218739478457e-05,
+      "loss": 0.5852,
+      "step": 9470
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.091591723560867,
+      "learning_rate": 1.0708536588975721e-05,
+      "loss": 0.5488,
+      "step": 9471
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.706371141067113,
+      "learning_rate": 1.0706854418322625e-05,
+      "loss": 0.57,
+      "step": 9472
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.67907114768848,
+      "learning_rate": 1.0705172227566996e-05,
+      "loss": 0.5924,
+      "step": 9473
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 25.80960656458376,
+      "learning_rate": 1.0703490016756688e-05,
+      "loss": 0.6129,
+      "step": 9474
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 27.17041024948368,
+      "learning_rate": 1.070180778593953e-05,
+      "loss": 0.7127,
+      "step": 9475
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.205340816419813,
+      "learning_rate": 1.0700125535163371e-05,
+      "loss": 0.65,
+      "step": 9476
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.982293288930224,
+      "learning_rate": 1.0698443264476051e-05,
+      "loss": 0.5533,
+      "step": 9477
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.031886832876715,
+      "learning_rate": 1.0696760973925413e-05,
+      "loss": 0.5697,
+      "step": 9478
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.236372267085525,
+      "learning_rate": 1.0695078663559302e-05,
+      "loss": 0.6043,
+      "step": 9479
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.486247462818596,
+      "learning_rate": 1.0693396333425559e-05,
+      "loss": 0.6826,
+      "step": 9480
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.5916181887827,
+      "learning_rate": 1.0691713983572028e-05,
+      "loss": 0.5957,
+      "step": 9481
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.68124914541815,
+      "learning_rate": 1.069003161404656e-05,
+      "loss": 0.6007,
+      "step": 9482
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.45992795001345,
+      "learning_rate": 1.0688349224896991e-05,
+      "loss": 0.6118,
+      "step": 9483
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.61345398813733,
+      "learning_rate": 1.0686666816171179e-05,
+      "loss": 0.6545,
+      "step": 9484
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.089503070496587,
+      "learning_rate": 1.0684984387916961e-05,
+      "loss": 0.5167,
+      "step": 9485
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 24.119442864267423,
+      "learning_rate": 1.068330194018219e-05,
+      "loss": 0.6023,
+      "step": 9486
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.792477187993107,
+      "learning_rate": 1.068161947301471e-05,
+      "loss": 0.6376,
+      "step": 9487
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 30.869152872577324,
+      "learning_rate": 1.067993698646237e-05,
+      "loss": 0.6115,
+      "step": 9488
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.931252073446462,
+      "learning_rate": 1.0678254480573023e-05,
+      "loss": 0.6299,
+      "step": 9489
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.67480474037045,
+      "learning_rate": 1.067657195539451e-05,
+      "loss": 0.5673,
+      "step": 9490
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.791128503380907,
+      "learning_rate": 1.0674889410974689e-05,
+      "loss": 0.6741,
+      "step": 9491
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.574516935055764,
+      "learning_rate": 1.0673206847361407e-05,
+      "loss": 0.5793,
+      "step": 9492
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.44956903891582,
+      "learning_rate": 1.067152426460252e-05,
+      "loss": 0.6147,
+      "step": 9493
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 25.631647409052487,
+      "learning_rate": 1.0669841662745874e-05,
+      "loss": 0.5661,
+      "step": 9494
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.798459480931673,
+      "learning_rate": 1.0668159041839321e-05,
+      "loss": 0.5407,
+      "step": 9495
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.97414835428551,
+      "learning_rate": 1.066647640193072e-05,
+      "loss": 0.6451,
+      "step": 9496
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 28.361269138596125,
+      "learning_rate": 1.0664793743067916e-05,
+      "loss": 0.6634,
+      "step": 9497
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.077724281735431,
+      "learning_rate": 1.0663111065298766e-05,
+      "loss": 0.5143,
+      "step": 9498
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.987191370713397,
+      "learning_rate": 1.066142836867113e-05,
+      "loss": 0.6361,
+      "step": 9499
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.609474364876062,
+      "learning_rate": 1.0659745653232851e-05,
+      "loss": 0.546,
+      "step": 9500
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.46218057470345,
+      "learning_rate": 1.06580629190318e-05,
+      "loss": 0.5314,
+      "step": 9501
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.539981436084707,
+      "learning_rate": 1.0656380166115818e-05,
+      "loss": 0.5221,
+      "step": 9502
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.82093256580783,
+      "learning_rate": 1.065469739453277e-05,
+      "loss": 0.5585,
+      "step": 9503
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 24.485869001173825,
+      "learning_rate": 1.0653014604330511e-05,
+      "loss": 0.6606,
+      "step": 9504
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.91504538604698,
+      "learning_rate": 1.06513317955569e-05,
+      "loss": 0.6476,
+      "step": 9505
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.354664022589944,
+      "learning_rate": 1.0649648968259793e-05,
+      "loss": 0.6103,
+      "step": 9506
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 19.770386579560626,
+      "learning_rate": 1.0647966122487048e-05,
+      "loss": 0.5838,
+      "step": 9507
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.284554726570335,
+      "learning_rate": 1.0646283258286524e-05,
+      "loss": 0.6077,
+      "step": 9508
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 25.092795261804845,
+      "learning_rate": 1.0644600375706087e-05,
+      "loss": 0.5511,
+      "step": 9509
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.014322802909845,
+      "learning_rate": 1.0642917474793591e-05,
+      "loss": 0.6354,
+      "step": 9510
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.469612704215882,
+      "learning_rate": 1.0641234555596898e-05,
+      "loss": 0.5781,
+      "step": 9511
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.41696005460803,
+      "learning_rate": 1.0639551618163869e-05,
+      "loss": 0.5833,
+      "step": 9512
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.243668227611723,
+      "learning_rate": 1.0637868662542364e-05,
+      "loss": 0.5534,
+      "step": 9513
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 26.532094638430074,
+      "learning_rate": 1.0636185688780253e-05,
+      "loss": 0.5653,
+      "step": 9514
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.319982775189816,
+      "learning_rate": 1.0634502696925387e-05,
+      "loss": 0.5705,
+      "step": 9515
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.95894349296172,
+      "learning_rate": 1.063281968702564e-05,
+      "loss": 0.6782,
+      "step": 9516
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.063393052919158,
+      "learning_rate": 1.0631136659128867e-05,
+      "loss": 0.5694,
+      "step": 9517
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.209102662146336,
+      "learning_rate": 1.062945361328294e-05,
+      "loss": 0.6827,
+      "step": 9518
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.475584376052943,
+      "learning_rate": 1.0627770549535716e-05,
+      "loss": 0.5483,
+      "step": 9519
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 23.914881369756916,
+      "learning_rate": 1.0626087467935069e-05,
+      "loss": 0.6358,
+      "step": 9520
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.735009675664795,
+      "learning_rate": 1.062440436852886e-05,
+      "loss": 0.5627,
+      "step": 9521
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.783430771235727,
+      "learning_rate": 1.0622721251364955e-05,
+      "loss": 0.6351,
+      "step": 9522
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.958779100398177,
+      "learning_rate": 1.0621038116491224e-05,
+      "loss": 0.5411,
+      "step": 9523
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.768034254116973,
+      "learning_rate": 1.061935496395553e-05,
+      "loss": 0.5827,
+      "step": 9524
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.388536022047735,
+      "learning_rate": 1.0617671793805743e-05,
+      "loss": 0.6838,
+      "step": 9525
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.26562426036298,
+      "learning_rate": 1.0615988606089733e-05,
+      "loss": 0.5505,
+      "step": 9526
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.565596474686416,
+      "learning_rate": 1.0614305400855367e-05,
+      "loss": 0.7014,
+      "step": 9527
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.974896961204593,
+      "learning_rate": 1.0612622178150515e-05,
+      "loss": 0.6463,
+      "step": 9528
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.238998235587502,
+      "learning_rate": 1.0610938938023047e-05,
+      "loss": 0.6253,
+      "step": 9529
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.923957649410756,
+      "learning_rate": 1.060925568052083e-05,
+      "loss": 0.6337,
+      "step": 9530
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.41670397875525,
+      "learning_rate": 1.0607572405691741e-05,
+      "loss": 0.6089,
+      "step": 9531
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.801162031710726,
+      "learning_rate": 1.0605889113583647e-05,
+      "loss": 0.5919,
+      "step": 9532
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 19.469134183065833,
+      "learning_rate": 1.0604205804244424e-05,
+      "loss": 0.5946,
+      "step": 9533
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.567008141113567,
+      "learning_rate": 1.0602522477721938e-05,
+      "loss": 0.6264,
+      "step": 9534
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.87978994529843,
+      "learning_rate": 1.0600839134064065e-05,
+      "loss": 0.7207,
+      "step": 9535
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 19.125638520654952,
+      "learning_rate": 1.0599155773318679e-05,
+      "loss": 0.5884,
+      "step": 9536
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 26.762880063197887,
+      "learning_rate": 1.0597472395533654e-05,
+      "loss": 0.6003,
+      "step": 9537
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.77436759191914,
+      "learning_rate": 1.0595789000756864e-05,
+      "loss": 0.6117,
+      "step": 9538
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.16403719104961,
+      "learning_rate": 1.0594105589036182e-05,
+      "loss": 0.4892,
+      "step": 9539
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.848443429109874,
+      "learning_rate": 1.0592422160419484e-05,
+      "loss": 0.6423,
+      "step": 9540
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 24.226938673467696,
+      "learning_rate": 1.0590738714954652e-05,
+      "loss": 0.6831,
+      "step": 9541
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.35935245390786,
+      "learning_rate": 1.0589055252689549e-05,
+      "loss": 0.5699,
+      "step": 9542
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.08264945118136,
+      "learning_rate": 1.0587371773672064e-05,
+      "loss": 0.6028,
+      "step": 9543
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 29.98770900765187,
+      "learning_rate": 1.0585688277950065e-05,
+      "loss": 0.7055,
+      "step": 9544
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.587539367293392,
+      "learning_rate": 1.058400476557144e-05,
+      "loss": 0.6007,
+      "step": 9545
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.978174737292955,
+      "learning_rate": 1.0582321236584057e-05,
+      "loss": 0.6042,
+      "step": 9546
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 27.281801388598783,
+      "learning_rate": 1.0580637691035799e-05,
+      "loss": 0.6295,
+      "step": 9547
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.556255756318215,
+      "learning_rate": 1.0578954128974546e-05,
+      "loss": 0.6692,
+      "step": 9548
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.6087337362928,
+      "learning_rate": 1.0577270550448175e-05,
+      "loss": 0.615,
+      "step": 9549
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.79501263445483,
+      "learning_rate": 1.0575586955504568e-05,
+      "loss": 0.6839,
+      "step": 9550
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.24112870045532,
+      "learning_rate": 1.0573903344191603e-05,
+      "loss": 0.6651,
+      "step": 9551
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.81415157755609,
+      "learning_rate": 1.0572219716557163e-05,
+      "loss": 0.6202,
+      "step": 9552
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.39179908597005,
+      "learning_rate": 1.0570536072649132e-05,
+      "loss": 0.5771,
+      "step": 9553
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.791777579774434,
+      "learning_rate": 1.0568852412515388e-05,
+      "loss": 0.5681,
+      "step": 9554
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.152439581947128,
+      "learning_rate": 1.0567168736203811e-05,
+      "loss": 0.6505,
+      "step": 9555
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.366736823599535,
+      "learning_rate": 1.0565485043762289e-05,
+      "loss": 0.6095,
+      "step": 9556
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.203889834130816,
+      "learning_rate": 1.0563801335238701e-05,
+      "loss": 0.6203,
+      "step": 9557
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.4492931060352,
+      "learning_rate": 1.0562117610680938e-05,
+      "loss": 0.584,
+      "step": 9558
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 24.79157263366647,
+      "learning_rate": 1.0560433870136872e-05,
+      "loss": 0.7251,
+      "step": 9559
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.480030784467429,
+      "learning_rate": 1.0558750113654404e-05,
+      "loss": 0.5692,
+      "step": 9560
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.889954727696995,
+      "learning_rate": 1.0557066341281402e-05,
+      "loss": 0.5767,
+      "step": 9561
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.438089618415116,
+      "learning_rate": 1.055538255306576e-05,
+      "loss": 0.6005,
+      "step": 9562
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.30729982622911,
+      "learning_rate": 1.0553698749055365e-05,
+      "loss": 0.5957,
+      "step": 9563
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 24.809103160319772,
+      "learning_rate": 1.05520149292981e-05,
+      "loss": 0.5624,
+      "step": 9564
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.908112386932967,
+      "learning_rate": 1.0550331093841855e-05,
+      "loss": 0.5825,
+      "step": 9565
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.515388137377007,
+      "learning_rate": 1.0548647242734516e-05,
+      "loss": 0.5655,
+      "step": 9566
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.679372729467417,
+      "learning_rate": 1.054696337602397e-05,
+      "loss": 0.6545,
+      "step": 9567
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.1263260896761,
+      "learning_rate": 1.0545279493758103e-05,
+      "loss": 0.56,
+      "step": 9568
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 24.053409338032864,
+      "learning_rate": 1.0543595595984806e-05,
+      "loss": 0.6447,
+      "step": 9569
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.056430020817814,
+      "learning_rate": 1.0541911682751971e-05,
+      "loss": 0.6836,
+      "step": 9570
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.916570906419091,
+      "learning_rate": 1.0540227754107483e-05,
+      "loss": 0.6045,
+      "step": 9571
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.304461112504757,
+      "learning_rate": 1.0538543810099236e-05,
+      "loss": 0.6043,
+      "step": 9572
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.328828931096027,
+      "learning_rate": 1.0536859850775117e-05,
+      "loss": 0.66,
+      "step": 9573
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.321388573598336,
+      "learning_rate": 1.0535175876183017e-05,
+      "loss": 0.6903,
+      "step": 9574
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.06154264029838,
+      "learning_rate": 1.0533491886370828e-05,
+      "loss": 0.594,
+      "step": 9575
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 28.40824175399614,
+      "learning_rate": 1.0531807881386442e-05,
+      "loss": 0.6531,
+      "step": 9576
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.491114377761125,
+      "learning_rate": 1.0530123861277752e-05,
+      "loss": 0.5708,
+      "step": 9577
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 20.745317276142934,
+      "learning_rate": 1.0528439826092649e-05,
+      "loss": 0.6742,
+      "step": 9578
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 26.819592832648894,
+      "learning_rate": 1.0526755775879024e-05,
+      "loss": 0.6408,
+      "step": 9579
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.210495938641355,
+      "learning_rate": 1.0525071710684777e-05,
+      "loss": 0.6574,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.755272966114962,
+      "learning_rate": 1.0523387630557797e-05,
+      "loss": 0.649,
+      "step": 9581
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.359098214031285,
+      "learning_rate": 1.0521703535545977e-05,
+      "loss": 0.5388,
+      "step": 9582
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.95129733191838,
+      "learning_rate": 1.0520019425697217e-05,
+      "loss": 0.5968,
+      "step": 9583
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.611221677474138,
+      "learning_rate": 1.0518335301059404e-05,
+      "loss": 0.6536,
+      "step": 9584
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 32.14734087634544,
+      "learning_rate": 1.0516651161680443e-05,
+      "loss": 0.6968,
+      "step": 9585
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.636278067544522,
+      "learning_rate": 1.0514967007608221e-05,
+      "loss": 0.5266,
+      "step": 9586
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.785277131063943,
+      "learning_rate": 1.0513282838890642e-05,
+      "loss": 0.5664,
+      "step": 9587
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.669899890082892,
+      "learning_rate": 1.0511598655575594e-05,
+      "loss": 0.537,
+      "step": 9588
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.3401458567208,
+      "learning_rate": 1.0509914457710986e-05,
+      "loss": 0.6169,
+      "step": 9589
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.943686247889197,
+      "learning_rate": 1.0508230245344707e-05,
+      "loss": 0.6526,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.400058174393402,
+      "learning_rate": 1.0506546018524656e-05,
+      "loss": 0.5398,
+      "step": 9591
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.38221197724052,
+      "learning_rate": 1.0504861777298732e-05,
+      "loss": 0.6262,
+      "step": 9592
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.8586760242244,
+      "learning_rate": 1.0503177521714836e-05,
+      "loss": 0.614,
+      "step": 9593
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.04308591355548,
+      "learning_rate": 1.0501493251820864e-05,
+      "loss": 0.6216,
+      "step": 9594
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.803606390929883,
+      "learning_rate": 1.0499808967664717e-05,
+      "loss": 0.5423,
+      "step": 9595
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 26.97331554537966,
+      "learning_rate": 1.0498124669294296e-05,
+      "loss": 0.6118,
+      "step": 9596
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 20.264583169691303,
+      "learning_rate": 1.0496440356757502e-05,
+      "loss": 0.557,
+      "step": 9597
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.91924624038395,
+      "learning_rate": 1.0494756030102236e-05,
+      "loss": 0.5841,
+      "step": 9598
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.78175202927894,
+      "learning_rate": 1.0493071689376396e-05,
+      "loss": 0.6376,
+      "step": 9599
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.143372913772797,
+      "learning_rate": 1.0491387334627884e-05,
+      "loss": 0.5542,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.07029006864258,
+      "learning_rate": 1.0489702965904605e-05,
+      "loss": 0.6338,
+      "step": 9601
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.01711013149036,
+      "learning_rate": 1.0488018583254463e-05,
+      "loss": 0.6842,
+      "step": 9602
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.069095836699088,
+      "learning_rate": 1.0486334186725353e-05,
+      "loss": 0.6613,
+      "step": 9603
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.043266099041606,
+      "learning_rate": 1.0484649776365189e-05,
+      "loss": 0.6321,
+      "step": 9604
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 23.33969613639088,
+      "learning_rate": 1.0482965352221863e-05,
+      "loss": 0.552,
+      "step": 9605
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.475998266671306,
+      "learning_rate": 1.0481280914343289e-05,
+      "loss": 0.5734,
+      "step": 9606
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 23.847528959992133,
+      "learning_rate": 1.0479596462777363e-05,
+      "loss": 0.5331,
+      "step": 9607
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.475077926024703,
+      "learning_rate": 1.0477911997571998e-05,
+      "loss": 0.6345,
+      "step": 9608
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.699984906394286,
+      "learning_rate": 1.0476227518775095e-05,
+      "loss": 0.5554,
+      "step": 9609
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.42390074512959,
+      "learning_rate": 1.0474543026434559e-05,
+      "loss": 0.6587,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.81933469467535,
+      "learning_rate": 1.0472858520598295e-05,
+      "loss": 0.638,
+      "step": 9611
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.350814940764952,
+      "learning_rate": 1.0471174001314215e-05,
+      "loss": 0.5785,
+      "step": 9612
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 28.812973572982727,
+      "learning_rate": 1.0469489468630218e-05,
+      "loss": 0.6487,
+      "step": 9613
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.2756985602967,
+      "learning_rate": 1.0467804922594218e-05,
+      "loss": 0.5419,
+      "step": 9614
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.91706930602366,
+      "learning_rate": 1.0466120363254119e-05,
+      "loss": 0.5769,
+      "step": 9615
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 20.756687172567716,
+      "learning_rate": 1.0464435790657833e-05,
+      "loss": 0.5769,
+      "step": 9616
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.618957540812264,
+      "learning_rate": 1.0462751204853262e-05,
+      "loss": 0.5736,
+      "step": 9617
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.11631268909565,
+      "learning_rate": 1.0461066605888314e-05,
+      "loss": 0.6871,
+      "step": 9618
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.81731376749823,
+      "learning_rate": 1.0459381993810906e-05,
+      "loss": 0.5945,
+      "step": 9619
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.639913323143578,
+      "learning_rate": 1.045769736866894e-05,
+      "loss": 0.6161,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.973731759155125,
+      "learning_rate": 1.0456012730510332e-05,
+      "loss": 0.6132,
+      "step": 9621
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.093217327408952,
+      "learning_rate": 1.0454328079382985e-05,
+      "loss": 0.5851,
+      "step": 9622
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.378059627342083,
+      "learning_rate": 1.0452643415334816e-05,
+      "loss": 0.5905,
+      "step": 9623
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.635351020585734,
+      "learning_rate": 1.045095873841373e-05,
+      "loss": 0.6017,
+      "step": 9624
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.787613591048734,
+      "learning_rate": 1.0449274048667644e-05,
+      "loss": 0.565,
+      "step": 9625
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 23.71521270771573,
+      "learning_rate": 1.0447589346144467e-05,
+      "loss": 0.629,
+      "step": 9626
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.837488716681479,
+      "learning_rate": 1.044590463089211e-05,
+      "loss": 0.6665,
+      "step": 9627
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.80534046064591,
+      "learning_rate": 1.0444219902958483e-05,
+      "loss": 0.6202,
+      "step": 9628
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.543291504337233,
+      "learning_rate": 1.0442535162391508e-05,
+      "loss": 0.5971,
+      "step": 9629
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.99881099247838,
+      "learning_rate": 1.0440850409239085e-05,
+      "loss": 0.5446,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.765826762735369,
+      "learning_rate": 1.043916564354914e-05,
+      "loss": 0.5486,
+      "step": 9631
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 20.81523294733946,
+      "learning_rate": 1.0437480865369577e-05,
+      "loss": 0.6378,
+      "step": 9632
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.563178170735709,
+      "learning_rate": 1.0435796074748316e-05,
+      "loss": 0.5161,
+      "step": 9633
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 18.48273473960778,
+      "learning_rate": 1.0434111271733271e-05,
+      "loss": 0.5999,
+      "step": 9634
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.222243011752166,
+      "learning_rate": 1.0432426456372356e-05,
+      "loss": 0.5401,
+      "step": 9635
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.073605794994442,
+      "learning_rate": 1.0430741628713482e-05,
+      "loss": 0.6115,
+      "step": 9636
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.715624759873922,
+      "learning_rate": 1.042905678880457e-05,
+      "loss": 0.5834,
+      "step": 9637
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 30.531316117768554,
+      "learning_rate": 1.0427371936693531e-05,
+      "loss": 0.6356,
+      "step": 9638
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.905429731539908,
+      "learning_rate": 1.0425687072428288e-05,
+      "loss": 0.5975,
+      "step": 9639
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 28.15837052136694,
+      "learning_rate": 1.042400219605675e-05,
+      "loss": 0.6498,
+      "step": 9640
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.623438171112767,
+      "learning_rate": 1.0422317307626842e-05,
+      "loss": 0.5523,
+      "step": 9641
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.70465619693213,
+      "learning_rate": 1.0420632407186475e-05,
+      "loss": 0.6668,
+      "step": 9642
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.58035066257976,
+      "learning_rate": 1.0418947494783567e-05,
+      "loss": 0.6056,
+      "step": 9643
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.5340884592538,
+      "learning_rate": 1.041726257046604e-05,
+      "loss": 0.5483,
+      "step": 9644
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.576004154279504,
+      "learning_rate": 1.0415577634281806e-05,
+      "loss": 0.544,
+      "step": 9645
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.879241438387634,
+      "learning_rate": 1.0413892686278791e-05,
+      "loss": 0.5846,
+      "step": 9646
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.072237066510255,
+      "learning_rate": 1.0412207726504906e-05,
+      "loss": 0.5579,
+      "step": 9647
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.506065142355302,
+      "learning_rate": 1.041052275500808e-05,
+      "loss": 0.596,
+      "step": 9648
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.049002962311587,
+      "learning_rate": 1.0408837771836222e-05,
+      "loss": 0.5536,
+      "step": 9649
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.397936068666075,
+      "learning_rate": 1.0407152777037258e-05,
+      "loss": 0.6283,
+      "step": 9650
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.235977550239177,
+      "learning_rate": 1.0405467770659107e-05,
+      "loss": 0.6029,
+      "step": 9651
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.61655253112933,
+      "learning_rate": 1.0403782752749692e-05,
+      "loss": 0.6601,
+      "step": 9652
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.83600702410067,
+      "learning_rate": 1.040209772335693e-05,
+      "loss": 0.5642,
+      "step": 9653
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.185093523963605,
+      "learning_rate": 1.0400412682528745e-05,
+      "loss": 0.6029,
+      "step": 9654
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.58923856554717,
+      "learning_rate": 1.0398727630313055e-05,
+      "loss": 0.5541,
+      "step": 9655
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.243442389314847,
+      "learning_rate": 1.0397042566757789e-05,
+      "loss": 0.6346,
+      "step": 9656
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.87343456668084,
+      "learning_rate": 1.039535749191086e-05,
+      "loss": 0.6218,
+      "step": 9657
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.669560513443166,
+      "learning_rate": 1.0393672405820197e-05,
+      "loss": 0.5286,
+      "step": 9658
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 24.856694818496706,
+      "learning_rate": 1.0391987308533722e-05,
+      "loss": 0.6374,
+      "step": 9659
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.917608547927562,
+      "learning_rate": 1.0390302200099355e-05,
+      "loss": 0.5294,
+      "step": 9660
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.980607044193555,
+      "learning_rate": 1.0388617080565024e-05,
+      "loss": 0.6468,
+      "step": 9661
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.413652209010364,
+      "learning_rate": 1.0386931949978649e-05,
+      "loss": 0.5366,
+      "step": 9662
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.617944231443634,
+      "learning_rate": 1.038524680838816e-05,
+      "loss": 0.5906,
+      "step": 9663
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.45518717600868,
+      "learning_rate": 1.038356165584147e-05,
+      "loss": 0.6459,
+      "step": 9664
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.670259877357442,
+      "learning_rate": 1.0381876492386516e-05,
+      "loss": 0.5649,
+      "step": 9665
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.59739134506479,
+      "learning_rate": 1.0380191318071215e-05,
+      "loss": 0.5716,
+      "step": 9666
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.128297436985324,
+      "learning_rate": 1.0378506132943497e-05,
+      "loss": 0.61,
+      "step": 9667
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.822045197951343,
+      "learning_rate": 1.0376820937051286e-05,
+      "loss": 0.6517,
+      "step": 9668
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.933589958913025,
+      "learning_rate": 1.0375135730442507e-05,
+      "loss": 0.5457,
+      "step": 9669
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 25.78790087367503,
+      "learning_rate": 1.0373450513165089e-05,
+      "loss": 0.5006,
+      "step": 9670
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 24.879292001607123,
+      "learning_rate": 1.0371765285266957e-05,
+      "loss": 0.7449,
+      "step": 9671
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.096462902150305,
+      "learning_rate": 1.0370080046796034e-05,
+      "loss": 0.5806,
+      "step": 9672
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.20883964014725,
+      "learning_rate": 1.0368394797800256e-05,
+      "loss": 0.5285,
+      "step": 9673
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.80245911197219,
+      "learning_rate": 1.0366709538327542e-05,
+      "loss": 0.5422,
+      "step": 9674
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.57351551851732,
+      "learning_rate": 1.0365024268425826e-05,
+      "loss": 0.6182,
+      "step": 9675
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.05579679874009,
+      "learning_rate": 1.036333898814303e-05,
+      "loss": 0.578,
+      "step": 9676
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.21522120266898,
+      "learning_rate": 1.0361653697527088e-05,
+      "loss": 0.5919,
+      "step": 9677
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.639236719654644,
+      "learning_rate": 1.0359968396625925e-05,
+      "loss": 0.5879,
+      "step": 9678
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.53797098980881,
+      "learning_rate": 1.0358283085487473e-05,
+      "loss": 0.5471,
+      "step": 9679
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.453302982016123,
+      "learning_rate": 1.0356597764159659e-05,
+      "loss": 0.6067,
+      "step": 9680
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 24.8029974288082,
+      "learning_rate": 1.0354912432690412e-05,
+      "loss": 0.6423,
+      "step": 9681
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.81667495944494,
+      "learning_rate": 1.0353227091127662e-05,
+      "loss": 0.5919,
+      "step": 9682
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.322391265133675,
+      "learning_rate": 1.0351541739519341e-05,
+      "loss": 0.6508,
+      "step": 9683
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.739329140593767,
+      "learning_rate": 1.034985637791338e-05,
+      "loss": 0.6226,
+      "step": 9684
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.4631897523884,
+      "learning_rate": 1.034817100635771e-05,
+      "loss": 0.6291,
+      "step": 9685
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 24.728615029887486,
+      "learning_rate": 1.0346485624900258e-05,
+      "loss": 0.5992,
+      "step": 9686
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 25.357985935608827,
+      "learning_rate": 1.0344800233588959e-05,
+      "loss": 0.6013,
+      "step": 9687
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 26.728368036087755,
+      "learning_rate": 1.0343114832471742e-05,
+      "loss": 0.5807,
+      "step": 9688
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.641954210006624,
+      "learning_rate": 1.0341429421596538e-05,
+      "loss": 0.5927,
+      "step": 9689
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.661158702157888,
+      "learning_rate": 1.0339744001011285e-05,
+      "loss": 0.6064,
+      "step": 9690
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.535338796147864,
+      "learning_rate": 1.0338058570763907e-05,
+      "loss": 0.593,
+      "step": 9691
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.891379072240596,
+      "learning_rate": 1.0336373130902346e-05,
+      "loss": 0.5489,
+      "step": 9692
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.70750703680616,
+      "learning_rate": 1.0334687681474524e-05,
+      "loss": 0.5354,
+      "step": 9693
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.374109674521367,
+      "learning_rate": 1.0333002222528384e-05,
+      "loss": 0.6331,
+      "step": 9694
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.964810037638912,
+      "learning_rate": 1.0331316754111855e-05,
+      "loss": 0.5579,
+      "step": 9695
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.67437888035102,
+      "learning_rate": 1.032963127627287e-05,
+      "loss": 0.5782,
+      "step": 9696
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.720702438182393,
+      "learning_rate": 1.0327945789059366e-05,
+      "loss": 0.5521,
+      "step": 9697
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.055180070100995,
+      "learning_rate": 1.0326260292519274e-05,
+      "loss": 0.6125,
+      "step": 9698
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.082120845605495,
+      "learning_rate": 1.032457478670053e-05,
+      "loss": 0.5681,
+      "step": 9699
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 27.765011397984036,
+      "learning_rate": 1.0322889271651073e-05,
+      "loss": 0.5868,
+      "step": 9700
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.549520035750025,
+      "learning_rate": 1.0321203747418829e-05,
+      "loss": 0.6422,
+      "step": 9701
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.63145328937301,
+      "learning_rate": 1.031951821405174e-05,
+      "loss": 0.5735,
+      "step": 9702
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.452257447679084,
+      "learning_rate": 1.031783267159774e-05,
+      "loss": 0.6196,
+      "step": 9703
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.704933470054325,
+      "learning_rate": 1.0316147120104767e-05,
+      "loss": 0.5698,
+      "step": 9704
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.007496829167547,
+      "learning_rate": 1.0314461559620752e-05,
+      "loss": 0.5526,
+      "step": 9705
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.626932779203724,
+      "learning_rate": 1.0312775990193635e-05,
+      "loss": 0.6084,
+      "step": 9706
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.193909153640224,
+      "learning_rate": 1.0311090411871354e-05,
+      "loss": 0.6134,
+      "step": 9707
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.992071896105898,
+      "learning_rate": 1.030940482470184e-05,
+      "loss": 0.653,
+      "step": 9708
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 23.32786916092986,
+      "learning_rate": 1.0307719228733034e-05,
+      "loss": 0.5311,
+      "step": 9709
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.04289358448721,
+      "learning_rate": 1.0306033624012875e-05,
+      "loss": 0.5914,
+      "step": 9710
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.094784621958546,
+      "learning_rate": 1.0304348010589298e-05,
+      "loss": 0.6218,
+      "step": 9711
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 26.112077356290023,
+      "learning_rate": 1.0302662388510243e-05,
+      "loss": 0.7058,
+      "step": 9712
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.5109792447267,
+      "learning_rate": 1.0300976757823643e-05,
+      "loss": 0.6278,
+      "step": 9713
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.711790830596787,
+      "learning_rate": 1.0299291118577445e-05,
+      "loss": 0.5624,
+      "step": 9714
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.12122326879831,
+      "learning_rate": 1.029760547081958e-05,
+      "loss": 0.6401,
+      "step": 9715
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.863390983195917,
+      "learning_rate": 1.0295919814597988e-05,
+      "loss": 0.5783,
+      "step": 9716
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.91616310685399,
+      "learning_rate": 1.0294234149960614e-05,
+      "loss": 0.6954,
+      "step": 9717
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.848052646670258,
+      "learning_rate": 1.0292548476955389e-05,
+      "loss": 0.6291,
+      "step": 9718
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.508515081407044,
+      "learning_rate": 1.0290862795630261e-05,
+      "loss": 0.6066,
+      "step": 9719
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.698917700203406,
+      "learning_rate": 1.028917710603316e-05,
+      "loss": 0.5862,
+      "step": 9720
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.166175449769707,
+      "learning_rate": 1.0287491408212031e-05,
+      "loss": 0.5832,
+      "step": 9721
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.907137928444762,
+      "learning_rate": 1.0285805702214817e-05,
+      "loss": 0.509,
+      "step": 9722
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.613246921192253,
+      "learning_rate": 1.0284119988089458e-05,
+      "loss": 0.621,
+      "step": 9723
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.33246406149417,
+      "learning_rate": 1.028243426588389e-05,
+      "loss": 0.6292,
+      "step": 9724
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.49944751511538,
+      "learning_rate": 1.0280748535646058e-05,
+      "loss": 0.6251,
+      "step": 9725
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.96674256182304,
+      "learning_rate": 1.0279062797423899e-05,
+      "loss": 0.5056,
+      "step": 9726
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.99551118157452,
+      "learning_rate": 1.0277377051265361e-05,
+      "loss": 0.5474,
+      "step": 9727
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.519572105791326,
+      "learning_rate": 1.027569129721838e-05,
+      "loss": 0.5619,
+      "step": 9728
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.134570249747842,
+      "learning_rate": 1.02740055353309e-05,
+      "loss": 0.5142,
+      "step": 9729
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.55793940025469,
+      "learning_rate": 1.0272319765650862e-05,
+      "loss": 0.551,
+      "step": 9730
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.757582540646,
+      "learning_rate": 1.0270633988226209e-05,
+      "loss": 0.5913,
+      "step": 9731
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.880201362449023,
+      "learning_rate": 1.0268948203104888e-05,
+      "loss": 0.5933,
+      "step": 9732
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.41890471454992,
+      "learning_rate": 1.026726241033483e-05,
+      "loss": 0.6509,
+      "step": 9733
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.561051001233153,
+      "learning_rate": 1.0265576609963994e-05,
+      "loss": 0.5906,
+      "step": 9734
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.22125263748386,
+      "learning_rate": 1.0263890802040308e-05,
+      "loss": 0.6282,
+      "step": 9735
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.599353912923885,
+      "learning_rate": 1.0262204986611726e-05,
+      "loss": 0.6209,
+      "step": 9736
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.644582756360847,
+      "learning_rate": 1.0260519163726183e-05,
+      "loss": 0.5601,
+      "step": 9737
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.62967762103688,
+      "learning_rate": 1.025883333343163e-05,
+      "loss": 0.6239,
+      "step": 9738
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.741691860675274,
+      "learning_rate": 1.025714749577601e-05,
+      "loss": 0.556,
+      "step": 9739
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.94721018892791,
+      "learning_rate": 1.0255461650807264e-05,
+      "loss": 0.5684,
+      "step": 9740
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.220464958127668,
+      "learning_rate": 1.025377579857334e-05,
+      "loss": 0.6047,
+      "step": 9741
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.498191713185314,
+      "learning_rate": 1.0252089939122179e-05,
+      "loss": 0.5286,
+      "step": 9742
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.820507427288362,
+      "learning_rate": 1.0250404072501724e-05,
+      "loss": 0.6005,
+      "step": 9743
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.218123499631304,
+      "learning_rate": 1.024871819875993e-05,
+      "loss": 0.5803,
+      "step": 9744
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.56208453483713,
+      "learning_rate": 1.0247032317944731e-05,
+      "loss": 0.6825,
+      "step": 9745
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.358616007679327,
+      "learning_rate": 1.0245346430104082e-05,
+      "loss": 0.5834,
+      "step": 9746
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.489950339584198,
+      "learning_rate": 1.024366053528592e-05,
+      "loss": 0.6595,
+      "step": 9747
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.524704981387046,
+      "learning_rate": 1.0241974633538198e-05,
+      "loss": 0.6182,
+      "step": 9748
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.349640465368353,
+      "learning_rate": 1.0240288724908858e-05,
+      "loss": 0.6046,
+      "step": 9749
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.977115747664868,
+      "learning_rate": 1.0238602809445846e-05,
+      "loss": 0.6919,
+      "step": 9750
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.00082695529993,
+      "learning_rate": 1.023691688719711e-05,
+      "loss": 0.6042,
+      "step": 9751
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.24340528159094,
+      "learning_rate": 1.0235230958210596e-05,
+      "loss": 0.546,
+      "step": 9752
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.84671425165797,
+      "learning_rate": 1.0233545022534252e-05,
+      "loss": 0.6432,
+      "step": 9753
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.40913042170192,
+      "learning_rate": 1.0231859080216025e-05,
+      "loss": 0.6516,
+      "step": 9754
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 26.12026396343735,
+      "learning_rate": 1.023017313130386e-05,
+      "loss": 0.5668,
+      "step": 9755
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.141763595544948,
+      "learning_rate": 1.0228487175845707e-05,
+      "loss": 0.6057,
+      "step": 9756
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.81871215724684,
+      "learning_rate": 1.0226801213889512e-05,
+      "loss": 0.6012,
+      "step": 9757
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 30.97131783389649,
+      "learning_rate": 1.0225115245483219e-05,
+      "loss": 0.6591,
+      "step": 9758
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.083537548890842,
+      "learning_rate": 1.0223429270674788e-05,
+      "loss": 0.6875,
+      "step": 9759
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.42947557206542,
+      "learning_rate": 1.0221743289512153e-05,
+      "loss": 0.6043,
+      "step": 9760
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 23.06860331990219,
+      "learning_rate": 1.0220057302043273e-05,
+      "loss": 0.6349,
+      "step": 9761
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.762559951179993,
+      "learning_rate": 1.0218371308316089e-05,
+      "loss": 0.5893,
+      "step": 9762
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.131425721705238,
+      "learning_rate": 1.0216685308378556e-05,
+      "loss": 0.6209,
+      "step": 9763
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.56900243104382,
+      "learning_rate": 1.0214999302278614e-05,
+      "loss": 0.5722,
+      "step": 9764
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.030102627278943,
+      "learning_rate": 1.0213313290064222e-05,
+      "loss": 0.5976,
+      "step": 9765
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.47728830861015,
+      "learning_rate": 1.0211627271783323e-05,
+      "loss": 0.5158,
+      "step": 9766
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.42929441760261,
+      "learning_rate": 1.0209941247483868e-05,
+      "loss": 0.574,
+      "step": 9767
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.15787797487095,
+      "learning_rate": 1.0208255217213809e-05,
+      "loss": 0.6116,
+      "step": 9768
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 25.321843004212354,
+      "learning_rate": 1.0206569181021092e-05,
+      "loss": 0.6432,
+      "step": 9769
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.74511198148553,
+      "learning_rate": 1.0204883138953666e-05,
+      "loss": 0.6437,
+      "step": 9770
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.193600862023636,
+      "learning_rate": 1.0203197091059485e-05,
+      "loss": 0.6095,
+      "step": 9771
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.42449413513622,
+      "learning_rate": 1.02015110373865e-05,
+      "loss": 0.5764,
+      "step": 9772
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.61353653422856,
+      "learning_rate": 1.0199824977982658e-05,
+      "loss": 0.5876,
+      "step": 9773
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.01537566618072,
+      "learning_rate": 1.0198138912895907e-05,
+      "loss": 0.5989,
+      "step": 9774
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.777430230276448,
+      "learning_rate": 1.0196452842174202e-05,
+      "loss": 0.5752,
+      "step": 9775
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 21.44965588164675,
+      "learning_rate": 1.0194766765865498e-05,
+      "loss": 0.6001,
+      "step": 9776
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 26.35495229814092,
+      "learning_rate": 1.0193080684017737e-05,
+      "loss": 0.6678,
+      "step": 9777
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.17195117393095,
+      "learning_rate": 1.0191394596678879e-05,
+      "loss": 0.6193,
+      "step": 9778
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 21.180960499109634,
+      "learning_rate": 1.0189708503896865e-05,
+      "loss": 0.5787,
+      "step": 9779
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.98836179604109,
+      "learning_rate": 1.0188022405719652e-05,
+      "loss": 0.6599,
+      "step": 9780
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.355594134001255,
+      "learning_rate": 1.0186336302195197e-05,
+      "loss": 0.655,
+      "step": 9781
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.255886298603194,
+      "learning_rate": 1.0184650193371444e-05,
+      "loss": 0.6365,
+      "step": 9782
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.904635072682954,
+      "learning_rate": 1.0182964079296347e-05,
+      "loss": 0.5763,
+      "step": 9783
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.628876284568133,
+      "learning_rate": 1.0181277960017856e-05,
+      "loss": 0.5941,
+      "step": 9784
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.917966169113157,
+      "learning_rate": 1.017959183558393e-05,
+      "loss": 0.5622,
+      "step": 9785
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.77326155944365,
+      "learning_rate": 1.0177905706042517e-05,
+      "loss": 0.623,
+      "step": 9786
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.67839546872158,
+      "learning_rate": 1.0176219571441565e-05,
+      "loss": 0.552,
+      "step": 9787
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.126178967570326,
+      "learning_rate": 1.0174533431829039e-05,
+      "loss": 0.6912,
+      "step": 9788
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 25.925455739672177,
+      "learning_rate": 1.0172847287252878e-05,
+      "loss": 0.5756,
+      "step": 9789
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.118236408110047,
+      "learning_rate": 1.0171161137761042e-05,
+      "loss": 0.5954,
+      "step": 9790
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.25787566846281,
+      "learning_rate": 1.0169474983401488e-05,
+      "loss": 0.5304,
+      "step": 9791
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.24041767631331,
+      "learning_rate": 1.016778882422216e-05,
+      "loss": 0.5783,
+      "step": 9792
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.971278162102124,
+      "learning_rate": 1.0166102660271018e-05,
+      "loss": 0.5405,
+      "step": 9793
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.206598312785587,
+      "learning_rate": 1.0164416491596014e-05,
+      "loss": 0.5874,
+      "step": 9794
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.00980527651768,
+      "learning_rate": 1.01627303182451e-05,
+      "loss": 0.5576,
+      "step": 9795
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.555684191899413,
+      "learning_rate": 1.016104414026623e-05,
+      "loss": 0.5853,
+      "step": 9796
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 26.607579592369152,
+      "learning_rate": 1.015935795770736e-05,
+      "loss": 0.6985,
+      "step": 9797
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.729322284245352,
+      "learning_rate": 1.0157671770616444e-05,
+      "loss": 0.659,
+      "step": 9798
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.630172165972308,
+      "learning_rate": 1.0155985579041434e-05,
+      "loss": 0.5162,
+      "step": 9799
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.524771154057753,
+      "learning_rate": 1.0154299383030287e-05,
+      "loss": 0.5533,
+      "step": 9800
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 21.371466626955154,
+      "learning_rate": 1.0152613182630953e-05,
+      "loss": 0.684,
+      "step": 9801
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.852210292883242,
+      "learning_rate": 1.0150926977891388e-05,
+      "loss": 0.6126,
+      "step": 9802
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 28.77283713568664,
+      "learning_rate": 1.0149240768859554e-05,
+      "loss": 0.6412,
+      "step": 9803
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.271240985111056,
+      "learning_rate": 1.0147554555583394e-05,
+      "loss": 0.6132,
+      "step": 9804
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.990411848866618,
+      "learning_rate": 1.0145868338110873e-05,
+      "loss": 0.585,
+      "step": 9805
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 21.628319372530903,
+      "learning_rate": 1.0144182116489938e-05,
+      "loss": 0.6258,
+      "step": 9806
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 30.878107570909272,
+      "learning_rate": 1.0142495890768551e-05,
+      "loss": 0.6224,
+      "step": 9807
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.719539571393554,
+      "learning_rate": 1.0140809660994663e-05,
+      "loss": 0.6045,
+      "step": 9808
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.16270071141561,
+      "learning_rate": 1.0139123427216231e-05,
+      "loss": 0.6214,
+      "step": 9809
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.264404714201888,
+      "learning_rate": 1.013743718948121e-05,
+      "loss": 0.5899,
+      "step": 9810
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.782291226396453,
+      "learning_rate": 1.0135750947837558e-05,
+      "loss": 0.5736,
+      "step": 9811
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.00041017878756,
+      "learning_rate": 1.0134064702333225e-05,
+      "loss": 0.5152,
+      "step": 9812
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.09147224906994,
+      "learning_rate": 1.0132378453016171e-05,
+      "loss": 0.6397,
+      "step": 9813
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.93878225236053,
+      "learning_rate": 1.013069219993435e-05,
+      "loss": 0.5874,
+      "step": 9814
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.597211456268017,
+      "learning_rate": 1.0129005943135721e-05,
+      "loss": 0.5429,
+      "step": 9815
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.85037321276858,
+      "learning_rate": 1.012731968266824e-05,
+      "loss": 0.5659,
+      "step": 9816
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.714440965821364,
+      "learning_rate": 1.012563341857986e-05,
+      "loss": 0.5713,
+      "step": 9817
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.695734829018026,
+      "learning_rate": 1.0123947150918539e-05,
+      "loss": 0.5793,
+      "step": 9818
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.717277534218006,
+      "learning_rate": 1.0122260879732231e-05,
+      "loss": 0.6535,
+      "step": 9819
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.35976604736056,
+      "learning_rate": 1.01205746050689e-05,
+      "loss": 0.7365,
+      "step": 9820
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 26.98959759102438,
+      "learning_rate": 1.0118888326976494e-05,
+      "loss": 0.6402,
+      "step": 9821
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.390692643532052,
+      "learning_rate": 1.0117202045502978e-05,
+      "loss": 0.6214,
+      "step": 9822
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 25.895291304471748,
+      "learning_rate": 1.01155157606963e-05,
+      "loss": 0.5579,
+      "step": 9823
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.940143999718817,
+      "learning_rate": 1.0113829472604422e-05,
+      "loss": 0.6262,
+      "step": 9824
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.693191947661525,
+      "learning_rate": 1.0112143181275302e-05,
+      "loss": 0.5833,
+      "step": 9825
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.92902628311404,
+      "learning_rate": 1.0110456886756894e-05,
+      "loss": 0.5885,
+      "step": 9826
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.488499609300288,
+      "learning_rate": 1.010877058909716e-05,
+      "loss": 0.6217,
+      "step": 9827
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.201822237923107,
+      "learning_rate": 1.0107084288344052e-05,
+      "loss": 0.6189,
+      "step": 9828
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.196419569325847,
+      "learning_rate": 1.0105397984545524e-05,
+      "loss": 0.6501,
+      "step": 9829
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.0816349336105,
+      "learning_rate": 1.0103711677749548e-05,
+      "loss": 0.5687,
+      "step": 9830
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 23.91595385392772,
+      "learning_rate": 1.0102025368004066e-05,
+      "loss": 0.5631,
+      "step": 9831
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.674529830649426,
+      "learning_rate": 1.0100339055357048e-05,
+      "loss": 0.6132,
+      "step": 9832
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.405248610498568,
+      "learning_rate": 1.0098652739856441e-05,
+      "loss": 0.6614,
+      "step": 9833
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.41468111471084,
+      "learning_rate": 1.0096966421550209e-05,
+      "loss": 0.5909,
+      "step": 9834
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.370706133186534,
+      "learning_rate": 1.0095280100486309e-05,
+      "loss": 0.6184,
+      "step": 9835
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.965911400366622,
+      "learning_rate": 1.00935937767127e-05,
+      "loss": 0.5726,
+      "step": 9836
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.131970856564038,
+      "learning_rate": 1.0091907450277338e-05,
+      "loss": 0.531,
+      "step": 9837
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.09407529094273,
+      "learning_rate": 1.0090221121228178e-05,
+      "loss": 0.5342,
+      "step": 9838
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.24465590330699,
+      "learning_rate": 1.0088534789613188e-05,
+      "loss": 0.6801,
+      "step": 9839
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.876139976588874,
+      "learning_rate": 1.0086848455480318e-05,
+      "loss": 0.6493,
+      "step": 9840
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.241602855250118,
+      "learning_rate": 1.0085162118877527e-05,
+      "loss": 0.5267,
+      "step": 9841
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 25.365857996756986,
+      "learning_rate": 1.0083475779852778e-05,
+      "loss": 0.6316,
+      "step": 9842
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.95590279190697,
+      "learning_rate": 1.0081789438454026e-05,
+      "loss": 0.5928,
+      "step": 9843
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.109198327474008,
+      "learning_rate": 1.0080103094729229e-05,
+      "loss": 0.5614,
+      "step": 9844
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.289320802512265,
+      "learning_rate": 1.007841674872635e-05,
+      "loss": 0.6248,
+      "step": 9845
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.843820585021568,
+      "learning_rate": 1.007673040049334e-05,
+      "loss": 0.5457,
+      "step": 9846
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.343440797542915,
+      "learning_rate": 1.0075044050078166e-05,
+      "loss": 0.6412,
+      "step": 9847
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.585986569009034,
+      "learning_rate": 1.0073357697528779e-05,
+      "loss": 0.6133,
+      "step": 9848
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.969192878190256,
+      "learning_rate": 1.0071671342893148e-05,
+      "loss": 0.5725,
+      "step": 9849
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.826596777377638,
+      "learning_rate": 1.0069984986219219e-05,
+      "loss": 0.6182,
+      "step": 9850
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.93768335125752,
+      "learning_rate": 1.0068298627554962e-05,
+      "loss": 0.5551,
+      "step": 9851
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.673201439254342,
+      "learning_rate": 1.0066612266948333e-05,
+      "loss": 0.5886,
+      "step": 9852
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.402230321188068,
+      "learning_rate": 1.0064925904447288e-05,
+      "loss": 0.6064,
+      "step": 9853
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.52270246081074,
+      "learning_rate": 1.006323954009979e-05,
+      "loss": 0.6101,
+      "step": 9854
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.883609994830113,
+      "learning_rate": 1.0061553173953794e-05,
+      "loss": 0.6599,
+      "step": 9855
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.423958749690176,
+      "learning_rate": 1.0059866806057263e-05,
+      "loss": 0.5467,
+      "step": 9856
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.857281790198073,
+      "learning_rate": 1.0058180436458156e-05,
+      "loss": 0.5315,
+      "step": 9857
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.796284954170929,
+      "learning_rate": 1.0056494065204428e-05,
+      "loss": 0.5457,
+      "step": 9858
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.452180206763426,
+      "learning_rate": 1.0054807692344045e-05,
+      "loss": 0.6638,
+      "step": 9859
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.82281976866086,
+      "learning_rate": 1.0053121317924963e-05,
+      "loss": 0.516,
+      "step": 9860
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.84377747911087,
+      "learning_rate": 1.0051434941995142e-05,
+      "loss": 0.6408,
+      "step": 9861
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.496517203922064,
+      "learning_rate": 1.004974856460254e-05,
+      "loss": 0.5712,
+      "step": 9862
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.528885550589,
+      "learning_rate": 1.0048062185795117e-05,
+      "loss": 0.6111,
+      "step": 9863
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.583634769062323,
+      "learning_rate": 1.0046375805620838e-05,
+      "loss": 0.5953,
+      "step": 9864
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.530671231331553,
+      "learning_rate": 1.004468942412765e-05,
+      "loss": 0.6013,
+      "step": 9865
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.883483566276666,
+      "learning_rate": 1.0043003041363531e-05,
+      "loss": 0.5862,
+      "step": 9866
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.668462437624736,
+      "learning_rate": 1.0041316657376426e-05,
+      "loss": 0.5047,
+      "step": 9867
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.723909886349947,
+      "learning_rate": 1.0039630272214299e-05,
+      "loss": 0.5517,
+      "step": 9868
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.050017125355206,
+      "learning_rate": 1.003794388592511e-05,
+      "loss": 0.5672,
+      "step": 9869
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.931392360492978,
+      "learning_rate": 1.0036257498556821e-05,
+      "loss": 0.6131,
+      "step": 9870
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.521188137504225,
+      "learning_rate": 1.0034571110157388e-05,
+      "loss": 0.5368,
+      "step": 9871
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.832271467979282,
+      "learning_rate": 1.0032884720774773e-05,
+      "loss": 0.6052,
+      "step": 9872
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 21.914611576973698,
+      "learning_rate": 1.0031198330456936e-05,
+      "loss": 0.5725,
+      "step": 9873
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.94124677633545,
+      "learning_rate": 1.002951193925184e-05,
+      "loss": 0.6069,
+      "step": 9874
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.148054018719586,
+      "learning_rate": 1.0027825547207435e-05,
+      "loss": 0.5259,
+      "step": 9875
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.810999591946036,
+      "learning_rate": 1.0026139154371694e-05,
+      "loss": 0.5989,
+      "step": 9876
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.218846892980409,
+      "learning_rate": 1.0024452760792566e-05,
+      "loss": 0.6176,
+      "step": 9877
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.009688907103346,
+      "learning_rate": 1.0022766366518018e-05,
+      "loss": 0.6078,
+      "step": 9878
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 26.936615283271358,
+      "learning_rate": 1.0021079971596009e-05,
+      "loss": 0.6418,
+      "step": 9879
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.099479744977542,
+      "learning_rate": 1.0019393576074497e-05,
+      "loss": 0.4746,
+      "step": 9880
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.946486662262915,
+      "learning_rate": 1.0017707180001443e-05,
+      "loss": 0.5979,
+      "step": 9881
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.441687089393028,
+      "learning_rate": 1.0016020783424805e-05,
+      "loss": 0.6242,
+      "step": 9882
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.200995576274416,
+      "learning_rate": 1.001433438639255e-05,
+      "loss": 0.5817,
+      "step": 9883
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 25.13079098251688,
+      "learning_rate": 1.0012647988952628e-05,
+      "loss": 0.7288,
+      "step": 9884
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.194260132108658,
+      "learning_rate": 1.0010961591153008e-05,
+      "loss": 0.6085,
+      "step": 9885
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.953252793714666,
+      "learning_rate": 1.0009275193041645e-05,
+      "loss": 0.6196,
+      "step": 9886
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 29.297800509520645,
+      "learning_rate": 1.0007588794666503e-05,
+      "loss": 0.6428,
+      "step": 9887
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.571010031342773,
+      "learning_rate": 1.000590239607554e-05,
+      "loss": 0.5867,
+      "step": 9888
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.933519257083526,
+      "learning_rate": 1.0004215997316715e-05,
+      "loss": 0.5932,
+      "step": 9889
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.45958911313961,
+      "learning_rate": 1.0002529598437988e-05,
+      "loss": 0.5464,
+      "step": 9890
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.00742481785263,
+      "learning_rate": 1.0000843199487325e-05,
+      "loss": 0.6235,
+      "step": 9891
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.62902593804824,
+      "learning_rate": 9.99915680051268e-06,
+      "loss": 0.6296,
+      "step": 9892
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.121323145343172,
+      "learning_rate": 9.997470401562015e-06,
+      "loss": 0.5748,
+      "step": 9893
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.20997601089092,
+      "learning_rate": 9.995784002683288e-06,
+      "loss": 0.6191,
+      "step": 9894
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.328343513168837,
+      "learning_rate": 9.994097603924462e-06,
+      "loss": 0.5634,
+      "step": 9895
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.472865471744875,
+      "learning_rate": 9.992411205333498e-06,
+      "loss": 0.6051,
+      "step": 9896
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.655178533817047,
+      "learning_rate": 9.990724806958358e-06,
+      "loss": 0.5335,
+      "step": 9897
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.636031693794646,
+      "learning_rate": 9.989038408846996e-06,
+      "loss": 0.6225,
+      "step": 9898
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.846735055191772,
+      "learning_rate": 9.987352011047374e-06,
+      "loss": 0.6031,
+      "step": 9899
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.382934042715444,
+      "learning_rate": 9.985665613607454e-06,
+      "loss": 0.5799,
+      "step": 9900
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.21675362455437,
+      "learning_rate": 9.983979216575195e-06,
+      "loss": 0.534,
+      "step": 9901
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.166205386545933,
+      "learning_rate": 9.982292819998562e-06,
+      "loss": 0.6495,
+      "step": 9902
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.94326492206648,
+      "learning_rate": 9.980606423925506e-06,
+      "loss": 0.7226,
+      "step": 9903
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.587813356402894,
+      "learning_rate": 9.978920028403995e-06,
+      "loss": 0.6983,
+      "step": 9904
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 23.414201246632665,
+      "learning_rate": 9.977233633481984e-06,
+      "loss": 0.5428,
+      "step": 9905
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.64245689516531,
+      "learning_rate": 9.975547239207435e-06,
+      "loss": 0.5448,
+      "step": 9906
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.025556916818953,
+      "learning_rate": 9.973860845628311e-06,
+      "loss": 0.6219,
+      "step": 9907
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.963216960531298,
+      "learning_rate": 9.972174452792568e-06,
+      "loss": 0.537,
+      "step": 9908
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.694681263352486,
+      "learning_rate": 9.970488060748164e-06,
+      "loss": 0.585,
+      "step": 9909
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.050040062475464,
+      "learning_rate": 9.968801669543066e-06,
+      "loss": 0.6326,
+      "step": 9910
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.306401981461846,
+      "learning_rate": 9.967115279225228e-06,
+      "loss": 0.6385,
+      "step": 9911
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.722230479390333,
+      "learning_rate": 9.965428889842617e-06,
+      "loss": 0.5485,
+      "step": 9912
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.268524689447355,
+      "learning_rate": 9.963742501443184e-06,
+      "loss": 0.5791,
+      "step": 9913
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.611129566668549,
+      "learning_rate": 9.962056114074893e-06,
+      "loss": 0.5641,
+      "step": 9914
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.79455886041507,
+      "learning_rate": 9.960369727785703e-06,
+      "loss": 0.687,
+      "step": 9915
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.246771988185504,
+      "learning_rate": 9.958683342623579e-06,
+      "loss": 0.5719,
+      "step": 9916
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.210505642019125,
+      "learning_rate": 9.956996958636474e-06,
+      "loss": 0.5923,
+      "step": 9917
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.489068636129506,
+      "learning_rate": 9.955310575872351e-06,
+      "loss": 0.6331,
+      "step": 9918
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.104498795216216,
+      "learning_rate": 9.953624194379165e-06,
+      "loss": 0.5477,
+      "step": 9919
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.122084049287043,
+      "learning_rate": 9.951937814204884e-06,
+      "loss": 0.5328,
+      "step": 9920
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.779803803681574,
+      "learning_rate": 9.950251435397466e-06,
+      "loss": 0.6101,
+      "step": 9921
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.895115824106036,
+      "learning_rate": 9.948565058004863e-06,
+      "loss": 0.6086,
+      "step": 9922
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.166533984156313,
+      "learning_rate": 9.94687868207504e-06,
+      "loss": 0.5932,
+      "step": 9923
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.573150429252294,
+      "learning_rate": 9.945192307655959e-06,
+      "loss": 0.5164,
+      "step": 9924
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.22028948068708,
+      "learning_rate": 9.943505934795572e-06,
+      "loss": 0.6668,
+      "step": 9925
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.384813425233407,
+      "learning_rate": 9.941819563541849e-06,
+      "loss": 0.644,
+      "step": 9926
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.500511899354784,
+      "learning_rate": 9.940133193942742e-06,
+      "loss": 0.576,
+      "step": 9927
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 26.124508283290584,
+      "learning_rate": 9.938446826046209e-06,
+      "loss": 0.614,
+      "step": 9928
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.123440994681633,
+      "learning_rate": 9.936760459900215e-06,
+      "loss": 0.5415,
+      "step": 9929
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.66765646729247,
+      "learning_rate": 9.935074095552714e-06,
+      "loss": 0.6245,
+      "step": 9930
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.512151878962634,
+      "learning_rate": 9.933387733051672e-06,
+      "loss": 0.6438,
+      "step": 9931
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.247368819620945,
+      "learning_rate": 9.93170137244504e-06,
+      "loss": 0.6089,
+      "step": 9932
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.046404442505159,
+      "learning_rate": 9.930015013780783e-06,
+      "loss": 0.5949,
+      "step": 9933
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.964185564414901,
+      "learning_rate": 9.928328657106855e-06,
+      "loss": 0.4891,
+      "step": 9934
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.93954144238853,
+      "learning_rate": 9.92664230247122e-06,
+      "loss": 0.644,
+      "step": 9935
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.822924170105942,
+      "learning_rate": 9.924955949921839e-06,
+      "loss": 0.5922,
+      "step": 9936
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 26.63956348843764,
+      "learning_rate": 9.923269599506664e-06,
+      "loss": 0.6541,
+      "step": 9937
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.025018378244766,
+      "learning_rate": 9.921583251273654e-06,
+      "loss": 0.5801,
+      "step": 9938
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.50971012217428,
+      "learning_rate": 9.919896905270772e-06,
+      "loss": 0.6773,
+      "step": 9939
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.90671493267816,
+      "learning_rate": 9.918210561545974e-06,
+      "loss": 0.559,
+      "step": 9940
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.98238292336455,
+      "learning_rate": 9.916524220147224e-06,
+      "loss": 0.6073,
+      "step": 9941
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.50327156007637,
+      "learning_rate": 9.914837881122474e-06,
+      "loss": 0.4855,
+      "step": 9942
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.27992628559294,
+      "learning_rate": 9.913151544519685e-06,
+      "loss": 0.5874,
+      "step": 9943
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.04891448232537,
+      "learning_rate": 9.911465210386813e-06,
+      "loss": 0.6289,
+      "step": 9944
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.440772380831724,
+      "learning_rate": 9.909778878771822e-06,
+      "loss": 0.5556,
+      "step": 9945
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.39156764333135,
+      "learning_rate": 9.908092549722667e-06,
+      "loss": 0.6782,
+      "step": 9946
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.937471535272632,
+      "learning_rate": 9.906406223287304e-06,
+      "loss": 0.6178,
+      "step": 9947
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.38511420886467,
+      "learning_rate": 9.904719899513693e-06,
+      "loss": 0.5479,
+      "step": 9948
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.071545944442061,
+      "learning_rate": 9.903033578449793e-06,
+      "loss": 0.5566,
+      "step": 9949
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.42644736235071,
+      "learning_rate": 9.90134726014356e-06,
+      "loss": 0.5795,
+      "step": 9950
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 26.848196280376573,
+      "learning_rate": 9.899660944642957e-06,
+      "loss": 0.5194,
+      "step": 9951
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.89129961708667,
+      "learning_rate": 9.897974631995937e-06,
+      "loss": 0.6012,
+      "step": 9952
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.843725417473586,
+      "learning_rate": 9.896288322250455e-06,
+      "loss": 0.5459,
+      "step": 9953
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 27.010483674494523,
+      "learning_rate": 9.894602015454476e-06,
+      "loss": 0.603,
+      "step": 9954
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.783049845427678,
+      "learning_rate": 9.892915711655953e-06,
+      "loss": 0.5753,
+      "step": 9955
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.184418986024316,
+      "learning_rate": 9.891229410902846e-06,
+      "loss": 0.6342,
+      "step": 9956
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 33.941184736712835,
+      "learning_rate": 9.88954311324311e-06,
+      "loss": 0.5585,
+      "step": 9957
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.565286535572284,
+      "learning_rate": 9.887856818724702e-06,
+      "loss": 0.6155,
+      "step": 9958
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.423762633025335,
+      "learning_rate": 9.88617052739558e-06,
+      "loss": 0.5807,
+      "step": 9959
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.059084364987115,
+      "learning_rate": 9.8844842393037e-06,
+      "loss": 0.6225,
+      "step": 9960
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.25280894387938,
+      "learning_rate": 9.882797954497028e-06,
+      "loss": 0.5436,
+      "step": 9961
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.743540728517914,
+      "learning_rate": 9.881111673023509e-06,
+      "loss": 0.5998,
+      "step": 9962
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.814166057385815,
+      "learning_rate": 9.879425394931103e-06,
+      "loss": 0.577,
+      "step": 9963
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.453038243484357,
+      "learning_rate": 9.877739120267769e-06,
+      "loss": 0.5843,
+      "step": 9964
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.67659237539244,
+      "learning_rate": 9.876052849081467e-06,
+      "loss": 0.6041,
+      "step": 9965
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.768110269139063,
+      "learning_rate": 9.874366581420144e-06,
+      "loss": 0.624,
+      "step": 9966
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.144665052704863,
+      "learning_rate": 9.872680317331764e-06,
+      "loss": 0.5898,
+      "step": 9967
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.512891213380023,
+      "learning_rate": 9.87099405686428e-06,
+      "loss": 0.5721,
+      "step": 9968
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.02635306769787,
+      "learning_rate": 9.869307800065651e-06,
+      "loss": 0.5754,
+      "step": 9969
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.452335579109473,
+      "learning_rate": 9.867621546983834e-06,
+      "loss": 0.5932,
+      "step": 9970
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.942141635817212,
+      "learning_rate": 9.86593529766678e-06,
+      "loss": 0.623,
+      "step": 9971
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.76215330603838,
+      "learning_rate": 9.864249052162447e-06,
+      "loss": 0.504,
+      "step": 9972
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.11229616887807,
+      "learning_rate": 9.862562810518792e-06,
+      "loss": 0.5537,
+      "step": 9973
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.006137157785712,
+      "learning_rate": 9.860876572783772e-06,
+      "loss": 0.6332,
+      "step": 9974
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.069077922642027,
+      "learning_rate": 9.859190339005342e-06,
+      "loss": 0.5788,
+      "step": 9975
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 21.022021846320726,
+      "learning_rate": 9.857504109231452e-06,
+      "loss": 0.6415,
+      "step": 9976
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.438184969558296,
+      "learning_rate": 9.855817883510063e-06,
+      "loss": 0.5972,
+      "step": 9977
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.07915214285458,
+      "learning_rate": 9.854131661889129e-06,
+      "loss": 0.627,
+      "step": 9978
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.43861554578508,
+      "learning_rate": 9.852445444416606e-06,
+      "loss": 0.6388,
+      "step": 9979
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.964867473436495,
+      "learning_rate": 9.850759231140451e-06,
+      "loss": 0.5946,
+      "step": 9980
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.13246155737529,
+      "learning_rate": 9.849073022108613e-06,
+      "loss": 0.5865,
+      "step": 9981
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.534122396204978,
+      "learning_rate": 9.84738681736905e-06,
+      "loss": 0.6329,
+      "step": 9982
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.400952487656827,
+      "learning_rate": 9.845700616969718e-06,
+      "loss": 0.7185,
+      "step": 9983
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.862095055385375,
+      "learning_rate": 9.844014420958567e-06,
+      "loss": 0.611,
+      "step": 9984
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.366300341082948,
+      "learning_rate": 9.84232822938356e-06,
+      "loss": 0.5397,
+      "step": 9985
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.757542661964088,
+      "learning_rate": 9.840642042292643e-06,
+      "loss": 0.6127,
+      "step": 9986
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.247783334322346,
+      "learning_rate": 9.838955859733773e-06,
+      "loss": 0.6427,
+      "step": 9987
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.24752014923365,
+      "learning_rate": 9.837269681754901e-06,
+      "loss": 0.6351,
+      "step": 9988
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.357579347341186,
+      "learning_rate": 9.83558350840399e-06,
+      "loss": 0.6147,
+      "step": 9989
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.714189794566042,
+      "learning_rate": 9.833897339728987e-06,
+      "loss": 0.6498,
+      "step": 9990
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 21.38621738142277,
+      "learning_rate": 9.832211175777841e-06,
+      "loss": 0.626,
+      "step": 9991
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.86324138088125,
+      "learning_rate": 9.830525016598515e-06,
+      "loss": 0.6015,
+      "step": 9992
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.786048721210143,
+      "learning_rate": 9.82883886223896e-06,
+      "loss": 0.6237,
+      "step": 9993
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.315495899431646,
+      "learning_rate": 9.827152712747122e-06,
+      "loss": 0.5222,
+      "step": 9994
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.484828884977205,
+      "learning_rate": 9.825466568170966e-06,
+      "loss": 0.5161,
+      "step": 9995
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 23.56019167002632,
+      "learning_rate": 9.823780428558437e-06,
+      "loss": 0.5658,
+      "step": 9996
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.197611497185168,
+      "learning_rate": 9.822094293957486e-06,
+      "loss": 0.6232,
+      "step": 9997
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.839589939946194,
+      "learning_rate": 9.820408164416071e-06,
+      "loss": 0.5993,
+      "step": 9998
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.453414585254144,
+      "learning_rate": 9.818722039982145e-06,
+      "loss": 0.5559,
+      "step": 9999
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.429415153109442,
+      "learning_rate": 9.81703592070366e-06,
+      "loss": 0.5934,
+      "step": 10000
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.833704595167802,
+      "learning_rate": 9.81534980662856e-06,
+      "loss": 0.627,
+      "step": 10001
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 22.602764708174913,
+      "learning_rate": 9.813663697804808e-06,
+      "loss": 0.6003,
+      "step": 10002
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.716150741705444,
+      "learning_rate": 9.811977594280348e-06,
+      "loss": 0.5747,
+      "step": 10003
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 23.605941266512165,
+      "learning_rate": 9.810291496103137e-06,
+      "loss": 0.6323,
+      "step": 10004
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 32.268516663109075,
+      "learning_rate": 9.808605403321128e-06,
+      "loss": 0.7072,
+      "step": 10005
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.580157368921157,
+      "learning_rate": 9.806919315982266e-06,
+      "loss": 0.709,
+      "step": 10006
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.00704672918317,
+      "learning_rate": 9.805233234134504e-06,
+      "loss": 0.5557,
+      "step": 10007
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.86993196473675,
+      "learning_rate": 9.803547157825796e-06,
+      "loss": 0.6805,
+      "step": 10008
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 23.02845911167214,
+      "learning_rate": 9.801861087104093e-06,
+      "loss": 0.5853,
+      "step": 10009
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.83648681667443,
+      "learning_rate": 9.800175022017346e-06,
+      "loss": 0.5768,
+      "step": 10010
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.236295876082348,
+      "learning_rate": 9.798488962613503e-06,
+      "loss": 0.5331,
+      "step": 10011
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.249045818244447,
+      "learning_rate": 9.796802908940516e-06,
+      "loss": 0.5372,
+      "step": 10012
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.572018369707923,
+      "learning_rate": 9.795116861046334e-06,
+      "loss": 0.6242,
+      "step": 10013
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.72488461600909,
+      "learning_rate": 9.793430818978913e-06,
+      "loss": 0.6959,
+      "step": 10014
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.468377731677613,
+      "learning_rate": 9.791744782786196e-06,
+      "loss": 0.5991,
+      "step": 10015
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.222245572684825,
+      "learning_rate": 9.790058752516134e-06,
+      "loss": 0.6418,
+      "step": 10016
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.746107808403732,
+      "learning_rate": 9.788372728216679e-06,
+      "loss": 0.5471,
+      "step": 10017
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.25826421638985,
+      "learning_rate": 9.786686709935781e-06,
+      "loss": 0.5759,
+      "step": 10018
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.3216778129315,
+      "learning_rate": 9.785000697721391e-06,
+      "loss": 0.5616,
+      "step": 10019
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.998181417608414,
+      "learning_rate": 9.783314691621451e-06,
+      "loss": 0.6413,
+      "step": 10020
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.35888629741709,
+      "learning_rate": 9.781628691683916e-06,
+      "loss": 0.6424,
+      "step": 10021
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.459458089193458,
+      "learning_rate": 9.77994269795673e-06,
+      "loss": 0.5769,
+      "step": 10022
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.25476294638253,
+      "learning_rate": 9.778256710487849e-06,
+      "loss": 0.5612,
+      "step": 10023
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 23.980834826409183,
+      "learning_rate": 9.776570729325217e-06,
+      "loss": 0.6082,
+      "step": 10024
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.90638193236052,
+      "learning_rate": 9.774884754516783e-06,
+      "loss": 0.6031,
+      "step": 10025
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.400623857259639,
+      "learning_rate": 9.773198786110492e-06,
+      "loss": 0.6128,
+      "step": 10026
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.086633911575074,
+      "learning_rate": 9.771512824154297e-06,
+      "loss": 0.5134,
+      "step": 10027
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.330108139346816,
+      "learning_rate": 9.76982686869614e-06,
+      "loss": 0.572,
+      "step": 10028
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.356847875643304,
+      "learning_rate": 9.768140919783979e-06,
+      "loss": 0.5971,
+      "step": 10029
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.6674742646926,
+      "learning_rate": 9.766454977465751e-06,
+      "loss": 0.6008,
+      "step": 10030
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 25.692674453288213,
+      "learning_rate": 9.764769041789408e-06,
+      "loss": 0.5852,
+      "step": 10031
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 23.52018672952823,
+      "learning_rate": 9.763083112802891e-06,
+      "loss": 0.558,
+      "step": 10032
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.73908533955712,
+      "learning_rate": 9.761397190554156e-06,
+      "loss": 0.6074,
+      "step": 10033
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.529178335437962,
+      "learning_rate": 9.759711275091149e-06,
+      "loss": 0.5673,
+      "step": 10034
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.023308214107018,
+      "learning_rate": 9.758025366461805e-06,
+      "loss": 0.5685,
+      "step": 10035
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.723803368653815,
+      "learning_rate": 9.756339464714081e-06,
+      "loss": 0.5399,
+      "step": 10036
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.34524316952747,
+      "learning_rate": 9.754653569895922e-06,
+      "loss": 0.5646,
+      "step": 10037
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.518265496216117,
+      "learning_rate": 9.752967682055269e-06,
+      "loss": 0.582,
+      "step": 10038
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.7688568391926,
+      "learning_rate": 9.751281801240075e-06,
+      "loss": 0.756,
+      "step": 10039
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.830683029405115,
+      "learning_rate": 9.749595927498277e-06,
+      "loss": 0.664,
+      "step": 10040
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.67313052897807,
+      "learning_rate": 9.747910060877824e-06,
+      "loss": 0.5482,
+      "step": 10041
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.71254616101865,
+      "learning_rate": 9.746224201426663e-06,
+      "loss": 0.566,
+      "step": 10042
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.922229060621065,
+      "learning_rate": 9.744538349192736e-06,
+      "loss": 0.5583,
+      "step": 10043
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.172194401263614,
+      "learning_rate": 9.742852504223995e-06,
+      "loss": 0.5571,
+      "step": 10044
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 21.68311685880578,
+      "learning_rate": 9.741166666568371e-06,
+      "loss": 0.5765,
+      "step": 10045
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.700981563864698,
+      "learning_rate": 9.739480836273819e-06,
+      "loss": 0.5952,
+      "step": 10046
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 26.562755268216712,
+      "learning_rate": 9.737795013388277e-06,
+      "loss": 0.6485,
+      "step": 10047
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 26.416795483710313,
+      "learning_rate": 9.736109197959693e-06,
+      "loss": 0.6672,
+      "step": 10048
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 20.315048663829586,
+      "learning_rate": 9.734423390036011e-06,
+      "loss": 0.5546,
+      "step": 10049
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.861351698157037,
+      "learning_rate": 9.732737589665171e-06,
+      "loss": 0.6097,
+      "step": 10050
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.340926464405918,
+      "learning_rate": 9.731051796895116e-06,
+      "loss": 0.55,
+      "step": 10051
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 24.10408875257175,
+      "learning_rate": 9.729366011773793e-06,
+      "loss": 0.6593,
+      "step": 10052
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.62370364724231,
+      "learning_rate": 9.727680234349138e-06,
+      "loss": 0.5941,
+      "step": 10053
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.556889430353527,
+      "learning_rate": 9.725994464669103e-06,
+      "loss": 0.592,
+      "step": 10054
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 24.830167263869914,
+      "learning_rate": 9.724308702781625e-06,
+      "loss": 0.617,
+      "step": 10055
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 22.948813933028337,
+      "learning_rate": 9.722622948734644e-06,
+      "loss": 0.5507,
+      "step": 10056
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.237573565261453,
+      "learning_rate": 9.720937202576101e-06,
+      "loss": 0.579,
+      "step": 10057
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.538579105661437,
+      "learning_rate": 9.719251464353944e-06,
+      "loss": 0.5689,
+      "step": 10058
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.259593188770477,
+      "learning_rate": 9.717565734116114e-06,
+      "loss": 0.5896,
+      "step": 10059
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.278848890750854,
+      "learning_rate": 9.715880011910545e-06,
+      "loss": 0.5417,
+      "step": 10060
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.032823222949075,
+      "learning_rate": 9.714194297785184e-06,
+      "loss": 0.5356,
+      "step": 10061
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.800956271110064,
+      "learning_rate": 9.71250859178797e-06,
+      "loss": 0.6126,
+      "step": 10062
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.2278365407668,
+      "learning_rate": 9.710822893966845e-06,
+      "loss": 0.6029,
+      "step": 10063
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.88481036378908,
+      "learning_rate": 9.709137204369746e-06,
+      "loss": 0.6065,
+      "step": 10064
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.89216972459948,
+      "learning_rate": 9.707451523044614e-06,
+      "loss": 0.6157,
+      "step": 10065
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.80632905571142,
+      "learning_rate": 9.705765850039388e-06,
+      "loss": 0.5651,
+      "step": 10066
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.021337088967,
+      "learning_rate": 9.704080185402012e-06,
+      "loss": 0.5574,
+      "step": 10067
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 20.36421736968079,
+      "learning_rate": 9.702394529180424e-06,
+      "loss": 0.5428,
+      "step": 10068
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.457185828416597,
+      "learning_rate": 9.70070888142256e-06,
+      "loss": 0.5958,
+      "step": 10069
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.866692643557005,
+      "learning_rate": 9.699023242176358e-06,
+      "loss": 0.5505,
+      "step": 10070
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.840051949893022,
+      "learning_rate": 9.69733761148976e-06,
+      "loss": 0.5661,
+      "step": 10071
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 23.585973973986793,
+      "learning_rate": 9.695651989410702e-06,
+      "loss": 0.6339,
+      "step": 10072
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 25.877395682531404,
+      "learning_rate": 9.693966375987128e-06,
+      "loss": 0.5814,
+      "step": 10073
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.72589435134867,
+      "learning_rate": 9.692280771266969e-06,
+      "loss": 0.503,
+      "step": 10074
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.221366208720603,
+      "learning_rate": 9.690595175298164e-06,
+      "loss": 0.4694,
+      "step": 10075
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.782637790812743,
+      "learning_rate": 9.68890958812865e-06,
+      "loss": 0.577,
+      "step": 10076
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 29.370909077281556,
+      "learning_rate": 9.687224009806366e-06,
+      "loss": 0.7095,
+      "step": 10077
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.02968395707663,
+      "learning_rate": 9.685538440379253e-06,
+      "loss": 0.5078,
+      "step": 10078
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.367498899872174,
+      "learning_rate": 9.683852879895236e-06,
+      "loss": 0.5463,
+      "step": 10079
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.613962039747314,
+      "learning_rate": 9.682167328402261e-06,
+      "loss": 0.4927,
+      "step": 10080
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.779577303303533,
+      "learning_rate": 9.680481785948263e-06,
+      "loss": 0.551,
+      "step": 10081
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.584457012445984,
+      "learning_rate": 9.678796252581171e-06,
+      "loss": 0.6895,
+      "step": 10082
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.44529614103979,
+      "learning_rate": 9.677110728348932e-06,
+      "loss": 0.6271,
+      "step": 10083
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.204992829401718,
+      "learning_rate": 9.675425213299471e-06,
+      "loss": 0.6812,
+      "step": 10084
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.054014717301907,
+      "learning_rate": 9.673739707480727e-06,
+      "loss": 0.6447,
+      "step": 10085
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.26093790209592,
+      "learning_rate": 9.672054210940638e-06,
+      "loss": 0.564,
+      "step": 10086
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.031045110184525,
+      "learning_rate": 9.670368723727131e-06,
+      "loss": 0.5727,
+      "step": 10087
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 21.579104807331625,
+      "learning_rate": 9.66868324588815e-06,
+      "loss": 0.5525,
+      "step": 10088
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.635028514488914,
+      "learning_rate": 9.66699777747162e-06,
+      "loss": 0.5841,
+      "step": 10089
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.389476404473285,
+      "learning_rate": 9.665312318525478e-06,
+      "loss": 0.5501,
+      "step": 10090
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.356013336116614,
+      "learning_rate": 9.663626869097657e-06,
+      "loss": 0.6834,
+      "step": 10091
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 22.067622225700482,
+      "learning_rate": 9.661941429236094e-06,
+      "loss": 0.6668,
+      "step": 10092
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.438632758348085,
+      "learning_rate": 9.66025599898872e-06,
+      "loss": 0.5709,
+      "step": 10093
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.151288601924566,
+      "learning_rate": 9.658570578403465e-06,
+      "loss": 0.5981,
+      "step": 10094
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.026308965783524,
+      "learning_rate": 9.656885167528261e-06,
+      "loss": 0.5739,
+      "step": 10095
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.103073550600058,
+      "learning_rate": 9.655199766411044e-06,
+      "loss": 0.6453,
+      "step": 10096
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.13750904108484,
+      "learning_rate": 9.653514375099742e-06,
+      "loss": 0.5524,
+      "step": 10097
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.252009208884083,
+      "learning_rate": 9.651828993642293e-06,
+      "loss": 0.6176,
+      "step": 10098
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.69996785979489,
+      "learning_rate": 9.650143622086621e-06,
+      "loss": 0.5696,
+      "step": 10099
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.49341584530932,
+      "learning_rate": 9.64845826048066e-06,
+      "loss": 0.6005,
+      "step": 10100
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.68649876027647,
+      "learning_rate": 9.646772908872338e-06,
+      "loss": 0.4704,
+      "step": 10101
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.16287216606658,
+      "learning_rate": 9.64508756730959e-06,
+      "loss": 0.6085,
+      "step": 10102
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.04830133901125,
+      "learning_rate": 9.643402235840346e-06,
+      "loss": 0.4905,
+      "step": 10103
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.06883674265519,
+      "learning_rate": 9.641716914512532e-06,
+      "loss": 0.55,
+      "step": 10104
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.27897008856571,
+      "learning_rate": 9.640031603374078e-06,
+      "loss": 0.5096,
+      "step": 10105
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.226691759753798,
+      "learning_rate": 9.638346302472916e-06,
+      "loss": 0.4599,
+      "step": 10106
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.126980546467149,
+      "learning_rate": 9.636661011856971e-06,
+      "loss": 0.543,
+      "step": 10107
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.663240713244512,
+      "learning_rate": 9.63497573157418e-06,
+      "loss": 0.5798,
+      "step": 10108
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.931403170806357,
+      "learning_rate": 9.633290461672463e-06,
+      "loss": 0.6677,
+      "step": 10109
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.958064945119073,
+      "learning_rate": 9.631605202199748e-06,
+      "loss": 0.619,
+      "step": 10110
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.973811260379875,
+      "learning_rate": 9.629919953203966e-06,
+      "loss": 0.5956,
+      "step": 10111
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.270058285730773,
+      "learning_rate": 9.628234714733048e-06,
+      "loss": 0.5969,
+      "step": 10112
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.911981900015906,
+      "learning_rate": 9.626549486834916e-06,
+      "loss": 0.6023,
+      "step": 10113
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.619658509364537,
+      "learning_rate": 9.624864269557495e-06,
+      "loss": 0.6119,
+      "step": 10114
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.55474463117469,
+      "learning_rate": 9.623179062948716e-06,
+      "loss": 0.5962,
+      "step": 10115
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.891516908957403,
+      "learning_rate": 9.621493867056505e-06,
+      "loss": 0.556,
+      "step": 10116
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.622133667595797,
+      "learning_rate": 9.619808681928788e-06,
+      "loss": 0.4833,
+      "step": 10117
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 30.217225328225783,
+      "learning_rate": 9.618123507613487e-06,
+      "loss": 0.5976,
+      "step": 10118
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.319260472563556,
+      "learning_rate": 9.616438344158533e-06,
+      "loss": 0.6221,
+      "step": 10119
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.064963212064075,
+      "learning_rate": 9.614753191611846e-06,
+      "loss": 0.4999,
+      "step": 10120
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.670909195385725,
+      "learning_rate": 9.613068050021353e-06,
+      "loss": 0.605,
+      "step": 10121
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.09931893999679,
+      "learning_rate": 9.61138291943498e-06,
+      "loss": 0.6798,
+      "step": 10122
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 21.721724371359706,
+      "learning_rate": 9.609697799900647e-06,
+      "loss": 0.6061,
+      "step": 10123
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.680934739331123,
+      "learning_rate": 9.608012691466281e-06,
+      "loss": 0.5393,
+      "step": 10124
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 22.99617350340761,
+      "learning_rate": 9.606327594179806e-06,
+      "loss": 0.5694,
+      "step": 10125
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.884327146865957,
+      "learning_rate": 9.604642508089142e-06,
+      "loss": 0.6111,
+      "step": 10126
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.003422813282558,
+      "learning_rate": 9.602957433242218e-06,
+      "loss": 0.6758,
+      "step": 10127
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 22.745249444803786,
+      "learning_rate": 9.601272369686948e-06,
+      "loss": 0.6849,
+      "step": 10128
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.502951511291759,
+      "learning_rate": 9.599587317471259e-06,
+      "loss": 0.5306,
+      "step": 10129
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.108790051654342,
+      "learning_rate": 9.597902276643074e-06,
+      "loss": 0.4684,
+      "step": 10130
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.67386112216378,
+      "learning_rate": 9.59621724725031e-06,
+      "loss": 0.5954,
+      "step": 10131
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 27.41976670422977,
+      "learning_rate": 9.594532229340898e-06,
+      "loss": 0.5848,
+      "step": 10132
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.702352275831599,
+      "learning_rate": 9.592847222962744e-06,
+      "loss": 0.543,
+      "step": 10133
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.863435959077865,
+      "learning_rate": 9.591162228163781e-06,
+      "loss": 0.6405,
+      "step": 10134
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.884657693043843,
+      "learning_rate": 9.589477244991924e-06,
+      "loss": 0.526,
+      "step": 10135
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 21.56997672941892,
+      "learning_rate": 9.587792273495095e-06,
+      "loss": 0.6221,
+      "step": 10136
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.7318703307114,
+      "learning_rate": 9.586107313721214e-06,
+      "loss": 0.5262,
+      "step": 10137
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.19203957877898,
+      "learning_rate": 9.584422365718197e-06,
+      "loss": 0.61,
+      "step": 10138
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.729414635116736,
+      "learning_rate": 9.582737429533964e-06,
+      "loss": 0.5727,
+      "step": 10139
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.388957499401354,
+      "learning_rate": 9.581052505216434e-06,
+      "loss": 0.5099,
+      "step": 10140
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.947607268739354,
+      "learning_rate": 9.579367592813526e-06,
+      "loss": 0.5567,
+      "step": 10141
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.280214226668772,
+      "learning_rate": 9.577682692373161e-06,
+      "loss": 0.5128,
+      "step": 10142
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.638860567901208,
+      "learning_rate": 9.575997803943251e-06,
+      "loss": 0.6247,
+      "step": 10143
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.800169965538153,
+      "learning_rate": 9.574312927571717e-06,
+      "loss": 0.6513,
+      "step": 10144
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.66225216641796,
+      "learning_rate": 9.572628063306469e-06,
+      "loss": 0.5887,
+      "step": 10145
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.067352312053234,
+      "learning_rate": 9.570943211195434e-06,
+      "loss": 0.6061,
+      "step": 10146
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.45671744872066,
+      "learning_rate": 9.569258371286523e-06,
+      "loss": 0.5249,
+      "step": 10147
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.880283281002388,
+      "learning_rate": 9.567573543627649e-06,
+      "loss": 0.5792,
+      "step": 10148
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 30.50596970329945,
+      "learning_rate": 9.565888728266732e-06,
+      "loss": 0.6021,
+      "step": 10149
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.981365633669338,
+      "learning_rate": 9.564203925251685e-06,
+      "loss": 0.6029,
+      "step": 10150
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.153166468078364,
+      "learning_rate": 9.562519134630423e-06,
+      "loss": 0.5361,
+      "step": 10151
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.73050006055113,
+      "learning_rate": 9.560834356450864e-06,
+      "loss": 0.614,
+      "step": 10152
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.54892938270337,
+      "learning_rate": 9.559149590760917e-06,
+      "loss": 0.6027,
+      "step": 10153
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.098824253816073,
+      "learning_rate": 9.557464837608496e-06,
+      "loss": 0.6092,
+      "step": 10154
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.676721788558627,
+      "learning_rate": 9.555780097041517e-06,
+      "loss": 0.6006,
+      "step": 10155
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.382317378740684,
+      "learning_rate": 9.554095369107892e-06,
+      "loss": 0.592,
+      "step": 10156
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.335653481955433,
+      "learning_rate": 9.55241065385554e-06,
+      "loss": 0.6207,
+      "step": 10157
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.10824624515318,
+      "learning_rate": 9.55072595133236e-06,
+      "loss": 0.5659,
+      "step": 10158
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.221329543216243,
+      "learning_rate": 9.549041261586273e-06,
+      "loss": 0.5967,
+      "step": 10159
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.922817758430813,
+      "learning_rate": 9.547356584665185e-06,
+      "loss": 0.6424,
+      "step": 10160
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 22.874842735698397,
+      "learning_rate": 9.545671920617018e-06,
+      "loss": 0.5457,
+      "step": 10161
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.082334397319096,
+      "learning_rate": 9.543987269489673e-06,
+      "loss": 0.5322,
+      "step": 10162
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.206464654208467,
+      "learning_rate": 9.542302631331063e-06,
+      "loss": 0.5889,
+      "step": 10163
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.305205657727395,
+      "learning_rate": 9.540618006189096e-06,
+      "loss": 0.5444,
+      "step": 10164
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.114067945740302,
+      "learning_rate": 9.538933394111687e-06,
+      "loss": 0.6235,
+      "step": 10165
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.498871242063327,
+      "learning_rate": 9.537248795146745e-06,
+      "loss": 0.5464,
+      "step": 10166
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.411073713948925,
+      "learning_rate": 9.535564209342172e-06,
+      "loss": 0.5993,
+      "step": 10167
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.395969403728373,
+      "learning_rate": 9.533879636745883e-06,
+      "loss": 0.704,
+      "step": 10168
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.301484906847854,
+      "learning_rate": 9.532195077405784e-06,
+      "loss": 0.5667,
+      "step": 10169
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.219251545566367,
+      "learning_rate": 9.53051053136978e-06,
+      "loss": 0.5725,
+      "step": 10170
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 23.52548695153024,
+      "learning_rate": 9.528825998685788e-06,
+      "loss": 0.6348,
+      "step": 10171
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 28.5215812639141,
+      "learning_rate": 9.527141479401708e-06,
+      "loss": 0.6318,
+      "step": 10172
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.428166581208274,
+      "learning_rate": 9.525456973565443e-06,
+      "loss": 0.5895,
+      "step": 10173
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 22.104338422728773,
+      "learning_rate": 9.523772481224908e-06,
+      "loss": 0.5357,
+      "step": 10174
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 23.521313544488233,
+      "learning_rate": 9.522088002428003e-06,
+      "loss": 0.5764,
+      "step": 10175
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.492518070911466,
+      "learning_rate": 9.520403537222642e-06,
+      "loss": 0.6186,
+      "step": 10176
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.469337373280638,
+      "learning_rate": 9.518719085656716e-06,
+      "loss": 0.5808,
+      "step": 10177
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.01032854613012,
+      "learning_rate": 9.51703464777814e-06,
+      "loss": 0.5827,
+      "step": 10178
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.454706514809335,
+      "learning_rate": 9.515350223634815e-06,
+      "loss": 0.5495,
+      "step": 10179
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.2733111774296,
+      "learning_rate": 9.513665813274647e-06,
+      "loss": 0.58,
+      "step": 10180
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 22.128237953431082,
+      "learning_rate": 9.511981416745542e-06,
+      "loss": 0.5409,
+      "step": 10181
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.02137841432915,
+      "learning_rate": 9.510297034095398e-06,
+      "loss": 0.5669,
+      "step": 10182
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.796168025320355,
+      "learning_rate": 9.508612665372117e-06,
+      "loss": 0.595,
+      "step": 10183
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.964037031263683,
+      "learning_rate": 9.506928310623608e-06,
+      "loss": 0.5848,
+      "step": 10184
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.160202415810765,
+      "learning_rate": 9.505243969897766e-06,
+      "loss": 0.5386,
+      "step": 10185
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.004701400328855,
+      "learning_rate": 9.5035596432425e-06,
+      "loss": 0.5449,
+      "step": 10186
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.47876873289876,
+      "learning_rate": 9.501875330705706e-06,
+      "loss": 0.5268,
+      "step": 10187
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.744135956666508,
+      "learning_rate": 9.500191032335286e-06,
+      "loss": 0.5836,
+      "step": 10188
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.923679271546753,
+      "learning_rate": 9.498506748179137e-06,
+      "loss": 0.5676,
+      "step": 10189
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.577169493325691,
+      "learning_rate": 9.496822478285167e-06,
+      "loss": 0.5193,
+      "step": 10190
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.95495695228826,
+      "learning_rate": 9.495138222701273e-06,
+      "loss": 0.6165,
+      "step": 10191
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 23.45231711942303,
+      "learning_rate": 9.493453981475348e-06,
+      "loss": 0.5866,
+      "step": 10192
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.599289539531185,
+      "learning_rate": 9.491769754655298e-06,
+      "loss": 0.5462,
+      "step": 10193
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.851061260612095,
+      "learning_rate": 9.490085542289016e-06,
+      "loss": 0.6032,
+      "step": 10194
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.497221020491342,
+      "learning_rate": 9.488401344424404e-06,
+      "loss": 0.5381,
+      "step": 10195
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.088175688090733,
+      "learning_rate": 9.486717161109363e-06,
+      "loss": 0.5695,
+      "step": 10196
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.392560985232791,
+      "learning_rate": 9.485032992391782e-06,
+      "loss": 0.531,
+      "step": 10197
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.9133322840861,
+      "learning_rate": 9.48334883831956e-06,
+      "loss": 0.5628,
+      "step": 10198
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.150192333934168,
+      "learning_rate": 9.481664698940598e-06,
+      "loss": 0.5566,
+      "step": 10199
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.434743320288796,
+      "learning_rate": 9.479980574302785e-06,
+      "loss": 0.5734,
+      "step": 10200
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.037543672653364,
+      "learning_rate": 9.478296464454028e-06,
+      "loss": 0.5512,
+      "step": 10201
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 26.748778909427127,
+      "learning_rate": 9.476612369442207e-06,
+      "loss": 0.5786,
+      "step": 10202
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.593833432818476,
+      "learning_rate": 9.474928289315224e-06,
+      "loss": 0.5504,
+      "step": 10203
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.797855319944865,
+      "learning_rate": 9.473244224120974e-06,
+      "loss": 0.6067,
+      "step": 10204
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 27.362343431585103,
+      "learning_rate": 9.471560173907353e-06,
+      "loss": 0.6348,
+      "step": 10205
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 23.205063470470282,
+      "learning_rate": 9.469876138722252e-06,
+      "loss": 0.5843,
+      "step": 10206
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.564455273848642,
+      "learning_rate": 9.46819211861356e-06,
+      "loss": 0.5634,
+      "step": 10207
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.979014592798812,
+      "learning_rate": 9.466508113629174e-06,
+      "loss": 0.5812,
+      "step": 10208
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.5927799162375,
+      "learning_rate": 9.464824123816986e-06,
+      "loss": 0.5749,
+      "step": 10209
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 23.331526192155177,
+      "learning_rate": 9.463140149224888e-06,
+      "loss": 0.6353,
+      "step": 10210
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.18905818317108,
+      "learning_rate": 9.461456189900767e-06,
+      "loss": 0.6544,
+      "step": 10211
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.73927568287242,
+      "learning_rate": 9.459772245892518e-06,
+      "loss": 0.5988,
+      "step": 10212
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.912787478362144,
+      "learning_rate": 9.458088317248032e-06,
+      "loss": 0.5653,
+      "step": 10213
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.484053835122573,
+      "learning_rate": 9.456404404015194e-06,
+      "loss": 0.6067,
+      "step": 10214
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.348222035518095,
+      "learning_rate": 9.454720506241902e-06,
+      "loss": 0.4943,
+      "step": 10215
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.978898508385594,
+      "learning_rate": 9.453036623976036e-06,
+      "loss": 0.59,
+      "step": 10216
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.48585443515506,
+      "learning_rate": 9.451352757265488e-06,
+      "loss": 0.5905,
+      "step": 10217
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.623685701407165,
+      "learning_rate": 9.449668906158149e-06,
+      "loss": 0.5601,
+      "step": 10218
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.59599014381848,
+      "learning_rate": 9.4479850707019e-06,
+      "loss": 0.5843,
+      "step": 10219
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.582674865188112,
+      "learning_rate": 9.446301250944641e-06,
+      "loss": 0.5157,
+      "step": 10220
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 26.282548043426566,
+      "learning_rate": 9.444617446934244e-06,
+      "loss": 0.5368,
+      "step": 10221
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.083932482277575,
+      "learning_rate": 9.442933658718603e-06,
+      "loss": 0.5931,
+      "step": 10222
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.950197334194463,
+      "learning_rate": 9.4412498863456e-06,
+      "loss": 0.5582,
+      "step": 10223
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.458807319213445,
+      "learning_rate": 9.439566129863126e-06,
+      "loss": 0.6908,
+      "step": 10224
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 19.418340543595967,
+      "learning_rate": 9.437882389319067e-06,
+      "loss": 0.5632,
+      "step": 10225
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.601004524550662,
+      "learning_rate": 9.436198664761302e-06,
+      "loss": 0.5978,
+      "step": 10226
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.946124442638643,
+      "learning_rate": 9.434514956237714e-06,
+      "loss": 0.5696,
+      "step": 10227
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.578052067031987,
+      "learning_rate": 9.43283126379619e-06,
+      "loss": 0.5703,
+      "step": 10228
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.660376256129847,
+      "learning_rate": 9.431147587484614e-06,
+      "loss": 0.5427,
+      "step": 10229
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.171986702980117,
+      "learning_rate": 9.429463927350872e-06,
+      "loss": 0.5693,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.131356021402402,
+      "learning_rate": 9.427780283442838e-06,
+      "loss": 0.6187,
+      "step": 10231
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.673093993805384,
+      "learning_rate": 9.4260966558084e-06,
+      "loss": 0.6402,
+      "step": 10232
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.810651001117048,
+      "learning_rate": 9.424413044495435e-06,
+      "loss": 0.6816,
+      "step": 10233
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.537029502815066,
+      "learning_rate": 9.422729449551828e-06,
+      "loss": 0.5753,
+      "step": 10234
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.174861105675564,
+      "learning_rate": 9.42104587102546e-06,
+      "loss": 0.5041,
+      "step": 10235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 24.18466711211708,
+      "learning_rate": 9.419362308964203e-06,
+      "loss": 0.6166,
+      "step": 10236
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.901788555175294,
+      "learning_rate": 9.417678763415948e-06,
+      "loss": 0.5414,
+      "step": 10237
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.198155662471024,
+      "learning_rate": 9.415995234428563e-06,
+      "loss": 0.6062,
+      "step": 10238
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.553335830466292,
+      "learning_rate": 9.414311722049935e-06,
+      "loss": 0.5701,
+      "step": 10239
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 19.9277178868868,
+      "learning_rate": 9.41262822632794e-06,
+      "loss": 0.6295,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 26.466209562185433,
+      "learning_rate": 9.410944747310454e-06,
+      "loss": 0.5657,
+      "step": 10241
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.962371337647367,
+      "learning_rate": 9.409261285045352e-06,
+      "loss": 0.5876,
+      "step": 10242
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 19.82596976378527,
+      "learning_rate": 9.407577839580516e-06,
+      "loss": 0.5939,
+      "step": 10243
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.59301528553886,
+      "learning_rate": 9.405894410963817e-06,
+      "loss": 0.622,
+      "step": 10244
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.98443494060237,
+      "learning_rate": 9.404210999243141e-06,
+      "loss": 0.6171,
+      "step": 10245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.873800868238796,
+      "learning_rate": 9.402527604466347e-06,
+      "loss": 0.5381,
+      "step": 10246
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.757366974704107,
+      "learning_rate": 9.400844226681324e-06,
+      "loss": 0.5755,
+      "step": 10247
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.078581079742847,
+      "learning_rate": 9.399160865935936e-06,
+      "loss": 0.5406,
+      "step": 10248
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.536315443335315,
+      "learning_rate": 9.397477522278064e-06,
+      "loss": 0.5606,
+      "step": 10249
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.518486491893302,
+      "learning_rate": 9.395794195755581e-06,
+      "loss": 0.6363,
+      "step": 10250
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.10358513276388,
+      "learning_rate": 9.394110886416355e-06,
+      "loss": 0.6003,
+      "step": 10251
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.416257672963312,
+      "learning_rate": 9.39242759430826e-06,
+      "loss": 0.526,
+      "step": 10252
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 24.113152665994722,
+      "learning_rate": 9.390744319479171e-06,
+      "loss": 0.6683,
+      "step": 10253
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.690606755692588,
+      "learning_rate": 9.389061061976958e-06,
+      "loss": 0.5294,
+      "step": 10254
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.46687030432764,
+      "learning_rate": 9.387377821849489e-06,
+      "loss": 0.7158,
+      "step": 10255
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 28.0977567736958,
+      "learning_rate": 9.385694599144636e-06,
+      "loss": 0.6548,
+      "step": 10256
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.389204608186773,
+      "learning_rate": 9.38401139391027e-06,
+      "loss": 0.5865,
+      "step": 10257
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.95918528178506,
+      "learning_rate": 9.382328206194259e-06,
+      "loss": 0.6258,
+      "step": 10258
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.568563356356565,
+      "learning_rate": 9.380645036044473e-06,
+      "loss": 0.502,
+      "step": 10259
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.88837911135321,
+      "learning_rate": 9.37896188350878e-06,
+      "loss": 0.6104,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.81118628233423,
+      "learning_rate": 9.377278748635046e-06,
+      "loss": 0.6765,
+      "step": 10261
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 19.87957169964455,
+      "learning_rate": 9.375595631471143e-06,
+      "loss": 0.578,
+      "step": 10262
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.256536082207402,
+      "learning_rate": 9.373912532064931e-06,
+      "loss": 0.5445,
+      "step": 10263
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.100147125699564,
+      "learning_rate": 9.372229450464287e-06,
+      "loss": 0.579,
+      "step": 10264
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.915311870614163,
+      "learning_rate": 9.370546386717065e-06,
+      "loss": 0.5259,
+      "step": 10265
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.69585569337966,
+      "learning_rate": 9.368863340871137e-06,
+      "loss": 0.5983,
+      "step": 10266
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.1828908723917,
+      "learning_rate": 9.367180312974364e-06,
+      "loss": 0.5635,
+      "step": 10267
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.697943695885563,
+      "learning_rate": 9.365497303074615e-06,
+      "loss": 0.5772,
+      "step": 10268
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 24.679322525325297,
+      "learning_rate": 9.363814311219754e-06,
+      "loss": 0.6061,
+      "step": 10269
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.954191442661077,
+      "learning_rate": 9.36213133745764e-06,
+      "loss": 0.5482,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.922208037579464,
+      "learning_rate": 9.360448381836134e-06,
+      "loss": 0.5827,
+      "step": 10271
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.75693757472985,
+      "learning_rate": 9.358765444403107e-06,
+      "loss": 0.7006,
+      "step": 10272
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.04679844827641,
+      "learning_rate": 9.35708252520641e-06,
+      "loss": 0.548,
+      "step": 10273
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.92478961980648,
+      "learning_rate": 9.355399624293917e-06,
+      "loss": 0.5839,
+      "step": 10274
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.703282960371414,
+      "learning_rate": 9.353716741713477e-06,
+      "loss": 0.6449,
+      "step": 10275
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.442910593222667,
+      "learning_rate": 9.352033877512957e-06,
+      "loss": 0.5102,
+      "step": 10276
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.887638379641604,
+      "learning_rate": 9.35035103174021e-06,
+      "loss": 0.6407,
+      "step": 10277
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.359682947376314,
+      "learning_rate": 9.348668204443103e-06,
+      "loss": 0.5838,
+      "step": 10278
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.313177205712227,
+      "learning_rate": 9.346985395669494e-06,
+      "loss": 0.5284,
+      "step": 10279
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.927042933145184,
+      "learning_rate": 9.345302605467232e-06,
+      "loss": 0.6332,
+      "step": 10280
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.39632241741886,
+      "learning_rate": 9.343619833884186e-06,
+      "loss": 0.5846,
+      "step": 10281
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.727835286498337,
+      "learning_rate": 9.341937080968204e-06,
+      "loss": 0.5179,
+      "step": 10282
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.1071667395017,
+      "learning_rate": 9.340254346767149e-06,
+      "loss": 0.6129,
+      "step": 10283
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.245811454546278,
+      "learning_rate": 9.338571631328877e-06,
+      "loss": 0.5398,
+      "step": 10284
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.270080420479953,
+      "learning_rate": 9.336888934701238e-06,
+      "loss": 0.6099,
+      "step": 10285
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.16792531977402,
+      "learning_rate": 9.335206256932088e-06,
+      "loss": 0.6419,
+      "step": 10286
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.657754697808723,
+      "learning_rate": 9.333523598069286e-06,
+      "loss": 0.5355,
+      "step": 10287
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.473529822096957,
+      "learning_rate": 9.331840958160679e-06,
+      "loss": 0.6247,
+      "step": 10288
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.271632504055837,
+      "learning_rate": 9.330158337254131e-06,
+      "loss": 0.4567,
+      "step": 10289
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.09334661726697,
+      "learning_rate": 9.328475735397483e-06,
+      "loss": 0.6163,
+      "step": 10290
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.25708516173471,
+      "learning_rate": 9.326793152638594e-06,
+      "loss": 0.5547,
+      "step": 10291
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.876319783460378,
+      "learning_rate": 9.325110589025311e-06,
+      "loss": 0.6222,
+      "step": 10292
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.7870769791939,
+      "learning_rate": 9.323428044605491e-06,
+      "loss": 0.5774,
+      "step": 10293
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.43082015578266,
+      "learning_rate": 9.321745519426984e-06,
+      "loss": 0.6114,
+      "step": 10294
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.380068503266173,
+      "learning_rate": 9.320063013537634e-06,
+      "loss": 0.5857,
+      "step": 10295
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.748713204256827,
+      "learning_rate": 9.318380526985293e-06,
+      "loss": 0.6269,
+      "step": 10296
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 22.65661006641371,
+      "learning_rate": 9.316698059817814e-06,
+      "loss": 0.5935,
+      "step": 10297
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.79811898427557,
+      "learning_rate": 9.315015612083039e-06,
+      "loss": 0.6588,
+      "step": 10298
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.167883736073115,
+      "learning_rate": 9.313333183828824e-06,
+      "loss": 0.6254,
+      "step": 10299
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.774981369258086,
+      "learning_rate": 9.31165077510301e-06,
+      "loss": 0.5082,
+      "step": 10300
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.732890730213768,
+      "learning_rate": 9.309968385953445e-06,
+      "loss": 0.6986,
+      "step": 10301
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.58489678773787,
+      "learning_rate": 9.308286016427974e-06,
+      "loss": 0.5383,
+      "step": 10302
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.716575974740497,
+      "learning_rate": 9.306603666574446e-06,
+      "loss": 0.5544,
+      "step": 10303
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.021266107854519,
+      "learning_rate": 9.304921336440704e-06,
+      "loss": 0.5209,
+      "step": 10304
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.063998376866504,
+      "learning_rate": 9.303239026074589e-06,
+      "loss": 0.5744,
+      "step": 10305
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.43306292145674,
+      "learning_rate": 9.301556735523952e-06,
+      "loss": 0.6,
+      "step": 10306
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.218778917099094,
+      "learning_rate": 9.29987446483663e-06,
+      "loss": 0.5853,
+      "step": 10307
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.283182325207553,
+      "learning_rate": 9.298192214060476e-06,
+      "loss": 0.5186,
+      "step": 10308
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 26.22087330342056,
+      "learning_rate": 9.296509983243319e-06,
+      "loss": 0.6404,
+      "step": 10309
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.1578283098604,
+      "learning_rate": 9.294827772433006e-06,
+      "loss": 0.5412,
+      "step": 10310
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.916514023106846,
+      "learning_rate": 9.293145581677378e-06,
+      "loss": 0.5919,
+      "step": 10311
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.71944951867443,
+      "learning_rate": 9.291463411024279e-06,
+      "loss": 0.648,
+      "step": 10312
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.223280638580093,
+      "learning_rate": 9.289781260521548e-06,
+      "loss": 0.5693,
+      "step": 10313
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.29924722303305,
+      "learning_rate": 9.288099130217018e-06,
+      "loss": 0.5729,
+      "step": 10314
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.990288877855676,
+      "learning_rate": 9.286417020158533e-06,
+      "loss": 0.6412,
+      "step": 10315
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.722236580115561,
+      "learning_rate": 9.284734930393931e-06,
+      "loss": 0.5158,
+      "step": 10316
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 27.177101535810518,
+      "learning_rate": 9.283052860971049e-06,
+      "loss": 0.6303,
+      "step": 10317
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.53566191274117,
+      "learning_rate": 9.281370811937726e-06,
+      "loss": 0.5773,
+      "step": 10318
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 24.816391767896608,
+      "learning_rate": 9.279688783341796e-06,
+      "loss": 0.5675,
+      "step": 10319
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.432124444753892,
+      "learning_rate": 9.278006775231097e-06,
+      "loss": 0.6174,
+      "step": 10320
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.286689421389843,
+      "learning_rate": 9.276324787653458e-06,
+      "loss": 0.5627,
+      "step": 10321
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 26.682728795398713,
+      "learning_rate": 9.274642820656722e-06,
+      "loss": 0.6473,
+      "step": 10322
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 23.291114155622914,
+      "learning_rate": 9.272960874288722e-06,
+      "loss": 0.6316,
+      "step": 10323
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.73245335851438,
+      "learning_rate": 9.271278948597285e-06,
+      "loss": 0.5468,
+      "step": 10324
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.81599449829306,
+      "learning_rate": 9.269597043630252e-06,
+      "loss": 0.6296,
+      "step": 10325
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.962175262337958,
+      "learning_rate": 9.267915159435447e-06,
+      "loss": 0.5747,
+      "step": 10326
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.38652485369521,
+      "learning_rate": 9.266233296060712e-06,
+      "loss": 0.6067,
+      "step": 10327
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.194172448009406,
+      "learning_rate": 9.264551453553872e-06,
+      "loss": 0.6119,
+      "step": 10328
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.653573340854845,
+      "learning_rate": 9.262869631962758e-06,
+      "loss": 0.5876,
+      "step": 10329
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.48438943970978,
+      "learning_rate": 9.261187831335197e-06,
+      "loss": 0.4992,
+      "step": 10330
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.082981925122127,
+      "learning_rate": 9.259506051719025e-06,
+      "loss": 0.6096,
+      "step": 10331
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.502835289291347,
+      "learning_rate": 9.257824293162066e-06,
+      "loss": 0.5047,
+      "step": 10332
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.420905425209333,
+      "learning_rate": 9.256142555712156e-06,
+      "loss": 0.6344,
+      "step": 10333
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.50588945073572,
+      "learning_rate": 9.25446083941711e-06,
+      "loss": 0.6006,
+      "step": 10334
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.39603997529431,
+      "learning_rate": 9.252779144324763e-06,
+      "loss": 0.5787,
+      "step": 10335
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 22.832102139952735,
+      "learning_rate": 9.251097470482938e-06,
+      "loss": 0.5998,
+      "step": 10336
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 26.953039387000295,
+      "learning_rate": 9.249415817939465e-06,
+      "loss": 0.5463,
+      "step": 10337
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 27.634023958526196,
+      "learning_rate": 9.24773418674217e-06,
+      "loss": 0.6105,
+      "step": 10338
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.144275607637432,
+      "learning_rate": 9.246052576938874e-06,
+      "loss": 0.6086,
+      "step": 10339
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.076926761811993,
+      "learning_rate": 9.244370988577396e-06,
+      "loss": 0.593,
+      "step": 10340
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.39934634324917,
+      "learning_rate": 9.24268942170557e-06,
+      "loss": 0.5132,
+      "step": 10341
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.67235096524518,
+      "learning_rate": 9.24100787637121e-06,
+      "loss": 0.5689,
+      "step": 10342
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.510444718569524,
+      "learning_rate": 9.239326352622146e-06,
+      "loss": 0.5262,
+      "step": 10343
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 26.69376720592254,
+      "learning_rate": 9.237644850506196e-06,
+      "loss": 0.6201,
+      "step": 10344
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.471294593275754,
+      "learning_rate": 9.235963370071176e-06,
+      "loss": 0.6485,
+      "step": 10345
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 29.448793617415234,
+      "learning_rate": 9.234281911364914e-06,
+      "loss": 0.6395,
+      "step": 10346
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.550272676024864,
+      "learning_rate": 9.232600474435226e-06,
+      "loss": 0.5727,
+      "step": 10347
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 29.68396072773288,
+      "learning_rate": 9.230919059329934e-06,
+      "loss": 0.6261,
+      "step": 10348
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.850594901922985,
+      "learning_rate": 9.229237666096848e-06,
+      "loss": 0.5576,
+      "step": 10349
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.38601101971303,
+      "learning_rate": 9.227556294783797e-06,
+      "loss": 0.5621,
+      "step": 10350
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.363081351609598,
+      "learning_rate": 9.22587494543859e-06,
+      "loss": 0.6408,
+      "step": 10351
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 28.623481177396595,
+      "learning_rate": 9.224193618109052e-06,
+      "loss": 0.5417,
+      "step": 10352
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.912896584319487,
+      "learning_rate": 9.222512312842988e-06,
+      "loss": 0.5638,
+      "step": 10353
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.642217769632843,
+      "learning_rate": 9.220831029688222e-06,
+      "loss": 0.6226,
+      "step": 10354
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.777555452694852,
+      "learning_rate": 9.219149768692563e-06,
+      "loss": 0.5332,
+      "step": 10355
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.636761167687975,
+      "learning_rate": 9.217468529903829e-06,
+      "loss": 0.596,
+      "step": 10356
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.28010750823603,
+      "learning_rate": 9.215787313369836e-06,
+      "loss": 0.576,
+      "step": 10357
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.846388950768564,
+      "learning_rate": 9.214106119138388e-06,
+      "loss": 0.5896,
+      "step": 10358
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.684558197454713,
+      "learning_rate": 9.212424947257302e-06,
+      "loss": 0.5528,
+      "step": 10359
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.25708366017856,
+      "learning_rate": 9.210743797774392e-06,
+      "loss": 0.559,
+      "step": 10360
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 26.2853679756167,
+      "learning_rate": 9.209062670737465e-06,
+      "loss": 0.5536,
+      "step": 10361
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.790140533854814,
+      "learning_rate": 9.207381566194336e-06,
+      "loss": 0.4897,
+      "step": 10362
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 30.496283108795332,
+      "learning_rate": 9.20570048419281e-06,
+      "loss": 0.5982,
+      "step": 10363
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.592540602423043,
+      "learning_rate": 9.2040194247807e-06,
+      "loss": 0.6106,
+      "step": 10364
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.310805134488394,
+      "learning_rate": 9.202338388005807e-06,
+      "loss": 0.5794,
+      "step": 10365
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.701146314865305,
+      "learning_rate": 9.200657373915946e-06,
+      "loss": 0.5905,
+      "step": 10366
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.291560195112899,
+      "learning_rate": 9.198976382558924e-06,
+      "loss": 0.5828,
+      "step": 10367
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.652736740745688,
+      "learning_rate": 9.19729541398254e-06,
+      "loss": 0.4989,
+      "step": 10368
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.452074973242688,
+      "learning_rate": 9.19561446823461e-06,
+      "loss": 0.5452,
+      "step": 10369
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.520115284540324,
+      "learning_rate": 9.19393354536293e-06,
+      "loss": 0.6126,
+      "step": 10370
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.569610225599869,
+      "learning_rate": 9.19225264541531e-06,
+      "loss": 0.4944,
+      "step": 10371
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.269979223842032,
+      "learning_rate": 9.190571768439554e-06,
+      "loss": 0.6588,
+      "step": 10372
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.070464657766006,
+      "learning_rate": 9.188890914483462e-06,
+      "loss": 0.5684,
+      "step": 10373
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.193578442888327,
+      "learning_rate": 9.187210083594835e-06,
+      "loss": 0.4986,
+      "step": 10374
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.756416496127612,
+      "learning_rate": 9.185529275821481e-06,
+      "loss": 0.4903,
+      "step": 10375
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.45561857070713,
+      "learning_rate": 9.183848491211196e-06,
+      "loss": 0.5849,
+      "step": 10376
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.188243915370844,
+      "learning_rate": 9.182167729811788e-06,
+      "loss": 0.5845,
+      "step": 10377
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.224510569736154,
+      "learning_rate": 9.180486991671046e-06,
+      "loss": 0.4929,
+      "step": 10378
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.828061450813465,
+      "learning_rate": 9.178806276836777e-06,
+      "loss": 0.5227,
+      "step": 10379
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.600475383227142,
+      "learning_rate": 9.177125585356773e-06,
+      "loss": 0.4719,
+      "step": 10380
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.9438112967308,
+      "learning_rate": 9.175444917278839e-06,
+      "loss": 0.5348,
+      "step": 10381
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 24.027224423666176,
+      "learning_rate": 9.173764272650771e-06,
+      "loss": 0.5896,
+      "step": 10382
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 24.211070452728517,
+      "learning_rate": 9.172083651520362e-06,
+      "loss": 0.683,
+      "step": 10383
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.378095623142716,
+      "learning_rate": 9.170403053935408e-06,
+      "loss": 0.5744,
+      "step": 10384
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.88163814186779,
+      "learning_rate": 9.168722479943707e-06,
+      "loss": 0.5265,
+      "step": 10385
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.597047180491167,
+      "learning_rate": 9.167041929593051e-06,
+      "loss": 0.5158,
+      "step": 10386
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.541592896320616,
+      "learning_rate": 9.16536140293124e-06,
+      "loss": 0.5669,
+      "step": 10387
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.123741828790738,
+      "learning_rate": 9.16368090000606e-06,
+      "loss": 0.6116,
+      "step": 10388
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.70749623638826,
+      "learning_rate": 9.162000420865303e-06,
+      "loss": 0.5384,
+      "step": 10389
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.061476114470807,
+      "learning_rate": 9.160319965556765e-06,
+      "loss": 0.5703,
+      "step": 10390
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.41892315738222,
+      "learning_rate": 9.158639534128239e-06,
+      "loss": 0.5311,
+      "step": 10391
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.964652011410568,
+      "learning_rate": 9.156959126627512e-06,
+      "loss": 0.5812,
+      "step": 10392
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.382889950618612,
+      "learning_rate": 9.155278743102372e-06,
+      "loss": 0.6956,
+      "step": 10393
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.852998802875378,
+      "learning_rate": 9.15359838360061e-06,
+      "loss": 0.5987,
+      "step": 10394
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 24.414755959389733,
+      "learning_rate": 9.151918048170015e-06,
+      "loss": 0.5732,
+      "step": 10395
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.29753428997181,
+      "learning_rate": 9.150237736858376e-06,
+      "loss": 0.5638,
+      "step": 10396
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.442121723980062,
+      "learning_rate": 9.148557449713481e-06,
+      "loss": 0.5895,
+      "step": 10397
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.856061955456987,
+      "learning_rate": 9.146877186783112e-06,
+      "loss": 0.5639,
+      "step": 10398
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.664234970078237,
+      "learning_rate": 9.145196948115053e-06,
+      "loss": 0.5464,
+      "step": 10399
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.588811622575363,
+      "learning_rate": 9.143516733757097e-06,
+      "loss": 0.5261,
+      "step": 10400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.106962277375844,
+      "learning_rate": 9.141836543757025e-06,
+      "loss": 0.6426,
+      "step": 10401
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 25.18564480262938,
+      "learning_rate": 9.140156378162619e-06,
+      "loss": 0.6504,
+      "step": 10402
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.423011503722876,
+      "learning_rate": 9.138476237021659e-06,
+      "loss": 0.5862,
+      "step": 10403
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 25.163164573407695,
+      "learning_rate": 9.136796120381933e-06,
+      "loss": 0.6266,
+      "step": 10404
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 22.171129386030675,
+      "learning_rate": 9.13511602829122e-06,
+      "loss": 0.6082,
+      "step": 10405
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.648151915025393,
+      "learning_rate": 9.133435960797304e-06,
+      "loss": 0.5139,
+      "step": 10406
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 34.61505707495903,
+      "learning_rate": 9.131755917947958e-06,
+      "loss": 0.6058,
+      "step": 10407
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 23.20616615479577,
+      "learning_rate": 9.13007589979097e-06,
+      "loss": 0.6402,
+      "step": 10408
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.013305828952747,
+      "learning_rate": 9.12839590637411e-06,
+      "loss": 0.5154,
+      "step": 10409
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.371907195460599,
+      "learning_rate": 9.126715937745163e-06,
+      "loss": 0.585,
+      "step": 10410
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.81631373047951,
+      "learning_rate": 9.125035993951907e-06,
+      "loss": 0.6006,
+      "step": 10411
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.413609229850618,
+      "learning_rate": 9.12335607504211e-06,
+      "loss": 0.581,
+      "step": 10412
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.397150209266,
+      "learning_rate": 9.121676181063556e-06,
+      "loss": 0.5582,
+      "step": 10413
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.1919420724341,
+      "learning_rate": 9.119996312064014e-06,
+      "loss": 0.5781,
+      "step": 10414
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.907842755811654,
+      "learning_rate": 9.118316468091267e-06,
+      "loss": 0.525,
+      "step": 10415
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.881090241103244,
+      "learning_rate": 9.116636649193085e-06,
+      "loss": 0.5209,
+      "step": 10416
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.368469373418282,
+      "learning_rate": 9.114956855417236e-06,
+      "loss": 0.5508,
+      "step": 10417
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 24.752123591082036,
+      "learning_rate": 9.113277086811495e-06,
+      "loss": 0.503,
+      "step": 10418
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.305718111075542,
+      "learning_rate": 9.11159734342364e-06,
+      "loss": 0.5498,
+      "step": 10419
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.812254986730656,
+      "learning_rate": 9.109917625301432e-06,
+      "loss": 0.5825,
+      "step": 10420
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.583515730608436,
+      "learning_rate": 9.108237932492652e-06,
+      "loss": 0.5404,
+      "step": 10421
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.505171441865173,
+      "learning_rate": 9.106558265045058e-06,
+      "loss": 0.5211,
+      "step": 10422
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 22.580835701071518,
+      "learning_rate": 9.104878623006427e-06,
+      "loss": 0.583,
+      "step": 10423
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.24192872894691,
+      "learning_rate": 9.103199006424521e-06,
+      "loss": 0.5479,
+      "step": 10424
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.851000703934886,
+      "learning_rate": 9.101519415347113e-06,
+      "loss": 0.5859,
+      "step": 10425
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.38091153155122,
+      "learning_rate": 9.099839849821971e-06,
+      "loss": 0.5956,
+      "step": 10426
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 62.709451744069504,
+      "learning_rate": 9.098160309896852e-06,
+      "loss": 0.5458,
+      "step": 10427
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.208093056625394,
+      "learning_rate": 9.096480795619527e-06,
+      "loss": 0.5215,
+      "step": 10428
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.630082533037218,
+      "learning_rate": 9.09480130703776e-06,
+      "loss": 0.5565,
+      "step": 10429
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 27.96890927061625,
+      "learning_rate": 9.093121844199312e-06,
+      "loss": 0.6058,
+      "step": 10430
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 22.805145402828263,
+      "learning_rate": 9.091442407151955e-06,
+      "loss": 0.5889,
+      "step": 10431
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 23.358026581697406,
+      "learning_rate": 9.08976299594344e-06,
+      "loss": 0.5732,
+      "step": 10432
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.60574966927669,
+      "learning_rate": 9.088083610621532e-06,
+      "loss": 0.5694,
+      "step": 10433
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 27.29064929366573,
+      "learning_rate": 9.086404251233993e-06,
+      "loss": 0.5882,
+      "step": 10434
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.62426147660861,
+      "learning_rate": 9.084724917828585e-06,
+      "loss": 0.5358,
+      "step": 10435
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.684225797940414,
+      "learning_rate": 9.083045610453065e-06,
+      "loss": 0.5923,
+      "step": 10436
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.8363555208507,
+      "learning_rate": 9.081366329155188e-06,
+      "loss": 0.6223,
+      "step": 10437
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.02217791405987,
+      "learning_rate": 9.079687073982719e-06,
+      "loss": 0.5745,
+      "step": 10438
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.561908722139812,
+      "learning_rate": 9.078007844983406e-06,
+      "loss": 0.6294,
+      "step": 10439
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 29.790703838079608,
+      "learning_rate": 9.076328642205015e-06,
+      "loss": 0.6192,
+      "step": 10440
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.040004356222106,
+      "learning_rate": 9.074649465695298e-06,
+      "loss": 0.5477,
+      "step": 10441
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.168233502881762,
+      "learning_rate": 9.072970315502008e-06,
+      "loss": 0.6195,
+      "step": 10442
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.03000217490477,
+      "learning_rate": 9.071291191672896e-06,
+      "loss": 0.621,
+      "step": 10443
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 25.93311040169013,
+      "learning_rate": 9.069612094255722e-06,
+      "loss": 0.588,
+      "step": 10444
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.210828988381525,
+      "learning_rate": 9.067933023298234e-06,
+      "loss": 0.5322,
+      "step": 10445
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.698145656581431,
+      "learning_rate": 9.066253978848192e-06,
+      "loss": 0.5393,
+      "step": 10446
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.219936736948377,
+      "learning_rate": 9.064574960953334e-06,
+      "loss": 0.5373,
+      "step": 10447
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 30.702779209820246,
+      "learning_rate": 9.06289596966142e-06,
+      "loss": 0.6247,
+      "step": 10448
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 27.47793611840959,
+      "learning_rate": 9.061217005020194e-06,
+      "loss": 0.5536,
+      "step": 10449
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 22.542843646619687,
+      "learning_rate": 9.05953806707741e-06,
+      "loss": 0.4856,
+      "step": 10450
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 29.10338815121388,
+      "learning_rate": 9.057859155880811e-06,
+      "loss": 0.5466,
+      "step": 10451
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.75676527896899,
+      "learning_rate": 9.05618027147815e-06,
+      "loss": 0.5425,
+      "step": 10452
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.432956507410985,
+      "learning_rate": 9.054501413917163e-06,
+      "loss": 0.5824,
+      "step": 10453
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 23.85335939962612,
+      "learning_rate": 9.052822583245608e-06,
+      "loss": 0.5536,
+      "step": 10454
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.364184912418985,
+      "learning_rate": 9.051143779511226e-06,
+      "loss": 0.5782,
+      "step": 10455
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 26.731295591658434,
+      "learning_rate": 9.049465002761756e-06,
+      "loss": 0.6328,
+      "step": 10456
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.87272332962382,
+      "learning_rate": 9.047786253044945e-06,
+      "loss": 0.5642,
+      "step": 10457
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 24.542904040311868,
+      "learning_rate": 9.046107530408537e-06,
+      "loss": 0.5402,
+      "step": 10458
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.292259341223126,
+      "learning_rate": 9.044428834900274e-06,
+      "loss": 0.5507,
+      "step": 10459
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.090231742617792,
+      "learning_rate": 9.042750166567898e-06,
+      "loss": 0.6195,
+      "step": 10460
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.310976359907976,
+      "learning_rate": 9.041071525459145e-06,
+      "loss": 0.564,
+      "step": 10461
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.651425541524818,
+      "learning_rate": 9.039392911621754e-06,
+      "loss": 0.5527,
+      "step": 10462
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.544464442729812,
+      "learning_rate": 9.03771432510347e-06,
+      "loss": 0.5885,
+      "step": 10463
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.909516456475053,
+      "learning_rate": 9.036035765952023e-06,
+      "loss": 0.5726,
+      "step": 10464
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.64231401738984,
+      "learning_rate": 9.034357234215164e-06,
+      "loss": 0.5706,
+      "step": 10465
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.89626764825909,
+      "learning_rate": 9.032678729940614e-06,
+      "loss": 0.5832,
+      "step": 10466
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 24.214936565917274,
+      "learning_rate": 9.031000253176117e-06,
+      "loss": 0.5661,
+      "step": 10467
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 24.995543725587893,
+      "learning_rate": 9.029321803969402e-06,
+      "loss": 0.6112,
+      "step": 10468
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 21.20016321517818,
+      "learning_rate": 9.02764338236821e-06,
+      "loss": 0.5867,
+      "step": 10469
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.50360846075877,
+      "learning_rate": 9.025964988420274e-06,
+      "loss": 0.6435,
+      "step": 10470
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 17.068692340826775,
+      "learning_rate": 9.024286622173323e-06,
+      "loss": 0.6163,
+      "step": 10471
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.56694383205472,
+      "learning_rate": 9.022608283675086e-06,
+      "loss": 0.6321,
+      "step": 10472
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.86294535479617,
+      "learning_rate": 9.0209299729733e-06,
+      "loss": 0.592,
+      "step": 10473
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.245385621951677,
+      "learning_rate": 9.019251690115692e-06,
+      "loss": 0.6287,
+      "step": 10474
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.085652060178237,
+      "learning_rate": 9.017573435149994e-06,
+      "loss": 0.606,
+      "step": 10475
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.030911748127032,
+      "learning_rate": 9.015895208123933e-06,
+      "loss": 0.6451,
+      "step": 10476
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 21.749620147928773,
+      "learning_rate": 9.014217009085232e-06,
+      "loss": 0.622,
+      "step": 10477
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 17.241953794194444,
+      "learning_rate": 9.012538838081626e-06,
+      "loss": 0.6088,
+      "step": 10478
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.791576828274543,
+      "learning_rate": 9.010860695160839e-06,
+      "loss": 0.5898,
+      "step": 10479
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.109710640467018,
+      "learning_rate": 9.009182580370596e-06,
+      "loss": 0.5762,
+      "step": 10480
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.69851413024346,
+      "learning_rate": 9.007504493758616e-06,
+      "loss": 0.502,
+      "step": 10481
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.708258823029182,
+      "learning_rate": 9.005826435372631e-06,
+      "loss": 0.5825,
+      "step": 10482
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.349294069837086,
+      "learning_rate": 9.004148405260358e-06,
+      "loss": 0.5218,
+      "step": 10483
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.486705073880735,
+      "learning_rate": 9.002470403469525e-06,
+      "loss": 0.5413,
+      "step": 10484
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 19.943930851873372,
+      "learning_rate": 9.000792430047852e-06,
+      "loss": 0.6559,
+      "step": 10485
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 17.393029896664174,
+      "learning_rate": 8.999114485043053e-06,
+      "loss": 0.6228,
+      "step": 10486
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.855324695579824,
+      "learning_rate": 8.997436568502852e-06,
+      "loss": 0.5897,
+      "step": 10487
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.82854908916895,
+      "learning_rate": 8.995758680474972e-06,
+      "loss": 0.6062,
+      "step": 10488
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 30.91201636922907,
+      "learning_rate": 8.994080821007124e-06,
+      "loss": 0.5471,
+      "step": 10489
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.952062258832903,
+      "learning_rate": 8.992402990147035e-06,
+      "loss": 0.596,
+      "step": 10490
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.700295361483954,
+      "learning_rate": 8.99072518794241e-06,
+      "loss": 0.5504,
+      "step": 10491
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 26.66921100374453,
+      "learning_rate": 8.989047414440973e-06,
+      "loss": 0.6041,
+      "step": 10492
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 23.783490233980313,
+      "learning_rate": 8.987369669690433e-06,
+      "loss": 0.6787,
+      "step": 10493
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 22.90884712817259,
+      "learning_rate": 8.98569195373851e-06,
+      "loss": 0.5389,
+      "step": 10494
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.559373820232615,
+      "learning_rate": 8.984014266632916e-06,
+      "loss": 0.5523,
+      "step": 10495
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 22.224304031512577,
+      "learning_rate": 8.98233660842136e-06,
+      "loss": 0.5691,
+      "step": 10496
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.789810081162711,
+      "learning_rate": 8.980658979151556e-06,
+      "loss": 0.6064,
+      "step": 10497
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.598343913717517,
+      "learning_rate": 8.978981378871213e-06,
+      "loss": 0.5818,
+      "step": 10498
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.935223357032948,
+      "learning_rate": 8.977303807628046e-06,
+      "loss": 0.5364,
+      "step": 10499
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.71346149770786,
+      "learning_rate": 8.975626265469755e-06,
+      "loss": 0.5531,
+      "step": 10500
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.912235980585484,
+      "learning_rate": 8.973948752444057e-06,
+      "loss": 0.5786,
+      "step": 10501
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.822584951635,
+      "learning_rate": 8.972271268598653e-06,
+      "loss": 0.5186,
+      "step": 10502
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.646987580410274,
+      "learning_rate": 8.970593813981257e-06,
+      "loss": 0.527,
+      "step": 10503
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 21.37215044870707,
+      "learning_rate": 8.96891638863957e-06,
+      "loss": 0.5504,
+      "step": 10504
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.32014397053585,
+      "learning_rate": 8.967238992621297e-06,
+      "loss": 0.5552,
+      "step": 10505
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.506098366750244,
+      "learning_rate": 8.965561625974142e-06,
+      "loss": 0.6567,
+      "step": 10506
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 22.63123209883333,
+      "learning_rate": 8.96388428874581e-06,
+      "loss": 0.5346,
+      "step": 10507
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 29.01387683201534,
+      "learning_rate": 8.962206980984001e-06,
+      "loss": 0.5601,
+      "step": 10508
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.579383760185625,
+      "learning_rate": 8.960529702736425e-06,
+      "loss": 0.6136,
+      "step": 10509
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 33.94865908966314,
+      "learning_rate": 8.95885245405077e-06,
+      "loss": 0.6253,
+      "step": 10510
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.326960453414634,
+      "learning_rate": 8.957175234974744e-06,
+      "loss": 0.5111,
+      "step": 10511
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.295587042864588,
+      "learning_rate": 8.955498045556041e-06,
+      "loss": 0.5929,
+      "step": 10512
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.619679746855416,
+      "learning_rate": 8.953820885842366e-06,
+      "loss": 0.5436,
+      "step": 10513
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.316659410239936,
+      "learning_rate": 8.952143755881416e-06,
+      "loss": 0.5652,
+      "step": 10514
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 19.005804529680333,
+      "learning_rate": 8.95046665572088e-06,
+      "loss": 0.5638,
+      "step": 10515
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 23.193856865295665,
+      "learning_rate": 8.948789585408457e-06,
+      "loss": 0.6839,
+      "step": 10516
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 29.485420764435943,
+      "learning_rate": 8.947112544991846e-06,
+      "loss": 0.691,
+      "step": 10517
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.825134452574616,
+      "learning_rate": 8.945435534518737e-06,
+      "loss": 0.535,
+      "step": 10518
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 17.698285867340033,
+      "learning_rate": 8.943758554036828e-06,
+      "loss": 0.6034,
+      "step": 10519
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.116419948227406,
+      "learning_rate": 8.942081603593805e-06,
+      "loss": 0.5716,
+      "step": 10520
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.646230375318012,
+      "learning_rate": 8.94040468323736e-06,
+      "loss": 0.4644,
+      "step": 10521
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.147505503801593,
+      "learning_rate": 8.93872779301519e-06,
+      "loss": 0.5879,
+      "step": 10522
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 23.506815008398426,
+      "learning_rate": 8.93705093297498e-06,
+      "loss": 0.6004,
+      "step": 10523
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 28.961036983470393,
+      "learning_rate": 8.93537410316442e-06,
+      "loss": 0.5787,
+      "step": 10524
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.411648502851467,
+      "learning_rate": 8.933697303631195e-06,
+      "loss": 0.5981,
+      "step": 10525
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.85277461758098,
+      "learning_rate": 8.932020534422997e-06,
+      "loss": 0.5784,
+      "step": 10526
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.567264533179394,
+      "learning_rate": 8.930343795587508e-06,
+      "loss": 0.6279,
+      "step": 10527
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 19.783303706361327,
+      "learning_rate": 8.928667087172417e-06,
+      "loss": 0.5405,
+      "step": 10528
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 26.242620891570777,
+      "learning_rate": 8.92699040922541e-06,
+      "loss": 0.598,
+      "step": 10529
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 24.446708085871386,
+      "learning_rate": 8.925313761794166e-06,
+      "loss": 0.649,
+      "step": 10530
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 26.43444869004903,
+      "learning_rate": 8.92363714492637e-06,
+      "loss": 0.5425,
+      "step": 10531
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.426807286335617,
+      "learning_rate": 8.921960558669704e-06,
+      "loss": 0.5696,
+      "step": 10532
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.3181047966698,
+      "learning_rate": 8.920284003071846e-06,
+      "loss": 0.5564,
+      "step": 10533
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.588846155782246,
+      "learning_rate": 8.918607478180487e-06,
+      "loss": 0.6613,
+      "step": 10534
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 21.827995510850453,
+      "learning_rate": 8.916930984043291e-06,
+      "loss": 0.6416,
+      "step": 10535
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.902758208505269,
+      "learning_rate": 8.915254520707948e-06,
+      "loss": 0.6196,
+      "step": 10536
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.381999434105282,
+      "learning_rate": 8.913578088222129e-06,
+      "loss": 0.627,
+      "step": 10537
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.943552720985142,
+      "learning_rate": 8.911901686633516e-06,
+      "loss": 0.5887,
+      "step": 10538
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.470312676219738,
+      "learning_rate": 8.910225315989786e-06,
+      "loss": 0.532,
+      "step": 10539
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.918104668553518,
+      "learning_rate": 8.908548976338604e-06,
+      "loss": 0.5603,
+      "step": 10540
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.809946006343832,
+      "learning_rate": 8.906872667727652e-06,
+      "loss": 0.5891,
+      "step": 10541
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.78282057566971,
+      "learning_rate": 8.905196390204604e-06,
+      "loss": 0.5432,
+      "step": 10542
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.873296661550523,
+      "learning_rate": 8.903520143817128e-06,
+      "loss": 0.4863,
+      "step": 10543
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.609685488784915,
+      "learning_rate": 8.901843928612902e-06,
+      "loss": 0.5901,
+      "step": 10544
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.091840284059156,
+      "learning_rate": 8.90016774463959e-06,
+      "loss": 0.5348,
+      "step": 10545
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 29.813559242190617,
+      "learning_rate": 8.89849159194486e-06,
+      "loss": 0.5508,
+      "step": 10546
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.20410714700571,
+      "learning_rate": 8.89681547057639e-06,
+      "loss": 0.5794,
+      "step": 10547
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 21.54428253732981,
+      "learning_rate": 8.895139380581843e-06,
+      "loss": 0.5544,
+      "step": 10548
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.134956746082892,
+      "learning_rate": 8.893463322008885e-06,
+      "loss": 0.5286,
+      "step": 10549
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.96443540760124,
+      "learning_rate": 8.891787294905178e-06,
+      "loss": 0.5442,
+      "step": 10550
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.288512659939148,
+      "learning_rate": 8.890111299318396e-06,
+      "loss": 0.5372,
+      "step": 10551
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.022725706215542,
+      "learning_rate": 8.888435335296197e-06,
+      "loss": 0.6456,
+      "step": 10552
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 25.12157082642232,
+      "learning_rate": 8.886759402886253e-06,
+      "loss": 0.621,
+      "step": 10553
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.40396885521449,
+      "learning_rate": 8.885083502136214e-06,
+      "loss": 0.6463,
+      "step": 10554
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.201191068195104,
+      "learning_rate": 8.883407633093751e-06,
+      "loss": 0.6021,
+      "step": 10555
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.472412542503527,
+      "learning_rate": 8.881731795806517e-06,
+      "loss": 0.5157,
+      "step": 10556
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 29.813747006053468,
+      "learning_rate": 8.88005599032218e-06,
+      "loss": 0.6652,
+      "step": 10557
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 24.104706436439592,
+      "learning_rate": 8.878380216688399e-06,
+      "loss": 0.5574,
+      "step": 10558
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.11552890909985,
+      "learning_rate": 8.876704474952826e-06,
+      "loss": 0.527,
+      "step": 10559
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 21.01454380969632,
+      "learning_rate": 8.875028765163117e-06,
+      "loss": 0.5797,
+      "step": 10560
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 46.84552553615185,
+      "learning_rate": 8.873353087366936e-06,
+      "loss": 0.6101,
+      "step": 10561
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.20300028154275,
+      "learning_rate": 8.87167744161193e-06,
+      "loss": 0.5251,
+      "step": 10562
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.605342130132044,
+      "learning_rate": 8.870001827945764e-06,
+      "loss": 0.6308,
+      "step": 10563
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.411500302351627,
+      "learning_rate": 8.86832624641608e-06,
+      "loss": 0.5474,
+      "step": 10564
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 21.25021404270756,
+      "learning_rate": 8.866650697070535e-06,
+      "loss": 0.5891,
+      "step": 10565
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.708686623737233,
+      "learning_rate": 8.864975179956784e-06,
+      "loss": 0.6242,
+      "step": 10566
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.342905163318473,
+      "learning_rate": 8.863299695122473e-06,
+      "loss": 0.6278,
+      "step": 10567
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.250185552111372,
+      "learning_rate": 8.861624242615258e-06,
+      "loss": 0.5149,
+      "step": 10568
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.698050931771068,
+      "learning_rate": 8.859948822482777e-06,
+      "loss": 0.6111,
+      "step": 10569
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.970391682744916,
+      "learning_rate": 8.85827343477269e-06,
+      "loss": 0.6045,
+      "step": 10570
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.980238250287478,
+      "learning_rate": 8.856598079532633e-06,
+      "loss": 0.5814,
+      "step": 10571
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.126611172746667,
+      "learning_rate": 8.854922756810263e-06,
+      "loss": 0.615,
+      "step": 10572
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.35933075934371,
+      "learning_rate": 8.85324746665322e-06,
+      "loss": 0.524,
+      "step": 10573
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.341025619636287,
+      "learning_rate": 8.851572209109147e-06,
+      "loss": 0.5864,
+      "step": 10574
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.709495743204776,
+      "learning_rate": 8.849896984225685e-06,
+      "loss": 0.5425,
+      "step": 10575
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.90207491511352,
+      "learning_rate": 8.848221792050484e-06,
+      "loss": 0.5611,
+      "step": 10576
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.862340315170645,
+      "learning_rate": 8.846546632631178e-06,
+      "loss": 0.5316,
+      "step": 10577
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.112449246456217,
+      "learning_rate": 8.844871506015417e-06,
+      "loss": 0.5679,
+      "step": 10578
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.957798917973193,
+      "learning_rate": 8.84319641225083e-06,
+      "loss": 0.642,
+      "step": 10579
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.40861249716418,
+      "learning_rate": 8.841521351385061e-06,
+      "loss": 0.6312,
+      "step": 10580
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.820813264069383,
+      "learning_rate": 8.839846323465745e-06,
+      "loss": 0.5613,
+      "step": 10581
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.41690770797413,
+      "learning_rate": 8.838171328540524e-06,
+      "loss": 0.4693,
+      "step": 10582
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.989719391963709,
+      "learning_rate": 8.836496366657032e-06,
+      "loss": 0.562,
+      "step": 10583
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.3326707033684,
+      "learning_rate": 8.834821437862899e-06,
+      "loss": 0.5911,
+      "step": 10584
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.576962384099044,
+      "learning_rate": 8.833146542205765e-06,
+      "loss": 0.6162,
+      "step": 10585
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.4642771952231,
+      "learning_rate": 8.831471679733262e-06,
+      "loss": 0.5255,
+      "step": 10586
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.243070515432077,
+      "learning_rate": 8.829796850493016e-06,
+      "loss": 0.5509,
+      "step": 10587
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.86933225291331,
+      "learning_rate": 8.828122054532669e-06,
+      "loss": 0.5437,
+      "step": 10588
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.259582024807646,
+      "learning_rate": 8.826447291899842e-06,
+      "loss": 0.5232,
+      "step": 10589
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.975049566519548,
+      "learning_rate": 8.824772562642166e-06,
+      "loss": 0.5528,
+      "step": 10590
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.093213555733982,
+      "learning_rate": 8.823097866807272e-06,
+      "loss": 0.6207,
+      "step": 10591
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.031383719263756,
+      "learning_rate": 8.82142320444279e-06,
+      "loss": 0.5709,
+      "step": 10592
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.665876311398485,
+      "learning_rate": 8.81974857559634e-06,
+      "loss": 0.5485,
+      "step": 10593
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.212568141739474,
+      "learning_rate": 8.818073980315547e-06,
+      "loss": 0.5291,
+      "step": 10594
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.588865287298297,
+      "learning_rate": 8.816399418648041e-06,
+      "loss": 0.5733,
+      "step": 10595
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.849516488449012,
+      "learning_rate": 8.814724890641444e-06,
+      "loss": 0.5425,
+      "step": 10596
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 21.87554829097664,
+      "learning_rate": 8.81305039634338e-06,
+      "loss": 0.5708,
+      "step": 10597
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.465789852701272,
+      "learning_rate": 8.811375935801463e-06,
+      "loss": 0.5993,
+      "step": 10598
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 28.559402022352625,
+      "learning_rate": 8.809701509063322e-06,
+      "loss": 0.5763,
+      "step": 10599
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.03008716436355,
+      "learning_rate": 8.808027116176572e-06,
+      "loss": 0.606,
+      "step": 10600
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.245845244938598,
+      "learning_rate": 8.806352757188835e-06,
+      "loss": 0.5874,
+      "step": 10601
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.14003062356893,
+      "learning_rate": 8.80467843214773e-06,
+      "loss": 0.6026,
+      "step": 10602
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.204199934927258,
+      "learning_rate": 8.803004141100869e-06,
+      "loss": 0.5457,
+      "step": 10603
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.529404622585364,
+      "learning_rate": 8.801329884095866e-06,
+      "loss": 0.5848,
+      "step": 10604
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.298156957054113,
+      "learning_rate": 8.799655661180344e-06,
+      "loss": 0.5413,
+      "step": 10605
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.737168289497772,
+      "learning_rate": 8.79798147240191e-06,
+      "loss": 0.5144,
+      "step": 10606
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 23.805284703584743,
+      "learning_rate": 8.796307317808183e-06,
+      "loss": 0.58,
+      "step": 10607
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 25.893084368389925,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.6639,
+      "step": 10608
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.783285408542032,
+      "learning_rate": 8.792959111365285e-06,
+      "loss": 0.5912,
+      "step": 10609
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 17.15976555473086,
+      "learning_rate": 8.791285059611335e-06,
+      "loss": 0.5486,
+      "step": 10610
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 58.08449727410499,
+      "learning_rate": 8.789611042232534e-06,
+      "loss": 0.5791,
+      "step": 10611
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.858577084603429,
+      "learning_rate": 8.787937059276486e-06,
+      "loss": 0.6202,
+      "step": 10612
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.93035156485214,
+      "learning_rate": 8.786263110790796e-06,
+      "loss": 0.5729,
+      "step": 10613
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.78064060734215,
+      "learning_rate": 8.784589196823075e-06,
+      "loss": 0.5756,
+      "step": 10614
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.37020895162406,
+      "learning_rate": 8.782915317420924e-06,
+      "loss": 0.6149,
+      "step": 10615
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.33864730379393,
+      "learning_rate": 8.781241472631952e-06,
+      "loss": 0.5862,
+      "step": 10616
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.152805862941696,
+      "learning_rate": 8.779567662503761e-06,
+      "loss": 0.5163,
+      "step": 10617
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.74101156120906,
+      "learning_rate": 8.77789388708395e-06,
+      "loss": 0.6642,
+      "step": 10618
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.764350783510245,
+      "learning_rate": 8.77622014642012e-06,
+      "loss": 0.5889,
+      "step": 10619
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.104367664027645,
+      "learning_rate": 8.774546440559874e-06,
+      "loss": 0.6471,
+      "step": 10620
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.159442081747295,
+      "learning_rate": 8.77287276955081e-06,
+      "loss": 0.5624,
+      "step": 10621
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 17.761670993650963,
+      "learning_rate": 8.771199133440533e-06,
+      "loss": 0.6047,
+      "step": 10622
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.538438136756746,
+      "learning_rate": 8.769525532276627e-06,
+      "loss": 0.5557,
+      "step": 10623
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.501896265708194,
+      "learning_rate": 8.767851966106696e-06,
+      "loss": 0.5563,
+      "step": 10624
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.367783037672163,
+      "learning_rate": 8.766178434978334e-06,
+      "loss": 0.5335,
+      "step": 10625
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.82841159677827,
+      "learning_rate": 8.764504938939137e-06,
+      "loss": 0.5462,
+      "step": 10626
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.8786251580896,
+      "learning_rate": 8.762831478036698e-06,
+      "loss": 0.5568,
+      "step": 10627
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.848564690399293,
+      "learning_rate": 8.761158052318605e-06,
+      "loss": 0.5251,
+      "step": 10628
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.7890619201901,
+      "learning_rate": 8.759484661832454e-06,
+      "loss": 0.6059,
+      "step": 10629
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.41973266728501,
+      "learning_rate": 8.757811306625833e-06,
+      "loss": 0.5684,
+      "step": 10630
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 26.163483730326483,
+      "learning_rate": 8.75613798674633e-06,
+      "loss": 0.6227,
+      "step": 10631
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.754502767668455,
+      "learning_rate": 8.754464702241539e-06,
+      "loss": 0.5203,
+      "step": 10632
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 33.083697424926214,
+      "learning_rate": 8.752791453159041e-06,
+      "loss": 0.6494,
+      "step": 10633
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.333628676123528,
+      "learning_rate": 8.751118239546424e-06,
+      "loss": 0.5658,
+      "step": 10634
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.01461747942771,
+      "learning_rate": 8.749445061451274e-06,
+      "loss": 0.5694,
+      "step": 10635
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.762830054612646,
+      "learning_rate": 8.747771918921176e-06,
+      "loss": 0.5754,
+      "step": 10636
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.04331935765732,
+      "learning_rate": 8.746098812003714e-06,
+      "loss": 0.5217,
+      "step": 10637
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 21.803258697348074,
+      "learning_rate": 8.744425740746464e-06,
+      "loss": 0.6335,
+      "step": 10638
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 35.64794688542407,
+      "learning_rate": 8.742752705197013e-06,
+      "loss": 0.6455,
+      "step": 10639
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.46138732668498,
+      "learning_rate": 8.741079705402939e-06,
+      "loss": 0.5211,
+      "step": 10640
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.39230899749596,
+      "learning_rate": 8.739406741411826e-06,
+      "loss": 0.6125,
+      "step": 10641
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.056763465866894,
+      "learning_rate": 8.737733813271244e-06,
+      "loss": 0.6236,
+      "step": 10642
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 23.11978770809794,
+      "learning_rate": 8.736060921028773e-06,
+      "loss": 0.6278,
+      "step": 10643
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.55293153906435,
+      "learning_rate": 8.73438806473199e-06,
+      "loss": 0.6099,
+      "step": 10644
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 31.703294073925896,
+      "learning_rate": 8.732715244428472e-06,
+      "loss": 0.5905,
+      "step": 10645
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.044599065678185,
+      "learning_rate": 8.731042460165793e-06,
+      "loss": 0.6493,
+      "step": 10646
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 17.53069151279472,
+      "learning_rate": 8.72936971199152e-06,
+      "loss": 0.5679,
+      "step": 10647
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.404807442570014,
+      "learning_rate": 8.72769699995323e-06,
+      "loss": 0.5657,
+      "step": 10648
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.64926481592163,
+      "learning_rate": 8.726024324098494e-06,
+      "loss": 0.5606,
+      "step": 10649
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.096280654487593,
+      "learning_rate": 8.724351684474878e-06,
+      "loss": 0.6145,
+      "step": 10650
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.983145720513793,
+      "learning_rate": 8.72267908112996e-06,
+      "loss": 0.5798,
+      "step": 10651
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.717085467814213,
+      "learning_rate": 8.721006514111299e-06,
+      "loss": 0.6065,
+      "step": 10652
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.342680007155987,
+      "learning_rate": 8.719333983466462e-06,
+      "loss": 0.573,
+      "step": 10653
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 17.836395298734068,
+      "learning_rate": 8.71766148924302e-06,
+      "loss": 0.5968,
+      "step": 10654
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.802264576610416,
+      "learning_rate": 8.715989031488534e-06,
+      "loss": 0.5596,
+      "step": 10655
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 31.443259978035083,
+      "learning_rate": 8.71431661025057e-06,
+      "loss": 0.6468,
+      "step": 10656
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 23.129382336842877,
+      "learning_rate": 8.712644225576687e-06,
+      "loss": 0.5922,
+      "step": 10657
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 28.295943588734268,
+      "learning_rate": 8.710971877514452e-06,
+      "loss": 0.6159,
+      "step": 10658
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.424142085832925,
+      "learning_rate": 8.709299566111418e-06,
+      "loss": 0.5413,
+      "step": 10659
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.136337155818854,
+      "learning_rate": 8.707627291415153e-06,
+      "loss": 0.5498,
+      "step": 10660
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.46602635314065,
+      "learning_rate": 8.705955053473215e-06,
+      "loss": 0.5709,
+      "step": 10661
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.894539876768423,
+      "learning_rate": 8.704282852333154e-06,
+      "loss": 0.5348,
+      "step": 10662
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.28782670389082,
+      "learning_rate": 8.702610688042531e-06,
+      "loss": 0.5645,
+      "step": 10663
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 24.659024680775442,
+      "learning_rate": 8.700938560648901e-06,
+      "loss": 0.5586,
+      "step": 10664
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.07548852227552,
+      "learning_rate": 8.699266470199817e-06,
+      "loss": 0.5867,
+      "step": 10665
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 24.687843035475538,
+      "learning_rate": 8.697594416742842e-06,
+      "loss": 0.567,
+      "step": 10666
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.690082866130616,
+      "learning_rate": 8.695922400325512e-06,
+      "loss": 0.6163,
+      "step": 10667
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.467159651740882,
+      "learning_rate": 8.694250420995387e-06,
+      "loss": 0.5532,
+      "step": 10668
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.305395753775834,
+      "learning_rate": 8.692578478800015e-06,
+      "loss": 0.545,
+      "step": 10669
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.87434396299192,
+      "learning_rate": 8.690906573786949e-06,
+      "loss": 0.5222,
+      "step": 10670
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.899565599573474,
+      "learning_rate": 8.689234706003735e-06,
+      "loss": 0.5471,
+      "step": 10671
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.70416979057833,
+      "learning_rate": 8.687562875497915e-06,
+      "loss": 0.5633,
+      "step": 10672
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.019830241904353,
+      "learning_rate": 8.685891082317043e-06,
+      "loss": 0.5505,
+      "step": 10673
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.802964553488113,
+      "learning_rate": 8.684219326508657e-06,
+      "loss": 0.5658,
+      "step": 10674
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 21.074838988021675,
+      "learning_rate": 8.682547608120305e-06,
+      "loss": 0.6734,
+      "step": 10675
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.346922444668317,
+      "learning_rate": 8.68087592719953e-06,
+      "loss": 0.5758,
+      "step": 10676
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.69825046983035,
+      "learning_rate": 8.679204283793872e-06,
+      "loss": 0.5865,
+      "step": 10677
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 21.119922815723925,
+      "learning_rate": 8.677532677950868e-06,
+      "loss": 0.5694,
+      "step": 10678
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.771917218837862,
+      "learning_rate": 8.675861109718064e-06,
+      "loss": 0.5888,
+      "step": 10679
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.020549771981512,
+      "learning_rate": 8.674189579142993e-06,
+      "loss": 0.5295,
+      "step": 10680
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.487481324367025,
+      "learning_rate": 8.672518086273199e-06,
+      "loss": 0.5509,
+      "step": 10681
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.103027997523373,
+      "learning_rate": 8.67084663115621e-06,
+      "loss": 0.6162,
+      "step": 10682
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.89042324849556,
+      "learning_rate": 8.669175213839566e-06,
+      "loss": 0.6217,
+      "step": 10683
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.26186337375132,
+      "learning_rate": 8.6675038343708e-06,
+      "loss": 0.5555,
+      "step": 10684
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.550121457856772,
+      "learning_rate": 8.665832492797447e-06,
+      "loss": 0.624,
+      "step": 10685
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 39.97594128846003,
+      "learning_rate": 8.664161189167039e-06,
+      "loss": 0.5122,
+      "step": 10686
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.74867915085563,
+      "learning_rate": 8.662489923527104e-06,
+      "loss": 0.6164,
+      "step": 10687
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 30.633618695158294,
+      "learning_rate": 8.66081869592517e-06,
+      "loss": 0.5545,
+      "step": 10688
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.27831230397225,
+      "learning_rate": 8.659147506408771e-06,
+      "loss": 0.5698,
+      "step": 10689
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.733921291307386,
+      "learning_rate": 8.657476355025436e-06,
+      "loss": 0.5785,
+      "step": 10690
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.975913549921286,
+      "learning_rate": 8.655805241822683e-06,
+      "loss": 0.5286,
+      "step": 10691
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 26.078959152841342,
+      "learning_rate": 8.654134166848045e-06,
+      "loss": 0.6512,
+      "step": 10692
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.678011360871974,
+      "learning_rate": 8.652463130149044e-06,
+      "loss": 0.5497,
+      "step": 10693
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.936375516138952,
+      "learning_rate": 8.6507921317732e-06,
+      "loss": 0.5309,
+      "step": 10694
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.004562876209715,
+      "learning_rate": 8.649121171768045e-06,
+      "loss": 0.5622,
+      "step": 10695
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.525471573860942,
+      "learning_rate": 8.64745025018109e-06,
+      "loss": 0.5093,
+      "step": 10696
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.033996785871246,
+      "learning_rate": 8.645779367059856e-06,
+      "loss": 0.5196,
+      "step": 10697
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.416615364736174,
+      "learning_rate": 8.644108522451868e-06,
+      "loss": 0.5451,
+      "step": 10698
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 24.653989905853603,
+      "learning_rate": 8.642437716404639e-06,
+      "loss": 0.6033,
+      "step": 10699
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 29.13510300191042,
+      "learning_rate": 8.64076694896569e-06,
+      "loss": 0.6148,
+      "step": 10700
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.439080940319844,
+      "learning_rate": 8.63909622018253e-06,
+      "loss": 0.5496,
+      "step": 10701
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.81245478191997,
+      "learning_rate": 8.63742553010268e-06,
+      "loss": 0.4813,
+      "step": 10702
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.147938339259284,
+      "learning_rate": 8.635754878773647e-06,
+      "loss": 0.5894,
+      "step": 10703
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.729983585734047,
+      "learning_rate": 8.63408426624295e-06,
+      "loss": 0.5983,
+      "step": 10704
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.491812111122975,
+      "learning_rate": 8.632413692558101e-06,
+      "loss": 0.6297,
+      "step": 10705
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.5818421440928,
+      "learning_rate": 8.630743157766603e-06,
+      "loss": 0.5013,
+      "step": 10706
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.654129999293646,
+      "learning_rate": 8.629072661915969e-06,
+      "loss": 0.5424,
+      "step": 10707
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 32.61416526818084,
+      "learning_rate": 8.627402205053705e-06,
+      "loss": 0.6774,
+      "step": 10708
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.55944077973013,
+      "learning_rate": 8.62573178722732e-06,
+      "loss": 0.6439,
+      "step": 10709
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 37.969366467341764,
+      "learning_rate": 8.624061408484323e-06,
+      "loss": 0.6523,
+      "step": 10710
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.16996973632973,
+      "learning_rate": 8.62239106887221e-06,
+      "loss": 0.5698,
+      "step": 10711
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.657162908334145,
+      "learning_rate": 8.620720768438491e-06,
+      "loss": 0.5486,
+      "step": 10712
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.22964584719875,
+      "learning_rate": 8.619050507230664e-06,
+      "loss": 0.512,
+      "step": 10713
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.45094193864924,
+      "learning_rate": 8.617380285296235e-06,
+      "loss": 0.5656,
+      "step": 10714
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.745405615218226,
+      "learning_rate": 8.615710102682704e-06,
+      "loss": 0.541,
+      "step": 10715
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.072208825359976,
+      "learning_rate": 8.614039959437563e-06,
+      "loss": 0.4817,
+      "step": 10716
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.896459487998627,
+      "learning_rate": 8.612369855608318e-06,
+      "loss": 0.5105,
+      "step": 10717
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.380521483569984,
+      "learning_rate": 8.610699791242463e-06,
+      "loss": 0.5645,
+      "step": 10718
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.418514226902982,
+      "learning_rate": 8.609029766387489e-06,
+      "loss": 0.5872,
+      "step": 10719
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.166731689153888,
+      "learning_rate": 8.6073597810909e-06,
+      "loss": 0.5525,
+      "step": 10720
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.744072309163542,
+      "learning_rate": 8.60568983540018e-06,
+      "loss": 0.5425,
+      "step": 10721
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.926109861894297,
+      "learning_rate": 8.604019929362826e-06,
+      "loss": 0.5313,
+      "step": 10722
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.165968500856906,
+      "learning_rate": 8.602350063026332e-06,
+      "loss": 0.6692,
+      "step": 10723
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 23.68867824160515,
+      "learning_rate": 8.600680236438182e-06,
+      "loss": 0.6164,
+      "step": 10724
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.540329024768237,
+      "learning_rate": 8.59901044964587e-06,
+      "loss": 0.5649,
+      "step": 10725
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.795344318937275,
+      "learning_rate": 8.597340702696879e-06,
+      "loss": 0.5707,
+      "step": 10726
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.778287348288444,
+      "learning_rate": 8.595670995638697e-06,
+      "loss": 0.6534,
+      "step": 10727
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.846555598264512,
+      "learning_rate": 8.59400132851881e-06,
+      "loss": 0.6313,
+      "step": 10728
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 60.657481253762995,
+      "learning_rate": 8.592331701384704e-06,
+      "loss": 0.5221,
+      "step": 10729
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.445630685882481,
+      "learning_rate": 8.590662114283864e-06,
+      "loss": 0.5594,
+      "step": 10730
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.772839558640303,
+      "learning_rate": 8.588992567263767e-06,
+      "loss": 0.5873,
+      "step": 10731
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.228508343026352,
+      "learning_rate": 8.587323060371893e-06,
+      "loss": 0.5972,
+      "step": 10732
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.554875978096312,
+      "learning_rate": 8.585653593655728e-06,
+      "loss": 0.4995,
+      "step": 10733
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.764185484088166,
+      "learning_rate": 8.583984167162744e-06,
+      "loss": 0.5808,
+      "step": 10734
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.758167474796675,
+      "learning_rate": 8.582314780940426e-06,
+      "loss": 0.6025,
+      "step": 10735
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 24.464260946486014,
+      "learning_rate": 8.580645435036246e-06,
+      "loss": 0.6673,
+      "step": 10736
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.463984912020386,
+      "learning_rate": 8.578976129497678e-06,
+      "loss": 0.6484,
+      "step": 10737
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.224395771497912,
+      "learning_rate": 8.577306864372196e-06,
+      "loss": 0.5168,
+      "step": 10738
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 22.178047040568764,
+      "learning_rate": 8.575637639707279e-06,
+      "loss": 0.5852,
+      "step": 10739
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.675893029240875,
+      "learning_rate": 8.573968455550393e-06,
+      "loss": 0.5868,
+      "step": 10740
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.05917112309347,
+      "learning_rate": 8.572299311949005e-06,
+      "loss": 0.5661,
+      "step": 10741
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.85306827364698,
+      "learning_rate": 8.570630208950593e-06,
+      "loss": 0.5354,
+      "step": 10742
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 24.342044462683226,
+      "learning_rate": 8.568961146602623e-06,
+      "loss": 0.7152,
+      "step": 10743
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.53403104858795,
+      "learning_rate": 8.567292124952562e-06,
+      "loss": 0.5077,
+      "step": 10744
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.330861984728294,
+      "learning_rate": 8.565623144047869e-06,
+      "loss": 0.6172,
+      "step": 10745
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.933010554071792,
+      "learning_rate": 8.563954203936018e-06,
+      "loss": 0.5138,
+      "step": 10746
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.617867703211141,
+      "learning_rate": 8.562285304664468e-06,
+      "loss": 0.5158,
+      "step": 10747
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.220326867581598,
+      "learning_rate": 8.560616446280685e-06,
+      "loss": 0.5887,
+      "step": 10748
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 21.163282920783544,
+      "learning_rate": 8.558947628832131e-06,
+      "loss": 0.547,
+      "step": 10749
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.178377573961438,
+      "learning_rate": 8.55727885236626e-06,
+      "loss": 0.501,
+      "step": 10750
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.539447230513943,
+      "learning_rate": 8.555610116930532e-06,
+      "loss": 0.5629,
+      "step": 10751
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.153772731149665,
+      "learning_rate": 8.55394142257241e-06,
+      "loss": 0.6401,
+      "step": 10752
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 22.523729855680674,
+      "learning_rate": 8.552272769339346e-06,
+      "loss": 0.5685,
+      "step": 10753
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.144733790670937,
+      "learning_rate": 8.550604157278804e-06,
+      "loss": 0.5591,
+      "step": 10754
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.987875599972817,
+      "learning_rate": 8.548935586438226e-06,
+      "loss": 0.5283,
+      "step": 10755
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 30.957807026477884,
+      "learning_rate": 8.547267056865072e-06,
+      "loss": 0.6091,
+      "step": 10756
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.298346499211956,
+      "learning_rate": 8.545598568606791e-06,
+      "loss": 0.5476,
+      "step": 10757
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 20.67125036199832,
+      "learning_rate": 8.54393012171084e-06,
+      "loss": 0.6027,
+      "step": 10758
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.273155631003064,
+      "learning_rate": 8.542261716224664e-06,
+      "loss": 0.602,
+      "step": 10759
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 29.03129477851625,
+      "learning_rate": 8.54059335219571e-06,
+      "loss": 0.7013,
+      "step": 10760
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.30868392574454,
+      "learning_rate": 8.538925029671428e-06,
+      "loss": 0.5596,
+      "step": 10761
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.693635683094321,
+      "learning_rate": 8.537256748699264e-06,
+      "loss": 0.5799,
+      "step": 10762
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 27.46103990950827,
+      "learning_rate": 8.53558850932666e-06,
+      "loss": 0.5265,
+      "step": 10763
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.19505132964049,
+      "learning_rate": 8.533920311601068e-06,
+      "loss": 0.592,
+      "step": 10764
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.687741969440236,
+      "learning_rate": 8.53225215556992e-06,
+      "loss": 0.5944,
+      "step": 10765
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.572429562279567,
+      "learning_rate": 8.530584041280661e-06,
+      "loss": 0.4828,
+      "step": 10766
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 24.930239366180977,
+      "learning_rate": 8.528915968780735e-06,
+      "loss": 0.5969,
+      "step": 10767
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 21.761951885231092,
+      "learning_rate": 8.527247938117577e-06,
+      "loss": 0.5746,
+      "step": 10768
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.295019880691335,
+      "learning_rate": 8.525579949338631e-06,
+      "loss": 0.557,
+      "step": 10769
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 22.540317906369257,
+      "learning_rate": 8.523912002491321e-06,
+      "loss": 0.5871,
+      "step": 10770
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.08280525004356,
+      "learning_rate": 8.522244097623095e-06,
+      "loss": 0.5634,
+      "step": 10771
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.854267211107338,
+      "learning_rate": 8.520576234781379e-06,
+      "loss": 0.6086,
+      "step": 10772
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 26.61927199983796,
+      "learning_rate": 8.518908414013612e-06,
+      "loss": 0.5546,
+      "step": 10773
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 20.498914969037678,
+      "learning_rate": 8.517240635367225e-06,
+      "loss": 0.5376,
+      "step": 10774
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.791312800465825,
+      "learning_rate": 8.515572898889645e-06,
+      "loss": 0.6344,
+      "step": 10775
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.368579324180516,
+      "learning_rate": 8.513905204628302e-06,
+      "loss": 0.596,
+      "step": 10776
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.730523377658212,
+      "learning_rate": 8.512237552630625e-06,
+      "loss": 0.5299,
+      "step": 10777
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.58309404923637,
+      "learning_rate": 8.510569942944043e-06,
+      "loss": 0.5225,
+      "step": 10778
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.148507973932883,
+      "learning_rate": 8.508902375615982e-06,
+      "loss": 0.5601,
+      "step": 10779
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 23.22874099415552,
+      "learning_rate": 8.507234850693864e-06,
+      "loss": 0.549,
+      "step": 10780
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.78281593201867,
+      "learning_rate": 8.505567368225116e-06,
+      "loss": 0.5283,
+      "step": 10781
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.257811337587967,
+      "learning_rate": 8.503899928257151e-06,
+      "loss": 0.5914,
+      "step": 10782
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.45303843612514,
+      "learning_rate": 8.502232530837404e-06,
+      "loss": 0.6183,
+      "step": 10783
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.4942155240457,
+      "learning_rate": 8.500565176013286e-06,
+      "loss": 0.5439,
+      "step": 10784
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 26.19587388704196,
+      "learning_rate": 8.498897863832215e-06,
+      "loss": 0.6469,
+      "step": 10785
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 28.86250391636306,
+      "learning_rate": 8.497230594341612e-06,
+      "loss": 0.55,
+      "step": 10786
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.092384997136644,
+      "learning_rate": 8.495563367588893e-06,
+      "loss": 0.5076,
+      "step": 10787
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.610718187934484,
+      "learning_rate": 8.493896183621474e-06,
+      "loss": 0.5576,
+      "step": 10788
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.281071674538994,
+      "learning_rate": 8.49222904248676e-06,
+      "loss": 0.5307,
+      "step": 10789
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.53231879887502,
+      "learning_rate": 8.490561944232173e-06,
+      "loss": 0.5903,
+      "step": 10790
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.234581263813762,
+      "learning_rate": 8.48889488890512e-06,
+      "loss": 0.5798,
+      "step": 10791
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.00969939803462,
+      "learning_rate": 8.487227876553012e-06,
+      "loss": 0.5739,
+      "step": 10792
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.15902377305113,
+      "learning_rate": 8.485560907223264e-06,
+      "loss": 0.54,
+      "step": 10793
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.751103074845837,
+      "learning_rate": 8.483893980963274e-06,
+      "loss": 0.5303,
+      "step": 10794
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.350415813533033,
+      "learning_rate": 8.48222709782045e-06,
+      "loss": 0.5704,
+      "step": 10795
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.66552754341187,
+      "learning_rate": 8.480560257842201e-06,
+      "loss": 0.5487,
+      "step": 10796
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.400866645839347,
+      "learning_rate": 8.478893461075929e-06,
+      "loss": 0.5899,
+      "step": 10797
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.682271762916503,
+      "learning_rate": 8.477226707569042e-06,
+      "loss": 0.4899,
+      "step": 10798
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.790982739180222,
+      "learning_rate": 8.475559997368932e-06,
+      "loss": 0.5653,
+      "step": 10799
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 22.081901957836713,
+      "learning_rate": 8.473893330523004e-06,
+      "loss": 0.5363,
+      "step": 10800
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 35.674554635624865,
+      "learning_rate": 8.472226707078655e-06,
+      "loss": 0.5174,
+      "step": 10801
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 23.486239177828867,
+      "learning_rate": 8.470560127083288e-06,
+      "loss": 0.7111,
+      "step": 10802
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.824546100939184,
+      "learning_rate": 8.468893590584299e-06,
+      "loss": 0.5417,
+      "step": 10803
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.76853911033336,
+      "learning_rate": 8.467227097629075e-06,
+      "loss": 0.5925,
+      "step": 10804
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.867199731510848,
+      "learning_rate": 8.465560648265017e-06,
+      "loss": 0.5799,
+      "step": 10805
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.116085691283033,
+      "learning_rate": 8.463894242539518e-06,
+      "loss": 0.5612,
+      "step": 10806
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.025610195884738,
+      "learning_rate": 8.462227880499967e-06,
+      "loss": 0.5085,
+      "step": 10807
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 31.240371653239404,
+      "learning_rate": 8.460561562193758e-06,
+      "loss": 0.6057,
+      "step": 10808
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.539389937135326,
+      "learning_rate": 8.458895287668277e-06,
+      "loss": 0.5588,
+      "step": 10809
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 20.450996276492145,
+      "learning_rate": 8.457229056970908e-06,
+      "loss": 0.4768,
+      "step": 10810
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.282977499772514,
+      "learning_rate": 8.455562870149046e-06,
+      "loss": 0.5475,
+      "step": 10811
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 25.531087087510674,
+      "learning_rate": 8.453896727250072e-06,
+      "loss": 0.5638,
+      "step": 10812
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.864832133815286,
+      "learning_rate": 8.452230628321373e-06,
+      "loss": 0.5117,
+      "step": 10813
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.856022948887485,
+      "learning_rate": 8.450564573410324e-06,
+      "loss": 0.5839,
+      "step": 10814
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.973730064148045,
+      "learning_rate": 8.448898562564316e-06,
+      "loss": 0.5701,
+      "step": 10815
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.4664471573499,
+      "learning_rate": 8.447232595830724e-06,
+      "loss": 0.5822,
+      "step": 10816
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 23.41364848235764,
+      "learning_rate": 8.445566673256928e-06,
+      "loss": 0.4963,
+      "step": 10817
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.801153053286477,
+      "learning_rate": 8.443900794890311e-06,
+      "loss": 0.6126,
+      "step": 10818
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.906502736191467,
+      "learning_rate": 8.442234960778242e-06,
+      "loss": 0.5466,
+      "step": 10819
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.460889732779616,
+      "learning_rate": 8.440569170968098e-06,
+      "loss": 0.5557,
+      "step": 10820
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.172252852193743,
+      "learning_rate": 8.438903425507257e-06,
+      "loss": 0.5847,
+      "step": 10821
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.27801680209744,
+      "learning_rate": 8.437237724443086e-06,
+      "loss": 0.5818,
+      "step": 10822
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.469337238790432,
+      "learning_rate": 8.435572067822964e-06,
+      "loss": 0.5871,
+      "step": 10823
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.37746908192004,
+      "learning_rate": 8.433906455694256e-06,
+      "loss": 0.5186,
+      "step": 10824
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.13722574333145,
+      "learning_rate": 8.432240888104331e-06,
+      "loss": 0.5721,
+      "step": 10825
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.924817153930416,
+      "learning_rate": 8.430575365100557e-06,
+      "loss": 0.526,
+      "step": 10826
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.800927392310008,
+      "learning_rate": 8.428909886730303e-06,
+      "loss": 0.578,
+      "step": 10827
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 22.606140092472344,
+      "learning_rate": 8.427244453040937e-06,
+      "loss": 0.5831,
+      "step": 10828
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 22.35319666936993,
+      "learning_rate": 8.425579064079811e-06,
+      "loss": 0.5031,
+      "step": 10829
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 25.111957761121204,
+      "learning_rate": 8.423913719894301e-06,
+      "loss": 0.6381,
+      "step": 10830
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 20.767845607352733,
+      "learning_rate": 8.42224842053176e-06,
+      "loss": 0.6046,
+      "step": 10831
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.303750973367805,
+      "learning_rate": 8.420583166039551e-06,
+      "loss": 0.552,
+      "step": 10832
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.924034830635621,
+      "learning_rate": 8.418917956465037e-06,
+      "loss": 0.5613,
+      "step": 10833
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.41116487785557,
+      "learning_rate": 8.417252791855566e-06,
+      "loss": 0.5754,
+      "step": 10834
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.25801904501597,
+      "learning_rate": 8.4155876722585e-06,
+      "loss": 0.659,
+      "step": 10835
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.872164505998658,
+      "learning_rate": 8.413922597721197e-06,
+      "loss": 0.5593,
+      "step": 10836
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.992201474627034,
+      "learning_rate": 8.41225756829101e-06,
+      "loss": 0.5238,
+      "step": 10837
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 28.645610031066756,
+      "learning_rate": 8.410592584015284e-06,
+      "loss": 0.6622,
+      "step": 10838
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.259955414190802,
+      "learning_rate": 8.408927644941373e-06,
+      "loss": 0.553,
+      "step": 10839
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.493482280724884,
+      "learning_rate": 8.407262751116633e-06,
+      "loss": 0.5705,
+      "step": 10840
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.236174416706437,
+      "learning_rate": 8.405597902588407e-06,
+      "loss": 0.5587,
+      "step": 10841
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.273024350079066,
+      "learning_rate": 8.403933099404047e-06,
+      "loss": 0.6023,
+      "step": 10842
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.518787243893357,
+      "learning_rate": 8.402268341610895e-06,
+      "loss": 0.5304,
+      "step": 10843
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.61697461072004,
+      "learning_rate": 8.400603629256294e-06,
+      "loss": 0.5343,
+      "step": 10844
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.30347646101296,
+      "learning_rate": 8.39893896238759e-06,
+      "loss": 0.59,
+      "step": 10845
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 31.802763261309277,
+      "learning_rate": 8.397274341052128e-06,
+      "loss": 0.6912,
+      "step": 10846
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 28.9689021548912,
+      "learning_rate": 8.395609765297249e-06,
+      "loss": 0.6296,
+      "step": 10847
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.51220117238763,
+      "learning_rate": 8.393945235170283e-06,
+      "loss": 0.5139,
+      "step": 10848
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 23.253475208923728,
+      "learning_rate": 8.392280750718577e-06,
+      "loss": 0.6178,
+      "step": 10849
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 40.39131619804619,
+      "learning_rate": 8.390616311989468e-06,
+      "loss": 0.5895,
+      "step": 10850
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.26240991027494,
+      "learning_rate": 8.388951919030287e-06,
+      "loss": 0.5462,
+      "step": 10851
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.505393947093093,
+      "learning_rate": 8.387287571888373e-06,
+      "loss": 0.5715,
+      "step": 10852
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 29.347960694063094,
+      "learning_rate": 8.385623270611058e-06,
+      "loss": 0.621,
+      "step": 10853
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.533378821857262,
+      "learning_rate": 8.383959015245669e-06,
+      "loss": 0.5487,
+      "step": 10854
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.14716468032409,
+      "learning_rate": 8.382294805839543e-06,
+      "loss": 0.6003,
+      "step": 10855
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 30.36464852636328,
+      "learning_rate": 8.380630642440006e-06,
+      "loss": 0.5592,
+      "step": 10856
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.669617198803504,
+      "learning_rate": 8.378966525094388e-06,
+      "loss": 0.5366,
+      "step": 10857
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.349752383971616,
+      "learning_rate": 8.377302453850012e-06,
+      "loss": 0.5522,
+      "step": 10858
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.422825150170677,
+      "learning_rate": 8.375638428754204e-06,
+      "loss": 0.4631,
+      "step": 10859
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.44501896224781,
+      "learning_rate": 8.373974449854288e-06,
+      "loss": 0.594,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.357106091811323,
+      "learning_rate": 8.37231051719759e-06,
+      "loss": 0.541,
+      "step": 10861
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.665164062469536,
+      "learning_rate": 8.37064663083143e-06,
+      "loss": 0.4646,
+      "step": 10862
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.425120253536413,
+      "learning_rate": 8.368982790803124e-06,
+      "loss": 0.5297,
+      "step": 10863
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.342211771554663,
+      "learning_rate": 8.367318997159992e-06,
+      "loss": 0.6073,
+      "step": 10864
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.74727391746629,
+      "learning_rate": 8.365655249949355e-06,
+      "loss": 0.5718,
+      "step": 10865
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.425060577013328,
+      "learning_rate": 8.363991549218525e-06,
+      "loss": 0.4823,
+      "step": 10866
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 22.13968438827088,
+      "learning_rate": 8.362327895014821e-06,
+      "loss": 0.6008,
+      "step": 10867
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.737013640701857,
+      "learning_rate": 8.360664287385552e-06,
+      "loss": 0.6384,
+      "step": 10868
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.90273808471432,
+      "learning_rate": 8.359000726378033e-06,
+      "loss": 0.6097,
+      "step": 10869
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.962587155541268,
+      "learning_rate": 8.35733721203957e-06,
+      "loss": 0.5433,
+      "step": 10870
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.109271828844765,
+      "learning_rate": 8.355673744417479e-06,
+      "loss": 0.5872,
+      "step": 10871
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.762518960996564,
+      "learning_rate": 8.354010323559065e-06,
+      "loss": 0.567,
+      "step": 10872
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.263026035887965,
+      "learning_rate": 8.352346949511632e-06,
+      "loss": 0.5727,
+      "step": 10873
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.912423635715026,
+      "learning_rate": 8.35068362232249e-06,
+      "loss": 0.5527,
+      "step": 10874
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.301853715817181,
+      "learning_rate": 8.34902034203894e-06,
+      "loss": 0.5117,
+      "step": 10875
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.68786742870043,
+      "learning_rate": 8.347357108708284e-06,
+      "loss": 0.6024,
+      "step": 10876
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.077632427469204,
+      "learning_rate": 8.345693922377829e-06,
+      "loss": 0.5627,
+      "step": 10877
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 26.141580930151132,
+      "learning_rate": 8.344030783094869e-06,
+      "loss": 0.5991,
+      "step": 10878
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.240592496703146,
+      "learning_rate": 8.342367690906702e-06,
+      "loss": 0.521,
+      "step": 10879
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.5548296749909,
+      "learning_rate": 8.340704645860632e-06,
+      "loss": 0.6031,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.405925888531877,
+      "learning_rate": 8.339041648003952e-06,
+      "loss": 0.5575,
+      "step": 10881
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.430878572772418,
+      "learning_rate": 8.337378697383956e-06,
+      "loss": 0.6364,
+      "step": 10882
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 27.631832317885262,
+      "learning_rate": 8.335715794047933e-06,
+      "loss": 0.6396,
+      "step": 10883
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.86446759722926,
+      "learning_rate": 8.33405293804318e-06,
+      "loss": 0.5707,
+      "step": 10884
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.897243542465798,
+      "learning_rate": 8.332390129416988e-06,
+      "loss": 0.5846,
+      "step": 10885
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.279694136198131,
+      "learning_rate": 8.330727368216647e-06,
+      "loss": 0.6303,
+      "step": 10886
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.776156146395545,
+      "learning_rate": 8.329064654489442e-06,
+      "loss": 0.488,
+      "step": 10887
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.848680261969525,
+      "learning_rate": 8.327401988282661e-06,
+      "loss": 0.5125,
+      "step": 10888
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.010487715179941,
+      "learning_rate": 8.325739369643588e-06,
+      "loss": 0.5131,
+      "step": 10889
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.783162116317529,
+      "learning_rate": 8.324076798619508e-06,
+      "loss": 0.5867,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.27836650840371,
+      "learning_rate": 8.322414275257707e-06,
+      "loss": 0.6239,
+      "step": 10891
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.562074142205898,
+      "learning_rate": 8.320751799605458e-06,
+      "loss": 0.5891,
+      "step": 10892
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.156873202026848,
+      "learning_rate": 8.319089371710048e-06,
+      "loss": 0.5657,
+      "step": 10893
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.307776239000717,
+      "learning_rate": 8.317426991618751e-06,
+      "loss": 0.5497,
+      "step": 10894
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.456159758357202,
+      "learning_rate": 8.315764659378845e-06,
+      "loss": 0.6096,
+      "step": 10895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.34234110323499,
+      "learning_rate": 8.314102375037611e-06,
+      "loss": 0.5512,
+      "step": 10896
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.335310418143504,
+      "learning_rate": 8.312440138642319e-06,
+      "loss": 0.5436,
+      "step": 10897
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.981360918954534,
+      "learning_rate": 8.310777950240238e-06,
+      "loss": 0.5684,
+      "step": 10898
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.586375878749468,
+      "learning_rate": 8.309115809878646e-06,
+      "loss": 0.5508,
+      "step": 10899
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.032988858217905,
+      "learning_rate": 8.307453717604811e-06,
+      "loss": 0.5258,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.502474815603955,
+      "learning_rate": 8.305791673466006e-06,
+      "loss": 0.5734,
+      "step": 10901
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.328851826862802,
+      "learning_rate": 8.304129677509491e-06,
+      "loss": 0.5672,
+      "step": 10902
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.237094704980088,
+      "learning_rate": 8.302467729782535e-06,
+      "loss": 0.5341,
+      "step": 10903
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.74807560803915,
+      "learning_rate": 8.300805830332404e-06,
+      "loss": 0.6041,
+      "step": 10904
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.99702414860258,
+      "learning_rate": 8.299143979206363e-06,
+      "loss": 0.5708,
+      "step": 10905
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 27.44408096202091,
+      "learning_rate": 8.297482176451675e-06,
+      "loss": 0.5676,
+      "step": 10906
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.123638509287442,
+      "learning_rate": 8.295820422115594e-06,
+      "loss": 0.5608,
+      "step": 10907
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.28172198974164,
+      "learning_rate": 8.294158716245381e-06,
+      "loss": 0.5305,
+      "step": 10908
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.354148299201572,
+      "learning_rate": 8.2924970588883e-06,
+      "loss": 0.5862,
+      "step": 10909
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 22.55309224210255,
+      "learning_rate": 8.290835450091601e-06,
+      "loss": 0.5633,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.20277382848681,
+      "learning_rate": 8.289173889902546e-06,
+      "loss": 0.5293,
+      "step": 10911
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.133885749202733,
+      "learning_rate": 8.287512378368384e-06,
+      "loss": 0.5776,
+      "step": 10912
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.523097200248724,
+      "learning_rate": 8.285850915536367e-06,
+      "loss": 0.5001,
+      "step": 10913
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.10297550986689,
+      "learning_rate": 8.284189501453745e-06,
+      "loss": 0.5975,
+      "step": 10914
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.999201031243887,
+      "learning_rate": 8.28252813616777e-06,
+      "loss": 0.5366,
+      "step": 10915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.51660198564347,
+      "learning_rate": 8.280866819725695e-06,
+      "loss": 0.5432,
+      "step": 10916
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.794595704593096,
+      "learning_rate": 8.279205552174758e-06,
+      "loss": 0.6163,
+      "step": 10917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.992468481262,
+      "learning_rate": 8.27754433356221e-06,
+      "loss": 0.6185,
+      "step": 10918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.496271285399576,
+      "learning_rate": 8.275883163935292e-06,
+      "loss": 0.5806,
+      "step": 10919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 22.46996722452469,
+      "learning_rate": 8.274222043341247e-06,
+      "loss": 0.6158,
+      "step": 10920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.716957907184327,
+      "learning_rate": 8.272560971827323e-06,
+      "loss": 0.5498,
+      "step": 10921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 20.150285231556925,
+      "learning_rate": 8.270899949440751e-06,
+      "loss": 0.4829,
+      "step": 10922
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.496475492785464,
+      "learning_rate": 8.269238976228771e-06,
+      "loss": 0.4964,
+      "step": 10923
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.502838690621552,
+      "learning_rate": 8.267578052238624e-06,
+      "loss": 0.5263,
+      "step": 10924
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.25713961797008,
+      "learning_rate": 8.265917177517545e-06,
+      "loss": 0.5561,
+      "step": 10925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 20.398053843916482,
+      "learning_rate": 8.264256352112768e-06,
+      "loss": 0.5813,
+      "step": 10926
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.41464395443639,
+      "learning_rate": 8.262595576071521e-06,
+      "loss": 0.4747,
+      "step": 10927
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.847317025928152,
+      "learning_rate": 8.260934849441042e-06,
+      "loss": 0.4976,
+      "step": 10928
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.635553731026775,
+      "learning_rate": 8.259274172268556e-06,
+      "loss": 0.5807,
+      "step": 10929
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 23.640106078867394,
+      "learning_rate": 8.2576135446013e-06,
+      "loss": 0.6105,
+      "step": 10930
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.383362906505662,
+      "learning_rate": 8.255952966486491e-06,
+      "loss": 0.6145,
+      "step": 10931
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.106375030346186,
+      "learning_rate": 8.25429243797136e-06,
+      "loss": 0.524,
+      "step": 10932
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.64138129670902,
+      "learning_rate": 8.252631959103129e-06,
+      "loss": 0.6047,
+      "step": 10933
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.419194950835276,
+      "learning_rate": 8.250971529929026e-06,
+      "loss": 0.5218,
+      "step": 10934
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 20.669199615435286,
+      "learning_rate": 8.249311150496271e-06,
+      "loss": 0.5736,
+      "step": 10935
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 26.472133173678266,
+      "learning_rate": 8.247650820852078e-06,
+      "loss": 0.6087,
+      "step": 10936
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 29.620755318065687,
+      "learning_rate": 8.245990541043672e-06,
+      "loss": 0.5594,
+      "step": 10937
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 25.733572474892174,
+      "learning_rate": 8.24433031111827e-06,
+      "loss": 0.5994,
+      "step": 10938
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.432677710961624,
+      "learning_rate": 8.242670131123085e-06,
+      "loss": 0.5492,
+      "step": 10939
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.384402292824348,
+      "learning_rate": 8.241010001105338e-06,
+      "loss": 0.6059,
+      "step": 10940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 20.774879041869337,
+      "learning_rate": 8.239349921112235e-06,
+      "loss": 0.5671,
+      "step": 10941
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.625399671269154,
+      "learning_rate": 8.237689891190988e-06,
+      "loss": 0.5578,
+      "step": 10942
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.42894168015925,
+      "learning_rate": 8.236029911388811e-06,
+      "loss": 0.5748,
+      "step": 10943
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.770875552165393,
+      "learning_rate": 8.234369981752913e-06,
+      "loss": 0.5848,
+      "step": 10944
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.350162655744025,
+      "learning_rate": 8.232710102330502e-06,
+      "loss": 0.5975,
+      "step": 10945
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.78063106926223,
+      "learning_rate": 8.231050273168776e-06,
+      "loss": 0.562,
+      "step": 10946
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.502297131688767,
+      "learning_rate": 8.229390494314949e-06,
+      "loss": 0.5946,
+      "step": 10947
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.264131519330666,
+      "learning_rate": 8.227730765816216e-06,
+      "loss": 0.5049,
+      "step": 10948
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.951478238830493,
+      "learning_rate": 8.226071087719789e-06,
+      "loss": 0.5444,
+      "step": 10949
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.942149590621575,
+      "learning_rate": 8.224411460072863e-06,
+      "loss": 0.5804,
+      "step": 10950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.968534621022663,
+      "learning_rate": 8.222751882922632e-06,
+      "loss": 0.6098,
+      "step": 10951
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.37887084960751,
+      "learning_rate": 8.221092356316298e-06,
+      "loss": 0.5557,
+      "step": 10952
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.608654212247025,
+      "learning_rate": 8.21943288030106e-06,
+      "loss": 0.6502,
+      "step": 10953
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.563353896621614,
+      "learning_rate": 8.217773454924105e-06,
+      "loss": 0.5859,
+      "step": 10954
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.051143185334322,
+      "learning_rate": 8.216114080232635e-06,
+      "loss": 0.6179,
+      "step": 10955
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.33739395143801,
+      "learning_rate": 8.214454756273836e-06,
+      "loss": 0.5227,
+      "step": 10956
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.930093385983994,
+      "learning_rate": 8.2127954830949e-06,
+      "loss": 0.571,
+      "step": 10957
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 27.78895490859864,
+      "learning_rate": 8.21113626074301e-06,
+      "loss": 0.6368,
+      "step": 10958
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.97172718471198,
+      "learning_rate": 8.20947708926536e-06,
+      "loss": 0.5462,
+      "step": 10959
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.152619516337708,
+      "learning_rate": 8.20781796870914e-06,
+      "loss": 0.4816,
+      "step": 10960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.294996336195624,
+      "learning_rate": 8.20615889912152e-06,
+      "loss": 0.5204,
+      "step": 10961
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.888929213512366,
+      "learning_rate": 8.204499880549699e-06,
+      "loss": 0.5903,
+      "step": 10962
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.16416867859061,
+      "learning_rate": 8.202840913040847e-06,
+      "loss": 0.593,
+      "step": 10963
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 34.83630193385815,
+      "learning_rate": 8.201181996642147e-06,
+      "loss": 0.6155,
+      "step": 10964
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.768592730950118,
+      "learning_rate": 8.199523131400783e-06,
+      "loss": 0.5226,
+      "step": 10965
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.132701464724713,
+      "learning_rate": 8.197864317363926e-06,
+      "loss": 0.5371,
+      "step": 10966
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.337489980956185,
+      "learning_rate": 8.19620555457875e-06,
+      "loss": 0.5384,
+      "step": 10967
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.497313984422355,
+      "learning_rate": 8.194546843092438e-06,
+      "loss": 0.5255,
+      "step": 10968
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.670657211968459,
+      "learning_rate": 8.192888182952155e-06,
+      "loss": 0.4893,
+      "step": 10969
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.008764870268358,
+      "learning_rate": 8.191229574205078e-06,
+      "loss": 0.5432,
+      "step": 10970
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.819143573333196,
+      "learning_rate": 8.18957101689837e-06,
+      "loss": 0.6081,
+      "step": 10971
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 26.228774009600517,
+      "learning_rate": 8.187912511079205e-06,
+      "loss": 0.5523,
+      "step": 10972
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.20772256216007,
+      "learning_rate": 8.186254056794747e-06,
+      "loss": 0.5818,
+      "step": 10973
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.06983739156714,
+      "learning_rate": 8.184595654092161e-06,
+      "loss": 0.5562,
+      "step": 10974
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 25.424231585890986,
+      "learning_rate": 8.182937303018619e-06,
+      "loss": 0.6361,
+      "step": 10975
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.16481744528762,
+      "learning_rate": 8.181279003621274e-06,
+      "loss": 0.5153,
+      "step": 10976
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.802604154974652,
+      "learning_rate": 8.179620755947287e-06,
+      "loss": 0.5733,
+      "step": 10977
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.742537468560272,
+      "learning_rate": 8.177962560043824e-06,
+      "loss": 0.6117,
+      "step": 10978
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 28.38632288276588,
+      "learning_rate": 8.17630441595804e-06,
+      "loss": 0.517,
+      "step": 10979
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.923263957167489,
+      "learning_rate": 8.17464632373709e-06,
+      "loss": 0.5239,
+      "step": 10980
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.490897551455543,
+      "learning_rate": 8.172988283428132e-06,
+      "loss": 0.5829,
+      "step": 10981
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.940319390755732,
+      "learning_rate": 8.171330295078318e-06,
+      "loss": 0.5117,
+      "step": 10982
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.43473985982462,
+      "learning_rate": 8.1696723587348e-06,
+      "loss": 0.6099,
+      "step": 10983
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.632587529878155,
+      "learning_rate": 8.168014474444731e-06,
+      "loss": 0.5982,
+      "step": 10984
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.675668228564096,
+      "learning_rate": 8.166356642255258e-06,
+      "loss": 0.584,
+      "step": 10985
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.01815195830891,
+      "learning_rate": 8.164698862213527e-06,
+      "loss": 0.6065,
+      "step": 10986
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.87360210561644,
+      "learning_rate": 8.16304113436669e-06,
+      "loss": 0.5353,
+      "step": 10987
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.041664347683344,
+      "learning_rate": 8.161383458761887e-06,
+      "loss": 0.5885,
+      "step": 10988
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.997703264741872,
+      "learning_rate": 8.159725835446266e-06,
+      "loss": 0.5921,
+      "step": 10989
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.729567331896416,
+      "learning_rate": 8.15806826446696e-06,
+      "loss": 0.5812,
+      "step": 10990
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.69375402328852,
+      "learning_rate": 8.156410745871119e-06,
+      "loss": 0.542,
+      "step": 10991
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.188270581676644,
+      "learning_rate": 8.154753279705875e-06,
+      "loss": 0.5461,
+      "step": 10992
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 24.136801629850993,
+      "learning_rate": 8.15309586601837e-06,
+      "loss": 0.5889,
+      "step": 10993
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.047274211675738,
+      "learning_rate": 8.151438504855742e-06,
+      "loss": 0.5705,
+      "step": 10994
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.155994891960784,
+      "learning_rate": 8.149781196265117e-06,
+      "loss": 0.5442,
+      "step": 10995
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.199399782235506,
+      "learning_rate": 8.148123940293632e-06,
+      "loss": 0.5261,
+      "step": 10996
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.092233468394266,
+      "learning_rate": 8.146466736988422e-06,
+      "loss": 0.5723,
+      "step": 10997
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.21096789645126,
+      "learning_rate": 8.144809586396608e-06,
+      "loss": 0.5656,
+      "step": 10998
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.233921976553873,
+      "learning_rate": 8.143152488565332e-06,
+      "loss": 0.5529,
+      "step": 10999
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.137995196396947,
+      "learning_rate": 8.141495443541708e-06,
+      "loss": 0.5669,
+      "step": 11000
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.661002369888248,
+      "learning_rate": 8.139838451372868e-06,
+      "loss": 0.534,
+      "step": 11001
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 25.811475719029517,
+      "learning_rate": 8.138181512105931e-06,
+      "loss": 0.5449,
+      "step": 11002
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.560522016809891,
+      "learning_rate": 8.136524625788026e-06,
+      "loss": 0.4659,
+      "step": 11003
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.104156449874377,
+      "learning_rate": 8.134867792466272e-06,
+      "loss": 0.5013,
+      "step": 11004
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.90485664669444,
+      "learning_rate": 8.133211012187783e-06,
+      "loss": 0.56,
+      "step": 11005
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.519362035032106,
+      "learning_rate": 8.131554284999683e-06,
+      "loss": 0.5678,
+      "step": 11006
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.27402890749563,
+      "learning_rate": 8.129897610949086e-06,
+      "loss": 0.502,
+      "step": 11007
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 25.254044915782583,
+      "learning_rate": 8.128240990083104e-06,
+      "loss": 0.6059,
+      "step": 11008
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.026716506973344,
+      "learning_rate": 8.12658442244886e-06,
+      "loss": 0.6353,
+      "step": 11009
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.869026824952083,
+      "learning_rate": 8.124927908093455e-06,
+      "loss": 0.4833,
+      "step": 11010
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 22.541212007840347,
+      "learning_rate": 8.123271447064001e-06,
+      "loss": 0.5512,
+      "step": 11011
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 24.436588693539342,
+      "learning_rate": 8.121615039407613e-06,
+      "loss": 0.5784,
+      "step": 11012
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.515478452900282,
+      "learning_rate": 8.119958685171392e-06,
+      "loss": 0.5607,
+      "step": 11013
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.27485765158441,
+      "learning_rate": 8.11830238440245e-06,
+      "loss": 0.5034,
+      "step": 11014
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.417919135522526,
+      "learning_rate": 8.116646137147884e-06,
+      "loss": 0.6027,
+      "step": 11015
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.62591573352725,
+      "learning_rate": 8.114989943454801e-06,
+      "loss": 0.5586,
+      "step": 11016
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.09641837558083,
+      "learning_rate": 8.113333803370297e-06,
+      "loss": 0.5503,
+      "step": 11017
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.03796482806511,
+      "learning_rate": 8.111677716941481e-06,
+      "loss": 0.5911,
+      "step": 11018
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.330280045255513,
+      "learning_rate": 8.110021684215448e-06,
+      "loss": 0.545,
+      "step": 11019
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 24.73453217201637,
+      "learning_rate": 8.10836570523929e-06,
+      "loss": 0.5973,
+      "step": 11020
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.325531198520604,
+      "learning_rate": 8.106709780060103e-06,
+      "loss": 0.5642,
+      "step": 11021
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 22.658649594400202,
+      "learning_rate": 8.105053908724985e-06,
+      "loss": 0.5774,
+      "step": 11022
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.606361811712446,
+      "learning_rate": 8.103398091281023e-06,
+      "loss": 0.5776,
+      "step": 11023
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.57523617190926,
+      "learning_rate": 8.101742327775312e-06,
+      "loss": 0.5409,
+      "step": 11024
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.01286506504692,
+      "learning_rate": 8.10008661825494e-06,
+      "loss": 0.5582,
+      "step": 11025
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.546346017841536,
+      "learning_rate": 8.09843096276699e-06,
+      "loss": 0.5599,
+      "step": 11026
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.43382008611894,
+      "learning_rate": 8.096775361358552e-06,
+      "loss": 0.5823,
+      "step": 11027
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.18612216552755,
+      "learning_rate": 8.095119814076711e-06,
+      "loss": 0.5712,
+      "step": 11028
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.592964890005003,
+      "learning_rate": 8.093464320968547e-06,
+      "loss": 0.5272,
+      "step": 11029
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 25.499410651680005,
+      "learning_rate": 8.09180888208114e-06,
+      "loss": 0.5841,
+      "step": 11030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.41703335771521,
+      "learning_rate": 8.090153497461573e-06,
+      "loss": 0.6221,
+      "step": 11031
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.54118439577805,
+      "learning_rate": 8.088498167156926e-06,
+      "loss": 0.4778,
+      "step": 11032
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.262342527585577,
+      "learning_rate": 8.086842891214274e-06,
+      "loss": 0.5956,
+      "step": 11033
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.358215950784594,
+      "learning_rate": 8.085187669680687e-06,
+      "loss": 0.5643,
+      "step": 11034
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.09042014762866,
+      "learning_rate": 8.083532502603243e-06,
+      "loss": 0.5982,
+      "step": 11035
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.472888105967124,
+      "learning_rate": 8.081877390029013e-06,
+      "loss": 0.5203,
+      "step": 11036
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.36259903488524,
+      "learning_rate": 8.08022233200507e-06,
+      "loss": 0.615,
+      "step": 11037
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.797824194055696,
+      "learning_rate": 8.078567328578482e-06,
+      "loss": 0.5085,
+      "step": 11038
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.77186851093607,
+      "learning_rate": 8.076912379796314e-06,
+      "loss": 0.5671,
+      "step": 11039
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.64235028247073,
+      "learning_rate": 8.07525748570563e-06,
+      "loss": 0.5634,
+      "step": 11040
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 30.995082032506957,
+      "learning_rate": 8.0736026463535e-06,
+      "loss": 0.6884,
+      "step": 11041
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.509702462811813,
+      "learning_rate": 8.071947861786982e-06,
+      "loss": 0.5493,
+      "step": 11042
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.382712219564755,
+      "learning_rate": 8.070293132053143e-06,
+      "loss": 0.5787,
+      "step": 11043
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.06460155802515,
+      "learning_rate": 8.068638457199037e-06,
+      "loss": 0.5473,
+      "step": 11044
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.342234244697979,
+      "learning_rate": 8.066983837271722e-06,
+      "loss": 0.5552,
+      "step": 11045
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.61376142491823,
+      "learning_rate": 8.065329272318255e-06,
+      "loss": 0.6151,
+      "step": 11046
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.626125978110455,
+      "learning_rate": 8.063674762385691e-06,
+      "loss": 0.553,
+      "step": 11047
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.643831769416725,
+      "learning_rate": 8.06202030752109e-06,
+      "loss": 0.5936,
+      "step": 11048
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 26.662655345895203,
+      "learning_rate": 8.06036590777149e-06,
+      "loss": 0.5426,
+      "step": 11049
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.247502883389618,
+      "learning_rate": 8.058711563183955e-06,
+      "loss": 0.4977,
+      "step": 11050
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.289681275470457,
+      "learning_rate": 8.057057273805525e-06,
+      "loss": 0.5044,
+      "step": 11051
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 28.097842980216058,
+      "learning_rate": 8.055403039683247e-06,
+      "loss": 0.661,
+      "step": 11052
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 31.535869156552145,
+      "learning_rate": 8.053748860864174e-06,
+      "loss": 0.6149,
+      "step": 11053
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.49580416106502,
+      "learning_rate": 8.052094737395343e-06,
+      "loss": 0.5196,
+      "step": 11054
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.72396495256864,
+      "learning_rate": 8.050440669323796e-06,
+      "loss": 0.5294,
+      "step": 11055
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.234785005909544,
+      "learning_rate": 8.048786656696579e-06,
+      "loss": 0.6251,
+      "step": 11056
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.220223077620975,
+      "learning_rate": 8.047132699560725e-06,
+      "loss": 0.5548,
+      "step": 11057
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.359246668096613,
+      "learning_rate": 8.04547879796328e-06,
+      "loss": 0.57,
+      "step": 11058
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.67673597067759,
+      "learning_rate": 8.04382495195127e-06,
+      "loss": 0.5596,
+      "step": 11059
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.909389040797807,
+      "learning_rate": 8.042171161571734e-06,
+      "loss": 0.5106,
+      "step": 11060
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.535049992591745,
+      "learning_rate": 8.040517426871703e-06,
+      "loss": 0.5295,
+      "step": 11061
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 21.898674181429076,
+      "learning_rate": 8.038863747898214e-06,
+      "loss": 0.5764,
+      "step": 11062
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.233164874520643,
+      "learning_rate": 8.037210124698294e-06,
+      "loss": 0.5359,
+      "step": 11063
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.149139086680933,
+      "learning_rate": 8.03555655731897e-06,
+      "loss": 0.4864,
+      "step": 11064
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 21.420859164603485,
+      "learning_rate": 8.033903045807262e-06,
+      "loss": 0.5552,
+      "step": 11065
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 26.332369953580862,
+      "learning_rate": 8.032249590210206e-06,
+      "loss": 0.5435,
+      "step": 11066
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.057074256631132,
+      "learning_rate": 8.030596190574818e-06,
+      "loss": 0.592,
+      "step": 11067
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.374759156789157,
+      "learning_rate": 8.028942846948126e-06,
+      "loss": 0.5417,
+      "step": 11068
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.16419267022803,
+      "learning_rate": 8.027289559377145e-06,
+      "loss": 0.5208,
+      "step": 11069
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.173438081687703,
+      "learning_rate": 8.025636327908895e-06,
+      "loss": 0.5357,
+      "step": 11070
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 21.554575367103908,
+      "learning_rate": 8.02398315259039e-06,
+      "loss": 0.5781,
+      "step": 11071
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.602296407876821,
+      "learning_rate": 8.02233003346865e-06,
+      "loss": 0.4713,
+      "step": 11072
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.593174744849943,
+      "learning_rate": 8.02067697059069e-06,
+      "loss": 0.6106,
+      "step": 11073
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 27.016106065666666,
+      "learning_rate": 8.019023964003513e-06,
+      "loss": 0.5307,
+      "step": 11074
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.047224747190242,
+      "learning_rate": 8.01737101375414e-06,
+      "loss": 0.5816,
+      "step": 11075
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.723276295792006,
+      "learning_rate": 8.015718119889576e-06,
+      "loss": 0.5206,
+      "step": 11076
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.158261089440472,
+      "learning_rate": 8.014065282456829e-06,
+      "loss": 0.5438,
+      "step": 11077
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.76313922040878,
+      "learning_rate": 8.0124125015029e-06,
+      "loss": 0.5314,
+      "step": 11078
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.643851173575644,
+      "learning_rate": 8.010759777074796e-06,
+      "loss": 0.622,
+      "step": 11079
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.529692492113387,
+      "learning_rate": 8.009107109219522e-06,
+      "loss": 0.5564,
+      "step": 11080
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.519515489222094,
+      "learning_rate": 8.007454497984078e-06,
+      "loss": 0.5922,
+      "step": 11081
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.377649951148507,
+      "learning_rate": 8.005801943415462e-06,
+      "loss": 0.5335,
+      "step": 11082
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 25.989080296956804,
+      "learning_rate": 8.004149445560675e-06,
+      "loss": 0.5507,
+      "step": 11083
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 26.98497781704211,
+      "learning_rate": 8.002497004466703e-06,
+      "loss": 0.6189,
+      "step": 11084
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.69837751890889,
+      "learning_rate": 8.000844620180553e-06,
+      "loss": 0.6157,
+      "step": 11085
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.52657075512683,
+      "learning_rate": 7.999192292749209e-06,
+      "loss": 0.5011,
+      "step": 11086
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.089404243929433,
+      "learning_rate": 7.997540022219672e-06,
+      "loss": 0.6038,
+      "step": 11087
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.70165493375579,
+      "learning_rate": 7.995887808638921e-06,
+      "loss": 0.6156,
+      "step": 11088
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 21.615571556863518,
+      "learning_rate": 7.99423565205395e-06,
+      "loss": 0.5861,
+      "step": 11089
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.954431413536655,
+      "learning_rate": 7.99258355251174e-06,
+      "loss": 0.5519,
+      "step": 11090
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.510564341840514,
+      "learning_rate": 7.990931510059281e-06,
+      "loss": 0.5643,
+      "step": 11091
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 26.357858273602496,
+      "learning_rate": 7.989279524743559e-06,
+      "loss": 0.532,
+      "step": 11092
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 22.877190438779387,
+      "learning_rate": 7.987627596611546e-06,
+      "loss": 0.625,
+      "step": 11093
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 28.456516538218356,
+      "learning_rate": 7.985975725710229e-06,
+      "loss": 0.6112,
+      "step": 11094
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.390847832925022,
+      "learning_rate": 7.984323912086584e-06,
+      "loss": 0.5136,
+      "step": 11095
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 28.019716719828942,
+      "learning_rate": 7.982672155787586e-06,
+      "loss": 0.58,
+      "step": 11096
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.003477062195799,
+      "learning_rate": 7.981020456860215e-06,
+      "loss": 0.583,
+      "step": 11097
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 25.403469419461544,
+      "learning_rate": 7.979368815351441e-06,
+      "loss": 0.5502,
+      "step": 11098
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.524190774060003,
+      "learning_rate": 7.977717231308233e-06,
+      "loss": 0.4403,
+      "step": 11099
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 27.22380789500585,
+      "learning_rate": 7.976065704777566e-06,
+      "loss": 0.5688,
+      "step": 11100
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.66113094994034,
+      "learning_rate": 7.974414235806403e-06,
+      "loss": 0.5801,
+      "step": 11101
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.72633698551709,
+      "learning_rate": 7.972762824441722e-06,
+      "loss": 0.5958,
+      "step": 11102
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.229137339523575,
+      "learning_rate": 7.971111470730474e-06,
+      "loss": 0.5563,
+      "step": 11103
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 21.096566558281705,
+      "learning_rate": 7.96946017471963e-06,
+      "loss": 0.565,
+      "step": 11104
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 21.96685706485057,
+      "learning_rate": 7.967808936456149e-06,
+      "loss": 0.5936,
+      "step": 11105
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.269102688477492,
+      "learning_rate": 7.966157755986994e-06,
+      "loss": 0.6045,
+      "step": 11106
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.209484271912398,
+      "learning_rate": 7.964506633359128e-06,
+      "loss": 0.5603,
+      "step": 11107
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.844999399586634,
+      "learning_rate": 7.962855568619499e-06,
+      "loss": 0.5773,
+      "step": 11108
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.72385245923531,
+      "learning_rate": 7.961204561815063e-06,
+      "loss": 0.4946,
+      "step": 11109
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.107827367319523,
+      "learning_rate": 7.95955361299278e-06,
+      "loss": 0.5071,
+      "step": 11110
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.403746216851392,
+      "learning_rate": 7.957902722199596e-06,
+      "loss": 0.6051,
+      "step": 11111
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 35.29953233308446,
+      "learning_rate": 7.956251889482467e-06,
+      "loss": 0.5955,
+      "step": 11112
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.51073173821916,
+      "learning_rate": 7.954601114888338e-06,
+      "loss": 0.5469,
+      "step": 11113
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.361051458775115,
+      "learning_rate": 7.952950398464156e-06,
+      "loss": 0.5672,
+      "step": 11114
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.68827777891515,
+      "learning_rate": 7.951299740256865e-06,
+      "loss": 0.585,
+      "step": 11115
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.981508707375458,
+      "learning_rate": 7.949649140313413e-06,
+      "loss": 0.5476,
+      "step": 11116
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.699210480390972,
+      "learning_rate": 7.947998598680743e-06,
+      "loss": 0.5574,
+      "step": 11117
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 22.554949270353937,
+      "learning_rate": 7.946348115405788e-06,
+      "loss": 0.5768,
+      "step": 11118
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.234936766024894,
+      "learning_rate": 7.944697690535493e-06,
+      "loss": 0.5614,
+      "step": 11119
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.38727312207389,
+      "learning_rate": 7.943047324116793e-06,
+      "loss": 0.5479,
+      "step": 11120
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.145180182217953,
+      "learning_rate": 7.941397016196622e-06,
+      "loss": 0.5471,
+      "step": 11121
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.59424449322826,
+      "learning_rate": 7.939746766821921e-06,
+      "loss": 0.6819,
+      "step": 11122
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 22.470554431222713,
+      "learning_rate": 7.938096576039613e-06,
+      "loss": 0.5799,
+      "step": 11123
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.87952576742975,
+      "learning_rate": 7.93644644389663e-06,
+      "loss": 0.5676,
+      "step": 11124
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.474331828724617,
+      "learning_rate": 7.934796370439907e-06,
+      "loss": 0.5295,
+      "step": 11125
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.151053640199148,
+      "learning_rate": 7.933146355716368e-06,
+      "loss": 0.5266,
+      "step": 11126
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.41669492814009,
+      "learning_rate": 7.931496399772935e-06,
+      "loss": 0.5283,
+      "step": 11127
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.02942631083082,
+      "learning_rate": 7.929846502656533e-06,
+      "loss": 0.5478,
+      "step": 11128
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.93103972345277,
+      "learning_rate": 7.928196664414088e-06,
+      "loss": 0.5782,
+      "step": 11129
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.59302446290554,
+      "learning_rate": 7.926546885092515e-06,
+      "loss": 0.5762,
+      "step": 11130
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.436897528615656,
+      "learning_rate": 7.92489716473874e-06,
+      "loss": 0.5658,
+      "step": 11131
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.341677152386612,
+      "learning_rate": 7.923247503399674e-06,
+      "loss": 0.5375,
+      "step": 11132
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.199581514459307,
+      "learning_rate": 7.921597901122234e-06,
+      "loss": 0.6023,
+      "step": 11133
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.14620217712568,
+      "learning_rate": 7.919948357953332e-06,
+      "loss": 0.5263,
+      "step": 11134
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.87994058848624,
+      "learning_rate": 7.918298873939882e-06,
+      "loss": 0.6351,
+      "step": 11135
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.894872621895544,
+      "learning_rate": 7.916649449128799e-06,
+      "loss": 0.5342,
+      "step": 11136
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.574830035472903,
+      "learning_rate": 7.915000083566981e-06,
+      "loss": 0.5463,
+      "step": 11137
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.873451409610581,
+      "learning_rate": 7.913350777301342e-06,
+      "loss": 0.5196,
+      "step": 11138
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.12385858459081,
+      "learning_rate": 7.911701530378789e-06,
+      "loss": 0.5146,
+      "step": 11139
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 25.481924751387364,
+      "learning_rate": 7.910052342846219e-06,
+      "loss": 0.5504,
+      "step": 11140
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.437353532569539,
+      "learning_rate": 7.908403214750544e-06,
+      "loss": 0.5345,
+      "step": 11141
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.383152699516856,
+      "learning_rate": 7.906754146138655e-06,
+      "loss": 0.5909,
+      "step": 11142
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 36.781543884667066,
+      "learning_rate": 7.905105137057451e-06,
+      "loss": 0.5767,
+      "step": 11143
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 22.89474109448701,
+      "learning_rate": 7.903456187553833e-06,
+      "loss": 0.6353,
+      "step": 11144
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.728992731936774,
+      "learning_rate": 7.901807297674693e-06,
+      "loss": 0.5532,
+      "step": 11145
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.184728292008584,
+      "learning_rate": 7.900158467466932e-06,
+      "loss": 0.4741,
+      "step": 11146
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.181395474145667,
+      "learning_rate": 7.89850969697743e-06,
+      "loss": 0.5903,
+      "step": 11147
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.120405041472555,
+      "learning_rate": 7.896860986253086e-06,
+      "loss": 0.5352,
+      "step": 11148
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.06258247740679,
+      "learning_rate": 7.89521233534078e-06,
+      "loss": 0.5398,
+      "step": 11149
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.13804934630115,
+      "learning_rate": 7.89356374428741e-06,
+      "loss": 0.5708,
+      "step": 11150
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.187892427720595,
+      "learning_rate": 7.891915213139855e-06,
+      "loss": 0.5356,
+      "step": 11151
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.343560072420896,
+      "learning_rate": 7.890266741944995e-06,
+      "loss": 0.5867,
+      "step": 11152
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.77213299013415,
+      "learning_rate": 7.888618330749715e-06,
+      "loss": 0.5711,
+      "step": 11153
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.805220979143153,
+      "learning_rate": 7.886969979600894e-06,
+      "loss": 0.5862,
+      "step": 11154
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.429068554675853,
+      "learning_rate": 7.885321688545412e-06,
+      "loss": 0.5516,
+      "step": 11155
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.692821746198884,
+      "learning_rate": 7.883673457630144e-06,
+      "loss": 0.503,
+      "step": 11156
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.367965746136644,
+      "learning_rate": 7.882025286901965e-06,
+      "loss": 0.5323,
+      "step": 11157
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 23.50679528370471,
+      "learning_rate": 7.880377176407749e-06,
+      "loss": 0.5922,
+      "step": 11158
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.70563036251502,
+      "learning_rate": 7.878729126194362e-06,
+      "loss": 0.5291,
+      "step": 11159
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.653253196150233,
+      "learning_rate": 7.877081136308681e-06,
+      "loss": 0.5786,
+      "step": 11160
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.693096101244436,
+      "learning_rate": 7.875433206797573e-06,
+      "loss": 0.6015,
+      "step": 11161
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.594477678776837,
+      "learning_rate": 7.873785337707897e-06,
+      "loss": 0.5464,
+      "step": 11162
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.992648678118947,
+      "learning_rate": 7.872137529086525e-06,
+      "loss": 0.594,
+      "step": 11163
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.856251143566597,
+      "learning_rate": 7.870489780980315e-06,
+      "loss": 0.5132,
+      "step": 11164
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.567361704222945,
+      "learning_rate": 7.86884209343613e-06,
+      "loss": 0.5292,
+      "step": 11165
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 23.6082497971756,
+      "learning_rate": 7.867194466500831e-06,
+      "loss": 0.5751,
+      "step": 11166
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 23.46617471518683,
+      "learning_rate": 7.865546900221274e-06,
+      "loss": 0.5088,
+      "step": 11167
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.359587600775672,
+      "learning_rate": 7.863899394644312e-06,
+      "loss": 0.6115,
+      "step": 11168
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.852719261937743,
+      "learning_rate": 7.862251949816803e-06,
+      "loss": 0.5957,
+      "step": 11169
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.061317520228739,
+      "learning_rate": 7.860604565785596e-06,
+      "loss": 0.5912,
+      "step": 11170
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.93777975066035,
+      "learning_rate": 7.85895724259755e-06,
+      "loss": 0.5291,
+      "step": 11171
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 25.539383992995287,
+      "learning_rate": 7.857309980299502e-06,
+      "loss": 0.5292,
+      "step": 11172
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.411902252602395,
+      "learning_rate": 7.855662778938306e-06,
+      "loss": 0.5422,
+      "step": 11173
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.621579744342313,
+      "learning_rate": 7.854015638560804e-06,
+      "loss": 0.5601,
+      "step": 11174
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.57756516314625,
+      "learning_rate": 7.852368559213846e-06,
+      "loss": 0.6073,
+      "step": 11175
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.900995452002077,
+      "learning_rate": 7.850721540944268e-06,
+      "loss": 0.5832,
+      "step": 11176
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.737593953849398,
+      "learning_rate": 7.849074583798912e-06,
+      "loss": 0.547,
+      "step": 11177
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 16.23704238232806,
+      "learning_rate": 7.847427687824615e-06,
+      "loss": 0.4946,
+      "step": 11178
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 25.272781334072516,
+      "learning_rate": 7.845780853068217e-06,
+      "loss": 0.5513,
+      "step": 11179
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.3597406125372,
+      "learning_rate": 7.844134079576553e-06,
+      "loss": 0.5056,
+      "step": 11180
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.034055145233486,
+      "learning_rate": 7.84248736739645e-06,
+      "loss": 0.4983,
+      "step": 11181
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.196566362067875,
+      "learning_rate": 7.840840716574748e-06,
+      "loss": 0.5368,
+      "step": 11182
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.627838873271484,
+      "learning_rate": 7.839194127158271e-06,
+      "loss": 0.5258,
+      "step": 11183
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.224867405885616,
+      "learning_rate": 7.837547599193849e-06,
+      "loss": 0.5772,
+      "step": 11184
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.95064474362585,
+      "learning_rate": 7.835901132728311e-06,
+      "loss": 0.4517,
+      "step": 11185
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 16.45076166153509,
+      "learning_rate": 7.834254727808477e-06,
+      "loss": 0.5257,
+      "step": 11186
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.820391013810106,
+      "learning_rate": 7.83260838448117e-06,
+      "loss": 0.5656,
+      "step": 11187
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.667336873699867,
+      "learning_rate": 7.830962102793214e-06,
+      "loss": 0.6235,
+      "step": 11188
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.754841785972543,
+      "learning_rate": 7.829315882791426e-06,
+      "loss": 0.5182,
+      "step": 11189
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.898584723934768,
+      "learning_rate": 7.827669724522632e-06,
+      "loss": 0.5112,
+      "step": 11190
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.03620929445909,
+      "learning_rate": 7.826023628033631e-06,
+      "loss": 0.5311,
+      "step": 11191
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 24.418773772718986,
+      "learning_rate": 7.824377593371248e-06,
+      "loss": 0.5441,
+      "step": 11192
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.220247998822664,
+      "learning_rate": 7.822731620582294e-06,
+      "loss": 0.562,
+      "step": 11193
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.274615775975082,
+      "learning_rate": 7.821085709713581e-06,
+      "loss": 0.6609,
+      "step": 11194
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.167098561320827,
+      "learning_rate": 7.819439860811915e-06,
+      "loss": 0.5042,
+      "step": 11195
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 21.093025764465423,
+      "learning_rate": 7.817794073924104e-06,
+      "loss": 0.5515,
+      "step": 11196
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.877998192084533,
+      "learning_rate": 7.816148349096951e-06,
+      "loss": 0.5659,
+      "step": 11197
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.18021799363955,
+      "learning_rate": 7.814502686377263e-06,
+      "loss": 0.5763,
+      "step": 11198
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.60497919127739,
+      "learning_rate": 7.812857085811837e-06,
+      "loss": 0.5274,
+      "step": 11199
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 24.518095578832366,
+      "learning_rate": 7.811211547447483e-06,
+      "loss": 0.5378,
+      "step": 11200
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.476006045798519,
+      "learning_rate": 7.809566071330987e-06,
+      "loss": 0.5343,
+      "step": 11201
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.502766660779288,
+      "learning_rate": 7.807920657509152e-06,
+      "loss": 0.5502,
+      "step": 11202
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.12532621615114,
+      "learning_rate": 7.80627530602877e-06,
+      "loss": 0.6286,
+      "step": 11203
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.686116191255312,
+      "learning_rate": 7.804630016936636e-06,
+      "loss": 0.5292,
+      "step": 11204
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.902276441245746,
+      "learning_rate": 7.802984790279542e-06,
+      "loss": 0.5782,
+      "step": 11205
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.75679950367359,
+      "learning_rate": 7.801339626104272e-06,
+      "loss": 0.5057,
+      "step": 11206
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 25.95664332252799,
+      "learning_rate": 7.799694524457616e-06,
+      "loss": 0.5221,
+      "step": 11207
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.29322166637064,
+      "learning_rate": 7.798049485386365e-06,
+      "loss": 0.6383,
+      "step": 11208
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.69645899817208,
+      "learning_rate": 7.796404508937295e-06,
+      "loss": 0.5263,
+      "step": 11209
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.98804970425537,
+      "learning_rate": 7.794759595157194e-06,
+      "loss": 0.5728,
+      "step": 11210
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.8381526613064,
+      "learning_rate": 7.79311474409284e-06,
+      "loss": 0.5681,
+      "step": 11211
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.553867530167153,
+      "learning_rate": 7.791469955791009e-06,
+      "loss": 0.5167,
+      "step": 11212
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 28.675711620388135,
+      "learning_rate": 7.789825230298481e-06,
+      "loss": 0.5042,
+      "step": 11213
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 24.621927666949734,
+      "learning_rate": 7.788180567662031e-06,
+      "loss": 0.5363,
+      "step": 11214
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.111166937280643,
+      "learning_rate": 7.786535967928437e-06,
+      "loss": 0.5568,
+      "step": 11215
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.22113113849329,
+      "learning_rate": 7.784891431144459e-06,
+      "loss": 0.5156,
+      "step": 11216
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 16.3711243235276,
+      "learning_rate": 7.783246957356876e-06,
+      "loss": 0.4867,
+      "step": 11217
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.481162776941453,
+      "learning_rate": 7.781602546612447e-06,
+      "loss": 0.5881,
+      "step": 11218
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.39490523396599,
+      "learning_rate": 7.779958198957951e-06,
+      "loss": 0.5699,
+      "step": 11219
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.988333990215676,
+      "learning_rate": 7.778313914440143e-06,
+      "loss": 0.5531,
+      "step": 11220
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 26.344981249615987,
+      "learning_rate": 7.776669693105786e-06,
+      "loss": 0.6281,
+      "step": 11221
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.98876758312302,
+      "learning_rate": 7.775025535001643e-06,
+      "loss": 0.6004,
+      "step": 11222
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.00271330621962,
+      "learning_rate": 7.773381440174473e-06,
+      "loss": 0.5866,
+      "step": 11223
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.538014427764445,
+      "learning_rate": 7.771737408671034e-06,
+      "loss": 0.5326,
+      "step": 11224
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.40138392845931,
+      "learning_rate": 7.770093440538076e-06,
+      "loss": 0.6238,
+      "step": 11225
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.467097297984274,
+      "learning_rate": 7.768449535822357e-06,
+      "loss": 0.5971,
+      "step": 11226
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 16.94261482295833,
+      "learning_rate": 7.766805694570629e-06,
+      "loss": 0.5379,
+      "step": 11227
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 16.03707658929044,
+      "learning_rate": 7.76516191682964e-06,
+      "loss": 0.5704,
+      "step": 11228
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.489801493630956,
+      "learning_rate": 7.76351820264614e-06,
+      "loss": 0.5618,
+      "step": 11229
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 39.26357738622217,
+      "learning_rate": 7.761874552066873e-06,
+      "loss": 0.6505,
+      "step": 11230
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 30.675719590576257,
+      "learning_rate": 7.760230965138582e-06,
+      "loss": 0.6188,
+      "step": 11231
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.640713466872427,
+      "learning_rate": 7.758587441908014e-06,
+      "loss": 0.5371,
+      "step": 11232
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 29.7058679482082,
+      "learning_rate": 7.756943982421907e-06,
+      "loss": 0.6301,
+      "step": 11233
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.67222206237001,
+      "learning_rate": 7.755300586727007e-06,
+      "loss": 0.5135,
+      "step": 11234
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.478443852101808,
+      "learning_rate": 7.753657254870038e-06,
+      "loss": 0.576,
+      "step": 11235
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 32.647392756461734,
+      "learning_rate": 7.752013986897747e-06,
+      "loss": 0.6269,
+      "step": 11236
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.111928499682369,
+      "learning_rate": 7.750370782856858e-06,
+      "loss": 0.4867,
+      "step": 11237
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.0842524601561,
+      "learning_rate": 7.748727642794111e-06,
+      "loss": 0.5763,
+      "step": 11238
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.249422957709474,
+      "learning_rate": 7.747084566756237e-06,
+      "loss": 0.6144,
+      "step": 11239
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.60997163917072,
+      "learning_rate": 7.745441554789956e-06,
+      "loss": 0.4776,
+      "step": 11240
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.54723510593718,
+      "learning_rate": 7.743798606941997e-06,
+      "loss": 0.5605,
+      "step": 11241
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.03156441258331,
+      "learning_rate": 7.742155723259086e-06,
+      "loss": 0.6029,
+      "step": 11242
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.844558310768933,
+      "learning_rate": 7.740512903787945e-06,
+      "loss": 0.5783,
+      "step": 11243
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.290430426064916,
+      "learning_rate": 7.738870148575299e-06,
+      "loss": 0.5824,
+      "step": 11244
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.021835985888213,
+      "learning_rate": 7.737227457667861e-06,
+      "loss": 0.6041,
+      "step": 11245
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.910213307840092,
+      "learning_rate": 7.735584831112352e-06,
+      "loss": 0.5481,
+      "step": 11246
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.54598192117341,
+      "learning_rate": 7.733942268955482e-06,
+      "loss": 0.5601,
+      "step": 11247
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.875992057808219,
+      "learning_rate": 7.732299771243972e-06,
+      "loss": 0.6435,
+      "step": 11248
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.111678844400902,
+      "learning_rate": 7.73065733802453e-06,
+      "loss": 0.5505,
+      "step": 11249
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.713366736911624,
+      "learning_rate": 7.729014969343863e-06,
+      "loss": 0.5367,
+      "step": 11250
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.51284570335611,
+      "learning_rate": 7.727372665248682e-06,
+      "loss": 0.5855,
+      "step": 11251
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 25.156948780045354,
+      "learning_rate": 7.725730425785694e-06,
+      "loss": 0.6051,
+      "step": 11252
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.73512527900115,
+      "learning_rate": 7.724088251001602e-06,
+      "loss": 0.5376,
+      "step": 11253
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.11760176748968,
+      "learning_rate": 7.722446140943111e-06,
+      "loss": 0.5813,
+      "step": 11254
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.03475619663871,
+      "learning_rate": 7.720804095656918e-06,
+      "loss": 0.4859,
+      "step": 11255
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.216014199055966,
+      "learning_rate": 7.71916211518972e-06,
+      "loss": 0.5153,
+      "step": 11256
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.174982535027215,
+      "learning_rate": 7.717520199588222e-06,
+      "loss": 0.5669,
+      "step": 11257
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.71978624911367,
+      "learning_rate": 7.71587834889911e-06,
+      "loss": 0.5883,
+      "step": 11258
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.758882855347803,
+      "learning_rate": 7.714236563169088e-06,
+      "loss": 0.4991,
+      "step": 11259
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.58478125235607,
+      "learning_rate": 7.712594842444836e-06,
+      "loss": 0.4861,
+      "step": 11260
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.183877571773113,
+      "learning_rate": 7.710953186773048e-06,
+      "loss": 0.6171,
+      "step": 11261
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.82640345035016,
+      "learning_rate": 7.709311596200412e-06,
+      "loss": 0.5654,
+      "step": 11262
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 30.559027839765587,
+      "learning_rate": 7.707670070773616e-06,
+      "loss": 0.5939,
+      "step": 11263
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 28.142003183966946,
+      "learning_rate": 7.706028610539345e-06,
+      "loss": 0.6379,
+      "step": 11264
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.554621440569086,
+      "learning_rate": 7.704387215544276e-06,
+      "loss": 0.5122,
+      "step": 11265
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.55907975554662,
+      "learning_rate": 7.70274588583509e-06,
+      "loss": 0.5367,
+      "step": 11266
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.607315172325947,
+      "learning_rate": 7.701104621458467e-06,
+      "loss": 0.5735,
+      "step": 11267
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.2734713567615,
+      "learning_rate": 7.69946342246109e-06,
+      "loss": 0.5533,
+      "step": 11268
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.717117067772456,
+      "learning_rate": 7.697822288889617e-06,
+      "loss": 0.5291,
+      "step": 11269
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.384736506442632,
+      "learning_rate": 7.69618122079074e-06,
+      "loss": 0.5018,
+      "step": 11270
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.973631029852697,
+      "learning_rate": 7.694540218211117e-06,
+      "loss": 0.563,
+      "step": 11271
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.628703270044507,
+      "learning_rate": 7.692899281197421e-06,
+      "loss": 0.4823,
+      "step": 11272
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.57420443406871,
+      "learning_rate": 7.691258409796324e-06,
+      "loss": 0.4832,
+      "step": 11273
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.447646746363091,
+      "learning_rate": 7.689617604054487e-06,
+      "loss": 0.5394,
+      "step": 11274
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.838827575073203,
+      "learning_rate": 7.68797686401857e-06,
+      "loss": 0.5503,
+      "step": 11275
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.091192857792542,
+      "learning_rate": 7.68633618973524e-06,
+      "loss": 0.4895,
+      "step": 11276
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.817442342235797,
+      "learning_rate": 7.684695581251153e-06,
+      "loss": 0.5912,
+      "step": 11277
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.593332278109564,
+      "learning_rate": 7.683055038612977e-06,
+      "loss": 0.6045,
+      "step": 11278
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.482524264990158,
+      "learning_rate": 7.681414561867353e-06,
+      "loss": 0.4914,
+      "step": 11279
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.650770275846927,
+      "learning_rate": 7.679774151060945e-06,
+      "loss": 0.5341,
+      "step": 11280
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 25.13054877720958,
+      "learning_rate": 7.6781338062404e-06,
+      "loss": 0.5412,
+      "step": 11281
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.12693658041485,
+      "learning_rate": 7.676493527452374e-06,
+      "loss": 0.5932,
+      "step": 11282
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.663614451730098,
+      "learning_rate": 7.674853314743516e-06,
+      "loss": 0.4899,
+      "step": 11283
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 24.244453653723927,
+      "learning_rate": 7.673213168160465e-06,
+      "loss": 0.5234,
+      "step": 11284
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.741632808532874,
+      "learning_rate": 7.671573087749871e-06,
+      "loss": 0.4883,
+      "step": 11285
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.626228054121917,
+      "learning_rate": 7.669933073558377e-06,
+      "loss": 0.6313,
+      "step": 11286
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.419755841583367,
+      "learning_rate": 7.668293125632621e-06,
+      "loss": 0.5558,
+      "step": 11287
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 24.203634339225513,
+      "learning_rate": 7.66665324401925e-06,
+      "loss": 0.6161,
+      "step": 11288
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 24.611350398632172,
+      "learning_rate": 7.665013428764892e-06,
+      "loss": 0.5002,
+      "step": 11289
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 24.788121986761205,
+      "learning_rate": 7.663373679916187e-06,
+      "loss": 0.5877,
+      "step": 11290
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 45.5453902615997,
+      "learning_rate": 7.661733997519765e-06,
+      "loss": 0.676,
+      "step": 11291
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.680113970711655,
+      "learning_rate": 7.660094381622265e-06,
+      "loss": 0.5259,
+      "step": 11292
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.286117485253868,
+      "learning_rate": 7.658454832270312e-06,
+      "loss": 0.5003,
+      "step": 11293
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.25049209012945,
+      "learning_rate": 7.656815349510528e-06,
+      "loss": 0.5075,
+      "step": 11294
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 24.28092716007329,
+      "learning_rate": 7.65517593338955e-06,
+      "loss": 0.5332,
+      "step": 11295
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 26.04700770455865,
+      "learning_rate": 7.653536583953993e-06,
+      "loss": 0.593,
+      "step": 11296
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.678745228459746,
+      "learning_rate": 7.651897301250484e-06,
+      "loss": 0.5501,
+      "step": 11297
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.003487137617729,
+      "learning_rate": 7.650258085325646e-06,
+      "loss": 0.5254,
+      "step": 11298
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.865165146944452,
+      "learning_rate": 7.64861893622609e-06,
+      "loss": 0.5383,
+      "step": 11299
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.664046491118665,
+      "learning_rate": 7.646979853998434e-06,
+      "loss": 0.5551,
+      "step": 11300
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.161354677915533,
+      "learning_rate": 7.645340838689297e-06,
+      "loss": 0.5478,
+      "step": 11301
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.640657624247467,
+      "learning_rate": 7.643701890345288e-06,
+      "loss": 0.5305,
+      "step": 11302
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.743547150944273,
+      "learning_rate": 7.642063009013025e-06,
+      "loss": 0.5061,
+      "step": 11303
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.06998142425038,
+      "learning_rate": 7.640424194739102e-06,
+      "loss": 0.5848,
+      "step": 11304
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.29649272435952,
+      "learning_rate": 7.63878544757014e-06,
+      "loss": 0.6403,
+      "step": 11305
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.715199545490968,
+      "learning_rate": 7.637146767552735e-06,
+      "loss": 0.5738,
+      "step": 11306
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 25.079013950718778,
+      "learning_rate": 7.635508154733494e-06,
+      "loss": 0.5993,
+      "step": 11307
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.884186294227668,
+      "learning_rate": 7.633869609159023e-06,
+      "loss": 0.5479,
+      "step": 11308
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.411200973428215,
+      "learning_rate": 7.632231130875912e-06,
+      "loss": 0.4966,
+      "step": 11309
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.025665686084285,
+      "learning_rate": 7.63059271993076e-06,
+      "loss": 0.493,
+      "step": 11310
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.17998847181433,
+      "learning_rate": 7.628954376370169e-06,
+      "loss": 0.5698,
+      "step": 11311
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.30178264902838,
+      "learning_rate": 7.627316100240726e-06,
+      "loss": 0.62,
+      "step": 11312
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.115320294558252,
+      "learning_rate": 7.625677891589029e-06,
+      "loss": 0.546,
+      "step": 11313
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.920930267645089,
+      "learning_rate": 7.6240397504616615e-06,
+      "loss": 0.5092,
+      "step": 11314
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 23.25450947570427,
+      "learning_rate": 7.622401676905214e-06,
+      "loss": 0.5265,
+      "step": 11315
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.207353591928037,
+      "learning_rate": 7.6207636709662694e-06,
+      "loss": 0.5301,
+      "step": 11316
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.44018048163546,
+      "learning_rate": 7.619125732691419e-06,
+      "loss": 0.5206,
+      "step": 11317
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 29.436667042284412,
+      "learning_rate": 7.617487862127238e-06,
+      "loss": 0.6249,
+      "step": 11318
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.89768052200801,
+      "learning_rate": 7.615850059320306e-06,
+      "loss": 0.5658,
+      "step": 11319
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.396333062932055,
+      "learning_rate": 7.614212324317205e-06,
+      "loss": 0.5533,
+      "step": 11320
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.466935711726588,
+      "learning_rate": 7.612574657164509e-06,
+      "loss": 0.5844,
+      "step": 11321
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.202536005017654,
+      "learning_rate": 7.610937057908798e-06,
+      "loss": 0.5077,
+      "step": 11322
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.427338583373395,
+      "learning_rate": 7.609299526596634e-06,
+      "loss": 0.5474,
+      "step": 11323
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.521451742194035,
+      "learning_rate": 7.607662063274595e-06,
+      "loss": 0.5075,
+      "step": 11324
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.76787564709022,
+      "learning_rate": 7.606024667989246e-06,
+      "loss": 0.5888,
+      "step": 11325
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.35670487486769,
+      "learning_rate": 7.604387340787156e-06,
+      "loss": 0.6043,
+      "step": 11326
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.377373961453696,
+      "learning_rate": 7.6027500817148915e-06,
+      "loss": 0.5456,
+      "step": 11327
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 24.70214588944195,
+      "learning_rate": 7.6011128908190105e-06,
+      "loss": 0.5512,
+      "step": 11328
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.17046677526658,
+      "learning_rate": 7.599475768146072e-06,
+      "loss": 0.5254,
+      "step": 11329
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 28.881291500838543,
+      "learning_rate": 7.5978387137426425e-06,
+      "loss": 0.5213,
+      "step": 11330
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 27.06366082855771,
+      "learning_rate": 7.59620172765527e-06,
+      "loss": 0.6224,
+      "step": 11331
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.986773955641482,
+      "learning_rate": 7.5945648099305206e-06,
+      "loss": 0.536,
+      "step": 11332
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.104959198871104,
+      "learning_rate": 7.592927960614939e-06,
+      "loss": 0.5259,
+      "step": 11333
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 27.427609701510946,
+      "learning_rate": 7.591291179755077e-06,
+      "loss": 0.5567,
+      "step": 11334
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 30.908048655047114,
+      "learning_rate": 7.5896544673974825e-06,
+      "loss": 0.6181,
+      "step": 11335
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.155516628645152,
+      "learning_rate": 7.588017823588707e-06,
+      "loss": 0.5286,
+      "step": 11336
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.862126498204118,
+      "learning_rate": 7.586381248375296e-06,
+      "loss": 0.542,
+      "step": 11337
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.68626587114657,
+      "learning_rate": 7.584744741803787e-06,
+      "loss": 0.6334,
+      "step": 11338
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.741204965630933,
+      "learning_rate": 7.583108303920726e-06,
+      "loss": 0.5653,
+      "step": 11339
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.622610320171088,
+      "learning_rate": 7.581471934772649e-06,
+      "loss": 0.5145,
+      "step": 11340
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 26.460607001777873,
+      "learning_rate": 7.579835634406097e-06,
+      "loss": 0.5746,
+      "step": 11341
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.09364872715884,
+      "learning_rate": 7.578199402867607e-06,
+      "loss": 0.5498,
+      "step": 11342
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.428643651102213,
+      "learning_rate": 7.576563240203707e-06,
+      "loss": 0.5437,
+      "step": 11343
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 27.391532778453815,
+      "learning_rate": 7.574927146460928e-06,
+      "loss": 0.7232,
+      "step": 11344
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.036940668630008,
+      "learning_rate": 7.573291121685805e-06,
+      "loss": 0.5555,
+      "step": 11345
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.541138106240838,
+      "learning_rate": 7.571655165924861e-06,
+      "loss": 0.5121,
+      "step": 11346
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.702217574801235,
+      "learning_rate": 7.570019279224629e-06,
+      "loss": 0.5851,
+      "step": 11347
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.994134952919435,
+      "learning_rate": 7.568383461631621e-06,
+      "loss": 0.5711,
+      "step": 11348
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.010437819121165,
+      "learning_rate": 7.566747713192368e-06,
+      "loss": 0.5566,
+      "step": 11349
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.631605681039963,
+      "learning_rate": 7.5651120339533836e-06,
+      "loss": 0.5361,
+      "step": 11350
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 26.537630574268555,
+      "learning_rate": 7.56347642396119e-06,
+      "loss": 0.4726,
+      "step": 11351
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.490625056277086,
+      "learning_rate": 7.5618408832623044e-06,
+      "loss": 0.5939,
+      "step": 11352
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.01667213887202,
+      "learning_rate": 7.5602054119032364e-06,
+      "loss": 0.581,
+      "step": 11353
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.18123170895015,
+      "learning_rate": 7.558570009930497e-06,
+      "loss": 0.5122,
+      "step": 11354
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.46694993671408,
+      "learning_rate": 7.556934677390599e-06,
+      "loss": 0.5276,
+      "step": 11355
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.678534258165081,
+      "learning_rate": 7.555299414330048e-06,
+      "loss": 0.5347,
+      "step": 11356
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.84877471547157,
+      "learning_rate": 7.553664220795357e-06,
+      "loss": 0.624,
+      "step": 11357
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.20675257685648,
+      "learning_rate": 7.5520290968330226e-06,
+      "loss": 0.5292,
+      "step": 11358
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.244811129173595,
+      "learning_rate": 7.550394042489547e-06,
+      "loss": 0.4978,
+      "step": 11359
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.691635399920745,
+      "learning_rate": 7.54875905781143e-06,
+      "loss": 0.5368,
+      "step": 11360
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.108495215344774,
+      "learning_rate": 7.547124142845174e-06,
+      "loss": 0.5894,
+      "step": 11361
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.787358983628412,
+      "learning_rate": 7.545489297637275e-06,
+      "loss": 0.5217,
+      "step": 11362
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.75864957839539,
+      "learning_rate": 7.54385452223422e-06,
+      "loss": 0.5147,
+      "step": 11363
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.827183936404236,
+      "learning_rate": 7.542219816682508e-06,
+      "loss": 0.5432,
+      "step": 11364
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.936623261335924,
+      "learning_rate": 7.5405851810286236e-06,
+      "loss": 0.52,
+      "step": 11365
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 31.788461221524262,
+      "learning_rate": 7.538950615319065e-06,
+      "loss": 0.5846,
+      "step": 11366
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.16915610236504,
+      "learning_rate": 7.537316119600305e-06,
+      "loss": 0.4864,
+      "step": 11367
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 26.443898616504786,
+      "learning_rate": 7.535681693918836e-06,
+      "loss": 0.6274,
+      "step": 11368
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 20.05773937853591,
+      "learning_rate": 7.534047338321135e-06,
+      "loss": 0.6017,
+      "step": 11369
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.0723196411894,
+      "learning_rate": 7.5324130528536885e-06,
+      "loss": 0.544,
+      "step": 11370
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.615076700172942,
+      "learning_rate": 7.530778837562973e-06,
+      "loss": 0.5261,
+      "step": 11371
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.103055651792662,
+      "learning_rate": 7.5291446924954604e-06,
+      "loss": 0.5812,
+      "step": 11372
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.764978800505,
+      "learning_rate": 7.527510617697627e-06,
+      "loss": 0.5989,
+      "step": 11373
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.735712433650383,
+      "learning_rate": 7.525876613215944e-06,
+      "loss": 0.5288,
+      "step": 11374
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.60952320707782,
+      "learning_rate": 7.524242679096884e-06,
+      "loss": 0.5349,
+      "step": 11375
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.33266028845027,
+      "learning_rate": 7.522608815386916e-06,
+      "loss": 0.5461,
+      "step": 11376
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 39.29299032375867,
+      "learning_rate": 7.520975022132503e-06,
+      "loss": 0.6018,
+      "step": 11377
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 20.944527482201728,
+      "learning_rate": 7.5193412993801096e-06,
+      "loss": 0.5803,
+      "step": 11378
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.319588604965567,
+      "learning_rate": 7.517707647176197e-06,
+      "loss": 0.5136,
+      "step": 11379
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.154907010275167,
+      "learning_rate": 7.5160740655672285e-06,
+      "loss": 0.5337,
+      "step": 11380
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.0472940895238,
+      "learning_rate": 7.514440554599662e-06,
+      "loss": 0.6092,
+      "step": 11381
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.427804427824013,
+      "learning_rate": 7.5128071143199485e-06,
+      "loss": 0.4463,
+      "step": 11382
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 28.468810734647416,
+      "learning_rate": 7.511173744774548e-06,
+      "loss": 0.6599,
+      "step": 11383
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 22.806755517760198,
+      "learning_rate": 7.509540446009909e-06,
+      "loss": 0.5261,
+      "step": 11384
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.650154336906553,
+      "learning_rate": 7.5079072180724834e-06,
+      "loss": 0.5738,
+      "step": 11385
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 20.552376672694674,
+      "learning_rate": 7.506274061008721e-06,
+      "loss": 0.5766,
+      "step": 11386
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.52665775612717,
+      "learning_rate": 7.504640974865065e-06,
+      "loss": 0.5652,
+      "step": 11387
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.364114213791307,
+      "learning_rate": 7.503007959687956e-06,
+      "loss": 0.5082,
+      "step": 11388
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.630305653625925,
+      "learning_rate": 7.501375015523843e-06,
+      "loss": 0.52,
+      "step": 11389
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.990781722460444,
+      "learning_rate": 7.4997421424191595e-06,
+      "loss": 0.5826,
+      "step": 11390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.385785030547925,
+      "learning_rate": 7.498109340420354e-06,
+      "loss": 0.5185,
+      "step": 11391
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 21.994585668737507,
+      "learning_rate": 7.496476609573851e-06,
+      "loss": 0.5332,
+      "step": 11392
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.288584503548858,
+      "learning_rate": 7.494843949926088e-06,
+      "loss": 0.5416,
+      "step": 11393
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.799751232809594,
+      "learning_rate": 7.493211361523496e-06,
+      "loss": 0.496,
+      "step": 11394
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.41847322767009,
+      "learning_rate": 7.49157884441251e-06,
+      "loss": 0.5305,
+      "step": 11395
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.666975920498224,
+      "learning_rate": 7.489946398639555e-06,
+      "loss": 0.557,
+      "step": 11396
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.181201309500835,
+      "learning_rate": 7.488314024251055e-06,
+      "loss": 0.5348,
+      "step": 11397
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.335047304902012,
+      "learning_rate": 7.486681721293432e-06,
+      "loss": 0.5491,
+      "step": 11398
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.997225798224527,
+      "learning_rate": 7.485049489813112e-06,
+      "loss": 0.5846,
+      "step": 11399
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 21.045463066134577,
+      "learning_rate": 7.483417329856513e-06,
+      "loss": 0.6344,
+      "step": 11400
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.379279360863514,
+      "learning_rate": 7.481785241470055e-06,
+      "loss": 0.5108,
+      "step": 11401
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 30.37656194493515,
+      "learning_rate": 7.48015322470015e-06,
+      "loss": 0.6035,
+      "step": 11402
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 28.551575829058095,
+      "learning_rate": 7.478521279593213e-06,
+      "loss": 0.5514,
+      "step": 11403
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 30.489734825170196,
+      "learning_rate": 7.476889406195656e-06,
+      "loss": 0.6415,
+      "step": 11404
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.29606572975482,
+      "learning_rate": 7.4752576045538894e-06,
+      "loss": 0.5571,
+      "step": 11405
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.47288837429332,
+      "learning_rate": 7.473625874714322e-06,
+      "loss": 0.5259,
+      "step": 11406
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 26.90683662127291,
+      "learning_rate": 7.471994216723352e-06,
+      "loss": 0.6256,
+      "step": 11407
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 28.548577318448874,
+      "learning_rate": 7.47036263062739e-06,
+      "loss": 0.5733,
+      "step": 11408
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.489789801600327,
+      "learning_rate": 7.4687311164728315e-06,
+      "loss": 0.6144,
+      "step": 11409
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 27.41217402065633,
+      "learning_rate": 7.467099674306083e-06,
+      "loss": 0.6344,
+      "step": 11410
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.61036505489193,
+      "learning_rate": 7.46546830417354e-06,
+      "loss": 0.6296,
+      "step": 11411
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.603542227338057,
+      "learning_rate": 7.463837006121593e-06,
+      "loss": 0.5792,
+      "step": 11412
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.910512073640668,
+      "learning_rate": 7.462205780196637e-06,
+      "loss": 0.4822,
+      "step": 11413
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.151463022633337,
+      "learning_rate": 7.460574626445065e-06,
+      "loss": 0.4946,
+      "step": 11414
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.15470215668673,
+      "learning_rate": 7.458943544913266e-06,
+      "loss": 0.4978,
+      "step": 11415
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.915036437568993,
+      "learning_rate": 7.457312535647627e-06,
+      "loss": 0.5425,
+      "step": 11416
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 29.983166331350837,
+      "learning_rate": 7.455681598694529e-06,
+      "loss": 0.5607,
+      "step": 11417
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.344378133859822,
+      "learning_rate": 7.454050734100358e-06,
+      "loss": 0.5259,
+      "step": 11418
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.006090705488193,
+      "learning_rate": 7.452419941911495e-06,
+      "loss": 0.4909,
+      "step": 11419
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 36.77931888197404,
+      "learning_rate": 7.450789222174322e-06,
+      "loss": 0.5748,
+      "step": 11420
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.665778530168314,
+      "learning_rate": 7.449158574935209e-06,
+      "loss": 0.5435,
+      "step": 11421
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.403401709261521,
+      "learning_rate": 7.4475280002405355e-06,
+      "loss": 0.4883,
+      "step": 11422
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.565923136502217,
+      "learning_rate": 7.445897498136671e-06,
+      "loss": 0.5907,
+      "step": 11423
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.554561143137953,
+      "learning_rate": 7.444267068669988e-06,
+      "loss": 0.5442,
+      "step": 11424
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.825035906550575,
+      "learning_rate": 7.442636711886857e-06,
+      "loss": 0.5484,
+      "step": 11425
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.329956392178588,
+      "learning_rate": 7.441006427833638e-06,
+      "loss": 0.4986,
+      "step": 11426
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.342584678444084,
+      "learning_rate": 7.439376216556702e-06,
+      "loss": 0.6236,
+      "step": 11427
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.12708209146642,
+      "learning_rate": 7.437746078102406e-06,
+      "loss": 0.5069,
+      "step": 11428
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.136200042452291,
+      "learning_rate": 7.4361160125171135e-06,
+      "loss": 0.522,
+      "step": 11429
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.614532100569896,
+      "learning_rate": 7.4344860198471845e-06,
+      "loss": 0.5555,
+      "step": 11430
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.02168763759666,
+      "learning_rate": 7.432856100138971e-06,
+      "loss": 0.5387,
+      "step": 11431
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 24.54406809162352,
+      "learning_rate": 7.431226253438826e-06,
+      "loss": 0.5536,
+      "step": 11432
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.297411764849363,
+      "learning_rate": 7.4295964797931045e-06,
+      "loss": 0.5904,
+      "step": 11433
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.62735902863111,
+      "learning_rate": 7.427966779248155e-06,
+      "loss": 0.5318,
+      "step": 11434
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.214972048562355,
+      "learning_rate": 7.4263371518503314e-06,
+      "loss": 0.544,
+      "step": 11435
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.4808209052573,
+      "learning_rate": 7.424707597645967e-06,
+      "loss": 0.5807,
+      "step": 11436
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 23.361949452178724,
+      "learning_rate": 7.423078116681415e-06,
+      "loss": 0.5074,
+      "step": 11437
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.813417483202834,
+      "learning_rate": 7.42144870900301e-06,
+      "loss": 0.5444,
+      "step": 11438
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.101800580091467,
+      "learning_rate": 7.419819374657099e-06,
+      "loss": 0.4982,
+      "step": 11439
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.75164397108254,
+      "learning_rate": 7.418190113690018e-06,
+      "loss": 0.5825,
+      "step": 11440
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 27.168074552109232,
+      "learning_rate": 7.416560926148096e-06,
+      "loss": 0.5696,
+      "step": 11441
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 24.425701566438384,
+      "learning_rate": 7.414931812077669e-06,
+      "loss": 0.5407,
+      "step": 11442
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.379802725722428,
+      "learning_rate": 7.413302771525071e-06,
+      "loss": 0.5773,
+      "step": 11443
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.548462701218444,
+      "learning_rate": 7.4116738045366275e-06,
+      "loss": 0.5163,
+      "step": 11444
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.31737987577868,
+      "learning_rate": 7.410044911158671e-06,
+      "loss": 0.4955,
+      "step": 11445
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.409990786943924,
+      "learning_rate": 7.408416091437519e-06,
+      "loss": 0.5088,
+      "step": 11446
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 21.85818226923015,
+      "learning_rate": 7.406787345419496e-06,
+      "loss": 0.4689,
+      "step": 11447
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.5033319472836,
+      "learning_rate": 7.405158673150925e-06,
+      "loss": 0.5989,
+      "step": 11448
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 23.966572658288015,
+      "learning_rate": 7.403530074678124e-06,
+      "loss": 0.6299,
+      "step": 11449
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 26.144250024422643,
+      "learning_rate": 7.401901550047412e-06,
+      "loss": 0.5756,
+      "step": 11450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 41.9853284824794,
+      "learning_rate": 7.400273099305095e-06,
+      "loss": 0.5159,
+      "step": 11451
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.188745112402138,
+      "learning_rate": 7.398644722497492e-06,
+      "loss": 0.58,
+      "step": 11452
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 23.155762553715245,
+      "learning_rate": 7.39701641967091e-06,
+      "loss": 0.6029,
+      "step": 11453
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.716441421418159,
+      "learning_rate": 7.39538819087166e-06,
+      "loss": 0.4462,
+      "step": 11454
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 21.049621076851558,
+      "learning_rate": 7.393760036146049e-06,
+      "loss": 0.5585,
+      "step": 11455
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.969821100140978,
+      "learning_rate": 7.392131955540375e-06,
+      "loss": 0.5433,
+      "step": 11456
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 29.09790260854883,
+      "learning_rate": 7.39050394910094e-06,
+      "loss": 0.5265,
+      "step": 11457
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 29.0838616271026,
+      "learning_rate": 7.388876016874049e-06,
+      "loss": 0.5435,
+      "step": 11458
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 23.600742562606495,
+      "learning_rate": 7.3872481589059955e-06,
+      "loss": 0.5656,
+      "step": 11459
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.187812296674355,
+      "learning_rate": 7.3856203752430815e-06,
+      "loss": 0.5298,
+      "step": 11460
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.464390609315895,
+      "learning_rate": 7.383992665931587e-06,
+      "loss": 0.5759,
+      "step": 11461
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 26.536347147839994,
+      "learning_rate": 7.382365031017815e-06,
+      "loss": 0.6713,
+      "step": 11462
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 26.35868706113572,
+      "learning_rate": 7.380737470548047e-06,
+      "loss": 0.5641,
+      "step": 11463
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.0942186094347,
+      "learning_rate": 7.379109984568578e-06,
+      "loss": 0.4985,
+      "step": 11464
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.604261128864643,
+      "learning_rate": 7.377482573125685e-06,
+      "loss": 0.5172,
+      "step": 11465
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.106487769309915,
+      "learning_rate": 7.375855236265653e-06,
+      "loss": 0.4756,
+      "step": 11466
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.625194477840566,
+      "learning_rate": 7.374227974034762e-06,
+      "loss": 0.5454,
+      "step": 11467
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.210040203417375,
+      "learning_rate": 7.372600786479292e-06,
+      "loss": 0.6038,
+      "step": 11468
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.766965638741333,
+      "learning_rate": 7.370973673645523e-06,
+      "loss": 0.5922,
+      "step": 11469
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.34738017269755,
+      "learning_rate": 7.369346635579719e-06,
+      "loss": 0.6014,
+      "step": 11470
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 21.399821268375863,
+      "learning_rate": 7.36771967232816e-06,
+      "loss": 0.6503,
+      "step": 11471
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.107412149332678,
+      "learning_rate": 7.366092783937112e-06,
+      "loss": 0.5697,
+      "step": 11472
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.495972166645576,
+      "learning_rate": 7.364465970452846e-06,
+      "loss": 0.5355,
+      "step": 11473
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.396476812833953,
+      "learning_rate": 7.362839231921629e-06,
+      "loss": 0.5564,
+      "step": 11474
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.382363428583044,
+      "learning_rate": 7.36121256838972e-06,
+      "loss": 0.5647,
+      "step": 11475
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 25.483142155800156,
+      "learning_rate": 7.359585979903378e-06,
+      "loss": 0.536,
+      "step": 11476
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 21.117138162908805,
+      "learning_rate": 7.357959466508869e-06,
+      "loss": 0.5983,
+      "step": 11477
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 24.221263222270228,
+      "learning_rate": 7.356333028252445e-06,
+      "loss": 0.6431,
+      "step": 11478
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.757440566532612,
+      "learning_rate": 7.35470666518037e-06,
+      "loss": 0.538,
+      "step": 11479
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.592670854771782,
+      "learning_rate": 7.353080377338884e-06,
+      "loss": 0.535,
+      "step": 11480
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 24.04995887655184,
+      "learning_rate": 7.351454164774246e-06,
+      "loss": 0.535,
+      "step": 11481
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.531234554569691,
+      "learning_rate": 7.349828027532697e-06,
+      "loss": 0.5391,
+      "step": 11482
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.007842090547893,
+      "learning_rate": 7.348201965660493e-06,
+      "loss": 0.4662,
+      "step": 11483
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.89742319054473,
+      "learning_rate": 7.346575979203876e-06,
+      "loss": 0.6283,
+      "step": 11484
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.997308287685787,
+      "learning_rate": 7.344950068209085e-06,
+      "loss": 0.5943,
+      "step": 11485
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.95218685340715,
+      "learning_rate": 7.343324232722358e-06,
+      "loss": 0.5205,
+      "step": 11486
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.901178688783077,
+      "learning_rate": 7.341698472789937e-06,
+      "loss": 0.5508,
+      "step": 11487
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 32.234490718439694,
+      "learning_rate": 7.340072788458054e-06,
+      "loss": 0.6428,
+      "step": 11488
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.13596092864975,
+      "learning_rate": 7.33844717977295e-06,
+      "loss": 0.5074,
+      "step": 11489
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.318878771675195,
+      "learning_rate": 7.336821646780848e-06,
+      "loss": 0.612,
+      "step": 11490
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.813702198757596,
+      "learning_rate": 7.33519618952798e-06,
+      "loss": 0.5537,
+      "step": 11491
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 33.8220451371144,
+      "learning_rate": 7.333570808060575e-06,
+      "loss": 0.6062,
+      "step": 11492
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.838004598079962,
+      "learning_rate": 7.331945502424856e-06,
+      "loss": 0.5966,
+      "step": 11493
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 23.606448647933572,
+      "learning_rate": 7.330320272667048e-06,
+      "loss": 0.5019,
+      "step": 11494
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.672707554502434,
+      "learning_rate": 7.328695118833366e-06,
+      "loss": 0.5954,
+      "step": 11495
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.290324147070788,
+      "learning_rate": 7.327070040970034e-06,
+      "loss": 0.5974,
+      "step": 11496
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.957349660297997,
+      "learning_rate": 7.325445039123264e-06,
+      "loss": 0.5039,
+      "step": 11497
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.023864402707884,
+      "learning_rate": 7.323820113339275e-06,
+      "loss": 0.5635,
+      "step": 11498
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.764988161436822,
+      "learning_rate": 7.32219526366428e-06,
+      "loss": 0.4926,
+      "step": 11499
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.2724340484563,
+      "learning_rate": 7.320570490144481e-06,
+      "loss": 0.56,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.689650348748872,
+      "learning_rate": 7.318945792826089e-06,
+      "loss": 0.4847,
+      "step": 11501
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.03860160808342,
+      "learning_rate": 7.317321171755312e-06,
+      "loss": 0.5869,
+      "step": 11502
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 21.71263439675476,
+      "learning_rate": 7.315696626978349e-06,
+      "loss": 0.582,
+      "step": 11503
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.618109297616618,
+      "learning_rate": 7.31407215854141e-06,
+      "loss": 0.5609,
+      "step": 11504
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.716229185459664,
+      "learning_rate": 7.312447766490682e-06,
+      "loss": 0.5531,
+      "step": 11505
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 28.59327396194591,
+      "learning_rate": 7.310823450872368e-06,
+      "loss": 0.6018,
+      "step": 11506
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 23.28109653729545,
+      "learning_rate": 7.309199211732662e-06,
+      "loss": 0.5789,
+      "step": 11507
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 25.067616773364556,
+      "learning_rate": 7.307575049117758e-06,
+      "loss": 0.529,
+      "step": 11508
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.129865554982795,
+      "learning_rate": 7.305950963073845e-06,
+      "loss": 0.5656,
+      "step": 11509
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.387748238030866,
+      "learning_rate": 7.30432695364711e-06,
+      "loss": 0.5259,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.88808603023079,
+      "learning_rate": 7.3027030208837365e-06,
+      "loss": 0.5565,
+      "step": 11511
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.058768320662143,
+      "learning_rate": 7.301079164829914e-06,
+      "loss": 0.5268,
+      "step": 11512
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.521769197177548,
+      "learning_rate": 7.299455385531824e-06,
+      "loss": 0.5398,
+      "step": 11513
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.795225292172237,
+      "learning_rate": 7.297831683035638e-06,
+      "loss": 0.5016,
+      "step": 11514
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.60226270664066,
+      "learning_rate": 7.296208057387539e-06,
+      "loss": 0.5119,
+      "step": 11515
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.838540825132426,
+      "learning_rate": 7.294584508633702e-06,
+      "loss": 0.4934,
+      "step": 11516
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.4410294206056,
+      "learning_rate": 7.292961036820299e-06,
+      "loss": 0.5948,
+      "step": 11517
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.67166955667262,
+      "learning_rate": 7.291337641993503e-06,
+      "loss": 0.5506,
+      "step": 11518
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 25.66219568880412,
+      "learning_rate": 7.289714324199477e-06,
+      "loss": 0.5738,
+      "step": 11519
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 25.18033534041247,
+      "learning_rate": 7.28809108348439e-06,
+      "loss": 0.5238,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 30.523125982537806,
+      "learning_rate": 7.2864679198944065e-06,
+      "loss": 0.5832,
+      "step": 11521
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.100377477474808,
+      "learning_rate": 7.284844833475687e-06,
+      "loss": 0.6028,
+      "step": 11522
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.453355127528393,
+      "learning_rate": 7.2832218242743976e-06,
+      "loss": 0.5583,
+      "step": 11523
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.263042211913042,
+      "learning_rate": 7.281598892336685e-06,
+      "loss": 0.4963,
+      "step": 11524
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.999787035483596,
+      "learning_rate": 7.279976037708712e-06,
+      "loss": 0.5325,
+      "step": 11525
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.882125789505263,
+      "learning_rate": 7.278353260436626e-06,
+      "loss": 0.4867,
+      "step": 11526
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.530496046922735,
+      "learning_rate": 7.276730560566583e-06,
+      "loss": 0.4782,
+      "step": 11527
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.621875188426237,
+      "learning_rate": 7.275107938144732e-06,
+      "loss": 0.5357,
+      "step": 11528
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.83509778942412,
+      "learning_rate": 7.273485393217217e-06,
+      "loss": 0.6171,
+      "step": 11529
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.770321480678714,
+      "learning_rate": 7.2718629258301786e-06,
+      "loss": 0.5575,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.559136687930513,
+      "learning_rate": 7.270240536029765e-06,
+      "loss": 0.5941,
+      "step": 11531
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 22.22612460263177,
+      "learning_rate": 7.268618223862112e-06,
+      "loss": 0.5206,
+      "step": 11532
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 30.797401464366427,
+      "learning_rate": 7.266995989373362e-06,
+      "loss": 0.6178,
+      "step": 11533
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.26367688901969,
+      "learning_rate": 7.265373832609647e-06,
+      "loss": 0.4966,
+      "step": 11534
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.968609561041333,
+      "learning_rate": 7.263751753617096e-06,
+      "loss": 0.4368,
+      "step": 11535
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.96177088798451,
+      "learning_rate": 7.262129752441849e-06,
+      "loss": 0.5324,
+      "step": 11536
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.13775319082707,
+      "learning_rate": 7.26050782913003e-06,
+      "loss": 0.5062,
+      "step": 11537
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.345901915224099,
+      "learning_rate": 7.258885983727767e-06,
+      "loss": 0.5407,
+      "step": 11538
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.20253414292048,
+      "learning_rate": 7.257264216281179e-06,
+      "loss": 0.5549,
+      "step": 11539
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.602094976076273,
+      "learning_rate": 7.255642526836395e-06,
+      "loss": 0.5279,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.932284910787157,
+      "learning_rate": 7.254020915439528e-06,
+      "loss": 0.5675,
+      "step": 11541
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 23.63453724990942,
+      "learning_rate": 7.252399382136703e-06,
+      "loss": 0.5577,
+      "step": 11542
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 29.52740181771988,
+      "learning_rate": 7.250777926974035e-06,
+      "loss": 0.5173,
+      "step": 11543
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.16748262686469,
+      "learning_rate": 7.2491565499976335e-06,
+      "loss": 0.4873,
+      "step": 11544
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 22.219209069393244,
+      "learning_rate": 7.247535251253606e-06,
+      "loss": 0.5414,
+      "step": 11545
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 22.595532408326665,
+      "learning_rate": 7.245914030788069e-06,
+      "loss": 0.6052,
+      "step": 11546
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.39046671400855,
+      "learning_rate": 7.2442928886471246e-06,
+      "loss": 0.4945,
+      "step": 11547
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.75083033600665,
+      "learning_rate": 7.242671824876884e-06,
+      "loss": 0.493,
+      "step": 11548
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.200690403828464,
+      "learning_rate": 7.241050839523437e-06,
+      "loss": 0.532,
+      "step": 11549
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.070537502632565,
+      "learning_rate": 7.239429932632892e-06,
+      "loss": 0.5129,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.56690887688945,
+      "learning_rate": 7.237809104251343e-06,
+      "loss": 0.5329,
+      "step": 11551
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 25.513129948030343,
+      "learning_rate": 7.23618835442489e-06,
+      "loss": 0.5995,
+      "step": 11552
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.327299232516882,
+      "learning_rate": 7.234567683199624e-06,
+      "loss": 0.5152,
+      "step": 11553
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 24.434607061388434,
+      "learning_rate": 7.232947090621633e-06,
+      "loss": 0.6099,
+      "step": 11554
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.68106319469668,
+      "learning_rate": 7.23132657673701e-06,
+      "loss": 0.5755,
+      "step": 11555
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.69053986252105,
+      "learning_rate": 7.2297061415918394e-06,
+      "loss": 0.5416,
+      "step": 11556
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 25.534710592366903,
+      "learning_rate": 7.228085785232206e-06,
+      "loss": 0.538,
+      "step": 11557
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.670140467561515,
+      "learning_rate": 7.226465507704189e-06,
+      "loss": 0.5287,
+      "step": 11558
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.034391567416467,
+      "learning_rate": 7.224845309053872e-06,
+      "loss": 0.5244,
+      "step": 11559
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.254307325144577,
+      "learning_rate": 7.2232251893273295e-06,
+      "loss": 0.554,
+      "step": 11560
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.428917244941026,
+      "learning_rate": 7.2216051485706385e-06,
+      "loss": 0.6522,
+      "step": 11561
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.35913087877865,
+      "learning_rate": 7.219985186829877e-06,
+      "loss": 0.5901,
+      "step": 11562
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.886915921771122,
+      "learning_rate": 7.2183653041511045e-06,
+      "loss": 0.4914,
+      "step": 11563
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.316887925659467,
+      "learning_rate": 7.216745500580396e-06,
+      "loss": 0.4751,
+      "step": 11564
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.381351965388394,
+      "learning_rate": 7.215125776163818e-06,
+      "loss": 0.53,
+      "step": 11565
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 22.229950311112308,
+      "learning_rate": 7.213506130947434e-06,
+      "loss": 0.5857,
+      "step": 11566
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.602523431748542,
+      "learning_rate": 7.21188656497731e-06,
+      "loss": 0.4648,
+      "step": 11567
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.704346830809087,
+      "learning_rate": 7.210267078299495e-06,
+      "loss": 0.6302,
+      "step": 11568
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.25239431136035,
+      "learning_rate": 7.208647670960055e-06,
+      "loss": 0.5672,
+      "step": 11569
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.816507427838555,
+      "learning_rate": 7.2070283430050405e-06,
+      "loss": 0.5629,
+      "step": 11570
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.329995378554532,
+      "learning_rate": 7.205409094480506e-06,
+      "loss": 0.6058,
+      "step": 11571
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.91797965965295,
+      "learning_rate": 7.203789925432507e-06,
+      "loss": 0.5032,
+      "step": 11572
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.999618177196492,
+      "learning_rate": 7.202170835907083e-06,
+      "loss": 0.512,
+      "step": 11573
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.35564411903794,
+      "learning_rate": 7.200551825950281e-06,
+      "loss": 0.5569,
+      "step": 11574
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 25.16508216059275,
+      "learning_rate": 7.198932895608153e-06,
+      "loss": 0.5442,
+      "step": 11575
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.32847007128881,
+      "learning_rate": 7.197314044926729e-06,
+      "loss": 0.5186,
+      "step": 11576
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 20.986982417213863,
+      "learning_rate": 7.1956952739520605e-06,
+      "loss": 0.6003,
+      "step": 11577
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.43661062931299,
+      "learning_rate": 7.194076582730174e-06,
+      "loss": 0.5839,
+      "step": 11578
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.077134116025597,
+      "learning_rate": 7.192457971307107e-06,
+      "loss": 0.4945,
+      "step": 11579
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.3660693365981,
+      "learning_rate": 7.190839439728896e-06,
+      "loss": 0.6116,
+      "step": 11580
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.944513625187975,
+      "learning_rate": 7.1892209880415665e-06,
+      "loss": 0.5307,
+      "step": 11581
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.509966789598643,
+      "learning_rate": 7.18760261629115e-06,
+      "loss": 0.5284,
+      "step": 11582
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.926267509656231,
+      "learning_rate": 7.185984324523667e-06,
+      "loss": 0.6299,
+      "step": 11583
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.47240983617567,
+      "learning_rate": 7.184366112785144e-06,
+      "loss": 0.5129,
+      "step": 11584
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.714475371439356,
+      "learning_rate": 7.1827479811216e-06,
+      "loss": 0.4596,
+      "step": 11585
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.759201687900756,
+      "learning_rate": 7.181129929579058e-06,
+      "loss": 0.6174,
+      "step": 11586
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.957772109820837,
+      "learning_rate": 7.179511958203535e-06,
+      "loss": 0.5165,
+      "step": 11587
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.264233700982434,
+      "learning_rate": 7.177894067041038e-06,
+      "loss": 0.495,
+      "step": 11588
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.501972470768314,
+      "learning_rate": 7.176276256137582e-06,
+      "loss": 0.5209,
+      "step": 11589
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.99326854017606,
+      "learning_rate": 7.174658525539179e-06,
+      "loss": 0.5556,
+      "step": 11590
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.133672481410485,
+      "learning_rate": 7.173040875291832e-06,
+      "loss": 0.5433,
+      "step": 11591
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 25.741492951767032,
+      "learning_rate": 7.171423305441556e-06,
+      "loss": 0.6295,
+      "step": 11592
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.711441463504482,
+      "learning_rate": 7.169805816034341e-06,
+      "loss": 0.5306,
+      "step": 11593
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.01586118994831,
+      "learning_rate": 7.168188407116194e-06,
+      "loss": 0.4987,
+      "step": 11594
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.13723441201615,
+      "learning_rate": 7.166571078733109e-06,
+      "loss": 0.5472,
+      "step": 11595
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 22.416324911487486,
+      "learning_rate": 7.164953830931089e-06,
+      "loss": 0.5724,
+      "step": 11596
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.745983958755303,
+      "learning_rate": 7.1633366637561245e-06,
+      "loss": 0.453,
+      "step": 11597
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.556714797830292,
+      "learning_rate": 7.161719577254203e-06,
+      "loss": 0.5831,
+      "step": 11598
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.194467292067234,
+      "learning_rate": 7.160102571471317e-06,
+      "loss": 0.5713,
+      "step": 11599
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.796453285077092,
+      "learning_rate": 7.158485646453452e-06,
+      "loss": 0.5181,
+      "step": 11600
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 31.683945862802087,
+      "learning_rate": 7.156868802246593e-06,
+      "loss": 0.6248,
+      "step": 11601
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.174943732778496,
+      "learning_rate": 7.155252038896725e-06,
+      "loss": 0.5787,
+      "step": 11602
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.143311059709728,
+      "learning_rate": 7.153635356449824e-06,
+      "loss": 0.5478,
+      "step": 11603
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.6078015893191,
+      "learning_rate": 7.152018754951864e-06,
+      "loss": 0.57,
+      "step": 11604
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.892040079862152,
+      "learning_rate": 7.150402234448829e-06,
+      "loss": 0.5277,
+      "step": 11605
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.082032184911203,
+      "learning_rate": 7.1487857949866875e-06,
+      "loss": 0.5345,
+      "step": 11606
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.600869583344398,
+      "learning_rate": 7.14716943661141e-06,
+      "loss": 0.484,
+      "step": 11607
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.463483634175,
+      "learning_rate": 7.145553159368962e-06,
+      "loss": 0.5432,
+      "step": 11608
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.550240639009058,
+      "learning_rate": 7.143936963305314e-06,
+      "loss": 0.4511,
+      "step": 11609
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.326305962561612,
+      "learning_rate": 7.142320848466427e-06,
+      "loss": 0.6216,
+      "step": 11610
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 20.925693796618475,
+      "learning_rate": 7.140704814898268e-06,
+      "loss": 0.5114,
+      "step": 11611
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.197542415629997,
+      "learning_rate": 7.1390888626467855e-06,
+      "loss": 0.5002,
+      "step": 11612
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.991477213955214,
+      "learning_rate": 7.137472991757945e-06,
+      "loss": 0.5065,
+      "step": 11613
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 20.739434624590164,
+      "learning_rate": 7.1358572022776965e-06,
+      "loss": 0.4908,
+      "step": 11614
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.775300688662563,
+      "learning_rate": 7.134241494251994e-06,
+      "loss": 0.5454,
+      "step": 11615
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.93031653162277,
+      "learning_rate": 7.1326258677267916e-06,
+      "loss": 0.5446,
+      "step": 11616
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.415612695785184,
+      "learning_rate": 7.131010322748029e-06,
+      "loss": 0.4873,
+      "step": 11617
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.02539989929117,
+      "learning_rate": 7.129394859361652e-06,
+      "loss": 0.559,
+      "step": 11618
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.872117397347287,
+      "learning_rate": 7.127779477613608e-06,
+      "loss": 0.5077,
+      "step": 11619
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 24.65724182914066,
+      "learning_rate": 7.126164177549832e-06,
+      "loss": 0.6218,
+      "step": 11620
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.759147034514728,
+      "learning_rate": 7.124548959216272e-06,
+      "loss": 0.5329,
+      "step": 11621
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.038905292185966,
+      "learning_rate": 7.122933822658855e-06,
+      "loss": 0.4956,
+      "step": 11622
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 29.62218741201781,
+      "learning_rate": 7.121318767923514e-06,
+      "loss": 0.4754,
+      "step": 11623
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.502984850110522,
+      "learning_rate": 7.1197037950561855e-06,
+      "loss": 0.4947,
+      "step": 11624
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.272880233786456,
+      "learning_rate": 7.118088904102798e-06,
+      "loss": 0.5762,
+      "step": 11625
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.44003194179857,
+      "learning_rate": 7.116474095109276e-06,
+      "loss": 0.6131,
+      "step": 11626
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 28.673043687165784,
+      "learning_rate": 7.11485936812154e-06,
+      "loss": 0.5435,
+      "step": 11627
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.987354130936072,
+      "learning_rate": 7.113244723185519e-06,
+      "loss": 0.5232,
+      "step": 11628
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.152391405969357,
+      "learning_rate": 7.111630160347127e-06,
+      "loss": 0.5511,
+      "step": 11629
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.23919644881341,
+      "learning_rate": 7.110015679652286e-06,
+      "loss": 0.5772,
+      "step": 11630
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.039844517465994,
+      "learning_rate": 7.1084012811469105e-06,
+      "loss": 0.5794,
+      "step": 11631
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.622477521039478,
+      "learning_rate": 7.106786964876909e-06,
+      "loss": 0.5101,
+      "step": 11632
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.455026629330053,
+      "learning_rate": 7.10517273088819e-06,
+      "loss": 0.5568,
+      "step": 11633
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.477765151496527,
+      "learning_rate": 7.103558579226668e-06,
+      "loss": 0.6092,
+      "step": 11634
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.827347758734685,
+      "learning_rate": 7.101944509938244e-06,
+      "loss": 0.5387,
+      "step": 11635
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.319090684113004,
+      "learning_rate": 7.100330523068829e-06,
+      "loss": 0.5566,
+      "step": 11636
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.285856624828533,
+      "learning_rate": 7.098716618664312e-06,
+      "loss": 0.5493,
+      "step": 11637
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.543538877210672,
+      "learning_rate": 7.097102796770598e-06,
+      "loss": 0.6521,
+      "step": 11638
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.06960988490768,
+      "learning_rate": 7.0954890574335814e-06,
+      "loss": 0.5151,
+      "step": 11639
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.641734816259934,
+      "learning_rate": 7.0938754006991584e-06,
+      "loss": 0.456,
+      "step": 11640
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.988555649461693,
+      "learning_rate": 7.092261826613221e-06,
+      "loss": 0.4699,
+      "step": 11641
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.667874990610347,
+      "learning_rate": 7.0906483352216525e-06,
+      "loss": 0.5523,
+      "step": 11642
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 26.79399074291014,
+      "learning_rate": 7.0890349265703455e-06,
+      "loss": 0.5444,
+      "step": 11643
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.387598566527046,
+      "learning_rate": 7.0874216007051826e-06,
+      "loss": 0.5266,
+      "step": 11644
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.485098884246934,
+      "learning_rate": 7.0858083576720416e-06,
+      "loss": 0.5479,
+      "step": 11645
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.359197228988386,
+      "learning_rate": 7.084195197516812e-06,
+      "loss": 0.5273,
+      "step": 11646
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.397537041574818,
+      "learning_rate": 7.082582120285363e-06,
+      "loss": 0.5431,
+      "step": 11647
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.962690917416055,
+      "learning_rate": 7.080969126023569e-06,
+      "loss": 0.5433,
+      "step": 11648
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.65443252515218,
+      "learning_rate": 7.079356214777309e-06,
+      "loss": 0.5319,
+      "step": 11649
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.089367226996785,
+      "learning_rate": 7.077743386592448e-06,
+      "loss": 0.5205,
+      "step": 11650
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.824622964645325,
+      "learning_rate": 7.07613064151486e-06,
+      "loss": 0.5486,
+      "step": 11651
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.933723391905808,
+      "learning_rate": 7.0745179795904005e-06,
+      "loss": 0.5626,
+      "step": 11652
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 22.729480015211607,
+      "learning_rate": 7.07290540086494e-06,
+      "loss": 0.5183,
+      "step": 11653
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.888364682400976,
+      "learning_rate": 7.071292905384336e-06,
+      "loss": 0.5771,
+      "step": 11654
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.516893766652776,
+      "learning_rate": 7.069680493194455e-06,
+      "loss": 0.5666,
+      "step": 11655
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.383855888560012,
+      "learning_rate": 7.068068164341139e-06,
+      "loss": 0.4768,
+      "step": 11656
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 26.092905400935205,
+      "learning_rate": 7.066455918870254e-06,
+      "loss": 0.6187,
+      "step": 11657
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.99880162031453,
+      "learning_rate": 7.064843756827641e-06,
+      "loss": 0.5263,
+      "step": 11658
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.83135382013805,
+      "learning_rate": 7.06323167825916e-06,
+      "loss": 0.4394,
+      "step": 11659
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.115081215869775,
+      "learning_rate": 7.0616196832106535e-06,
+      "loss": 0.5675,
+      "step": 11660
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.19547223039035,
+      "learning_rate": 7.060007771727961e-06,
+      "loss": 0.5267,
+      "step": 11661
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.36535662075284,
+      "learning_rate": 7.058395943856926e-06,
+      "loss": 0.4983,
+      "step": 11662
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.881207718754755,
+      "learning_rate": 7.056784199643392e-06,
+      "loss": 0.615,
+      "step": 11663
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.98192371899534,
+      "learning_rate": 7.05517253913319e-06,
+      "loss": 0.7017,
+      "step": 11664
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.144408628215704,
+      "learning_rate": 7.053560962372166e-06,
+      "loss": 0.5324,
+      "step": 11665
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 41.209823439305936,
+      "learning_rate": 7.051949469406139e-06,
+      "loss": 0.6122,
+      "step": 11666
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.731807680259813,
+      "learning_rate": 7.0503380602809435e-06,
+      "loss": 0.532,
+      "step": 11667
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.82789525494998,
+      "learning_rate": 7.04872673504241e-06,
+      "loss": 0.4863,
+      "step": 11668
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.934443048076906,
+      "learning_rate": 7.047115493736363e-06,
+      "loss": 0.5891,
+      "step": 11669
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 25.771589173584143,
+      "learning_rate": 7.045504336408625e-06,
+      "loss": 0.5339,
+      "step": 11670
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.56569298862143,
+      "learning_rate": 7.043893263105012e-06,
+      "loss": 0.5708,
+      "step": 11671
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 32.98548931595973,
+      "learning_rate": 7.042282273871346e-06,
+      "loss": 0.5527,
+      "step": 11672
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 23.69957079357906,
+      "learning_rate": 7.04067136875344e-06,
+      "loss": 0.5612,
+      "step": 11673
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.38158897442927,
+      "learning_rate": 7.039060547797112e-06,
+      "loss": 0.6501,
+      "step": 11674
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.600871070000377,
+      "learning_rate": 7.037449811048172e-06,
+      "loss": 0.5599,
+      "step": 11675
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.809189468344815,
+      "learning_rate": 7.035839158552424e-06,
+      "loss": 0.5837,
+      "step": 11676
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.348039741530055,
+      "learning_rate": 7.034228590355674e-06,
+      "loss": 0.5329,
+      "step": 11677
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.98229254698548,
+      "learning_rate": 7.03261810650373e-06,
+      "loss": 0.6243,
+      "step": 11678
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 24.72234095593728,
+      "learning_rate": 7.03100770704239e-06,
+      "loss": 0.636,
+      "step": 11679
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.60186140150378,
+      "learning_rate": 7.029397392017461e-06,
+      "loss": 0.4827,
+      "step": 11680
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.502826870839673,
+      "learning_rate": 7.027787161474724e-06,
+      "loss": 0.5632,
+      "step": 11681
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.442930845763893,
+      "learning_rate": 7.026177015459984e-06,
+      "loss": 0.4844,
+      "step": 11682
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.428943477520082,
+      "learning_rate": 7.024566954019028e-06,
+      "loss": 0.5357,
+      "step": 11683
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.090731012707813,
+      "learning_rate": 7.0229569771976505e-06,
+      "loss": 0.4444,
+      "step": 11684
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.401333984377093,
+      "learning_rate": 7.021347085041636e-06,
+      "loss": 0.4988,
+      "step": 11685
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.31659277496896,
+      "learning_rate": 7.019737277596763e-06,
+      "loss": 0.5421,
+      "step": 11686
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.635348465567645,
+      "learning_rate": 7.018127554908823e-06,
+      "loss": 0.5676,
+      "step": 11687
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.71528537624553,
+      "learning_rate": 7.016517917023588e-06,
+      "loss": 0.5824,
+      "step": 11688
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.41710880322763,
+      "learning_rate": 7.014908363986839e-06,
+      "loss": 0.5984,
+      "step": 11689
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.762436271924965,
+      "learning_rate": 7.013298895844352e-06,
+      "loss": 0.5494,
+      "step": 11690
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.50616798945086,
+      "learning_rate": 7.011689512641896e-06,
+      "loss": 0.5149,
+      "step": 11691
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.878454605473662,
+      "learning_rate": 7.01008021442524e-06,
+      "loss": 0.5081,
+      "step": 11692
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 26.63039866301779,
+      "learning_rate": 7.008471001240156e-06,
+      "loss": 0.6059,
+      "step": 11693
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.296200632127004,
+      "learning_rate": 7.006861873132406e-06,
+      "loss": 0.5353,
+      "step": 11694
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.308023601777744,
+      "learning_rate": 7.005252830147756e-06,
+      "loss": 0.4995,
+      "step": 11695
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.613516585780115,
+      "learning_rate": 7.00364387233196e-06,
+      "loss": 0.5259,
+      "step": 11696
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.367796623519187,
+      "learning_rate": 7.00203499973078e-06,
+      "loss": 0.5112,
+      "step": 11697
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.999478345894985,
+      "learning_rate": 7.00042621238997e-06,
+      "loss": 0.5607,
+      "step": 11698
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.2043551788129,
+      "learning_rate": 6.998817510355285e-06,
+      "loss": 0.4995,
+      "step": 11699
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.29378257709534,
+      "learning_rate": 6.997208893672478e-06,
+      "loss": 0.5981,
+      "step": 11700
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.001025599766976,
+      "learning_rate": 6.9956003623872895e-06,
+      "loss": 0.5671,
+      "step": 11701
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.665063599420087,
+      "learning_rate": 6.993991916545468e-06,
+      "loss": 0.541,
+      "step": 11702
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.794120899011663,
+      "learning_rate": 6.99238355619276e-06,
+      "loss": 0.5195,
+      "step": 11703
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.06642043260306,
+      "learning_rate": 6.990775281374907e-06,
+      "loss": 0.5742,
+      "step": 11704
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.570380307139004,
+      "learning_rate": 6.989167092137639e-06,
+      "loss": 0.5394,
+      "step": 11705
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.745364045974263,
+      "learning_rate": 6.987558988526702e-06,
+      "loss": 0.5228,
+      "step": 11706
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.606251088876917,
+      "learning_rate": 6.985950970587823e-06,
+      "loss": 0.5828,
+      "step": 11707
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.743709976491427,
+      "learning_rate": 6.984343038366733e-06,
+      "loss": 0.5984,
+      "step": 11708
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.52210655499473,
+      "learning_rate": 6.982735191909168e-06,
+      "loss": 0.5115,
+      "step": 11709
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.525566713777664,
+      "learning_rate": 6.981127431260846e-06,
+      "loss": 0.5047,
+      "step": 11710
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.782272385537256,
+      "learning_rate": 6.9795197564674935e-06,
+      "loss": 0.5157,
+      "step": 11711
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 28.74157636220329,
+      "learning_rate": 6.977912167574833e-06,
+      "loss": 0.4437,
+      "step": 11712
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.292225681735477,
+      "learning_rate": 6.9763046646285836e-06,
+      "loss": 0.5652,
+      "step": 11713
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.522042738018346,
+      "learning_rate": 6.974697247674462e-06,
+      "loss": 0.5312,
+      "step": 11714
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 27.193514223376404,
+      "learning_rate": 6.973089916758177e-06,
+      "loss": 0.5129,
+      "step": 11715
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.90361913170271,
+      "learning_rate": 6.971482671925446e-06,
+      "loss": 0.5479,
+      "step": 11716
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 28.873867614946988,
+      "learning_rate": 6.969875513221974e-06,
+      "loss": 0.5831,
+      "step": 11717
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.782363282017418,
+      "learning_rate": 6.968268440693472e-06,
+      "loss": 0.5187,
+      "step": 11718
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.702251415175613,
+      "learning_rate": 6.966661454385644e-06,
+      "loss": 0.5572,
+      "step": 11719
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.902716813603863,
+      "learning_rate": 6.965054554344188e-06,
+      "loss": 0.5399,
+      "step": 11720
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.625732002420186,
+      "learning_rate": 6.963447740614804e-06,
+      "loss": 0.5546,
+      "step": 11721
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.208926691726003,
+      "learning_rate": 6.96184101324319e-06,
+      "loss": 0.5155,
+      "step": 11722
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.97838514141988,
+      "learning_rate": 6.960234372275039e-06,
+      "loss": 0.4979,
+      "step": 11723
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.449116927089577,
+      "learning_rate": 6.958627817756052e-06,
+      "loss": 0.5685,
+      "step": 11724
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.50965678122747,
+      "learning_rate": 6.957021349731902e-06,
+      "loss": 0.4852,
+      "step": 11725
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 27.984516194560854,
+      "learning_rate": 6.955414968248289e-06,
+      "loss": 0.5442,
+      "step": 11726
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.848203408271477,
+      "learning_rate": 6.953808673350889e-06,
+      "loss": 0.5545,
+      "step": 11727
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.553414325506136,
+      "learning_rate": 6.952202465085391e-06,
+      "loss": 0.5492,
+      "step": 11728
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 23.345977558031564,
+      "learning_rate": 6.950596343497475e-06,
+      "loss": 0.5099,
+      "step": 11729
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.98191970118878,
+      "learning_rate": 6.9489903086328106e-06,
+      "loss": 0.5756,
+      "step": 11730
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 35.24253605048893,
+      "learning_rate": 6.94738436053708e-06,
+      "loss": 0.4795,
+      "step": 11731
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.177290150022948,
+      "learning_rate": 6.945778499255951e-06,
+      "loss": 0.5683,
+      "step": 11732
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.957408071805656,
+      "learning_rate": 6.944172724835093e-06,
+      "loss": 0.6336,
+      "step": 11733
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.72423147978298,
+      "learning_rate": 6.94256703732018e-06,
+      "loss": 0.5884,
+      "step": 11734
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 26.55956095816552,
+      "learning_rate": 6.94096143675687e-06,
+      "loss": 0.5428,
+      "step": 11735
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.664277537580542,
+      "learning_rate": 6.939355923190823e-06,
+      "loss": 0.5702,
+      "step": 11736
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.13854389946916,
+      "learning_rate": 6.9377504966677075e-06,
+      "loss": 0.4787,
+      "step": 11737
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.473880337545149,
+      "learning_rate": 6.936145157233176e-06,
+      "loss": 0.5476,
+      "step": 11738
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.92305827347941,
+      "learning_rate": 6.934539904932887e-06,
+      "loss": 0.5446,
+      "step": 11739
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.016222793545023,
+      "learning_rate": 6.9329347398124825e-06,
+      "loss": 0.4996,
+      "step": 11740
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.33511644389723,
+      "learning_rate": 6.9313296619176255e-06,
+      "loss": 0.5833,
+      "step": 11741
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.728913362676044,
+      "learning_rate": 6.929724671293954e-06,
+      "loss": 0.5381,
+      "step": 11742
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.216073908335357,
+      "learning_rate": 6.928119767987118e-06,
+      "loss": 0.5825,
+      "step": 11743
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.530455562260306,
+      "learning_rate": 6.926514952042762e-06,
+      "loss": 0.5411,
+      "step": 11744
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.774554030780376,
+      "learning_rate": 6.9249102235065205e-06,
+      "loss": 0.5609,
+      "step": 11745
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.532517514202322,
+      "learning_rate": 6.923305582424031e-06,
+      "loss": 0.4781,
+      "step": 11746
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 28.746084930422715,
+      "learning_rate": 6.9217010288409335e-06,
+      "loss": 0.5183,
+      "step": 11747
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.037601828894335,
+      "learning_rate": 6.920096562802856e-06,
+      "loss": 0.5638,
+      "step": 11748
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 25.63200652236906,
+      "learning_rate": 6.918492184355434e-06,
+      "loss": 0.5939,
+      "step": 11749
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.951470353109308,
+      "learning_rate": 6.9168878935442905e-06,
+      "loss": 0.5804,
+      "step": 11750
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.41521443287422,
+      "learning_rate": 6.915283690415051e-06,
+      "loss": 0.6116,
+      "step": 11751
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.170861591270473,
+      "learning_rate": 6.913679575013338e-06,
+      "loss": 0.5257,
+      "step": 11752
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.677820440642755,
+      "learning_rate": 6.9120755473847755e-06,
+      "loss": 0.5979,
+      "step": 11753
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.912057962892913,
+      "learning_rate": 6.9104716075749755e-06,
+      "loss": 0.5724,
+      "step": 11754
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.12830214182361,
+      "learning_rate": 6.908867755629555e-06,
+      "loss": 0.4904,
+      "step": 11755
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.174629708546878,
+      "learning_rate": 6.907263991594129e-06,
+      "loss": 0.4541,
+      "step": 11756
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.04055811483667,
+      "learning_rate": 6.905660315514304e-06,
+      "loss": 0.554,
+      "step": 11757
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.510908032083996,
+      "learning_rate": 6.904056727435694e-06,
+      "loss": 0.4692,
+      "step": 11758
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.405438107846976,
+      "learning_rate": 6.9024532274038935e-06,
+      "loss": 0.5845,
+      "step": 11759
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 49.87042182361547,
+      "learning_rate": 6.900849815464515e-06,
+      "loss": 0.5984,
+      "step": 11760
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.197483714257224,
+      "learning_rate": 6.8992464916631515e-06,
+      "loss": 0.5514,
+      "step": 11761
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.98222614662644,
+      "learning_rate": 6.897643256045407e-06,
+      "loss": 0.5275,
+      "step": 11762
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.394025212454915,
+      "learning_rate": 6.896040108656874e-06,
+      "loss": 0.5384,
+      "step": 11763
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.957535848675242,
+      "learning_rate": 6.894437049543144e-06,
+      "loss": 0.5513,
+      "step": 11764
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.594980462785127,
+      "learning_rate": 6.892834078749805e-06,
+      "loss": 0.5522,
+      "step": 11765
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.172844572303415,
+      "learning_rate": 6.89123119632245e-06,
+      "loss": 0.5277,
+      "step": 11766
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.18619036645395,
+      "learning_rate": 6.8896284023066585e-06,
+      "loss": 0.5676,
+      "step": 11767
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.165212246182225,
+      "learning_rate": 6.888025696748021e-06,
+      "loss": 0.5836,
+      "step": 11768
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.903844497147052,
+      "learning_rate": 6.886423079692106e-06,
+      "loss": 0.4911,
+      "step": 11769
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.889666266293897,
+      "learning_rate": 6.884820551184501e-06,
+      "loss": 0.5259,
+      "step": 11770
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 59.62691749432491,
+      "learning_rate": 6.883218111270774e-06,
+      "loss": 0.6656,
+      "step": 11771
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.16000784953565,
+      "learning_rate": 6.8816157599965025e-06,
+      "loss": 0.5916,
+      "step": 11772
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.41139580947437,
+      "learning_rate": 6.880013497407257e-06,
+      "loss": 0.5374,
+      "step": 11773
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.153799914221107,
+      "learning_rate": 6.878411323548599e-06,
+      "loss": 0.565,
+      "step": 11774
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.224518948631207,
+      "learning_rate": 6.8768092384660975e-06,
+      "loss": 0.5482,
+      "step": 11775
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.499790664511192,
+      "learning_rate": 6.875207242205316e-06,
+      "loss": 0.544,
+      "step": 11776
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.849667799125257,
+      "learning_rate": 6.8736053348118106e-06,
+      "loss": 0.552,
+      "step": 11777
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.69364399524402,
+      "learning_rate": 6.8720035163311425e-06,
+      "loss": 0.5775,
+      "step": 11778
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.042298678013218,
+      "learning_rate": 6.870401786808865e-06,
+      "loss": 0.5117,
+      "step": 11779
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 23.697660090276667,
+      "learning_rate": 6.868800146290526e-06,
+      "loss": 0.5191,
+      "step": 11780
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.543585380670198,
+      "learning_rate": 6.867198594821683e-06,
+      "loss": 0.6217,
+      "step": 11781
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.137636251504162,
+      "learning_rate": 6.8655971324478764e-06,
+      "loss": 0.5715,
+      "step": 11782
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.380098124966214,
+      "learning_rate": 6.863995759214659e-06,
+      "loss": 0.5126,
+      "step": 11783
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.56032807183403,
+      "learning_rate": 6.862394475167562e-06,
+      "loss": 0.6195,
+      "step": 11784
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.78594554179766,
+      "learning_rate": 6.860793280352132e-06,
+      "loss": 0.4863,
+      "step": 11785
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.041057080859524,
+      "learning_rate": 6.859192174813905e-06,
+      "loss": 0.5814,
+      "step": 11786
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.23394728968936,
+      "learning_rate": 6.857591158598415e-06,
+      "loss": 0.4819,
+      "step": 11787
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.65858811681614,
+      "learning_rate": 6.855990231751197e-06,
+      "loss": 0.5488,
+      "step": 11788
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 30.701395605851804,
+      "learning_rate": 6.854389394317776e-06,
+      "loss": 0.6204,
+      "step": 11789
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.008493917346742,
+      "learning_rate": 6.852788646343678e-06,
+      "loss": 0.5578,
+      "step": 11790
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.51846641001329,
+      "learning_rate": 6.851187987874432e-06,
+      "loss": 0.5132,
+      "step": 11791
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.708946223078858,
+      "learning_rate": 6.849587418955554e-06,
+      "loss": 0.5397,
+      "step": 11792
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.564967495303193,
+      "learning_rate": 6.847986939632574e-06,
+      "loss": 0.4913,
+      "step": 11793
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.902148200622115,
+      "learning_rate": 6.846386549950998e-06,
+      "loss": 0.5779,
+      "step": 11794
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.169481890982244,
+      "learning_rate": 6.844786249956343e-06,
+      "loss": 0.5989,
+      "step": 11795
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.29195636306816,
+      "learning_rate": 6.8431860396941205e-06,
+      "loss": 0.5063,
+      "step": 11796
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.995587490005533,
+      "learning_rate": 6.841585919209842e-06,
+      "loss": 0.4843,
+      "step": 11797
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.299465661928968,
+      "learning_rate": 6.839985888549015e-06,
+      "loss": 0.4884,
+      "step": 11798
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.042442160481034,
+      "learning_rate": 6.838385947757136e-06,
+      "loss": 0.4733,
+      "step": 11799
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.234073573858344,
+      "learning_rate": 6.836786096879713e-06,
+      "loss": 0.4793,
+      "step": 11800
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.637346360869344,
+      "learning_rate": 6.8351863359622424e-06,
+      "loss": 0.5195,
+      "step": 11801
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.11047107254283,
+      "learning_rate": 6.833586665050225e-06,
+      "loss": 0.5508,
+      "step": 11802
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.622951219543683,
+      "learning_rate": 6.831987084189144e-06,
+      "loss": 0.5396,
+      "step": 11803
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.988993444953024,
+      "learning_rate": 6.830387593424501e-06,
+      "loss": 0.5571,
+      "step": 11804
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.64473752405375,
+      "learning_rate": 6.828788192801779e-06,
+      "loss": 0.5243,
+      "step": 11805
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.002541021424317,
+      "learning_rate": 6.827188882366467e-06,
+      "loss": 0.5972,
+      "step": 11806
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.862318553427183,
+      "learning_rate": 6.825589662164049e-06,
+      "loss": 0.4388,
+      "step": 11807
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 17.240215211181084,
+      "learning_rate": 6.823990532240001e-06,
+      "loss": 0.5396,
+      "step": 11808
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.415745600830192,
+      "learning_rate": 6.822391492639804e-06,
+      "loss": 0.5604,
+      "step": 11809
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.99398208284026,
+      "learning_rate": 6.820792543408935e-06,
+      "loss": 0.5525,
+      "step": 11810
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.17063712961563,
+      "learning_rate": 6.819193684592864e-06,
+      "loss": 0.5205,
+      "step": 11811
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.598539516072515,
+      "learning_rate": 6.817594916237071e-06,
+      "loss": 0.5296,
+      "step": 11812
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.106494857291004,
+      "learning_rate": 6.815996238387011e-06,
+      "loss": 0.5202,
+      "step": 11813
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.761730921744594,
+      "learning_rate": 6.814397651088156e-06,
+      "loss": 0.5598,
+      "step": 11814
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.505484719089626,
+      "learning_rate": 6.812799154385967e-06,
+      "loss": 0.5047,
+      "step": 11815
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 26.217824920729925,
+      "learning_rate": 6.811200748325908e-06,
+      "loss": 0.5667,
+      "step": 11816
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.357447255129127,
+      "learning_rate": 6.8096024329534376e-06,
+      "loss": 0.5417,
+      "step": 11817
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.722383620791216,
+      "learning_rate": 6.808004208314001e-06,
+      "loss": 0.5542,
+      "step": 11818
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.253706437495703,
+      "learning_rate": 6.806406074453061e-06,
+      "loss": 0.5359,
+      "step": 11819
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.598648277932625,
+      "learning_rate": 6.804808031416064e-06,
+      "loss": 0.5253,
+      "step": 11820
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 28.686280740433627,
+      "learning_rate": 6.803210079248454e-06,
+      "loss": 0.5844,
+      "step": 11821
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.462911925171134,
+      "learning_rate": 6.8016122179956855e-06,
+      "loss": 0.5614,
+      "step": 11822
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.004407491641857,
+      "learning_rate": 6.800014447703191e-06,
+      "loss": 0.6014,
+      "step": 11823
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.385313763184758,
+      "learning_rate": 6.798416768416412e-06,
+      "loss": 0.5567,
+      "step": 11824
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.957210859857593,
+      "learning_rate": 6.7968191801807894e-06,
+      "loss": 0.5931,
+      "step": 11825
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.17053963594402,
+      "learning_rate": 6.795221683041756e-06,
+      "loss": 0.5224,
+      "step": 11826
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.049286904697535,
+      "learning_rate": 6.793624277044745e-06,
+      "loss": 0.6028,
+      "step": 11827
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.505051819156407,
+      "learning_rate": 6.792026962235178e-06,
+      "loss": 0.551,
+      "step": 11828
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.627967005178707,
+      "learning_rate": 6.790429738658493e-06,
+      "loss": 0.5519,
+      "step": 11829
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 34.89721143971475,
+      "learning_rate": 6.788832606360104e-06,
+      "loss": 0.5496,
+      "step": 11830
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.622222009458074,
+      "learning_rate": 6.787235565385439e-06,
+      "loss": 0.5669,
+      "step": 11831
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.831788367733381,
+      "learning_rate": 6.785638615779919e-06,
+      "loss": 0.4961,
+      "step": 11832
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.221369891869672,
+      "learning_rate": 6.784041757588954e-06,
+      "loss": 0.4772,
+      "step": 11833
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.665110942426757,
+      "learning_rate": 6.782444990857957e-06,
+      "loss": 0.5192,
+      "step": 11834
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.58718310719083,
+      "learning_rate": 6.780848315632344e-06,
+      "loss": 0.5618,
+      "step": 11835
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.653080811745662,
+      "learning_rate": 6.779251731957522e-06,
+      "loss": 0.4769,
+      "step": 11836
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.05049685073707,
+      "learning_rate": 6.7776552398788995e-06,
+      "loss": 0.5144,
+      "step": 11837
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.362459280753125,
+      "learning_rate": 6.776058839441875e-06,
+      "loss": 0.6139,
+      "step": 11838
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.530316306937102,
+      "learning_rate": 6.77446253069185e-06,
+      "loss": 0.5723,
+      "step": 11839
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.7788743057357,
+      "learning_rate": 6.772866313674223e-06,
+      "loss": 0.5287,
+      "step": 11840
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.03395121596905,
+      "learning_rate": 6.771270188434392e-06,
+      "loss": 0.5809,
+      "step": 11841
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.256488767245816,
+      "learning_rate": 6.7696741550177505e-06,
+      "loss": 0.5326,
+      "step": 11842
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.767931118769736,
+      "learning_rate": 6.768078213469682e-06,
+      "loss": 0.526,
+      "step": 11843
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.13373259923442,
+      "learning_rate": 6.76648236383558e-06,
+      "loss": 0.5511,
+      "step": 11844
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.087789328284906,
+      "learning_rate": 6.764886606160828e-06,
+      "loss": 0.6134,
+      "step": 11845
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.54309932356595,
+      "learning_rate": 6.763290940490807e-06,
+      "loss": 0.5314,
+      "step": 11846
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 26.978805143862637,
+      "learning_rate": 6.761695366870902e-06,
+      "loss": 0.5518,
+      "step": 11847
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.006521624635567,
+      "learning_rate": 6.760099885346485e-06,
+      "loss": 0.5675,
+      "step": 11848
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.727629834750847,
+      "learning_rate": 6.758504495962929e-06,
+      "loss": 0.5407,
+      "step": 11849
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.421520238950748,
+      "learning_rate": 6.756909198765611e-06,
+      "loss": 0.5079,
+      "step": 11850
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.702382093039137,
+      "learning_rate": 6.755313993799901e-06,
+      "loss": 0.5862,
+      "step": 11851
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.637445580775932,
+      "learning_rate": 6.75371888111116e-06,
+      "loss": 0.598,
+      "step": 11852
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.19318262773005,
+      "learning_rate": 6.7521238607447536e-06,
+      "loss": 0.5322,
+      "step": 11853
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.023378612685022,
+      "learning_rate": 6.750528932746047e-06,
+      "loss": 0.5321,
+      "step": 11854
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.177275724017136,
+      "learning_rate": 6.7489340971603935e-06,
+      "loss": 0.5131,
+      "step": 11855
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.2220833791688,
+      "learning_rate": 6.747339354033158e-06,
+      "loss": 0.591,
+      "step": 11856
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.46797529746724,
+      "learning_rate": 6.745744703409685e-06,
+      "loss": 0.4741,
+      "step": 11857
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.043386639357967,
+      "learning_rate": 6.74415014533533e-06,
+      "loss": 0.5669,
+      "step": 11858
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.599418667835762,
+      "learning_rate": 6.742555679855438e-06,
+      "loss": 0.5139,
+      "step": 11859
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.676745969282397,
+      "learning_rate": 6.740961307015359e-06,
+      "loss": 0.5714,
+      "step": 11860
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.501435230617314,
+      "learning_rate": 6.739367026860435e-06,
+      "loss": 0.5779,
+      "step": 11861
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 22.944800473462895,
+      "learning_rate": 6.737772839436003e-06,
+      "loss": 0.5178,
+      "step": 11862
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.066471334164344,
+      "learning_rate": 6.736178744787403e-06,
+      "loss": 0.5734,
+      "step": 11863
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.15696972152039,
+      "learning_rate": 6.734584742959972e-06,
+      "loss": 0.5637,
+      "step": 11864
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.14204942502275,
+      "learning_rate": 6.732990833999038e-06,
+      "loss": 0.5619,
+      "step": 11865
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 27.13584361484866,
+      "learning_rate": 6.731397017949937e-06,
+      "loss": 0.5247,
+      "step": 11866
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.631445407909792,
+      "learning_rate": 6.7298032948579925e-06,
+      "loss": 0.4517,
+      "step": 11867
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.83198742297547,
+      "learning_rate": 6.728209664768525e-06,
+      "loss": 0.5518,
+      "step": 11868
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.781420560544717,
+      "learning_rate": 6.726616127726863e-06,
+      "loss": 0.4721,
+      "step": 11869
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.300662697260762,
+      "learning_rate": 6.725022683778326e-06,
+      "loss": 0.524,
+      "step": 11870
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.412647118527575,
+      "learning_rate": 6.723429332968227e-06,
+      "loss": 0.4846,
+      "step": 11871
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.994617002918577,
+      "learning_rate": 6.721836075341879e-06,
+      "loss": 0.5776,
+      "step": 11872
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.225958231151278,
+      "learning_rate": 6.720242910944596e-06,
+      "loss": 0.5125,
+      "step": 11873
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.809634243317463,
+      "learning_rate": 6.718649839821686e-06,
+      "loss": 0.5774,
+      "step": 11874
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.659502838570962,
+      "learning_rate": 6.717056862018456e-06,
+      "loss": 0.5375,
+      "step": 11875
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.93651919531852,
+      "learning_rate": 6.71546397758021e-06,
+      "loss": 0.4983,
+      "step": 11876
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 32.62854315940374,
+      "learning_rate": 6.713871186552246e-06,
+      "loss": 0.5915,
+      "step": 11877
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.088300326161473,
+      "learning_rate": 6.7122784889798596e-06,
+      "loss": 0.4891,
+      "step": 11878
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.14300445188734,
+      "learning_rate": 6.710685884908352e-06,
+      "loss": 0.5869,
+      "step": 11879
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.329452902371624,
+      "learning_rate": 6.709093374383012e-06,
+      "loss": 0.5971,
+      "step": 11880
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.83084185153524,
+      "learning_rate": 6.7075009574491355e-06,
+      "loss": 0.4683,
+      "step": 11881
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.93847415733815,
+      "learning_rate": 6.7059086341520045e-06,
+      "loss": 0.5264,
+      "step": 11882
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.919098557608034,
+      "learning_rate": 6.704316404536904e-06,
+      "loss": 0.523,
+      "step": 11883
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.745501041267495,
+      "learning_rate": 6.702724268649114e-06,
+      "loss": 0.5162,
+      "step": 11884
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.12752802442818,
+      "learning_rate": 6.70113222653392e-06,
+      "loss": 0.5366,
+      "step": 11885
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.86254287679728,
+      "learning_rate": 6.699540278236598e-06,
+      "loss": 0.536,
+      "step": 11886
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 31.498146309124703,
+      "learning_rate": 6.697948423802416e-06,
+      "loss": 0.5708,
+      "step": 11887
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.63014622476013,
+      "learning_rate": 6.696356663276651e-06,
+      "loss": 0.5801,
+      "step": 11888
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.92612413246314,
+      "learning_rate": 6.694764996704569e-06,
+      "loss": 0.6027,
+      "step": 11889
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.431043361548078,
+      "learning_rate": 6.6931734241314365e-06,
+      "loss": 0.5927,
+      "step": 11890
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.900931632374874,
+      "learning_rate": 6.69158194560252e-06,
+      "loss": 0.5302,
+      "step": 11891
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.44972644217673,
+      "learning_rate": 6.689990561163075e-06,
+      "loss": 0.483,
+      "step": 11892
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 25.53810183053534,
+      "learning_rate": 6.688399270858362e-06,
+      "loss": 0.5208,
+      "step": 11893
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.32393672489673,
+      "learning_rate": 6.686808074733638e-06,
+      "loss": 0.6336,
+      "step": 11894
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.550682988187054,
+      "learning_rate": 6.6852169728341565e-06,
+      "loss": 0.5505,
+      "step": 11895
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 28.163360533517643,
+      "learning_rate": 6.683625965205162e-06,
+      "loss": 0.5482,
+      "step": 11896
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.439995082846295,
+      "learning_rate": 6.6820350518919044e-06,
+      "loss": 0.5604,
+      "step": 11897
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.926460986782294,
+      "learning_rate": 6.680444232939629e-06,
+      "loss": 0.5129,
+      "step": 11898
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.509597936661928,
+      "learning_rate": 6.6788535083935764e-06,
+      "loss": 0.4443,
+      "step": 11899
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.854731308975214,
+      "learning_rate": 6.677262878298993e-06,
+      "loss": 0.5309,
+      "step": 11900
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.2357011253982,
+      "learning_rate": 6.675672342701107e-06,
+      "loss": 0.5588,
+      "step": 11901
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.498992077658713,
+      "learning_rate": 6.674081901645153e-06,
+      "loss": 0.5059,
+      "step": 11902
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.24220808973923,
+      "learning_rate": 6.672491555176363e-06,
+      "loss": 0.548,
+      "step": 11903
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.62993069630609,
+      "learning_rate": 6.6709013033399695e-06,
+      "loss": 0.5014,
+      "step": 11904
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.23498499360388,
+      "learning_rate": 6.6693111461811975e-06,
+      "loss": 0.5233,
+      "step": 11905
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.433866590705799,
+      "learning_rate": 6.6677210837452645e-06,
+      "loss": 0.5078,
+      "step": 11906
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.8680186939078,
+      "learning_rate": 6.6661311160773945e-06,
+      "loss": 0.531,
+      "step": 11907
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 27.097072217239205,
+      "learning_rate": 6.664541243222806e-06,
+      "loss": 0.5464,
+      "step": 11908
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.21038081373631,
+      "learning_rate": 6.662951465226713e-06,
+      "loss": 0.5657,
+      "step": 11909
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.1176575815036,
+      "learning_rate": 6.66136178213433e-06,
+      "loss": 0.5954,
+      "step": 11910
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.379564099253406,
+      "learning_rate": 6.659772193990865e-06,
+      "loss": 0.4821,
+      "step": 11911
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.86837813145166,
+      "learning_rate": 6.658182700841521e-06,
+      "loss": 0.5862,
+      "step": 11912
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 24.83615664193141,
+      "learning_rate": 6.656593302731509e-06,
+      "loss": 0.5975,
+      "step": 11913
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 27.50810701342046,
+      "learning_rate": 6.655003999706027e-06,
+      "loss": 0.5414,
+      "step": 11914
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 26.595309570946146,
+      "learning_rate": 6.653414791810277e-06,
+      "loss": 0.5938,
+      "step": 11915
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.01384182990767,
+      "learning_rate": 6.6518256790894484e-06,
+      "loss": 0.5025,
+      "step": 11916
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.942821893884318,
+      "learning_rate": 6.6502366615887415e-06,
+      "loss": 0.4598,
+      "step": 11917
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.74590904804912,
+      "learning_rate": 6.648647739353342e-06,
+      "loss": 0.6213,
+      "step": 11918
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.531460046782406,
+      "learning_rate": 6.6470589124284414e-06,
+      "loss": 0.5528,
+      "step": 11919
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.49027521427508,
+      "learning_rate": 6.645470180859227e-06,
+      "loss": 0.5561,
+      "step": 11920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.957432561850307,
+      "learning_rate": 6.643881544690876e-06,
+      "loss": 0.5572,
+      "step": 11921
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.841515348867155,
+      "learning_rate": 6.6422930039685695e-06,
+      "loss": 0.5858,
+      "step": 11922
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.962875077399254,
+      "learning_rate": 6.640704558737487e-06,
+      "loss": 0.5437,
+      "step": 11923
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.801763986755386,
+      "learning_rate": 6.639116209042798e-06,
+      "loss": 0.5837,
+      "step": 11924
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.553471164792626,
+      "learning_rate": 6.637527954929685e-06,
+      "loss": 0.5125,
+      "step": 11925
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.343271173697165,
+      "learning_rate": 6.635939796443306e-06,
+      "loss": 0.5093,
+      "step": 11926
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.16702925649879,
+      "learning_rate": 6.634351733628832e-06,
+      "loss": 0.4375,
+      "step": 11927
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.231680656249214,
+      "learning_rate": 6.632763766531423e-06,
+      "loss": 0.5767,
+      "step": 11928
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.807245196131904,
+      "learning_rate": 6.631175895196245e-06,
+      "loss": 0.4982,
+      "step": 11929
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.58400584091568,
+      "learning_rate": 6.629588119668458e-06,
+      "loss": 0.5558,
+      "step": 11930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 25.211921388987264,
+      "learning_rate": 6.628000439993207e-06,
+      "loss": 0.581,
+      "step": 11931
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.20751505527256,
+      "learning_rate": 6.626412856215653e-06,
+      "loss": 0.5769,
+      "step": 11932
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.05257301884939,
+      "learning_rate": 6.624825368380944e-06,
+      "loss": 0.5327,
+      "step": 11933
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.78064218766533,
+      "learning_rate": 6.623237976534225e-06,
+      "loss": 0.5307,
+      "step": 11934
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.038876802576763,
+      "learning_rate": 6.6216506807206475e-06,
+      "loss": 0.5385,
+      "step": 11935
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 27.964040149215766,
+      "learning_rate": 6.620063480985346e-06,
+      "loss": 0.591,
+      "step": 11936
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.065915284779983,
+      "learning_rate": 6.61847637737346e-06,
+      "loss": 0.5,
+      "step": 11937
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 25.694112426091873,
+      "learning_rate": 6.6168893699301286e-06,
+      "loss": 0.6282,
+      "step": 11938
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.02899333407536,
+      "learning_rate": 6.6153024587004855e-06,
+      "loss": 0.5439,
+      "step": 11939
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.54619790065212,
+      "learning_rate": 6.613715643729661e-06,
+      "loss": 0.52,
+      "step": 11940
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.125489193222258,
+      "learning_rate": 6.612128925062781e-06,
+      "loss": 0.5128,
+      "step": 11941
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.600904049204516,
+      "learning_rate": 6.610542302744973e-06,
+      "loss": 0.4713,
+      "step": 11942
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.244118339483844,
+      "learning_rate": 6.608955776821357e-06,
+      "loss": 0.5841,
+      "step": 11943
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.010179813372737,
+      "learning_rate": 6.60736934733706e-06,
+      "loss": 0.5324,
+      "step": 11944
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.17303308343498,
+      "learning_rate": 6.605783014337192e-06,
+      "loss": 0.58,
+      "step": 11945
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 24.99588703267362,
+      "learning_rate": 6.6041967778668695e-06,
+      "loss": 0.5321,
+      "step": 11946
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.759150767172045,
+      "learning_rate": 6.602610637971201e-06,
+      "loss": 0.5313,
+      "step": 11947
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.140154150185321,
+      "learning_rate": 6.601024594695302e-06,
+      "loss": 0.5859,
+      "step": 11948
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 26.09471790037621,
+      "learning_rate": 6.599438648084277e-06,
+      "loss": 0.5384,
+      "step": 11949
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 22.696077316843915,
+      "learning_rate": 6.597852798183222e-06,
+      "loss": 0.5021,
+      "step": 11950
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 22.99404933000497,
+      "learning_rate": 6.596267045037247e-06,
+      "loss": 0.4996,
+      "step": 11951
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.844594934604636,
+      "learning_rate": 6.594681388691445e-06,
+      "loss": 0.5581,
+      "step": 11952
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.823244354999975,
+      "learning_rate": 6.593095829190911e-06,
+      "loss": 0.5456,
+      "step": 11953
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.63540224995561,
+      "learning_rate": 6.591510366580742e-06,
+      "loss": 0.5803,
+      "step": 11954
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 25.34283976913497,
+      "learning_rate": 6.589925000906023e-06,
+      "loss": 0.6194,
+      "step": 11955
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.68611024235323,
+      "learning_rate": 6.588339732211839e-06,
+      "loss": 0.5292,
+      "step": 11956
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.242855904238755,
+      "learning_rate": 6.586754560543281e-06,
+      "loss": 0.5524,
+      "step": 11957
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.19504177016277,
+      "learning_rate": 6.585169485945425e-06,
+      "loss": 0.5544,
+      "step": 11958
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.036393206621653,
+      "learning_rate": 6.583584508463354e-06,
+      "loss": 0.5281,
+      "step": 11959
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.209893802917698,
+      "learning_rate": 6.581999628142137e-06,
+      "loss": 0.5102,
+      "step": 11960
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.44764380716063,
+      "learning_rate": 6.580414845026853e-06,
+      "loss": 0.6065,
+      "step": 11961
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.453680625751677,
+      "learning_rate": 6.578830159162568e-06,
+      "loss": 0.5405,
+      "step": 11962
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.52716847491306,
+      "learning_rate": 6.577245570594355e-06,
+      "loss": 0.5187,
+      "step": 11963
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.793590153852755,
+      "learning_rate": 6.575661079367277e-06,
+      "loss": 0.5581,
+      "step": 11964
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.835090525624217,
+      "learning_rate": 6.574076685526392e-06,
+      "loss": 0.5464,
+      "step": 11965
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.724116884577217,
+      "learning_rate": 6.572492389116761e-06,
+      "loss": 0.5823,
+      "step": 11966
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.399030876242543,
+      "learning_rate": 6.570908190183442e-06,
+      "loss": 0.575,
+      "step": 11967
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.239860788201426,
+      "learning_rate": 6.569324088771488e-06,
+      "loss": 0.5382,
+      "step": 11968
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.204370771723788,
+      "learning_rate": 6.5677400849259535e-06,
+      "loss": 0.5312,
+      "step": 11969
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.074923913451805,
+      "learning_rate": 6.566156178691882e-06,
+      "loss": 0.5588,
+      "step": 11970
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.565930506182092,
+      "learning_rate": 6.564572370114318e-06,
+      "loss": 0.5463,
+      "step": 11971
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.476208829247197,
+      "learning_rate": 6.562988659238305e-06,
+      "loss": 0.4763,
+      "step": 11972
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.253953913921627,
+      "learning_rate": 6.561405046108886e-06,
+      "loss": 0.5453,
+      "step": 11973
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 31.662751848120546,
+      "learning_rate": 6.5598215307710975e-06,
+      "loss": 0.5514,
+      "step": 11974
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.379043718630413,
+      "learning_rate": 6.55823811326997e-06,
+      "loss": 0.4958,
+      "step": 11975
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.306475433529185,
+      "learning_rate": 6.5566547936505375e-06,
+      "loss": 0.4861,
+      "step": 11976
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.781586888938156,
+      "learning_rate": 6.5550715719578295e-06,
+      "loss": 0.5202,
+      "step": 11977
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.454892737963295,
+      "learning_rate": 6.5534884482368665e-06,
+      "loss": 0.5321,
+      "step": 11978
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 30.386922081657993,
+      "learning_rate": 6.551905422532683e-06,
+      "loss": 0.5595,
+      "step": 11979
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.129213434522658,
+      "learning_rate": 6.550322494890289e-06,
+      "loss": 0.5749,
+      "step": 11980
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 31.121168642850527,
+      "learning_rate": 6.548739665354702e-06,
+      "loss": 0.6267,
+      "step": 11981
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 26.492847846690314,
+      "learning_rate": 6.547156933970943e-06,
+      "loss": 0.4809,
+      "step": 11982
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.42820274279954,
+      "learning_rate": 6.545574300784021e-06,
+      "loss": 0.5006,
+      "step": 11983
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.74027669340231,
+      "learning_rate": 6.543991765838946e-06,
+      "loss": 0.4851,
+      "step": 11984
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.6468925805872,
+      "learning_rate": 6.542409329180721e-06,
+      "loss": 0.5231,
+      "step": 11985
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.411628075991704,
+      "learning_rate": 6.540826990854353e-06,
+      "loss": 0.4959,
+      "step": 11986
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.310887480824633,
+      "learning_rate": 6.53924475090484e-06,
+      "loss": 0.641,
+      "step": 11987
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.091250560901004,
+      "learning_rate": 6.537662609377184e-06,
+      "loss": 0.4852,
+      "step": 11988
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.301149613096703,
+      "learning_rate": 6.53608056631638e-06,
+      "loss": 0.536,
+      "step": 11989
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.0324518614721,
+      "learning_rate": 6.534498621767418e-06,
+      "loss": 0.568,
+      "step": 11990
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.685525064723162,
+      "learning_rate": 6.5329167757752835e-06,
+      "loss": 0.5087,
+      "step": 11991
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.139792904897877,
+      "learning_rate": 6.531335028384971e-06,
+      "loss": 0.5631,
+      "step": 11992
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 30.70380287198627,
+      "learning_rate": 6.529753379641462e-06,
+      "loss": 0.5151,
+      "step": 11993
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.420217631754886,
+      "learning_rate": 6.528171829589734e-06,
+      "loss": 0.5444,
+      "step": 11994
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.095784423662707,
+      "learning_rate": 6.52659037827477e-06,
+      "loss": 0.5532,
+      "step": 11995
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.702125868921176,
+      "learning_rate": 6.525009025741543e-06,
+      "loss": 0.5825,
+      "step": 11996
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 22.49969019246063,
+      "learning_rate": 6.523427772035024e-06,
+      "loss": 0.569,
+      "step": 11997
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.672128056363587,
+      "learning_rate": 6.521846617200191e-06,
+      "loss": 0.5452,
+      "step": 11998
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.031964799807888,
+      "learning_rate": 6.520265561282004e-06,
+      "loss": 0.4891,
+      "step": 11999
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.824851233042082,
+      "learning_rate": 6.5186846043254246e-06,
+      "loss": 0.5455,
+      "step": 12000
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.722747281923528,
+      "learning_rate": 6.5171037463754224e-06,
+      "loss": 0.4926,
+      "step": 12001
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.635381943777627,
+      "learning_rate": 6.515522987476952e-06,
+      "loss": 0.5578,
+      "step": 12002
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.64048423798505,
+      "learning_rate": 6.513942327674971e-06,
+      "loss": 0.5128,
+      "step": 12003
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.054445207350017,
+      "learning_rate": 6.512361767014427e-06,
+      "loss": 0.4807,
+      "step": 12004
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 26.537284136804757,
+      "learning_rate": 6.510781305540277e-06,
+      "loss": 0.5757,
+      "step": 12005
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 24.2552813390368,
+      "learning_rate": 6.509200943297463e-06,
+      "loss": 0.5545,
+      "step": 12006
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.618808972273953,
+      "learning_rate": 6.507620680330933e-06,
+      "loss": 0.558,
+      "step": 12007
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.196480106087815,
+      "learning_rate": 6.50604051668563e-06,
+      "loss": 0.4874,
+      "step": 12008
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.234040437030032,
+      "learning_rate": 6.504460452406489e-06,
+      "loss": 0.5577,
+      "step": 12009
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.218088589459896,
+      "learning_rate": 6.502880487538448e-06,
+      "loss": 0.5357,
+      "step": 12010
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.23254004343529,
+      "learning_rate": 6.50130062212644e-06,
+      "loss": 0.4843,
+      "step": 12011
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.948171562680933,
+      "learning_rate": 6.499720856215395e-06,
+      "loss": 0.5707,
+      "step": 12012
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.484562485962197,
+      "learning_rate": 6.498141189850243e-06,
+      "loss": 0.5083,
+      "step": 12013
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 22.168233041406186,
+      "learning_rate": 6.496561623075907e-06,
+      "loss": 0.6047,
+      "step": 12014
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.80088044407739,
+      "learning_rate": 6.494982155937308e-06,
+      "loss": 0.5362,
+      "step": 12015
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 24.10934125733502,
+      "learning_rate": 6.493402788479365e-06,
+      "loss": 0.614,
+      "step": 12016
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.099278274792237,
+      "learning_rate": 6.491823520746996e-06,
+      "loss": 0.4768,
+      "step": 12017
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.835058640332303,
+      "learning_rate": 6.4902443527851175e-06,
+      "loss": 0.4981,
+      "step": 12018
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.48080520220848,
+      "learning_rate": 6.488665284638631e-06,
+      "loss": 0.5364,
+      "step": 12019
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.531853567924134,
+      "learning_rate": 6.487086316352453e-06,
+      "loss": 0.4992,
+      "step": 12020
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.790879642131458,
+      "learning_rate": 6.4855074479714845e-06,
+      "loss": 0.588,
+      "step": 12021
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.51269961058938,
+      "learning_rate": 6.483928679540627e-06,
+      "loss": 0.5755,
+      "step": 12022
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 27.226335474214363,
+      "learning_rate": 6.482350011104785e-06,
+      "loss": 0.6003,
+      "step": 12023
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.332537974278345,
+      "learning_rate": 6.480771442708847e-06,
+      "loss": 0.4889,
+      "step": 12024
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 20.805310986666903,
+      "learning_rate": 6.47919297439771e-06,
+      "loss": 0.581,
+      "step": 12025
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 20.600358841777485,
+      "learning_rate": 6.477614606216267e-06,
+      "loss": 0.5528,
+      "step": 12026
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.761596828209886,
+      "learning_rate": 6.476036338209404e-06,
+      "loss": 0.613,
+      "step": 12027
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.9691599011309,
+      "learning_rate": 6.474458170422009e-06,
+      "loss": 0.5067,
+      "step": 12028
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 22.95342617798133,
+      "learning_rate": 6.4728801028989556e-06,
+      "loss": 0.584,
+      "step": 12029
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.458857956363488,
+      "learning_rate": 6.471302135685131e-06,
+      "loss": 0.5122,
+      "step": 12030
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 20.434727997712663,
+      "learning_rate": 6.4697242688254095e-06,
+      "loss": 0.5729,
+      "step": 12031
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.332635926544123,
+      "learning_rate": 6.468146502364665e-06,
+      "loss": 0.5738,
+      "step": 12032
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.16065222525277,
+      "learning_rate": 6.46656883634777e-06,
+      "loss": 0.5258,
+      "step": 12033
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.332691181118758,
+      "learning_rate": 6.464991270819587e-06,
+      "loss": 0.5065,
+      "step": 12034
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.425591228550076,
+      "learning_rate": 6.463413805824985e-06,
+      "loss": 0.5185,
+      "step": 12035
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 25.27596465451524,
+      "learning_rate": 6.4618364414088264e-06,
+      "loss": 0.5113,
+      "step": 12036
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 25.13137946699103,
+      "learning_rate": 6.460259177615966e-06,
+      "loss": 0.5017,
+      "step": 12037
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.078665961157515,
+      "learning_rate": 6.458682014491271e-06,
+      "loss": 0.5111,
+      "step": 12038
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 22.039267404476618,
+      "learning_rate": 6.4571049520795836e-06,
+      "loss": 0.6013,
+      "step": 12039
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.297309446350216,
+      "learning_rate": 6.45552799042576e-06,
+      "loss": 0.4705,
+      "step": 12040
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.868311334626522,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.5293,
+      "step": 12041
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.002769508056005,
+      "learning_rate": 6.452374369571088e-06,
+      "loss": 0.4852,
+      "step": 12042
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.531897925985582,
+      "learning_rate": 6.4507977104599275e-06,
+      "loss": 0.5412,
+      "step": 12043
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.400456852549771,
+      "learning_rate": 6.449221152286001e-06,
+      "loss": 0.552,
+      "step": 12044
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.919692352216606,
+      "learning_rate": 6.447644695094151e-06,
+      "loss": 0.5007,
+      "step": 12045
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.454047197725393,
+      "learning_rate": 6.446068338929208e-06,
+      "loss": 0.4966,
+      "step": 12046
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.529276509635686,
+      "learning_rate": 6.444492083836004e-06,
+      "loss": 0.5484,
+      "step": 12047
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.080353742944991,
+      "learning_rate": 6.442915929859361e-06,
+      "loss": 0.5258,
+      "step": 12048
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.57137340668942,
+      "learning_rate": 6.44133987704411e-06,
+      "loss": 0.4786,
+      "step": 12049
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 20.809829112026197,
+      "learning_rate": 6.43976392543507e-06,
+      "loss": 0.5064,
+      "step": 12050
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.35816842140582,
+      "learning_rate": 6.438188075077062e-06,
+      "loss": 0.5983,
+      "step": 12051
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.744822945028577,
+      "learning_rate": 6.436612326014904e-06,
+      "loss": 0.4876,
+      "step": 12052
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.071159078858736,
+      "learning_rate": 6.435036678293405e-06,
+      "loss": 0.4982,
+      "step": 12053
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.206086003062207,
+      "learning_rate": 6.433461131957376e-06,
+      "loss": 0.5777,
+      "step": 12054
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.565210411623358,
+      "learning_rate": 6.431885687051628e-06,
+      "loss": 0.5207,
+      "step": 12055
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.904620525914225,
+      "learning_rate": 6.43031034362096e-06,
+      "loss": 0.521,
+      "step": 12056
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.268963539661176,
+      "learning_rate": 6.428735101710184e-06,
+      "loss": 0.5785,
+      "step": 12057
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.266283307890372,
+      "learning_rate": 6.427159961364089e-06,
+      "loss": 0.5278,
+      "step": 12058
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.946035343048436,
+      "learning_rate": 6.425584922627476e-06,
+      "loss": 0.6558,
+      "step": 12059
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 20.4069344170661,
+      "learning_rate": 6.424009985545134e-06,
+      "loss": 0.5802,
+      "step": 12060
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 26.070471611371907,
+      "learning_rate": 6.422435150161859e-06,
+      "loss": 0.5259,
+      "step": 12061
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.002173108227096,
+      "learning_rate": 6.4208604165224366e-06,
+      "loss": 0.5054,
+      "step": 12062
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 29.44864237138187,
+      "learning_rate": 6.419285784671645e-06,
+      "loss": 0.5329,
+      "step": 12063
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.694834885448381,
+      "learning_rate": 6.417711254654275e-06,
+      "loss": 0.5357,
+      "step": 12064
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.190361391668823,
+      "learning_rate": 6.4161368265151e-06,
+      "loss": 0.5836,
+      "step": 12065
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.61209067726471,
+      "learning_rate": 6.414562500298896e-06,
+      "loss": 0.5106,
+      "step": 12066
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.921850059005422,
+      "learning_rate": 6.412988276050441e-06,
+      "loss": 0.5209,
+      "step": 12067
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.85677272130333,
+      "learning_rate": 6.411414153814499e-06,
+      "loss": 0.5596,
+      "step": 12068
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 26.65474902585938,
+      "learning_rate": 6.409840133635837e-06,
+      "loss": 0.5578,
+      "step": 12069
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.700509780796263,
+      "learning_rate": 6.408266215559225e-06,
+      "loss": 0.5157,
+      "step": 12070
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.823636016248656,
+      "learning_rate": 6.406692399629418e-06,
+      "loss": 0.5578,
+      "step": 12071
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.114453731968332,
+      "learning_rate": 6.405118685891181e-06,
+      "loss": 0.5969,
+      "step": 12072
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.228801245186332,
+      "learning_rate": 6.403545074389261e-06,
+      "loss": 0.6072,
+      "step": 12073
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.891383374977051,
+      "learning_rate": 6.401971565168416e-06,
+      "loss": 0.5766,
+      "step": 12074
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 19.5082393992152,
+      "learning_rate": 6.400398158273395e-06,
+      "loss": 0.613,
+      "step": 12075
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.321157001973747,
+      "learning_rate": 6.398824853748946e-06,
+      "loss": 0.491,
+      "step": 12076
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.11424374184085,
+      "learning_rate": 6.397251651639813e-06,
+      "loss": 0.5516,
+      "step": 12077
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.155877556280513,
+      "learning_rate": 6.395678551990735e-06,
+      "loss": 0.5044,
+      "step": 12078
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.834961961666178,
+      "learning_rate": 6.394105554846447e-06,
+      "loss": 0.5645,
+      "step": 12079
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 27.123841943828932,
+      "learning_rate": 6.3925326602516904e-06,
+      "loss": 0.5897,
+      "step": 12080
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 28.80286386540562,
+      "learning_rate": 6.390959868251192e-06,
+      "loss": 0.5403,
+      "step": 12081
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.309276944286207,
+      "learning_rate": 6.389387178889688e-06,
+      "loss": 0.5799,
+      "step": 12082
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.868157630875439,
+      "learning_rate": 6.3878145922118985e-06,
+      "loss": 0.5274,
+      "step": 12083
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.53124128967179,
+      "learning_rate": 6.3862421082625495e-06,
+      "loss": 0.4512,
+      "step": 12084
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.752414053072727,
+      "learning_rate": 6.3846697270863576e-06,
+      "loss": 0.557,
+      "step": 12085
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.740643366025363,
+      "learning_rate": 6.383097448728046e-06,
+      "loss": 0.5528,
+      "step": 12086
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.205294359221694,
+      "learning_rate": 6.3815252732323304e-06,
+      "loss": 0.5682,
+      "step": 12087
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.470416003879741,
+      "learning_rate": 6.379953200643915e-06,
+      "loss": 0.5613,
+      "step": 12088
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.122439780259775,
+      "learning_rate": 6.378381231007514e-06,
+      "loss": 0.472,
+      "step": 12089
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.79463706046399,
+      "learning_rate": 6.376809364367833e-06,
+      "loss": 0.5723,
+      "step": 12090
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.655461484572395,
+      "learning_rate": 6.375237600769575e-06,
+      "loss": 0.5336,
+      "step": 12091
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.200802081020502,
+      "learning_rate": 6.373665940257436e-06,
+      "loss": 0.4941,
+      "step": 12092
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 27.97962465335117,
+      "learning_rate": 6.37209438287612e-06,
+      "loss": 0.5426,
+      "step": 12093
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.244034309401638,
+      "learning_rate": 6.370522928670313e-06,
+      "loss": 0.5486,
+      "step": 12094
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.552523041911183,
+      "learning_rate": 6.368951577684713e-06,
+      "loss": 0.5198,
+      "step": 12095
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.469898056492717,
+      "learning_rate": 6.367380329964009e-06,
+      "loss": 0.534,
+      "step": 12096
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 19.157528948134143,
+      "learning_rate": 6.36580918555288e-06,
+      "loss": 0.5386,
+      "step": 12097
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.377336372007253,
+      "learning_rate": 6.364238144496011e-06,
+      "loss": 0.5521,
+      "step": 12098
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.245526389650948,
+      "learning_rate": 6.362667206838085e-06,
+      "loss": 0.5953,
+      "step": 12099
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.804114905249342,
+      "learning_rate": 6.361096372623773e-06,
+      "loss": 0.5421,
+      "step": 12100
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.342537276912662,
+      "learning_rate": 6.359525641897755e-06,
+      "loss": 0.5526,
+      "step": 12101
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.77301974320269,
+      "learning_rate": 6.357955014704698e-06,
+      "loss": 0.4785,
+      "step": 12102
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.091841981690127,
+      "learning_rate": 6.356384491089267e-06,
+      "loss": 0.4871,
+      "step": 12103
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.33809527510175,
+      "learning_rate": 6.354814071096129e-06,
+      "loss": 0.593,
+      "step": 12104
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.102190394788074,
+      "learning_rate": 6.353243754769948e-06,
+      "loss": 0.5027,
+      "step": 12105
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.060306430312203,
+      "learning_rate": 6.351673542155385e-06,
+      "loss": 0.4962,
+      "step": 12106
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.6670855382289,
+      "learning_rate": 6.350103433297087e-06,
+      "loss": 0.5085,
+      "step": 12107
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.680164550481418,
+      "learning_rate": 6.348533428239714e-06,
+      "loss": 0.634,
+      "step": 12108
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.489945181727462,
+      "learning_rate": 6.346963527027914e-06,
+      "loss": 0.5525,
+      "step": 12109
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.19784392701362,
+      "learning_rate": 6.345393729706333e-06,
+      "loss": 0.546,
+      "step": 12110
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 19.954025546209017,
+      "learning_rate": 6.343824036319622e-06,
+      "loss": 0.5526,
+      "step": 12111
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.34881135030741,
+      "learning_rate": 6.3422544469124116e-06,
+      "loss": 0.5157,
+      "step": 12112
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.118293256647107,
+      "learning_rate": 6.340684961529344e-06,
+      "loss": 0.5003,
+      "step": 12113
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 29.154600797479603,
+      "learning_rate": 6.3391155802150585e-06,
+      "loss": 0.5921,
+      "step": 12114
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.12816342673484,
+      "learning_rate": 6.337546303014182e-06,
+      "loss": 0.58,
+      "step": 12115
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.339794443128238,
+      "learning_rate": 6.3359771299713515e-06,
+      "loss": 0.5267,
+      "step": 12116
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.634952107227992,
+      "learning_rate": 6.334408061131182e-06,
+      "loss": 0.6128,
+      "step": 12117
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.08689824993263,
+      "learning_rate": 6.332839096538306e-06,
+      "loss": 0.5128,
+      "step": 12118
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 32.108857005412496,
+      "learning_rate": 6.331270236237337e-06,
+      "loss": 0.6248,
+      "step": 12119
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.527985321504907,
+      "learning_rate": 6.3297014802729e-06,
+      "loss": 0.5667,
+      "step": 12120
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.067530380097878,
+      "learning_rate": 6.328132828689607e-06,
+      "loss": 0.5479,
+      "step": 12121
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.4930213472013,
+      "learning_rate": 6.326564281532066e-06,
+      "loss": 0.5874,
+      "step": 12122
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.555763306628627,
+      "learning_rate": 6.324995838844887e-06,
+      "loss": 0.5761,
+      "step": 12123
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.541109242948549,
+      "learning_rate": 6.3234275006726785e-06,
+      "loss": 0.5583,
+      "step": 12124
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 22.056090374825953,
+      "learning_rate": 6.321859267060039e-06,
+      "loss": 0.5633,
+      "step": 12125
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 22.46125347346427,
+      "learning_rate": 6.320291138051574e-06,
+      "loss": 0.5139,
+      "step": 12126
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.564255330388068,
+      "learning_rate": 6.318723113691874e-06,
+      "loss": 0.5711,
+      "step": 12127
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.729528991683672,
+      "learning_rate": 6.317155194025536e-06,
+      "loss": 0.514,
+      "step": 12128
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.953630778900546,
+      "learning_rate": 6.315587379097148e-06,
+      "loss": 0.4967,
+      "step": 12129
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 19.196151424732893,
+      "learning_rate": 6.314019668951301e-06,
+      "loss": 0.4802,
+      "step": 12130
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 28.745074756786046,
+      "learning_rate": 6.312452063632581e-06,
+      "loss": 0.6421,
+      "step": 12131
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.89685114620413,
+      "learning_rate": 6.310884563185562e-06,
+      "loss": 0.5391,
+      "step": 12132
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.51968135207296,
+      "learning_rate": 6.309317167654832e-06,
+      "loss": 0.5695,
+      "step": 12133
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.21754217152341,
+      "learning_rate": 6.307749877084961e-06,
+      "loss": 0.5368,
+      "step": 12134
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 15.802725015764079,
+      "learning_rate": 6.306182691520522e-06,
+      "loss": 0.5337,
+      "step": 12135
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 21.475203937732335,
+      "learning_rate": 6.304615611006091e-06,
+      "loss": 0.5122,
+      "step": 12136
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.803607309789403,
+      "learning_rate": 6.303048635586226e-06,
+      "loss": 0.5234,
+      "step": 12137
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.6900878393442,
+      "learning_rate": 6.301481765305495e-06,
+      "loss": 0.4845,
+      "step": 12138
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.536274392563245,
+      "learning_rate": 6.2999150002084605e-06,
+      "loss": 0.4481,
+      "step": 12139
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.34631101077166,
+      "learning_rate": 6.298348340339681e-06,
+      "loss": 0.4614,
+      "step": 12140
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.733572981706885,
+      "learning_rate": 6.296781785743708e-06,
+      "loss": 0.4746,
+      "step": 12141
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.696732479182458,
+      "learning_rate": 6.295215336465094e-06,
+      "loss": 0.5664,
+      "step": 12142
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 23.773605769800138,
+      "learning_rate": 6.293648992548389e-06,
+      "loss": 0.5913,
+      "step": 12143
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 27.63732583878765,
+      "learning_rate": 6.292082754038135e-06,
+      "loss": 0.5831,
+      "step": 12144
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.396089979180317,
+      "learning_rate": 6.2905166209788845e-06,
+      "loss": 0.552,
+      "step": 12145
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.33877369740794,
+      "learning_rate": 6.288950593415168e-06,
+      "loss": 0.4786,
+      "step": 12146
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 22.863567746112626,
+      "learning_rate": 6.2873846713915276e-06,
+      "loss": 0.5624,
+      "step": 12147
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 35.01740979961583,
+      "learning_rate": 6.285818854952492e-06,
+      "loss": 0.5152,
+      "step": 12148
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 26.63655857307103,
+      "learning_rate": 6.284253144142598e-06,
+      "loss": 0.5377,
+      "step": 12149
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.547387084538498,
+      "learning_rate": 6.282687539006371e-06,
+      "loss": 0.4411,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 22.093835304451684,
+      "learning_rate": 6.2811220395883346e-06,
+      "loss": 0.5887,
+      "step": 12151
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.83116213730814,
+      "learning_rate": 6.279556645933011e-06,
+      "loss": 0.4992,
+      "step": 12152
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.305860641229682,
+      "learning_rate": 6.277991358084923e-06,
+      "loss": 0.4904,
+      "step": 12153
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 23.656308158944118,
+      "learning_rate": 6.276426176088579e-06,
+      "loss": 0.5123,
+      "step": 12154
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 37.919970446881415,
+      "learning_rate": 6.274861099988502e-06,
+      "loss": 0.4972,
+      "step": 12155
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 22.10413727202575,
+      "learning_rate": 6.273296129829194e-06,
+      "loss": 0.4898,
+      "step": 12156
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.42871961564321,
+      "learning_rate": 6.271731265655161e-06,
+      "loss": 0.4595,
+      "step": 12157
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.745079032566384,
+      "learning_rate": 6.270166507510913e-06,
+      "loss": 0.5628,
+      "step": 12158
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.419873320898557,
+      "learning_rate": 6.268601855440944e-06,
+      "loss": 0.5304,
+      "step": 12159
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.7436566319673,
+      "learning_rate": 6.267037309489763e-06,
+      "loss": 0.5514,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 30.18045657385437,
+      "learning_rate": 6.265472869701851e-06,
+      "loss": 0.5706,
+      "step": 12161
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.326844633949047,
+      "learning_rate": 6.2639085361217075e-06,
+      "loss": 0.5038,
+      "step": 12162
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.806943304338365,
+      "learning_rate": 6.262344308793818e-06,
+      "loss": 0.5406,
+      "step": 12163
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.500973005289495,
+      "learning_rate": 6.2607801877626715e-06,
+      "loss": 0.5545,
+      "step": 12164
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.16940634386228,
+      "learning_rate": 6.259216173072751e-06,
+      "loss": 0.5182,
+      "step": 12165
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.769873816215153,
+      "learning_rate": 6.257652264768532e-06,
+      "loss": 0.5135,
+      "step": 12166
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 22.583136667244826,
+      "learning_rate": 6.256088462894492e-06,
+      "loss": 0.4973,
+      "step": 12167
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.71962303544123,
+      "learning_rate": 6.254524767495108e-06,
+      "loss": 0.5374,
+      "step": 12168
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 15.828915716531528,
+      "learning_rate": 6.252961178614847e-06,
+      "loss": 0.5273,
+      "step": 12169
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.488331191192067,
+      "learning_rate": 6.251397696298181e-06,
+      "loss": 0.5523,
+      "step": 12170
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 25.26685122142957,
+      "learning_rate": 6.24983432058957e-06,
+      "loss": 0.4952,
+      "step": 12171
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 22.96508958934432,
+      "learning_rate": 6.2482710515334775e-06,
+      "loss": 0.5503,
+      "step": 12172
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 25.732194557995523,
+      "learning_rate": 6.246707889174359e-06,
+      "loss": 0.6013,
+      "step": 12173
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.552280537380508,
+      "learning_rate": 6.245144833556675e-06,
+      "loss": 0.5129,
+      "step": 12174
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 23.539379121541863,
+      "learning_rate": 6.243581884724879e-06,
+      "loss": 0.5449,
+      "step": 12175
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.429760648118778,
+      "learning_rate": 6.242019042723411e-06,
+      "loss": 0.4941,
+      "step": 12176
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.374425665741384,
+      "learning_rate": 6.240456307596725e-06,
+      "loss": 0.5319,
+      "step": 12177
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.315272535988825,
+      "learning_rate": 6.238893679389262e-06,
+      "loss": 0.5312,
+      "step": 12178
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.871648856812556,
+      "learning_rate": 6.237331158145461e-06,
+      "loss": 0.5329,
+      "step": 12179
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.468476990643177,
+      "learning_rate": 6.2357687439097645e-06,
+      "loss": 0.5216,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.178170901760858,
+      "learning_rate": 6.234206436726602e-06,
+      "loss": 0.5029,
+      "step": 12181
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.410333693919632,
+      "learning_rate": 6.232644236640403e-06,
+      "loss": 0.468,
+      "step": 12182
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.020185020285673,
+      "learning_rate": 6.231082143695599e-06,
+      "loss": 0.4947,
+      "step": 12183
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.449760025430884,
+      "learning_rate": 6.229520157936617e-06,
+      "loss": 0.538,
+      "step": 12184
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.12944773331522,
+      "learning_rate": 6.2279582794078755e-06,
+      "loss": 0.4938,
+      "step": 12185
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.59313569376448,
+      "learning_rate": 6.226396508153789e-06,
+      "loss": 0.5377,
+      "step": 12186
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.33269462872682,
+      "learning_rate": 6.2248348442187825e-06,
+      "loss": 0.5134,
+      "step": 12187
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.566752857629098,
+      "learning_rate": 6.223273287647262e-06,
+      "loss": 0.5049,
+      "step": 12188
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.16600884295014,
+      "learning_rate": 6.221711838483644e-06,
+      "loss": 0.569,
+      "step": 12189
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.89655233224021,
+      "learning_rate": 6.220150496772328e-06,
+      "loss": 0.4793,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.262054701911623,
+      "learning_rate": 6.2185892625577215e-06,
+      "loss": 0.5134,
+      "step": 12191
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.098562206397553,
+      "learning_rate": 6.217028135884222e-06,
+      "loss": 0.5987,
+      "step": 12192
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.694832545628266,
+      "learning_rate": 6.215467116796232e-06,
+      "loss": 0.5321,
+      "step": 12193
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 26.71760695616933,
+      "learning_rate": 6.213906205338145e-06,
+      "loss": 0.6175,
+      "step": 12194
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.29930276882442,
+      "learning_rate": 6.212345401554347e-06,
+      "loss": 0.493,
+      "step": 12195
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.941679654082067,
+      "learning_rate": 6.2107847054892325e-06,
+      "loss": 0.5197,
+      "step": 12196
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 29.9872709792075,
+      "learning_rate": 6.209224117187184e-06,
+      "loss": 0.5919,
+      "step": 12197
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 27.558368416452314,
+      "learning_rate": 6.207663636692581e-06,
+      "loss": 0.5752,
+      "step": 12198
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.595299440629072,
+      "learning_rate": 6.2061032640498095e-06,
+      "loss": 0.4722,
+      "step": 12199
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.196052407349097,
+      "learning_rate": 6.204542999303242e-06,
+      "loss": 0.5753,
+      "step": 12200
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.92598435466475,
+      "learning_rate": 6.202982842497247e-06,
+      "loss": 0.5,
+      "step": 12201
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.196931852423365,
+      "learning_rate": 6.201422793676203e-06,
+      "loss": 0.5354,
+      "step": 12202
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 25.322023037926495,
+      "learning_rate": 6.1998628528844685e-06,
+      "loss": 0.5301,
+      "step": 12203
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.072040389281618,
+      "learning_rate": 6.198303020166418e-06,
+      "loss": 0.4978,
+      "step": 12204
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.72833382236004,
+      "learning_rate": 6.196743295566401e-06,
+      "loss": 0.5288,
+      "step": 12205
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.988414108663815,
+      "learning_rate": 6.195183679128781e-06,
+      "loss": 0.5215,
+      "step": 12206
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.723997816160026,
+      "learning_rate": 6.19362417089791e-06,
+      "loss": 0.461,
+      "step": 12207
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.68155221802161,
+      "learning_rate": 6.1920647709181425e-06,
+      "loss": 0.5168,
+      "step": 12208
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.470433069431035,
+      "learning_rate": 6.190505479233826e-06,
+      "loss": 0.5784,
+      "step": 12209
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 31.764688419846152,
+      "learning_rate": 6.188946295889304e-06,
+      "loss": 0.5499,
+      "step": 12210
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 52.303380862688996,
+      "learning_rate": 6.18738722092892e-06,
+      "loss": 0.5165,
+      "step": 12211
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.527049362147796,
+      "learning_rate": 6.185828254397012e-06,
+      "loss": 0.5718,
+      "step": 12212
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 24.46383839094235,
+      "learning_rate": 6.184269396337916e-06,
+      "loss": 0.544,
+      "step": 12213
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.638061376217884,
+      "learning_rate": 6.182710646795972e-06,
+      "loss": 0.5732,
+      "step": 12214
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 28.396531618539328,
+      "learning_rate": 6.181152005815501e-06,
+      "loss": 0.6066,
+      "step": 12215
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 25.555496253753066,
+      "learning_rate": 6.179593473440833e-06,
+      "loss": 0.5093,
+      "step": 12216
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.767739276362914,
+      "learning_rate": 6.178035049716289e-06,
+      "loss": 0.5647,
+      "step": 12217
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.583684578647365,
+      "learning_rate": 6.1764767346861955e-06,
+      "loss": 0.5463,
+      "step": 12218
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 26.37654932045606,
+      "learning_rate": 6.174918528394868e-06,
+      "loss": 0.5759,
+      "step": 12219
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.013167751793873,
+      "learning_rate": 6.173360430886617e-06,
+      "loss": 0.4951,
+      "step": 12220
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 25.632850781812976,
+      "learning_rate": 6.171802442205759e-06,
+      "loss": 0.5838,
+      "step": 12221
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 29.75469137214884,
+      "learning_rate": 6.170244562396599e-06,
+      "loss": 0.6622,
+      "step": 12222
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.208161660740046,
+      "learning_rate": 6.1686867915034425e-06,
+      "loss": 0.5114,
+      "step": 12223
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.09749480076051,
+      "learning_rate": 6.167129129570595e-06,
+      "loss": 0.5446,
+      "step": 12224
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 21.50417245428113,
+      "learning_rate": 6.165571576642351e-06,
+      "loss": 0.5308,
+      "step": 12225
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.708743762687742,
+      "learning_rate": 6.164014132763008e-06,
+      "loss": 0.5796,
+      "step": 12226
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.435235427625951,
+      "learning_rate": 6.162456797976859e-06,
+      "loss": 0.4247,
+      "step": 12227
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.944371516691,
+      "learning_rate": 6.1608995723281925e-06,
+      "loss": 0.584,
+      "step": 12228
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.31472861152017,
+      "learning_rate": 6.159342455861304e-06,
+      "loss": 0.4976,
+      "step": 12229
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.092443165619823,
+      "learning_rate": 6.157785448620461e-06,
+      "loss": 0.5718,
+      "step": 12230
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.706088023786904,
+      "learning_rate": 6.156228550649954e-06,
+      "loss": 0.4975,
+      "step": 12231
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 25.66864942322433,
+      "learning_rate": 6.154671761994058e-06,
+      "loss": 0.6087,
+      "step": 12232
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.474472487283657,
+      "learning_rate": 6.153115082697052e-06,
+      "loss": 0.4825,
+      "step": 12233
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.818215031021587,
+      "learning_rate": 6.1515585128032e-06,
+      "loss": 0.5691,
+      "step": 12234
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.501704291506618,
+      "learning_rate": 6.150002052356772e-06,
+      "loss": 0.5322,
+      "step": 12235
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.194695851182095,
+      "learning_rate": 6.148445701402032e-06,
+      "loss": 0.5778,
+      "step": 12236
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.76790458525802,
+      "learning_rate": 6.1468894599832455e-06,
+      "loss": 0.5472,
+      "step": 12237
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.855332722142574,
+      "learning_rate": 6.145333328144669e-06,
+      "loss": 0.4362,
+      "step": 12238
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.30343730953803,
+      "learning_rate": 6.143777305930556e-06,
+      "loss": 0.5867,
+      "step": 12239
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.508747785039262,
+      "learning_rate": 6.142221393385161e-06,
+      "loss": 0.4782,
+      "step": 12240
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.681509216897357,
+      "learning_rate": 6.140665590552735e-06,
+      "loss": 0.5278,
+      "step": 12241
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 21.12231851381048,
+      "learning_rate": 6.139109897477517e-06,
+      "loss": 0.499,
+      "step": 12242
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.80724646913212,
+      "learning_rate": 6.137554314203762e-06,
+      "loss": 0.5027,
+      "step": 12243
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.74158231929338,
+      "learning_rate": 6.135998840775699e-06,
+      "loss": 0.4842,
+      "step": 12244
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.675780274552274,
+      "learning_rate": 6.134443477237566e-06,
+      "loss": 0.5263,
+      "step": 12245
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.52745672419481,
+      "learning_rate": 6.132888223633602e-06,
+      "loss": 0.5024,
+      "step": 12246
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.004724199356977,
+      "learning_rate": 6.131333080008033e-06,
+      "loss": 0.4343,
+      "step": 12247
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.733886475834016,
+      "learning_rate": 6.129778046405095e-06,
+      "loss": 0.5556,
+      "step": 12248
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.8758789978313,
+      "learning_rate": 6.128223122868997e-06,
+      "loss": 0.5325,
+      "step": 12249
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.666615292904387,
+      "learning_rate": 6.126668309443973e-06,
+      "loss": 0.4913,
+      "step": 12250
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.96252052545536,
+      "learning_rate": 6.125113606174233e-06,
+      "loss": 0.4449,
+      "step": 12251
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.053416924322395,
+      "learning_rate": 6.123559013103997e-06,
+      "loss": 0.5862,
+      "step": 12252
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.105968538219564,
+      "learning_rate": 6.122004530277476e-06,
+      "loss": 0.5546,
+      "step": 12253
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.602386607329148,
+      "learning_rate": 6.1204501577388775e-06,
+      "loss": 0.5671,
+      "step": 12254
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.400506416576217,
+      "learning_rate": 6.118895895532405e-06,
+      "loss": 0.4884,
+      "step": 12255
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.01123151614482,
+      "learning_rate": 6.117341743702265e-06,
+      "loss": 0.5161,
+      "step": 12256
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.794278583194735,
+      "learning_rate": 6.115787702292651e-06,
+      "loss": 0.5328,
+      "step": 12257
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.752328734634908,
+      "learning_rate": 6.1142337713477685e-06,
+      "loss": 0.4854,
+      "step": 12258
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.945047820567815,
+      "learning_rate": 6.112679950911802e-06,
+      "loss": 0.5748,
+      "step": 12259
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.582051918908004,
+      "learning_rate": 6.111126241028942e-06,
+      "loss": 0.4871,
+      "step": 12260
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.339243105559248,
+      "learning_rate": 6.1095726417433765e-06,
+      "loss": 0.5092,
+      "step": 12261
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.14320682434728,
+      "learning_rate": 6.10801915309929e-06,
+      "loss": 0.546,
+      "step": 12262
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.53945640434366,
+      "learning_rate": 6.106465775140865e-06,
+      "loss": 0.5227,
+      "step": 12263
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.576915372890655,
+      "learning_rate": 6.104912507912272e-06,
+      "loss": 0.4823,
+      "step": 12264
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.49053580678368,
+      "learning_rate": 6.1033593514576895e-06,
+      "loss": 0.5771,
+      "step": 12265
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.141406715618732,
+      "learning_rate": 6.101806305821288e-06,
+      "loss": 0.4874,
+      "step": 12266
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.96907863864128,
+      "learning_rate": 6.100253371047233e-06,
+      "loss": 0.5957,
+      "step": 12267
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.537269674362996,
+      "learning_rate": 6.098700547179695e-06,
+      "loss": 0.559,
+      "step": 12268
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.360425778300087,
+      "learning_rate": 6.0971478342628285e-06,
+      "loss": 0.5299,
+      "step": 12269
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.44405969940147,
+      "learning_rate": 6.095595232340794e-06,
+      "loss": 0.488,
+      "step": 12270
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.080720858805734,
+      "learning_rate": 6.094042741457749e-06,
+      "loss": 0.5644,
+      "step": 12271
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.63498438983016,
+      "learning_rate": 6.0924903616578394e-06,
+      "loss": 0.5554,
+      "step": 12272
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.989096452521217,
+      "learning_rate": 6.090938092985226e-06,
+      "loss": 0.546,
+      "step": 12273
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.255078179482236,
+      "learning_rate": 6.08938593548404e-06,
+      "loss": 0.4866,
+      "step": 12274
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.732335913330358,
+      "learning_rate": 6.087833889198431e-06,
+      "loss": 0.4742,
+      "step": 12275
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.263861475589888,
+      "learning_rate": 6.086281954172537e-06,
+      "loss": 0.5251,
+      "step": 12276
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.329257043351287,
+      "learning_rate": 6.084730130450497e-06,
+      "loss": 0.5346,
+      "step": 12277
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.871124746430304,
+      "learning_rate": 6.083178418076443e-06,
+      "loss": 0.5068,
+      "step": 12278
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.967703829454823,
+      "learning_rate": 6.081626817094502e-06,
+      "loss": 0.4768,
+      "step": 12279
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.833495254505166,
+      "learning_rate": 6.0800753275488e-06,
+      "loss": 0.5139,
+      "step": 12280
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.877622778332483,
+      "learning_rate": 6.078523949483465e-06,
+      "loss": 0.473,
+      "step": 12281
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.82979669122015,
+      "learning_rate": 6.076972682942617e-06,
+      "loss": 0.5988,
+      "step": 12282
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 35.828291760774825,
+      "learning_rate": 6.075421527970366e-06,
+      "loss": 0.5938,
+      "step": 12283
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.579769136467974,
+      "learning_rate": 6.073870484610833e-06,
+      "loss": 0.4486,
+      "step": 12284
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.436900194562686,
+      "learning_rate": 6.072319552908128e-06,
+      "loss": 0.5465,
+      "step": 12285
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.484018060529767,
+      "learning_rate": 6.070768732906353e-06,
+      "loss": 0.5934,
+      "step": 12286
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.157476642748698,
+      "learning_rate": 6.069218024649622e-06,
+      "loss": 0.526,
+      "step": 12287
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.04431796631102,
+      "learning_rate": 6.0676674281820294e-06,
+      "loss": 0.5257,
+      "step": 12288
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.817975569553894,
+      "learning_rate": 6.066116943547673e-06,
+      "loss": 0.5153,
+      "step": 12289
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 26.919243660167318,
+      "learning_rate": 6.064566570790651e-06,
+      "loss": 0.5411,
+      "step": 12290
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.007482900307902,
+      "learning_rate": 6.063016309955051e-06,
+      "loss": 0.5515,
+      "step": 12291
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.910521128157438,
+      "learning_rate": 6.06146616108497e-06,
+      "loss": 0.5549,
+      "step": 12292
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.922693304899678,
+      "learning_rate": 6.059916124224483e-06,
+      "loss": 0.5857,
+      "step": 12293
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.935425710856062,
+      "learning_rate": 6.058366199417676e-06,
+      "loss": 0.5129,
+      "step": 12294
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.906686051767723,
+      "learning_rate": 6.056816386708628e-06,
+      "loss": 0.5667,
+      "step": 12295
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.12557874118802,
+      "learning_rate": 6.055266686141418e-06,
+      "loss": 0.5148,
+      "step": 12296
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.554082193330338,
+      "learning_rate": 6.053717097760117e-06,
+      "loss": 0.6317,
+      "step": 12297
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.323846431926478,
+      "learning_rate": 6.052167621608791e-06,
+      "loss": 0.4545,
+      "step": 12298
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.044315923526742,
+      "learning_rate": 6.050618257731507e-06,
+      "loss": 0.588,
+      "step": 12299
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 21.87047303360174,
+      "learning_rate": 6.049069006172331e-06,
+      "loss": 0.6194,
+      "step": 12300
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.025605170167378,
+      "learning_rate": 6.04751986697532e-06,
+      "loss": 0.5724,
+      "step": 12301
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.72592136212466,
+      "learning_rate": 6.045970840184534e-06,
+      "loss": 0.561,
+      "step": 12302
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.776708591536986,
+      "learning_rate": 6.044421925844024e-06,
+      "loss": 0.577,
+      "step": 12303
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.23633041943759,
+      "learning_rate": 6.0428731239978415e-06,
+      "loss": 0.5055,
+      "step": 12304
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.852625788170199,
+      "learning_rate": 6.041324434690028e-06,
+      "loss": 0.5349,
+      "step": 12305
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.325067232427184,
+      "learning_rate": 6.0397758579646335e-06,
+      "loss": 0.5686,
+      "step": 12306
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.65158011481245,
+      "learning_rate": 6.038227393865701e-06,
+      "loss": 0.5007,
+      "step": 12307
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.732066002306635,
+      "learning_rate": 6.036679042437259e-06,
+      "loss": 0.5151,
+      "step": 12308
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.113880423875756,
+      "learning_rate": 6.035130803723349e-06,
+      "loss": 0.5191,
+      "step": 12309
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.109060157198147,
+      "learning_rate": 6.033582677767997e-06,
+      "loss": 0.4982,
+      "step": 12310
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.981712496775902,
+      "learning_rate": 6.032034664615236e-06,
+      "loss": 0.5105,
+      "step": 12311
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.46247899208498,
+      "learning_rate": 6.030486764309089e-06,
+      "loss": 0.4542,
+      "step": 12312
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.68847238236059,
+      "learning_rate": 6.028938976893575e-06,
+      "loss": 0.5514,
+      "step": 12313
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.300428745900966,
+      "learning_rate": 6.027391302412711e-06,
+      "loss": 0.4951,
+      "step": 12314
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 27.54649278545439,
+      "learning_rate": 6.025843740910518e-06,
+      "loss": 0.5783,
+      "step": 12315
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.536354303130112,
+      "learning_rate": 6.024296292431e-06,
+      "loss": 0.4886,
+      "step": 12316
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.83737265854583,
+      "learning_rate": 6.022748957018176e-06,
+      "loss": 0.5059,
+      "step": 12317
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.784168414085872,
+      "learning_rate": 6.02120173471604e-06,
+      "loss": 0.5209,
+      "step": 12318
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.342161685798253,
+      "learning_rate": 6.0196546255686016e-06,
+      "loss": 0.5904,
+      "step": 12319
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.253202693117608,
+      "learning_rate": 6.018107629619854e-06,
+      "loss": 0.5323,
+      "step": 12320
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.855083079835506,
+      "learning_rate": 6.0165607469137984e-06,
+      "loss": 0.5526,
+      "step": 12321
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.258805807173978,
+      "learning_rate": 6.015013977494429e-06,
+      "loss": 0.4956,
+      "step": 12322
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.880375996946327,
+      "learning_rate": 6.013467321405726e-06,
+      "loss": 0.5218,
+      "step": 12323
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.002634858758192,
+      "learning_rate": 6.011920778691679e-06,
+      "loss": 0.5165,
+      "step": 12324
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 23.227114741114562,
+      "learning_rate": 6.010374349396274e-06,
+      "loss": 0.558,
+      "step": 12325
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.40746386292164,
+      "learning_rate": 6.008828033563487e-06,
+      "loss": 0.5004,
+      "step": 12326
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.305761380857877,
+      "learning_rate": 6.0072818312373e-06,
+      "loss": 0.4808,
+      "step": 12327
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.371554597161667,
+      "learning_rate": 6.005735742461681e-06,
+      "loss": 0.4352,
+      "step": 12328
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.46274298321452,
+      "learning_rate": 6.004189767280599e-06,
+      "loss": 0.4739,
+      "step": 12329
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 20.270289648941908,
+      "learning_rate": 6.002643905738022e-06,
+      "loss": 0.5381,
+      "step": 12330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.718616033827995,
+      "learning_rate": 6.0010981578779175e-06,
+      "loss": 0.4453,
+      "step": 12331
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.35109225747623,
+      "learning_rate": 5.9995525237442395e-06,
+      "loss": 0.4429,
+      "step": 12332
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.473640577903073,
+      "learning_rate": 5.9980070033809455e-06,
+      "loss": 0.5189,
+      "step": 12333
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.72697366906619,
+      "learning_rate": 5.996461596831995e-06,
+      "loss": 0.5608,
+      "step": 12334
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.514171329426617,
+      "learning_rate": 5.9949163041413315e-06,
+      "loss": 0.4483,
+      "step": 12335
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.781371233548622,
+      "learning_rate": 5.9933711253529115e-06,
+      "loss": 0.5658,
+      "step": 12336
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 20.80831062263204,
+      "learning_rate": 5.991826060510666e-06,
+      "loss": 0.5643,
+      "step": 12337
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.85296637077086,
+      "learning_rate": 5.990281109658545e-06,
+      "loss": 0.5359,
+      "step": 12338
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 28.357611902460572,
+      "learning_rate": 5.988736272840483e-06,
+      "loss": 0.5674,
+      "step": 12339
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 20.35844605610881,
+      "learning_rate": 5.987191550100415e-06,
+      "loss": 0.5319,
+      "step": 12340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 22.82902687218169,
+      "learning_rate": 5.985646941482275e-06,
+      "loss": 0.5442,
+      "step": 12341
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.13460488522913,
+      "learning_rate": 5.984102447029986e-06,
+      "loss": 0.5105,
+      "step": 12342
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.501221263792665,
+      "learning_rate": 5.9825580667874715e-06,
+      "loss": 0.4686,
+      "step": 12343
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.02378032069939,
+      "learning_rate": 5.981013800798657e-06,
+      "loss": 0.4675,
+      "step": 12344
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.986058449378904,
+      "learning_rate": 5.9794696491074576e-06,
+      "loss": 0.4835,
+      "step": 12345
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.695985333615948,
+      "learning_rate": 5.977925611757794e-06,
+      "loss": 0.4871,
+      "step": 12346
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.67034463970409,
+      "learning_rate": 5.976381688793571e-06,
+      "loss": 0.5343,
+      "step": 12347
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.512696596220714,
+      "learning_rate": 5.974837880258697e-06,
+      "loss": 0.487,
+      "step": 12348
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.34622884831244,
+      "learning_rate": 5.973294186197079e-06,
+      "loss": 0.6084,
+      "step": 12349
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.68531712609804,
+      "learning_rate": 5.971750606652621e-06,
+      "loss": 0.5363,
+      "step": 12350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 20.314529720279726,
+      "learning_rate": 5.97020714166922e-06,
+      "loss": 0.5161,
+      "step": 12351
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.415219791041318,
+      "learning_rate": 5.9686637912907675e-06,
+      "loss": 0.5051,
+      "step": 12352
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.814206456863005,
+      "learning_rate": 5.96712055556116e-06,
+      "loss": 0.545,
+      "step": 12353
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.706709550619333,
+      "learning_rate": 5.965577434524283e-06,
+      "loss": 0.4624,
+      "step": 12354
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.581329845397073,
+      "learning_rate": 5.964034428224026e-06,
+      "loss": 0.5049,
+      "step": 12355
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.117178325638507,
+      "learning_rate": 5.9624915367042705e-06,
+      "loss": 0.492,
+      "step": 12356
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.125009604859244,
+      "learning_rate": 5.960948760008891e-06,
+      "loss": 0.5081,
+      "step": 12357
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 23.20827378614795,
+      "learning_rate": 5.959406098181766e-06,
+      "loss": 0.5871,
+      "step": 12358
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.45444301351702,
+      "learning_rate": 5.9578635512667686e-06,
+      "loss": 0.5759,
+      "step": 12359
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 27.47755231964424,
+      "learning_rate": 5.9563211193077644e-06,
+      "loss": 0.5985,
+      "step": 12360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.32948695242102,
+      "learning_rate": 5.9547788023486296e-06,
+      "loss": 0.5493,
+      "step": 12361
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 28.60447621319841,
+      "learning_rate": 5.953236600433212e-06,
+      "loss": 0.6283,
+      "step": 12362
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 22.70753012069293,
+      "learning_rate": 5.951694513605381e-06,
+      "loss": 0.541,
+      "step": 12363
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.780237856413557,
+      "learning_rate": 5.950152541908988e-06,
+      "loss": 0.4784,
+      "step": 12364
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.073217379480376,
+      "learning_rate": 5.94861068538789e-06,
+      "loss": 0.4856,
+      "step": 12365
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.70224930033288,
+      "learning_rate": 5.947068944085934e-06,
+      "loss": 0.5123,
+      "step": 12366
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.355420320444683,
+      "learning_rate": 5.9455273180469666e-06,
+      "loss": 0.5719,
+      "step": 12367
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.02773195553403,
+      "learning_rate": 5.943985807314827e-06,
+      "loss": 0.4973,
+      "step": 12368
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.900555744632838,
+      "learning_rate": 5.942444411933361e-06,
+      "loss": 0.5651,
+      "step": 12369
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.365440457697158,
+      "learning_rate": 5.940903131946401e-06,
+      "loss": 0.5803,
+      "step": 12370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.917058685633137,
+      "learning_rate": 5.939361967397785e-06,
+      "loss": 0.6036,
+      "step": 12371
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.248547168624674,
+      "learning_rate": 5.937820918331336e-06,
+      "loss": 0.458,
+      "step": 12372
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.06914817004524,
+      "learning_rate": 5.9362799847908845e-06,
+      "loss": 0.4907,
+      "step": 12373
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 23.20746151830743,
+      "learning_rate": 5.934739166820252e-06,
+      "loss": 0.6095,
+      "step": 12374
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.4524110427762,
+      "learning_rate": 5.9331984644632615e-06,
+      "loss": 0.5806,
+      "step": 12375
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 26.480536295771884,
+      "learning_rate": 5.931657877763728e-06,
+      "loss": 0.4927,
+      "step": 12376
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 22.82148895182661,
+      "learning_rate": 5.930117406765462e-06,
+      "loss": 0.6186,
+      "step": 12377
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.16527062831763,
+      "learning_rate": 5.928577051512277e-06,
+      "loss": 0.5995,
+      "step": 12378
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.246434517017754,
+      "learning_rate": 5.927036812047978e-06,
+      "loss": 0.5412,
+      "step": 12379
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 22.70932936674966,
+      "learning_rate": 5.925496688416374e-06,
+      "loss": 0.5898,
+      "step": 12380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.30599333081173,
+      "learning_rate": 5.923956680661257e-06,
+      "loss": 0.5145,
+      "step": 12381
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 26.839283217938156,
+      "learning_rate": 5.922416788826429e-06,
+      "loss": 0.5827,
+      "step": 12382
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.621006501287976,
+      "learning_rate": 5.920877012955679e-06,
+      "loss": 0.5624,
+      "step": 12383
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.77983446013133,
+      "learning_rate": 5.919337353092802e-06,
+      "loss": 0.5661,
+      "step": 12384
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.844974066521289,
+      "learning_rate": 5.917797809281587e-06,
+      "loss": 0.4736,
+      "step": 12385
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.97412940604854,
+      "learning_rate": 5.916258381565811e-06,
+      "loss": 0.4732,
+      "step": 12386
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.183570081700054,
+      "learning_rate": 5.914719069989257e-06,
+      "loss": 0.4821,
+      "step": 12387
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.848445057527147,
+      "learning_rate": 5.913179874595705e-06,
+      "loss": 0.542,
+      "step": 12388
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.3524597269773,
+      "learning_rate": 5.911640795428923e-06,
+      "loss": 0.5594,
+      "step": 12389
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 25.311570329430268,
+      "learning_rate": 5.910101832532691e-06,
+      "loss": 0.5641,
+      "step": 12390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.574542657705813,
+      "learning_rate": 5.908562985950768e-06,
+      "loss": 0.5433,
+      "step": 12391
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.314135868731526,
+      "learning_rate": 5.907024255726919e-06,
+      "loss": 0.5528,
+      "step": 12392
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.87727639547193,
+      "learning_rate": 5.905485641904904e-06,
+      "loss": 0.531,
+      "step": 12393
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.0783898903737,
+      "learning_rate": 5.903947144528486e-06,
+      "loss": 0.5324,
+      "step": 12394
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 24.25238748815649,
+      "learning_rate": 5.902408763641416e-06,
+      "loss": 0.5721,
+      "step": 12395
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.923012906656197,
+      "learning_rate": 5.9008704992874384e-06,
+      "loss": 0.5137,
+      "step": 12396
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.70871145537197,
+      "learning_rate": 5.89933235151031e-06,
+      "loss": 0.4857,
+      "step": 12397
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.25025273029509,
+      "learning_rate": 5.897794320353768e-06,
+      "loss": 0.5418,
+      "step": 12398
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.853369733894144,
+      "learning_rate": 5.896256405861558e-06,
+      "loss": 0.5445,
+      "step": 12399
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.20968496516432,
+      "learning_rate": 5.8947186080774165e-06,
+      "loss": 0.4262,
+      "step": 12400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.339365901195418,
+      "learning_rate": 5.8931809270450746e-06,
+      "loss": 0.4989,
+      "step": 12401
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.328759201097878,
+      "learning_rate": 5.8916433628082626e-06,
+      "loss": 0.5199,
+      "step": 12402
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.190822115815195,
+      "learning_rate": 5.890105915410712e-06,
+      "loss": 0.5436,
+      "step": 12403
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.322672321278281,
+      "learning_rate": 5.8885685848961424e-06,
+      "loss": 0.4906,
+      "step": 12404
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.316512197750402,
+      "learning_rate": 5.8870313713082845e-06,
+      "loss": 0.4905,
+      "step": 12405
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 24.476398468076734,
+      "learning_rate": 5.885494274690842e-06,
+      "loss": 0.5817,
+      "step": 12406
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.981877596757233,
+      "learning_rate": 5.883957295087537e-06,
+      "loss": 0.5369,
+      "step": 12407
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.790324660696253,
+      "learning_rate": 5.882420432542077e-06,
+      "loss": 0.6368,
+      "step": 12408
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.776741641622746,
+      "learning_rate": 5.8808836870981725e-06,
+      "loss": 0.5382,
+      "step": 12409
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 28.8304841045478,
+      "learning_rate": 5.87934705879953e-06,
+      "loss": 0.4809,
+      "step": 12410
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.49730136397342,
+      "learning_rate": 5.8778105476898415e-06,
+      "loss": 0.569,
+      "step": 12411
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 24.831637009879625,
+      "learning_rate": 5.87627415381281e-06,
+      "loss": 0.5326,
+      "step": 12412
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.52488268034948,
+      "learning_rate": 5.874737877212131e-06,
+      "loss": 0.5532,
+      "step": 12413
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.600109363752566,
+      "learning_rate": 5.873201717931492e-06,
+      "loss": 0.4851,
+      "step": 12414
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.071249126516147,
+      "learning_rate": 5.871665676014584e-06,
+      "loss": 0.4928,
+      "step": 12415
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.959499104560754,
+      "learning_rate": 5.870129751505089e-06,
+      "loss": 0.5253,
+      "step": 12416
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.189228694467246,
+      "learning_rate": 5.868593944446688e-06,
+      "loss": 0.5779,
+      "step": 12417
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.355434399466912,
+      "learning_rate": 5.867058254883056e-06,
+      "loss": 0.5513,
+      "step": 12418
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.660743889390798,
+      "learning_rate": 5.8655226828578715e-06,
+      "loss": 0.4516,
+      "step": 12419
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.286007793706315,
+      "learning_rate": 5.863987228414805e-06,
+      "loss": 0.5183,
+      "step": 12420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.253188452800565,
+      "learning_rate": 5.8624518915975185e-06,
+      "loss": 0.5251,
+      "step": 12421
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.41976905119453,
+      "learning_rate": 5.860916672449682e-06,
+      "loss": 0.5562,
+      "step": 12422
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.102702458140445,
+      "learning_rate": 5.859381571014951e-06,
+      "loss": 0.5111,
+      "step": 12423
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 26.352578266560638,
+      "learning_rate": 5.857846587336989e-06,
+      "loss": 0.5896,
+      "step": 12424
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.961129258199225,
+      "learning_rate": 5.856311721459449e-06,
+      "loss": 0.5243,
+      "step": 12425
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.310855484682733,
+      "learning_rate": 5.854776973425978e-06,
+      "loss": 0.6174,
+      "step": 12426
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.128513415428728,
+      "learning_rate": 5.853242343280222e-06,
+      "loss": 0.4883,
+      "step": 12427
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.5058426664811,
+      "learning_rate": 5.85170783106583e-06,
+      "loss": 0.5232,
+      "step": 12428
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.839485978626312,
+      "learning_rate": 5.8501734368264425e-06,
+      "loss": 0.5185,
+      "step": 12429
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.23613590313458,
+      "learning_rate": 5.848639160605694e-06,
+      "loss": 0.5017,
+      "step": 12430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.551606047043208,
+      "learning_rate": 5.847105002447218e-06,
+      "loss": 0.5182,
+      "step": 12431
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.34640627229531,
+      "learning_rate": 5.845570962394647e-06,
+      "loss": 0.569,
+      "step": 12432
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.691517026642444,
+      "learning_rate": 5.8440370404916035e-06,
+      "loss": 0.4934,
+      "step": 12433
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.552155417085057,
+      "learning_rate": 5.842503236781722e-06,
+      "loss": 0.4917,
+      "step": 12434
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.389523060572717,
+      "learning_rate": 5.840969551308614e-06,
+      "loss": 0.4808,
+      "step": 12435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.692684038938735,
+      "learning_rate": 5.839435984115899e-06,
+      "loss": 0.5329,
+      "step": 12436
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 24.00601334414222,
+      "learning_rate": 5.8379025352471905e-06,
+      "loss": 0.5689,
+      "step": 12437
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.812143364337903,
+      "learning_rate": 5.836369204746097e-06,
+      "loss": 0.5513,
+      "step": 12438
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.104422500629422,
+      "learning_rate": 5.834835992656232e-06,
+      "loss": 0.5193,
+      "step": 12439
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.40942907408857,
+      "learning_rate": 5.833302899021191e-06,
+      "loss": 0.5167,
+      "step": 12440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.757174954111548,
+      "learning_rate": 5.831769923884579e-06,
+      "loss": 0.5556,
+      "step": 12441
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.52624071654354,
+      "learning_rate": 5.830237067289993e-06,
+      "loss": 0.51,
+      "step": 12442
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.920302141397592,
+      "learning_rate": 5.828704329281024e-06,
+      "loss": 0.5263,
+      "step": 12443
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.55781347183722,
+      "learning_rate": 5.827171709901267e-06,
+      "loss": 0.5711,
+      "step": 12444
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 28.532334896719814,
+      "learning_rate": 5.825639209194302e-06,
+      "loss": 0.5597,
+      "step": 12445
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.352330935389503,
+      "learning_rate": 5.824106827203719e-06,
+      "loss": 0.5696,
+      "step": 12446
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.64854656392188,
+      "learning_rate": 5.822574563973091e-06,
+      "loss": 0.5965,
+      "step": 12447
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.898015455986016,
+      "learning_rate": 5.8210424195460005e-06,
+      "loss": 0.5788,
+      "step": 12448
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.882164503947042,
+      "learning_rate": 5.8195103939660214e-06,
+      "loss": 0.4394,
+      "step": 12449
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.478105566625683,
+      "learning_rate": 5.817978487276722e-06,
+      "loss": 0.4859,
+      "step": 12450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.224604271771039,
+      "learning_rate": 5.816446699521663e-06,
+      "loss": 0.5257,
+      "step": 12451
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 24.33418179052358,
+      "learning_rate": 5.814915030744414e-06,
+      "loss": 0.4796,
+      "step": 12452
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.10577788189234,
+      "learning_rate": 5.813383480988533e-06,
+      "loss": 0.4948,
+      "step": 12453
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.879545363926923,
+      "learning_rate": 5.811852050297579e-06,
+      "loss": 0.5374,
+      "step": 12454
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.97696263006178,
+      "learning_rate": 5.810320738715104e-06,
+      "loss": 0.506,
+      "step": 12455
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.110892698703324,
+      "learning_rate": 5.808789546284652e-06,
+      "loss": 0.476,
+      "step": 12456
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 24.43157343451516,
+      "learning_rate": 5.807258473049774e-06,
+      "loss": 0.5584,
+      "step": 12457
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 26.064357158531386,
+      "learning_rate": 5.805727519054017e-06,
+      "loss": 0.5798,
+      "step": 12458
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.64100087892689,
+      "learning_rate": 5.8041966843409135e-06,
+      "loss": 0.4604,
+      "step": 12459
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.739041076827338,
+      "learning_rate": 5.802665968954e-06,
+      "loss": 0.5202,
+      "step": 12460
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 24.643162402588274,
+      "learning_rate": 5.801135372936809e-06,
+      "loss": 0.5942,
+      "step": 12461
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.75121216164051,
+      "learning_rate": 5.7996048963328775e-06,
+      "loss": 0.4859,
+      "step": 12462
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.670004865447886,
+      "learning_rate": 5.798074539185721e-06,
+      "loss": 0.5276,
+      "step": 12463
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.13919389795665,
+      "learning_rate": 5.79654430153887e-06,
+      "loss": 0.493,
+      "step": 12464
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.92708306068029,
+      "learning_rate": 5.7950141834358365e-06,
+      "loss": 0.5175,
+      "step": 12465
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.451526262504714,
+      "learning_rate": 5.793484184920139e-06,
+      "loss": 0.5432,
+      "step": 12466
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.563684716292574,
+      "learning_rate": 5.7919543060352965e-06,
+      "loss": 0.5377,
+      "step": 12467
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 24.197960593496667,
+      "learning_rate": 5.790424546824806e-06,
+      "loss": 0.476,
+      "step": 12468
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.942537784489552,
+      "learning_rate": 5.788894907332184e-06,
+      "loss": 0.4994,
+      "step": 12469
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.053730136514798,
+      "learning_rate": 5.787365387600923e-06,
+      "loss": 0.4765,
+      "step": 12470
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 27.214363119087164,
+      "learning_rate": 5.78583598767453e-06,
+      "loss": 0.5595,
+      "step": 12471
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.915624174855418,
+      "learning_rate": 5.784306707596492e-06,
+      "loss": 0.5166,
+      "step": 12472
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.060402106052805,
+      "learning_rate": 5.782777547410305e-06,
+      "loss": 0.5342,
+      "step": 12473
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.914268547678464,
+      "learning_rate": 5.781248507159463e-06,
+      "loss": 0.5307,
+      "step": 12474
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.899785463421146,
+      "learning_rate": 5.7797195868874445e-06,
+      "loss": 0.5482,
+      "step": 12475
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.001721067835165,
+      "learning_rate": 5.778190786637729e-06,
+      "loss": 0.5411,
+      "step": 12476
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.444153370730817,
+      "learning_rate": 5.776662106453797e-06,
+      "loss": 0.4678,
+      "step": 12477
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.353632849090427,
+      "learning_rate": 5.775133546379128e-06,
+      "loss": 0.5077,
+      "step": 12478
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.834760320730712,
+      "learning_rate": 5.773605106457185e-06,
+      "loss": 0.4954,
+      "step": 12479
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.308350709486163,
+      "learning_rate": 5.7720767867314464e-06,
+      "loss": 0.5926,
+      "step": 12480
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.609722802013273,
+      "learning_rate": 5.7705485872453645e-06,
+      "loss": 0.5273,
+      "step": 12481
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.27119436783684,
+      "learning_rate": 5.769020508042408e-06,
+      "loss": 0.5826,
+      "step": 12482
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.239027591810153,
+      "learning_rate": 5.7674925491660365e-06,
+      "loss": 0.5343,
+      "step": 12483
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 22.908663200060055,
+      "learning_rate": 5.765964710659702e-06,
+      "loss": 0.5001,
+      "step": 12484
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.89939788976453,
+      "learning_rate": 5.764436992566849e-06,
+      "loss": 0.5365,
+      "step": 12485
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.136950492813643,
+      "learning_rate": 5.762909394930931e-06,
+      "loss": 0.4934,
+      "step": 12486
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.235445422265318,
+      "learning_rate": 5.761381917795394e-06,
+      "loss": 0.6292,
+      "step": 12487
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.285681019017932,
+      "learning_rate": 5.759854561203677e-06,
+      "loss": 0.5552,
+      "step": 12488
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.197960103509278,
+      "learning_rate": 5.7583273251992115e-06,
+      "loss": 0.4705,
+      "step": 12489
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.631962528254324,
+      "learning_rate": 5.756800209825435e-06,
+      "loss": 0.5682,
+      "step": 12490
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.57362087103549,
+      "learning_rate": 5.75527321512578e-06,
+      "loss": 0.5323,
+      "step": 12491
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.030932422969112,
+      "learning_rate": 5.753746341143674e-06,
+      "loss": 0.5403,
+      "step": 12492
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.337260179503378,
+      "learning_rate": 5.752219587922538e-06,
+      "loss": 0.5667,
+      "step": 12493
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.85073213202555,
+      "learning_rate": 5.75069295550579e-06,
+      "loss": 0.5049,
+      "step": 12494
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.73827537788363,
+      "learning_rate": 5.749166443936847e-06,
+      "loss": 0.5169,
+      "step": 12495
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 28.790778922850944,
+      "learning_rate": 5.747640053259127e-06,
+      "loss": 0.5765,
+      "step": 12496
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.127927161419743,
+      "learning_rate": 5.746113783516034e-06,
+      "loss": 0.5171,
+      "step": 12497
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.800661820548402,
+      "learning_rate": 5.74458763475098e-06,
+      "loss": 0.4633,
+      "step": 12498
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.60879236975523,
+      "learning_rate": 5.743061607007359e-06,
+      "loss": 0.4808,
+      "step": 12499
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.959288420981103,
+      "learning_rate": 5.741535700328581e-06,
+      "loss": 0.5148,
+      "step": 12500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.711860192404695,
+      "learning_rate": 5.740009914758032e-06,
+      "loss": 0.5283,
+      "step": 12501
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 26.531583622159197,
+      "learning_rate": 5.738484250339109e-06,
+      "loss": 0.5762,
+      "step": 12502
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.132486755725896,
+      "learning_rate": 5.7369587071152055e-06,
+      "loss": 0.4862,
+      "step": 12503
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 23.722684751858132,
+      "learning_rate": 5.735433285129699e-06,
+      "loss": 0.5213,
+      "step": 12504
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 28.15577624268529,
+      "learning_rate": 5.733907984425979e-06,
+      "loss": 0.5755,
+      "step": 12505
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.741721753286174,
+      "learning_rate": 5.732382805047416e-06,
+      "loss": 0.5924,
+      "step": 12506
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 33.274203655077116,
+      "learning_rate": 5.730857747037389e-06,
+      "loss": 0.5293,
+      "step": 12507
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.296892509717193,
+      "learning_rate": 5.729332810439274e-06,
+      "loss": 0.4954,
+      "step": 12508
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.053697510222843,
+      "learning_rate": 5.727807995296437e-06,
+      "loss": 0.5509,
+      "step": 12509
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.581892688055213,
+      "learning_rate": 5.7262833016522366e-06,
+      "loss": 0.4933,
+      "step": 12510
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.662274235466818,
+      "learning_rate": 5.72475872955004e-06,
+      "loss": 0.4975,
+      "step": 12511
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.684752196093253,
+      "learning_rate": 5.723234279033207e-06,
+      "loss": 0.5613,
+      "step": 12512
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.316687034363607,
+      "learning_rate": 5.721709950145089e-06,
+      "loss": 0.4734,
+      "step": 12513
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.93026752592397,
+      "learning_rate": 5.720185742929034e-06,
+      "loss": 0.4768,
+      "step": 12514
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.1710369393287,
+      "learning_rate": 5.718661657428393e-06,
+      "loss": 0.568,
+      "step": 12515
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 23.876264666595794,
+      "learning_rate": 5.717137693686509e-06,
+      "loss": 0.557,
+      "step": 12516
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.55901711471774,
+      "learning_rate": 5.7156138517467285e-06,
+      "loss": 0.5792,
+      "step": 12517
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.886133900409952,
+      "learning_rate": 5.714090131652385e-06,
+      "loss": 0.4748,
+      "step": 12518
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.304057770134662,
+      "learning_rate": 5.712566533446805e-06,
+      "loss": 0.5342,
+      "step": 12519
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 29.1376990048434,
+      "learning_rate": 5.711043057173326e-06,
+      "loss": 0.5208,
+      "step": 12520
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.883859511732858,
+      "learning_rate": 5.709519702875277e-06,
+      "loss": 0.5307,
+      "step": 12521
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.384569057745946,
+      "learning_rate": 5.707996470595977e-06,
+      "loss": 0.5088,
+      "step": 12522
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 23.603573246586848,
+      "learning_rate": 5.706473360378743e-06,
+      "loss": 0.5319,
+      "step": 12523
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.797908347876046,
+      "learning_rate": 5.704950372266895e-06,
+      "loss": 0.5121,
+      "step": 12524
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.752522660479476,
+      "learning_rate": 5.70342750630375e-06,
+      "loss": 0.5141,
+      "step": 12525
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.492231915303353,
+      "learning_rate": 5.70190476253261e-06,
+      "loss": 0.5276,
+      "step": 12526
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.262984210376192,
+      "learning_rate": 5.700382140996787e-06,
+      "loss": 0.5458,
+      "step": 12527
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.646237699000984,
+      "learning_rate": 5.698859641739578e-06,
+      "loss": 0.5303,
+      "step": 12528
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.33721121600938,
+      "learning_rate": 5.697337264804283e-06,
+      "loss": 0.5405,
+      "step": 12529
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.62713172184386,
+      "learning_rate": 5.695815010234204e-06,
+      "loss": 0.5091,
+      "step": 12530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 15.779170723887841,
+      "learning_rate": 5.694292878072625e-06,
+      "loss": 0.5264,
+      "step": 12531
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.651083869925394,
+      "learning_rate": 5.6927708683628415e-06,
+      "loss": 0.4928,
+      "step": 12532
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.997205040947005,
+      "learning_rate": 5.6912489811481295e-06,
+      "loss": 0.5407,
+      "step": 12533
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.428851128556964,
+      "learning_rate": 5.689727216471782e-06,
+      "loss": 0.5003,
+      "step": 12534
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 27.057714057259094,
+      "learning_rate": 5.688205574377066e-06,
+      "loss": 0.5358,
+      "step": 12535
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.40300789085051,
+      "learning_rate": 5.686684054907261e-06,
+      "loss": 0.5165,
+      "step": 12536
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.309578242343726,
+      "learning_rate": 5.685162658105643e-06,
+      "loss": 0.5323,
+      "step": 12537
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.518517897377475,
+      "learning_rate": 5.683641384015475e-06,
+      "loss": 0.5376,
+      "step": 12538
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.366438832991074,
+      "learning_rate": 5.682120232680015e-06,
+      "loss": 0.4875,
+      "step": 12539
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.21755931963509,
+      "learning_rate": 5.6805992041425315e-06,
+      "loss": 0.4978,
+      "step": 12540
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.553028789026607,
+      "learning_rate": 5.679078298446279e-06,
+      "loss": 0.4798,
+      "step": 12541
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.120778420877425,
+      "learning_rate": 5.677557515634517e-06,
+      "loss": 0.5202,
+      "step": 12542
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.682213563651375,
+      "learning_rate": 5.67603685575049e-06,
+      "loss": 0.498,
+      "step": 12543
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.76760529484355,
+      "learning_rate": 5.674516318837442e-06,
+      "loss": 0.5423,
+      "step": 12544
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 26.01898118499872,
+      "learning_rate": 5.6729959049386185e-06,
+      "loss": 0.4723,
+      "step": 12545
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.906333548665486,
+      "learning_rate": 5.6714756140972645e-06,
+      "loss": 0.5309,
+      "step": 12546
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 23.24678918499856,
+      "learning_rate": 5.669955446356612e-06,
+      "loss": 0.5461,
+      "step": 12547
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.26010737006464,
+      "learning_rate": 5.668435401759891e-06,
+      "loss": 0.5405,
+      "step": 12548
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.728930170174033,
+      "learning_rate": 5.666915480350332e-06,
+      "loss": 0.4694,
+      "step": 12549
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.655714290122393,
+      "learning_rate": 5.665395682171166e-06,
+      "loss": 0.56,
+      "step": 12550
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 28.772033980332132,
+      "learning_rate": 5.6638760072656075e-06,
+      "loss": 0.6019,
+      "step": 12551
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.37753722750831,
+      "learning_rate": 5.662356455676882e-06,
+      "loss": 0.5345,
+      "step": 12552
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.169553315475262,
+      "learning_rate": 5.660837027448198e-06,
+      "loss": 0.5226,
+      "step": 12553
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.76049640319283,
+      "learning_rate": 5.65931772262277e-06,
+      "loss": 0.5174,
+      "step": 12554
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.494866472412605,
+      "learning_rate": 5.657798541243812e-06,
+      "loss": 0.5609,
+      "step": 12555
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.063626158757213,
+      "learning_rate": 5.656279483354519e-06,
+      "loss": 0.5339,
+      "step": 12556
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.754580984163205,
+      "learning_rate": 5.6547605489981e-06,
+      "loss": 0.5504,
+      "step": 12557
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 27.982667719669724,
+      "learning_rate": 5.653241738217745e-06,
+      "loss": 0.5279,
+      "step": 12558
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.728815621649602,
+      "learning_rate": 5.651723051056657e-06,
+      "loss": 0.4881,
+      "step": 12559
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 23.595384337175233,
+      "learning_rate": 5.650204487558016e-06,
+      "loss": 0.5373,
+      "step": 12560
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 30.876869151838356,
+      "learning_rate": 5.648686047765017e-06,
+      "loss": 0.5955,
+      "step": 12561
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.985678155480695,
+      "learning_rate": 5.647167731720844e-06,
+      "loss": 0.5058,
+      "step": 12562
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.238870857978693,
+      "learning_rate": 5.645649539468675e-06,
+      "loss": 0.5549,
+      "step": 12563
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.525420028330117,
+      "learning_rate": 5.644131471051681e-06,
+      "loss": 0.4633,
+      "step": 12564
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.56415096756817,
+      "learning_rate": 5.642613526513041e-06,
+      "loss": 0.4476,
+      "step": 12565
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.182992717464638,
+      "learning_rate": 5.641095705895923e-06,
+      "loss": 0.4575,
+      "step": 12566
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.488173297783145,
+      "learning_rate": 5.639578009243496e-06,
+      "loss": 0.5252,
+      "step": 12567
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 24.166150841195538,
+      "learning_rate": 5.638060436598922e-06,
+      "loss": 0.5236,
+      "step": 12568
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.61527252302159,
+      "learning_rate": 5.636542988005351e-06,
+      "loss": 0.5729,
+      "step": 12569
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.638191243712967,
+      "learning_rate": 5.635025663505946e-06,
+      "loss": 0.5291,
+      "step": 12570
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.562467559199376,
+      "learning_rate": 5.633508463143862e-06,
+      "loss": 0.6122,
+      "step": 12571
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.02950975459572,
+      "learning_rate": 5.631991386962243e-06,
+      "loss": 0.5858,
+      "step": 12572
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.289733018226038,
+      "learning_rate": 5.6304744350042295e-06,
+      "loss": 0.6044,
+      "step": 12573
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.089825940804054,
+      "learning_rate": 5.628957607312967e-06,
+      "loss": 0.501,
+      "step": 12574
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.87329719303149,
+      "learning_rate": 5.627440903931598e-06,
+      "loss": 0.5379,
+      "step": 12575
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.764890367891415,
+      "learning_rate": 5.62592432490325e-06,
+      "loss": 0.5123,
+      "step": 12576
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.741319689084367,
+      "learning_rate": 5.624407870271052e-06,
+      "loss": 0.496,
+      "step": 12577
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.34575921924254,
+      "learning_rate": 5.622891540078135e-06,
+      "loss": 0.5614,
+      "step": 12578
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.20204664593074,
+      "learning_rate": 5.621375334367622e-06,
+      "loss": 0.5523,
+      "step": 12579
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.332139501443475,
+      "learning_rate": 5.619859253182638e-06,
+      "loss": 0.4856,
+      "step": 12580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.622692533943,
+      "learning_rate": 5.618343296566293e-06,
+      "loss": 0.4649,
+      "step": 12581
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.584562452381421,
+      "learning_rate": 5.616827464561698e-06,
+      "loss": 0.5671,
+      "step": 12582
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.894356240485116,
+      "learning_rate": 5.615311757211965e-06,
+      "loss": 0.468,
+      "step": 12583
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.573504952153666,
+      "learning_rate": 5.613796174560207e-06,
+      "loss": 0.4361,
+      "step": 12584
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.565973201874296,
+      "learning_rate": 5.612280716649514e-06,
+      "loss": 0.5162,
+      "step": 12585
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.376518821859634,
+      "learning_rate": 5.6107653835229954e-06,
+      "loss": 0.554,
+      "step": 12586
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.229109762060997,
+      "learning_rate": 5.609250175223737e-06,
+      "loss": 0.4645,
+      "step": 12587
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.404187656848006,
+      "learning_rate": 5.607735091794839e-06,
+      "loss": 0.5476,
+      "step": 12588
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.088113569729536,
+      "learning_rate": 5.606220133279383e-06,
+      "loss": 0.512,
+      "step": 12589
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.489813118036757,
+      "learning_rate": 5.604705299720455e-06,
+      "loss": 0.4846,
+      "step": 12590
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.402525817803287,
+      "learning_rate": 5.603190591161141e-06,
+      "loss": 0.5612,
+      "step": 12591
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.287011542456632,
+      "learning_rate": 5.601676007644511e-06,
+      "loss": 0.5346,
+      "step": 12592
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 25.26303079305595,
+      "learning_rate": 5.600161549213647e-06,
+      "loss": 0.6273,
+      "step": 12593
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.737577954910897,
+      "learning_rate": 5.598647215911609e-06,
+      "loss": 0.4365,
+      "step": 12594
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.259214475553833,
+      "learning_rate": 5.59713300778147e-06,
+      "loss": 0.536,
+      "step": 12595
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.789865134882238,
+      "learning_rate": 5.595618924866298e-06,
+      "loss": 0.5557,
+      "step": 12596
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.925207838876528,
+      "learning_rate": 5.594104967209146e-06,
+      "loss": 0.5139,
+      "step": 12597
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.160048251129773,
+      "learning_rate": 5.592591134853067e-06,
+      "loss": 0.4601,
+      "step": 12598
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 33.102190886856214,
+      "learning_rate": 5.591077427841118e-06,
+      "loss": 0.5281,
+      "step": 12599
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 32.24142014145156,
+      "learning_rate": 5.5895638462163536e-06,
+      "loss": 0.5306,
+      "step": 12600
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.513560001143286,
+      "learning_rate": 5.588050390021812e-06,
+      "loss": 0.5155,
+      "step": 12601
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.357463083037707,
+      "learning_rate": 5.586537059300532e-06,
+      "loss": 0.5161,
+      "step": 12602
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.789975246355397,
+      "learning_rate": 5.585023854095557e-06,
+      "loss": 0.5413,
+      "step": 12603
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.613335914478983,
+      "learning_rate": 5.5835107744499205e-06,
+      "loss": 0.4683,
+      "step": 12604
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 24.80942499891135,
+      "learning_rate": 5.581997820406659e-06,
+      "loss": 0.5057,
+      "step": 12605
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.141944585903207,
+      "learning_rate": 5.580484992008795e-06,
+      "loss": 0.5648,
+      "step": 12606
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 25.615456157019434,
+      "learning_rate": 5.5789722892993466e-06,
+      "loss": 0.6131,
+      "step": 12607
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 36.313936375821235,
+      "learning_rate": 5.577459712321341e-06,
+      "loss": 0.5668,
+      "step": 12608
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.30870921782114,
+      "learning_rate": 5.575947261117798e-06,
+      "loss": 0.5608,
+      "step": 12609
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 22.2800572030574,
+      "learning_rate": 5.574434935731723e-06,
+      "loss": 0.4784,
+      "step": 12610
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 26.826145475115595,
+      "learning_rate": 5.572922736206135e-06,
+      "loss": 0.502,
+      "step": 12611
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.839644765217468,
+      "learning_rate": 5.571410662584029e-06,
+      "loss": 0.4732,
+      "step": 12612
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.342510382058945,
+      "learning_rate": 5.5698987149084174e-06,
+      "loss": 0.5158,
+      "step": 12613
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.544607838055274,
+      "learning_rate": 5.568386893222291e-06,
+      "loss": 0.5141,
+      "step": 12614
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.28824598855327,
+      "learning_rate": 5.5668751975686485e-06,
+      "loss": 0.6036,
+      "step": 12615
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 35.074450238780216,
+      "learning_rate": 5.565363627990485e-06,
+      "loss": 0.5349,
+      "step": 12616
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.567612451319665,
+      "learning_rate": 5.563852184530784e-06,
+      "loss": 0.4689,
+      "step": 12617
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.957472684793785,
+      "learning_rate": 5.562340867232535e-06,
+      "loss": 0.6178,
+      "step": 12618
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.258346950241314,
+      "learning_rate": 5.560829676138712e-06,
+      "loss": 0.5912,
+      "step": 12619
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.857396440666978,
+      "learning_rate": 5.559318611292299e-06,
+      "loss": 0.5387,
+      "step": 12620
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 22.970008307004164,
+      "learning_rate": 5.557807672736264e-06,
+      "loss": 0.5527,
+      "step": 12621
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 25.550777935103092,
+      "learning_rate": 5.556296860513584e-06,
+      "loss": 0.4918,
+      "step": 12622
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.66604013467039,
+      "learning_rate": 5.554786174667217e-06,
+      "loss": 0.5096,
+      "step": 12623
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.533930410269456,
+      "learning_rate": 5.553275615240132e-06,
+      "loss": 0.5125,
+      "step": 12624
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.539640807673022,
+      "learning_rate": 5.551765182275292e-06,
+      "loss": 0.4658,
+      "step": 12625
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.966272654662532,
+      "learning_rate": 5.550254875815646e-06,
+      "loss": 0.519,
+      "step": 12626
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.805065626939996,
+      "learning_rate": 5.548744695904145e-06,
+      "loss": 0.4964,
+      "step": 12627
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.900625258009978,
+      "learning_rate": 5.5472346425837405e-06,
+      "loss": 0.56,
+      "step": 12628
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.063630668201945,
+      "learning_rate": 5.545724715897381e-06,
+      "loss": 0.5138,
+      "step": 12629
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.596001596001994,
+      "learning_rate": 5.544214915888006e-06,
+      "loss": 0.5456,
+      "step": 12630
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.340409046731004,
+      "learning_rate": 5.542705242598552e-06,
+      "loss": 0.5157,
+      "step": 12631
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.661060137275072,
+      "learning_rate": 5.54119569607195e-06,
+      "loss": 0.5138,
+      "step": 12632
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 22.36441971592129,
+      "learning_rate": 5.539686276351135e-06,
+      "loss": 0.495,
+      "step": 12633
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.35537810670071,
+      "learning_rate": 5.538176983479036e-06,
+      "loss": 0.5449,
+      "step": 12634
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.332534711553333,
+      "learning_rate": 5.536667817498573e-06,
+      "loss": 0.4595,
+      "step": 12635
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 25.484735979836614,
+      "learning_rate": 5.535158778452664e-06,
+      "loss": 0.5956,
+      "step": 12636
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.2746785231696,
+      "learning_rate": 5.533649866384226e-06,
+      "loss": 0.5036,
+      "step": 12637
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.945022102042472,
+      "learning_rate": 5.532141081336177e-06,
+      "loss": 0.5224,
+      "step": 12638
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.935658346049768,
+      "learning_rate": 5.530632423351421e-06,
+      "loss": 0.4443,
+      "step": 12639
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.51050762449517,
+      "learning_rate": 5.5291238924728654e-06,
+      "loss": 0.4302,
+      "step": 12640
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.027996218024988,
+      "learning_rate": 5.5276154887434075e-06,
+      "loss": 0.5304,
+      "step": 12641
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.159128770952982,
+      "learning_rate": 5.526107212205949e-06,
+      "loss": 0.5162,
+      "step": 12642
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.50456143483398,
+      "learning_rate": 5.524599062903388e-06,
+      "loss": 0.512,
+      "step": 12643
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.944525703500673,
+      "learning_rate": 5.523091040878608e-06,
+      "loss": 0.5381,
+      "step": 12644
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.98428168009199,
+      "learning_rate": 5.521583146174503e-06,
+      "loss": 0.4688,
+      "step": 12645
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.4442419198208,
+      "learning_rate": 5.5200753788339515e-06,
+      "loss": 0.5068,
+      "step": 12646
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.575833924074296,
+      "learning_rate": 5.518567738899838e-06,
+      "loss": 0.5214,
+      "step": 12647
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.677035146254948,
+      "learning_rate": 5.517060226415032e-06,
+      "loss": 0.4799,
+      "step": 12648
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.97043014080274,
+      "learning_rate": 5.515552841422412e-06,
+      "loss": 0.509,
+      "step": 12649
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.025432938561142,
+      "learning_rate": 5.514045583964848e-06,
+      "loss": 0.4991,
+      "step": 12650
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.943064070286184,
+      "learning_rate": 5.512538454085206e-06,
+      "loss": 0.5119,
+      "step": 12651
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.590342849984218,
+      "learning_rate": 5.51103145182634e-06,
+      "loss": 0.493,
+      "step": 12652
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.1280554974113,
+      "learning_rate": 5.509524577231114e-06,
+      "loss": 0.5113,
+      "step": 12653
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.01443975128287,
+      "learning_rate": 5.50801783034238e-06,
+      "loss": 0.4957,
+      "step": 12654
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.94761608709194,
+      "learning_rate": 5.506511211202997e-06,
+      "loss": 0.5017,
+      "step": 12655
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.506820794649318,
+      "learning_rate": 5.505004719855806e-06,
+      "loss": 0.4814,
+      "step": 12656
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.043393147162146,
+      "learning_rate": 5.503498356343648e-06,
+      "loss": 0.4844,
+      "step": 12657
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.508233658479071,
+      "learning_rate": 5.501992120709367e-06,
+      "loss": 0.4701,
+      "step": 12658
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.224943003368498,
+      "learning_rate": 5.500486012995801e-06,
+      "loss": 0.4635,
+      "step": 12659
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.72161179764742,
+      "learning_rate": 5.498980033245782e-06,
+      "loss": 0.5972,
+      "step": 12660
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.86460772424473,
+      "learning_rate": 5.4974741815021336e-06,
+      "loss": 0.5446,
+      "step": 12661
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.606701164842864,
+      "learning_rate": 5.4959684578076855e-06,
+      "loss": 0.4804,
+      "step": 12662
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 23.86005248636155,
+      "learning_rate": 5.494462862205263e-06,
+      "loss": 0.5453,
+      "step": 12663
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.910636331555022,
+      "learning_rate": 5.492957394737677e-06,
+      "loss": 0.5336,
+      "step": 12664
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.01784423432177,
+      "learning_rate": 5.49145205544775e-06,
+      "loss": 0.4901,
+      "step": 12665
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.87061895546319,
+      "learning_rate": 5.4899468443782864e-06,
+      "loss": 0.4757,
+      "step": 12666
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 31.910062698953737,
+      "learning_rate": 5.488441761572096e-06,
+      "loss": 0.5706,
+      "step": 12667
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 25.82587874949454,
+      "learning_rate": 5.486936807071986e-06,
+      "loss": 0.532,
+      "step": 12668
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.129131046965366,
+      "learning_rate": 5.485431980920753e-06,
+      "loss": 0.5269,
+      "step": 12669
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 23.77725639534025,
+      "learning_rate": 5.4839272831611905e-06,
+      "loss": 0.5058,
+      "step": 12670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.692392764572553,
+      "learning_rate": 5.482422713836094e-06,
+      "loss": 0.4975,
+      "step": 12671
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.37550091415856,
+      "learning_rate": 5.480918272988256e-06,
+      "loss": 0.5867,
+      "step": 12672
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.749561082540527,
+      "learning_rate": 5.4794139606604556e-06,
+      "loss": 0.4644,
+      "step": 12673
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.30856260188741,
+      "learning_rate": 5.477909776895481e-06,
+      "loss": 0.5362,
+      "step": 12674
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.235047056584289,
+      "learning_rate": 5.476405721736104e-06,
+      "loss": 0.5213,
+      "step": 12675
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.307820831949023,
+      "learning_rate": 5.474901795225105e-06,
+      "loss": 0.4949,
+      "step": 12676
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.21044316117008,
+      "learning_rate": 5.473397997405249e-06,
+      "loss": 0.591,
+      "step": 12677
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 24.42783966066164,
+      "learning_rate": 5.471894328319305e-06,
+      "loss": 0.5754,
+      "step": 12678
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.827380236440565,
+      "learning_rate": 5.470390788010042e-06,
+      "loss": 0.4074,
+      "step": 12679
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.652490441152498,
+      "learning_rate": 5.4688873765202114e-06,
+      "loss": 0.476,
+      "step": 12680
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.85093569175643,
+      "learning_rate": 5.467384093892576e-06,
+      "loss": 0.5046,
+      "step": 12681
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.883627884861234,
+      "learning_rate": 5.465880940169881e-06,
+      "loss": 0.4855,
+      "step": 12682
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.787404535248456,
+      "learning_rate": 5.464377915394882e-06,
+      "loss": 0.4596,
+      "step": 12683
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 23.843085502224415,
+      "learning_rate": 5.4628750196103245e-06,
+      "loss": 0.5642,
+      "step": 12684
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.258187935913085,
+      "learning_rate": 5.461372252858948e-06,
+      "loss": 0.4234,
+      "step": 12685
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.812965002678975,
+      "learning_rate": 5.459869615183484e-06,
+      "loss": 0.5264,
+      "step": 12686
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.075905620799684,
+      "learning_rate": 5.458367106626674e-06,
+      "loss": 0.4839,
+      "step": 12687
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 31.14433998849154,
+      "learning_rate": 5.456864727231246e-06,
+      "loss": 0.5968,
+      "step": 12688
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.173216462374063,
+      "learning_rate": 5.455362477039935e-06,
+      "loss": 0.4966,
+      "step": 12689
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 23.078196366625278,
+      "learning_rate": 5.453860356095448e-06,
+      "loss": 0.515,
+      "step": 12690
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.39923988531013,
+      "learning_rate": 5.452358364440515e-06,
+      "loss": 0.4839,
+      "step": 12691
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 34.13339292234002,
+      "learning_rate": 5.450856502117849e-06,
+      "loss": 0.5461,
+      "step": 12692
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 46.60784001688967,
+      "learning_rate": 5.449354769170168e-06,
+      "loss": 0.4725,
+      "step": 12693
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.84592823884347,
+      "learning_rate": 5.447853165640173e-06,
+      "loss": 0.4901,
+      "step": 12694
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.6065309328846,
+      "learning_rate": 5.4463516915705684e-06,
+      "loss": 0.5196,
+      "step": 12695
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.75612199254925,
+      "learning_rate": 5.444850347004058e-06,
+      "loss": 0.4887,
+      "step": 12696
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.72679947134772,
+      "learning_rate": 5.443349131983343e-06,
+      "loss": 0.5495,
+      "step": 12697
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.374805875166246,
+      "learning_rate": 5.441848046551108e-06,
+      "loss": 0.4772,
+      "step": 12698
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 23.827177509168333,
+      "learning_rate": 5.440347090750053e-06,
+      "loss": 0.4797,
+      "step": 12699
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.745513655058687,
+      "learning_rate": 5.438846264622857e-06,
+      "loss": 0.5091,
+      "step": 12700
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.107582728295293,
+      "learning_rate": 5.437345568212207e-06,
+      "loss": 0.4981,
+      "step": 12701
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 27.19873679217685,
+      "learning_rate": 5.435845001560775e-06,
+      "loss": 0.6202,
+      "step": 12702
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.438463846506803,
+      "learning_rate": 5.434344564711244e-06,
+      "loss": 0.534,
+      "step": 12703
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 34.288939828016595,
+      "learning_rate": 5.432844257706285e-06,
+      "loss": 0.5175,
+      "step": 12704
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.999118775092427,
+      "learning_rate": 5.431344080588561e-06,
+      "loss": 0.479,
+      "step": 12705
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.19710137900284,
+      "learning_rate": 5.4298440334007415e-06,
+      "loss": 0.4579,
+      "step": 12706
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.66095380333795,
+      "learning_rate": 5.42834411618548e-06,
+      "loss": 0.5239,
+      "step": 12707
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.466056785167183,
+      "learning_rate": 5.426844328985439e-06,
+      "loss": 0.4562,
+      "step": 12708
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.137920294845857,
+      "learning_rate": 5.4253446718432724e-06,
+      "loss": 0.5183,
+      "step": 12709
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.31210413416006,
+      "learning_rate": 5.423845144801627e-06,
+      "loss": 0.449,
+      "step": 12710
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.312527322911524,
+      "learning_rate": 5.422345747903146e-06,
+      "loss": 0.581,
+      "step": 12711
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.636423452958091,
+      "learning_rate": 5.420846481190473e-06,
+      "loss": 0.4649,
+      "step": 12712
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 20.911916545742397,
+      "learning_rate": 5.41934734470625e-06,
+      "loss": 0.443,
+      "step": 12713
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.308307962485706,
+      "learning_rate": 5.417848338493114e-06,
+      "loss": 0.4783,
+      "step": 12714
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.178909942491316,
+      "learning_rate": 5.416349462593684e-06,
+      "loss": 0.4977,
+      "step": 12715
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.61768315827325,
+      "learning_rate": 5.414850717050593e-06,
+      "loss": 0.567,
+      "step": 12716
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 31.15318871120512,
+      "learning_rate": 5.413352101906466e-06,
+      "loss": 0.5068,
+      "step": 12717
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.16018160429881,
+      "learning_rate": 5.411853617203926e-06,
+      "loss": 0.5339,
+      "step": 12718
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.35363165088891,
+      "learning_rate": 5.410355262985585e-06,
+      "loss": 0.6084,
+      "step": 12719
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 24.946300817947172,
+      "learning_rate": 5.408857039294052e-06,
+      "loss": 0.606,
+      "step": 12720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.25809714569129,
+      "learning_rate": 5.407358946171939e-06,
+      "loss": 0.5121,
+      "step": 12721
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 20.165364431807888,
+      "learning_rate": 5.405860983661854e-06,
+      "loss": 0.5182,
+      "step": 12722
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.306420280230563,
+      "learning_rate": 5.404363151806397e-06,
+      "loss": 0.5631,
+      "step": 12723
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 21.31934463978182,
+      "learning_rate": 5.402865450648158e-06,
+      "loss": 0.5226,
+      "step": 12724
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.09855279634109,
+      "learning_rate": 5.401367880229737e-06,
+      "loss": 0.4902,
+      "step": 12725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 20.092424936343008,
+      "learning_rate": 5.399870440593726e-06,
+      "loss": 0.4783,
+      "step": 12726
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 24.567248699527127,
+      "learning_rate": 5.3983731317827075e-06,
+      "loss": 0.4904,
+      "step": 12727
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 24.319415452431414,
+      "learning_rate": 5.396875953839267e-06,
+      "loss": 0.5109,
+      "step": 12728
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.834170988666447,
+      "learning_rate": 5.3953789068059785e-06,
+      "loss": 0.47,
+      "step": 12729
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.93428220054176,
+      "learning_rate": 5.3938819907254204e-06,
+      "loss": 0.499,
+      "step": 12730
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.772634888247694,
+      "learning_rate": 5.392385205640167e-06,
+      "loss": 0.4686,
+      "step": 12731
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 28.175442800283278,
+      "learning_rate": 5.390888551592779e-06,
+      "loss": 0.5549,
+      "step": 12732
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.662389325299316,
+      "learning_rate": 5.3893920286258285e-06,
+      "loss": 0.4863,
+      "step": 12733
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.161478477555537,
+      "learning_rate": 5.387895636781868e-06,
+      "loss": 0.531,
+      "step": 12734
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.2288172786045,
+      "learning_rate": 5.386399376103462e-06,
+      "loss": 0.5023,
+      "step": 12735
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 29.600610627620956,
+      "learning_rate": 5.3849032466331555e-06,
+      "loss": 0.5666,
+      "step": 12736
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.337931923220637,
+      "learning_rate": 5.3834072484134995e-06,
+      "loss": 0.4963,
+      "step": 12737
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 30.882099074854956,
+      "learning_rate": 5.381911381487044e-06,
+      "loss": 0.5255,
+      "step": 12738
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.9306973341253,
+      "learning_rate": 5.380415645896329e-06,
+      "loss": 0.5578,
+      "step": 12739
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 23.92875505300171,
+      "learning_rate": 5.378920041683886e-06,
+      "loss": 0.5569,
+      "step": 12740
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 25.04182570708978,
+      "learning_rate": 5.377424568892253e-06,
+      "loss": 0.5425,
+      "step": 12741
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.041537775774856,
+      "learning_rate": 5.375929227563963e-06,
+      "loss": 0.4817,
+      "step": 12742
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.55573722194232,
+      "learning_rate": 5.374434017741543e-06,
+      "loss": 0.5313,
+      "step": 12743
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 23.24591995787106,
+      "learning_rate": 5.372938939467514e-06,
+      "loss": 0.5362,
+      "step": 12744
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 23.268212947544505,
+      "learning_rate": 5.371443992784391e-06,
+      "loss": 0.5324,
+      "step": 12745
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.161217592649967,
+      "learning_rate": 5.3699491777346935e-06,
+      "loss": 0.4901,
+      "step": 12746
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.94016474127129,
+      "learning_rate": 5.3684544943609375e-06,
+      "loss": 0.5367,
+      "step": 12747
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.399479291889282,
+      "learning_rate": 5.366959942705625e-06,
+      "loss": 0.567,
+      "step": 12748
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.848442611949254,
+      "learning_rate": 5.3654655228112574e-06,
+      "loss": 0.5463,
+      "step": 12749
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 21.540081386106916,
+      "learning_rate": 5.36397123472034e-06,
+      "loss": 0.5711,
+      "step": 12750
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.756696683615296,
+      "learning_rate": 5.362477078475372e-06,
+      "loss": 0.4785,
+      "step": 12751
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 27.4221936579608,
+      "learning_rate": 5.36098305411884e-06,
+      "loss": 0.5048,
+      "step": 12752
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.355368281571305,
+      "learning_rate": 5.35948916169324e-06,
+      "loss": 0.4539,
+      "step": 12753
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.192543244568437,
+      "learning_rate": 5.357995401241049e-06,
+      "loss": 0.5127,
+      "step": 12754
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.156664353024194,
+      "learning_rate": 5.3565017728047545e-06,
+      "loss": 0.5508,
+      "step": 12755
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.408044845592613,
+      "learning_rate": 5.355008276426836e-06,
+      "loss": 0.5727,
+      "step": 12756
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 27.378817416501576,
+      "learning_rate": 5.353514912149761e-06,
+      "loss": 0.5143,
+      "step": 12757
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.563650196440346,
+      "learning_rate": 5.3520216800160085e-06,
+      "loss": 0.489,
+      "step": 12758
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.96476733199368,
+      "learning_rate": 5.350528580068035e-06,
+      "loss": 0.545,
+      "step": 12759
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 23.688638459101494,
+      "learning_rate": 5.349035612348314e-06,
+      "loss": 0.4843,
+      "step": 12760
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.728049830840234,
+      "learning_rate": 5.347542776899295e-06,
+      "loss": 0.4693,
+      "step": 12761
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 23.359305906672024,
+      "learning_rate": 5.346050073763438e-06,
+      "loss": 0.5438,
+      "step": 12762
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.39442096690375,
+      "learning_rate": 5.344557502983198e-06,
+      "loss": 0.6188,
+      "step": 12763
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.823107072761507,
+      "learning_rate": 5.343065064601021e-06,
+      "loss": 0.5734,
+      "step": 12764
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 24.274501514472956,
+      "learning_rate": 5.3415727586593435e-06,
+      "loss": 0.5177,
+      "step": 12765
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.339050194956446,
+      "learning_rate": 5.340080585200612e-06,
+      "loss": 0.5186,
+      "step": 12766
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.009594334932329,
+      "learning_rate": 5.338588544267267e-06,
+      "loss": 0.5078,
+      "step": 12767
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 21.263938690498964,
+      "learning_rate": 5.3370966359017325e-06,
+      "loss": 0.5454,
+      "step": 12768
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.471205769851714,
+      "learning_rate": 5.335604860146446e-06,
+      "loss": 0.4558,
+      "step": 12769
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.348412244402393,
+      "learning_rate": 5.334113217043826e-06,
+      "loss": 0.4622,
+      "step": 12770
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 28.880391896280504,
+      "learning_rate": 5.332621706636295e-06,
+      "loss": 0.5254,
+      "step": 12771
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 18.312226957834174,
+      "learning_rate": 5.331130328966276e-06,
+      "loss": 0.5503,
+      "step": 12772
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.21481608547601,
+      "learning_rate": 5.329639084076181e-06,
+      "loss": 0.5063,
+      "step": 12773
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 23.63318441555478,
+      "learning_rate": 5.3281479720084125e-06,
+      "loss": 0.4939,
+      "step": 12774
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.814329478358903,
+      "learning_rate": 5.326656992805384e-06,
+      "loss": 0.4858,
+      "step": 12775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.21986106136835,
+      "learning_rate": 5.325166146509497e-06,
+      "loss": 0.4915,
+      "step": 12776
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.45065739390407,
+      "learning_rate": 5.323675433163158e-06,
+      "loss": 0.5109,
+      "step": 12777
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 18.690695171323338,
+      "learning_rate": 5.322184852808745e-06,
+      "loss": 0.5427,
+      "step": 12778
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.935566199145306,
+      "learning_rate": 5.32069440548866e-06,
+      "loss": 0.4739,
+      "step": 12779
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.900462675804087,
+      "learning_rate": 5.319204091245287e-06,
+      "loss": 0.5493,
+      "step": 12780
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.957816886321336,
+      "learning_rate": 5.317713910121016e-06,
+      "loss": 0.5106,
+      "step": 12781
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.700480878612106,
+      "learning_rate": 5.316223862158223e-06,
+      "loss": 0.4849,
+      "step": 12782
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.165340633469405,
+      "learning_rate": 5.31473394739928e-06,
+      "loss": 0.486,
+      "step": 12783
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.21583252138848,
+      "learning_rate": 5.313244165886562e-06,
+      "loss": 0.5137,
+      "step": 12784
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.628529801300667,
+      "learning_rate": 5.311754517662442e-06,
+      "loss": 0.5165,
+      "step": 12785
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 28.03665938858337,
+      "learning_rate": 5.310265002769277e-06,
+      "loss": 0.5386,
+      "step": 12786
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 26.411248315271912,
+      "learning_rate": 5.308775621249435e-06,
+      "loss": 0.5839,
+      "step": 12787
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.33712262346072,
+      "learning_rate": 5.3072863731452674e-06,
+      "loss": 0.5538,
+      "step": 12788
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.014096949390407,
+      "learning_rate": 5.305797258499134e-06,
+      "loss": 0.5904,
+      "step": 12789
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 24.258354472127433,
+      "learning_rate": 5.304308277353376e-06,
+      "loss": 0.5926,
+      "step": 12790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 28.179351920976487,
+      "learning_rate": 5.3028194297503445e-06,
+      "loss": 0.5284,
+      "step": 12791
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 23.945392149252918,
+      "learning_rate": 5.301330715732385e-06,
+      "loss": 0.5319,
+      "step": 12792
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 18.357873436416508,
+      "learning_rate": 5.299842135341825e-06,
+      "loss": 0.5489,
+      "step": 12793
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.088004487734466,
+      "learning_rate": 5.29835368862101e-06,
+      "loss": 0.5624,
+      "step": 12794
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.059239912012366,
+      "learning_rate": 5.296865375612264e-06,
+      "loss": 0.5117,
+      "step": 12795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.88174735815471,
+      "learning_rate": 5.295377196357914e-06,
+      "loss": 0.5244,
+      "step": 12796
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 24.131125571783016,
+      "learning_rate": 5.293889150900287e-06,
+      "loss": 0.5465,
+      "step": 12797
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.5814515033147,
+      "learning_rate": 5.2924012392817014e-06,
+      "loss": 0.5624,
+      "step": 12798
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.22358408168549,
+      "learning_rate": 5.290913461544466e-06,
+      "loss": 0.5068,
+      "step": 12799
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.86146255158927,
+      "learning_rate": 5.289425817730897e-06,
+      "loss": 0.4743,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.716995784140252,
+      "learning_rate": 5.287938307883302e-06,
+      "loss": 0.5308,
+      "step": 12801
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.709844466138364,
+      "learning_rate": 5.286450932043994e-06,
+      "loss": 0.4986,
+      "step": 12802
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.515052494807907,
+      "learning_rate": 5.284963690255254e-06,
+      "loss": 0.4749,
+      "step": 12803
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.939860179421112,
+      "learning_rate": 5.28347658255939e-06,
+      "loss": 0.5057,
+      "step": 12804
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.851529592571353,
+      "learning_rate": 5.281989608998693e-06,
+      "loss": 0.5132,
+      "step": 12805
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.747981189919482,
+      "learning_rate": 5.280502769615456e-06,
+      "loss": 0.4984,
+      "step": 12806
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.71458497372247,
+      "learning_rate": 5.279016064451959e-06,
+      "loss": 0.5556,
+      "step": 12807
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.483231372346973,
+      "learning_rate": 5.277529493550478e-06,
+      "loss": 0.5781,
+      "step": 12808
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.501442855437578,
+      "learning_rate": 5.276043056953297e-06,
+      "loss": 0.4606,
+      "step": 12809
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 34.23938197375652,
+      "learning_rate": 5.274556754702691e-06,
+      "loss": 0.521,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.855925610686196,
+      "learning_rate": 5.273070586840925e-06,
+      "loss": 0.5214,
+      "step": 12811
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.92509189624372,
+      "learning_rate": 5.271584553410267e-06,
+      "loss": 0.5487,
+      "step": 12812
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.282506018392025,
+      "learning_rate": 5.270098654452977e-06,
+      "loss": 0.5312,
+      "step": 12813
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.534534498495953,
+      "learning_rate": 5.268612890011318e-06,
+      "loss": 0.4754,
+      "step": 12814
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.077242536635687,
+      "learning_rate": 5.267127260127536e-06,
+      "loss": 0.5234,
+      "step": 12815
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.95447469177707,
+      "learning_rate": 5.26564176484389e-06,
+      "loss": 0.5423,
+      "step": 12816
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.944773742849737,
+      "learning_rate": 5.264156404202618e-06,
+      "loss": 0.5707,
+      "step": 12817
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.50944322472432,
+      "learning_rate": 5.262671178245968e-06,
+      "loss": 0.5324,
+      "step": 12818
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.443451093896535,
+      "learning_rate": 5.261186087016183e-06,
+      "loss": 0.533,
+      "step": 12819
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.053129724739613,
+      "learning_rate": 5.259701130555489e-06,
+      "loss": 0.4852,
+      "step": 12820
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.67333586642384,
+      "learning_rate": 5.2582163089061255e-06,
+      "loss": 0.5785,
+      "step": 12821
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.678764346717594,
+      "learning_rate": 5.256731622110314e-06,
+      "loss": 0.4252,
+      "step": 12822
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.637510476205016,
+      "learning_rate": 5.2552470702102835e-06,
+      "loss": 0.5379,
+      "step": 12823
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.331672049249445,
+      "learning_rate": 5.253762653248245e-06,
+      "loss": 0.526,
+      "step": 12824
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.219938923691952,
+      "learning_rate": 5.252278371266422e-06,
+      "loss": 0.449,
+      "step": 12825
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.178244468847648,
+      "learning_rate": 5.250794224307028e-06,
+      "loss": 0.4775,
+      "step": 12826
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.999490883948951,
+      "learning_rate": 5.249310212412269e-06,
+      "loss": 0.503,
+      "step": 12827
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.402767786844866,
+      "learning_rate": 5.247826335624344e-06,
+      "loss": 0.5055,
+      "step": 12828
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.37863443841836,
+      "learning_rate": 5.246342593985458e-06,
+      "loss": 0.5109,
+      "step": 12829
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.144753117740438,
+      "learning_rate": 5.244858987537808e-06,
+      "loss": 0.482,
+      "step": 12830
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.419721878328797,
+      "learning_rate": 5.243375516323591e-06,
+      "loss": 0.5129,
+      "step": 12831
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 18.602522359449324,
+      "learning_rate": 5.241892180384991e-06,
+      "loss": 0.511,
+      "step": 12832
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.599632476802746,
+      "learning_rate": 5.240408979764191e-06,
+      "loss": 0.4747,
+      "step": 12833
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.39301927661398,
+      "learning_rate": 5.2389259145033744e-06,
+      "loss": 0.5152,
+      "step": 12834
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.495350274620353,
+      "learning_rate": 5.237442984644724e-06,
+      "loss": 0.4846,
+      "step": 12835
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.860988160947693,
+      "learning_rate": 5.235960190230409e-06,
+      "loss": 0.4342,
+      "step": 12836
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.270513452675903,
+      "learning_rate": 5.234477531302595e-06,
+      "loss": 0.5372,
+      "step": 12837
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.419396131694374,
+      "learning_rate": 5.232995007903453e-06,
+      "loss": 0.534,
+      "step": 12838
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 33.4992244745453,
+      "learning_rate": 5.231512620075143e-06,
+      "loss": 0.566,
+      "step": 12839
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.161195104967208,
+      "learning_rate": 5.23003036785983e-06,
+      "loss": 0.4307,
+      "step": 12840
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.680810002096436,
+      "learning_rate": 5.228548251299663e-06,
+      "loss": 0.429,
+      "step": 12841
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 22.076746387999645,
+      "learning_rate": 5.227066270436788e-06,
+      "loss": 0.5123,
+      "step": 12842
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.717862502103365,
+      "learning_rate": 5.225584425313357e-06,
+      "loss": 0.5131,
+      "step": 12843
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.42245341178339,
+      "learning_rate": 5.224102715971515e-06,
+      "loss": 0.4998,
+      "step": 12844
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.873891543429119,
+      "learning_rate": 5.222621142453394e-06,
+      "loss": 0.506,
+      "step": 12845
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.739271958784194,
+      "learning_rate": 5.221139704801137e-06,
+      "loss": 0.4974,
+      "step": 12846
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.322773141463294,
+      "learning_rate": 5.219658403056867e-06,
+      "loss": 0.4874,
+      "step": 12847
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.303040486121585,
+      "learning_rate": 5.218177237262721e-06,
+      "loss": 0.5105,
+      "step": 12848
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.723988803755613,
+      "learning_rate": 5.216696207460812e-06,
+      "loss": 0.5677,
+      "step": 12849
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.079114685983859,
+      "learning_rate": 5.215215313693265e-06,
+      "loss": 0.4846,
+      "step": 12850
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.86693381951216,
+      "learning_rate": 5.213734556002199e-06,
+      "loss": 0.5356,
+      "step": 12851
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.769928964439313,
+      "learning_rate": 5.212253934429723e-06,
+      "loss": 0.5335,
+      "step": 12852
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.150151082419267,
+      "learning_rate": 5.210773449017939e-06,
+      "loss": 0.5059,
+      "step": 12853
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.212346614400335,
+      "learning_rate": 5.209293099808959e-06,
+      "loss": 0.4648,
+      "step": 12854
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.03153420504713,
+      "learning_rate": 5.207812886844879e-06,
+      "loss": 0.5475,
+      "step": 12855
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.486372312635062,
+      "learning_rate": 5.2063328101678004e-06,
+      "loss": 0.5562,
+      "step": 12856
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.595165511905176,
+      "learning_rate": 5.204852869819814e-06,
+      "loss": 0.4704,
+      "step": 12857
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.74521933199456,
+      "learning_rate": 5.203373065843003e-06,
+      "loss": 0.5267,
+      "step": 12858
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 22.16987454864099,
+      "learning_rate": 5.201893398279454e-06,
+      "loss": 0.5634,
+      "step": 12859
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 55.38955412584117,
+      "learning_rate": 5.2004138671712555e-06,
+      "loss": 0.5711,
+      "step": 12860
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.553992401398364,
+      "learning_rate": 5.198934472560479e-06,
+      "loss": 0.5495,
+      "step": 12861
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.043493991234104,
+      "learning_rate": 5.197455214489193e-06,
+      "loss": 0.4802,
+      "step": 12862
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.467267454489008,
+      "learning_rate": 5.195976092999472e-06,
+      "loss": 0.5754,
+      "step": 12863
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.545558530574674,
+      "learning_rate": 5.194497108133381e-06,
+      "loss": 0.5344,
+      "step": 12864
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.526643115518555,
+      "learning_rate": 5.193018259932987e-06,
+      "loss": 0.4985,
+      "step": 12865
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.22003788214398,
+      "learning_rate": 5.191539548440336e-06,
+      "loss": 0.5001,
+      "step": 12866
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 22.295615203409778,
+      "learning_rate": 5.190060973697486e-06,
+      "loss": 0.5713,
+      "step": 12867
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.880064723342546,
+      "learning_rate": 5.188582535746489e-06,
+      "loss": 0.4764,
+      "step": 12868
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.816532976525576,
+      "learning_rate": 5.187104234629394e-06,
+      "loss": 0.4935,
+      "step": 12869
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.477431978855728,
+      "learning_rate": 5.185626070388239e-06,
+      "loss": 0.4337,
+      "step": 12870
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.118029455207974,
+      "learning_rate": 5.184148043065058e-06,
+      "loss": 0.5271,
+      "step": 12871
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 17.57273564775897,
+      "learning_rate": 5.182670152701889e-06,
+      "loss": 0.496,
+      "step": 12872
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.403528377403912,
+      "learning_rate": 5.181192399340768e-06,
+      "loss": 0.561,
+      "step": 12873
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.00317202921992,
+      "learning_rate": 5.179714783023711e-06,
+      "loss": 0.5042,
+      "step": 12874
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.11761677830964,
+      "learning_rate": 5.17823730379275e-06,
+      "loss": 0.4443,
+      "step": 12875
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.650502661525685,
+      "learning_rate": 5.1767599616898965e-06,
+      "loss": 0.5648,
+      "step": 12876
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.093180487730304,
+      "learning_rate": 5.175282756757172e-06,
+      "loss": 0.5177,
+      "step": 12877
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.59031068543,
+      "learning_rate": 5.1738056890365775e-06,
+      "loss": 0.4623,
+      "step": 12878
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 30.362432434131314,
+      "learning_rate": 5.1723287585701285e-06,
+      "loss": 0.5351,
+      "step": 12879
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.879641186058198,
+      "learning_rate": 5.1708519653998275e-06,
+      "loss": 0.5023,
+      "step": 12880
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.007617434015483,
+      "learning_rate": 5.169375309567669e-06,
+      "loss": 0.5119,
+      "step": 12881
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 31.5426782331286,
+      "learning_rate": 5.167898791115654e-06,
+      "loss": 0.5462,
+      "step": 12882
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.6085095704948,
+      "learning_rate": 5.166422410085767e-06,
+      "loss": 0.498,
+      "step": 12883
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.62602963228964,
+      "learning_rate": 5.16494616652e-06,
+      "loss": 0.5543,
+      "step": 12884
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 27.532708059870203,
+      "learning_rate": 5.163470060460338e-06,
+      "loss": 0.5182,
+      "step": 12885
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.561550753199148,
+      "learning_rate": 5.161994091948759e-06,
+      "loss": 0.5083,
+      "step": 12886
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.228462194603114,
+      "learning_rate": 5.160518261027234e-06,
+      "loss": 0.5143,
+      "step": 12887
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.861113874822482,
+      "learning_rate": 5.159042567737739e-06,
+      "loss": 0.5461,
+      "step": 12888
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 43.83926517526602,
+      "learning_rate": 5.157567012122241e-06,
+      "loss": 0.5495,
+      "step": 12889
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.808685678283762,
+      "learning_rate": 5.1560915942227115e-06,
+      "loss": 0.4999,
+      "step": 12890
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 27.285094730905854,
+      "learning_rate": 5.154616314081098e-06,
+      "loss": 0.4821,
+      "step": 12891
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.975509671121834,
+      "learning_rate": 5.15314117173936e-06,
+      "loss": 0.4987,
+      "step": 12892
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.795217110536342,
+      "learning_rate": 5.151666167239452e-06,
+      "loss": 0.576,
+      "step": 12893
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.030537334907187,
+      "learning_rate": 5.150191300623326e-06,
+      "loss": 0.5055,
+      "step": 12894
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.954326292021857,
+      "learning_rate": 5.1487165719329225e-06,
+      "loss": 0.4767,
+      "step": 12895
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.781848346946983,
+      "learning_rate": 5.147241981210178e-06,
+      "loss": 0.4882,
+      "step": 12896
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 26.721712180230348,
+      "learning_rate": 5.145767528497032e-06,
+      "loss": 0.495,
+      "step": 12897
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.71099786092121,
+      "learning_rate": 5.1442932138354225e-06,
+      "loss": 0.5591,
+      "step": 12898
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.50726376843265,
+      "learning_rate": 5.142819037267271e-06,
+      "loss": 0.4738,
+      "step": 12899
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.43351219853536,
+      "learning_rate": 5.1413449988345055e-06,
+      "loss": 0.446,
+      "step": 12900
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.86939762003122,
+      "learning_rate": 5.139871098579045e-06,
+      "loss": 0.4269,
+      "step": 12901
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.580438248351527,
+      "learning_rate": 5.138397336542809e-06,
+      "loss": 0.4845,
+      "step": 12902
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 30.333315181639414,
+      "learning_rate": 5.136923712767706e-06,
+      "loss": 0.4742,
+      "step": 12903
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.771202178233892,
+      "learning_rate": 5.1354502272956486e-06,
+      "loss": 0.5283,
+      "step": 12904
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 26.203682428366672,
+      "learning_rate": 5.133976880168543e-06,
+      "loss": 0.5423,
+      "step": 12905
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.895838072514895,
+      "learning_rate": 5.132503671428286e-06,
+      "loss": 0.4642,
+      "step": 12906
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.36616706681718,
+      "learning_rate": 5.13103060111678e-06,
+      "loss": 0.4566,
+      "step": 12907
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 15.740179368285217,
+      "learning_rate": 5.129557669275913e-06,
+      "loss": 0.5351,
+      "step": 12908
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.187535147887303,
+      "learning_rate": 5.128084875947579e-06,
+      "loss": 0.4946,
+      "step": 12909
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.08145395379845,
+      "learning_rate": 5.126612221173659e-06,
+      "loss": 0.4965,
+      "step": 12910
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.949772486213806,
+      "learning_rate": 5.125139704996038e-06,
+      "loss": 0.5379,
+      "step": 12911
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.47096047268034,
+      "learning_rate": 5.123667327456591e-06,
+      "loss": 0.4741,
+      "step": 12912
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.483268776302502,
+      "learning_rate": 5.12219508859719e-06,
+      "loss": 0.5142,
+      "step": 12913
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.783053883549393,
+      "learning_rate": 5.1207229884597135e-06,
+      "loss": 0.5842,
+      "step": 12914
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.618585379566344,
+      "learning_rate": 5.119251027086019e-06,
+      "loss": 0.4189,
+      "step": 12915
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.84419775846366,
+      "learning_rate": 5.117779204517967e-06,
+      "loss": 0.4689,
+      "step": 12916
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.524304974100705,
+      "learning_rate": 5.116307520797419e-06,
+      "loss": 0.5453,
+      "step": 12917
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.32835194981006,
+      "learning_rate": 5.114835975966228e-06,
+      "loss": 0.4066,
+      "step": 12918
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 24.56690370841834,
+      "learning_rate": 5.113364570066248e-06,
+      "loss": 0.5124,
+      "step": 12919
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.504089780884723,
+      "learning_rate": 5.11189330313932e-06,
+      "loss": 0.4682,
+      "step": 12920
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.077138459862923,
+      "learning_rate": 5.110422175227284e-06,
+      "loss": 0.5939,
+      "step": 12921
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.649965935713773,
+      "learning_rate": 5.1089511863719824e-06,
+      "loss": 0.4629,
+      "step": 12922
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 32.15808636562768,
+      "learning_rate": 5.10748033661525e-06,
+      "loss": 0.5891,
+      "step": 12923
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.038751400052202,
+      "learning_rate": 5.106009625998913e-06,
+      "loss": 0.4786,
+      "step": 12924
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 28.30107336689926,
+      "learning_rate": 5.104539054564799e-06,
+      "loss": 0.4997,
+      "step": 12925
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 15.10869506323825,
+      "learning_rate": 5.103068622354729e-06,
+      "loss": 0.5007,
+      "step": 12926
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 28.111912110840457,
+      "learning_rate": 5.101598329410522e-06,
+      "loss": 0.4682,
+      "step": 12927
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.111038363739276,
+      "learning_rate": 5.100128175773998e-06,
+      "loss": 0.4905,
+      "step": 12928
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.552109248155773,
+      "learning_rate": 5.098658161486962e-06,
+      "loss": 0.5043,
+      "step": 12929
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.00488729032167,
+      "learning_rate": 5.097188286591217e-06,
+      "loss": 0.4849,
+      "step": 12930
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 22.301912926835485,
+      "learning_rate": 5.095718551128569e-06,
+      "loss": 0.579,
+      "step": 12931
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.263912913502743,
+      "learning_rate": 5.09424895514082e-06,
+      "loss": 0.5646,
+      "step": 12932
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 26.040774268245876,
+      "learning_rate": 5.092779498669758e-06,
+      "loss": 0.4881,
+      "step": 12933
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 25.22300925885845,
+      "learning_rate": 5.091310181757178e-06,
+      "loss": 0.4967,
+      "step": 12934
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 15.216214411333613,
+      "learning_rate": 5.089841004444864e-06,
+      "loss": 0.5145,
+      "step": 12935
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 25.828643816694733,
+      "learning_rate": 5.088371966774601e-06,
+      "loss": 0.5384,
+      "step": 12936
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.97114113562809,
+      "learning_rate": 5.086903068788163e-06,
+      "loss": 0.5048,
+      "step": 12937
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.293208981455534,
+      "learning_rate": 5.085434310527329e-06,
+      "loss": 0.4853,
+      "step": 12938
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.845856726872723,
+      "learning_rate": 5.083965692033871e-06,
+      "loss": 0.4793,
+      "step": 12939
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 15.982750894963123,
+      "learning_rate": 5.0824972133495535e-06,
+      "loss": 0.5018,
+      "step": 12940
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 15.962981736401444,
+      "learning_rate": 5.081028874516134e-06,
+      "loss": 0.4511,
+      "step": 12941
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 25.991586146780836,
+      "learning_rate": 5.0795606755753745e-06,
+      "loss": 0.4761,
+      "step": 12942
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.59586374946476,
+      "learning_rate": 5.0780926165690325e-06,
+      "loss": 0.4913,
+      "step": 12943
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 26.261921909221112,
+      "learning_rate": 5.076624697538861e-06,
+      "loss": 0.5173,
+      "step": 12944
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 23.850943563273177,
+      "learning_rate": 5.075156918526604e-06,
+      "loss": 0.5424,
+      "step": 12945
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.629691340431602,
+      "learning_rate": 5.073689279573999e-06,
+      "loss": 0.4819,
+      "step": 12946
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.535175868433708,
+      "learning_rate": 5.072221780722788e-06,
+      "loss": 0.4638,
+      "step": 12947
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.420423019238626,
+      "learning_rate": 5.070754422014711e-06,
+      "loss": 0.4447,
+      "step": 12948
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 24.360977288989826,
+      "learning_rate": 5.069287203491495e-06,
+      "loss": 0.5058,
+      "step": 12949
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.359228791145526,
+      "learning_rate": 5.0678201251948625e-06,
+      "loss": 0.5314,
+      "step": 12950
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.21461601000422,
+      "learning_rate": 5.066353187166542e-06,
+      "loss": 0.5161,
+      "step": 12951
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 22.560840450553414,
+      "learning_rate": 5.064886389448249e-06,
+      "loss": 0.4922,
+      "step": 12952
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.521824713462816,
+      "learning_rate": 5.063419732081704e-06,
+      "loss": 0.5454,
+      "step": 12953
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 23.041056475886474,
+      "learning_rate": 5.061953215108614e-06,
+      "loss": 0.511,
+      "step": 12954
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.945096319738319,
+      "learning_rate": 5.060486838570682e-06,
+      "loss": 0.44,
+      "step": 12955
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.523696602369135,
+      "learning_rate": 5.059020602509616e-06,
+      "loss": 0.5487,
+      "step": 12956
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 26.189207472023572,
+      "learning_rate": 5.057554506967116e-06,
+      "loss": 0.5123,
+      "step": 12957
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.71969545792104,
+      "learning_rate": 5.056088551984876e-06,
+      "loss": 0.5108,
+      "step": 12958
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.007571547783737,
+      "learning_rate": 5.0546227376045796e-06,
+      "loss": 0.4888,
+      "step": 12959
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.61477990856951,
+      "learning_rate": 5.053157063867922e-06,
+      "loss": 0.5057,
+      "step": 12960
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.44929291894792,
+      "learning_rate": 5.051691530816586e-06,
+      "loss": 0.5329,
+      "step": 12961
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.669504545917782,
+      "learning_rate": 5.050226138492245e-06,
+      "loss": 0.4971,
+      "step": 12962
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.916932108291594,
+      "learning_rate": 5.048760886936581e-06,
+      "loss": 0.6495,
+      "step": 12963
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.93328422811248,
+      "learning_rate": 5.0472957761912565e-06,
+      "loss": 0.4565,
+      "step": 12964
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.554134242636305,
+      "learning_rate": 5.0458308062979466e-06,
+      "loss": 0.5064,
+      "step": 12965
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 31.569692142568133,
+      "learning_rate": 5.0443659772983064e-06,
+      "loss": 0.4578,
+      "step": 12966
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.86317067754007,
+      "learning_rate": 5.042901289234001e-06,
+      "loss": 0.4656,
+      "step": 12967
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.796203357704613,
+      "learning_rate": 5.041436742146685e-06,
+      "loss": 0.4735,
+      "step": 12968
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.234503629593902,
+      "learning_rate": 5.039972336078004e-06,
+      "loss": 0.5157,
+      "step": 12969
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.97618756199585,
+      "learning_rate": 5.038508071069612e-06,
+      "loss": 0.4808,
+      "step": 12970
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.701606671683642,
+      "learning_rate": 5.037043947163145e-06,
+      "loss": 0.5264,
+      "step": 12971
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.91798585272234,
+      "learning_rate": 5.035579964400244e-06,
+      "loss": 0.4856,
+      "step": 12972
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.52615643650113,
+      "learning_rate": 5.034116122822547e-06,
+      "loss": 0.4913,
+      "step": 12973
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.718765009371467,
+      "learning_rate": 5.032652422471685e-06,
+      "loss": 0.5045,
+      "step": 12974
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.632496651204306,
+      "learning_rate": 5.031188863389278e-06,
+      "loss": 0.5444,
+      "step": 12975
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.311159396900184,
+      "learning_rate": 5.029725445616953e-06,
+      "loss": 0.4897,
+      "step": 12976
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.556354788440725,
+      "learning_rate": 5.028262169196329e-06,
+      "loss": 0.4971,
+      "step": 12977
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.03218709035426,
+      "learning_rate": 5.0267990341690275e-06,
+      "loss": 0.5537,
+      "step": 12978
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.586155611199526,
+      "learning_rate": 5.025336040576645e-06,
+      "loss": 0.475,
+      "step": 12979
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.218267322505994,
+      "learning_rate": 5.023873188460795e-06,
+      "loss": 0.5585,
+      "step": 12980
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.621293800068443,
+      "learning_rate": 5.02241047786308e-06,
+      "loss": 0.5174,
+      "step": 12981
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 14.338429233924517,
+      "learning_rate": 5.020947908825104e-06,
+      "loss": 0.463,
+      "step": 12982
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.489757113012747,
+      "learning_rate": 5.019485481388457e-06,
+      "loss": 0.4849,
+      "step": 12983
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.381331134448654,
+      "learning_rate": 5.018023195594726e-06,
+      "loss": 0.4659,
+      "step": 12984
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.696198371658305,
+      "learning_rate": 5.0165610514855005e-06,
+      "loss": 0.545,
+      "step": 12985
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 27.280164352917183,
+      "learning_rate": 5.0150990491023655e-06,
+      "loss": 0.5153,
+      "step": 12986
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.186738514157582,
+      "learning_rate": 5.013637188486895e-06,
+      "loss": 0.4551,
+      "step": 12987
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.209177101321878,
+      "learning_rate": 5.0121754696806705e-06,
+      "loss": 0.5723,
+      "step": 12988
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.233382817101287,
+      "learning_rate": 5.010713892725253e-06,
+      "loss": 0.5307,
+      "step": 12989
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.51935531142266,
+      "learning_rate": 5.009252457662215e-06,
+      "loss": 0.528,
+      "step": 12990
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.229875616045803,
+      "learning_rate": 5.00779116453312e-06,
+      "loss": 0.5104,
+      "step": 12991
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 21.693085909591737,
+      "learning_rate": 5.00633001337952e-06,
+      "loss": 0.517,
+      "step": 12992
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 25.112386556125134,
+      "learning_rate": 5.004869004242978e-06,
+      "loss": 0.5073,
+      "step": 12993
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.9949028866268,
+      "learning_rate": 5.0034081371650355e-06,
+      "loss": 0.4546,
+      "step": 12994
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.344539489310822,
+      "learning_rate": 5.001947412187246e-06,
+      "loss": 0.5482,
+      "step": 12995
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.425448719903216,
+      "learning_rate": 5.000486829351145e-06,
+      "loss": 0.4497,
+      "step": 12996
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.94215794828907,
+      "learning_rate": 4.999026388698274e-06,
+      "loss": 0.4724,
+      "step": 12997
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.277678036389432,
+      "learning_rate": 4.9975660902701695e-06,
+      "loss": 0.5749,
+      "step": 12998
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.646998735564413,
+      "learning_rate": 4.996105934108361e-06,
+      "loss": 0.5397,
+      "step": 12999
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.888224960434503,
+      "learning_rate": 4.9946459202543675e-06,
+      "loss": 0.4837,
+      "step": 13000
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.404559647187043,
+      "learning_rate": 4.993186048749717e-06,
+      "loss": 0.4851,
+      "step": 13001
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.407126168639,
+      "learning_rate": 4.991726319635925e-06,
+      "loss": 0.5178,
+      "step": 13002
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.285891033021166,
+      "learning_rate": 4.990266732954515e-06,
+      "loss": 0.5457,
+      "step": 13003
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.235065248718087,
+      "learning_rate": 4.98880728874698e-06,
+      "loss": 0.5051,
+      "step": 13004
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.03958188909016,
+      "learning_rate": 4.987347987054835e-06,
+      "loss": 0.4714,
+      "step": 13005
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.90473123471736,
+      "learning_rate": 4.9858888279195815e-06,
+      "loss": 0.4831,
+      "step": 13006
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.53985441781149,
+      "learning_rate": 4.984429811382721e-06,
+      "loss": 0.468,
+      "step": 13007
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.609928158114585,
+      "learning_rate": 4.982970937485741e-06,
+      "loss": 0.5079,
+      "step": 13008
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.073784736051554,
+      "learning_rate": 4.981512206270129e-06,
+      "loss": 0.4932,
+      "step": 13009
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 22.075368560333942,
+      "learning_rate": 4.980053617777375e-06,
+      "loss": 0.4835,
+      "step": 13010
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.64095544623726,
+      "learning_rate": 4.978595172048963e-06,
+      "loss": 0.5551,
+      "step": 13011
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.01881777896616,
+      "learning_rate": 4.977136869126366e-06,
+      "loss": 0.4885,
+      "step": 13012
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.801097233917018,
+      "learning_rate": 4.975678709051056e-06,
+      "loss": 0.4988,
+      "step": 13013
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 22.741799172034856,
+      "learning_rate": 4.974220691864503e-06,
+      "loss": 0.4692,
+      "step": 13014
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.919911083698452,
+      "learning_rate": 4.972762817608174e-06,
+      "loss": 0.4664,
+      "step": 13015
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 29.579764216190913,
+      "learning_rate": 4.971305086323534e-06,
+      "loss": 0.5139,
+      "step": 13016
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.2319391323564,
+      "learning_rate": 4.969847498052035e-06,
+      "loss": 0.474,
+      "step": 13017
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.199246220057237,
+      "learning_rate": 4.968390052835127e-06,
+      "loss": 0.5309,
+      "step": 13018
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.906873815366412,
+      "learning_rate": 4.966932750714262e-06,
+      "loss": 0.4528,
+      "step": 13019
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.783104280692573,
+      "learning_rate": 4.96547559173089e-06,
+      "loss": 0.5059,
+      "step": 13020
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.0585573699973,
+      "learning_rate": 4.964018575926442e-06,
+      "loss": 0.4826,
+      "step": 13021
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.384747476709627,
+      "learning_rate": 4.962561703342364e-06,
+      "loss": 0.4838,
+      "step": 13022
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.105715717920873,
+      "learning_rate": 4.961104974020081e-06,
+      "loss": 0.5308,
+      "step": 13023
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.207606773542302,
+      "learning_rate": 4.959648388001028e-06,
+      "loss": 0.4892,
+      "step": 13024
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.095427183483118,
+      "learning_rate": 4.958191945326624e-06,
+      "loss": 0.4947,
+      "step": 13025
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 25.755325375488038,
+      "learning_rate": 4.9567356460382895e-06,
+      "loss": 0.4677,
+      "step": 13026
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.235071031858206,
+      "learning_rate": 4.955279490177448e-06,
+      "loss": 0.515,
+      "step": 13027
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.819117463320943,
+      "learning_rate": 4.953823477785507e-06,
+      "loss": 0.4928,
+      "step": 13028
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.049638209353752,
+      "learning_rate": 4.952367608903871e-06,
+      "loss": 0.4752,
+      "step": 13029
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 20.620339938601706,
+      "learning_rate": 4.9509118835739465e-06,
+      "loss": 0.5261,
+      "step": 13030
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 20.11174514781223,
+      "learning_rate": 4.949456301837134e-06,
+      "loss": 0.4821,
+      "step": 13031
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.63069351136901,
+      "learning_rate": 4.948000863734834e-06,
+      "loss": 0.4823,
+      "step": 13032
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.461578355081606,
+      "learning_rate": 4.946545569308436e-06,
+      "loss": 0.4603,
+      "step": 13033
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.697868419755885,
+      "learning_rate": 4.94509041859932e-06,
+      "loss": 0.4841,
+      "step": 13034
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 23.784272872751654,
+      "learning_rate": 4.943635411648875e-06,
+      "loss": 0.5846,
+      "step": 13035
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 23.470043085795815,
+      "learning_rate": 4.942180548498485e-06,
+      "loss": 0.4584,
+      "step": 13036
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.63525237817207,
+      "learning_rate": 4.940725829189523e-06,
+      "loss": 0.5578,
+      "step": 13037
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 21.616966067201265,
+      "learning_rate": 4.939271253763354e-06,
+      "loss": 0.4953,
+      "step": 13038
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.966022760732278,
+      "learning_rate": 4.93781682226135e-06,
+      "loss": 0.4928,
+      "step": 13039
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.258842047480677,
+      "learning_rate": 4.936362534724874e-06,
+      "loss": 0.4683,
+      "step": 13040
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.542794078296232,
+      "learning_rate": 4.93490839119529e-06,
+      "loss": 0.5421,
+      "step": 13041
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.147552542113733,
+      "learning_rate": 4.933454391713948e-06,
+      "loss": 0.5064,
+      "step": 13042
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 20.626187050500658,
+      "learning_rate": 4.932000536322196e-06,
+      "loss": 0.4845,
+      "step": 13043
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 21.654846846411715,
+      "learning_rate": 4.930546825061383e-06,
+      "loss": 0.4447,
+      "step": 13044
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.889471016107183,
+      "learning_rate": 4.929093257972857e-06,
+      "loss": 0.425,
+      "step": 13045
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.998922219144394,
+      "learning_rate": 4.927639835097948e-06,
+      "loss": 0.617,
+      "step": 13046
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 43.946394668478995,
+      "learning_rate": 4.926186556478e-06,
+      "loss": 0.5313,
+      "step": 13047
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.354368619718787,
+      "learning_rate": 4.924733422154333e-06,
+      "loss": 0.4354,
+      "step": 13048
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.574407637122906,
+      "learning_rate": 4.9232804321682835e-06,
+      "loss": 0.6144,
+      "step": 13049
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.508368702453623,
+      "learning_rate": 4.921827586561164e-06,
+      "loss": 0.5333,
+      "step": 13050
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.70410935391592,
+      "learning_rate": 4.9203748853742986e-06,
+      "loss": 0.4878,
+      "step": 13051
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 15.846075923695057,
+      "learning_rate": 4.918922328649003e-06,
+      "loss": 0.5222,
+      "step": 13052
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 25.70642182532485,
+      "learning_rate": 4.917469916426584e-06,
+      "loss": 0.4936,
+      "step": 13053
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 29.422650817755528,
+      "learning_rate": 4.916017648748342e-06,
+      "loss": 0.4421,
+      "step": 13054
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.939127934634389,
+      "learning_rate": 4.914565525655587e-06,
+      "loss": 0.46,
+      "step": 13055
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.252299034270948,
+      "learning_rate": 4.913113547189615e-06,
+      "loss": 0.4824,
+      "step": 13056
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.09355082792873,
+      "learning_rate": 4.911661713391716e-06,
+      "loss": 0.4894,
+      "step": 13057
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.789493086642114,
+      "learning_rate": 4.9102100243031835e-06,
+      "loss": 0.4764,
+      "step": 13058
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.479533381129174,
+      "learning_rate": 4.9087584799652976e-06,
+      "loss": 0.4745,
+      "step": 13059
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.950148162202517,
+      "learning_rate": 4.907307080419341e-06,
+      "loss": 0.4781,
+      "step": 13060
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 21.722010297797063,
+      "learning_rate": 4.905855825706597e-06,
+      "loss": 0.4307,
+      "step": 13061
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.026886453767627,
+      "learning_rate": 4.904404715868333e-06,
+      "loss": 0.4888,
+      "step": 13062
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.4081582668416,
+      "learning_rate": 4.902953750945814e-06,
+      "loss": 0.5287,
+      "step": 13063
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.495950213857196,
+      "learning_rate": 4.90150293098031e-06,
+      "loss": 0.482,
+      "step": 13064
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.550060380763576,
+      "learning_rate": 4.900052256013078e-06,
+      "loss": 0.4732,
+      "step": 13065
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.514216969149448,
+      "learning_rate": 4.898601726085385e-06,
+      "loss": 0.5483,
+      "step": 13066
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 25.881271781050895,
+      "learning_rate": 4.897151341238468e-06,
+      "loss": 0.5618,
+      "step": 13067
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.2862616224728,
+      "learning_rate": 4.89570110151358e-06,
+      "loss": 0.5062,
+      "step": 13068
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.636930824770552,
+      "learning_rate": 4.894251006951968e-06,
+      "loss": 0.4818,
+      "step": 13069
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 39.210850895103675,
+      "learning_rate": 4.892801057594874e-06,
+      "loss": 0.5468,
+      "step": 13070
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.21575051696606,
+      "learning_rate": 4.891351253483529e-06,
+      "loss": 0.5224,
+      "step": 13071
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.857765260387817,
+      "learning_rate": 4.8899015946591645e-06,
+      "loss": 0.4473,
+      "step": 13072
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 21.41204916929831,
+      "learning_rate": 4.888452081163007e-06,
+      "loss": 0.6123,
+      "step": 13073
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.0952821628801,
+      "learning_rate": 4.8870027130362885e-06,
+      "loss": 0.4573,
+      "step": 13074
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.262400707975143,
+      "learning_rate": 4.8855534903202165e-06,
+      "loss": 0.523,
+      "step": 13075
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 27.61781785777043,
+      "learning_rate": 4.884104413056015e-06,
+      "loss": 0.5429,
+      "step": 13076
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 29.266838248685563,
+      "learning_rate": 4.882655481284888e-06,
+      "loss": 0.5581,
+      "step": 13077
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.540519602827015,
+      "learning_rate": 4.881206695048046e-06,
+      "loss": 0.5006,
+      "step": 13078
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.898472919000717,
+      "learning_rate": 4.879758054386694e-06,
+      "loss": 0.4967,
+      "step": 13079
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.996039434218446,
+      "learning_rate": 4.8783095593420234e-06,
+      "loss": 0.6114,
+      "step": 13080
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.872649070934497,
+      "learning_rate": 4.876861209955238e-06,
+      "loss": 0.4357,
+      "step": 13081
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.342807587975297,
+      "learning_rate": 4.875413006267518e-06,
+      "loss": 0.4837,
+      "step": 13082
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.1471705149605,
+      "learning_rate": 4.873964948320059e-06,
+      "loss": 0.5326,
+      "step": 13083
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.54139311762064,
+      "learning_rate": 4.8725170361540345e-06,
+      "loss": 0.5116,
+      "step": 13084
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.049095065687027,
+      "learning_rate": 4.871069269810626e-06,
+      "loss": 0.4685,
+      "step": 13085
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.9775664480053,
+      "learning_rate": 4.869621649331009e-06,
+      "loss": 0.5759,
+      "step": 13086
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.462539014305552,
+      "learning_rate": 4.868174174756353e-06,
+      "loss": 0.5591,
+      "step": 13087
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.31731374249837,
+      "learning_rate": 4.866726846127817e-06,
+      "loss": 0.4638,
+      "step": 13088
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.251425048652923,
+      "learning_rate": 4.865279663486567e-06,
+      "loss": 0.5928,
+      "step": 13089
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.183029788047318,
+      "learning_rate": 4.863832626873759e-06,
+      "loss": 0.5242,
+      "step": 13090
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.660302493324785,
+      "learning_rate": 4.862385736330555e-06,
+      "loss": 0.5345,
+      "step": 13091
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.32070945925332,
+      "learning_rate": 4.860938991898088e-06,
+      "loss": 0.5238,
+      "step": 13092
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.224183372765292,
+      "learning_rate": 4.859492393617509e-06,
+      "loss": 0.4141,
+      "step": 13093
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.471059750200975,
+      "learning_rate": 4.8580459415299585e-06,
+      "loss": 0.6052,
+      "step": 13094
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.971992064574472,
+      "learning_rate": 4.856599635676578e-06,
+      "loss": 0.5123,
+      "step": 13095
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.373945844321241,
+      "learning_rate": 4.8551534760984954e-06,
+      "loss": 0.4766,
+      "step": 13096
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.025852800447048,
+      "learning_rate": 4.853707462836834e-06,
+      "loss": 0.4895,
+      "step": 13097
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.710918251264056,
+      "learning_rate": 4.852261595932721e-06,
+      "loss": 0.5298,
+      "step": 13098
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.674850255173887,
+      "learning_rate": 4.850815875427283e-06,
+      "loss": 0.4622,
+      "step": 13099
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.87107094045906,
+      "learning_rate": 4.849370301361623e-06,
+      "loss": 0.5502,
+      "step": 13100
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.496457546868037,
+      "learning_rate": 4.8479248737768644e-06,
+      "loss": 0.4528,
+      "step": 13101
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.58280930364715,
+      "learning_rate": 4.846479592714104e-06,
+      "loss": 0.4763,
+      "step": 13102
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.63788096808724,
+      "learning_rate": 4.845034458214449e-06,
+      "loss": 0.4793,
+      "step": 13103
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.55182021770802,
+      "learning_rate": 4.8435894703190026e-06,
+      "loss": 0.4668,
+      "step": 13104
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.505994733864373,
+      "learning_rate": 4.842144629068854e-06,
+      "loss": 0.476,
+      "step": 13105
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.48734237116329,
+      "learning_rate": 4.8406999345050934e-06,
+      "loss": 0.5079,
+      "step": 13106
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.49507249221655,
+      "learning_rate": 4.839255386668806e-06,
+      "loss": 0.4895,
+      "step": 13107
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.317551708540375,
+      "learning_rate": 4.8378109856010825e-06,
+      "loss": 0.5622,
+      "step": 13108
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.86885238236193,
+      "learning_rate": 4.8363667313429896e-06,
+      "loss": 0.5886,
+      "step": 13109
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.538093366415712,
+      "learning_rate": 4.83492262393561e-06,
+      "loss": 0.4805,
+      "step": 13110
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.875291456058891,
+      "learning_rate": 4.833478663420007e-06,
+      "loss": 0.5045,
+      "step": 13111
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.29727931910288,
+      "learning_rate": 4.832034849837252e-06,
+      "loss": 0.5028,
+      "step": 13112
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.391570617988279,
+      "learning_rate": 4.830591183228398e-06,
+      "loss": 0.5129,
+      "step": 13113
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.781565945849648,
+      "learning_rate": 4.829147663634507e-06,
+      "loss": 0.475,
+      "step": 13114
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 21.271750278507955,
+      "learning_rate": 4.827704291096635e-06,
+      "loss": 0.5685,
+      "step": 13115
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.570186180615327,
+      "learning_rate": 4.826261065655828e-06,
+      "loss": 0.5113,
+      "step": 13116
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 24.171336103891505,
+      "learning_rate": 4.824817987353126e-06,
+      "loss": 0.4951,
+      "step": 13117
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 25.022904260856897,
+      "learning_rate": 4.823375056229573e-06,
+      "loss": 0.4439,
+      "step": 13118
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.6026222659298,
+      "learning_rate": 4.821932272326204e-06,
+      "loss": 0.445,
+      "step": 13119
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 21.25628924003059,
+      "learning_rate": 4.820489635684057e-06,
+      "loss": 0.5066,
+      "step": 13120
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.04497601537879,
+      "learning_rate": 4.819047146344154e-06,
+      "loss": 0.4717,
+      "step": 13121
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.909828417992323,
+      "learning_rate": 4.817604804347517e-06,
+      "loss": 0.4742,
+      "step": 13122
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.844475594447097,
+      "learning_rate": 4.816162609735167e-06,
+      "loss": 0.5907,
+      "step": 13123
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 28.258651678751484,
+      "learning_rate": 4.814720562548123e-06,
+      "loss": 0.5051,
+      "step": 13124
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 29.189905794985034,
+      "learning_rate": 4.8132786628273945e-06,
+      "loss": 0.4528,
+      "step": 13125
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.55679207420983,
+      "learning_rate": 4.811836910613982e-06,
+      "loss": 0.4475,
+      "step": 13126
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.25663783067621,
+      "learning_rate": 4.810395305948892e-06,
+      "loss": 0.4709,
+      "step": 13127
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.461609445830266,
+      "learning_rate": 4.8089538488731244e-06,
+      "loss": 0.4838,
+      "step": 13128
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.427072531874575,
+      "learning_rate": 4.807512539427677e-06,
+      "loss": 0.514,
+      "step": 13129
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.82424251842245,
+      "learning_rate": 4.806071377653534e-06,
+      "loss": 0.4567,
+      "step": 13130
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.610791011663075,
+      "learning_rate": 4.80463036359168e-06,
+      "loss": 0.5322,
+      "step": 13131
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.69776457447809,
+      "learning_rate": 4.803189497283099e-06,
+      "loss": 0.4828,
+      "step": 13132
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.26149548849849,
+      "learning_rate": 4.801748778768772e-06,
+      "loss": 0.5067,
+      "step": 13133
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.147832899069844,
+      "learning_rate": 4.800308208089666e-06,
+      "loss": 0.4711,
+      "step": 13134
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.39570157316439,
+      "learning_rate": 4.798867785286756e-06,
+      "loss": 0.4451,
+      "step": 13135
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.326819546857475,
+      "learning_rate": 4.797427510401001e-06,
+      "loss": 0.4599,
+      "step": 13136
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.71725273558126,
+      "learning_rate": 4.795987383473366e-06,
+      "loss": 0.53,
+      "step": 13137
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.39146658252815,
+      "learning_rate": 4.794547404544802e-06,
+      "loss": 0.5757,
+      "step": 13138
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.27019751220681,
+      "learning_rate": 4.793107573656266e-06,
+      "loss": 0.4434,
+      "step": 13139
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.784911968176218,
+      "learning_rate": 4.791667890848708e-06,
+      "loss": 0.5118,
+      "step": 13140
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.650569769826888,
+      "learning_rate": 4.790228356163065e-06,
+      "loss": 0.496,
+      "step": 13141
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.366631113469268,
+      "learning_rate": 4.788788969640283e-06,
+      "loss": 0.5115,
+      "step": 13142
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.826351075873383,
+      "learning_rate": 4.787349731321291e-06,
+      "loss": 0.4305,
+      "step": 13143
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.161040994357972,
+      "learning_rate": 4.785910641247022e-06,
+      "loss": 0.5053,
+      "step": 13144
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.033952075636236,
+      "learning_rate": 4.78447169945841e-06,
+      "loss": 0.511,
+      "step": 13145
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.50736386909542,
+      "learning_rate": 4.78303290599637e-06,
+      "loss": 0.5146,
+      "step": 13146
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 29.35894583936064,
+      "learning_rate": 4.78159426090182e-06,
+      "loss": 0.5213,
+      "step": 13147
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.52666531820444,
+      "learning_rate": 4.780155764215676e-06,
+      "loss": 0.4146,
+      "step": 13148
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 24.906601628725465,
+      "learning_rate": 4.7787174159788514e-06,
+      "loss": 0.4949,
+      "step": 13149
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.291174776819464,
+      "learning_rate": 4.7772792162322505e-06,
+      "loss": 0.4918,
+      "step": 13150
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.13246260432585,
+      "learning_rate": 4.775841165016769e-06,
+      "loss": 0.4286,
+      "step": 13151
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.380334995204837,
+      "learning_rate": 4.774403262373308e-06,
+      "loss": 0.4636,
+      "step": 13152
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 24.10448252085768,
+      "learning_rate": 4.772965508342763e-06,
+      "loss": 0.4568,
+      "step": 13153
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 25.580935339550056,
+      "learning_rate": 4.771527902966028e-06,
+      "loss": 0.4981,
+      "step": 13154
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.631121422976026,
+      "learning_rate": 4.7700904462839735e-06,
+      "loss": 0.4566,
+      "step": 13155
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 21.537421263248078,
+      "learning_rate": 4.7686531383374865e-06,
+      "loss": 0.5223,
+      "step": 13156
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 24.658410671153835,
+      "learning_rate": 4.767215979167445e-06,
+      "loss": 0.4918,
+      "step": 13157
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.134114788897476,
+      "learning_rate": 4.765778968814724e-06,
+      "loss": 0.4591,
+      "step": 13158
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 25.363092060755346,
+      "learning_rate": 4.764342107320186e-06,
+      "loss": 0.5202,
+      "step": 13159
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 22.8174915006642,
+      "learning_rate": 4.762905394724691e-06,
+      "loss": 0.4964,
+      "step": 13160
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 29.040706876349738,
+      "learning_rate": 4.761468831069106e-06,
+      "loss": 0.5863,
+      "step": 13161
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.987952909326758,
+      "learning_rate": 4.7600324163942845e-06,
+      "loss": 0.4474,
+      "step": 13162
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 24.6635138154888,
+      "learning_rate": 4.758596150741073e-06,
+      "loss": 0.4867,
+      "step": 13163
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.164974662691726,
+      "learning_rate": 4.757160034150324e-06,
+      "loss": 0.4996,
+      "step": 13164
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 28.721278538236707,
+      "learning_rate": 4.7557240666628735e-06,
+      "loss": 0.5115,
+      "step": 13165
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 22.04895459389824,
+      "learning_rate": 4.754288248319563e-06,
+      "loss": 0.4863,
+      "step": 13166
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 23.734362158354347,
+      "learning_rate": 4.752852579161229e-06,
+      "loss": 0.5033,
+      "step": 13167
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.837193166256506,
+      "learning_rate": 4.751417059228695e-06,
+      "loss": 0.5426,
+      "step": 13168
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.734374005602717,
+      "learning_rate": 4.749981688562792e-06,
+      "loss": 0.4455,
+      "step": 13169
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.256447441515185,
+      "learning_rate": 4.748546467204336e-06,
+      "loss": 0.4139,
+      "step": 13170
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.689209340064366,
+      "learning_rate": 4.747111395194149e-06,
+      "loss": 0.472,
+      "step": 13171
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.07914701686187,
+      "learning_rate": 4.7456764725730385e-06,
+      "loss": 0.474,
+      "step": 13172
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.888771738598066,
+      "learning_rate": 4.744241699381816e-06,
+      "loss": 0.4863,
+      "step": 13173
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.261870969048253,
+      "learning_rate": 4.742807075661288e-06,
+      "loss": 0.5512,
+      "step": 13174
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.238974395861504,
+      "learning_rate": 4.741372601452251e-06,
+      "loss": 0.52,
+      "step": 13175
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.092522108737782,
+      "learning_rate": 4.739938276795498e-06,
+      "loss": 0.5036,
+      "step": 13176
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.593687492845298,
+      "learning_rate": 4.738504101731824e-06,
+      "loss": 0.4155,
+      "step": 13177
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 21.249497735875668,
+      "learning_rate": 4.737070076302014e-06,
+      "loss": 0.4787,
+      "step": 13178
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.961678617833368,
+      "learning_rate": 4.735636200546859e-06,
+      "loss": 0.4691,
+      "step": 13179
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.321922905751014,
+      "learning_rate": 4.734202474507126e-06,
+      "loss": 0.5002,
+      "step": 13180
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 28.150140581110264,
+      "learning_rate": 4.7327688982235916e-06,
+      "loss": 0.4799,
+      "step": 13181
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 25.732883946361678,
+      "learning_rate": 4.731335471737031e-06,
+      "loss": 0.525,
+      "step": 13182
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.31312165323388,
+      "learning_rate": 4.72990219508821e-06,
+      "loss": 0.5178,
+      "step": 13183
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.750640312960908,
+      "learning_rate": 4.728469068317887e-06,
+      "loss": 0.5771,
+      "step": 13184
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 20.267775892763424,
+      "learning_rate": 4.727036091466816e-06,
+      "loss": 0.5371,
+      "step": 13185
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.92896450978873,
+      "learning_rate": 4.7256032645757555e-06,
+      "loss": 0.4909,
+      "step": 13186
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.03684667372953,
+      "learning_rate": 4.724170587685456e-06,
+      "loss": 0.496,
+      "step": 13187
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 25.026947929011325,
+      "learning_rate": 4.722738060836654e-06,
+      "loss": 0.6308,
+      "step": 13188
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.34865867297164,
+      "learning_rate": 4.7213056840701e-06,
+      "loss": 0.4742,
+      "step": 13189
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 25.45888529978834,
+      "learning_rate": 4.719873457426519e-06,
+      "loss": 0.5104,
+      "step": 13190
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.4916437604768,
+      "learning_rate": 4.718441380946649e-06,
+      "loss": 0.448,
+      "step": 13191
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.640372947789828,
+      "learning_rate": 4.717009454671221e-06,
+      "loss": 0.4748,
+      "step": 13192
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.926840234874444,
+      "learning_rate": 4.715577678640948e-06,
+      "loss": 0.5192,
+      "step": 13193
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.48118511662771,
+      "learning_rate": 4.714146052896559e-06,
+      "loss": 0.4897,
+      "step": 13194
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.640480900530083,
+      "learning_rate": 4.7127145774787604e-06,
+      "loss": 0.5584,
+      "step": 13195
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 31.634562008018673,
+      "learning_rate": 4.71128325242827e-06,
+      "loss": 0.5316,
+      "step": 13196
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.578033159314309,
+      "learning_rate": 4.709852077785787e-06,
+      "loss": 0.5237,
+      "step": 13197
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 25.785091844179796,
+      "learning_rate": 4.708421053592019e-06,
+      "loss": 0.4978,
+      "step": 13198
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 22.001608689512103,
+      "learning_rate": 4.706990179887658e-06,
+      "loss": 0.4839,
+      "step": 13199
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.638333979575602,
+      "learning_rate": 4.705559456713403e-06,
+      "loss": 0.456,
+      "step": 13200
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 23.374595600342712,
+      "learning_rate": 4.704128884109936e-06,
+      "loss": 0.5992,
+      "step": 13201
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.259441866014114,
+      "learning_rate": 4.702698462117946e-06,
+      "loss": 0.5045,
+      "step": 13202
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.117531366897676,
+      "learning_rate": 4.701268190778117e-06,
+      "loss": 0.4629,
+      "step": 13203
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.42074207069252,
+      "learning_rate": 4.6998380701311196e-06,
+      "loss": 0.4112,
+      "step": 13204
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.662762698111324,
+      "learning_rate": 4.6984081002176256e-06,
+      "loss": 0.484,
+      "step": 13205
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.72600408840137,
+      "learning_rate": 4.696978281078302e-06,
+      "loss": 0.4784,
+      "step": 13206
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.682652369952763,
+      "learning_rate": 4.695548612753814e-06,
+      "loss": 0.5063,
+      "step": 13207
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 20.305321777816935,
+      "learning_rate": 4.694119095284825e-06,
+      "loss": 0.5034,
+      "step": 13208
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 21.726412949223977,
+      "learning_rate": 4.6926897287119845e-06,
+      "loss": 0.5166,
+      "step": 13209
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.03169614977865,
+      "learning_rate": 4.6912605130759396e-06,
+      "loss": 0.4976,
+      "step": 13210
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 22.463874376767897,
+      "learning_rate": 4.68983144841734e-06,
+      "loss": 0.4748,
+      "step": 13211
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 18.00008583131168,
+      "learning_rate": 4.688402534776832e-06,
+      "loss": 0.5078,
+      "step": 13212
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.768752517691908,
+      "learning_rate": 4.686973772195048e-06,
+      "loss": 0.4953,
+      "step": 13213
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.489240478831842,
+      "learning_rate": 4.685545160712619e-06,
+      "loss": 0.4815,
+      "step": 13214
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 22.387293940516482,
+      "learning_rate": 4.684116700370176e-06,
+      "loss": 0.4804,
+      "step": 13215
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.5873130809419,
+      "learning_rate": 4.682688391208345e-06,
+      "loss": 0.4599,
+      "step": 13216
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.849759722572045,
+      "learning_rate": 4.681260233267749e-06,
+      "loss": 0.4729,
+      "step": 13217
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 32.49318304002005,
+      "learning_rate": 4.679832226589001e-06,
+      "loss": 0.5261,
+      "step": 13218
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.763241618638316,
+      "learning_rate": 4.6784043712127084e-06,
+      "loss": 0.4391,
+      "step": 13219
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 21.010813152628668,
+      "learning_rate": 4.676976667179482e-06,
+      "loss": 0.4879,
+      "step": 13220
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.51285047854393,
+      "learning_rate": 4.675549114529929e-06,
+      "loss": 0.4461,
+      "step": 13221
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.272838068293154,
+      "learning_rate": 4.6741217133046415e-06,
+      "loss": 0.4978,
+      "step": 13222
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.640360671110585,
+      "learning_rate": 4.67269446354422e-06,
+      "loss": 0.4897,
+      "step": 13223
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.475874627240326,
+      "learning_rate": 4.671267365289247e-06,
+      "loss": 0.5642,
+      "step": 13224
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.514364817556473,
+      "learning_rate": 4.669840418580318e-06,
+      "loss": 0.5352,
+      "step": 13225
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.470593647954168,
+      "learning_rate": 4.668413623458006e-06,
+      "loss": 0.5017,
+      "step": 13226
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.7617868985732,
+      "learning_rate": 4.666986979962891e-06,
+      "loss": 0.4471,
+      "step": 13227
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 23.45572303852927,
+      "learning_rate": 4.6655604881355495e-06,
+      "loss": 0.5002,
+      "step": 13228
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 20.608994056732502,
+      "learning_rate": 4.664134148016545e-06,
+      "loss": 0.5192,
+      "step": 13229
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 42.26989431695043,
+      "learning_rate": 4.662707959646446e-06,
+      "loss": 0.53,
+      "step": 13230
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.433607868562383,
+      "learning_rate": 4.661281923065808e-06,
+      "loss": 0.5339,
+      "step": 13231
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.983701504211606,
+      "learning_rate": 4.659856038315188e-06,
+      "loss": 0.5411,
+      "step": 13232
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.651993867294077,
+      "learning_rate": 4.658430305435143e-06,
+      "loss": 0.5539,
+      "step": 13233
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.453543533221193,
+      "learning_rate": 4.657004724466216e-06,
+      "loss": 0.4818,
+      "step": 13234
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 20.137962185257834,
+      "learning_rate": 4.655579295448944e-06,
+      "loss": 0.4841,
+      "step": 13235
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.48465977418263,
+      "learning_rate": 4.654154018423871e-06,
+      "loss": 0.4693,
+      "step": 13236
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.99336340332264,
+      "learning_rate": 4.652728893431534e-06,
+      "loss": 0.5159,
+      "step": 13237
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.420228762644086,
+      "learning_rate": 4.651303920512459e-06,
+      "loss": 0.5163,
+      "step": 13238
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 27.681051048955716,
+      "learning_rate": 4.649879099707168e-06,
+      "loss": 0.4894,
+      "step": 13239
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.828555037257193,
+      "learning_rate": 4.648454431056186e-06,
+      "loss": 0.5398,
+      "step": 13240
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.027205134078546,
+      "learning_rate": 4.64702991460003e-06,
+      "loss": 0.4548,
+      "step": 13241
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 26.730373530683064,
+      "learning_rate": 4.645605550379214e-06,
+      "loss": 0.5444,
+      "step": 13242
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.670789631336856,
+      "learning_rate": 4.644181338434245e-06,
+      "loss": 0.4692,
+      "step": 13243
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.388460423661712,
+      "learning_rate": 4.642757278805622e-06,
+      "loss": 0.4848,
+      "step": 13244
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.01839281634388,
+      "learning_rate": 4.641333371533848e-06,
+      "loss": 0.4741,
+      "step": 13245
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.859873675000927,
+      "learning_rate": 4.6399096166594215e-06,
+      "loss": 0.565,
+      "step": 13246
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 20.440530171019105,
+      "learning_rate": 4.638486014222831e-06,
+      "loss": 0.558,
+      "step": 13247
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.89752531235786,
+      "learning_rate": 4.6370625642645565e-06,
+      "loss": 0.5609,
+      "step": 13248
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.214453554088877,
+      "learning_rate": 4.635639266825086e-06,
+      "loss": 0.4755,
+      "step": 13249
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.558230368107555,
+      "learning_rate": 4.634216121944901e-06,
+      "loss": 0.4882,
+      "step": 13250
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.937022105566925,
+      "learning_rate": 4.632793129664466e-06,
+      "loss": 0.5074,
+      "step": 13251
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.170384326781141,
+      "learning_rate": 4.63137029002426e-06,
+      "loss": 0.4778,
+      "step": 13252
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.679424250309843,
+      "learning_rate": 4.629947603064737e-06,
+      "loss": 0.4871,
+      "step": 13253
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.605160811761372,
+      "learning_rate": 4.6285250688263625e-06,
+      "loss": 0.478,
+      "step": 13254
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.4407472684658,
+      "learning_rate": 4.627102687349599e-06,
+      "loss": 0.4601,
+      "step": 13255
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.16531912004903,
+      "learning_rate": 4.625680458674886e-06,
+      "loss": 0.4667,
+      "step": 13256
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.87519743062133,
+      "learning_rate": 4.624258382842681e-06,
+      "loss": 0.5047,
+      "step": 13257
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.26658181477529,
+      "learning_rate": 4.622836459893419e-06,
+      "loss": 0.5049,
+      "step": 13258
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.944766816321177,
+      "learning_rate": 4.621414689867547e-06,
+      "loss": 0.5219,
+      "step": 13259
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.991635752752213,
+      "learning_rate": 4.619993072805491e-06,
+      "loss": 0.4795,
+      "step": 13260
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.84476442191785,
+      "learning_rate": 4.618571608747685e-06,
+      "loss": 0.5565,
+      "step": 13261
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.35563997458557,
+      "learning_rate": 4.617150297734557e-06,
+      "loss": 0.5183,
+      "step": 13262
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.05433216185614,
+      "learning_rate": 4.615729139806527e-06,
+      "loss": 0.5606,
+      "step": 13263
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.802504232805475,
+      "learning_rate": 4.614308135004006e-06,
+      "loss": 0.4718,
+      "step": 13264
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 23.50101327740678,
+      "learning_rate": 4.612887283367411e-06,
+      "loss": 0.5271,
+      "step": 13265
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.01612142358905,
+      "learning_rate": 4.611466584937151e-06,
+      "loss": 0.515,
+      "step": 13266
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.621576061979578,
+      "learning_rate": 4.610046039753636e-06,
+      "loss": 0.5371,
+      "step": 13267
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.89313635461756,
+      "learning_rate": 4.608625647857251e-06,
+      "loss": 0.4967,
+      "step": 13268
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.527823419300745,
+      "learning_rate": 4.6072054092884e-06,
+      "loss": 0.5123,
+      "step": 13269
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.466714567903377,
+      "learning_rate": 4.60578532408747e-06,
+      "loss": 0.4647,
+      "step": 13270
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 28.490170040516933,
+      "learning_rate": 4.604365392294856e-06,
+      "loss": 0.5206,
+      "step": 13271
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.953122054179286,
+      "learning_rate": 4.602945613950933e-06,
+      "loss": 0.5469,
+      "step": 13272
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 23.444795847522744,
+      "learning_rate": 4.6015259890960765e-06,
+      "loss": 0.4765,
+      "step": 13273
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 33.2344108369721,
+      "learning_rate": 4.600106517770662e-06,
+      "loss": 0.5262,
+      "step": 13274
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 25.078606215283795,
+      "learning_rate": 4.598687200015063e-06,
+      "loss": 0.5199,
+      "step": 13275
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.847109196383375,
+      "learning_rate": 4.597268035869636e-06,
+      "loss": 0.4408,
+      "step": 13276
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.273272357843112,
+      "learning_rate": 4.59584902537475e-06,
+      "loss": 0.48,
+      "step": 13277
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 20.891020378788014,
+      "learning_rate": 4.594430168570753e-06,
+      "loss": 0.5481,
+      "step": 13278
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 35.46058444922919,
+      "learning_rate": 4.593011465498e-06,
+      "loss": 0.5408,
+      "step": 13279
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.44592376198636,
+      "learning_rate": 4.591592916196841e-06,
+      "loss": 0.5321,
+      "step": 13280
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.537570353016452,
+      "learning_rate": 4.590174520707612e-06,
+      "loss": 0.4878,
+      "step": 13281
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.821963766807723,
+      "learning_rate": 4.58875627907066e-06,
+      "loss": 0.4343,
+      "step": 13282
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.44724914384452,
+      "learning_rate": 4.587338191326308e-06,
+      "loss": 0.449,
+      "step": 13283
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 30.103092306209025,
+      "learning_rate": 4.585920257514897e-06,
+      "loss": 0.4938,
+      "step": 13284
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.819799432641574,
+      "learning_rate": 4.584502477676742e-06,
+      "loss": 0.5244,
+      "step": 13285
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.27474386383267,
+      "learning_rate": 4.583084851852169e-06,
+      "loss": 0.5628,
+      "step": 13286
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 22.788345147770126,
+      "learning_rate": 4.581667380081497e-06,
+      "loss": 0.4706,
+      "step": 13287
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.6980046119383,
+      "learning_rate": 4.580250062405036e-06,
+      "loss": 0.4697,
+      "step": 13288
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.228160775360433,
+      "learning_rate": 4.578832898863088e-06,
+      "loss": 0.4847,
+      "step": 13289
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 32.10342693043912,
+      "learning_rate": 4.577415889495962e-06,
+      "loss": 0.5837,
+      "step": 13290
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.98573331491425,
+      "learning_rate": 4.575999034343957e-06,
+      "loss": 0.5588,
+      "step": 13291
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 24.260059327578197,
+      "learning_rate": 4.5745823334473685e-06,
+      "loss": 0.4674,
+      "step": 13292
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.381798869962854,
+      "learning_rate": 4.573165786846485e-06,
+      "loss": 0.5284,
+      "step": 13293
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 24.528074104199415,
+      "learning_rate": 4.571749394581588e-06,
+      "loss": 0.5152,
+      "step": 13294
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.18882387638936,
+      "learning_rate": 4.570333156692963e-06,
+      "loss": 0.4948,
+      "step": 13295
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.183718532373142,
+      "learning_rate": 4.568917073220891e-06,
+      "loss": 0.4781,
+      "step": 13296
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 26.013057232625478,
+      "learning_rate": 4.567501144205639e-06,
+      "loss": 0.5371,
+      "step": 13297
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.105198645333353,
+      "learning_rate": 4.566085369687475e-06,
+      "loss": 0.4781,
+      "step": 13298
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.81561349514247,
+      "learning_rate": 4.564669749706663e-06,
+      "loss": 0.4589,
+      "step": 13299
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.75364095533292,
+      "learning_rate": 4.563254284303468e-06,
+      "loss": 0.5483,
+      "step": 13300
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.414261980529798,
+      "learning_rate": 4.56183897351814e-06,
+      "loss": 0.4292,
+      "step": 13301
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 28.169139517149215,
+      "learning_rate": 4.560423817390927e-06,
+      "loss": 0.5308,
+      "step": 13302
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.864232061451382,
+      "learning_rate": 4.559008815962078e-06,
+      "loss": 0.4324,
+      "step": 13303
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.460231120868325,
+      "learning_rate": 4.557593969271836e-06,
+      "loss": 0.528,
+      "step": 13304
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.877083541629784,
+      "learning_rate": 4.556179277360442e-06,
+      "loss": 0.4705,
+      "step": 13305
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 28.45493918474053,
+      "learning_rate": 4.554764740268124e-06,
+      "loss": 0.525,
+      "step": 13306
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 30.070636441276545,
+      "learning_rate": 4.553350358035108e-06,
+      "loss": 0.5167,
+      "step": 13307
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.855782165752,
+      "learning_rate": 4.551936130701622e-06,
+      "loss": 0.4943,
+      "step": 13308
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 18.96931533348104,
+      "learning_rate": 4.550522058307888e-06,
+      "loss": 0.4773,
+      "step": 13309
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.585443390966994,
+      "learning_rate": 4.549108140894117e-06,
+      "loss": 0.473,
+      "step": 13310
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.401202400835626,
+      "learning_rate": 4.547694378500523e-06,
+      "loss": 0.4329,
+      "step": 13311
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.704926195938093,
+      "learning_rate": 4.54628077116731e-06,
+      "loss": 0.5286,
+      "step": 13312
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 18.107338179007186,
+      "learning_rate": 4.544867318934684e-06,
+      "loss": 0.4678,
+      "step": 13313
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.743244162718618,
+      "learning_rate": 4.543454021842836e-06,
+      "loss": 0.4638,
+      "step": 13314
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.776271012553853,
+      "learning_rate": 4.5420408799319646e-06,
+      "loss": 0.45,
+      "step": 13315
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.885712340231894,
+      "learning_rate": 4.540627893242261e-06,
+      "loss": 0.4533,
+      "step": 13316
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.790081447185585,
+      "learning_rate": 4.539215061813903e-06,
+      "loss": 0.5177,
+      "step": 13317
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 35.4578125192979,
+      "learning_rate": 4.537802385687078e-06,
+      "loss": 0.4472,
+      "step": 13318
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.718204867464728,
+      "learning_rate": 4.5363898649019545e-06,
+      "loss": 0.4364,
+      "step": 13319
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.74725590503136,
+      "learning_rate": 4.534977499498706e-06,
+      "loss": 0.4313,
+      "step": 13320
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.19691321457073,
+      "learning_rate": 4.533565289517506e-06,
+      "loss": 0.5193,
+      "step": 13321
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.306957474357283,
+      "learning_rate": 4.5321532349985095e-06,
+      "loss": 0.4197,
+      "step": 13322
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.334155559786513,
+      "learning_rate": 4.530741335981874e-06,
+      "loss": 0.4571,
+      "step": 13323
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.120207059017911,
+      "learning_rate": 4.529329592507755e-06,
+      "loss": 0.4476,
+      "step": 13324
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 18.673845227092745,
+      "learning_rate": 4.527918004616305e-06,
+      "loss": 0.46,
+      "step": 13325
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 22.132544220666155,
+      "learning_rate": 4.5265065723476675e-06,
+      "loss": 0.5036,
+      "step": 13326
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.62823446652317,
+      "learning_rate": 4.525095295741976e-06,
+      "loss": 0.5255,
+      "step": 13327
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 28.649807073615964,
+      "learning_rate": 4.523684174839372e-06,
+      "loss": 0.5218,
+      "step": 13328
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.79047333020339,
+      "learning_rate": 4.5222732096799875e-06,
+      "loss": 0.431,
+      "step": 13329
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.665507084310978,
+      "learning_rate": 4.520862400303951e-06,
+      "loss": 0.5162,
+      "step": 13330
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.49819586497654,
+      "learning_rate": 4.519451746751383e-06,
+      "loss": 0.4882,
+      "step": 13331
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.160024828916969,
+      "learning_rate": 4.518041249062398e-06,
+      "loss": 0.5908,
+      "step": 13332
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.103632722671467,
+      "learning_rate": 4.516630907277113e-06,
+      "loss": 0.4805,
+      "step": 13333
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.395934853428344,
+      "learning_rate": 4.5152207214356406e-06,
+      "loss": 0.5737,
+      "step": 13334
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 18.24113983478634,
+      "learning_rate": 4.513810691578079e-06,
+      "loss": 0.5527,
+      "step": 13335
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.32820993767835,
+      "learning_rate": 4.512400817744535e-06,
+      "loss": 0.4805,
+      "step": 13336
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.00100803889626,
+      "learning_rate": 4.510991099975098e-06,
+      "loss": 0.4277,
+      "step": 13337
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 31.046224498989424,
+      "learning_rate": 4.509581538309867e-06,
+      "loss": 0.5508,
+      "step": 13338
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.51467379080833,
+      "learning_rate": 4.508172132788923e-06,
+      "loss": 0.5532,
+      "step": 13339
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.355440968077888,
+      "learning_rate": 4.506762883452349e-06,
+      "loss": 0.4422,
+      "step": 13340
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.804332656805983,
+      "learning_rate": 4.50535379034023e-06,
+      "loss": 0.4851,
+      "step": 13341
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 34.906183021177206,
+      "learning_rate": 4.503944853492631e-06,
+      "loss": 0.542,
+      "step": 13342
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.73444871325246,
+      "learning_rate": 4.502536072949628e-06,
+      "loss": 0.4982,
+      "step": 13343
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 28.978346040808887,
+      "learning_rate": 4.5011274487512806e-06,
+      "loss": 0.4867,
+      "step": 13344
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 20.602943925251328,
+      "learning_rate": 4.4997189809376554e-06,
+      "loss": 0.5291,
+      "step": 13345
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.135052178232254,
+      "learning_rate": 4.498310669548801e-06,
+      "loss": 0.4534,
+      "step": 13346
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.818277409403674,
+      "learning_rate": 4.496902514624775e-06,
+      "loss": 0.5213,
+      "step": 13347
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 19.446230919170148,
+      "learning_rate": 4.495494516205621e-06,
+      "loss": 0.5111,
+      "step": 13348
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.29595463617056,
+      "learning_rate": 4.494086674331382e-06,
+      "loss": 0.4961,
+      "step": 13349
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.931181608762111,
+      "learning_rate": 4.492678989042099e-06,
+      "loss": 0.4893,
+      "step": 13350
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 20.38331068840339,
+      "learning_rate": 4.491271460377806e-06,
+      "loss": 0.5184,
+      "step": 13351
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.977967451624647,
+      "learning_rate": 4.489864088378526e-06,
+      "loss": 0.4909,
+      "step": 13352
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 27.461327212784816,
+      "learning_rate": 4.488456873084288e-06,
+      "loss": 0.4533,
+      "step": 13353
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 30.113063848583863,
+      "learning_rate": 4.487049814535112e-06,
+      "loss": 0.4814,
+      "step": 13354
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 23.05653861450223,
+      "learning_rate": 4.4856429127710224e-06,
+      "loss": 0.5623,
+      "step": 13355
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.800967790895884,
+      "learning_rate": 4.484236167832015e-06,
+      "loss": 0.4503,
+      "step": 13356
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.436910201979742,
+      "learning_rate": 4.482829579758103e-06,
+      "loss": 0.4479,
+      "step": 13357
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.097455583234527,
+      "learning_rate": 4.481423148589292e-06,
+      "loss": 0.5083,
+      "step": 13358
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.07132455874364,
+      "learning_rate": 4.4800168743655814e-06,
+      "loss": 0.5065,
+      "step": 13359
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 18.15619025901775,
+      "learning_rate": 4.478610757126962e-06,
+      "loss": 0.4788,
+      "step": 13360
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 23.43294365272516,
+      "learning_rate": 4.477204796913419e-06,
+      "loss": 0.4948,
+      "step": 13361
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.072391561296246,
+      "learning_rate": 4.475798993764941e-06,
+      "loss": 0.4632,
+      "step": 13362
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.355250719725868,
+      "learning_rate": 4.474393347721511e-06,
+      "loss": 0.472,
+      "step": 13363
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.806147210482905,
+      "learning_rate": 4.472987858823099e-06,
+      "loss": 0.4732,
+      "step": 13364
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 30.2367927836394,
+      "learning_rate": 4.471582527109683e-06,
+      "loss": 0.5207,
+      "step": 13365
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.766143714746494,
+      "learning_rate": 4.470177352621221e-06,
+      "loss": 0.5428,
+      "step": 13366
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.279677872452869,
+      "learning_rate": 4.468772335397681e-06,
+      "loss": 0.5143,
+      "step": 13367
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 30.028609946729038,
+      "learning_rate": 4.467367475479023e-06,
+      "loss": 0.5886,
+      "step": 13368
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.703485283549373,
+      "learning_rate": 4.465962772905195e-06,
+      "loss": 0.47,
+      "step": 13369
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.02288981834509,
+      "learning_rate": 4.464558227716152e-06,
+      "loss": 0.4235,
+      "step": 13370
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.08937653998288,
+      "learning_rate": 4.463153839951829e-06,
+      "loss": 0.4421,
+      "step": 13371
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 25.071619293970567,
+      "learning_rate": 4.461749609652179e-06,
+      "loss": 0.5131,
+      "step": 13372
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 20.046640020481444,
+      "learning_rate": 4.460345536857124e-06,
+      "loss": 0.481,
+      "step": 13373
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.804403857693078,
+      "learning_rate": 4.458941621606602e-06,
+      "loss": 0.4692,
+      "step": 13374
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 28.887366923500668,
+      "learning_rate": 4.457537863940543e-06,
+      "loss": 0.4929,
+      "step": 13375
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.737847099771972,
+      "learning_rate": 4.4561342638988645e-06,
+      "loss": 0.4731,
+      "step": 13376
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.01579895841372,
+      "learning_rate": 4.4547308215214815e-06,
+      "loss": 0.4864,
+      "step": 13377
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.57507625521603,
+      "learning_rate": 4.453327536848309e-06,
+      "loss": 0.5053,
+      "step": 13378
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.295256455841429,
+      "learning_rate": 4.451924409919257e-06,
+      "loss": 0.4498,
+      "step": 13379
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.032001856742218,
+      "learning_rate": 4.450521440774233e-06,
+      "loss": 0.4372,
+      "step": 13380
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.853086634235515,
+      "learning_rate": 4.449118629453133e-06,
+      "loss": 0.4574,
+      "step": 13381
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.162096127968514,
+      "learning_rate": 4.447715975995848e-06,
+      "loss": 0.4785,
+      "step": 13382
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.77159990249389,
+      "learning_rate": 4.446313480442272e-06,
+      "loss": 0.516,
+      "step": 13383
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.727921858709138,
+      "learning_rate": 4.444911142832297e-06,
+      "loss": 0.5095,
+      "step": 13384
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.91745334847935,
+      "learning_rate": 4.4435089632058e-06,
+      "loss": 0.545,
+      "step": 13385
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 20.937821783610683,
+      "learning_rate": 4.442106941602652e-06,
+      "loss": 0.3844,
+      "step": 13386
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.151497540172905,
+      "learning_rate": 4.440705078062732e-06,
+      "loss": 0.4793,
+      "step": 13387
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.833597902207083,
+      "learning_rate": 4.4393033726259116e-06,
+      "loss": 0.4458,
+      "step": 13388
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.930245868701412,
+      "learning_rate": 4.437901825332046e-06,
+      "loss": 0.5224,
+      "step": 13389
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 35.93038643051206,
+      "learning_rate": 4.436500436221003e-06,
+      "loss": 0.5245,
+      "step": 13390
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.300825086649915,
+      "learning_rate": 4.4350992053326295e-06,
+      "loss": 0.4769,
+      "step": 13391
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.15378430110901,
+      "learning_rate": 4.433698132706779e-06,
+      "loss": 0.4482,
+      "step": 13392
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 23.1344869407657,
+      "learning_rate": 4.4322972183833e-06,
+      "loss": 0.4981,
+      "step": 13393
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.851429273959964,
+      "learning_rate": 4.430896462402033e-06,
+      "loss": 0.4633,
+      "step": 13394
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.324897390681926,
+      "learning_rate": 4.429495864802808e-06,
+      "loss": 0.5283,
+      "step": 13395
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.50423002562317,
+      "learning_rate": 4.428095425625462e-06,
+      "loss": 0.5526,
+      "step": 13396
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.829657741191447,
+      "learning_rate": 4.426695144909826e-06,
+      "loss": 0.4498,
+      "step": 13397
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.97235357949818,
+      "learning_rate": 4.425295022695716e-06,
+      "loss": 0.4363,
+      "step": 13398
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.15532984810896,
+      "learning_rate": 4.423895059022959e-06,
+      "loss": 0.5423,
+      "step": 13399
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 29.374390386789514,
+      "learning_rate": 4.42249525393136e-06,
+      "loss": 0.4868,
+      "step": 13400
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.370441946581966,
+      "learning_rate": 4.421095607460738e-06,
+      "loss": 0.5241,
+      "step": 13401
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 24.046294385656193,
+      "learning_rate": 4.419696119650889e-06,
+      "loss": 0.5022,
+      "step": 13402
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.471081369764708,
+      "learning_rate": 4.418296790541618e-06,
+      "loss": 0.453,
+      "step": 13403
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.778671540287874,
+      "learning_rate": 4.4168976201727255e-06,
+      "loss": 0.5484,
+      "step": 13404
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.642828665612175,
+      "learning_rate": 4.415498608583993e-06,
+      "loss": 0.498,
+      "step": 13405
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.196734116869575,
+      "learning_rate": 4.414099755815219e-06,
+      "loss": 0.5427,
+      "step": 13406
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 24.414930914730913,
+      "learning_rate": 4.412701061906176e-06,
+      "loss": 0.4866,
+      "step": 13407
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.613471442240435,
+      "learning_rate": 4.411302526896646e-06,
+      "loss": 0.4846,
+      "step": 13408
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.281869920587228,
+      "learning_rate": 4.409904150826408e-06,
+      "loss": 0.487,
+      "step": 13409
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.914756785320684,
+      "learning_rate": 4.408505933735223e-06,
+      "loss": 0.4572,
+      "step": 13410
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 31.268555643467717,
+      "learning_rate": 4.407107875662857e-06,
+      "loss": 0.5634,
+      "step": 13411
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.664296413192254,
+      "learning_rate": 4.405709976649069e-06,
+      "loss": 0.4784,
+      "step": 13412
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 26.10843387513603,
+      "learning_rate": 4.404312236733623e-06,
+      "loss": 0.5861,
+      "step": 13413
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.622896061499468,
+      "learning_rate": 4.402914655956262e-06,
+      "loss": 0.4474,
+      "step": 13414
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.505891463197088,
+      "learning_rate": 4.401517234356731e-06,
+      "loss": 0.469,
+      "step": 13415
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.116856751582834,
+      "learning_rate": 4.400119971974774e-06,
+      "loss": 0.4995,
+      "step": 13416
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.487069225015794,
+      "learning_rate": 4.398722868850131e-06,
+      "loss": 0.5086,
+      "step": 13417
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 27.17987141460619,
+      "learning_rate": 4.3973259250225355e-06,
+      "loss": 0.5204,
+      "step": 13418
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.958732782735332,
+      "learning_rate": 4.395929140531713e-06,
+      "loss": 0.5012,
+      "step": 13419
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.360661793517504,
+      "learning_rate": 4.3945325154173834e-06,
+      "loss": 0.4872,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.569555511957606,
+      "learning_rate": 4.39313604971927e-06,
+      "loss": 0.487,
+      "step": 13421
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.395513677277506,
+      "learning_rate": 4.391739743477092e-06,
+      "loss": 0.4899,
+      "step": 13422
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.362455836163539,
+      "learning_rate": 4.390343596730551e-06,
+      "loss": 0.4725,
+      "step": 13423
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.110999586455385,
+      "learning_rate": 4.38894760951936e-06,
+      "loss": 0.4512,
+      "step": 13424
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 34.52331145484327,
+      "learning_rate": 4.387551781883213e-06,
+      "loss": 0.4879,
+      "step": 13425
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.1126875864702,
+      "learning_rate": 4.386156113861814e-06,
+      "loss": 0.514,
+      "step": 13426
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.91621939521298,
+      "learning_rate": 4.384760605494847e-06,
+      "loss": 0.4568,
+      "step": 13427
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.71690157702032,
+      "learning_rate": 4.383365256822003e-06,
+      "loss": 0.5882,
+      "step": 13428
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 28.110465908033323,
+      "learning_rate": 4.3819700678829705e-06,
+      "loss": 0.5001,
+      "step": 13429
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 24.28365850715801,
+      "learning_rate": 4.380575038717419e-06,
+      "loss": 0.4645,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.778307378775217,
+      "learning_rate": 4.37918016936503e-06,
+      "loss": 0.4423,
+      "step": 13431
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.601300170022775,
+      "learning_rate": 4.377785459865463e-06,
+      "loss": 0.4522,
+      "step": 13432
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.816120012352762,
+      "learning_rate": 4.376390910258391e-06,
+      "loss": 0.5215,
+      "step": 13433
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.929450510871082,
+      "learning_rate": 4.374996520583474e-06,
+      "loss": 0.4921,
+      "step": 13434
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.439954293978737,
+      "learning_rate": 4.373602290880367e-06,
+      "loss": 0.5448,
+      "step": 13435
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.480410820748844,
+      "learning_rate": 4.372208221188715e-06,
+      "loss": 0.5549,
+      "step": 13436
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.031498653574978,
+      "learning_rate": 4.370814311548168e-06,
+      "loss": 0.5153,
+      "step": 13437
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.58080708478236,
+      "learning_rate": 4.3694205619983744e-06,
+      "loss": 0.4444,
+      "step": 13438
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.081969274360034,
+      "learning_rate": 4.368026972578965e-06,
+      "loss": 0.4861,
+      "step": 13439
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.746778438933905,
+      "learning_rate": 4.366633543329571e-06,
+      "loss": 0.4883,
+      "step": 13440
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.18715196160379,
+      "learning_rate": 4.365240274289824e-06,
+      "loss": 0.5169,
+      "step": 13441
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.94904135873195,
+      "learning_rate": 4.363847165499347e-06,
+      "loss": 0.5225,
+      "step": 13442
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.299634151607687,
+      "learning_rate": 4.362454216997761e-06,
+      "loss": 0.5753,
+      "step": 13443
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.082429075885264,
+      "learning_rate": 4.361061428824682e-06,
+      "loss": 0.4669,
+      "step": 13444
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.104073344794825,
+      "learning_rate": 4.359668801019713e-06,
+      "loss": 0.4348,
+      "step": 13445
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.718910643004673,
+      "learning_rate": 4.358276333622463e-06,
+      "loss": 0.4432,
+      "step": 13446
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.711772740623914,
+      "learning_rate": 4.356884026672537e-06,
+      "loss": 0.5723,
+      "step": 13447
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 25.126831003982765,
+      "learning_rate": 4.35549188020953e-06,
+      "loss": 0.5013,
+      "step": 13448
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.404262327988558,
+      "learning_rate": 4.354099894273027e-06,
+      "loss": 0.4589,
+      "step": 13449
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.73178193177861,
+      "learning_rate": 4.352708068902621e-06,
+      "loss": 0.5353,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.986171861251247,
+      "learning_rate": 4.351316404137898e-06,
+      "loss": 0.4418,
+      "step": 13451
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.675801292465934,
+      "learning_rate": 4.349924900018427e-06,
+      "loss": 0.4957,
+      "step": 13452
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 26.15825558052395,
+      "learning_rate": 4.3485335565837905e-06,
+      "loss": 0.467,
+      "step": 13453
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.360053995739154,
+      "learning_rate": 4.347142373873551e-06,
+      "loss": 0.5268,
+      "step": 13454
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.354302844395008,
+      "learning_rate": 4.345751351927275e-06,
+      "loss": 0.4907,
+      "step": 13455
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.607745848555282,
+      "learning_rate": 4.344360490784526e-06,
+      "loss": 0.3871,
+      "step": 13456
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.682853313083948,
+      "learning_rate": 4.342969790484853e-06,
+      "loss": 0.507,
+      "step": 13457
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.511992220193015,
+      "learning_rate": 4.341579251067815e-06,
+      "loss": 0.5151,
+      "step": 13458
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.429363591706245,
+      "learning_rate": 4.3401888725729465e-06,
+      "loss": 0.4728,
+      "step": 13459
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.251778645359995,
+      "learning_rate": 4.338798655039802e-06,
+      "loss": 0.5422,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 28.283733727263364,
+      "learning_rate": 4.3374085985079075e-06,
+      "loss": 0.4668,
+      "step": 13461
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 27.149063186570853,
+      "learning_rate": 4.336018703016799e-06,
+      "loss": 0.6065,
+      "step": 13462
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.774726462637535,
+      "learning_rate": 4.33462896860601e-06,
+      "loss": 0.5066,
+      "step": 13463
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 25.6124587660728,
+      "learning_rate": 4.3332393953150574e-06,
+      "loss": 0.5152,
+      "step": 13464
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.04154843680813,
+      "learning_rate": 4.331849983183459e-06,
+      "loss": 0.4485,
+      "step": 13465
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.307506383220737,
+      "learning_rate": 4.330460732250732e-06,
+      "loss": 0.5035,
+      "step": 13466
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.697163901889468,
+      "learning_rate": 4.329071642556384e-06,
+      "loss": 0.504,
+      "step": 13467
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.534570776876762,
+      "learning_rate": 4.327682714139925e-06,
+      "loss": 0.5129,
+      "step": 13468
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.343498125849504,
+      "learning_rate": 4.326293947040851e-06,
+      "loss": 0.5047,
+      "step": 13469
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.346328722351895,
+      "learning_rate": 4.324905341298655e-06,
+      "loss": 0.4915,
+      "step": 13470
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.040168434746302,
+      "learning_rate": 4.323516896952832e-06,
+      "loss": 0.4433,
+      "step": 13471
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.014367330946357,
+      "learning_rate": 4.3221286140428695e-06,
+      "loss": 0.5257,
+      "step": 13472
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.90869229918086,
+      "learning_rate": 4.32074049260825e-06,
+      "loss": 0.4855,
+      "step": 13473
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.19758465507929,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.4622,
+      "step": 13474
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.783367663592294,
+      "learning_rate": 4.317964734322928e-06,
+      "loss": 0.4441,
+      "step": 13475
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.85024333048001,
+      "learning_rate": 4.316577097551176e-06,
+      "loss": 0.4744,
+      "step": 13476
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.207394375839547,
+      "learning_rate": 4.315189622412642e-06,
+      "loss": 0.4636,
+      "step": 13477
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 23.860458434674364,
+      "learning_rate": 4.313802308946794e-06,
+      "loss": 0.5367,
+      "step": 13478
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.696411620464664,
+      "learning_rate": 4.312415157193078e-06,
+      "loss": 0.4785,
+      "step": 13479
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.291071754104014,
+      "learning_rate": 4.3110281671909484e-06,
+      "loss": 0.5058,
+      "step": 13480
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.7198849206344,
+      "learning_rate": 4.309641338979853e-06,
+      "loss": 0.5133,
+      "step": 13481
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 14.280293193668603,
+      "learning_rate": 4.308254672599225e-06,
+      "loss": 0.47,
+      "step": 13482
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.55920090012835,
+      "learning_rate": 4.306868168088508e-06,
+      "loss": 0.4729,
+      "step": 13483
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.34642124592735,
+      "learning_rate": 4.305481825487128e-06,
+      "loss": 0.4877,
+      "step": 13484
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.00180157595025,
+      "learning_rate": 4.304095644834516e-06,
+      "loss": 0.5126,
+      "step": 13485
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.27862461031544,
+      "learning_rate": 4.302709626170089e-06,
+      "loss": 0.5341,
+      "step": 13486
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.11646340000066,
+      "learning_rate": 4.30132376953327e-06,
+      "loss": 0.3857,
+      "step": 13487
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.20590603693105,
+      "learning_rate": 4.299938074963465e-06,
+      "loss": 0.4701,
+      "step": 13488
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.47539332846092,
+      "learning_rate": 4.298552542500093e-06,
+      "loss": 0.502,
+      "step": 13489
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.398325496980426,
+      "learning_rate": 4.297167172182546e-06,
+      "loss": 0.5287,
+      "step": 13490
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.68794039397025,
+      "learning_rate": 4.295781964050229e-06,
+      "loss": 0.5194,
+      "step": 13491
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.276072685491886,
+      "learning_rate": 4.2943969181425395e-06,
+      "loss": 0.4434,
+      "step": 13492
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.53435643712485,
+      "learning_rate": 4.2930120344988605e-06,
+      "loss": 0.4705,
+      "step": 13493
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 24.75365813631957,
+      "learning_rate": 4.291627313158584e-06,
+      "loss": 0.5483,
+      "step": 13494
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.243731719798483,
+      "learning_rate": 4.290242754161084e-06,
+      "loss": 0.493,
+      "step": 13495
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.60385053777235,
+      "learning_rate": 4.28885835754574e-06,
+      "loss": 0.4554,
+      "step": 13496
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 32.31787061053695,
+      "learning_rate": 4.287474123351928e-06,
+      "loss": 0.541,
+      "step": 13497
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.96526104970375,
+      "learning_rate": 4.286090051619008e-06,
+      "loss": 0.4559,
+      "step": 13498
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.100422787102886,
+      "learning_rate": 4.284706142386342e-06,
+      "loss": 0.5356,
+      "step": 13499
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.722016532497694,
+      "learning_rate": 4.2833223956932916e-06,
+      "loss": 0.5214,
+      "step": 13500
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.78271697635198,
+      "learning_rate": 4.2819388115792095e-06,
+      "loss": 0.4669,
+      "step": 13501
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.97942187587495,
+      "learning_rate": 4.280555390083443e-06,
+      "loss": 0.4304,
+      "step": 13502
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.659132246796773,
+      "learning_rate": 4.279172131245332e-06,
+      "loss": 0.4608,
+      "step": 13503
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.692096957162224,
+      "learning_rate": 4.27778903510422e-06,
+      "loss": 0.5481,
+      "step": 13504
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 24.752425590297722,
+      "learning_rate": 4.2764061016994386e-06,
+      "loss": 0.538,
+      "step": 13505
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.18439080037953,
+      "learning_rate": 4.275023331070324e-06,
+      "loss": 0.5294,
+      "step": 13506
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.458798299958378,
+      "learning_rate": 4.2736407232561975e-06,
+      "loss": 0.4268,
+      "step": 13507
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 23.10961786649343,
+      "learning_rate": 4.272258278296374e-06,
+      "loss": 0.4963,
+      "step": 13508
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.70235840348874,
+      "learning_rate": 4.270875996230176e-06,
+      "loss": 0.5808,
+      "step": 13509
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 28.62649659082115,
+      "learning_rate": 4.269493877096915e-06,
+      "loss": 0.5652,
+      "step": 13510
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.99787160008829,
+      "learning_rate": 4.268111920935893e-06,
+      "loss": 0.4426,
+      "step": 13511
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.321086393620956,
+      "learning_rate": 4.266730127786419e-06,
+      "loss": 0.4878,
+      "step": 13512
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 27.343704060107193,
+      "learning_rate": 4.265348497687784e-06,
+      "loss": 0.495,
+      "step": 13513
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 24.921945488091644,
+      "learning_rate": 4.263967030679284e-06,
+      "loss": 0.4941,
+      "step": 13514
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.520433813811366,
+      "learning_rate": 4.262585726800204e-06,
+      "loss": 0.5092,
+      "step": 13515
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.808686122499925,
+      "learning_rate": 4.26120458608983e-06,
+      "loss": 0.5165,
+      "step": 13516
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.30822553360003,
+      "learning_rate": 4.259823608587443e-06,
+      "loss": 0.505,
+      "step": 13517
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.909309206936634,
+      "learning_rate": 4.258442794332311e-06,
+      "loss": 0.4927,
+      "step": 13518
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 28.05286728858601,
+      "learning_rate": 4.2570621433637116e-06,
+      "loss": 0.5266,
+      "step": 13519
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 28.709137145186844,
+      "learning_rate": 4.255681655720901e-06,
+      "loss": 0.6329,
+      "step": 13520
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.58938696985488,
+      "learning_rate": 4.254301331443145e-06,
+      "loss": 0.5211,
+      "step": 13521
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.575775673928216,
+      "learning_rate": 4.252921170569701e-06,
+      "loss": 0.4887,
+      "step": 13522
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.976739653636162,
+      "learning_rate": 4.251541173139816e-06,
+      "loss": 0.4615,
+      "step": 13523
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.604192278645606,
+      "learning_rate": 4.250161339192734e-06,
+      "loss": 0.4779,
+      "step": 13524
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.061330305781027,
+      "learning_rate": 4.2487816687677006e-06,
+      "loss": 0.4923,
+      "step": 13525
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 22.94314081684299,
+      "learning_rate": 4.247402161903956e-06,
+      "loss": 0.4928,
+      "step": 13526
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.352750244607382,
+      "learning_rate": 4.246022818640728e-06,
+      "loss": 0.5382,
+      "step": 13527
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.118142313798433,
+      "learning_rate": 4.244643639017242e-06,
+      "loss": 0.513,
+      "step": 13528
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.816358548555687,
+      "learning_rate": 4.243264623072725e-06,
+      "loss": 0.5031,
+      "step": 13529
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 28.05101485501066,
+      "learning_rate": 4.241885770846392e-06,
+      "loss": 0.5212,
+      "step": 13530
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.47354865135328,
+      "learning_rate": 4.240507082377465e-06,
+      "loss": 0.5088,
+      "step": 13531
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 26.268185417557632,
+      "learning_rate": 4.239128557705146e-06,
+      "loss": 0.4698,
+      "step": 13532
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 28.84944940364741,
+      "learning_rate": 4.237750196868639e-06,
+      "loss": 0.5317,
+      "step": 13533
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.476987320569467,
+      "learning_rate": 4.236371999907144e-06,
+      "loss": 0.4917,
+      "step": 13534
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 14.891998469844339,
+      "learning_rate": 4.234993966859862e-06,
+      "loss": 0.4485,
+      "step": 13535
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.146306924811462,
+      "learning_rate": 4.233616097765979e-06,
+      "loss": 0.5415,
+      "step": 13536
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.932540197922695,
+      "learning_rate": 4.232238392664677e-06,
+      "loss": 0.4883,
+      "step": 13537
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.163851399526058,
+      "learning_rate": 4.2308608515951414e-06,
+      "loss": 0.4369,
+      "step": 13538
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.091191036385965,
+      "learning_rate": 4.229483474596552e-06,
+      "loss": 0.4475,
+      "step": 13539
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 23.806749958748377,
+      "learning_rate": 4.228106261708073e-06,
+      "loss": 0.4853,
+      "step": 13540
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.165884479575702,
+      "learning_rate": 4.22672921296888e-06,
+      "loss": 0.4928,
+      "step": 13541
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 21.29297327151099,
+      "learning_rate": 4.225352328418126e-06,
+      "loss": 0.5042,
+      "step": 13542
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.2225418156734,
+      "learning_rate": 4.223975608094975e-06,
+      "loss": 0.5866,
+      "step": 13543
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.721819138129119,
+      "learning_rate": 4.222599052038581e-06,
+      "loss": 0.4423,
+      "step": 13544
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.805819150086123,
+      "learning_rate": 4.221222660288087e-06,
+      "loss": 0.4885,
+      "step": 13545
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.034232436800302,
+      "learning_rate": 4.219846432882644e-06,
+      "loss": 0.4366,
+      "step": 13546
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 29.151975361289153,
+      "learning_rate": 4.218470369861384e-06,
+      "loss": 0.6061,
+      "step": 13547
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 26.154425497011548,
+      "learning_rate": 4.217094471263447e-06,
+      "loss": 0.5386,
+      "step": 13548
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.32975497437622,
+      "learning_rate": 4.2157187371279585e-06,
+      "loss": 0.4813,
+      "step": 13549
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 26.540746655056783,
+      "learning_rate": 4.214343167494044e-06,
+      "loss": 0.4965,
+      "step": 13550
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.448068747658212,
+      "learning_rate": 4.2129677624008294e-06,
+      "loss": 0.4188,
+      "step": 13551
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.746551157949762,
+      "learning_rate": 4.211592521887426e-06,
+      "loss": 0.4561,
+      "step": 13552
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.01736470911061,
+      "learning_rate": 4.2102174459929424e-06,
+      "loss": 0.4634,
+      "step": 13553
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.461717377866623,
+      "learning_rate": 4.2088425347564866e-06,
+      "loss": 0.4935,
+      "step": 13554
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.913210044486908,
+      "learning_rate": 4.207467788217162e-06,
+      "loss": 0.4567,
+      "step": 13555
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.193984821439942,
+      "learning_rate": 4.206093206414069e-06,
+      "loss": 0.4453,
+      "step": 13556
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 35.940176690803796,
+      "learning_rate": 4.204718789386295e-06,
+      "loss": 0.5471,
+      "step": 13557
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.406448302584195,
+      "learning_rate": 4.203344537172925e-06,
+      "loss": 0.4823,
+      "step": 13558
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 17.516960817506494,
+      "learning_rate": 4.201970449813045e-06,
+      "loss": 0.427,
+      "step": 13559
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.517757946971244,
+      "learning_rate": 4.200596527345738e-06,
+      "loss": 0.4655,
+      "step": 13560
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.124877200012897,
+      "learning_rate": 4.199222769810074e-06,
+      "loss": 0.5051,
+      "step": 13561
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.653873230921988,
+      "learning_rate": 4.197849177245117e-06,
+      "loss": 0.5575,
+      "step": 13562
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.744192549943213,
+      "learning_rate": 4.196475749689934e-06,
+      "loss": 0.5311,
+      "step": 13563
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.507579584161434,
+      "learning_rate": 4.19510248718359e-06,
+      "loss": 0.5005,
+      "step": 13564
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 25.340815557395395,
+      "learning_rate": 4.1937293897651324e-06,
+      "loss": 0.4535,
+      "step": 13565
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.670890821891394,
+      "learning_rate": 4.192356457473617e-06,
+      "loss": 0.4888,
+      "step": 13566
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.3920332498085,
+      "learning_rate": 4.190983690348084e-06,
+      "loss": 0.5754,
+      "step": 13567
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.55890911024046,
+      "learning_rate": 4.189611088427576e-06,
+      "loss": 0.5032,
+      "step": 13568
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.953563424492723,
+      "learning_rate": 4.188238651751134e-06,
+      "loss": 0.4359,
+      "step": 13569
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.11835830233997,
+      "learning_rate": 4.186866380357782e-06,
+      "loss": 0.4383,
+      "step": 13570
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.627039716220475,
+      "learning_rate": 4.1854942742865514e-06,
+      "loss": 0.4585,
+      "step": 13571
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.443986626544245,
+      "learning_rate": 4.184122333576459e-06,
+      "loss": 0.4323,
+      "step": 13572
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 25.55811324404679,
+      "learning_rate": 4.1827505582665295e-06,
+      "loss": 0.5244,
+      "step": 13573
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.951393495349535,
+      "learning_rate": 4.181378948395767e-06,
+      "loss": 0.4781,
+      "step": 13574
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.91530885927774,
+      "learning_rate": 4.180007504003183e-06,
+      "loss": 0.4867,
+      "step": 13575
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.51367503981383,
+      "learning_rate": 4.178636225127786e-06,
+      "loss": 0.5121,
+      "step": 13576
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 21.956888057565543,
+      "learning_rate": 4.177265111808568e-06,
+      "loss": 0.5423,
+      "step": 13577
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 32.657642296492234,
+      "learning_rate": 4.175894164084521e-06,
+      "loss": 0.4856,
+      "step": 13578
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.39869156722586,
+      "learning_rate": 4.174523381994637e-06,
+      "loss": 0.5382,
+      "step": 13579
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.979040378898556,
+      "learning_rate": 4.1731527655779e-06,
+      "loss": 0.4163,
+      "step": 13580
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.721661672297405,
+      "learning_rate": 4.171782314873294e-06,
+      "loss": 0.5212,
+      "step": 13581
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.207655940404102,
+      "learning_rate": 4.170412029919788e-06,
+      "loss": 0.4455,
+      "step": 13582
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.62061479067443,
+      "learning_rate": 4.169041910756352e-06,
+      "loss": 0.4677,
+      "step": 13583
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 28.295346041299343,
+      "learning_rate": 4.167671957421952e-06,
+      "loss": 0.4471,
+      "step": 13584
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.790220923991725,
+      "learning_rate": 4.166302169955553e-06,
+      "loss": 0.4968,
+      "step": 13585
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 17.582258738732836,
+      "learning_rate": 4.164932548396108e-06,
+      "loss": 0.5303,
+      "step": 13586
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 21.44405705715216,
+      "learning_rate": 4.163563092782564e-06,
+      "loss": 0.4366,
+      "step": 13587
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.801282382393591,
+      "learning_rate": 4.162193803153872e-06,
+      "loss": 0.545,
+      "step": 13588
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.452182708128388,
+      "learning_rate": 4.160824679548977e-06,
+      "loss": 0.4392,
+      "step": 13589
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 30.298217108255777,
+      "learning_rate": 4.1594557220068125e-06,
+      "loss": 0.4864,
+      "step": 13590
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 26.0626089839349,
+      "learning_rate": 4.158086930566307e-06,
+      "loss": 0.61,
+      "step": 13591
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 29.86377432037528,
+      "learning_rate": 4.15671830526639e-06,
+      "loss": 0.4633,
+      "step": 13592
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.084666647865678,
+      "learning_rate": 4.155349846145988e-06,
+      "loss": 0.4707,
+      "step": 13593
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.109577938535764,
+      "learning_rate": 4.153981553244021e-06,
+      "loss": 0.4985,
+      "step": 13594
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.339063911976808,
+      "learning_rate": 4.152613426599398e-06,
+      "loss": 0.6181,
+      "step": 13595
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.537277732569315,
+      "learning_rate": 4.151245466251025e-06,
+      "loss": 0.4634,
+      "step": 13596
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.67606763859391,
+      "learning_rate": 4.14987767223781e-06,
+      "loss": 0.4626,
+      "step": 13597
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 17.820083548125744,
+      "learning_rate": 4.148510044598655e-06,
+      "loss": 0.4513,
+      "step": 13598
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.20304431619588,
+      "learning_rate": 4.147142583372448e-06,
+      "loss": 0.4403,
+      "step": 13599
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.956380531482687,
+      "learning_rate": 4.145775288598085e-06,
+      "loss": 0.4714,
+      "step": 13600
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.326358972281653,
+      "learning_rate": 4.1444081603144445e-06,
+      "loss": 0.5179,
+      "step": 13601
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.695428501969044,
+      "learning_rate": 4.1430411985604145e-06,
+      "loss": 0.4596,
+      "step": 13602
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.93097910976537,
+      "learning_rate": 4.141674403374864e-06,
+      "loss": 0.5294,
+      "step": 13603
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 22.41000683498459,
+      "learning_rate": 4.1403077747966646e-06,
+      "loss": 0.5187,
+      "step": 13604
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.835924438098633,
+      "learning_rate": 4.13894131286469e-06,
+      "loss": 0.4493,
+      "step": 13605
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.468391398315873,
+      "learning_rate": 4.13757501761779e-06,
+      "loss": 0.5106,
+      "step": 13606
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.169558327599734,
+      "learning_rate": 4.136208889094832e-06,
+      "loss": 0.5085,
+      "step": 13607
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 30.760694053416294,
+      "learning_rate": 4.1348429273346595e-06,
+      "loss": 0.4817,
+      "step": 13608
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.130547217234085,
+      "learning_rate": 4.133477132376122e-06,
+      "loss": 0.4943,
+      "step": 13609
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.134660244498594,
+      "learning_rate": 4.132111504258067e-06,
+      "loss": 0.4764,
+      "step": 13610
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.11397962311846,
+      "learning_rate": 4.130746043019329e-06,
+      "loss": 0.47,
+      "step": 13611
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.031046037436356,
+      "learning_rate": 4.129380748698737e-06,
+      "loss": 0.51,
+      "step": 13612
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.06777510111775,
+      "learning_rate": 4.128015621335121e-06,
+      "loss": 0.5151,
+      "step": 13613
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.195392775139672,
+      "learning_rate": 4.12665066096731e-06,
+      "loss": 0.5055,
+      "step": 13614
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.38425398003947,
+      "learning_rate": 4.1252858676341175e-06,
+      "loss": 0.4768,
+      "step": 13615
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.492607801792264,
+      "learning_rate": 4.123921241374355e-06,
+      "loss": 0.516,
+      "step": 13616
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.164307575955736,
+      "learning_rate": 4.1225567822268365e-06,
+      "loss": 0.4869,
+      "step": 13617
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.589675209840214,
+      "learning_rate": 4.121192490230363e-06,
+      "loss": 0.4767,
+      "step": 13618
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 16.242954502096712,
+      "learning_rate": 4.11982836542374e-06,
+      "loss": 0.4503,
+      "step": 13619
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.02628616834752,
+      "learning_rate": 4.118464407845759e-06,
+      "loss": 0.5123,
+      "step": 13620
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.525161193664482,
+      "learning_rate": 4.117100617535207e-06,
+      "loss": 0.4797,
+      "step": 13621
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.23901406084867,
+      "learning_rate": 4.11573699453087e-06,
+      "loss": 0.4386,
+      "step": 13622
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.815428997454436,
+      "learning_rate": 4.114373538871535e-06,
+      "loss": 0.4999,
+      "step": 13623
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.88352985085275,
+      "learning_rate": 4.1130102505959715e-06,
+      "loss": 0.4624,
+      "step": 13624
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.60170207397827,
+      "learning_rate": 4.111647129742954e-06,
+      "loss": 0.5037,
+      "step": 13625
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.241897778712433,
+      "learning_rate": 4.110284176351245e-06,
+      "loss": 0.4716,
+      "step": 13626
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.909195686506425,
+      "learning_rate": 4.108921390459612e-06,
+      "loss": 0.536,
+      "step": 13627
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.83249178786733,
+      "learning_rate": 4.107558772106805e-06,
+      "loss": 0.5445,
+      "step": 13628
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.100331026489787,
+      "learning_rate": 4.106196321331581e-06,
+      "loss": 0.4798,
+      "step": 13629
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.76686234714087,
+      "learning_rate": 4.104834038172687e-06,
+      "loss": 0.5654,
+      "step": 13630
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 16.727406996709465,
+      "learning_rate": 4.103471922668862e-06,
+      "loss": 0.4911,
+      "step": 13631
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.848115422367645,
+      "learning_rate": 4.10210997485885e-06,
+      "loss": 0.5234,
+      "step": 13632
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.365964335774482,
+      "learning_rate": 4.100748194781376e-06,
+      "loss": 0.4239,
+      "step": 13633
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.449688821689175,
+      "learning_rate": 4.099386582475175e-06,
+      "loss": 0.5003,
+      "step": 13634
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 31.09837584950147,
+      "learning_rate": 4.0980251379789655e-06,
+      "loss": 0.4586,
+      "step": 13635
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.19372386342677,
+      "learning_rate": 4.096663861331472e-06,
+      "loss": 0.4624,
+      "step": 13636
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.027454313568263,
+      "learning_rate": 4.095302752571402e-06,
+      "loss": 0.4547,
+      "step": 13637
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 30.405046929771117,
+      "learning_rate": 4.093941811737466e-06,
+      "loss": 0.5329,
+      "step": 13638
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.270770578509428,
+      "learning_rate": 4.092581038868375e-06,
+      "loss": 0.4578,
+      "step": 13639
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 24.343442198750296,
+      "learning_rate": 4.0912204340028224e-06,
+      "loss": 0.521,
+      "step": 13640
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.023524792437934,
+      "learning_rate": 4.089859997179502e-06,
+      "loss": 0.4562,
+      "step": 13641
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.9503108633932,
+      "learning_rate": 4.088499728437104e-06,
+      "loss": 0.5366,
+      "step": 13642
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.036514319506963,
+      "learning_rate": 4.087139627814317e-06,
+      "loss": 0.4861,
+      "step": 13643
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.313620862264777,
+      "learning_rate": 4.085779695349823e-06,
+      "loss": 0.5295,
+      "step": 13644
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 26.80087441051656,
+      "learning_rate": 4.084419931082295e-06,
+      "loss": 0.5164,
+      "step": 13645
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.693608031999965,
+      "learning_rate": 4.083060335050401e-06,
+      "loss": 0.5041,
+      "step": 13646
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.884248269917485,
+      "learning_rate": 4.0817009072928085e-06,
+      "loss": 0.3982,
+      "step": 13647
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.967372945924364,
+      "learning_rate": 4.080341647848184e-06,
+      "loss": 0.4749,
+      "step": 13648
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.246629287595802,
+      "learning_rate": 4.07898255675518e-06,
+      "loss": 0.4649,
+      "step": 13649
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.401588402596303,
+      "learning_rate": 4.077623634052445e-06,
+      "loss": 0.5824,
+      "step": 13650
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.589883329087062,
+      "learning_rate": 4.076264879778629e-06,
+      "loss": 0.4992,
+      "step": 13651
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.885499242978852,
+      "learning_rate": 4.074906293972378e-06,
+      "loss": 0.4625,
+      "step": 13652
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.665082111681873,
+      "learning_rate": 4.073547876672323e-06,
+      "loss": 0.5504,
+      "step": 13653
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.166712005149897,
+      "learning_rate": 4.072189627917102e-06,
+      "loss": 0.4586,
+      "step": 13654
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.487445647461794,
+      "learning_rate": 4.070831547745338e-06,
+      "loss": 0.4891,
+      "step": 13655
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.933707751744688,
+      "learning_rate": 4.0694736361956555e-06,
+      "loss": 0.4906,
+      "step": 13656
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.028510284517882,
+      "learning_rate": 4.068115893306678e-06,
+      "loss": 0.4744,
+      "step": 13657
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.826276854669484,
+      "learning_rate": 4.06675831911701e-06,
+      "loss": 0.4593,
+      "step": 13658
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 28.371458327776494,
+      "learning_rate": 4.0654009136652685e-06,
+      "loss": 0.4562,
+      "step": 13659
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.532878240921082,
+      "learning_rate": 4.06404367699005e-06,
+      "loss": 0.4931,
+      "step": 13660
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.0797959864339,
+      "learning_rate": 4.0626866091299595e-06,
+      "loss": 0.4673,
+      "step": 13661
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.074621787791056,
+      "learning_rate": 4.061329710123586e-06,
+      "loss": 0.4647,
+      "step": 13662
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.404031571572393,
+      "learning_rate": 4.059972980009522e-06,
+      "loss": 0.4751,
+      "step": 13663
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.388421726196704,
+      "learning_rate": 4.058616418826355e-06,
+      "loss": 0.4938,
+      "step": 13664
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 29.828047124322392,
+      "learning_rate": 4.05726002661266e-06,
+      "loss": 0.5372,
+      "step": 13665
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.08296897638668,
+      "learning_rate": 4.055903803407011e-06,
+      "loss": 0.4706,
+      "step": 13666
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.473197783283254,
+      "learning_rate": 4.05454774924798e-06,
+      "loss": 0.5085,
+      "step": 13667
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 28.580788632144923,
+      "learning_rate": 4.0531918641741344e-06,
+      "loss": 0.4934,
+      "step": 13668
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 21.64413508858331,
+      "learning_rate": 4.051836148224035e-06,
+      "loss": 0.5595,
+      "step": 13669
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 23.084197259882018,
+      "learning_rate": 4.050480601436237e-06,
+      "loss": 0.4821,
+      "step": 13670
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.58997925229432,
+      "learning_rate": 4.049125223849287e-06,
+      "loss": 0.4646,
+      "step": 13671
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 31.123418109979106,
+      "learning_rate": 4.047770015501734e-06,
+      "loss": 0.5362,
+      "step": 13672
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.521583464744715,
+      "learning_rate": 4.046414976432124e-06,
+      "loss": 0.4303,
+      "step": 13673
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.726309800748407,
+      "learning_rate": 4.045060106678989e-06,
+      "loss": 0.5317,
+      "step": 13674
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.419603646916862,
+      "learning_rate": 4.043705406280857e-06,
+      "loss": 0.494,
+      "step": 13675
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.07930658199789,
+      "learning_rate": 4.04235087527626e-06,
+      "loss": 0.4651,
+      "step": 13676
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.151086766839835,
+      "learning_rate": 4.040996513703721e-06,
+      "loss": 0.5083,
+      "step": 13677
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.109216745552544,
+      "learning_rate": 4.039642321601753e-06,
+      "loss": 0.4849,
+      "step": 13678
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.10280049311328,
+      "learning_rate": 4.0382882990088735e-06,
+      "loss": 0.4456,
+      "step": 13679
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 29.768161518423963,
+      "learning_rate": 4.036934445963584e-06,
+      "loss": 0.4621,
+      "step": 13680
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.009388870482887,
+      "learning_rate": 4.035580762504391e-06,
+      "loss": 0.5105,
+      "step": 13681
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.039032976600957,
+      "learning_rate": 4.034227248669794e-06,
+      "loss": 0.5935,
+      "step": 13682
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.264620074412747,
+      "learning_rate": 4.032873904498286e-06,
+      "loss": 0.5224,
+      "step": 13683
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 22.122891115448716,
+      "learning_rate": 4.031520730028348e-06,
+      "loss": 0.4953,
+      "step": 13684
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.5715564868856,
+      "learning_rate": 4.030167725298472e-06,
+      "loss": 0.5739,
+      "step": 13685
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 21.738966846193932,
+      "learning_rate": 4.028814890347134e-06,
+      "loss": 0.5055,
+      "step": 13686
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.60996665555389,
+      "learning_rate": 4.027462225212806e-06,
+      "loss": 0.4751,
+      "step": 13687
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.13213868708344,
+      "learning_rate": 4.026109729933962e-06,
+      "loss": 0.4357,
+      "step": 13688
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 23.05932294470333,
+      "learning_rate": 4.024757404549058e-06,
+      "loss": 0.5339,
+      "step": 13689
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.319212736306508,
+      "learning_rate": 4.023405249096561e-06,
+      "loss": 0.5883,
+      "step": 13690
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.859497610802075,
+      "learning_rate": 4.022053263614921e-06,
+      "loss": 0.4219,
+      "step": 13691
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 30.85810576931222,
+      "learning_rate": 4.0207014481425875e-06,
+      "loss": 0.5327,
+      "step": 13692
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.55922623288415,
+      "learning_rate": 4.019349802718011e-06,
+      "loss": 0.4878,
+      "step": 13693
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.247477699593244,
+      "learning_rate": 4.017998327379626e-06,
+      "loss": 0.4509,
+      "step": 13694
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.51693598173506,
+      "learning_rate": 4.01664702216587e-06,
+      "loss": 0.4849,
+      "step": 13695
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.843596417485822,
+      "learning_rate": 4.015295887115169e-06,
+      "loss": 0.5144,
+      "step": 13696
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.52286735209337,
+      "learning_rate": 4.013944922265953e-06,
+      "loss": 0.4234,
+      "step": 13697
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.705827313065996,
+      "learning_rate": 4.012594127656646e-06,
+      "loss": 0.4958,
+      "step": 13698
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.127258333558165,
+      "learning_rate": 4.011243503325658e-06,
+      "loss": 0.5063,
+      "step": 13699
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.312314186023546,
+      "learning_rate": 4.009893049311399e-06,
+      "loss": 0.4913,
+      "step": 13700
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.626372023357817,
+      "learning_rate": 4.008542765652277e-06,
+      "loss": 0.4502,
+      "step": 13701
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 21.61066268606607,
+      "learning_rate": 4.007192652386695e-06,
+      "loss": 0.5332,
+      "step": 13702
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.217029566634995,
+      "learning_rate": 4.005842709553053e-06,
+      "loss": 0.5055,
+      "step": 13703
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 31.327060903518884,
+      "learning_rate": 4.0044929371897325e-06,
+      "loss": 0.4788,
+      "step": 13704
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.38063009206433,
+      "learning_rate": 4.003143335335124e-06,
+      "loss": 0.4758,
+      "step": 13705
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.571766981610075,
+      "learning_rate": 4.0017939040276136e-06,
+      "loss": 0.4352,
+      "step": 13706
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.570630728357337,
+      "learning_rate": 4.000444643305577e-06,
+      "loss": 0.5377,
+      "step": 13707
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.537448221804265,
+      "learning_rate": 3.999095553207386e-06,
+      "loss": 0.4685,
+      "step": 13708
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.748541355499185,
+      "learning_rate": 3.997746633771403e-06,
+      "loss": 0.4966,
+      "step": 13709
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.882406352810595,
+      "learning_rate": 3.996397885035995e-06,
+      "loss": 0.4785,
+      "step": 13710
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.273152824164345,
+      "learning_rate": 3.9950493070395235e-06,
+      "loss": 0.4481,
+      "step": 13711
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.574758165891964,
+      "learning_rate": 3.993700899820332e-06,
+      "loss": 0.4331,
+      "step": 13712
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 21.10326420723819,
+      "learning_rate": 3.9923526634167775e-06,
+      "loss": 0.5087,
+      "step": 13713
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.482406304830853,
+      "learning_rate": 3.991004597867195e-06,
+      "loss": 0.5555,
+      "step": 13714
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.86213545143624,
+      "learning_rate": 3.989656703209931e-06,
+      "loss": 0.5258,
+      "step": 13715
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.263404000223726,
+      "learning_rate": 3.9883089794833094e-06,
+      "loss": 0.4341,
+      "step": 13716
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.18826280236325,
+      "learning_rate": 3.986961426725665e-06,
+      "loss": 0.5836,
+      "step": 13717
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 28.938851417639622,
+      "learning_rate": 3.985614044975323e-06,
+      "loss": 0.5013,
+      "step": 13718
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.860513615254497,
+      "learning_rate": 3.9842668342705956e-06,
+      "loss": 0.4221,
+      "step": 13719
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.91892328610376,
+      "learning_rate": 3.982919794649805e-06,
+      "loss": 0.5074,
+      "step": 13720
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.067444423764387,
+      "learning_rate": 3.98157292615125e-06,
+      "loss": 0.4327,
+      "step": 13721
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.54187269306718,
+      "learning_rate": 3.980226228813242e-06,
+      "loss": 0.4802,
+      "step": 13722
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 22.339117434908303,
+      "learning_rate": 3.978879702674081e-06,
+      "loss": 0.4776,
+      "step": 13723
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.754255318646006,
+      "learning_rate": 3.97753334777206e-06,
+      "loss": 0.4561,
+      "step": 13724
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.517606919861695,
+      "learning_rate": 3.976187164145463e-06,
+      "loss": 0.4778,
+      "step": 13725
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.43586719082391,
+      "learning_rate": 3.97484115183258e-06,
+      "loss": 0.4658,
+      "step": 13726
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.798569817845294,
+      "learning_rate": 3.97349531087169e-06,
+      "loss": 0.5521,
+      "step": 13727
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 25.420551036386378,
+      "learning_rate": 3.972149641301075e-06,
+      "loss": 0.4836,
+      "step": 13728
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.13754603500082,
+      "learning_rate": 3.97080414315899e-06,
+      "loss": 0.4835,
+      "step": 13729
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.881668349463343,
+      "learning_rate": 3.969458816483709e-06,
+      "loss": 0.4099,
+      "step": 13730
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.04242205769323,
+      "learning_rate": 3.968113661313492e-06,
+      "loss": 0.4811,
+      "step": 13731
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.945751120612176,
+      "learning_rate": 3.966768677686597e-06,
+      "loss": 0.5218,
+      "step": 13732
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.603341586719335,
+      "learning_rate": 3.9654238656412715e-06,
+      "loss": 0.434,
+      "step": 13733
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.39277913818106,
+      "learning_rate": 3.9640792252157564e-06,
+      "loss": 0.5095,
+      "step": 13734
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.45804126437764,
+      "learning_rate": 3.962734756448299e-06,
+      "loss": 0.4849,
+      "step": 13735
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.199613024017832,
+      "learning_rate": 3.961390459377136e-06,
+      "loss": 0.4819,
+      "step": 13736
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.847953142935125,
+      "learning_rate": 3.960046334040496e-06,
+      "loss": 0.4422,
+      "step": 13737
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 29.35578340017858,
+      "learning_rate": 3.958702380476602e-06,
+      "loss": 0.5015,
+      "step": 13738
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.535883657896086,
+      "learning_rate": 3.957358598723677e-06,
+      "loss": 0.4936,
+      "step": 13739
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.06538448052129,
+      "learning_rate": 3.956014988819944e-06,
+      "loss": 0.437,
+      "step": 13740
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 33.44518593485961,
+      "learning_rate": 3.954671550803604e-06,
+      "loss": 0.584,
+      "step": 13741
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 28.37576228800434,
+      "learning_rate": 3.953328284712873e-06,
+      "loss": 0.4868,
+      "step": 13742
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.72052935358145,
+      "learning_rate": 3.951985190585944e-06,
+      "loss": 0.4664,
+      "step": 13743
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.911912218378284,
+      "learning_rate": 3.950642268461019e-06,
+      "loss": 0.4844,
+      "step": 13744
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.67270996938472,
+      "learning_rate": 3.9492995183762925e-06,
+      "loss": 0.5117,
+      "step": 13745
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 34.96199081283332,
+      "learning_rate": 3.947956940369944e-06,
+      "loss": 0.519,
+      "step": 13746
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 25.21025821466923,
+      "learning_rate": 3.946614534480164e-06,
+      "loss": 0.5094,
+      "step": 13747
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 19.969785803733526,
+      "learning_rate": 3.945272300745122e-06,
+      "loss": 0.5673,
+      "step": 13748
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.7217589835283,
+      "learning_rate": 3.943930239202996e-06,
+      "loss": 0.4993,
+      "step": 13749
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.70408681628699,
+      "learning_rate": 3.94258834989195e-06,
+      "loss": 0.4487,
+      "step": 13750
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.773733450972227,
+      "learning_rate": 3.941246632850146e-06,
+      "loss": 0.4801,
+      "step": 13751
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.074960599819416,
+      "learning_rate": 3.939905088115749e-06,
+      "loss": 0.5416,
+      "step": 13752
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.103058469709426,
+      "learning_rate": 3.938563715726904e-06,
+      "loss": 0.4651,
+      "step": 13753
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.58381256575493,
+      "learning_rate": 3.937222515721759e-06,
+      "loss": 0.4427,
+      "step": 13754
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.944561510052417,
+      "learning_rate": 3.935881488138459e-06,
+      "loss": 0.4169,
+      "step": 13755
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.743133884370458,
+      "learning_rate": 3.934540633015141e-06,
+      "loss": 0.4329,
+      "step": 13756
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.08461541190231,
+      "learning_rate": 3.933199950389944e-06,
+      "loss": 0.4711,
+      "step": 13757
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 37.73339844539813,
+      "learning_rate": 3.931859440300991e-06,
+      "loss": 0.5231,
+      "step": 13758
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.505439799912597,
+      "learning_rate": 3.930519102786402e-06,
+      "loss": 0.4466,
+      "step": 13759
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 25.546909714888105,
+      "learning_rate": 3.9291789378843e-06,
+      "loss": 0.4311,
+      "step": 13760
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.2298276042473,
+      "learning_rate": 3.927838945632799e-06,
+      "loss": 0.5691,
+      "step": 13761
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.52777270267802,
+      "learning_rate": 3.926499126070008e-06,
+      "loss": 0.461,
+      "step": 13762
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.432323475238753,
+      "learning_rate": 3.925159479234025e-06,
+      "loss": 0.4728,
+      "step": 13763
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.988836755522414,
+      "learning_rate": 3.923820005162954e-06,
+      "loss": 0.4175,
+      "step": 13764
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.657416414800597,
+      "learning_rate": 3.92248070389489e-06,
+      "loss": 0.5605,
+      "step": 13765
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.00641292382174,
+      "learning_rate": 3.9211415754679164e-06,
+      "loss": 0.4619,
+      "step": 13766
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 23.691330852215604,
+      "learning_rate": 3.919802619920122e-06,
+      "loss": 0.492,
+      "step": 13767
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 26.29551294015667,
+      "learning_rate": 3.9184638372895835e-06,
+      "loss": 0.4238,
+      "step": 13768
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 38.15365430428977,
+      "learning_rate": 3.9171252276143745e-06,
+      "loss": 0.498,
+      "step": 13769
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 20.59322324956117,
+      "learning_rate": 3.9157867909325684e-06,
+      "loss": 0.4847,
+      "step": 13770
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.18020199795622,
+      "learning_rate": 3.914448527282225e-06,
+      "loss": 0.4729,
+      "step": 13771
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.175924663004682,
+      "learning_rate": 3.9131104367014085e-06,
+      "loss": 0.4657,
+      "step": 13772
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.820109088673199,
+      "learning_rate": 3.9117725192281664e-06,
+      "loss": 0.5235,
+      "step": 13773
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 30.926313180072196,
+      "learning_rate": 3.910434774900555e-06,
+      "loss": 0.452,
+      "step": 13774
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.734153041526948,
+      "learning_rate": 3.909097203756615e-06,
+      "loss": 0.4625,
+      "step": 13775
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.465166973454593,
+      "learning_rate": 3.907759805834387e-06,
+      "loss": 0.4865,
+      "step": 13776
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.328716967956478,
+      "learning_rate": 3.906422581171908e-06,
+      "loss": 0.4502,
+      "step": 13777
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 33.67026838819909,
+      "learning_rate": 3.905085529807208e-06,
+      "loss": 0.4552,
+      "step": 13778
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.702986331894113,
+      "learning_rate": 3.903748651778306e-06,
+      "loss": 0.5045,
+      "step": 13779
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.366622138092893,
+      "learning_rate": 3.902411947123226e-06,
+      "loss": 0.4295,
+      "step": 13780
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.022108940959637,
+      "learning_rate": 3.901075415879987e-06,
+      "loss": 0.4728,
+      "step": 13781
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 19.92704491944772,
+      "learning_rate": 3.8997390580865915e-06,
+      "loss": 0.5135,
+      "step": 13782
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.0508792601334,
+      "learning_rate": 3.898402873781052e-06,
+      "loss": 0.5274,
+      "step": 13783
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.433416757073212,
+      "learning_rate": 3.897066863001363e-06,
+      "loss": 0.4083,
+      "step": 13784
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.850326068728911,
+      "learning_rate": 3.895731025785521e-06,
+      "loss": 0.401,
+      "step": 13785
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.382674369902773,
+      "learning_rate": 3.894395362171523e-06,
+      "loss": 0.4531,
+      "step": 13786
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 23.78860367330921,
+      "learning_rate": 3.893059872197347e-06,
+      "loss": 0.5513,
+      "step": 13787
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.47919211095475,
+      "learning_rate": 3.8917245559009735e-06,
+      "loss": 0.4,
+      "step": 13788
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.180328592974686,
+      "learning_rate": 3.89038941332038e-06,
+      "loss": 0.4821,
+      "step": 13789
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 23.990397153291415,
+      "learning_rate": 3.889054444493537e-06,
+      "loss": 0.4787,
+      "step": 13790
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.76833099618584,
+      "learning_rate": 3.887719649458419e-06,
+      "loss": 0.4812,
+      "step": 13791
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.83301022059549,
+      "learning_rate": 3.8863850282529715e-06,
+      "loss": 0.5184,
+      "step": 13792
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 20.655549034598337,
+      "learning_rate": 3.8850505809151574e-06,
+      "loss": 0.5091,
+      "step": 13793
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 37.917152107933724,
+      "learning_rate": 3.883716307482928e-06,
+      "loss": 0.4166,
+      "step": 13794
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.995171467479597,
+      "learning_rate": 3.882382207994231e-06,
+      "loss": 0.4807,
+      "step": 13795
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.210267603273245,
+      "learning_rate": 3.881048282487005e-06,
+      "loss": 0.4251,
+      "step": 13796
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.194573550702966,
+      "learning_rate": 3.879714530999185e-06,
+      "loss": 0.5272,
+      "step": 13797
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 31.432147372145565,
+      "learning_rate": 3.878380953568702e-06,
+      "loss": 0.4632,
+      "step": 13798
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.33612165889988,
+      "learning_rate": 3.877047550233488e-06,
+      "loss": 0.4312,
+      "step": 13799
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.39905848497902,
+      "learning_rate": 3.875714321031455e-06,
+      "loss": 0.489,
+      "step": 13800
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.022634447605213,
+      "learning_rate": 3.874381266000529e-06,
+      "loss": 0.4236,
+      "step": 13801
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.33871689375885,
+      "learning_rate": 3.873048385178613e-06,
+      "loss": 0.4222,
+      "step": 13802
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.614947080571543,
+      "learning_rate": 3.8717156786036194e-06,
+      "loss": 0.4792,
+      "step": 13803
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.12220059720625,
+      "learning_rate": 3.870383146313444e-06,
+      "loss": 0.478,
+      "step": 13804
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.6458914932082,
+      "learning_rate": 3.869050788345985e-06,
+      "loss": 0.5189,
+      "step": 13805
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.606458162450263,
+      "learning_rate": 3.86771860473914e-06,
+      "loss": 0.5411,
+      "step": 13806
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.049167188300654,
+      "learning_rate": 3.866386595530787e-06,
+      "loss": 0.5451,
+      "step": 13807
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.226807028144304,
+      "learning_rate": 3.865054760758813e-06,
+      "loss": 0.462,
+      "step": 13808
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.703487373353195,
+      "learning_rate": 3.863723100461091e-06,
+      "loss": 0.5312,
+      "step": 13809
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.720219999366716,
+      "learning_rate": 3.8623916146754935e-06,
+      "loss": 0.4453,
+      "step": 13810
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.43698700216267,
+      "learning_rate": 3.861060303439892e-06,
+      "loss": 0.5466,
+      "step": 13811
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.113386443644814,
+      "learning_rate": 3.859729166792144e-06,
+      "loss": 0.5718,
+      "step": 13812
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.54450793873083,
+      "learning_rate": 3.858398204770104e-06,
+      "loss": 0.4871,
+      "step": 13813
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.94733574632788,
+      "learning_rate": 3.8570674174116256e-06,
+      "loss": 0.4339,
+      "step": 13814
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.756846983481807,
+      "learning_rate": 3.855736804754556e-06,
+      "loss": 0.4843,
+      "step": 13815
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.177096987878823,
+      "learning_rate": 3.854406366836745e-06,
+      "loss": 0.4191,
+      "step": 13816
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 27.469610137296737,
+      "learning_rate": 3.853076103696015e-06,
+      "loss": 0.4479,
+      "step": 13817
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.972465716415062,
+      "learning_rate": 3.851746015370204e-06,
+      "loss": 0.4721,
+      "step": 13818
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.57357047932348,
+      "learning_rate": 3.8504161018971395e-06,
+      "loss": 0.532,
+      "step": 13819
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.066666471030683,
+      "learning_rate": 3.849086363314648e-06,
+      "loss": 0.4783,
+      "step": 13820
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.251182363377925,
+      "learning_rate": 3.847756799660543e-06,
+      "loss": 0.5119,
+      "step": 13821
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.0781719236785,
+      "learning_rate": 3.84642741097263e-06,
+      "loss": 0.505,
+      "step": 13822
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.662032923425876,
+      "learning_rate": 3.845098197288724e-06,
+      "loss": 0.4965,
+      "step": 13823
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.116391376852972,
+      "learning_rate": 3.843769158646626e-06,
+      "loss": 0.4629,
+      "step": 13824
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.739721609921304,
+      "learning_rate": 3.842440295084133e-06,
+      "loss": 0.4611,
+      "step": 13825
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.773359891389546,
+      "learning_rate": 3.841111606639032e-06,
+      "loss": 0.5289,
+      "step": 13826
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.240824536072004,
+      "learning_rate": 3.839783093349114e-06,
+      "loss": 0.5214,
+      "step": 13827
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.626175148974685,
+      "learning_rate": 3.838454755252166e-06,
+      "loss": 0.5176,
+      "step": 13828
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 24.017673629666405,
+      "learning_rate": 3.837126592385956e-06,
+      "loss": 0.5024,
+      "step": 13829
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.998351240671198,
+      "learning_rate": 3.8357986047882635e-06,
+      "loss": 0.4954,
+      "step": 13830
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.535264263659025,
+      "learning_rate": 3.8344707924968494e-06,
+      "loss": 0.4635,
+      "step": 13831
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 29.739895736581154,
+      "learning_rate": 3.833143155549479e-06,
+      "loss": 0.5204,
+      "step": 13832
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 26.93640113567417,
+      "learning_rate": 3.831815693983915e-06,
+      "loss": 0.5732,
+      "step": 13833
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 29.060360410338088,
+      "learning_rate": 3.830488407837899e-06,
+      "loss": 0.4942,
+      "step": 13834
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 34.74055317559641,
+      "learning_rate": 3.829161297149187e-06,
+      "loss": 0.4653,
+      "step": 13835
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.806554128051562,
+      "learning_rate": 3.8278343619555155e-06,
+      "loss": 0.5492,
+      "step": 13836
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 26.49196965975394,
+      "learning_rate": 3.826507602294628e-06,
+      "loss": 0.5491,
+      "step": 13837
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.616057039250588,
+      "learning_rate": 3.825181018204248e-06,
+      "loss": 0.5572,
+      "step": 13838
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.009056243128285,
+      "learning_rate": 3.823854609722108e-06,
+      "loss": 0.5575,
+      "step": 13839
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.919704019155592,
+      "learning_rate": 3.8225283768859345e-06,
+      "loss": 0.5026,
+      "step": 13840
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.00686820611522,
+      "learning_rate": 3.8212023197334395e-06,
+      "loss": 0.5281,
+      "step": 13841
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.563345590388938,
+      "learning_rate": 3.819876438302332e-06,
+      "loss": 0.5126,
+      "step": 13842
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.395522280632267,
+      "learning_rate": 3.8185507326303236e-06,
+      "loss": 0.4583,
+      "step": 13843
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.343331406311542,
+      "learning_rate": 3.817225202755117e-06,
+      "loss": 0.4793,
+      "step": 13844
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.359305721413385,
+      "learning_rate": 3.815899848714412e-06,
+      "loss": 0.4604,
+      "step": 13845
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 35.65342793509411,
+      "learning_rate": 3.8145746705458976e-06,
+      "loss": 0.5287,
+      "step": 13846
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 24.1185799540555,
+      "learning_rate": 3.8132496682872576e-06,
+      "loss": 0.4408,
+      "step": 13847
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 20.168136318174888,
+      "learning_rate": 3.811924841976178e-06,
+      "loss": 0.4905,
+      "step": 13848
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 30.395899087003755,
+      "learning_rate": 3.81060019165034e-06,
+      "loss": 0.4961,
+      "step": 13849
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 26.04443045554107,
+      "learning_rate": 3.8092757173474114e-06,
+      "loss": 0.4972,
+      "step": 13850
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.447656623516373,
+      "learning_rate": 3.807951419105056e-06,
+      "loss": 0.4863,
+      "step": 13851
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.089494493786223,
+      "learning_rate": 3.8066272969609407e-06,
+      "loss": 0.499,
+      "step": 13852
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.885366479673738,
+      "learning_rate": 3.8053033509527214e-06,
+      "loss": 0.4918,
+      "step": 13853
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.0940775160618,
+      "learning_rate": 3.8039795811180547e-06,
+      "loss": 0.5497,
+      "step": 13854
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 30.623567587576517,
+      "learning_rate": 3.8026559874945844e-06,
+      "loss": 0.4488,
+      "step": 13855
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.204946987146617,
+      "learning_rate": 3.801332570119949e-06,
+      "loss": 0.523,
+      "step": 13856
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 33.90051573857576,
+      "learning_rate": 3.8000093290317886e-06,
+      "loss": 0.5347,
+      "step": 13857
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 27.360146308496923,
+      "learning_rate": 3.7986862642677402e-06,
+      "loss": 0.5333,
+      "step": 13858
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.113558550125575,
+      "learning_rate": 3.7973633758654225e-06,
+      "loss": 0.5234,
+      "step": 13859
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.666177408655454,
+      "learning_rate": 3.796040663862467e-06,
+      "loss": 0.5465,
+      "step": 13860
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 24.423976211983565,
+      "learning_rate": 3.7947181282964806e-06,
+      "loss": 0.5468,
+      "step": 13861
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.03031949267016,
+      "learning_rate": 3.793395769205085e-06,
+      "loss": 0.4362,
+      "step": 13862
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 27.25323634877798,
+      "learning_rate": 3.7920735866258796e-06,
+      "loss": 0.4984,
+      "step": 13863
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.221239961071152,
+      "learning_rate": 3.790751580596469e-06,
+      "loss": 0.5372,
+      "step": 13864
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 27.802734656379144,
+      "learning_rate": 3.7894297511544553e-06,
+      "loss": 0.5342,
+      "step": 13865
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.28318000315127,
+      "learning_rate": 3.7881080983374263e-06,
+      "loss": 0.4139,
+      "step": 13866
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.103619880902194,
+      "learning_rate": 3.786786622182966e-06,
+      "loss": 0.5184,
+      "step": 13867
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.232947886224364,
+      "learning_rate": 3.7854653227286586e-06,
+      "loss": 0.4749,
+      "step": 13868
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.69926999905366,
+      "learning_rate": 3.784144200012083e-06,
+      "loss": 0.5395,
+      "step": 13869
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 26.86781979166525,
+      "learning_rate": 3.782823254070813e-06,
+      "loss": 0.4364,
+      "step": 13870
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.672117333855482,
+      "learning_rate": 3.781502484942413e-06,
+      "loss": 0.5121,
+      "step": 13871
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 30.376665610301327,
+      "learning_rate": 3.7801818926644416e-06,
+      "loss": 0.5527,
+      "step": 13872
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.870592599040634,
+      "learning_rate": 3.778861477274458e-06,
+      "loss": 0.4935,
+      "step": 13873
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.214417900681852,
+      "learning_rate": 3.7775412388100186e-06,
+      "loss": 0.4581,
+      "step": 13874
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.50413907142526,
+      "learning_rate": 3.7762211773086656e-06,
+      "loss": 0.4296,
+      "step": 13875
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.086538747282624,
+      "learning_rate": 3.7749012928079387e-06,
+      "loss": 0.4754,
+      "step": 13876
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.153668502430836,
+      "learning_rate": 3.773581585345377e-06,
+      "loss": 0.518,
+      "step": 13877
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.565319608962314,
+      "learning_rate": 3.772262054958513e-06,
+      "loss": 0.4488,
+      "step": 13878
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.002450101918896,
+      "learning_rate": 3.77094270168488e-06,
+      "loss": 0.5108,
+      "step": 13879
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 21.186942853466455,
+      "learning_rate": 3.769623525561986e-06,
+      "loss": 0.4658,
+      "step": 13880
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.06550099585411,
+      "learning_rate": 3.768304526627354e-06,
+      "loss": 0.4874,
+      "step": 13881
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.429163810757675,
+      "learning_rate": 3.7669857049184953e-06,
+      "loss": 0.5271,
+      "step": 13882
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.727649414142682,
+      "learning_rate": 3.765667060472922e-06,
+      "loss": 0.4413,
+      "step": 13883
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 27.112849995071905,
+      "learning_rate": 3.764348593328129e-06,
+      "loss": 0.4824,
+      "step": 13884
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 21.889569403433157,
+      "learning_rate": 3.763030303521611e-06,
+      "loss": 0.5256,
+      "step": 13885
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.050841578719153,
+      "learning_rate": 3.7617121910908627e-06,
+      "loss": 0.4345,
+      "step": 13886
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.25122012117449,
+      "learning_rate": 3.760394256073374e-06,
+      "loss": 0.4876,
+      "step": 13887
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.327057029496785,
+      "learning_rate": 3.7590764985066187e-06,
+      "loss": 0.4769,
+      "step": 13888
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.126113154129019,
+      "learning_rate": 3.7577589184280817e-06,
+      "loss": 0.4417,
+      "step": 13889
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.267813302315506,
+      "learning_rate": 3.7564415158752244e-06,
+      "loss": 0.5399,
+      "step": 13890
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 15.934242441075234,
+      "learning_rate": 3.755124290885522e-06,
+      "loss": 0.4572,
+      "step": 13891
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 25.72777193854207,
+      "learning_rate": 3.753807243496429e-06,
+      "loss": 0.5363,
+      "step": 13892
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.68732316912678,
+      "learning_rate": 3.752490373745403e-06,
+      "loss": 0.4636,
+      "step": 13893
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.195430534114255,
+      "learning_rate": 3.751173681669901e-06,
+      "loss": 0.4551,
+      "step": 13894
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 25.770825307151473,
+      "learning_rate": 3.749857167307359e-06,
+      "loss": 0.492,
+      "step": 13895
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.805678432736446,
+      "learning_rate": 3.7485408306952263e-06,
+      "loss": 0.4557,
+      "step": 13896
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.36118923347034,
+      "learning_rate": 3.7472246718709325e-06,
+      "loss": 0.4547,
+      "step": 13897
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.697549062766893,
+      "learning_rate": 3.7459086908719124e-06,
+      "loss": 0.4868,
+      "step": 13898
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.198767491674225,
+      "learning_rate": 3.744592887735592e-06,
+      "loss": 0.4764,
+      "step": 13899
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 26.00196618316421,
+      "learning_rate": 3.7432772624993917e-06,
+      "loss": 0.4639,
+      "step": 13900
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 25.156482588836,
+      "learning_rate": 3.741961815200723e-06,
+      "loss": 0.4867,
+      "step": 13901
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.34395287012653,
+      "learning_rate": 3.740646545876998e-06,
+      "loss": 0.493,
+      "step": 13902
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.052540452182416,
+      "learning_rate": 3.7393314545656243e-06,
+      "loss": 0.5237,
+      "step": 13903
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 27.072504637392218,
+      "learning_rate": 3.7380165413040092e-06,
+      "loss": 0.4997,
+      "step": 13904
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.62259429581696,
+      "learning_rate": 3.7367018061295335e-06,
+      "loss": 0.5068,
+      "step": 13905
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.021429367556948,
+      "learning_rate": 3.735387249079594e-06,
+      "loss": 0.415,
+      "step": 13906
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.917348662659062,
+      "learning_rate": 3.7340728701915764e-06,
+      "loss": 0.4499,
+      "step": 13907
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 15.856144422902132,
+      "learning_rate": 3.732758669502864e-06,
+      "loss": 0.4747,
+      "step": 13908
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.802182053117328,
+      "learning_rate": 3.731444647050829e-06,
+      "loss": 0.478,
+      "step": 13909
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.648342386425874,
+      "learning_rate": 3.730130802872839e-06,
+      "loss": 0.5104,
+      "step": 13910
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.40224088064852,
+      "learning_rate": 3.72881713700626e-06,
+      "loss": 0.4798,
+      "step": 13911
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.22119826652589,
+      "learning_rate": 3.7275036494884563e-06,
+      "loss": 0.4925,
+      "step": 13912
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.040495374066186,
+      "learning_rate": 3.7261903403567767e-06,
+      "loss": 0.4351,
+      "step": 13913
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.476561855614552,
+      "learning_rate": 3.7248772096485773e-06,
+      "loss": 0.4969,
+      "step": 13914
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 29.656008212929844,
+      "learning_rate": 3.7235642574011955e-06,
+      "loss": 0.5233,
+      "step": 13915
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.42924940373642,
+      "learning_rate": 3.7222514836519784e-06,
+      "loss": 0.418,
+      "step": 13916
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.142868668758883,
+      "learning_rate": 3.7209388884382526e-06,
+      "loss": 0.4841,
+      "step": 13917
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.182305342595264,
+      "learning_rate": 3.719626471797352e-06,
+      "loss": 0.5755,
+      "step": 13918
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.20518447272807,
+      "learning_rate": 3.7183142337666045e-06,
+      "loss": 0.4834,
+      "step": 13919
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.57457942554384,
+      "learning_rate": 3.7170021743833216e-06,
+      "loss": 0.3893,
+      "step": 13920
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 21.992605063775144,
+      "learning_rate": 3.715690293684825e-06,
+      "loss": 0.4778,
+      "step": 13921
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 15.958230542950803,
+      "learning_rate": 3.7143785917084163e-06,
+      "loss": 0.3823,
+      "step": 13922
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.730368526606386,
+      "learning_rate": 3.7130670684914083e-06,
+      "loss": 0.5007,
+      "step": 13923
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 28.378319004228697,
+      "learning_rate": 3.71175572407109e-06,
+      "loss": 0.5193,
+      "step": 13924
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.744855643999525,
+      "learning_rate": 3.7104445584847647e-06,
+      "loss": 0.442,
+      "step": 13925
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.921790652791913,
+      "learning_rate": 3.709133571769713e-06,
+      "loss": 0.4352,
+      "step": 13926
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 31.174525690791548,
+      "learning_rate": 3.7078227639632234e-06,
+      "loss": 0.5035,
+      "step": 13927
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.620553649009429,
+      "learning_rate": 3.7065121351025758e-06,
+      "loss": 0.4392,
+      "step": 13928
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 29.90004100392075,
+      "learning_rate": 3.705201685225043e-06,
+      "loss": 0.6798,
+      "step": 13929
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.527296288113387,
+      "learning_rate": 3.7038914143678874e-06,
+      "loss": 0.4699,
+      "step": 13930
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 22.30784560191277,
+      "learning_rate": 3.702581322568376e-06,
+      "loss": 0.5408,
+      "step": 13931
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.861843915730232,
+      "learning_rate": 3.701271409863769e-06,
+      "loss": 0.4325,
+      "step": 13932
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.81794535048324,
+      "learning_rate": 3.6999616762913226e-06,
+      "loss": 0.4409,
+      "step": 13933
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 22.7578377024586,
+      "learning_rate": 3.6986521218882798e-06,
+      "loss": 0.5059,
+      "step": 13934
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.806213747355129,
+      "learning_rate": 3.697342746691881e-06,
+      "loss": 0.4334,
+      "step": 13935
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 26.173465803394198,
+      "learning_rate": 3.6960335507393672e-06,
+      "loss": 0.5271,
+      "step": 13936
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.64270022441161,
+      "learning_rate": 3.694724534067976e-06,
+      "loss": 0.4328,
+      "step": 13937
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.916564799516234,
+      "learning_rate": 3.693415696714929e-06,
+      "loss": 0.45,
+      "step": 13938
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.592918441737208,
+      "learning_rate": 3.6921070387174484e-06,
+      "loss": 0.5372,
+      "step": 13939
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.651884181892083,
+      "learning_rate": 3.690798560112754e-06,
+      "loss": 0.4198,
+      "step": 13940
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.509405655447036,
+      "learning_rate": 3.6894902609380566e-06,
+      "loss": 0.5016,
+      "step": 13941
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.27971133497859,
+      "learning_rate": 3.6881821412305683e-06,
+      "loss": 0.4705,
+      "step": 13942
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.39685677519553,
+      "learning_rate": 3.686874201027487e-06,
+      "loss": 0.4337,
+      "step": 13943
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.002955028784868,
+      "learning_rate": 3.6855664403660073e-06,
+      "loss": 0.4793,
+      "step": 13944
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 32.031929770676626,
+      "learning_rate": 3.684258859283325e-06,
+      "loss": 0.4854,
+      "step": 13945
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.166517287957952,
+      "learning_rate": 3.6829514578166293e-06,
+      "loss": 0.433,
+      "step": 13946
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 26.03985727769971,
+      "learning_rate": 3.6816442360030946e-06,
+      "loss": 0.5459,
+      "step": 13947
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.580314624301515,
+      "learning_rate": 3.680337193879906e-06,
+      "loss": 0.4546,
+      "step": 13948
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.81311772462806,
+      "learning_rate": 3.679030331484227e-06,
+      "loss": 0.4476,
+      "step": 13949
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.472065570750203,
+      "learning_rate": 3.677723648853231e-06,
+      "loss": 0.4231,
+      "step": 13950
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.031267716612696,
+      "learning_rate": 3.676417146024073e-06,
+      "loss": 0.4422,
+      "step": 13951
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.97955650829259,
+      "learning_rate": 3.6751108230339116e-06,
+      "loss": 0.4984,
+      "step": 13952
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.24901122721355,
+      "learning_rate": 3.673804679919901e-06,
+      "loss": 0.5214,
+      "step": 13953
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.329999098499357,
+      "learning_rate": 3.672498716719186e-06,
+      "loss": 0.4278,
+      "step": 13954
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 29.249313411642508,
+      "learning_rate": 3.6711929334689024e-06,
+      "loss": 0.5053,
+      "step": 13955
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.4968305344219,
+      "learning_rate": 3.6698873302061888e-06,
+      "loss": 0.505,
+      "step": 13956
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.45472467240973,
+      "learning_rate": 3.668581906968177e-06,
+      "loss": 0.4573,
+      "step": 13957
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.04060933564639,
+      "learning_rate": 3.6672766637919967e-06,
+      "loss": 0.4659,
+      "step": 13958
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.05959855520941,
+      "learning_rate": 3.6659716007147616e-06,
+      "loss": 0.5075,
+      "step": 13959
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.095191734239712,
+      "learning_rate": 3.664666717773586e-06,
+      "loss": 0.5157,
+      "step": 13960
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.977994121589116,
+      "learning_rate": 3.6633620150055835e-06,
+      "loss": 0.4636,
+      "step": 13961
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.195037928153063,
+      "learning_rate": 3.6620574924478615e-06,
+      "loss": 0.4557,
+      "step": 13962
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.064549108242435,
+      "learning_rate": 3.660753150137517e-06,
+      "loss": 0.4572,
+      "step": 13963
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.54747176460059,
+      "learning_rate": 3.65944898811164e-06,
+      "loss": 0.5224,
+      "step": 13964
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 25.491890843892687,
+      "learning_rate": 3.6581450064073265e-06,
+      "loss": 0.4815,
+      "step": 13965
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.26828795487254,
+      "learning_rate": 3.6568412050616574e-06,
+      "loss": 0.5211,
+      "step": 13966
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.160762294224657,
+      "learning_rate": 3.655537584111718e-06,
+      "loss": 0.4263,
+      "step": 13967
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.093935823248746,
+      "learning_rate": 3.654234143594578e-06,
+      "loss": 0.458,
+      "step": 13968
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.317338236743893,
+      "learning_rate": 3.652930883547304e-06,
+      "loss": 0.4909,
+      "step": 13969
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.61116934100178,
+      "learning_rate": 3.651627804006963e-06,
+      "loss": 0.4267,
+      "step": 13970
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.55893812941986,
+      "learning_rate": 3.650324905010616e-06,
+      "loss": 0.4397,
+      "step": 13971
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 27.044436155293575,
+      "learning_rate": 3.6490221865953146e-06,
+      "loss": 0.5589,
+      "step": 13972
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.380076979906473,
+      "learning_rate": 3.6477196487981036e-06,
+      "loss": 0.5088,
+      "step": 13973
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.389732167335104,
+      "learning_rate": 3.6464172916560305e-06,
+      "loss": 0.4705,
+      "step": 13974
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.589091577782902,
+      "learning_rate": 3.645115115206136e-06,
+      "loss": 0.4937,
+      "step": 13975
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 29.01265602554712,
+      "learning_rate": 3.643813119485445e-06,
+      "loss": 0.4521,
+      "step": 13976
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.64416319719017,
+      "learning_rate": 3.642511304530996e-06,
+      "loss": 0.4766,
+      "step": 13977
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.352320732961957,
+      "learning_rate": 3.641209670379803e-06,
+      "loss": 0.4317,
+      "step": 13978
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.73647408334706,
+      "learning_rate": 3.6399082170688893e-06,
+      "loss": 0.5278,
+      "step": 13979
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.12766188186991,
+      "learning_rate": 3.6386069446352632e-06,
+      "loss": 0.5229,
+      "step": 13980
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.355657629092867,
+      "learning_rate": 3.6373058531159332e-06,
+      "loss": 0.4636,
+      "step": 13981
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.65096667990874,
+      "learning_rate": 3.636004942547907e-06,
+      "loss": 0.4917,
+      "step": 13982
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.37103479607225,
+      "learning_rate": 3.634704212968174e-06,
+      "loss": 0.4963,
+      "step": 13983
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.788994113711732,
+      "learning_rate": 3.6334036644137328e-06,
+      "loss": 0.4964,
+      "step": 13984
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.224723738366823,
+      "learning_rate": 3.632103296921563e-06,
+      "loss": 0.4695,
+      "step": 13985
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.8523101380616,
+      "learning_rate": 3.630803110528651e-06,
+      "loss": 0.5534,
+      "step": 13986
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.84165634201634,
+      "learning_rate": 3.6295031052719766e-06,
+      "loss": 0.5135,
+      "step": 13987
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.956733448610258,
+      "learning_rate": 3.628203281188506e-06,
+      "loss": 0.4013,
+      "step": 13988
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 35.406843996915335,
+      "learning_rate": 3.6269036383152032e-06,
+      "loss": 0.5603,
+      "step": 13989
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 23.49508580228283,
+      "learning_rate": 3.6256041766890337e-06,
+      "loss": 0.4794,
+      "step": 13990
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 25.701856949435804,
+      "learning_rate": 3.624304896346953e-06,
+      "loss": 0.43,
+      "step": 13991
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 26.33248877711425,
+      "learning_rate": 3.623005797325917e-06,
+      "loss": 0.4812,
+      "step": 13992
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.093145042456435,
+      "learning_rate": 3.6217068796628597e-06,
+      "loss": 0.4626,
+      "step": 13993
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.07658711791868,
+      "learning_rate": 3.6204081433947268e-06,
+      "loss": 0.4784,
+      "step": 13994
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.614420989407897,
+      "learning_rate": 3.619109588558455e-06,
+      "loss": 0.5323,
+      "step": 13995
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 23.058644835738473,
+      "learning_rate": 3.6178112151909763e-06,
+      "loss": 0.4095,
+      "step": 13996
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.812510704389624,
+      "learning_rate": 3.6165130233292133e-06,
+      "loss": 0.5238,
+      "step": 13997
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 23.198471185753913,
+      "learning_rate": 3.6152150130100825e-06,
+      "loss": 0.5558,
+      "step": 13998
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 25.21144552460755,
+      "learning_rate": 3.6139171842705024e-06,
+      "loss": 0.5057,
+      "step": 13999
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.192608312420063,
+      "learning_rate": 3.612619537147385e-06,
+      "loss": 0.5618,
+      "step": 14000
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.8332970868708,
+      "learning_rate": 3.611322071677628e-06,
+      "loss": 0.477,
+      "step": 14001
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.833045054788366,
+      "learning_rate": 3.6100247878981365e-06,
+      "loss": 0.389,
+      "step": 14002
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.222395402544583,
+      "learning_rate": 3.6087276858458e-06,
+      "loss": 0.4335,
+      "step": 14003
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.119977674108185,
+      "learning_rate": 3.607430765557508e-06,
+      "loss": 0.5015,
+      "step": 14004
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.618093805031508,
+      "learning_rate": 3.606134027070151e-06,
+      "loss": 0.4526,
+      "step": 14005
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.710519401435027,
+      "learning_rate": 3.604837470420596e-06,
+      "loss": 0.4316,
+      "step": 14006
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.891181656083717,
+      "learning_rate": 3.603541095645727e-06,
+      "loss": 0.4568,
+      "step": 14007
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 27.573844077065104,
+      "learning_rate": 3.6022449027824035e-06,
+      "loss": 0.4954,
+      "step": 14008
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.6284477473315,
+      "learning_rate": 3.600948891867496e-06,
+      "loss": 0.5733,
+      "step": 14009
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.12823187197001,
+      "learning_rate": 3.599653062937856e-06,
+      "loss": 0.5448,
+      "step": 14010
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.564854645638107,
+      "learning_rate": 3.598357416030338e-06,
+      "loss": 0.4427,
+      "step": 14011
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 21.107277889542466,
+      "learning_rate": 3.597061951181794e-06,
+      "loss": 0.498,
+      "step": 14012
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.69182937783355,
+      "learning_rate": 3.5957666684290626e-06,
+      "loss": 0.508,
+      "step": 14013
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.85321189709224,
+      "learning_rate": 3.594471567808977e-06,
+      "loss": 0.4794,
+      "step": 14014
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.42785927445323,
+      "learning_rate": 3.593176649358373e-06,
+      "loss": 0.4412,
+      "step": 14015
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 34.66148731861961,
+      "learning_rate": 3.591881913114077e-06,
+      "loss": 0.4727,
+      "step": 14016
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.502460511120397,
+      "learning_rate": 3.590587359112917e-06,
+      "loss": 0.4875,
+      "step": 14017
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.031174962728825,
+      "learning_rate": 3.589292987391697e-06,
+      "loss": 0.4412,
+      "step": 14018
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.394031926978407,
+      "learning_rate": 3.5879987979872342e-06,
+      "loss": 0.5227,
+      "step": 14019
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.69219731579314,
+      "learning_rate": 3.586704790936335e-06,
+      "loss": 0.4277,
+      "step": 14020
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.419275426006198,
+      "learning_rate": 3.5854109662758043e-06,
+      "loss": 0.4646,
+      "step": 14021
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.173975894212102,
+      "learning_rate": 3.584117324042431e-06,
+      "loss": 0.4418,
+      "step": 14022
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.775346166971545,
+      "learning_rate": 3.5828238642730063e-06,
+      "loss": 0.5118,
+      "step": 14023
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.174961773109384,
+      "learning_rate": 3.581530587004316e-06,
+      "loss": 0.4566,
+      "step": 14024
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.649343930034874,
+      "learning_rate": 3.580237492273144e-06,
+      "loss": 0.5392,
+      "step": 14025
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.97642455476868,
+      "learning_rate": 3.578944580116264e-06,
+      "loss": 0.4857,
+      "step": 14026
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 29.003735713778482,
+      "learning_rate": 3.5776518505704383e-06,
+      "loss": 0.5662,
+      "step": 14027
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 21.726996247050618,
+      "learning_rate": 3.5763593036724387e-06,
+      "loss": 0.4362,
+      "step": 14028
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.769312331371186,
+      "learning_rate": 3.575066939459022e-06,
+      "loss": 0.5074,
+      "step": 14029
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.008584782554852,
+      "learning_rate": 3.5737747579669468e-06,
+      "loss": 0.4826,
+      "step": 14030
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 31.0675369195081,
+      "learning_rate": 3.572482759232958e-06,
+      "loss": 0.5604,
+      "step": 14031
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.034944622705922,
+      "learning_rate": 3.5711909432937964e-06,
+      "loss": 0.4204,
+      "step": 14032
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.918155319393698,
+      "learning_rate": 3.5698993101862034e-06,
+      "loss": 0.4572,
+      "step": 14033
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.828490565598276,
+      "learning_rate": 3.5686078599469166e-06,
+      "loss": 0.4274,
+      "step": 14034
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.154704688221837,
+      "learning_rate": 3.5673165926126553e-06,
+      "loss": 0.4479,
+      "step": 14035
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.335644179274187,
+      "learning_rate": 3.566025508220151e-06,
+      "loss": 0.4564,
+      "step": 14036
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.158496724080653,
+      "learning_rate": 3.5647346068061152e-06,
+      "loss": 0.4476,
+      "step": 14037
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.988967250210617,
+      "learning_rate": 3.5634438884072653e-06,
+      "loss": 0.4973,
+      "step": 14038
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.731892454006914,
+      "learning_rate": 3.5621533530603024e-06,
+      "loss": 0.5157,
+      "step": 14039
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.45232979270624,
+      "learning_rate": 3.560863000801932e-06,
+      "loss": 0.4275,
+      "step": 14040
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 21.361726270071014,
+      "learning_rate": 3.559572831668855e-06,
+      "loss": 0.4794,
+      "step": 14041
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.57647011807346,
+      "learning_rate": 3.558282845697758e-06,
+      "loss": 0.4836,
+      "step": 14042
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.13212645690718,
+      "learning_rate": 3.5569930429253263e-06,
+      "loss": 0.4892,
+      "step": 14043
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.072692573341673,
+      "learning_rate": 3.5557034233882426e-06,
+      "loss": 0.4775,
+      "step": 14044
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.109924302234905,
+      "learning_rate": 3.5544139871231842e-06,
+      "loss": 0.4951,
+      "step": 14045
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 26.79231211125581,
+      "learning_rate": 3.5531247341668253e-06,
+      "loss": 0.4937,
+      "step": 14046
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.738250184690546,
+      "learning_rate": 3.551835664555827e-06,
+      "loss": 0.5187,
+      "step": 14047
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.87518642146174,
+      "learning_rate": 3.5505467783268465e-06,
+      "loss": 0.3845,
+      "step": 14048
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.32296510824594,
+      "learning_rate": 3.549258075516544e-06,
+      "loss": 0.4773,
+      "step": 14049
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.356241592345082,
+      "learning_rate": 3.54796955616157e-06,
+      "loss": 0.559,
+      "step": 14050
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.505514144658026,
+      "learning_rate": 3.546681220298569e-06,
+      "loss": 0.4726,
+      "step": 14051
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 27.18552797005648,
+      "learning_rate": 3.545393067964176e-06,
+      "loss": 0.4635,
+      "step": 14052
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.11100749671277,
+      "learning_rate": 3.5441050991950264e-06,
+      "loss": 0.4708,
+      "step": 14053
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.290062820689727,
+      "learning_rate": 3.5428173140277523e-06,
+      "loss": 0.4224,
+      "step": 14054
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.972685798429524,
+      "learning_rate": 3.54152971249898e-06,
+      "loss": 0.4253,
+      "step": 14055
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.22771292387033,
+      "learning_rate": 3.5402422946453242e-06,
+      "loss": 0.4478,
+      "step": 14056
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.75963982291962,
+      "learning_rate": 3.5389550605033963e-06,
+      "loss": 0.4764,
+      "step": 14057
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.94142548189347,
+      "learning_rate": 3.537668010109805e-06,
+      "loss": 0.4443,
+      "step": 14058
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.8522734309868,
+      "learning_rate": 3.5363811435011598e-06,
+      "loss": 0.4693,
+      "step": 14059
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.312032718622437,
+      "learning_rate": 3.53509446071405e-06,
+      "loss": 0.4541,
+      "step": 14060
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.733736502176065,
+      "learning_rate": 3.5338079617850762e-06,
+      "loss": 0.5451,
+      "step": 14061
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.812510162470874,
+      "learning_rate": 3.5325216467508172e-06,
+      "loss": 0.4658,
+      "step": 14062
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 26.776922491698524,
+      "learning_rate": 3.5312355156478616e-06,
+      "loss": 0.5021,
+      "step": 14063
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.546906582921906,
+      "learning_rate": 3.5299495685127816e-06,
+      "loss": 0.5025,
+      "step": 14064
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.470269510054994,
+      "learning_rate": 3.528663805382151e-06,
+      "loss": 0.4629,
+      "step": 14065
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.87054624667286,
+      "learning_rate": 3.5273782262925392e-06,
+      "loss": 0.4486,
+      "step": 14066
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.641677316730252,
+      "learning_rate": 3.5260928312805042e-06,
+      "loss": 0.4712,
+      "step": 14067
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.922997956361876,
+      "learning_rate": 3.5248076203825987e-06,
+      "loss": 0.4194,
+      "step": 14068
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.050922270157553,
+      "learning_rate": 3.5235225936353767e-06,
+      "loss": 0.4295,
+      "step": 14069
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.062036976125352,
+      "learning_rate": 3.522237751075387e-06,
+      "loss": 0.5112,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.444524533294278,
+      "learning_rate": 3.5209530927391632e-06,
+      "loss": 0.4451,
+      "step": 14071
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 25.213072462088874,
+      "learning_rate": 3.5196686186632457e-06,
+      "loss": 0.4932,
+      "step": 14072
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 30.96334830925032,
+      "learning_rate": 3.518384328884159e-06,
+      "loss": 0.4499,
+      "step": 14073
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.86835308699928,
+      "learning_rate": 3.517100223438431e-06,
+      "loss": 0.5184,
+      "step": 14074
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.84810799443279,
+      "learning_rate": 3.5158163023625825e-06,
+      "loss": 0.5209,
+      "step": 14075
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.432368447763544,
+      "learning_rate": 3.514532565693126e-06,
+      "loss": 0.4604,
+      "step": 14076
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.796697835065537,
+      "learning_rate": 3.513249013466565e-06,
+      "loss": 0.5287,
+      "step": 14077
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 23.644830117321195,
+      "learning_rate": 3.5119656457194086e-06,
+      "loss": 0.5124,
+      "step": 14078
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.887887654731266,
+      "learning_rate": 3.510682462488153e-06,
+      "loss": 0.4844,
+      "step": 14079
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 27.712158869350212,
+      "learning_rate": 3.5093994638093e-06,
+      "loss": 0.4934,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.37520570542549,
+      "learning_rate": 3.5081166497193208e-06,
+      "loss": 0.4673,
+      "step": 14081
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.733917116285223,
+      "learning_rate": 3.506834020254708e-06,
+      "loss": 0.4224,
+      "step": 14082
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 28.46560013468761,
+      "learning_rate": 3.5055515754519363e-06,
+      "loss": 0.5899,
+      "step": 14083
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.0879471205812,
+      "learning_rate": 3.504269315347483e-06,
+      "loss": 0.4107,
+      "step": 14084
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 23.056191021463018,
+      "learning_rate": 3.50298723997781e-06,
+      "loss": 0.5308,
+      "step": 14085
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.03328130569702,
+      "learning_rate": 3.501705349379375e-06,
+      "loss": 0.5034,
+      "step": 14086
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.221858021841893,
+      "learning_rate": 3.5004236435886395e-06,
+      "loss": 0.4313,
+      "step": 14087
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 28.614419932964108,
+      "learning_rate": 3.4991421226420563e-06,
+      "loss": 0.5076,
+      "step": 14088
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.563292301814137,
+      "learning_rate": 3.497860786576065e-06,
+      "loss": 0.4621,
+      "step": 14089
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.712796202145565,
+      "learning_rate": 3.496579635427113e-06,
+      "loss": 0.5354,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.276930815999908,
+      "learning_rate": 3.4952986692316283e-06,
+      "loss": 0.4711,
+      "step": 14091
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.864254241709762,
+      "learning_rate": 3.4940178880260446e-06,
+      "loss": 0.4399,
+      "step": 14092
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.459176633574064,
+      "learning_rate": 3.4927372918467893e-06,
+      "loss": 0.4139,
+      "step": 14093
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.626736464080295,
+      "learning_rate": 3.491456880730275e-06,
+      "loss": 0.4625,
+      "step": 14094
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.652212032909171,
+      "learning_rate": 3.4901766547129234e-06,
+      "loss": 0.3932,
+      "step": 14095
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.399589499208492,
+      "learning_rate": 3.4888966138311363e-06,
+      "loss": 0.4526,
+      "step": 14096
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 23.10224476859147,
+      "learning_rate": 3.487616758121324e-06,
+      "loss": 0.501,
+      "step": 14097
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 26.427406652385116,
+      "learning_rate": 3.4863370876198777e-06,
+      "loss": 0.4978,
+      "step": 14098
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.963691570911946,
+      "learning_rate": 3.4850576023631942e-06,
+      "loss": 0.4076,
+      "step": 14099
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.642865323550817,
+      "learning_rate": 3.483778302387665e-06,
+      "loss": 0.4195,
+      "step": 14100
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.715794370707034,
+      "learning_rate": 3.4824991877296687e-06,
+      "loss": 0.5094,
+      "step": 14101
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.806229377481852,
+      "learning_rate": 3.4812202584255805e-06,
+      "loss": 0.5296,
+      "step": 14102
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.787694634893695,
+      "learning_rate": 3.4799415145117743e-06,
+      "loss": 0.5387,
+      "step": 14103
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.17787377227078,
+      "learning_rate": 3.4786629560246166e-06,
+      "loss": 0.4788,
+      "step": 14104
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.324468011522743,
+      "learning_rate": 3.477384583000477e-06,
+      "loss": 0.4665,
+      "step": 14105
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.87162415806508,
+      "learning_rate": 3.4761063954756967e-06,
+      "loss": 0.4935,
+      "step": 14106
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.881506885263878,
+      "learning_rate": 3.4748283934866354e-06,
+      "loss": 0.4713,
+      "step": 14107
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.943740553383154,
+      "learning_rate": 3.4735505770696365e-06,
+      "loss": 0.5316,
+      "step": 14108
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.812310162099454,
+      "learning_rate": 3.472272946261045e-06,
+      "loss": 0.4961,
+      "step": 14109
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.55296125191944,
+      "learning_rate": 3.4709955010971928e-06,
+      "loss": 0.4191,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.224162283638634,
+      "learning_rate": 3.4697182416144047e-06,
+      "loss": 0.4456,
+      "step": 14111
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.882020101803512,
+      "learning_rate": 3.4684411678490104e-06,
+      "loss": 0.4243,
+      "step": 14112
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 24.784082809121216,
+      "learning_rate": 3.4671642798373316e-06,
+      "loss": 0.5193,
+      "step": 14113
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.32385857737832,
+      "learning_rate": 3.4658875776156763e-06,
+      "loss": 0.4954,
+      "step": 14114
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.0715255234308,
+      "learning_rate": 3.4646110612203586e-06,
+      "loss": 0.3921,
+      "step": 14115
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.061098432941833,
+      "learning_rate": 3.4633347306876763e-06,
+      "loss": 0.4915,
+      "step": 14116
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 15.773273701924971,
+      "learning_rate": 3.4620585860539303e-06,
+      "loss": 0.4743,
+      "step": 14117
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 34.74314170763986,
+      "learning_rate": 3.460782627355417e-06,
+      "loss": 0.4791,
+      "step": 14118
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.900922653014524,
+      "learning_rate": 3.4595068546284207e-06,
+      "loss": 0.4289,
+      "step": 14119
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.947930371650763,
+      "learning_rate": 3.458231267909219e-06,
+      "loss": 0.4502,
+      "step": 14120
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.370486386702545,
+      "learning_rate": 3.4569558672340943e-06,
+      "loss": 0.4923,
+      "step": 14121
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.91746958684292,
+      "learning_rate": 3.455680652639319e-06,
+      "loss": 0.4596,
+      "step": 14122
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 25.278477306884255,
+      "learning_rate": 3.4544056241611556e-06,
+      "loss": 0.5325,
+      "step": 14123
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 24.186439899382407,
+      "learning_rate": 3.4531307818358705e-06,
+      "loss": 0.468,
+      "step": 14124
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.243319270050254,
+      "learning_rate": 3.451856125699713e-06,
+      "loss": 0.4905,
+      "step": 14125
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.746459396353284,
+      "learning_rate": 3.4505816557889393e-06,
+      "loss": 0.4814,
+      "step": 14126
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.467693026117686,
+      "learning_rate": 3.4493073721397895e-06,
+      "loss": 0.5108,
+      "step": 14127
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.660902901563624,
+      "learning_rate": 3.4480332747885047e-06,
+      "loss": 0.4562,
+      "step": 14128
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 24.14380614360332,
+      "learning_rate": 3.446759363771326e-06,
+      "loss": 0.4643,
+      "step": 14129
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.309130709073447,
+      "learning_rate": 3.445485639124476e-06,
+      "loss": 0.4471,
+      "step": 14130
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.896155444501538,
+      "learning_rate": 3.444212100884177e-06,
+      "loss": 0.4594,
+      "step": 14131
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.981652467205958,
+      "learning_rate": 3.4429387490866506e-06,
+      "loss": 0.5047,
+      "step": 14132
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 19.197004302292797,
+      "learning_rate": 3.4416655837681113e-06,
+      "loss": 0.4436,
+      "step": 14133
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.28478034953453,
+      "learning_rate": 3.4403926049647684e-06,
+      "loss": 0.4566,
+      "step": 14134
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 32.408551509154584,
+      "learning_rate": 3.4391198127128232e-06,
+      "loss": 0.5468,
+      "step": 14135
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.636295686396245,
+      "learning_rate": 3.4378472070484704e-06,
+      "loss": 0.462,
+      "step": 14136
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.610298432492023,
+      "learning_rate": 3.4365747880079027e-06,
+      "loss": 0.4794,
+      "step": 14137
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 32.36538118814052,
+      "learning_rate": 3.435302555627312e-06,
+      "loss": 0.511,
+      "step": 14138
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 28.38483892126981,
+      "learning_rate": 3.4340305099428774e-06,
+      "loss": 0.4952,
+      "step": 14139
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.034427193231025,
+      "learning_rate": 3.4327586509907695e-06,
+      "loss": 0.4728,
+      "step": 14140
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.309065176248062,
+      "learning_rate": 3.431486978807165e-06,
+      "loss": 0.4917,
+      "step": 14141
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.13643800481047,
+      "learning_rate": 3.4302154934282272e-06,
+      "loss": 0.4239,
+      "step": 14142
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.289588030006954,
+      "learning_rate": 3.4289441948901214e-06,
+      "loss": 0.5347,
+      "step": 14143
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 27.630478786016152,
+      "learning_rate": 3.4276730832289997e-06,
+      "loss": 0.541,
+      "step": 14144
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.937499302404888,
+      "learning_rate": 3.4264021584810058e-06,
+      "loss": 0.4982,
+      "step": 14145
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.395984152637848,
+      "learning_rate": 3.4251314206822894e-06,
+      "loss": 0.4164,
+      "step": 14146
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 26.127649796864702,
+      "learning_rate": 3.4238608698689933e-06,
+      "loss": 0.4489,
+      "step": 14147
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.792726029359738,
+      "learning_rate": 3.4225905060772436e-06,
+      "loss": 0.5104,
+      "step": 14148
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.616790053119878,
+      "learning_rate": 3.421320329343175e-06,
+      "loss": 0.4441,
+      "step": 14149
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 32.529334857498206,
+      "learning_rate": 3.420050339702905e-06,
+      "loss": 0.4474,
+      "step": 14150
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 31.246698348868602,
+      "learning_rate": 3.4187805371925575e-06,
+      "loss": 0.5011,
+      "step": 14151
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 15.64517496239613,
+      "learning_rate": 3.4175109218482383e-06,
+      "loss": 0.4612,
+      "step": 14152
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.80270826449453,
+      "learning_rate": 3.4162414937060584e-06,
+      "loss": 0.4647,
+      "step": 14153
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.96895271150894,
+      "learning_rate": 3.414972252802121e-06,
+      "loss": 0.4539,
+      "step": 14154
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.328766394329747,
+      "learning_rate": 3.413703199172519e-06,
+      "loss": 0.4923,
+      "step": 14155
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.68504558291149,
+      "learning_rate": 3.4124343328533473e-06,
+      "loss": 0.5028,
+      "step": 14156
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 25.70867271556117,
+      "learning_rate": 3.411165653880686e-06,
+      "loss": 0.4463,
+      "step": 14157
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.55297796394052,
+      "learning_rate": 3.40989716229062e-06,
+      "loss": 0.5025,
+      "step": 14158
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.30916769874984,
+      "learning_rate": 3.408628858119226e-06,
+      "loss": 0.4585,
+      "step": 14159
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.766876085226116,
+      "learning_rate": 3.4073607414025723e-06,
+      "loss": 0.4396,
+      "step": 14160
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.99950692394386,
+      "learning_rate": 3.406092812176719e-06,
+      "loss": 0.4851,
+      "step": 14161
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.813836869884632,
+      "learning_rate": 3.404825070477729e-06,
+      "loss": 0.4791,
+      "step": 14162
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.32462966381847,
+      "learning_rate": 3.403557516341658e-06,
+      "loss": 0.504,
+      "step": 14163
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.55743554542306,
+      "learning_rate": 3.4022901498045536e-06,
+      "loss": 0.4485,
+      "step": 14164
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 27.66898178751445,
+      "learning_rate": 3.4010229709024545e-06,
+      "loss": 0.462,
+      "step": 14165
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 27.434941073638207,
+      "learning_rate": 3.399755979671401e-06,
+      "loss": 0.4548,
+      "step": 14166
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 19.783287359241562,
+      "learning_rate": 3.398489176147427e-06,
+      "loss": 0.4205,
+      "step": 14167
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.71350405721002,
+      "learning_rate": 3.3972225603665653e-06,
+      "loss": 0.4487,
+      "step": 14168
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 25.63923409250133,
+      "learning_rate": 3.395956132364825e-06,
+      "loss": 0.5299,
+      "step": 14169
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.060362616664143,
+      "learning_rate": 3.394689892178229e-06,
+      "loss": 0.4602,
+      "step": 14170
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.928723529214714,
+      "learning_rate": 3.3934238398427886e-06,
+      "loss": 0.4494,
+      "step": 14171
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 19.48356566778734,
+      "learning_rate": 3.3921579753945132e-06,
+      "loss": 0.4911,
+      "step": 14172
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.46514293542787,
+      "learning_rate": 3.3908922988693995e-06,
+      "loss": 0.4977,
+      "step": 14173
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.995072645330605,
+      "learning_rate": 3.389626810303439e-06,
+      "loss": 0.4918,
+      "step": 14174
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.043434922224902,
+      "learning_rate": 3.388361509732625e-06,
+      "loss": 0.3946,
+      "step": 14175
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 25.518325539529023,
+      "learning_rate": 3.387096397192945e-06,
+      "loss": 0.5225,
+      "step": 14176
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.24785047840458,
+      "learning_rate": 3.3858314727203724e-06,
+      "loss": 0.4514,
+      "step": 14177
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 19.376635347076668,
+      "learning_rate": 3.384566736350886e-06,
+      "loss": 0.4733,
+      "step": 14178
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.212229146527353,
+      "learning_rate": 3.3833021881204487e-06,
+      "loss": 0.4301,
+      "step": 14179
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.13177084131551,
+      "learning_rate": 3.3820378280650267e-06,
+      "loss": 0.5764,
+      "step": 14180
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.98748482006631,
+      "learning_rate": 3.3807736562205805e-06,
+      "loss": 0.4465,
+      "step": 14181
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.408600330064246,
+      "learning_rate": 3.3795096726230558e-06,
+      "loss": 0.4226,
+      "step": 14182
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.44167334212344,
+      "learning_rate": 3.3782458773084072e-06,
+      "loss": 0.4467,
+      "step": 14183
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.3822110439587,
+      "learning_rate": 3.3769822703125686e-06,
+      "loss": 0.4714,
+      "step": 14184
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.633649028751396,
+      "learning_rate": 3.3757188516714833e-06,
+      "loss": 0.5298,
+      "step": 14185
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.040769237496903,
+      "learning_rate": 3.3744556214210756e-06,
+      "loss": 0.5229,
+      "step": 14186
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.53573065318118,
+      "learning_rate": 3.3731925795972743e-06,
+      "loss": 0.4111,
+      "step": 14187
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.835639341237453,
+      "learning_rate": 3.371929726236002e-06,
+      "loss": 0.4892,
+      "step": 14188
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.292185956232,
+      "learning_rate": 3.370667061373172e-06,
+      "loss": 0.4838,
+      "step": 14189
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.642229667564486,
+      "learning_rate": 3.369404585044689e-06,
+      "loss": 0.4707,
+      "step": 14190
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.12206559686894,
+      "learning_rate": 3.3681422972864598e-06,
+      "loss": 0.4677,
+      "step": 14191
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.649980219197595,
+      "learning_rate": 3.366880198134386e-06,
+      "loss": 0.5158,
+      "step": 14192
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.220100786178264,
+      "learning_rate": 3.3656182876243636e-06,
+      "loss": 0.4718,
+      "step": 14193
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 29.121310873264775,
+      "learning_rate": 3.3643565657922704e-06,
+      "loss": 0.5004,
+      "step": 14194
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 22.303880807214757,
+      "learning_rate": 3.3630950326739954e-06,
+      "loss": 0.4644,
+      "step": 14195
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 23.889365711292687,
+      "learning_rate": 3.3618336883054146e-06,
+      "loss": 0.4265,
+      "step": 14196
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.29784727567313,
+      "learning_rate": 3.360572532722404e-06,
+      "loss": 0.4522,
+      "step": 14197
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.26974162583579,
+      "learning_rate": 3.359311565960827e-06,
+      "loss": 0.3747,
+      "step": 14198
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 28.503402181682258,
+      "learning_rate": 3.3580507880565406e-06,
+      "loss": 0.5239,
+      "step": 14199
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 31.36989860932342,
+      "learning_rate": 3.3567901990454043e-06,
+      "loss": 0.5603,
+      "step": 14200
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.294901828499274,
+      "learning_rate": 3.355529798963272e-06,
+      "loss": 0.498,
+      "step": 14201
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.89706129689245,
+      "learning_rate": 3.354269587845982e-06,
+      "loss": 0.4806,
+      "step": 14202
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.72062211116757,
+      "learning_rate": 3.3530095657293803e-06,
+      "loss": 0.4153,
+      "step": 14203
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.057245536089876,
+      "learning_rate": 3.3517497326492952e-06,
+      "loss": 0.5043,
+      "step": 14204
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.124092487109326,
+      "learning_rate": 3.350490088641558e-06,
+      "loss": 0.4259,
+      "step": 14205
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.50410646427527,
+      "learning_rate": 3.3492306337419967e-06,
+      "loss": 0.4986,
+      "step": 14206
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.564168096832738,
+      "learning_rate": 3.3479713679864223e-06,
+      "loss": 0.4927,
+      "step": 14207
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 24.4743156375904,
+      "learning_rate": 3.346712291410654e-06,
+      "loss": 0.5083,
+      "step": 14208
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.03059779145361,
+      "learning_rate": 3.345453404050493e-06,
+      "loss": 0.4428,
+      "step": 14209
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.726774609971446,
+      "learning_rate": 3.344194705941748e-06,
+      "loss": 0.4248,
+      "step": 14210
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.253499794708,
+      "learning_rate": 3.342936197120208e-06,
+      "loss": 0.4438,
+      "step": 14211
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.849995816354262,
+      "learning_rate": 3.341677877621672e-06,
+      "loss": 0.4895,
+      "step": 14212
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.839533799346864,
+      "learning_rate": 3.340419747481919e-06,
+      "loss": 0.4433,
+      "step": 14213
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 24.960924142884693,
+      "learning_rate": 3.3391618067367347e-06,
+      "loss": 0.4278,
+      "step": 14214
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.56766996589953,
+      "learning_rate": 3.33790405542189e-06,
+      "loss": 0.5227,
+      "step": 14215
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.760033100704494,
+      "learning_rate": 3.3366464935731566e-06,
+      "loss": 0.4874,
+      "step": 14216
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.74126868121376,
+      "learning_rate": 3.3353891212263023e-06,
+      "loss": 0.4802,
+      "step": 14217
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 26.367003042635503,
+      "learning_rate": 3.3341319384170824e-06,
+      "loss": 0.5098,
+      "step": 14218
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.144786309187655,
+      "learning_rate": 3.332874945181248e-06,
+      "loss": 0.4565,
+      "step": 14219
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.01811562315315,
+      "learning_rate": 3.33161814155455e-06,
+      "loss": 0.452,
+      "step": 14220
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 24.598941370793774,
+      "learning_rate": 3.330361527572731e-06,
+      "loss": 0.4083,
+      "step": 14221
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.59739929756781,
+      "learning_rate": 3.3291051032715317e-06,
+      "loss": 0.4356,
+      "step": 14222
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.428833061356688,
+      "learning_rate": 3.3278488686866807e-06,
+      "loss": 0.4809,
+      "step": 14223
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.427374598967635,
+      "learning_rate": 3.3265928238539015e-06,
+      "loss": 0.4905,
+      "step": 14224
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.67122250775418,
+      "learning_rate": 3.3253369688089186e-06,
+      "loss": 0.4164,
+      "step": 14225
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 23.149478178291545,
+      "learning_rate": 3.324081303587451e-06,
+      "loss": 0.4805,
+      "step": 14226
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.9630818272006,
+      "learning_rate": 3.3228258282252056e-06,
+      "loss": 0.4986,
+      "step": 14227
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.311415303319023,
+      "learning_rate": 3.321570542757885e-06,
+      "loss": 0.4531,
+      "step": 14228
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.470234888394227,
+      "learning_rate": 3.320315447221191e-06,
+      "loss": 0.479,
+      "step": 14229
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 26.985279540786987,
+      "learning_rate": 3.3190605416508182e-06,
+      "loss": 0.4606,
+      "step": 14230
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 23.78908937762767,
+      "learning_rate": 3.3178058260824586e-06,
+      "loss": 0.4094,
+      "step": 14231
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.78107048204239,
+      "learning_rate": 3.3165513005517925e-06,
+      "loss": 0.5466,
+      "step": 14232
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.38778097645092,
+      "learning_rate": 3.3152969650944943e-06,
+      "loss": 0.4879,
+      "step": 14233
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.175010316681853,
+      "learning_rate": 3.3140428197462406e-06,
+      "loss": 0.4802,
+      "step": 14234
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.772871014297134,
+      "learning_rate": 3.312788864542701e-06,
+      "loss": 0.4472,
+      "step": 14235
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.060031068760054,
+      "learning_rate": 3.3115350995195293e-06,
+      "loss": 0.4908,
+      "step": 14236
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 29.84109969878593,
+      "learning_rate": 3.310281524712392e-06,
+      "loss": 0.4939,
+      "step": 14237
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.69208608810965,
+      "learning_rate": 3.3090281401569302e-06,
+      "loss": 0.4536,
+      "step": 14238
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 27.794362378113373,
+      "learning_rate": 3.307774945888799e-06,
+      "loss": 0.507,
+      "step": 14239
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.428111228643306,
+      "learning_rate": 3.306521941943628e-06,
+      "loss": 0.4862,
+      "step": 14240
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.649227753240204,
+      "learning_rate": 3.3052691283570593e-06,
+      "loss": 0.5169,
+      "step": 14241
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.769626700522476,
+      "learning_rate": 3.304016505164722e-06,
+      "loss": 0.4817,
+      "step": 14242
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.108087637774734,
+      "learning_rate": 3.3027640724022357e-06,
+      "loss": 0.4683,
+      "step": 14243
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.233353210137388,
+      "learning_rate": 3.301511830105225e-06,
+      "loss": 0.4916,
+      "step": 14244
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 25.25821851126149,
+      "learning_rate": 3.300259778309296e-06,
+      "loss": 0.5044,
+      "step": 14245
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.10562831236159,
+      "learning_rate": 3.2990079170500587e-06,
+      "loss": 0.4736,
+      "step": 14246
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 27.241259908897085,
+      "learning_rate": 3.297756246363121e-06,
+      "loss": 0.4797,
+      "step": 14247
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.764458495389457,
+      "learning_rate": 3.2965047662840744e-06,
+      "loss": 0.4578,
+      "step": 14248
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.533729231931737,
+      "learning_rate": 3.2952534768485066e-06,
+      "loss": 0.4996,
+      "step": 14249
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 33.62612132149112,
+      "learning_rate": 3.294002378092008e-06,
+      "loss": 0.6179,
+      "step": 14250
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.766559414729766,
+      "learning_rate": 3.292751470050164e-06,
+      "loss": 0.3969,
+      "step": 14251
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.868531518816555,
+      "learning_rate": 3.291500752758543e-06,
+      "loss": 0.5056,
+      "step": 14252
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.674855912340163,
+      "learning_rate": 3.2902502262527124e-06,
+      "loss": 0.5254,
+      "step": 14253
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.660867243506882,
+      "learning_rate": 3.288999890568242e-06,
+      "loss": 0.5089,
+      "step": 14254
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.399274103197467,
+      "learning_rate": 3.287749745740687e-06,
+      "loss": 0.5188,
+      "step": 14255
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 30.479011577077774,
+      "learning_rate": 3.2864997918056075e-06,
+      "loss": 0.5113,
+      "step": 14256
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 25.150002384944067,
+      "learning_rate": 3.2852500287985455e-06,
+      "loss": 0.5246,
+      "step": 14257
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.388451270724662,
+      "learning_rate": 3.284000456755042e-06,
+      "loss": 0.4819,
+      "step": 14258
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.012601831596356,
+      "learning_rate": 3.282751075710637e-06,
+      "loss": 0.4771,
+      "step": 14259
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.74239715452549,
+      "learning_rate": 3.2815018857008653e-06,
+      "loss": 0.41,
+      "step": 14260
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.97333561576591,
+      "learning_rate": 3.2802528867612494e-06,
+      "loss": 0.5099,
+      "step": 14261
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.139812240992367,
+      "learning_rate": 3.279004078927307e-06,
+      "loss": 0.4836,
+      "step": 14262
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.145610888241855,
+      "learning_rate": 3.2777554622345563e-06,
+      "loss": 0.4683,
+      "step": 14263
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 25.393912403621083,
+      "learning_rate": 3.2765070367185126e-06,
+      "loss": 0.5101,
+      "step": 14264
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.908920665919428,
+      "learning_rate": 3.2752588024146714e-06,
+      "loss": 0.4527,
+      "step": 14265
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 25.12345903330322,
+      "learning_rate": 3.274010759358539e-06,
+      "loss": 0.478,
+      "step": 14266
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.874270747172954,
+      "learning_rate": 3.272762907585604e-06,
+      "loss": 0.456,
+      "step": 14267
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.42134939908545,
+      "learning_rate": 3.271515247131356e-06,
+      "loss": 0.4637,
+      "step": 14268
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.79649799246871,
+      "learning_rate": 3.270267778031281e-06,
+      "loss": 0.4652,
+      "step": 14269
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.23659705366554,
+      "learning_rate": 3.2690205003208508e-06,
+      "loss": 0.4655,
+      "step": 14270
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.757774548523162,
+      "learning_rate": 3.2677734140355434e-06,
+      "loss": 0.4932,
+      "step": 14271
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.056692628958345,
+      "learning_rate": 3.2665265192108186e-06,
+      "loss": 0.4383,
+      "step": 14272
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.407862603978565,
+      "learning_rate": 3.265279815882144e-06,
+      "loss": 0.5083,
+      "step": 14273
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.034531624729695,
+      "learning_rate": 3.264033304084968e-06,
+      "loss": 0.5098,
+      "step": 14274
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.019349531035484,
+      "learning_rate": 3.2627869838547454e-06,
+      "loss": 0.4443,
+      "step": 14275
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.186384294640565,
+      "learning_rate": 3.2615408552269223e-06,
+      "loss": 0.4873,
+      "step": 14276
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.194729800490645,
+      "learning_rate": 3.260294918236935e-06,
+      "loss": 0.4052,
+      "step": 14277
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 20.94586027824207,
+      "learning_rate": 3.2590491729202146e-06,
+      "loss": 0.427,
+      "step": 14278
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.449611347066973,
+      "learning_rate": 3.257803619312193e-06,
+      "loss": 0.4804,
+      "step": 14279
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.939002629026504,
+      "learning_rate": 3.256558257448292e-06,
+      "loss": 0.4719,
+      "step": 14280
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.642706943152906,
+      "learning_rate": 3.255313087363936e-06,
+      "loss": 0.4459,
+      "step": 14281
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.899193243596365,
+      "learning_rate": 3.2540681090945235e-06,
+      "loss": 0.5118,
+      "step": 14282
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.39349957168287,
+      "learning_rate": 3.252823322675468e-06,
+      "loss": 0.4634,
+      "step": 14283
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.177561836556563,
+      "learning_rate": 3.2515787281421697e-06,
+      "loss": 0.587,
+      "step": 14284
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.530327032668758,
+      "learning_rate": 3.250334325530029e-06,
+      "loss": 0.4582,
+      "step": 14285
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.160174986162176,
+      "learning_rate": 3.24909011487443e-06,
+      "loss": 0.4731,
+      "step": 14286
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.226236104355927,
+      "learning_rate": 3.247846096210757e-06,
+      "loss": 0.5326,
+      "step": 14287
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 20.96990783169642,
+      "learning_rate": 3.2466022695743905e-06,
+      "loss": 0.4769,
+      "step": 14288
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 14.472697040948256,
+      "learning_rate": 3.2453586350007084e-06,
+      "loss": 0.4267,
+      "step": 14289
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.966648976634477,
+      "learning_rate": 3.2441151925250724e-06,
+      "loss": 0.4255,
+      "step": 14290
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.185181563092526,
+      "learning_rate": 3.2428719421828515e-06,
+      "loss": 0.4687,
+      "step": 14291
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.446849352864483,
+      "learning_rate": 3.2416288840093955e-06,
+      "loss": 0.5577,
+      "step": 14292
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.577778325834785,
+      "learning_rate": 3.2403860180400615e-06,
+      "loss": 0.4748,
+      "step": 14293
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 27.08563303329335,
+      "learning_rate": 3.2391433443101984e-06,
+      "loss": 0.4926,
+      "step": 14294
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.28430442522078,
+      "learning_rate": 3.237900862855139e-06,
+      "loss": 0.3785,
+      "step": 14295
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.686519460034482,
+      "learning_rate": 3.236658573710227e-06,
+      "loss": 0.4874,
+      "step": 14296
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.982863507525025,
+      "learning_rate": 3.2354164769107866e-06,
+      "loss": 0.5085,
+      "step": 14297
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.35585327265751,
+      "learning_rate": 3.234174572492147e-06,
+      "loss": 0.4952,
+      "step": 14298
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.697228652851315,
+      "learning_rate": 3.2329328604896217e-06,
+      "loss": 0.472,
+      "step": 14299
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.83689424639408,
+      "learning_rate": 3.2316913409385263e-06,
+      "loss": 0.4262,
+      "step": 14300
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 20.1018932877213,
+      "learning_rate": 3.2304500138741736e-06,
+      "loss": 0.5658,
+      "step": 14301
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.114858241742233,
+      "learning_rate": 3.229208879331862e-06,
+      "loss": 0.4258,
+      "step": 14302
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 26.834629937623834,
+      "learning_rate": 3.2279679373468866e-06,
+      "loss": 0.4812,
+      "step": 14303
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.88807828150885,
+      "learning_rate": 3.2267271879545413e-06,
+      "loss": 0.5089,
+      "step": 14304
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.1155344353335,
+      "learning_rate": 3.2254866311901125e-06,
+      "loss": 0.4876,
+      "step": 14305
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.317829679443097,
+      "learning_rate": 3.224246267088884e-06,
+      "loss": 0.5531,
+      "step": 14306
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 33.593595767886946,
+      "learning_rate": 3.2230060956861285e-06,
+      "loss": 0.5258,
+      "step": 14307
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.675784659747722,
+      "learning_rate": 3.221766117017111e-06,
+      "loss": 0.4679,
+      "step": 14308
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 17.130109254529824,
+      "learning_rate": 3.220526331117101e-06,
+      "loss": 0.4636,
+      "step": 14309
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.189185045228566,
+      "learning_rate": 3.219286738021359e-06,
+      "loss": 0.4753,
+      "step": 14310
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.383957187050587,
+      "learning_rate": 3.2180473377651355e-06,
+      "loss": 0.5106,
+      "step": 14311
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.01716723519588,
+      "learning_rate": 3.216808130383675e-06,
+      "loss": 0.4941,
+      "step": 14312
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 26.421769824909966,
+      "learning_rate": 3.2155691159122236e-06,
+      "loss": 0.5627,
+      "step": 14313
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.824993583398598,
+      "learning_rate": 3.214330294386021e-06,
+      "loss": 0.5243,
+      "step": 14314
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.072516955797546,
+      "learning_rate": 3.2130916658402956e-06,
+      "loss": 0.5274,
+      "step": 14315
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 17.153106139095012,
+      "learning_rate": 3.211853230310269e-06,
+      "loss": 0.5096,
+      "step": 14316
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.136361590087546,
+      "learning_rate": 3.2106149878311655e-06,
+      "loss": 0.3973,
+      "step": 14317
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 24.237866439159284,
+      "learning_rate": 3.209376938438201e-06,
+      "loss": 0.4129,
+      "step": 14318
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.565251571926947,
+      "learning_rate": 3.208139082166587e-06,
+      "loss": 0.4767,
+      "step": 14319
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.440590835511284,
+      "learning_rate": 3.2069014190515237e-06,
+      "loss": 0.5808,
+      "step": 14320
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 16.453265147866635,
+      "learning_rate": 3.205663949128207e-06,
+      "loss": 0.4778,
+      "step": 14321
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 331.2600062478427,
+      "learning_rate": 3.204426672431834e-06,
+      "loss": 0.5467,
+      "step": 14322
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 27.51695419872413,
+      "learning_rate": 3.2031895889975926e-06,
+      "loss": 0.4934,
+      "step": 14323
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.623731021886798,
+      "learning_rate": 3.201952698860662e-06,
+      "loss": 0.5092,
+      "step": 14324
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.77207562124895,
+      "learning_rate": 3.200716002056222e-06,
+      "loss": 0.496,
+      "step": 14325
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.125458357699145,
+      "learning_rate": 3.199479498619439e-06,
+      "loss": 0.5078,
+      "step": 14326
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 24.328589947618013,
+      "learning_rate": 3.1982431885854837e-06,
+      "loss": 0.457,
+      "step": 14327
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 14.784039748250384,
+      "learning_rate": 3.19700707198951e-06,
+      "loss": 0.4783,
+      "step": 14328
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.502971612606256,
+      "learning_rate": 3.1957711488666756e-06,
+      "loss": 0.4858,
+      "step": 14329
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 25.170697397193248,
+      "learning_rate": 3.1945354192521316e-06,
+      "loss": 0.47,
+      "step": 14330
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.939671265055,
+      "learning_rate": 3.193299883181017e-06,
+      "loss": 0.4573,
+      "step": 14331
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.847025219725495,
+      "learning_rate": 3.192064540688475e-06,
+      "loss": 0.5015,
+      "step": 14332
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 14.79447557840037,
+      "learning_rate": 3.190829391809631e-06,
+      "loss": 0.4768,
+      "step": 14333
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.246256180200069,
+      "learning_rate": 3.189594436579617e-06,
+      "loss": 0.3996,
+      "step": 14334
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.952643798576595,
+      "learning_rate": 3.1883596750335554e-06,
+      "loss": 0.4828,
+      "step": 14335
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.894220802905878,
+      "learning_rate": 3.1871251072065613e-06,
+      "loss": 0.4751,
+      "step": 14336
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.197414232890242,
+      "learning_rate": 3.1858907331337397e-06,
+      "loss": 0.4944,
+      "step": 14337
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.07762755501611,
+      "learning_rate": 3.1846565528501993e-06,
+      "loss": 0.4641,
+      "step": 14338
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 24.98848517833969,
+      "learning_rate": 3.1834225663910436e-06,
+      "loss": 0.547,
+      "step": 14339
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.141554250910612,
+      "learning_rate": 3.1821887737913613e-06,
+      "loss": 0.3794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 16.376573819594306,
+      "learning_rate": 3.18095517508624e-06,
+      "loss": 0.3943,
+      "step": 14341
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 34.161243254792005,
+      "learning_rate": 3.1797217703107643e-06,
+      "loss": 0.5386,
+      "step": 14342
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 31.969370179206262,
+      "learning_rate": 3.1784885595000114e-06,
+      "loss": 0.5642,
+      "step": 14343
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 14.332285108628783,
+      "learning_rate": 3.177255542689056e-06,
+      "loss": 0.5116,
+      "step": 14344
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.646392088672556,
+      "learning_rate": 3.1760227199129623e-06,
+      "loss": 0.4162,
+      "step": 14345
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.409295328701088,
+      "learning_rate": 3.1747900912067877e-06,
+      "loss": 0.451,
+      "step": 14346
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 31.4810930893793,
+      "learning_rate": 3.17355765660559e-06,
+      "loss": 0.5056,
+      "step": 14347
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.86838834835201,
+      "learning_rate": 3.172325416144423e-06,
+      "loss": 0.4957,
+      "step": 14348
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.439822092020375,
+      "learning_rate": 3.171093369858322e-06,
+      "loss": 0.4592,
+      "step": 14349
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.048567435999786,
+      "learning_rate": 3.1698615177823357e-06,
+      "loss": 0.4631,
+      "step": 14350
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.6653666146451,
+      "learning_rate": 3.1686298599514887e-06,
+      "loss": 0.5,
+      "step": 14351
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.060286325369788,
+      "learning_rate": 3.167398396400816e-06,
+      "loss": 0.4454,
+      "step": 14352
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.511651940746397,
+      "learning_rate": 3.1661671271653317e-06,
+      "loss": 0.4407,
+      "step": 14353
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.465336946040818,
+      "learning_rate": 3.1649360522800578e-06,
+      "loss": 0.509,
+      "step": 14354
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.67990111678465,
+      "learning_rate": 3.163705171780006e-06,
+      "loss": 0.454,
+      "step": 14355
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 14.015911591618107,
+      "learning_rate": 3.1624744857001776e-06,
+      "loss": 0.4932,
+      "step": 14356
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 28.546697440375567,
+      "learning_rate": 3.161243994075578e-06,
+      "loss": 0.4614,
+      "step": 14357
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.98538722177228,
+      "learning_rate": 3.1600136969411954e-06,
+      "loss": 0.4479,
+      "step": 14358
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.639863646021713,
+      "learning_rate": 3.1587835943320256e-06,
+      "loss": 0.4667,
+      "step": 14359
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.654913511470536,
+      "learning_rate": 3.1575536862830445e-06,
+      "loss": 0.4452,
+      "step": 14360
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.19234611852943,
+      "learning_rate": 3.156323972829237e-06,
+      "loss": 0.4799,
+      "step": 14361
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.202184871932534,
+      "learning_rate": 3.1550944540055693e-06,
+      "loss": 0.4994,
+      "step": 14362
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.69935143448063,
+      "learning_rate": 3.1538651298470126e-06,
+      "loss": 0.485,
+      "step": 14363
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 17.52985505896903,
+      "learning_rate": 3.1526360003885283e-06,
+      "loss": 0.4512,
+      "step": 14364
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 24.83612723456546,
+      "learning_rate": 3.151407065665071e-06,
+      "loss": 0.5222,
+      "step": 14365
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.234942443733026,
+      "learning_rate": 3.150178325711587e-06,
+      "loss": 0.4532,
+      "step": 14366
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 53.68903504176631,
+      "learning_rate": 3.1489497805630243e-06,
+      "loss": 0.4524,
+      "step": 14367
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.256895821842843,
+      "learning_rate": 3.1477214302543225e-06,
+      "loss": 0.51,
+      "step": 14368
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 17.933806884037544,
+      "learning_rate": 3.1464932748204215e-06,
+      "loss": 0.4733,
+      "step": 14369
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.700028696694204,
+      "learning_rate": 3.1452653142962353e-06,
+      "loss": 0.4283,
+      "step": 14370
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 34.95162712369848,
+      "learning_rate": 3.144037548716694e-06,
+      "loss": 0.5144,
+      "step": 14371
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.451144056398366,
+      "learning_rate": 3.142809978116714e-06,
+      "loss": 0.5001,
+      "step": 14372
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.573215571137066,
+      "learning_rate": 3.1415826025312103e-06,
+      "loss": 0.412,
+      "step": 14373
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 29.050526628485077,
+      "learning_rate": 3.140355421995086e-06,
+      "loss": 0.5389,
+      "step": 14374
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.248755929833392,
+      "learning_rate": 3.139128436543236e-06,
+      "loss": 0.468,
+      "step": 14375
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.8957388994746,
+      "learning_rate": 3.1379016462105617e-06,
+      "loss": 0.3903,
+      "step": 14376
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.93119597084899,
+      "learning_rate": 3.136675051031952e-06,
+      "loss": 0.3946,
+      "step": 14377
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.80356864297048,
+      "learning_rate": 3.1354486510422865e-06,
+      "loss": 0.4358,
+      "step": 14378
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 26.34947878145194,
+      "learning_rate": 3.134222446276448e-06,
+      "loss": 0.4947,
+      "step": 14379
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.728335436484356,
+      "learning_rate": 3.1329964367693042e-06,
+      "loss": 0.4604,
+      "step": 14380
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.813707834769023,
+      "learning_rate": 3.1317706225557254e-06,
+      "loss": 0.4187,
+      "step": 14381
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.45159610225471,
+      "learning_rate": 3.130545003670574e-06,
+      "loss": 0.4904,
+      "step": 14382
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.566818894589602,
+      "learning_rate": 3.129319580148702e-06,
+      "loss": 0.5048,
+      "step": 14383
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 23.911631635401953,
+      "learning_rate": 3.128094352024965e-06,
+      "loss": 0.4119,
+      "step": 14384
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 26.220533922953038,
+      "learning_rate": 3.1268693193342014e-06,
+      "loss": 0.4568,
+      "step": 14385
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.011667214802133,
+      "learning_rate": 3.1256444821112573e-06,
+      "loss": 0.4442,
+      "step": 14386
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 28.49043746878992,
+      "learning_rate": 3.1244198403909587e-06,
+      "loss": 0.4872,
+      "step": 14387
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.50332712568118,
+      "learning_rate": 3.1231953942081385e-06,
+      "loss": 0.4169,
+      "step": 14388
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.501733297614937,
+      "learning_rate": 3.1219711435976207e-06,
+      "loss": 0.4344,
+      "step": 14389
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 32.946003909564176,
+      "learning_rate": 3.1207470885942213e-06,
+      "loss": 0.4389,
+      "step": 14390
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.428630474746697,
+      "learning_rate": 3.119523229232746e-06,
+      "loss": 0.4551,
+      "step": 14391
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 28.246480268878578,
+      "learning_rate": 3.1182995655480065e-06,
+      "loss": 0.4917,
+      "step": 14392
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.894016106189262,
+      "learning_rate": 3.1170760975748006e-06,
+      "loss": 0.4509,
+      "step": 14393
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.825113068778577,
+      "learning_rate": 3.115852825347927e-06,
+      "loss": 0.4263,
+      "step": 14394
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.229917379061057,
+      "learning_rate": 3.114629748902173e-06,
+      "loss": 0.449,
+      "step": 14395
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.263691951888067,
+      "learning_rate": 3.1134068682723173e-06,
+      "loss": 0.4603,
+      "step": 14396
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 22.43525973906959,
+      "learning_rate": 3.112184183493142e-06,
+      "loss": 0.473,
+      "step": 14397
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.820768709080474,
+      "learning_rate": 3.1109616945994225e-06,
+      "loss": 0.4678,
+      "step": 14398
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 33.057930132058054,
+      "learning_rate": 3.109739401625922e-06,
+      "loss": 0.5474,
+      "step": 14399
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 23.04651367615477,
+      "learning_rate": 3.108517304607399e-06,
+      "loss": 0.5315,
+      "step": 14400
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 19.247604440900464,
+      "learning_rate": 3.107295403578613e-06,
+      "loss": 0.4507,
+      "step": 14401
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.72954990365314,
+      "learning_rate": 3.1060736985743166e-06,
+      "loss": 0.6107,
+      "step": 14402
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.574510548656622,
+      "learning_rate": 3.1048521896292483e-06,
+      "loss": 0.4751,
+      "step": 14403
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 22.150621748501255,
+      "learning_rate": 3.103630876778153e-06,
+      "loss": 0.4813,
+      "step": 14404
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.581079570714497,
+      "learning_rate": 3.102409760055758e-06,
+      "loss": 0.4464,
+      "step": 14405
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 26.680110779472226,
+      "learning_rate": 3.101188839496795e-06,
+      "loss": 0.5548,
+      "step": 14406
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.091697645504162,
+      "learning_rate": 3.099968115135988e-06,
+      "loss": 0.5001,
+      "step": 14407
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 27.30240287882424,
+      "learning_rate": 3.098747587008053e-06,
+      "loss": 0.5052,
+      "step": 14408
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.883326689529127,
+      "learning_rate": 3.0975272551476953e-06,
+      "loss": 0.4574,
+      "step": 14409
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.06161724272564,
+      "learning_rate": 3.0963071195896243e-06,
+      "loss": 0.5273,
+      "step": 14410
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 29.230492738792673,
+      "learning_rate": 3.0950871803685435e-06,
+      "loss": 0.4941,
+      "step": 14411
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 38.369405423690246,
+      "learning_rate": 3.0938674375191415e-06,
+      "loss": 0.5153,
+      "step": 14412
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.70718803467509,
+      "learning_rate": 3.0926478910761127e-06,
+      "loss": 0.4729,
+      "step": 14413
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 27.151596696762624,
+      "learning_rate": 3.091428541074134e-06,
+      "loss": 0.4964,
+      "step": 14414
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 30.89761117922585,
+      "learning_rate": 3.090209387547889e-06,
+      "loss": 0.5068,
+      "step": 14415
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 22.268542439245305,
+      "learning_rate": 3.088990430532044e-06,
+      "loss": 0.5266,
+      "step": 14416
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.64029710015681,
+      "learning_rate": 3.0877716700612683e-06,
+      "loss": 0.4689,
+      "step": 14417
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.933843844827553,
+      "learning_rate": 3.086553106170226e-06,
+      "loss": 0.4081,
+      "step": 14418
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.797660422602728,
+      "learning_rate": 3.085334738893565e-06,
+      "loss": 0.4636,
+      "step": 14419
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 19.296978701571533,
+      "learning_rate": 3.0841165682659436e-06,
+      "loss": 0.431,
+      "step": 14420
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.908002233275102,
+      "learning_rate": 3.082898594321998e-06,
+      "loss": 0.4549,
+      "step": 14421
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.254789524200397,
+      "learning_rate": 3.0816808170963697e-06,
+      "loss": 0.4357,
+      "step": 14422
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.83150593497665,
+      "learning_rate": 3.080463236623694e-06,
+      "loss": 0.4414,
+      "step": 14423
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.63845862288901,
+      "learning_rate": 3.079245852938597e-06,
+      "loss": 0.471,
+      "step": 14424
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.340703832860676,
+      "learning_rate": 3.0780286660756964e-06,
+      "loss": 0.5241,
+      "step": 14425
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.455858806933296,
+      "learning_rate": 3.0768116760696108e-06,
+      "loss": 0.4209,
+      "step": 14426
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.35202524830817,
+      "learning_rate": 3.0755948829549552e-06,
+      "loss": 0.505,
+      "step": 14427
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 22.689368047929896,
+      "learning_rate": 3.0743782867663297e-06,
+      "loss": 0.5471,
+      "step": 14428
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.994484678204449,
+      "learning_rate": 3.0731618875383318e-06,
+      "loss": 0.5364,
+      "step": 14429
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 26.42212318905037,
+      "learning_rate": 3.071945685305557e-06,
+      "loss": 0.5211,
+      "step": 14430
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.191327406256336,
+      "learning_rate": 3.0707296801025954e-06,
+      "loss": 0.5204,
+      "step": 14431
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.991302221039223,
+      "learning_rate": 3.0695138719640307e-06,
+      "loss": 0.4586,
+      "step": 14432
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.888557334610837,
+      "learning_rate": 3.068298260924436e-06,
+      "loss": 0.501,
+      "step": 14433
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.2318372367857,
+      "learning_rate": 3.0670828470183823e-06,
+      "loss": 0.4796,
+      "step": 14434
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.624803972164138,
+      "learning_rate": 3.065867630280436e-06,
+      "loss": 0.4517,
+      "step": 14435
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.187882251486435,
+      "learning_rate": 3.0646526107451615e-06,
+      "loss": 0.4536,
+      "step": 14436
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.443662702139886,
+      "learning_rate": 3.0634377884471055e-06,
+      "loss": 0.4794,
+      "step": 14437
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 26.660167494551413,
+      "learning_rate": 3.0622231634208254e-06,
+      "loss": 0.4736,
+      "step": 14438
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 24.6690069782025,
+      "learning_rate": 3.0610087357008567e-06,
+      "loss": 0.4688,
+      "step": 14439
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.528152093975113,
+      "learning_rate": 3.0597945053217435e-06,
+      "loss": 0.4876,
+      "step": 14440
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.846604173722543,
+      "learning_rate": 3.0585804723180115e-06,
+      "loss": 0.473,
+      "step": 14441
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.46423862074733,
+      "learning_rate": 3.0573666367241893e-06,
+      "loss": 0.47,
+      "step": 14442
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.817058423026626,
+      "learning_rate": 3.056152998574803e-06,
+      "loss": 0.4354,
+      "step": 14443
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 28.24880661384868,
+      "learning_rate": 3.0549395579043593e-06,
+      "loss": 0.4707,
+      "step": 14444
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.011777732696835,
+      "learning_rate": 3.053726314747376e-06,
+      "loss": 0.4621,
+      "step": 14445
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 37.511315925657414,
+      "learning_rate": 3.0525132691383486e-06,
+      "loss": 0.5196,
+      "step": 14446
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.35770411127352,
+      "learning_rate": 3.0513004211117805e-06,
+      "loss": 0.5356,
+      "step": 14447
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 26.269570943242257,
+      "learning_rate": 3.050087770702167e-06,
+      "loss": 0.5597,
+      "step": 14448
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.864862249370894,
+      "learning_rate": 3.048875317943991e-06,
+      "loss": 0.4689,
+      "step": 14449
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.075530169860757,
+      "learning_rate": 3.047663062871731e-06,
+      "loss": 0.4594,
+      "step": 14450
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.95961564916996,
+      "learning_rate": 3.046451005519867e-06,
+      "loss": 0.4948,
+      "step": 14451
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.648139900230777,
+      "learning_rate": 3.0452391459228726e-06,
+      "loss": 0.4036,
+      "step": 14452
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.6742962693194,
+      "learning_rate": 3.0440274841152072e-06,
+      "loss": 0.3943,
+      "step": 14453
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.8202832262932,
+      "learning_rate": 3.0428160201313295e-06,
+      "loss": 0.4934,
+      "step": 14454
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 25.224949656697138,
+      "learning_rate": 3.0416047540056924e-06,
+      "loss": 0.5326,
+      "step": 14455
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.351467316124854,
+      "learning_rate": 3.040393685772747e-06,
+      "loss": 0.451,
+      "step": 14456
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.01128670964009,
+      "learning_rate": 3.0391828154669367e-06,
+      "loss": 0.4905,
+      "step": 14457
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.223739676859527,
+      "learning_rate": 3.037972143122695e-06,
+      "loss": 0.5233,
+      "step": 14458
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.127069508035094,
+      "learning_rate": 3.0367616687744504e-06,
+      "loss": 0.5177,
+      "step": 14459
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.784847572725155,
+      "learning_rate": 3.035551392456629e-06,
+      "loss": 0.5001,
+      "step": 14460
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.118037957009825,
+      "learning_rate": 3.0343413142036573e-06,
+      "loss": 0.4609,
+      "step": 14461
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 24.86200295673555,
+      "learning_rate": 3.0331314340499416e-06,
+      "loss": 0.3973,
+      "step": 14462
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 27.53457395620035,
+      "learning_rate": 3.0319217520298904e-06,
+      "loss": 0.5044,
+      "step": 14463
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.766893642032095,
+      "learning_rate": 3.0307122681779088e-06,
+      "loss": 0.4725,
+      "step": 14464
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.08258026387261,
+      "learning_rate": 3.0295029825283952e-06,
+      "loss": 0.4086,
+      "step": 14465
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.277191850390157,
+      "learning_rate": 3.028293895115737e-06,
+      "loss": 0.5337,
+      "step": 14466
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.337090513963576,
+      "learning_rate": 3.027085005974325e-06,
+      "loss": 0.4672,
+      "step": 14467
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.845246605093315,
+      "learning_rate": 3.025876315138533e-06,
+      "loss": 0.47,
+      "step": 14468
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 30.07893215334974,
+      "learning_rate": 3.0246678226427396e-06,
+      "loss": 0.514,
+      "step": 14469
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.79710428473836,
+      "learning_rate": 3.0234595285213154e-06,
+      "loss": 0.4601,
+      "step": 14470
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 25.590624931709048,
+      "learning_rate": 3.022251432808618e-06,
+      "loss": 0.4572,
+      "step": 14471
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.60547368954316,
+      "learning_rate": 3.021043535539011e-06,
+      "loss": 0.492,
+      "step": 14472
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.489112097051457,
+      "learning_rate": 3.0198358367468407e-06,
+      "loss": 0.5081,
+      "step": 14473
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.924892681158354,
+      "learning_rate": 3.018628336466458e-06,
+      "loss": 0.4881,
+      "step": 14474
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 27.067810916223017,
+      "learning_rate": 3.0174210347322e-06,
+      "loss": 0.4992,
+      "step": 14475
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.110252014848722,
+      "learning_rate": 3.016213931578401e-06,
+      "loss": 0.3848,
+      "step": 14476
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.053493181843194,
+      "learning_rate": 3.0150070270393973e-06,
+      "loss": 0.4943,
+      "step": 14477
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.591367301746114,
+      "learning_rate": 3.013800321149506e-06,
+      "loss": 0.4513,
+      "step": 14478
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 27.25668355364714,
+      "learning_rate": 3.012593813943044e-06,
+      "loss": 0.4823,
+      "step": 14479
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 24.236099586389955,
+      "learning_rate": 3.0113875054543273e-06,
+      "loss": 0.4419,
+      "step": 14480
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.793343441803447,
+      "learning_rate": 3.0101813957176617e-06,
+      "loss": 0.4485,
+      "step": 14481
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.148656435629523,
+      "learning_rate": 3.008975484767351e-06,
+      "loss": 0.4498,
+      "step": 14482
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 26.316310479237725,
+      "learning_rate": 3.0077697726376873e-06,
+      "loss": 0.4948,
+      "step": 14483
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.967292667981102,
+      "learning_rate": 3.0065642593629597e-06,
+      "loss": 0.4571,
+      "step": 14484
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.033081494290624,
+      "learning_rate": 3.005358944977452e-06,
+      "loss": 0.422,
+      "step": 14485
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.42182222943798,
+      "learning_rate": 3.004153829515447e-06,
+      "loss": 0.3917,
+      "step": 14486
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.248015574051667,
+      "learning_rate": 3.0029489130112155e-06,
+      "loss": 0.4413,
+      "step": 14487
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 24.56216708532793,
+      "learning_rate": 3.00174419549902e-06,
+      "loss": 0.5005,
+      "step": 14488
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.607647846384427,
+      "learning_rate": 3.0005396770131266e-06,
+      "loss": 0.4137,
+      "step": 14489
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.657152513001353,
+      "learning_rate": 2.9993353575877926e-06,
+      "loss": 0.5015,
+      "step": 14490
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.530034866680325,
+      "learning_rate": 2.9981312372572624e-06,
+      "loss": 0.4064,
+      "step": 14491
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 27.66560088660425,
+      "learning_rate": 2.9969273160557877e-06,
+      "loss": 0.4708,
+      "step": 14492
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.908500331863422,
+      "learning_rate": 2.9957235940176e-06,
+      "loss": 0.4383,
+      "step": 14493
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.767054634609384,
+      "learning_rate": 2.9945200711769362e-06,
+      "loss": 0.4238,
+      "step": 14494
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.695565296883993,
+      "learning_rate": 2.993316747568026e-06,
+      "loss": 0.4209,
+      "step": 14495
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.785178843667577,
+      "learning_rate": 2.992113623225086e-06,
+      "loss": 0.5235,
+      "step": 14496
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.993745771290246,
+      "learning_rate": 2.990910698182338e-06,
+      "loss": 0.4338,
+      "step": 14497
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.035853734141558,
+      "learning_rate": 2.9897079724739866e-06,
+      "loss": 0.486,
+      "step": 14498
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.84878956762315,
+      "learning_rate": 2.9885054461342423e-06,
+      "loss": 0.4751,
+      "step": 14499
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.292491253505947,
+      "learning_rate": 2.987303119197299e-06,
+      "loss": 0.571,
+      "step": 14500
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 26.748964980414147,
+      "learning_rate": 2.986100991697355e-06,
+      "loss": 0.503,
+      "step": 14501
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.626106715288486,
+      "learning_rate": 2.9848990636685937e-06,
+      "loss": 0.4401,
+      "step": 14502
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.825547914341616,
+      "learning_rate": 2.9836973351452014e-06,
+      "loss": 0.4166,
+      "step": 14503
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.36021487039967,
+      "learning_rate": 2.9824958061613497e-06,
+      "loss": 0.5354,
+      "step": 14504
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 14.280289947932223,
+      "learning_rate": 2.9812944767512132e-06,
+      "loss": 0.4738,
+      "step": 14505
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.28615169192425,
+      "learning_rate": 2.980093346948958e-06,
+      "loss": 0.4233,
+      "step": 14506
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 14.652010355293557,
+      "learning_rate": 2.978892416788738e-06,
+      "loss": 0.4452,
+      "step": 14507
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.86114506152174,
+      "learning_rate": 2.9776916863047146e-06,
+      "loss": 0.513,
+      "step": 14508
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.08327401077484,
+      "learning_rate": 2.976491155531027e-06,
+      "loss": 0.3844,
+      "step": 14509
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.971513364263334,
+      "learning_rate": 2.975290824501824e-06,
+      "loss": 0.4355,
+      "step": 14510
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.137544789759183,
+      "learning_rate": 2.9740906932512427e-06,
+      "loss": 0.4916,
+      "step": 14511
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 21.506983227994724,
+      "learning_rate": 2.9728907618134127e-06,
+      "loss": 0.4789,
+      "step": 14512
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 26.674275112045073,
+      "learning_rate": 2.9716910302224556e-06,
+      "loss": 0.4751,
+      "step": 14513
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.903675721692522,
+      "learning_rate": 2.970491498512493e-06,
+      "loss": 0.4841,
+      "step": 14514
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 27.091565882834058,
+      "learning_rate": 2.9692921667176443e-06,
+      "loss": 0.495,
+      "step": 14515
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.164713719495037,
+      "learning_rate": 2.968093034872014e-06,
+      "loss": 0.4176,
+      "step": 14516
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.855802217633038,
+      "learning_rate": 2.9668941030097e-06,
+      "loss": 0.5089,
+      "step": 14517
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.513754404920117,
+      "learning_rate": 2.965695371164804e-06,
+      "loss": 0.4824,
+      "step": 14518
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.153989606483556,
+      "learning_rate": 2.964496839371417e-06,
+      "loss": 0.475,
+      "step": 14519
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.54719415535711,
+      "learning_rate": 2.963298507663627e-06,
+      "loss": 0.4146,
+      "step": 14520
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 15.587431353251151,
+      "learning_rate": 2.9621003760755105e-06,
+      "loss": 0.4653,
+      "step": 14521
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.765670173238664,
+      "learning_rate": 2.9609024446411394e-06,
+      "loss": 0.4635,
+      "step": 14522
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 33.189888607785285,
+      "learning_rate": 2.9597047133945855e-06,
+      "loss": 0.4649,
+      "step": 14523
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.530166295839656,
+      "learning_rate": 2.9585071823699143e-06,
+      "loss": 0.5239,
+      "step": 14524
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.76675658021288,
+      "learning_rate": 2.9573098516011756e-06,
+      "loss": 0.437,
+      "step": 14525
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 21.74801631731782,
+      "learning_rate": 2.9561127211224283e-06,
+      "loss": 0.4819,
+      "step": 14526
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.54625519662564,
+      "learning_rate": 2.9549157909677116e-06,
+      "loss": 0.4643,
+      "step": 14527
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.086998798894857,
+      "learning_rate": 2.953719061171072e-06,
+      "loss": 0.5299,
+      "step": 14528
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 21.47009228448441,
+      "learning_rate": 2.9525225317665375e-06,
+      "loss": 0.4952,
+      "step": 14529
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.299977458215515,
+      "learning_rate": 2.9513262027881396e-06,
+      "loss": 0.4498,
+      "step": 14530
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.708110349709333,
+      "learning_rate": 2.9501300742699036e-06,
+      "loss": 0.4598,
+      "step": 14531
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.90220833643359,
+      "learning_rate": 2.948934146245841e-06,
+      "loss": 0.4472,
+      "step": 14532
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 29.114908057001653,
+      "learning_rate": 2.94773841874997e-06,
+      "loss": 0.5214,
+      "step": 14533
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.30533818311973,
+      "learning_rate": 2.94654289181629e-06,
+      "loss": 0.5607,
+      "step": 14534
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.211873931090388,
+      "learning_rate": 2.9453475654788032e-06,
+      "loss": 0.4365,
+      "step": 14535
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.13345323023959,
+      "learning_rate": 2.9441524397715094e-06,
+      "loss": 0.4351,
+      "step": 14536
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.27511627659148,
+      "learning_rate": 2.942957514728392e-06,
+      "loss": 0.5023,
+      "step": 14537
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.422833012431038,
+      "learning_rate": 2.9417627903834313e-06,
+      "loss": 0.46,
+      "step": 14538
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 22.46845957842912,
+      "learning_rate": 2.940568266770607e-06,
+      "loss": 0.5018,
+      "step": 14539
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 30.015772882203493,
+      "learning_rate": 2.9393739439238955e-06,
+      "loss": 0.463,
+      "step": 14540
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.04022529065963,
+      "learning_rate": 2.9381798218772594e-06,
+      "loss": 0.4812,
+      "step": 14541
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.619839533049767,
+      "learning_rate": 2.936985900664654e-06,
+      "loss": 0.4966,
+      "step": 14542
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.654944393936418,
+      "learning_rate": 2.9357921803200386e-06,
+      "loss": 0.4499,
+      "step": 14543
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.553267308308644,
+      "learning_rate": 2.93459866087736e-06,
+      "loss": 0.5098,
+      "step": 14544
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.593171497541185,
+      "learning_rate": 2.9334053423705666e-06,
+      "loss": 0.4264,
+      "step": 14545
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 22.457023025407675,
+      "learning_rate": 2.9322122248335905e-06,
+      "loss": 0.4872,
+      "step": 14546
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.542750911649627,
+      "learning_rate": 2.9310193083003614e-06,
+      "loss": 0.4392,
+      "step": 14547
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 21.489082211932182,
+      "learning_rate": 2.9298265928048086e-06,
+      "loss": 0.4723,
+      "step": 14548
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 22.081831631044146,
+      "learning_rate": 2.9286340783808543e-06,
+      "loss": 0.5261,
+      "step": 14549
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 28.07680276868442,
+      "learning_rate": 2.9274417650624097e-06,
+      "loss": 0.4656,
+      "step": 14550
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.449567615829828,
+      "learning_rate": 2.926249652883382e-06,
+      "loss": 0.5031,
+      "step": 14551
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.441363449929163,
+      "learning_rate": 2.925057741877675e-06,
+      "loss": 0.431,
+      "step": 14552
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 27.545289395895157,
+      "learning_rate": 2.92386603207919e-06,
+      "loss": 0.4956,
+      "step": 14553
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.43183017108756,
+      "learning_rate": 2.9226745235218134e-06,
+      "loss": 0.446,
+      "step": 14554
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.270468638981548,
+      "learning_rate": 2.921483216239436e-06,
+      "loss": 0.4876,
+      "step": 14555
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.45729308615934,
+      "learning_rate": 2.920292110265932e-06,
+      "loss": 0.4861,
+      "step": 14556
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 27.16373185797766,
+      "learning_rate": 2.9191012056351787e-06,
+      "loss": 0.4482,
+      "step": 14557
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.588607991492953,
+      "learning_rate": 2.917910502381048e-06,
+      "loss": 0.4362,
+      "step": 14558
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 33.51284107837178,
+      "learning_rate": 2.9167200005373976e-06,
+      "loss": 0.5505,
+      "step": 14559
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.74736918384356,
+      "learning_rate": 2.915529700138089e-06,
+      "loss": 0.442,
+      "step": 14560
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.059512406185483,
+      "learning_rate": 2.9143396012169677e-06,
+      "loss": 0.4956,
+      "step": 14561
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.80184154521149,
+      "learning_rate": 2.913149703807887e-06,
+      "loss": 0.4861,
+      "step": 14562
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.87458520970228,
+      "learning_rate": 2.9119600079446807e-06,
+      "loss": 0.5143,
+      "step": 14563
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.69032047095443,
+      "learning_rate": 2.910770513661184e-06,
+      "loss": 0.4832,
+      "step": 14564
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.498750248229264,
+      "learning_rate": 2.909581220991231e-06,
+      "loss": 0.5434,
+      "step": 14565
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.0219330277558,
+      "learning_rate": 2.9083921299686403e-06,
+      "loss": 0.442,
+      "step": 14566
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.758304284939708,
+      "learning_rate": 2.9072032406272243e-06,
+      "loss": 0.4921,
+      "step": 14567
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.624430396421985,
+      "learning_rate": 2.9060145530008013e-06,
+      "loss": 0.4476,
+      "step": 14568
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.387851107157815,
+      "learning_rate": 2.9048260671231733e-06,
+      "loss": 0.4768,
+      "step": 14569
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.096355680163803,
+      "learning_rate": 2.9036377830281435e-06,
+      "loss": 0.4922,
+      "step": 14570
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.7804446582132,
+      "learning_rate": 2.902449700749506e-06,
+      "loss": 0.4775,
+      "step": 14571
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.879931748187758,
+      "learning_rate": 2.901261820321042e-06,
+      "loss": 0.5038,
+      "step": 14572
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.898715280707513,
+      "learning_rate": 2.90007414177654e-06,
+      "loss": 0.4704,
+      "step": 14573
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 35.882618256385136,
+      "learning_rate": 2.8988866651497793e-06,
+      "loss": 0.558,
+      "step": 14574
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 25.85289553343663,
+      "learning_rate": 2.8976993904745277e-06,
+      "loss": 0.4738,
+      "step": 14575
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.857581692419895,
+      "learning_rate": 2.8965123177845476e-06,
+      "loss": 0.479,
+      "step": 14576
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.481075737700465,
+      "learning_rate": 2.8953254471136017e-06,
+      "loss": 0.475,
+      "step": 14577
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 22.950848007157003,
+      "learning_rate": 2.8941387784954476e-06,
+      "loss": 0.5141,
+      "step": 14578
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.821943100227863,
+      "learning_rate": 2.8929523119638257e-06,
+      "loss": 0.4449,
+      "step": 14579
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.016087738776136,
+      "learning_rate": 2.8917660475524867e-06,
+      "loss": 0.5287,
+      "step": 14580
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.85752968946482,
+      "learning_rate": 2.89057998529516e-06,
+      "loss": 0.4505,
+      "step": 14581
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.64946738988201,
+      "learning_rate": 2.8893941252255784e-06,
+      "loss": 0.4415,
+      "step": 14582
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.013762917310192,
+      "learning_rate": 2.8882084673774722e-06,
+      "loss": 0.4651,
+      "step": 14583
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 33.37288981278104,
+      "learning_rate": 2.8870230117845543e-06,
+      "loss": 0.4923,
+      "step": 14584
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 26.872206953092036,
+      "learning_rate": 2.885837758480543e-06,
+      "loss": 0.4836,
+      "step": 14585
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.644142551234598,
+      "learning_rate": 2.8846527074991427e-06,
+      "loss": 0.4874,
+      "step": 14586
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.325005004417585,
+      "learning_rate": 2.8834678588740585e-06,
+      "loss": 0.4944,
+      "step": 14587
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.662272654185333,
+      "learning_rate": 2.8822832126389834e-06,
+      "loss": 0.4829,
+      "step": 14588
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.74556383159142,
+      "learning_rate": 2.8810987688276092e-06,
+      "loss": 0.4504,
+      "step": 14589
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.91071675869982,
+      "learning_rate": 2.8799145274736253e-06,
+      "loss": 0.4825,
+      "step": 14590
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.21911702937514,
+      "learning_rate": 2.878730488610706e-06,
+      "loss": 0.4193,
+      "step": 14591
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.372383373215637,
+      "learning_rate": 2.8775466522725238e-06,
+      "loss": 0.4663,
+      "step": 14592
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.75069859854102,
+      "learning_rate": 2.876363018492747e-06,
+      "loss": 0.4074,
+      "step": 14593
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 27.8066160873516,
+      "learning_rate": 2.8751795873050405e-06,
+      "loss": 0.5877,
+      "step": 14594
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.661622441894131,
+      "learning_rate": 2.873996358743061e-06,
+      "loss": 0.3917,
+      "step": 14595
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.017143054186315,
+      "learning_rate": 2.872813332840456e-06,
+      "loss": 0.4441,
+      "step": 14596
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.565455806318877,
+      "learning_rate": 2.8716305096308674e-06,
+      "loss": 0.4235,
+      "step": 14597
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.927970018637607,
+      "learning_rate": 2.870447889147938e-06,
+      "loss": 0.5055,
+      "step": 14598
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.692718635637814,
+      "learning_rate": 2.869265471425302e-06,
+      "loss": 0.4674,
+      "step": 14599
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.29028742650249,
+      "learning_rate": 2.8680832564965853e-06,
+      "loss": 0.4597,
+      "step": 14600
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.760241108927712,
+      "learning_rate": 2.866901244395406e-06,
+      "loss": 0.4893,
+      "step": 14601
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.41395928274524,
+      "learning_rate": 2.8657194351553817e-06,
+      "loss": 0.5354,
+      "step": 14602
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.196611814185918,
+      "learning_rate": 2.8645378288101267e-06,
+      "loss": 0.4533,
+      "step": 14603
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.016700890852174,
+      "learning_rate": 2.863356425393242e-06,
+      "loss": 0.4446,
+      "step": 14604
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.901529817688725,
+      "learning_rate": 2.862175224938323e-06,
+      "loss": 0.4981,
+      "step": 14605
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.93669328376025,
+      "learning_rate": 2.860994227478964e-06,
+      "loss": 0.4859,
+      "step": 14606
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.275653836699915,
+      "learning_rate": 2.8598134330487536e-06,
+      "loss": 0.4262,
+      "step": 14607
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 29.23866465692385,
+      "learning_rate": 2.8586328416812746e-06,
+      "loss": 0.479,
+      "step": 14608
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.303862175792137,
+      "learning_rate": 2.857452453410101e-06,
+      "loss": 0.5062,
+      "step": 14609
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.070503608147156,
+      "learning_rate": 2.856272268268798e-06,
+      "loss": 0.4274,
+      "step": 14610
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.416829497740473,
+      "learning_rate": 2.8550922862909327e-06,
+      "loss": 0.4651,
+      "step": 14611
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 53.24683257429877,
+      "learning_rate": 2.853912507510066e-06,
+      "loss": 0.4817,
+      "step": 14612
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.6163698181681,
+      "learning_rate": 2.852732931959744e-06,
+      "loss": 0.5131,
+      "step": 14613
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.85994082182788,
+      "learning_rate": 2.85155355967352e-06,
+      "loss": 0.4233,
+      "step": 14614
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 25.748390077405716,
+      "learning_rate": 2.850374390684928e-06,
+      "loss": 0.4581,
+      "step": 14615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.71247043596305,
+      "learning_rate": 2.849195425027509e-06,
+      "loss": 0.577,
+      "step": 14616
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.36669564409993,
+      "learning_rate": 2.8480166627347847e-06,
+      "loss": 0.4809,
+      "step": 14617
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.013068291095262,
+      "learning_rate": 2.8468381038402836e-06,
+      "loss": 0.4681,
+      "step": 14618
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.458456391071184,
+      "learning_rate": 2.845659748377526e-06,
+      "loss": 0.5167,
+      "step": 14619
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.23870538157785,
+      "learning_rate": 2.8444815963800164e-06,
+      "loss": 0.5133,
+      "step": 14620
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.536674247422006,
+      "learning_rate": 2.843303647881267e-06,
+      "loss": 0.4429,
+      "step": 14621
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.70212843796634,
+      "learning_rate": 2.842125902914773e-06,
+      "loss": 0.5036,
+      "step": 14622
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 37.736611541272914,
+      "learning_rate": 2.840948361514031e-06,
+      "loss": 0.5101,
+      "step": 14623
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.606048121652947,
+      "learning_rate": 2.839771023712532e-06,
+      "loss": 0.4418,
+      "step": 14624
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.13825398493852,
+      "learning_rate": 2.838593889543757e-06,
+      "loss": 0.4611,
+      "step": 14625
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.694140225746896,
+      "learning_rate": 2.8374169590411793e-06,
+      "loss": 0.4813,
+      "step": 14626
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.938039316712928,
+      "learning_rate": 2.836240232238274e-06,
+      "loss": 0.49,
+      "step": 14627
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.672149234018818,
+      "learning_rate": 2.8350637091685085e-06,
+      "loss": 0.4539,
+      "step": 14628
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.839086796349058,
+      "learning_rate": 2.8338873898653394e-06,
+      "loss": 0.4864,
+      "step": 14629
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.876517253987576,
+      "learning_rate": 2.8327112743622176e-06,
+      "loss": 0.4772,
+      "step": 14630
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 29.03922370566224,
+      "learning_rate": 2.8315353626925943e-06,
+      "loss": 0.4416,
+      "step": 14631
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 25.506289814340988,
+      "learning_rate": 2.8303596548899127e-06,
+      "loss": 0.5167,
+      "step": 14632
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 23.45906272877004,
+      "learning_rate": 2.829184150987612e-06,
+      "loss": 0.4793,
+      "step": 14633
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.315959965353123,
+      "learning_rate": 2.8280088510191174e-06,
+      "loss": 0.5183,
+      "step": 14634
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 25.80685360739133,
+      "learning_rate": 2.8268337550178537e-06,
+      "loss": 0.4263,
+      "step": 14635
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 23.11465568640033,
+      "learning_rate": 2.8256588630172412e-06,
+      "loss": 0.4544,
+      "step": 14636
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 23.103817308213625,
+      "learning_rate": 2.8244841750506967e-06,
+      "loss": 0.4515,
+      "step": 14637
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.62563470366555,
+      "learning_rate": 2.823309691151622e-06,
+      "loss": 0.4442,
+      "step": 14638
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 26.325458061004305,
+      "learning_rate": 2.8221354113534248e-06,
+      "loss": 0.4463,
+      "step": 14639
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 25.48125682982296,
+      "learning_rate": 2.8209613356894948e-06,
+      "loss": 0.4377,
+      "step": 14640
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 20.810976844044117,
+      "learning_rate": 2.8197874641932276e-06,
+      "loss": 0.4796,
+      "step": 14641
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.447786302624664,
+      "learning_rate": 2.818613796898001e-06,
+      "loss": 0.4873,
+      "step": 14642
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 20.518754891044363,
+      "learning_rate": 2.8174403338371968e-06,
+      "loss": 0.4801,
+      "step": 14643
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 23.140094227148214,
+      "learning_rate": 2.8162670750441924e-06,
+      "loss": 0.487,
+      "step": 14644
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.97101558398097,
+      "learning_rate": 2.815094020552346e-06,
+      "loss": 0.5129,
+      "step": 14645
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.28130223562377,
+      "learning_rate": 2.8139211703950255e-06,
+      "loss": 0.6144,
+      "step": 14646
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.737472836752787,
+      "learning_rate": 2.81274852460558e-06,
+      "loss": 0.4402,
+      "step": 14647
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.490587807186053,
+      "learning_rate": 2.8115760832173657e-06,
+      "loss": 0.4549,
+      "step": 14648
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.184511076016577,
+      "learning_rate": 2.8104038462637195e-06,
+      "loss": 0.4527,
+      "step": 14649
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.66197647132942,
+      "learning_rate": 2.809231813777984e-06,
+      "loss": 0.4474,
+      "step": 14650
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 20.516430213623416,
+      "learning_rate": 2.8080599857934876e-06,
+      "loss": 0.5242,
+      "step": 14651
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.360133258856674,
+      "learning_rate": 2.806888362343557e-06,
+      "loss": 0.5067,
+      "step": 14652
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.378408133727232,
+      "learning_rate": 2.805716943461517e-06,
+      "loss": 0.4669,
+      "step": 14653
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.125692484808297,
+      "learning_rate": 2.8045457291806787e-06,
+      "loss": 0.3607,
+      "step": 14654
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.92809975105807,
+      "learning_rate": 2.8033747195343474e-06,
+      "loss": 0.5073,
+      "step": 14655
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.888604003515766,
+      "learning_rate": 2.802203914555829e-06,
+      "loss": 0.4903,
+      "step": 14656
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.927222034641275,
+      "learning_rate": 2.8010333142784217e-06,
+      "loss": 0.5385,
+      "step": 14657
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.71245044373287,
+      "learning_rate": 2.7998629187354178e-06,
+      "loss": 0.4453,
+      "step": 14658
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.554409654087648,
+      "learning_rate": 2.798692727960101e-06,
+      "loss": 0.4856,
+      "step": 14659
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.986017007424994,
+      "learning_rate": 2.7975227419857466e-06,
+      "loss": 0.4349,
+      "step": 14660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.87303181061066,
+      "learning_rate": 2.796352960845632e-06,
+      "loss": 0.4533,
+      "step": 14661
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 29.373021142333087,
+      "learning_rate": 2.795183384573029e-06,
+      "loss": 0.4723,
+      "step": 14662
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.222709169900291,
+      "learning_rate": 2.7940140132011963e-06,
+      "loss": 0.4216,
+      "step": 14663
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 23.032443900070554,
+      "learning_rate": 2.792844846763385e-06,
+      "loss": 0.4792,
+      "step": 14664
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.9586632307366,
+      "learning_rate": 2.7916758852928514e-06,
+      "loss": 0.4608,
+      "step": 14665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.399868598568535,
+      "learning_rate": 2.7905071288228415e-06,
+      "loss": 0.4484,
+      "step": 14666
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 28.753317653444675,
+      "learning_rate": 2.789338577386589e-06,
+      "loss": 0.5067,
+      "step": 14667
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.383428350103912,
+      "learning_rate": 2.788170231017332e-06,
+      "loss": 0.4084,
+      "step": 14668
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 31.82335384846937,
+      "learning_rate": 2.787002089748291e-06,
+      "loss": 0.4722,
+      "step": 14669
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.38579140469859,
+      "learning_rate": 2.785834153612692e-06,
+      "loss": 0.4914,
+      "step": 14670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.433195744947856,
+      "learning_rate": 2.7846664226437524e-06,
+      "loss": 0.4063,
+      "step": 14671
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.670153166576075,
+      "learning_rate": 2.783498896874676e-06,
+      "loss": 0.4822,
+      "step": 14672
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 26.748641396273573,
+      "learning_rate": 2.7823315763386715e-06,
+      "loss": 0.4548,
+      "step": 14673
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.134115833809354,
+      "learning_rate": 2.7811644610689327e-06,
+      "loss": 0.4557,
+      "step": 14674
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.29928887699729,
+      "learning_rate": 2.7799975510986566e-06,
+      "loss": 0.4867,
+      "step": 14675
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.249070043826354,
+      "learning_rate": 2.7788308464610225e-06,
+      "loss": 0.4484,
+      "step": 14676
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.684559613134798,
+      "learning_rate": 2.777664347189216e-06,
+      "loss": 0.4743,
+      "step": 14677
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.679171289820214,
+      "learning_rate": 2.776498053316413e-06,
+      "loss": 0.4336,
+      "step": 14678
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.897845134899944,
+      "learning_rate": 2.7753319648757802e-06,
+      "loss": 0.4376,
+      "step": 14679
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 23.94907233449032,
+      "learning_rate": 2.7741660819004766e-06,
+      "loss": 0.493,
+      "step": 14680
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.118537295277786,
+      "learning_rate": 2.7730004044236636e-06,
+      "loss": 0.3929,
+      "step": 14681
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.018668022160856,
+      "learning_rate": 2.7718349324784897e-06,
+      "loss": 0.4568,
+      "step": 14682
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.37691721558114,
+      "learning_rate": 2.770669666098107e-06,
+      "loss": 0.545,
+      "step": 14683
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 28.157284896494332,
+      "learning_rate": 2.769504605315648e-06,
+      "loss": 0.4321,
+      "step": 14684
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.420540112757475,
+      "learning_rate": 2.7683397501642463e-06,
+      "loss": 0.4862,
+      "step": 14685
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.18713358101687,
+      "learning_rate": 2.767175100677032e-06,
+      "loss": 0.4293,
+      "step": 14686
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.996247025641516,
+      "learning_rate": 2.7660106568871293e-06,
+      "loss": 0.4966,
+      "step": 14687
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.484005550675867,
+      "learning_rate": 2.7648464188276514e-06,
+      "loss": 0.4995,
+      "step": 14688
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.23665222348996,
+      "learning_rate": 2.7636823865317065e-06,
+      "loss": 0.4938,
+      "step": 14689
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.48146588649827,
+      "learning_rate": 2.7625185600324013e-06,
+      "loss": 0.4615,
+      "step": 14690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.65652048721396,
+      "learning_rate": 2.761354939362838e-06,
+      "loss": 0.4101,
+      "step": 14691
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.754384520145393,
+      "learning_rate": 2.7601915245561017e-06,
+      "loss": 0.4562,
+      "step": 14692
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.28091433127257,
+      "learning_rate": 2.7590283156452867e-06,
+      "loss": 0.4447,
+      "step": 14693
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.169771750070954,
+      "learning_rate": 2.757865312663468e-06,
+      "loss": 0.5174,
+      "step": 14694
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.597543016264888,
+      "learning_rate": 2.7567025156437223e-06,
+      "loss": 0.4409,
+      "step": 14695
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 16.911334087744464,
+      "learning_rate": 2.755539924619124e-06,
+      "loss": 0.4581,
+      "step": 14696
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.502788704677496,
+      "learning_rate": 2.754377539622731e-06,
+      "loss": 0.4437,
+      "step": 14697
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 24.431116527730488,
+      "learning_rate": 2.7532153606876e-06,
+      "loss": 0.4441,
+      "step": 14698
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.527162249277215,
+      "learning_rate": 2.7520533878467846e-06,
+      "loss": 0.4626,
+      "step": 14699
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.97678055389082,
+      "learning_rate": 2.7508916211333346e-06,
+      "loss": 0.4527,
+      "step": 14700
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 24.68243090762094,
+      "learning_rate": 2.7497300605802823e-06,
+      "loss": 0.4519,
+      "step": 14701
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.909088308455285,
+      "learning_rate": 2.74856870622067e-06,
+      "loss": 0.5036,
+      "step": 14702
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.995748140782542,
+      "learning_rate": 2.7474075580875182e-06,
+      "loss": 0.5154,
+      "step": 14703
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.834694657208953,
+      "learning_rate": 2.7462466162138557e-06,
+      "loss": 0.4256,
+      "step": 14704
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 27.08003131299765,
+      "learning_rate": 2.745085880632693e-06,
+      "loss": 0.5153,
+      "step": 14705
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.49231549338284,
+      "learning_rate": 2.7439253513770434e-06,
+      "loss": 0.484,
+      "step": 14706
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.393991719809545,
+      "learning_rate": 2.7427650284799157e-06,
+      "loss": 0.4556,
+      "step": 14707
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.16881591696427,
+      "learning_rate": 2.7416049119743025e-06,
+      "loss": 0.4203,
+      "step": 14708
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.339026026426843,
+      "learning_rate": 2.7404450018932027e-06,
+      "loss": 0.5047,
+      "step": 14709
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 26.102454892123095,
+      "learning_rate": 2.739285298269596e-06,
+      "loss": 0.5579,
+      "step": 14710
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.587943772420655,
+      "learning_rate": 2.7381258011364697e-06,
+      "loss": 0.4283,
+      "step": 14711
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.980272688650185,
+      "learning_rate": 2.7369665105267995e-06,
+      "loss": 0.4591,
+      "step": 14712
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 27.826995916663137,
+      "learning_rate": 2.7358074264735534e-06,
+      "loss": 0.4661,
+      "step": 14713
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 28.78130568686368,
+      "learning_rate": 2.7346485490096908e-06,
+      "loss": 0.4651,
+      "step": 14714
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.15933784413965,
+      "learning_rate": 2.733489878168175e-06,
+      "loss": 0.4658,
+      "step": 14715
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.930780814506697,
+      "learning_rate": 2.732331413981958e-06,
+      "loss": 0.4281,
+      "step": 14716
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.150278522548328,
+      "learning_rate": 2.731173156483985e-06,
+      "loss": 0.3993,
+      "step": 14717
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 25.11578189245044,
+      "learning_rate": 2.730015105707191e-06,
+      "loss": 0.5256,
+      "step": 14718
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.060144420985747,
+      "learning_rate": 2.7288572616845154e-06,
+      "loss": 0.4226,
+      "step": 14719
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.270587037624495,
+      "learning_rate": 2.7276996244488864e-06,
+      "loss": 0.4851,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.69489096465625,
+      "learning_rate": 2.7265421940332283e-06,
+      "loss": 0.4996,
+      "step": 14721
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.65475279781367,
+      "learning_rate": 2.725384970470455e-06,
+      "loss": 0.4392,
+      "step": 14722
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.89639241926165,
+      "learning_rate": 2.724227953793475e-06,
+      "loss": 0.4943,
+      "step": 14723
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.830173977256038,
+      "learning_rate": 2.7230711440351952e-06,
+      "loss": 0.4506,
+      "step": 14724
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.002047370994212,
+      "learning_rate": 2.7219145412285187e-06,
+      "loss": 0.5654,
+      "step": 14725
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 30.83761012448493,
+      "learning_rate": 2.7207581454063314e-06,
+      "loss": 0.4628,
+      "step": 14726
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.420598344895975,
+      "learning_rate": 2.7196019566015286e-06,
+      "loss": 0.4241,
+      "step": 14727
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.726341596820642,
+      "learning_rate": 2.7184459748469836e-06,
+      "loss": 0.442,
+      "step": 14728
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 28.157998006417163,
+      "learning_rate": 2.7172902001755787e-06,
+      "loss": 0.4655,
+      "step": 14729
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.637164559726063,
+      "learning_rate": 2.7161346326201763e-06,
+      "loss": 0.4612,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.990402636917032,
+      "learning_rate": 2.7149792722136446e-06,
+      "loss": 0.424,
+      "step": 14731
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.8277824629726,
+      "learning_rate": 2.713824118988844e-06,
+      "loss": 0.4213,
+      "step": 14732
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.966103460986012,
+      "learning_rate": 2.7126691729786202e-06,
+      "loss": 0.3914,
+      "step": 14733
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.871155453871058,
+      "learning_rate": 2.711514434215825e-06,
+      "loss": 0.4956,
+      "step": 14734
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.273026851950597,
+      "learning_rate": 2.710359902733293e-06,
+      "loss": 0.4185,
+      "step": 14735
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.39572743871331,
+      "learning_rate": 2.709205578563859e-06,
+      "loss": 0.4317,
+      "step": 14736
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.75973473988637,
+      "learning_rate": 2.7080514617403584e-06,
+      "loss": 0.485,
+      "step": 14737
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.986755810526127,
+      "learning_rate": 2.706897552295609e-06,
+      "loss": 0.4726,
+      "step": 14738
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.84436533763036,
+      "learning_rate": 2.705743850262422e-06,
+      "loss": 0.4531,
+      "step": 14739
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 21.006405353056806,
+      "learning_rate": 2.704590355673614e-06,
+      "loss": 0.4994,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.168732031082126,
+      "learning_rate": 2.7034370685619894e-06,
+      "loss": 0.4639,
+      "step": 14741
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 25.70326264063381,
+      "learning_rate": 2.7022839889603514e-06,
+      "loss": 0.5261,
+      "step": 14742
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 26.200516599920157,
+      "learning_rate": 2.7011311169014807e-06,
+      "loss": 0.5424,
+      "step": 14743
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.137120982314368,
+      "learning_rate": 2.6999784524181727e-06,
+      "loss": 0.4332,
+      "step": 14744
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.503260156884863,
+      "learning_rate": 2.698825995543207e-06,
+      "loss": 0.4426,
+      "step": 14745
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.890467453975056,
+      "learning_rate": 2.6976737463093617e-06,
+      "loss": 0.5504,
+      "step": 14746
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.410389858341277,
+      "learning_rate": 2.696521704749403e-06,
+      "loss": 0.4595,
+      "step": 14747
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 28.532383847138505,
+      "learning_rate": 2.6953698708960916e-06,
+      "loss": 0.4939,
+      "step": 14748
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 24.730443736033425,
+      "learning_rate": 2.6942182447821884e-06,
+      "loss": 0.5174,
+      "step": 14749
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.030762659920697,
+      "learning_rate": 2.693066826440447e-06,
+      "loss": 0.3762,
+      "step": 14750
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.309325667082327,
+      "learning_rate": 2.691915615903611e-06,
+      "loss": 0.4684,
+      "step": 14751
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.336966322184672,
+      "learning_rate": 2.6907646132044174e-06,
+      "loss": 0.4399,
+      "step": 14752
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.96643673056194,
+      "learning_rate": 2.6896138183756015e-06,
+      "loss": 0.466,
+      "step": 14753
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 28.01286912651057,
+      "learning_rate": 2.6884632314498947e-06,
+      "loss": 0.5197,
+      "step": 14754
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 28.32817280360805,
+      "learning_rate": 2.6873128524600144e-06,
+      "loss": 0.4914,
+      "step": 14755
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 24.038694696894883,
+      "learning_rate": 2.686162681438682e-06,
+      "loss": 0.4478,
+      "step": 14756
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.72828883096704,
+      "learning_rate": 2.6850127184186015e-06,
+      "loss": 0.4786,
+      "step": 14757
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.499255774401032,
+      "learning_rate": 2.6838629634324796e-06,
+      "loss": 0.4649,
+      "step": 14758
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.49377016157838,
+      "learning_rate": 2.6827134165130175e-06,
+      "loss": 0.4751,
+      "step": 14759
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 33.80068617339733,
+      "learning_rate": 2.681564077692903e-06,
+      "loss": 0.465,
+      "step": 14760
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.10800867350413,
+      "learning_rate": 2.6804149470048293e-06,
+      "loss": 0.4257,
+      "step": 14761
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.92176748177484,
+      "learning_rate": 2.6792660244814683e-06,
+      "loss": 0.5182,
+      "step": 14762
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.61863688678932,
+      "learning_rate": 2.6781173101555014e-06,
+      "loss": 0.5147,
+      "step": 14763
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 22.51446621022123,
+      "learning_rate": 2.6769688040595933e-06,
+      "loss": 0.418,
+      "step": 14764
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 35.37467467424671,
+      "learning_rate": 2.6758205062264076e-06,
+      "loss": 0.6174,
+      "step": 14765
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.19856184104096,
+      "learning_rate": 2.6746724166886063e-06,
+      "loss": 0.4986,
+      "step": 14766
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.73393693066997,
+      "learning_rate": 2.6735245354788354e-06,
+      "loss": 0.4074,
+      "step": 14767
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 29.977719126804683,
+      "learning_rate": 2.6723768626297387e-06,
+      "loss": 0.5491,
+      "step": 14768
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.036516328754196,
+      "learning_rate": 2.671229398173957e-06,
+      "loss": 0.4181,
+      "step": 14769
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.795193939475638,
+      "learning_rate": 2.670082142144124e-06,
+      "loss": 0.4695,
+      "step": 14770
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.281777913103817,
+      "learning_rate": 2.66893509457287e-06,
+      "loss": 0.4283,
+      "step": 14771
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.715814940672992,
+      "learning_rate": 2.6677882554928126e-06,
+      "loss": 0.4336,
+      "step": 14772
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.29970615856538,
+      "learning_rate": 2.666641624936566e-06,
+      "loss": 0.4705,
+      "step": 14773
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 25.7762620869203,
+      "learning_rate": 2.665495202936741e-06,
+      "loss": 0.3934,
+      "step": 14774
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.45398289686087,
+      "learning_rate": 2.6643489895259454e-06,
+      "loss": 0.4339,
+      "step": 14775
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 31.787859629787935,
+      "learning_rate": 2.6632029847367734e-06,
+      "loss": 0.4828,
+      "step": 14776
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.304642750443232,
+      "learning_rate": 2.662057188601813e-06,
+      "loss": 0.4754,
+      "step": 14777
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.273346968341556,
+      "learning_rate": 2.660911601153654e-06,
+      "loss": 0.4228,
+      "step": 14778
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.552807188520944,
+      "learning_rate": 2.659766222424879e-06,
+      "loss": 0.4679,
+      "step": 14779
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.11000004850306,
+      "learning_rate": 2.658621052448055e-06,
+      "loss": 0.4796,
+      "step": 14780
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.583140371604497,
+      "learning_rate": 2.6574760912557575e-06,
+      "loss": 0.4617,
+      "step": 14781
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 22.13267904780665,
+      "learning_rate": 2.656331338880541e-06,
+      "loss": 0.4948,
+      "step": 14782
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.483142060774096,
+      "learning_rate": 2.6551867953549657e-06,
+      "loss": 0.4608,
+      "step": 14783
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.424854003297458,
+      "learning_rate": 2.6540424607115845e-06,
+      "loss": 0.4336,
+      "step": 14784
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.966072058776174,
+      "learning_rate": 2.6528983349829353e-06,
+      "loss": 0.4938,
+      "step": 14785
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 25.3838812854603,
+      "learning_rate": 2.651754418201563e-06,
+      "loss": 0.5059,
+      "step": 14786
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.55757092470731,
+      "learning_rate": 2.6506107103999924e-06,
+      "loss": 0.5013,
+      "step": 14787
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.67083547969026,
+      "learning_rate": 2.6494672116107577e-06,
+      "loss": 0.4672,
+      "step": 14788
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.768631326653288,
+      "learning_rate": 2.6483239218663724e-06,
+      "loss": 0.4431,
+      "step": 14789
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.775635248215437,
+      "learning_rate": 2.647180841199354e-06,
+      "loss": 0.4165,
+      "step": 14790
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.37594359719969,
+      "learning_rate": 2.6460379696422146e-06,
+      "loss": 0.5215,
+      "step": 14791
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.149163491907018,
+      "learning_rate": 2.6448953072274532e-06,
+      "loss": 0.4261,
+      "step": 14792
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 27.66035867979796,
+      "learning_rate": 2.6437528539875633e-06,
+      "loss": 0.4685,
+      "step": 14793
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.53604255743931,
+      "learning_rate": 2.6426106099550396e-06,
+      "loss": 0.4199,
+      "step": 14794
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.808610632910877,
+      "learning_rate": 2.6414685751623682e-06,
+      "loss": 0.5242,
+      "step": 14795
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 26.38101476407635,
+      "learning_rate": 2.6403267496420238e-06,
+      "loss": 0.4226,
+      "step": 14796
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.953883919708073,
+      "learning_rate": 2.639185133426484e-06,
+      "loss": 0.4586,
+      "step": 14797
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 24.96119778576508,
+      "learning_rate": 2.6380437265482107e-06,
+      "loss": 0.4967,
+      "step": 14798
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.063426896917715,
+      "learning_rate": 2.6369025290396664e-06,
+      "loss": 0.471,
+      "step": 14799
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 24.121129323338522,
+      "learning_rate": 2.6357615409333092e-06,
+      "loss": 0.5235,
+      "step": 14800
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 28.479236777572076,
+      "learning_rate": 2.634620762261587e-06,
+      "loss": 0.5212,
+      "step": 14801
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 26.14978696533939,
+      "learning_rate": 2.6334801930569387e-06,
+      "loss": 0.4359,
+      "step": 14802
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.385133531371865,
+      "learning_rate": 2.6323398333518047e-06,
+      "loss": 0.3778,
+      "step": 14803
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.217725480424736,
+      "learning_rate": 2.631199683178615e-06,
+      "loss": 0.4403,
+      "step": 14804
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.753549080849846,
+      "learning_rate": 2.6300597425698015e-06,
+      "loss": 0.4559,
+      "step": 14805
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.199802247465566,
+      "learning_rate": 2.6289200115577727e-06,
+      "loss": 0.469,
+      "step": 14806
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.925972175709823,
+      "learning_rate": 2.627780490174946e-06,
+      "loss": 0.4779,
+      "step": 14807
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.888678641712463,
+      "learning_rate": 2.6266411784537294e-06,
+      "loss": 0.4734,
+      "step": 14808
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 29.56573948742745,
+      "learning_rate": 2.625502076426528e-06,
+      "loss": 0.4755,
+      "step": 14809
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 24.176629647454032,
+      "learning_rate": 2.6243631841257332e-06,
+      "loss": 0.486,
+      "step": 14810
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.585281193322523,
+      "learning_rate": 2.6232245015837323e-06,
+      "loss": 0.4237,
+      "step": 14811
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.804607391848606,
+      "learning_rate": 2.6220860288329108e-06,
+      "loss": 0.4012,
+      "step": 14812
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.361003595040447,
+      "learning_rate": 2.6209477659056503e-06,
+      "loss": 0.4458,
+      "step": 14813
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 29.78269713009057,
+      "learning_rate": 2.6198097128343147e-06,
+      "loss": 0.4823,
+      "step": 14814
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 27.436299516911774,
+      "learning_rate": 2.618671869651278e-06,
+      "loss": 0.4859,
+      "step": 14815
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 26.03135524398759,
+      "learning_rate": 2.617534236388891e-06,
+      "loss": 0.426,
+      "step": 14816
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 24.54870931885095,
+      "learning_rate": 2.616396813079516e-06,
+      "loss": 0.5183,
+      "step": 14817
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.097912615598116,
+      "learning_rate": 2.615259599755492e-06,
+      "loss": 0.4959,
+      "step": 14818
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.04726304797163,
+      "learning_rate": 2.614122596449166e-06,
+      "loss": 0.4563,
+      "step": 14819
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 22.833553091845644,
+      "learning_rate": 2.6129858031928746e-06,
+      "loss": 0.4487,
+      "step": 14820
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.616677906779472,
+      "learning_rate": 2.611849220018944e-06,
+      "loss": 0.4976,
+      "step": 14821
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.643266251972197,
+      "learning_rate": 2.610712846959702e-06,
+      "loss": 0.4414,
+      "step": 14822
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.073181399367094,
+      "learning_rate": 2.609576684047461e-06,
+      "loss": 0.4214,
+      "step": 14823
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 27.644867918566085,
+      "learning_rate": 2.6084407313145354e-06,
+      "loss": 0.5168,
+      "step": 14824
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.276763378764976,
+      "learning_rate": 2.6073049887932357e-06,
+      "loss": 0.4718,
+      "step": 14825
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.10417396931099,
+      "learning_rate": 2.606169456515857e-06,
+      "loss": 0.4461,
+      "step": 14826
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.48619740757821,
+      "learning_rate": 2.6050341345146902e-06,
+      "loss": 0.4862,
+      "step": 14827
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 25.439154093150872,
+      "learning_rate": 2.6038990228220275e-06,
+      "loss": 0.478,
+      "step": 14828
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.906699752389017,
+      "learning_rate": 2.6027641214701495e-06,
+      "loss": 0.4522,
+      "step": 14829
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.301037713919737,
+      "learning_rate": 2.601629430491338e-06,
+      "loss": 0.4749,
+      "step": 14830
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.233602452374694,
+      "learning_rate": 2.6004949499178522e-06,
+      "loss": 0.4284,
+      "step": 14831
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.47842621460263,
+      "learning_rate": 2.5993606797819616e-06,
+      "loss": 0.4687,
+      "step": 14832
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.425854689750977,
+      "learning_rate": 2.5982266201159236e-06,
+      "loss": 0.4504,
+      "step": 14833
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.461040963172433,
+      "learning_rate": 2.5970927709519942e-06,
+      "loss": 0.5189,
+      "step": 14834
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.111971924386424,
+      "learning_rate": 2.5959591323224152e-06,
+      "loss": 0.4433,
+      "step": 14835
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.55038206833877,
+      "learning_rate": 2.5948257042594248e-06,
+      "loss": 0.4911,
+      "step": 14836
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.941442181129986,
+      "learning_rate": 2.5936924867952596e-06,
+      "loss": 0.491,
+      "step": 14837
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 28.30753601001517,
+      "learning_rate": 2.5925594799621503e-06,
+      "loss": 0.4588,
+      "step": 14838
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.024396194283737,
+      "learning_rate": 2.5914266837923153e-06,
+      "loss": 0.4093,
+      "step": 14839
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.974316825679026,
+      "learning_rate": 2.5902940983179694e-06,
+      "loss": 0.5072,
+      "step": 14840
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.16822372357109,
+      "learning_rate": 2.589161723571324e-06,
+      "loss": 0.4847,
+      "step": 14841
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.215014280924883,
+      "learning_rate": 2.5880295595845885e-06,
+      "loss": 0.5208,
+      "step": 14842
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.876328135326183,
+      "learning_rate": 2.586897606389952e-06,
+      "loss": 0.4503,
+      "step": 14843
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 34.072124884356555,
+      "learning_rate": 2.585765864019615e-06,
+      "loss": 0.408,
+      "step": 14844
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.428078083498058,
+      "learning_rate": 2.5846343325057556e-06,
+      "loss": 0.4284,
+      "step": 14845
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.794546595276005,
+      "learning_rate": 2.583503011880558e-06,
+      "loss": 0.4672,
+      "step": 14846
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 30.61878933990655,
+      "learning_rate": 2.5823719021761993e-06,
+      "loss": 0.5086,
+      "step": 14847
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.845127475043842,
+      "learning_rate": 2.5812410034248426e-06,
+      "loss": 0.4499,
+      "step": 14848
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.689449787403717,
+      "learning_rate": 2.5801103156586536e-06,
+      "loss": 0.4536,
+      "step": 14849
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.768953118210405,
+      "learning_rate": 2.5789798389097855e-06,
+      "loss": 0.4558,
+      "step": 14850
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.603273176706368,
+      "learning_rate": 2.5778495732103914e-06,
+      "loss": 0.4389,
+      "step": 14851
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.59337215787625,
+      "learning_rate": 2.5767195185926098e-06,
+      "loss": 0.4187,
+      "step": 14852
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.49086633224654,
+      "learning_rate": 2.5755896750885834e-06,
+      "loss": 0.4934,
+      "step": 14853
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.443562248250814,
+      "learning_rate": 2.5744600427304467e-06,
+      "loss": 0.4737,
+      "step": 14854
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 27.265632918228267,
+      "learning_rate": 2.573330621550323e-06,
+      "loss": 0.4643,
+      "step": 14855
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.102236767700052,
+      "learning_rate": 2.5722014115803284e-06,
+      "loss": 0.5043,
+      "step": 14856
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 42.18891286850637,
+      "learning_rate": 2.57107241285258e-06,
+      "loss": 0.4472,
+      "step": 14857
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.464202794195778,
+      "learning_rate": 2.569943625399187e-06,
+      "loss": 0.4397,
+      "step": 14858
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.301314694360217,
+      "learning_rate": 2.5688150492522533e-06,
+      "loss": 0.4834,
+      "step": 14859
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 38.41842983805012,
+      "learning_rate": 2.567686684443873e-06,
+      "loss": 0.5286,
+      "step": 14860
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.73256821324706,
+      "learning_rate": 2.5665585310061327e-06,
+      "loss": 0.4968,
+      "step": 14861
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.696909525666674,
+      "learning_rate": 2.5654305889711182e-06,
+      "loss": 0.3983,
+      "step": 14862
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.193662637424165,
+      "learning_rate": 2.564302858370913e-06,
+      "loss": 0.5031,
+      "step": 14863
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 23.51864185688541,
+      "learning_rate": 2.5631753392375846e-06,
+      "loss": 0.438,
+      "step": 14864
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.578503771494018,
+      "learning_rate": 2.562048031603196e-06,
+      "loss": 0.4669,
+      "step": 14865
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.248392692124835,
+      "learning_rate": 2.5609209354998097e-06,
+      "loss": 0.5014,
+      "step": 14866
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.274470300706877,
+      "learning_rate": 2.5597940509594845e-06,
+      "loss": 0.4124,
+      "step": 14867
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.404641410115246,
+      "learning_rate": 2.55866737801426e-06,
+      "loss": 0.4448,
+      "step": 14868
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.988768344908845,
+      "learning_rate": 2.557540916696185e-06,
+      "loss": 0.4244,
+      "step": 14869
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.524144897910464,
+      "learning_rate": 2.5564146670372904e-06,
+      "loss": 0.5063,
+      "step": 14870
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.849123852546313,
+      "learning_rate": 2.5552886290696075e-06,
+      "loss": 0.491,
+      "step": 14871
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.02197733938225,
+      "learning_rate": 2.554162802825165e-06,
+      "loss": 0.5005,
+      "step": 14872
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 28.52587704033205,
+      "learning_rate": 2.553037188335973e-06,
+      "loss": 0.5299,
+      "step": 14873
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 30.571850955114936,
+      "learning_rate": 2.5519117856340503e-06,
+      "loss": 0.444,
+      "step": 14874
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 31.508207762774102,
+      "learning_rate": 2.5507865947513966e-06,
+      "loss": 0.5001,
+      "step": 14875
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.745713434179905,
+      "learning_rate": 2.549661615720017e-06,
+      "loss": 0.4894,
+      "step": 14876
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.509532882464033,
+      "learning_rate": 2.5485368485719e-06,
+      "loss": 0.4895,
+      "step": 14877
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 27.204400296694757,
+      "learning_rate": 2.5474122933390355e-06,
+      "loss": 0.4808,
+      "step": 14878
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.238772374540634,
+      "learning_rate": 2.54628795005341e-06,
+      "loss": 0.4826,
+      "step": 14879
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.43901433442772,
+      "learning_rate": 2.545163818746994e-06,
+      "loss": 0.4471,
+      "step": 14880
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.967828119451124,
+      "learning_rate": 2.5440398994517546e-06,
+      "loss": 0.4474,
+      "step": 14881
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 23.352975465021924,
+      "learning_rate": 2.54291619219966e-06,
+      "loss": 0.4161,
+      "step": 14882
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 20.75014815507415,
+      "learning_rate": 2.541792697022666e-06,
+      "loss": 0.508,
+      "step": 14883
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 32.89221903569449,
+      "learning_rate": 2.540669413952728e-06,
+      "loss": 0.5474,
+      "step": 14884
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.21871007766676,
+      "learning_rate": 2.5395463430217883e-06,
+      "loss": 0.4734,
+      "step": 14885
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.473334047437193,
+      "learning_rate": 2.538423484261783e-06,
+      "loss": 0.4387,
+      "step": 14886
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 23.590986072886295,
+      "learning_rate": 2.537300837704648e-06,
+      "loss": 0.4992,
+      "step": 14887
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.750634755861585,
+      "learning_rate": 2.536178403382317e-06,
+      "loss": 0.4633,
+      "step": 14888
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.826223110092187,
+      "learning_rate": 2.535056181326704e-06,
+      "loss": 0.4211,
+      "step": 14889
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 25.123645260806665,
+      "learning_rate": 2.533934171569724e-06,
+      "loss": 0.4863,
+      "step": 14890
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.663848545394245,
+      "learning_rate": 2.5328123741432887e-06,
+      "loss": 0.4774,
+      "step": 14891
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.10033263763426,
+      "learning_rate": 2.5316907890793007e-06,
+      "loss": 0.4488,
+      "step": 14892
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.353011283085245,
+      "learning_rate": 2.530569416409665e-06,
+      "loss": 0.43,
+      "step": 14893
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.914013441645707,
+      "learning_rate": 2.529448256166259e-06,
+      "loss": 0.5523,
+      "step": 14894
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 22.388623450954917,
+      "learning_rate": 2.5283273083809746e-06,
+      "loss": 0.4799,
+      "step": 14895
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 28.360423670314166,
+      "learning_rate": 2.52720657308569e-06,
+      "loss": 0.5104,
+      "step": 14896
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.648751782449107,
+      "learning_rate": 2.526086050312283e-06,
+      "loss": 0.4867,
+      "step": 14897
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 23.202380727461488,
+      "learning_rate": 2.5249657400926165e-06,
+      "loss": 0.4579,
+      "step": 14898
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.208175456570928,
+      "learning_rate": 2.5238456424585477e-06,
+      "loss": 0.5395,
+      "step": 14899
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.22642357450763,
+      "learning_rate": 2.522725757441936e-06,
+      "loss": 0.523,
+      "step": 14900
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 23.96946688690302,
+      "learning_rate": 2.5216060850746326e-06,
+      "loss": 0.4238,
+      "step": 14901
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.70682862873449,
+      "learning_rate": 2.520486625388475e-06,
+      "loss": 0.5179,
+      "step": 14902
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.26506501552932,
+      "learning_rate": 2.519367378415305e-06,
+      "loss": 0.4175,
+      "step": 14903
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 28.265522268687448,
+      "learning_rate": 2.5182483441869477e-06,
+      "loss": 0.4958,
+      "step": 14904
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.64060572681296,
+      "learning_rate": 2.5171295227352332e-06,
+      "loss": 0.4378,
+      "step": 14905
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.50999705361037,
+      "learning_rate": 2.516010914091975e-06,
+      "loss": 0.5289,
+      "step": 14906
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.060710645198675,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.5828,
+      "step": 14907
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.108628868188458,
+      "learning_rate": 2.5137743353580833e-06,
+      "loss": 0.4716,
+      "step": 14908
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 35.21383427481106,
+      "learning_rate": 2.512656365331053e-06,
+      "loss": 0.5287,
+      "step": 14909
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 22.300545590803143,
+      "learning_rate": 2.5115386082396996e-06,
+      "loss": 0.4895,
+      "step": 14910
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.891203170996736,
+      "learning_rate": 2.5104210641158045e-06,
+      "loss": 0.4978,
+      "step": 14911
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.390122345186295,
+      "learning_rate": 2.509303732991153e-06,
+      "loss": 0.4928,
+      "step": 14912
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.891904415135357,
+      "learning_rate": 2.508186614897524e-06,
+      "loss": 0.5075,
+      "step": 14913
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.8286407349919,
+      "learning_rate": 2.507069709866685e-06,
+      "loss": 0.5637,
+      "step": 14914
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.558267714111223,
+      "learning_rate": 2.5059530179303983e-06,
+      "loss": 0.509,
+      "step": 14915
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.68190799579549,
+      "learning_rate": 2.5048365391204233e-06,
+      "loss": 0.4622,
+      "step": 14916
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 14.198108666433214,
+      "learning_rate": 2.5037202734685117e-06,
+      "loss": 0.4392,
+      "step": 14917
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 44.612409332816014,
+      "learning_rate": 2.502604221006417e-06,
+      "loss": 0.5159,
+      "step": 14918
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.79677335389258,
+      "learning_rate": 2.5014883817658662e-06,
+      "loss": 0.4584,
+      "step": 14919
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.223816181167074,
+      "learning_rate": 2.5003727557785997e-06,
+      "loss": 0.4899,
+      "step": 14920
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.338412216169772,
+      "learning_rate": 2.4992573430763447e-06,
+      "loss": 0.4704,
+      "step": 14921
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.266150324296394,
+      "learning_rate": 2.498142143690826e-06,
+      "loss": 0.4251,
+      "step": 14922
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 26.36628844981907,
+      "learning_rate": 2.4970271576537563e-06,
+      "loss": 0.4463,
+      "step": 14923
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.369843771302754,
+      "learning_rate": 2.4959123849968427e-06,
+      "loss": 0.4128,
+      "step": 14924
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.515989629680922,
+      "learning_rate": 2.4947978257517903e-06,
+      "loss": 0.4476,
+      "step": 14925
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.347493744395427,
+      "learning_rate": 2.4936834799503005e-06,
+      "loss": 0.4164,
+      "step": 14926
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.77561540779018,
+      "learning_rate": 2.4925693476240587e-06,
+      "loss": 0.4677,
+      "step": 14927
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.057707899789197,
+      "learning_rate": 2.4914554288047553e-06,
+      "loss": 0.5281,
+      "step": 14928
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.505151410565166,
+      "learning_rate": 2.4903417235240647e-06,
+      "loss": 0.4587,
+      "step": 14929
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 22.570211867494013,
+      "learning_rate": 2.489228231813665e-06,
+      "loss": 0.4689,
+      "step": 14930
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 26.963006089739416,
+      "learning_rate": 2.4881149537052183e-06,
+      "loss": 0.4297,
+      "step": 14931
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.576649079476116,
+      "learning_rate": 2.4870018892303883e-06,
+      "loss": 0.4819,
+      "step": 14932
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.821338733951984,
+      "learning_rate": 2.485889038420831e-06,
+      "loss": 0.4319,
+      "step": 14933
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.333859014522588,
+      "learning_rate": 2.4847764013081923e-06,
+      "loss": 0.4518,
+      "step": 14934
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.750719861759315,
+      "learning_rate": 2.483663977924119e-06,
+      "loss": 0.5296,
+      "step": 14935
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.035616928027732,
+      "learning_rate": 2.482551768300242e-06,
+      "loss": 0.4475,
+      "step": 14936
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.39575879709027,
+      "learning_rate": 2.4814397724681995e-06,
+      "loss": 0.4389,
+      "step": 14937
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 26.18367423355775,
+      "learning_rate": 2.4803279904596067e-06,
+      "loss": 0.5237,
+      "step": 14938
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.494608101793883,
+      "learning_rate": 2.479216422306091e-06,
+      "loss": 0.3787,
+      "step": 14939
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.0936613904269,
+      "learning_rate": 2.4781050680392572e-06,
+      "loss": 0.4652,
+      "step": 14940
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.019365501805396,
+      "learning_rate": 2.4769939276907163e-06,
+      "loss": 0.4932,
+      "step": 14941
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.607130837220044,
+      "learning_rate": 2.4758830012920688e-06,
+      "loss": 0.4356,
+      "step": 14942
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 32.46969584848756,
+      "learning_rate": 2.4747722888749082e-06,
+      "loss": 0.5453,
+      "step": 14943
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.45369661778227,
+      "learning_rate": 2.4736617904708184e-06,
+      "loss": 0.4634,
+      "step": 14944
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.218689123468057,
+      "learning_rate": 2.4725515061113835e-06,
+      "loss": 0.5306,
+      "step": 14945
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.013655443471304,
+      "learning_rate": 2.471441435828181e-06,
+      "loss": 0.4706,
+      "step": 14946
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.41876768538191,
+      "learning_rate": 2.470331579652784e-06,
+      "loss": 0.5321,
+      "step": 14947
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.032984111707147,
+      "learning_rate": 2.4692219376167504e-06,
+      "loss": 0.417,
+      "step": 14948
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.417769252703222,
+      "learning_rate": 2.4681125097516376e-06,
+      "loss": 0.4101,
+      "step": 14949
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.586297617976975,
+      "learning_rate": 2.467003296088999e-06,
+      "loss": 0.3846,
+      "step": 14950
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.620353305817154,
+      "learning_rate": 2.465894296660383e-06,
+      "loss": 0.4324,
+      "step": 14951
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 18.054366319771617,
+      "learning_rate": 2.464785511497325e-06,
+      "loss": 0.5084,
+      "step": 14952
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 28.02748522360581,
+      "learning_rate": 2.4636769406313576e-06,
+      "loss": 0.4487,
+      "step": 14953
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.005331870006021,
+      "learning_rate": 2.4625685840940083e-06,
+      "loss": 0.419,
+      "step": 14954
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 28.732713900766228,
+      "learning_rate": 2.4614604419167997e-06,
+      "loss": 0.4639,
+      "step": 14955
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.951818094790486,
+      "learning_rate": 2.460352514131249e-06,
+      "loss": 0.4381,
+      "step": 14956
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.367471802114718,
+      "learning_rate": 2.459244800768862e-06,
+      "loss": 0.4936,
+      "step": 14957
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 20.924971382003047,
+      "learning_rate": 2.458137301861139e-06,
+      "loss": 0.4543,
+      "step": 14958
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.393652212000195,
+      "learning_rate": 2.457030017439579e-06,
+      "loss": 0.4744,
+      "step": 14959
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 31.35668543375843,
+      "learning_rate": 2.455922947535675e-06,
+      "loss": 0.4553,
+      "step": 14960
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 21.097606762702817,
+      "learning_rate": 2.4548160921809073e-06,
+      "loss": 0.5306,
+      "step": 14961
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.973161160730598,
+      "learning_rate": 2.4537094514067596e-06,
+      "loss": 0.4066,
+      "step": 14962
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.01353757137663,
+      "learning_rate": 2.4526030252446963e-06,
+      "loss": 0.4172,
+      "step": 14963
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.944544113328586,
+      "learning_rate": 2.4514968137261918e-06,
+      "loss": 0.436,
+      "step": 14964
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 29.676451297500044,
+      "learning_rate": 2.4503908168826985e-06,
+      "loss": 0.4781,
+      "step": 14965
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.34743399293661,
+      "learning_rate": 2.449285034745674e-06,
+      "loss": 0.4258,
+      "step": 14966
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.51332246192332,
+      "learning_rate": 2.4481794673465697e-06,
+      "loss": 0.4292,
+      "step": 14967
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.55031654431513,
+      "learning_rate": 2.4470741147168242e-06,
+      "loss": 0.423,
+      "step": 14968
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.123962968071744,
+      "learning_rate": 2.445968976887868e-06,
+      "loss": 0.4538,
+      "step": 14969
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.20522426428361,
+      "learning_rate": 2.4448640538911363e-06,
+      "loss": 0.4326,
+      "step": 14970
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 21.067320327973807,
+      "learning_rate": 2.443759345758051e-06,
+      "loss": 0.4336,
+      "step": 14971
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.172972162745936,
+      "learning_rate": 2.442654852520032e-06,
+      "loss": 0.4923,
+      "step": 14972
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.07625028037539,
+      "learning_rate": 2.441550574208489e-06,
+      "loss": 0.5377,
+      "step": 14973
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.12302248948293,
+      "learning_rate": 2.4404465108548226e-06,
+      "loss": 0.534,
+      "step": 14974
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.095100328031485,
+      "learning_rate": 2.4393426624904347e-06,
+      "loss": 0.5104,
+      "step": 14975
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.292043007967393,
+      "learning_rate": 2.438239029146723e-06,
+      "loss": 0.4153,
+      "step": 14976
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 31.048400311565103,
+      "learning_rate": 2.437135610855068e-06,
+      "loss": 0.4563,
+      "step": 14977
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.961414223713774,
+      "learning_rate": 2.436032407646851e-06,
+      "loss": 0.4316,
+      "step": 14978
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.57240716296963,
+      "learning_rate": 2.434929419553447e-06,
+      "loss": 0.4457,
+      "step": 14979
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.265405732012503,
+      "learning_rate": 2.4338266466062242e-06,
+      "loss": 0.4394,
+      "step": 14980
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 27.10760443124565,
+      "learning_rate": 2.432724088836549e-06,
+      "loss": 0.4543,
+      "step": 14981
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 34.44885512425127,
+      "learning_rate": 2.431621746275773e-06,
+      "loss": 0.5066,
+      "step": 14982
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 27.4942005230256,
+      "learning_rate": 2.430519618955245e-06,
+      "loss": 0.4659,
+      "step": 14983
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.874659131397916,
+      "learning_rate": 2.429417706906311e-06,
+      "loss": 0.4105,
+      "step": 14984
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.379084559060704,
+      "learning_rate": 2.428316010160312e-06,
+      "loss": 0.5344,
+      "step": 14985
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.240202050230563,
+      "learning_rate": 2.427214528748576e-06,
+      "loss": 0.4811,
+      "step": 14986
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 31.282736381387473,
+      "learning_rate": 2.4261132627024254e-06,
+      "loss": 0.497,
+      "step": 14987
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 26.072178603805863,
+      "learning_rate": 2.425012212053184e-06,
+      "loss": 0.4797,
+      "step": 14988
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.55136766776394,
+      "learning_rate": 2.4239113768321666e-06,
+      "loss": 0.4536,
+      "step": 14989
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.996251690024376,
+      "learning_rate": 2.422810757070675e-06,
+      "loss": 0.4558,
+      "step": 14990
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.55493883545863,
+      "learning_rate": 2.4217103528000153e-06,
+      "loss": 0.4331,
+      "step": 14991
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 21.073009212313934,
+      "learning_rate": 2.4206101640514778e-06,
+      "loss": 0.4405,
+      "step": 14992
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.764856347388857,
+      "learning_rate": 2.4195101908563556e-06,
+      "loss": 0.5174,
+      "step": 14993
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.598587684706972,
+      "learning_rate": 2.4184104332459267e-06,
+      "loss": 0.4127,
+      "step": 14994
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.26352661808254,
+      "learning_rate": 2.4173108912514696e-06,
+      "loss": 0.4606,
+      "step": 14995
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 27.17008360660415,
+      "learning_rate": 2.4162115649042584e-06,
+      "loss": 0.516,
+      "step": 14996
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.425129906671753,
+      "learning_rate": 2.415112454235551e-06,
+      "loss": 0.4293,
+      "step": 14997
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.44199735875639,
+      "learning_rate": 2.41401355927661e-06,
+      "loss": 0.4917,
+      "step": 14998
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.851185743651754,
+      "learning_rate": 2.4129148800586846e-06,
+      "loss": 0.4333,
+      "step": 14999
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.86725538508925,
+      "learning_rate": 2.411816416613021e-06,
+      "loss": 0.4597,
+      "step": 15000
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.98778962989567,
+      "learning_rate": 2.4107181689708614e-06,
+      "loss": 0.5131,
+      "step": 15001
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.802075302152716,
+      "learning_rate": 2.4096201371634387e-06,
+      "loss": 0.4535,
+      "step": 15002
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.372028739991396,
+      "learning_rate": 2.4085223212219767e-06,
+      "loss": 0.4252,
+      "step": 15003
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 20.576019779724053,
+      "learning_rate": 2.407424721177698e-06,
+      "loss": 0.4688,
+      "step": 15004
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.118607587257904,
+      "learning_rate": 2.406327337061818e-06,
+      "loss": 0.4039,
+      "step": 15005
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.77158144537063,
+      "learning_rate": 2.4052301689055536e-06,
+      "loss": 0.5015,
+      "step": 15006
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.119996615644077,
+      "learning_rate": 2.4041332167400944e-06,
+      "loss": 0.5551,
+      "step": 15007
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.989182301652665,
+      "learning_rate": 2.4030364805966423e-06,
+      "loss": 0.4872,
+      "step": 15008
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 20.823894124675242,
+      "learning_rate": 2.401939960506389e-06,
+      "loss": 0.4754,
+      "step": 15009
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.71575180119783,
+      "learning_rate": 2.400843656500521e-06,
+      "loss": 0.4743,
+      "step": 15010
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.538955506135007,
+      "learning_rate": 2.3997475686102147e-06,
+      "loss": 0.4732,
+      "step": 15011
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.860486072159112,
+      "learning_rate": 2.3986516968666385e-06,
+      "loss": 0.4324,
+      "step": 15012
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.621210674465352,
+      "learning_rate": 2.397556041300961e-06,
+      "loss": 0.4393,
+      "step": 15013
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.63412330997982,
+      "learning_rate": 2.396460601944346e-06,
+      "loss": 0.414,
+      "step": 15014
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 25.2569740656539,
+      "learning_rate": 2.3953653788279395e-06,
+      "loss": 0.5455,
+      "step": 15015
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 27.54138978255306,
+      "learning_rate": 2.3942703719828965e-06,
+      "loss": 0.531,
+      "step": 15016
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.553068131286206,
+      "learning_rate": 2.393175581440352e-06,
+      "loss": 0.4352,
+      "step": 15017
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 24.916161770207733,
+      "learning_rate": 2.3920810072314473e-06,
+      "loss": 0.5124,
+      "step": 15018
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.102864184970702,
+      "learning_rate": 2.3909866493873048e-06,
+      "loss": 0.4557,
+      "step": 15019
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.91727923458385,
+      "learning_rate": 2.389892507939051e-06,
+      "loss": 0.4645,
+      "step": 15020
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.530191500294986,
+      "learning_rate": 2.388798582917806e-06,
+      "loss": 0.4213,
+      "step": 15021
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.047844349602368,
+      "learning_rate": 2.387704874354674e-06,
+      "loss": 0.4513,
+      "step": 15022
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 23.088735549655407,
+      "learning_rate": 2.3866113822807645e-06,
+      "loss": 0.508,
+      "step": 15023
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.924472581369244,
+      "learning_rate": 2.3855181067271703e-06,
+      "loss": 0.4427,
+      "step": 15024
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.682589026198844,
+      "learning_rate": 2.3844250477249877e-06,
+      "loss": 0.4234,
+      "step": 15025
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 30.908038435221503,
+      "learning_rate": 2.3833322053053045e-06,
+      "loss": 0.4948,
+      "step": 15026
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.868917020265116,
+      "learning_rate": 2.382239579499198e-06,
+      "loss": 0.4493,
+      "step": 15027
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 25.83841338404869,
+      "learning_rate": 2.3811471703377376e-06,
+      "loss": 0.4761,
+      "step": 15028
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.192612713754983,
+      "learning_rate": 2.3800549778519956e-06,
+      "loss": 0.4263,
+      "step": 15029
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.85421361513422,
+      "learning_rate": 2.3789630020730327e-06,
+      "loss": 0.5324,
+      "step": 15030
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.814362989740594,
+      "learning_rate": 2.3778712430319095e-06,
+      "loss": 0.419,
+      "step": 15031
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 24.775991052577602,
+      "learning_rate": 2.376779700759664e-06,
+      "loss": 0.4638,
+      "step": 15032
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 26.492217591205897,
+      "learning_rate": 2.3756883752873426e-06,
+      "loss": 0.4974,
+      "step": 15033
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.598084423178136,
+      "learning_rate": 2.3745972666459847e-06,
+      "loss": 0.4467,
+      "step": 15034
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.346625671265098,
+      "learning_rate": 2.373506374866623e-06,
+      "loss": 0.4888,
+      "step": 15035
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.6641484038716,
+      "learning_rate": 2.3724156999802795e-06,
+      "loss": 0.4539,
+      "step": 15036
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.10505575538282,
+      "learning_rate": 2.3713252420179668e-06,
+      "loss": 0.4668,
+      "step": 15037
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.419118771799873,
+      "learning_rate": 2.3702350010107023e-06,
+      "loss": 0.4351,
+      "step": 15038
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.85497914012074,
+      "learning_rate": 2.3691449769894946e-06,
+      "loss": 0.4623,
+      "step": 15039
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.339209468551555,
+      "learning_rate": 2.368055169985339e-06,
+      "loss": 0.41,
+      "step": 15040
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.87039009224215,
+      "learning_rate": 2.366965580029227e-06,
+      "loss": 0.4823,
+      "step": 15041
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.988464601056158,
+      "learning_rate": 2.365876207152149e-06,
+      "loss": 0.4696,
+      "step": 15042
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.337716538432867,
+      "learning_rate": 2.3647870513850857e-06,
+      "loss": 0.4161,
+      "step": 15043
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.433713748895297,
+      "learning_rate": 2.3636981127590155e-06,
+      "loss": 0.4285,
+      "step": 15044
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.199931924794557,
+      "learning_rate": 2.362609391304903e-06,
+      "loss": 0.419,
+      "step": 15045
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 29.65361755865329,
+      "learning_rate": 2.361520887053709e-06,
+      "loss": 0.4675,
+      "step": 15046
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 33.776858864836804,
+      "learning_rate": 2.360432600036393e-06,
+      "loss": 0.4308,
+      "step": 15047
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 32.109737641080365,
+      "learning_rate": 2.359344530283908e-06,
+      "loss": 0.6344,
+      "step": 15048
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.87092471627326,
+      "learning_rate": 2.358256677827191e-06,
+      "loss": 0.3783,
+      "step": 15049
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 25.31178849185736,
+      "learning_rate": 2.357169042697187e-06,
+      "loss": 0.5111,
+      "step": 15050
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 21.185630606155875,
+      "learning_rate": 2.356081624924822e-06,
+      "loss": 0.423,
+      "step": 15051
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.59275958254778,
+      "learning_rate": 2.354994424541027e-06,
+      "loss": 0.4502,
+      "step": 15052
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 26.230387707866985,
+      "learning_rate": 2.353907441576715e-06,
+      "loss": 0.4714,
+      "step": 15053
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 23.36885248844274,
+      "learning_rate": 2.3528206760628026e-06,
+      "loss": 0.5002,
+      "step": 15054
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.731884886518342,
+      "learning_rate": 2.3517341280301997e-06,
+      "loss": 0.5039,
+      "step": 15055
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.666482198200214,
+      "learning_rate": 2.350647797509804e-06,
+      "loss": 0.5093,
+      "step": 15056
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.761040849150536,
+      "learning_rate": 2.3495616845325074e-06,
+      "loss": 0.4546,
+      "step": 15057
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.422941218373836,
+      "learning_rate": 2.348475789129202e-06,
+      "loss": 0.4277,
+      "step": 15058
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.727040200133143,
+      "learning_rate": 2.3473901113307685e-06,
+      "loss": 0.4747,
+      "step": 15059
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.24599733125058,
+      "learning_rate": 2.3463046511680864e-06,
+      "loss": 0.4769,
+      "step": 15060
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.143046850149158,
+      "learning_rate": 2.3452194086720225e-06,
+      "loss": 0.462,
+      "step": 15061
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.73811209328546,
+      "learning_rate": 2.3441343838734375e-06,
+      "loss": 0.4421,
+      "step": 15062
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.13671426798326,
+      "learning_rate": 2.3430495768031923e-06,
+      "loss": 0.4325,
+      "step": 15063
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.814185855241742,
+      "learning_rate": 2.3419649874921413e-06,
+      "loss": 0.4137,
+      "step": 15064
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.830912419901717,
+      "learning_rate": 2.3408806159711262e-06,
+      "loss": 0.3923,
+      "step": 15065
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 27.002734095504596,
+      "learning_rate": 2.3397964622709825e-06,
+      "loss": 0.5272,
+      "step": 15066
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.237510153146669,
+      "learning_rate": 2.3387125264225475e-06,
+      "loss": 0.4649,
+      "step": 15067
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 21.016762106384046,
+      "learning_rate": 2.3376288084566445e-06,
+      "loss": 0.4739,
+      "step": 15068
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 24.655839518356228,
+      "learning_rate": 2.3365453084041e-06,
+      "loss": 0.5133,
+      "step": 15069
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 23.933735255113675,
+      "learning_rate": 2.3354620262957238e-06,
+      "loss": 0.468,
+      "step": 15070
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 25.10690323699414,
+      "learning_rate": 2.3343789621623213e-06,
+      "loss": 0.4302,
+      "step": 15071
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 27.086627692293746,
+      "learning_rate": 2.3332961160346966e-06,
+      "loss": 0.4648,
+      "step": 15072
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 25.18997946904562,
+      "learning_rate": 2.3322134879436487e-06,
+      "loss": 0.4564,
+      "step": 15073
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 23.97426075171063,
+      "learning_rate": 2.3311310779199603e-06,
+      "loss": 0.4732,
+      "step": 15074
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.09005984624344,
+      "learning_rate": 2.3300488859944217e-06,
+      "loss": 0.4992,
+      "step": 15075
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 32.03438821631756,
+      "learning_rate": 2.328966912197802e-06,
+      "loss": 0.5262,
+      "step": 15076
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.092950243447078,
+      "learning_rate": 2.3278851565608782e-06,
+      "loss": 0.4489,
+      "step": 15077
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.48221056060619,
+      "learning_rate": 2.3268036191144117e-06,
+      "loss": 0.4537,
+      "step": 15078
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.29270156384741,
+      "learning_rate": 2.3257222998891603e-06,
+      "loss": 0.4318,
+      "step": 15079
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 28.135608742821066,
+      "learning_rate": 2.32464119891588e-06,
+      "loss": 0.5001,
+      "step": 15080
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.235043348267983,
+      "learning_rate": 2.323560316225314e-06,
+      "loss": 0.4727,
+      "step": 15081
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.84446794752678,
+      "learning_rate": 2.3224796518481995e-06,
+      "loss": 0.4851,
+      "step": 15082
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.97587809733731,
+      "learning_rate": 2.3213992058152733e-06,
+      "loss": 0.5008,
+      "step": 15083
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 26.42728099584931,
+      "learning_rate": 2.320318978157263e-06,
+      "loss": 0.514,
+      "step": 15084
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 34.64607227686141,
+      "learning_rate": 2.319238968904888e-06,
+      "loss": 0.4717,
+      "step": 15085
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.462435583824682,
+      "learning_rate": 2.318159178088865e-06,
+      "loss": 0.4446,
+      "step": 15086
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 27.888941684649076,
+      "learning_rate": 2.317079605739898e-06,
+      "loss": 0.4867,
+      "step": 15087
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.10364849224933,
+      "learning_rate": 2.3160002518886927e-06,
+      "loss": 0.498,
+      "step": 15088
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.953757768548204,
+      "learning_rate": 2.314921116565948e-06,
+      "loss": 0.4684,
+      "step": 15089
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.420475021522208,
+      "learning_rate": 2.313842199802352e-06,
+      "loss": 0.4878,
+      "step": 15090
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.036588174673145,
+      "learning_rate": 2.312763501628584e-06,
+      "loss": 0.5103,
+      "step": 15091
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 27.239336427453367,
+      "learning_rate": 2.3116850220753253e-06,
+      "loss": 0.4904,
+      "step": 15092
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.92729935643449,
+      "learning_rate": 2.3106067611732473e-06,
+      "loss": 0.4889,
+      "step": 15093
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.614072179281564,
+      "learning_rate": 2.3095287189530203e-06,
+      "loss": 0.5348,
+      "step": 15094
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 26.012088577086438,
+      "learning_rate": 2.3084508954452923e-06,
+      "loss": 0.4741,
+      "step": 15095
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.222409866091148,
+      "learning_rate": 2.307373290680721e-06,
+      "loss": 0.424,
+      "step": 15096
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 36.56650173711479,
+      "learning_rate": 2.3062959046899535e-06,
+      "loss": 0.5467,
+      "step": 15097
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.061730655844066,
+      "learning_rate": 2.3052187375036327e-06,
+      "loss": 0.4801,
+      "step": 15098
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 27.289106672585966,
+      "learning_rate": 2.3041417891523897e-06,
+      "loss": 0.4666,
+      "step": 15099
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.583160254711412,
+      "learning_rate": 2.303065059666849e-06,
+      "loss": 0.4162,
+      "step": 15100
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.483103717779745,
+      "learning_rate": 2.301988549077636e-06,
+      "loss": 0.4439,
+      "step": 15101
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.947887461905324,
+      "learning_rate": 2.3009122574153673e-06,
+      "loss": 0.4959,
+      "step": 15102
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.980107709409623,
+      "learning_rate": 2.2998361847106487e-06,
+      "loss": 0.45,
+      "step": 15103
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.862130669069074,
+      "learning_rate": 2.2987603309940866e-06,
+      "loss": 0.4614,
+      "step": 15104
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.08831933421522,
+      "learning_rate": 2.297684696296273e-06,
+      "loss": 0.4327,
+      "step": 15105
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.927605998576137,
+      "learning_rate": 2.2966092806478e-06,
+      "loss": 0.4253,
+      "step": 15106
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 19.731646935165134,
+      "learning_rate": 2.2955340840792563e-06,
+      "loss": 0.4872,
+      "step": 15107
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.502861993208516,
+      "learning_rate": 2.294459106621214e-06,
+      "loss": 0.505,
+      "step": 15108
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 27.69800669017182,
+      "learning_rate": 2.2933843483042495e-06,
+      "loss": 0.5039,
+      "step": 15109
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 19.162776566417605,
+      "learning_rate": 2.2923098091589224e-06,
+      "loss": 0.4361,
+      "step": 15110
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.103642994710736,
+      "learning_rate": 2.2912354892157995e-06,
+      "loss": 0.4665,
+      "step": 15111
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.710251316525817,
+      "learning_rate": 2.2901613885054252e-06,
+      "loss": 0.4314,
+      "step": 15112
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.402614824715194,
+      "learning_rate": 2.289087507058353e-06,
+      "loss": 0.4764,
+      "step": 15113
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.036047549123868,
+      "learning_rate": 2.2880138449051227e-06,
+      "loss": 0.4791,
+      "step": 15114
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.06097541706953,
+      "learning_rate": 2.2869404020762676e-06,
+      "loss": 0.4918,
+      "step": 15115
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.834690349172593,
+      "learning_rate": 2.285867178602312e-06,
+      "loss": 0.4796,
+      "step": 15116
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.41086027666266,
+      "learning_rate": 2.2847941745137826e-06,
+      "loss": 0.4671,
+      "step": 15117
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.05177698987236,
+      "learning_rate": 2.2837213898411925e-06,
+      "loss": 0.3635,
+      "step": 15118
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 19.04124619855663,
+      "learning_rate": 2.282648824615059e-06,
+      "loss": 0.5058,
+      "step": 15119
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 27.816393747904115,
+      "learning_rate": 2.281576478865872e-06,
+      "loss": 0.4795,
+      "step": 15120
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.70489589203088,
+      "learning_rate": 2.2805043526241342e-06,
+      "loss": 0.485,
+      "step": 15121
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 28.742116212043584,
+      "learning_rate": 2.2794324459203377e-06,
+      "loss": 0.4964,
+      "step": 15122
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.5325057086598,
+      "learning_rate": 2.278360758784969e-06,
+      "loss": 0.4472,
+      "step": 15123
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.80214907767873,
+      "learning_rate": 2.277289291248502e-06,
+      "loss": 0.4132,
+      "step": 15124
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.10269729138953,
+      "learning_rate": 2.2762180433414083e-06,
+      "loss": 0.4791,
+      "step": 15125
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.031731506011766,
+      "learning_rate": 2.2751470150941558e-06,
+      "loss": 0.5251,
+      "step": 15126
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 26.074122114823965,
+      "learning_rate": 2.2740762065372046e-06,
+      "loss": 0.3994,
+      "step": 15127
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.40802495369327,
+      "learning_rate": 2.2730056177010075e-06,
+      "loss": 0.4494,
+      "step": 15128
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.072359591101424,
+      "learning_rate": 2.2719352486160074e-06,
+      "loss": 0.468,
+      "step": 15129
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.795083219011353,
+      "learning_rate": 2.270865099312648e-06,
+      "loss": 0.3959,
+      "step": 15130
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 34.55590289826952,
+      "learning_rate": 2.2697951698213628e-06,
+      "loss": 0.4944,
+      "step": 15131
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.966692484312077,
+      "learning_rate": 2.268725460172585e-06,
+      "loss": 0.4744,
+      "step": 15132
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.845703329157487,
+      "learning_rate": 2.2676559703967317e-06,
+      "loss": 0.5224,
+      "step": 15133
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 25.79507267395847,
+      "learning_rate": 2.266586700524217e-06,
+      "loss": 0.4531,
+      "step": 15134
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.88192124289535,
+      "learning_rate": 2.2655176505854527e-06,
+      "loss": 0.5386,
+      "step": 15135
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 26.416832025854777,
+      "learning_rate": 2.264448820610844e-06,
+      "loss": 0.4835,
+      "step": 15136
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 27.42077197222985,
+      "learning_rate": 2.2633802106307835e-06,
+      "loss": 0.4408,
+      "step": 15137
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.94313458717503,
+      "learning_rate": 2.2623118206756668e-06,
+      "loss": 0.4415,
+      "step": 15138
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.750516646427478,
+      "learning_rate": 2.2612436507758726e-06,
+      "loss": 0.4662,
+      "step": 15139
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 19.288046312969854,
+      "learning_rate": 2.260175700961785e-06,
+      "loss": 0.4662,
+      "step": 15140
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.896086577545045,
+      "learning_rate": 2.2591079712637698e-06,
+      "loss": 0.4487,
+      "step": 15141
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.10662859500066,
+      "learning_rate": 2.2580404617121954e-06,
+      "loss": 0.4259,
+      "step": 15142
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 25.261615140085144,
+      "learning_rate": 2.2569731723374243e-06,
+      "loss": 0.4717,
+      "step": 15143
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 26.266033891062857,
+      "learning_rate": 2.255906103169806e-06,
+      "loss": 0.4583,
+      "step": 15144
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 22.01899804807438,
+      "learning_rate": 2.2548392542396856e-06,
+      "loss": 0.5174,
+      "step": 15145
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 25.078344651921057,
+      "learning_rate": 2.2537726255774063e-06,
+      "loss": 0.5355,
+      "step": 15146
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.49513815217672,
+      "learning_rate": 2.252706217213302e-06,
+      "loss": 0.4529,
+      "step": 15147
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.20640744803605,
+      "learning_rate": 2.251640029177704e-06,
+      "loss": 0.4513,
+      "step": 15148
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 22.20882558426196,
+      "learning_rate": 2.250574061500931e-06,
+      "loss": 0.4744,
+      "step": 15149
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.609538546018904,
+      "learning_rate": 2.2495083142132944e-06,
+      "loss": 0.4265,
+      "step": 15150
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 26.405454795195126,
+      "learning_rate": 2.2484427873451086e-06,
+      "loss": 0.475,
+      "step": 15151
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.365715220181023,
+      "learning_rate": 2.247377480926678e-06,
+      "loss": 0.4119,
+      "step": 15152
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.081622743119825,
+      "learning_rate": 2.246312394988296e-06,
+      "loss": 0.4392,
+      "step": 15153
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.932720703552434,
+      "learning_rate": 2.2452475295602518e-06,
+      "loss": 0.4564,
+      "step": 15154
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.11219220962849,
+      "learning_rate": 2.2441828846728307e-06,
+      "loss": 0.4462,
+      "step": 15155
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.680290516208032,
+      "learning_rate": 2.243118460356312e-06,
+      "loss": 0.4261,
+      "step": 15156
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.008526355659011,
+      "learning_rate": 2.2420542566409686e-06,
+      "loss": 0.4033,
+      "step": 15157
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.946891196248757,
+      "learning_rate": 2.2409902735570643e-06,
+      "loss": 0.4615,
+      "step": 15158
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.7169714617309,
+      "learning_rate": 2.2399265111348558e-06,
+      "loss": 0.5126,
+      "step": 15159
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 24.489491198999364,
+      "learning_rate": 2.238862969404596e-06,
+      "loss": 0.4774,
+      "step": 15160
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.558503354134128,
+      "learning_rate": 2.2377996483965368e-06,
+      "loss": 0.4408,
+      "step": 15161
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 24.781019896440743,
+      "learning_rate": 2.236736548140913e-06,
+      "loss": 0.5461,
+      "step": 15162
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.67023745713467,
+      "learning_rate": 2.2356736686679624e-06,
+      "loss": 0.504,
+      "step": 15163
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.553010393576393,
+      "learning_rate": 2.2346110100079076e-06,
+      "loss": 0.4241,
+      "step": 15164
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.042129224106823,
+      "learning_rate": 2.2335485721909766e-06,
+      "loss": 0.4849,
+      "step": 15165
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.516321256112047,
+      "learning_rate": 2.2324863552473776e-06,
+      "loss": 0.4786,
+      "step": 15166
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 30.363196988407843,
+      "learning_rate": 2.2314243592073226e-06,
+      "loss": 0.4663,
+      "step": 15167
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.034170561528624,
+      "learning_rate": 2.230362584101018e-06,
+      "loss": 0.4426,
+      "step": 15168
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.68042147095207,
+      "learning_rate": 2.229301029958656e-06,
+      "loss": 0.4172,
+      "step": 15169
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.598652276949753,
+      "learning_rate": 2.2282396968104235e-06,
+      "loss": 0.4323,
+      "step": 15170
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.253130448943175,
+      "learning_rate": 2.227178584686509e-06,
+      "loss": 0.4624,
+      "step": 15171
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 28.68904261295678,
+      "learning_rate": 2.226117693617088e-06,
+      "loss": 0.4532,
+      "step": 15172
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.15407636569071,
+      "learning_rate": 2.2250570236323344e-06,
+      "loss": 0.4531,
+      "step": 15173
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.526510357836578,
+      "learning_rate": 2.2239965747624117e-06,
+      "loss": 0.4421,
+      "step": 15174
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 22.863900434829066,
+      "learning_rate": 2.222936347037474e-06,
+      "loss": 0.5133,
+      "step": 15175
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 37.661697840448895,
+      "learning_rate": 2.2218763404876775e-06,
+      "loss": 0.5123,
+      "step": 15176
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 35.003177265298156,
+      "learning_rate": 2.2208165551431706e-06,
+      "loss": 0.4869,
+      "step": 15177
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.896395629073762,
+      "learning_rate": 2.21975699103409e-06,
+      "loss": 0.4571,
+      "step": 15178
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.09295527842895,
+      "learning_rate": 2.2186976481905663e-06,
+      "loss": 0.4676,
+      "step": 15179
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 33.75960256301567,
+      "learning_rate": 2.2176385266427302e-06,
+      "loss": 0.5048,
+      "step": 15180
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.864206347047276,
+      "learning_rate": 2.216579626420702e-06,
+      "loss": 0.4616,
+      "step": 15181
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.9554156526153,
+      "learning_rate": 2.2155209475546013e-06,
+      "loss": 0.4541,
+      "step": 15182
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 27.950230723146994,
+      "learning_rate": 2.2144624900745247e-06,
+      "loss": 0.4778,
+      "step": 15183
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.409156613119322,
+      "learning_rate": 2.2134042540105814e-06,
+      "loss": 0.4772,
+      "step": 15184
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.49391084901629,
+      "learning_rate": 2.2123462393928663e-06,
+      "loss": 0.4264,
+      "step": 15185
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.893617807793394,
+      "learning_rate": 2.2112884462514707e-06,
+      "loss": 0.4835,
+      "step": 15186
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.10285137495659,
+      "learning_rate": 2.210230874616475e-06,
+      "loss": 0.4756,
+      "step": 15187
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.442786691519522,
+      "learning_rate": 2.2091735245179548e-06,
+      "loss": 0.4194,
+      "step": 15188
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 25.597548588922628,
+      "learning_rate": 2.208116395985981e-06,
+      "loss": 0.4776,
+      "step": 15189
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.568131821060923,
+      "learning_rate": 2.2070594890506216e-06,
+      "loss": 0.4597,
+      "step": 15190
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 31.14080444529234,
+      "learning_rate": 2.206002803741929e-06,
+      "loss": 0.4679,
+      "step": 15191
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.002903793213502,
+      "learning_rate": 2.2049463400899606e-06,
+      "loss": 0.4415,
+      "step": 15192
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.99219630971295,
+      "learning_rate": 2.2038900981247545e-06,
+      "loss": 0.455,
+      "step": 15193
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.539661793398427,
+      "learning_rate": 2.2028340778763544e-06,
+      "loss": 0.4065,
+      "step": 15194
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.403671257978583,
+      "learning_rate": 2.201778279374794e-06,
+      "loss": 0.5221,
+      "step": 15195
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.47130216310369,
+      "learning_rate": 2.2007227026500956e-06,
+      "loss": 0.4356,
+      "step": 15196
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.045397497759183,
+      "learning_rate": 2.199667347732284e-06,
+      "loss": 0.5095,
+      "step": 15197
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.87522694578659,
+      "learning_rate": 2.1986122146513654e-06,
+      "loss": 0.4589,
+      "step": 15198
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.537142542464714,
+      "learning_rate": 2.1975573034373563e-06,
+      "loss": 0.4946,
+      "step": 15199
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.27352048351801,
+      "learning_rate": 2.1965026141202495e-06,
+      "loss": 0.4413,
+      "step": 15200
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.423561029841085,
+      "learning_rate": 2.195448146730044e-06,
+      "loss": 0.4912,
+      "step": 15201
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 26.706838841712543,
+      "learning_rate": 2.1943939012967295e-06,
+      "loss": 0.5352,
+      "step": 15202
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.325223293023893,
+      "learning_rate": 2.193339877850288e-06,
+      "loss": 0.4296,
+      "step": 15203
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 30.31120406408753,
+      "learning_rate": 2.19228607642069e-06,
+      "loss": 0.4846,
+      "step": 15204
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 22.647088281393838,
+      "learning_rate": 2.1912324970379084e-06,
+      "loss": 0.4687,
+      "step": 15205
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.04969849370558,
+      "learning_rate": 2.190179139731906e-06,
+      "loss": 0.4253,
+      "step": 15206
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 27.21354589410533,
+      "learning_rate": 2.1891260045326467e-06,
+      "loss": 0.4336,
+      "step": 15207
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.34410482362443,
+      "learning_rate": 2.1880730914700687e-06,
+      "loss": 0.3635,
+      "step": 15208
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.928483493779746,
+      "learning_rate": 2.1870204005741226e-06,
+      "loss": 0.4417,
+      "step": 15209
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 26.76386835443305,
+      "learning_rate": 2.185967931874745e-06,
+      "loss": 0.5163,
+      "step": 15210
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 23.337665815050382,
+      "learning_rate": 2.1849156854018715e-06,
+      "loss": 0.4789,
+      "step": 15211
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.98255458471436,
+      "learning_rate": 2.1838636611854246e-06,
+      "loss": 0.4906,
+      "step": 15212
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.883022576260995,
+      "learning_rate": 2.1828118592553195e-06,
+      "loss": 0.4344,
+      "step": 15213
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.319707726045827,
+      "learning_rate": 2.181760279641473e-06,
+      "loss": 0.4343,
+      "step": 15214
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 24.73896608619982,
+      "learning_rate": 2.180708922373792e-06,
+      "loss": 0.5158,
+      "step": 15215
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.008709923474523,
+      "learning_rate": 2.1796577874821734e-06,
+      "loss": 0.4283,
+      "step": 15216
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 26.025005388396746,
+      "learning_rate": 2.178606874996515e-06,
+      "loss": 0.5685,
+      "step": 15217
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 34.11862513931091,
+      "learning_rate": 2.1775561849466987e-06,
+      "loss": 0.5322,
+      "step": 15218
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.529657402317886,
+      "learning_rate": 2.176505717362609e-06,
+      "loss": 0.3853,
+      "step": 15219
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 23.28974909187821,
+      "learning_rate": 2.1754554722741227e-06,
+      "loss": 0.469,
+      "step": 15220
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 24.327234389088307,
+      "learning_rate": 2.1744054497111033e-06,
+      "loss": 0.5213,
+      "step": 15221
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.708528128637457,
+      "learning_rate": 2.173355649703417e-06,
+      "loss": 0.4262,
+      "step": 15222
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 25.981680667860044,
+      "learning_rate": 2.1723060722809155e-06,
+      "loss": 0.4433,
+      "step": 15223
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.021885487173225,
+      "learning_rate": 2.1712567174734523e-06,
+      "loss": 0.4914,
+      "step": 15224
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.080261608435933,
+      "learning_rate": 2.170207585310865e-06,
+      "loss": 0.4246,
+      "step": 15225
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 20.35669418406093,
+      "learning_rate": 2.1691586758229976e-06,
+      "loss": 0.4319,
+      "step": 15226
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 24.717864791727223,
+      "learning_rate": 2.168109989039674e-06,
+      "loss": 0.5165,
+      "step": 15227
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.313844705728073,
+      "learning_rate": 2.1670615249907234e-06,
+      "loss": 0.4082,
+      "step": 15228
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 22.27455009804577,
+      "learning_rate": 2.1660132837059576e-06,
+      "loss": 0.427,
+      "step": 15229
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 28.38154265305997,
+      "learning_rate": 2.164965265215191e-06,
+      "loss": 0.5739,
+      "step": 15230
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.880419604734247,
+      "learning_rate": 2.1639174695482322e-06,
+      "loss": 0.4886,
+      "step": 15231
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 20.93977163218643,
+      "learning_rate": 2.162869896734876e-06,
+      "loss": 0.5195,
+      "step": 15232
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.304286216475976,
+      "learning_rate": 2.161822546804914e-06,
+      "loss": 0.4787,
+      "step": 15233
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.907581546883872,
+      "learning_rate": 2.1607754197881324e-06,
+      "loss": 0.4781,
+      "step": 15234
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.136662726528343,
+      "learning_rate": 2.1597285157143122e-06,
+      "loss": 0.529,
+      "step": 15235
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.70373016542226,
+      "learning_rate": 2.15868183461323e-06,
+      "loss": 0.4257,
+      "step": 15236
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.803857775908057,
+      "learning_rate": 2.1576353765146486e-06,
+      "loss": 0.4258,
+      "step": 15237
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.926739208091387,
+      "learning_rate": 2.1565891414483266e-06,
+      "loss": 0.5076,
+      "step": 15238
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.914958365823445,
+      "learning_rate": 2.155543129444021e-06,
+      "loss": 0.4534,
+      "step": 15239
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.620965244687518,
+      "learning_rate": 2.154497340531484e-06,
+      "loss": 0.4115,
+      "step": 15240
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 20.489008931491618,
+      "learning_rate": 2.1534517747404516e-06,
+      "loss": 0.4078,
+      "step": 15241
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.020740197787905,
+      "learning_rate": 2.152406432100659e-06,
+      "loss": 0.4945,
+      "step": 15242
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 34.305891782958305,
+      "learning_rate": 2.1513613126418364e-06,
+      "loss": 0.5067,
+      "step": 15243
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 32.25065685682638,
+      "learning_rate": 2.1503164163937062e-06,
+      "loss": 0.5569,
+      "step": 15244
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.246492199405413,
+      "learning_rate": 2.1492717433859887e-06,
+      "loss": 0.4588,
+      "step": 15245
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.875655993289474,
+      "learning_rate": 2.1482272936483896e-06,
+      "loss": 0.4507,
+      "step": 15246
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 25.398889432586618,
+      "learning_rate": 2.147183067210611e-06,
+      "loss": 0.4635,
+      "step": 15247
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.829173364897283,
+      "learning_rate": 2.1461390641023516e-06,
+      "loss": 0.3928,
+      "step": 15248
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.466488108280227,
+      "learning_rate": 2.145095284353307e-06,
+      "loss": 0.4859,
+      "step": 15249
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 22.865140941282668,
+      "learning_rate": 2.1440517279931528e-06,
+      "loss": 0.4691,
+      "step": 15250
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 26.47114574214595,
+      "learning_rate": 2.1430083950515755e-06,
+      "loss": 0.5508,
+      "step": 15251
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 20.067723503341373,
+      "learning_rate": 2.1419652855582406e-06,
+      "loss": 0.4572,
+      "step": 15252
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.574947489603044,
+      "learning_rate": 2.1409223995428187e-06,
+      "loss": 0.4094,
+      "step": 15253
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.520740203244234,
+      "learning_rate": 2.1398797370349644e-06,
+      "loss": 0.4278,
+      "step": 15254
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.120394540791928,
+      "learning_rate": 2.1388372980643315e-06,
+      "loss": 0.4548,
+      "step": 15255
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.811261880598597,
+      "learning_rate": 2.13779508266057e-06,
+      "loss": 0.4311,
+      "step": 15256
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.838750567266487,
+      "learning_rate": 2.136753090853314e-06,
+      "loss": 0.4487,
+      "step": 15257
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 27.622548763507183,
+      "learning_rate": 2.1357113226722036e-06,
+      "loss": 0.4885,
+      "step": 15258
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.994836262835964,
+      "learning_rate": 2.1346697781468593e-06,
+      "loss": 0.4761,
+      "step": 15259
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.120743627985972,
+      "learning_rate": 2.1336284573069067e-06,
+      "loss": 0.4331,
+      "step": 15260
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.730562044693688,
+      "learning_rate": 2.1325873601819613e-06,
+      "loss": 0.4798,
+      "step": 15261
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 20.816124974843554,
+      "learning_rate": 2.1315464868016287e-06,
+      "loss": 0.4176,
+      "step": 15262
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.066580148553282,
+      "learning_rate": 2.130505837195508e-06,
+      "loss": 0.4189,
+      "step": 15263
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 24.81745196480578,
+      "learning_rate": 2.129465411393198e-06,
+      "loss": 0.4104,
+      "step": 15264
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.185868853329204,
+      "learning_rate": 2.1284252094242908e-06,
+      "loss": 0.4024,
+      "step": 15265
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 28.804102140810347,
+      "learning_rate": 2.1273852313183663e-06,
+      "loss": 0.5208,
+      "step": 15266
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 23.183263133904518,
+      "learning_rate": 2.126345477104996e-06,
+      "loss": 0.4094,
+      "step": 15267
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.447683793891619,
+      "learning_rate": 2.125305946813756e-06,
+      "loss": 0.464,
+      "step": 15268
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.236747514415185,
+      "learning_rate": 2.1242666404742074e-06,
+      "loss": 0.4207,
+      "step": 15269
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 35.1851084101385,
+      "learning_rate": 2.1232275581159123e-06,
+      "loss": 0.4921,
+      "step": 15270
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 27.142526780668916,
+      "learning_rate": 2.122188699768416e-06,
+      "loss": 0.5261,
+      "step": 15271
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 26.914903823591946,
+      "learning_rate": 2.1211500654612625e-06,
+      "loss": 0.4496,
+      "step": 15272
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.448980001916134,
+      "learning_rate": 2.120111655223993e-06,
+      "loss": 0.449,
+      "step": 15273
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 25.74322992625142,
+      "learning_rate": 2.1190734690861403e-06,
+      "loss": 0.4286,
+      "step": 15274
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.213273706092936,
+      "learning_rate": 2.1180355070772287e-06,
+      "loss": 0.3943,
+      "step": 15275
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 20.0684325212691,
+      "learning_rate": 2.116997769226773e-06,
+      "loss": 0.4632,
+      "step": 15276
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.165732116302507,
+      "learning_rate": 2.11596025556429e-06,
+      "loss": 0.4407,
+      "step": 15277
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 20.299523644748298,
+      "learning_rate": 2.114922966119287e-06,
+      "loss": 0.4801,
+      "step": 15278
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.816497956679257,
+      "learning_rate": 2.113885900921261e-06,
+      "loss": 0.4531,
+      "step": 15279
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.627782371408447,
+      "learning_rate": 2.1128490599997078e-06,
+      "loss": 0.4729,
+      "step": 15280
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 23.13071447853857,
+      "learning_rate": 2.1118124433841114e-06,
+      "loss": 0.4311,
+      "step": 15281
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.15286762311381,
+      "learning_rate": 2.1107760511039553e-06,
+      "loss": 0.401,
+      "step": 15282
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.096559783668733,
+      "learning_rate": 2.109739883188715e-06,
+      "loss": 0.4717,
+      "step": 15283
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.369880701111178,
+      "learning_rate": 2.1087039396678544e-06,
+      "loss": 0.4749,
+      "step": 15284
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.97325918166824,
+      "learning_rate": 2.107668220570841e-06,
+      "loss": 0.4723,
+      "step": 15285
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.392319367006277,
+      "learning_rate": 2.1066327259271223e-06,
+      "loss": 0.4417,
+      "step": 15286
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.358544998449325,
+      "learning_rate": 2.1055974557661553e-06,
+      "loss": 0.4639,
+      "step": 15287
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.538118530338991,
+      "learning_rate": 2.1045624101173754e-06,
+      "loss": 0.48,
+      "step": 15288
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.54379026748019,
+      "learning_rate": 2.1035275890102214e-06,
+      "loss": 0.4564,
+      "step": 15289
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.67548195167444,
+      "learning_rate": 2.1024929924741265e-06,
+      "loss": 0.4887,
+      "step": 15290
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 24.105772885345147,
+      "learning_rate": 2.1014586205385113e-06,
+      "loss": 0.451,
+      "step": 15291
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.880563801004168,
+      "learning_rate": 2.1004244732327896e-06,
+      "loss": 0.4955,
+      "step": 15292
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.33069642415794,
+      "learning_rate": 2.0993905505863755e-06,
+      "loss": 0.4027,
+      "step": 15293
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 20.187647428690234,
+      "learning_rate": 2.098356852628671e-06,
+      "loss": 0.4164,
+      "step": 15294
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.326887277138487,
+      "learning_rate": 2.0973233793890812e-06,
+      "loss": 0.4642,
+      "step": 15295
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 29.452539743735503,
+      "learning_rate": 2.0962901308969864e-06,
+      "loss": 0.4929,
+      "step": 15296
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.54224098553883,
+      "learning_rate": 2.0952571071817763e-06,
+      "loss": 0.491,
+      "step": 15297
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.126878092015584,
+      "learning_rate": 2.094224308272831e-06,
+      "loss": 0.4714,
+      "step": 15298
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.508026746074151,
+      "learning_rate": 2.0931917341995233e-06,
+      "loss": 0.4208,
+      "step": 15299
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.945344411732755,
+      "learning_rate": 2.092159384991217e-06,
+      "loss": 0.4564,
+      "step": 15300
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 29.7161328941377,
+      "learning_rate": 2.09112726067727e-06,
+      "loss": 0.4838,
+      "step": 15301
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.722281553197112,
+      "learning_rate": 2.0900953612870367e-06,
+      "loss": 0.4828,
+      "step": 15302
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.280879863388925,
+      "learning_rate": 2.089063686849867e-06,
+      "loss": 0.4541,
+      "step": 15303
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 24.265620390676048,
+      "learning_rate": 2.0880322373950957e-06,
+      "loss": 0.4254,
+      "step": 15304
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.847108449072214,
+      "learning_rate": 2.0870010129520624e-06,
+      "loss": 0.3639,
+      "step": 15305
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.175332719037446,
+      "learning_rate": 2.085970013550088e-06,
+      "loss": 0.4502,
+      "step": 15306
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.620323661778215,
+      "learning_rate": 2.0849392392184963e-06,
+      "loss": 0.4247,
+      "step": 15307
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.89964126482681,
+      "learning_rate": 2.0839086899866055e-06,
+      "loss": 0.3961,
+      "step": 15308
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 26.751031064724515,
+      "learning_rate": 2.0828783658837194e-06,
+      "loss": 0.503,
+      "step": 15309
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.3922401173626,
+      "learning_rate": 2.0818482669391428e-06,
+      "loss": 0.4951,
+      "step": 15310
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 29.168647014905293,
+      "learning_rate": 2.080818393182167e-06,
+      "loss": 0.4679,
+      "step": 15311
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.284134225318383,
+      "learning_rate": 2.079788744642085e-06,
+      "loss": 0.3895,
+      "step": 15312
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.468573998432976,
+      "learning_rate": 2.078759321348177e-06,
+      "loss": 0.4335,
+      "step": 15313
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.35319610402778,
+      "learning_rate": 2.0777301233297187e-06,
+      "loss": 0.4567,
+      "step": 15314
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 23.803433324202466,
+      "learning_rate": 2.076701150615985e-06,
+      "loss": 0.4792,
+      "step": 15315
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 25.24004927339067,
+      "learning_rate": 2.075672403236235e-06,
+      "loss": 0.4346,
+      "step": 15316
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.991296868892224,
+      "learning_rate": 2.0746438812197244e-06,
+      "loss": 0.456,
+      "step": 15317
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 27.957388735536725,
+      "learning_rate": 2.0736155845957053e-06,
+      "loss": 0.4773,
+      "step": 15318
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.605519259565634,
+      "learning_rate": 2.0725875133934216e-06,
+      "loss": 0.4481,
+      "step": 15319
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.73151073978322,
+      "learning_rate": 2.0715596676421167e-06,
+      "loss": 0.4388,
+      "step": 15320
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.991339745138355,
+      "learning_rate": 2.070532047371012e-06,
+      "loss": 0.4013,
+      "step": 15321
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 23.16828528988592,
+      "learning_rate": 2.0695046526093375e-06,
+      "loss": 0.4142,
+      "step": 15322
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 20.058899684159,
+      "learning_rate": 2.0684774833863117e-06,
+      "loss": 0.4261,
+      "step": 15323
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 21.119604420841213,
+      "learning_rate": 2.067450539731148e-06,
+      "loss": 0.465,
+      "step": 15324
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 24.24926345421895,
+      "learning_rate": 2.0664238216730513e-06,
+      "loss": 0.4469,
+      "step": 15325
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.12030764577204,
+      "learning_rate": 2.0653973292412177e-06,
+      "loss": 0.4801,
+      "step": 15326
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.922300739106715,
+      "learning_rate": 2.0643710624648427e-06,
+      "loss": 0.3977,
+      "step": 15327
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.896278905716102,
+      "learning_rate": 2.063345021373114e-06,
+      "loss": 0.4221,
+      "step": 15328
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.350592302390524,
+      "learning_rate": 2.0623192059952114e-06,
+      "loss": 0.4511,
+      "step": 15329
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.65658045513267,
+      "learning_rate": 2.061293616360304e-06,
+      "loss": 0.4579,
+      "step": 15330
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 32.541728066465026,
+      "learning_rate": 2.0602682524975616e-06,
+      "loss": 0.4114,
+      "step": 15331
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 24.310236105285252,
+      "learning_rate": 2.0592431144361458e-06,
+      "loss": 0.4857,
+      "step": 15332
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 25.37305633906512,
+      "learning_rate": 2.0582182022052134e-06,
+      "loss": 0.5092,
+      "step": 15333
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.066039909978777,
+      "learning_rate": 2.0571935158339084e-06,
+      "loss": 0.4638,
+      "step": 15334
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 24.331466619873176,
+      "learning_rate": 2.0561690553513725e-06,
+      "loss": 0.4662,
+      "step": 15335
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 28.017903718223245,
+      "learning_rate": 2.0551448207867407e-06,
+      "loss": 0.5036,
+      "step": 15336
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.294786548508146,
+      "learning_rate": 2.0541208121691458e-06,
+      "loss": 0.4489,
+      "step": 15337
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.567544250155628,
+      "learning_rate": 2.0530970295277032e-06,
+      "loss": 0.4863,
+      "step": 15338
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.931100286867284,
+      "learning_rate": 2.0520734728915358e-06,
+      "loss": 0.4708,
+      "step": 15339
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 37.74441704810459,
+      "learning_rate": 2.0510501422897466e-06,
+      "loss": 0.4936,
+      "step": 15340
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.6797677173722,
+      "learning_rate": 2.050027037751444e-06,
+      "loss": 0.4403,
+      "step": 15341
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.973351363146076,
+      "learning_rate": 2.0490041593057185e-06,
+      "loss": 0.4396,
+      "step": 15342
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.134208738840865,
+      "learning_rate": 2.0479815069816643e-06,
+      "loss": 0.4349,
+      "step": 15343
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 25.92914842273367,
+      "learning_rate": 2.0469590808083674e-06,
+      "loss": 0.4598,
+      "step": 15344
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.291027910721137,
+      "learning_rate": 2.0459368808148983e-06,
+      "loss": 0.4279,
+      "step": 15345
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 19.97231052934327,
+      "learning_rate": 2.0449149070303344e-06,
+      "loss": 0.5042,
+      "step": 15346
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.486329738739798,
+      "learning_rate": 2.043893159483734e-06,
+      "loss": 0.4563,
+      "step": 15347
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 32.7448715576632,
+      "learning_rate": 2.042871638204158e-06,
+      "loss": 0.4979,
+      "step": 15348
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.77696737173364,
+      "learning_rate": 2.0418503432206604e-06,
+      "loss": 0.5243,
+      "step": 15349
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.805140429332745,
+      "learning_rate": 2.040829274562284e-06,
+      "loss": 0.4392,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.66522374520072,
+      "learning_rate": 2.0398084322580634e-06,
+      "loss": 0.4811,
+      "step": 15351
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 24.87259473505785,
+      "learning_rate": 2.0387878163370354e-06,
+      "loss": 0.4471,
+      "step": 15352
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.038117560517303,
+      "learning_rate": 2.0377674268282275e-06,
+      "loss": 0.4631,
+      "step": 15353
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.373652843686266,
+      "learning_rate": 2.0367472637606554e-06,
+      "loss": 0.4535,
+      "step": 15354
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.776490327793674,
+      "learning_rate": 2.0357273271633304e-06,
+      "loss": 0.4974,
+      "step": 15355
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.12868131708863,
+      "learning_rate": 2.0347076170652624e-06,
+      "loss": 0.4719,
+      "step": 15356
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.62670633619575,
+      "learning_rate": 2.0336881334954484e-06,
+      "loss": 0.4814,
+      "step": 15357
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.178459496376743,
+      "learning_rate": 2.0326688764828873e-06,
+      "loss": 0.4184,
+      "step": 15358
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 16.562927807375075,
+      "learning_rate": 2.0316498460565627e-06,
+      "loss": 0.4516,
+      "step": 15359
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.897289915633605,
+      "learning_rate": 2.030631042245452e-06,
+      "loss": 0.4528,
+      "step": 15360
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.380814730794636,
+      "learning_rate": 2.0296124650785333e-06,
+      "loss": 0.4434,
+      "step": 15361
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 37.66922588717112,
+      "learning_rate": 2.0285941145847754e-06,
+      "loss": 0.5892,
+      "step": 15362
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 19.163152481233393,
+      "learning_rate": 2.0275759907931356e-06,
+      "loss": 0.4204,
+      "step": 15363
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 16.885929421890374,
+      "learning_rate": 2.0265580937325735e-06,
+      "loss": 0.4409,
+      "step": 15364
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.469361192656066,
+      "learning_rate": 2.025540423432032e-06,
+      "loss": 0.4698,
+      "step": 15365
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.994192649943056,
+      "learning_rate": 2.024522979920458e-06,
+      "loss": 0.4465,
+      "step": 15366
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.51679327306867,
+      "learning_rate": 2.023505763226783e-06,
+      "loss": 0.4988,
+      "step": 15367
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.77590399343352,
+      "learning_rate": 2.0224887733799383e-06,
+      "loss": 0.4089,
+      "step": 15368
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 32.02313432459231,
+      "learning_rate": 2.021472010408848e-06,
+      "loss": 0.4216,
+      "step": 15369
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 30.690017619258448,
+      "learning_rate": 2.020455474342424e-06,
+      "loss": 0.5699,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.010036847206578,
+      "learning_rate": 2.019439165209581e-06,
+      "loss": 0.4813,
+      "step": 15371
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.770772061073693,
+      "learning_rate": 2.018423083039218e-06,
+      "loss": 0.3951,
+      "step": 15372
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 25.315667106740317,
+      "learning_rate": 2.0174072278602352e-06,
+      "loss": 0.4572,
+      "step": 15373
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 16.238974084083967,
+      "learning_rate": 2.0163915997015195e-06,
+      "loss": 0.4257,
+      "step": 15374
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.102981106203305,
+      "learning_rate": 2.0153761985919575e-06,
+      "loss": 0.4603,
+      "step": 15375
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 28.179075897026795,
+      "learning_rate": 2.014361024560424e-06,
+      "loss": 0.52,
+      "step": 15376
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.324719065050832,
+      "learning_rate": 2.0133460776357906e-06,
+      "loss": 0.4679,
+      "step": 15377
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.442114778843433,
+      "learning_rate": 2.012331357846926e-06,
+      "loss": 0.4396,
+      "step": 15378
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 27.113192681389247,
+      "learning_rate": 2.0113168652226843e-06,
+      "loss": 0.5031,
+      "step": 15379
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.800191970076195,
+      "learning_rate": 2.0103025997919155e-06,
+      "loss": 0.4541,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 19.773256372641818,
+      "learning_rate": 2.009288561583468e-06,
+      "loss": 0.4666,
+      "step": 15381
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 25.079896243995304,
+      "learning_rate": 2.008274750626178e-06,
+      "loss": 0.4205,
+      "step": 15382
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 24.364590584423745,
+      "learning_rate": 2.007261166948885e-06,
+      "loss": 0.4824,
+      "step": 15383
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.51123863787077,
+      "learning_rate": 2.006247810580403e-06,
+      "loss": 0.4745,
+      "step": 15384
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.29417493985364,
+      "learning_rate": 2.0052346815495573e-06,
+      "loss": 0.5009,
+      "step": 15385
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.300373253885418,
+      "learning_rate": 2.0042217798851606e-06,
+      "loss": 0.4424,
+      "step": 15386
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.944193459635352,
+      "learning_rate": 2.003209105616021e-06,
+      "loss": 0.4249,
+      "step": 15387
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 19.081763144789335,
+      "learning_rate": 2.0021966587709372e-06,
+      "loss": 0.4179,
+      "step": 15388
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.19094276075015,
+      "learning_rate": 2.0011844393786983e-06,
+      "loss": 0.4952,
+      "step": 15389
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.347019968802632,
+      "learning_rate": 2.0001724474680963e-06,
+      "loss": 0.4252,
+      "step": 15390
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.528058081918658,
+      "learning_rate": 1.999160683067911e-06,
+      "loss": 0.4597,
+      "step": 15391
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.026360461180957,
+      "learning_rate": 1.9981491462069146e-06,
+      "loss": 0.4793,
+      "step": 15392
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.33475373845935,
+      "learning_rate": 1.9971378369138773e-06,
+      "loss": 0.4627,
+      "step": 15393
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 24.78181475636807,
+      "learning_rate": 1.9961267552175558e-06,
+      "loss": 0.4858,
+      "step": 15394
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.11313778125388,
+      "learning_rate": 1.995115901146707e-06,
+      "loss": 0.4304,
+      "step": 15395
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.311099526541714,
+      "learning_rate": 1.9941052747300826e-06,
+      "loss": 0.4489,
+      "step": 15396
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 19.42199272970177,
+      "learning_rate": 1.9930948759964175e-06,
+      "loss": 0.4383,
+      "step": 15397
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 28.49084642087205,
+      "learning_rate": 1.992084704974453e-06,
+      "loss": 0.4936,
+      "step": 15398
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.074301459503367,
+      "learning_rate": 1.991074761692913e-06,
+      "loss": 0.4381,
+      "step": 15399
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.3196861184539,
+      "learning_rate": 1.9900650461805237e-06,
+      "loss": 0.492,
+      "step": 15400
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.923558758099947,
+      "learning_rate": 1.9890555584659965e-06,
+      "loss": 0.4317,
+      "step": 15401
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 29.10296726190204,
+      "learning_rate": 1.9880462985780423e-06,
+      "loss": 0.4126,
+      "step": 15402
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.012345671436195,
+      "learning_rate": 1.9870372665453673e-06,
+      "loss": 0.5412,
+      "step": 15403
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.319882534691205,
+      "learning_rate": 1.986028462396666e-06,
+      "loss": 0.426,
+      "step": 15404
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 29.70867586217058,
+      "learning_rate": 1.985019886160624e-06,
+      "loss": 0.4587,
+      "step": 15405
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.001231797939358,
+      "learning_rate": 1.9840115378659275e-06,
+      "loss": 0.4341,
+      "step": 15406
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.576686755284673,
+      "learning_rate": 1.983003417541254e-06,
+      "loss": 0.4102,
+      "step": 15407
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.271254128035764,
+      "learning_rate": 1.9819955252152755e-06,
+      "loss": 0.4084,
+      "step": 15408
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.68404562964947,
+      "learning_rate": 1.9809878609166546e-06,
+      "loss": 0.4174,
+      "step": 15409
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 21.624490099336295,
+      "learning_rate": 1.979980424674045e-06,
+      "loss": 0.367,
+      "step": 15410
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.412502316791894,
+      "learning_rate": 1.9789732165161e-06,
+      "loss": 0.4526,
+      "step": 15411
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.2091591234103,
+      "learning_rate": 1.977966236471468e-06,
+      "loss": 0.4503,
+      "step": 15412
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.325366069873303,
+      "learning_rate": 1.9769594845687833e-06,
+      "loss": 0.3721,
+      "step": 15413
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 23.835925868983953,
+      "learning_rate": 1.9759529608366744e-06,
+      "loss": 0.4913,
+      "step": 15414
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 25.879553597181058,
+      "learning_rate": 1.9749466653037707e-06,
+      "loss": 0.4376,
+      "step": 15415
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.57163658773314,
+      "learning_rate": 1.9739405979986904e-06,
+      "loss": 0.4387,
+      "step": 15416
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.763706329939858,
+      "learning_rate": 1.9729347589500426e-06,
+      "loss": 0.468,
+      "step": 15417
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.819709399160494,
+      "learning_rate": 1.9719291481864366e-06,
+      "loss": 0.4323,
+      "step": 15418
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 19.61366202921762,
+      "learning_rate": 1.970923765736468e-06,
+      "loss": 0.4705,
+      "step": 15419
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.109965891069336,
+      "learning_rate": 1.9699186116287295e-06,
+      "loss": 0.4518,
+      "step": 15420
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.525286626730985,
+      "learning_rate": 1.9689136858918112e-06,
+      "loss": 0.4076,
+      "step": 15421
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 21.754325400294373,
+      "learning_rate": 1.967908988554289e-06,
+      "loss": 0.4334,
+      "step": 15422
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 35.943351270917994,
+      "learning_rate": 1.9669045196447345e-06,
+      "loss": 0.5155,
+      "step": 15423
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 21.22633731443197,
+      "learning_rate": 1.9659002791917157e-06,
+      "loss": 0.4026,
+      "step": 15424
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 24.75152008489357,
+      "learning_rate": 1.964896267223797e-06,
+      "loss": 0.5159,
+      "step": 15425
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.93363529949157,
+      "learning_rate": 1.963892483769524e-06,
+      "loss": 0.525,
+      "step": 15426
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 26.46723593622914,
+      "learning_rate": 1.9628889288574514e-06,
+      "loss": 0.5061,
+      "step": 15427
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 23.668548761977448,
+      "learning_rate": 1.9618856025161127e-06,
+      "loss": 0.4853,
+      "step": 15428
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.384002698692992,
+      "learning_rate": 1.9608825047740486e-06,
+      "loss": 0.4093,
+      "step": 15429
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.924566578618645,
+      "learning_rate": 1.9598796356597806e-06,
+      "loss": 0.4646,
+      "step": 15430
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 26.824214996964244,
+      "learning_rate": 1.958876995201833e-06,
+      "loss": 0.4867,
+      "step": 15431
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.065254880886613,
+      "learning_rate": 1.9578745834287204e-06,
+      "loss": 0.4278,
+      "step": 15432
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.789933509528122,
+      "learning_rate": 1.9568724003689486e-06,
+      "loss": 0.4337,
+      "step": 15433
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 30.280174263820577,
+      "learning_rate": 1.9558704460510235e-06,
+      "loss": 0.4887,
+      "step": 15434
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 30.785945669720157,
+      "learning_rate": 1.9548687205034346e-06,
+      "loss": 0.4184,
+      "step": 15435
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.040069254900477,
+      "learning_rate": 1.953867223754673e-06,
+      "loss": 0.4437,
+      "step": 15436
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.663784536914285,
+      "learning_rate": 1.952865955833223e-06,
+      "loss": 0.4312,
+      "step": 15437
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.958025664367458,
+      "learning_rate": 1.9518649167675585e-06,
+      "loss": 0.4752,
+      "step": 15438
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.61297948338482,
+      "learning_rate": 1.9508641065861445e-06,
+      "loss": 0.4297,
+      "step": 15439
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.158109998640352,
+      "learning_rate": 1.949863525317447e-06,
+      "loss": 0.4255,
+      "step": 15440
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 19.182126192799373,
+      "learning_rate": 1.9488631729899243e-06,
+      "loss": 0.4966,
+      "step": 15441
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 24.806468850398485,
+      "learning_rate": 1.9478630496320227e-06,
+      "loss": 0.4453,
+      "step": 15442
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.316897872565225,
+      "learning_rate": 1.946863155272183e-06,
+      "loss": 0.4356,
+      "step": 15443
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.5711907342403,
+      "learning_rate": 1.9458634899388453e-06,
+      "loss": 0.4,
+      "step": 15444
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.81115641857474,
+      "learning_rate": 1.944864053660437e-06,
+      "loss": 0.4453,
+      "step": 15445
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.257289195039252,
+      "learning_rate": 1.9438648464653865e-06,
+      "loss": 0.4563,
+      "step": 15446
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.242501838211776,
+      "learning_rate": 1.9428658683821066e-06,
+      "loss": 0.4796,
+      "step": 15447
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 21.693119120209765,
+      "learning_rate": 1.9418671194390047e-06,
+      "loss": 0.4608,
+      "step": 15448
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.888288650158334,
+      "learning_rate": 1.9408685996644893e-06,
+      "loss": 0.4281,
+      "step": 15449
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 23.671266004074816,
+      "learning_rate": 1.9398703090869585e-06,
+      "loss": 0.4544,
+      "step": 15450
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 26.165442782295703,
+      "learning_rate": 1.938872247734799e-06,
+      "loss": 0.4851,
+      "step": 15451
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.86494189204278,
+      "learning_rate": 1.9378744156363993e-06,
+      "loss": 0.4449,
+      "step": 15452
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 26.405277811584256,
+      "learning_rate": 1.936876812820133e-06,
+      "loss": 0.4911,
+      "step": 15453
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 23.213567757429583,
+      "learning_rate": 1.9358794393143755e-06,
+      "loss": 0.4307,
+      "step": 15454
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.93357932709829,
+      "learning_rate": 1.934882295147487e-06,
+      "loss": 0.4249,
+      "step": 15455
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.539829820614916,
+      "learning_rate": 1.9338853803478274e-06,
+      "loss": 0.4973,
+      "step": 15456
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.161147379474386,
+      "learning_rate": 1.932888694943752e-06,
+      "loss": 0.445,
+      "step": 15457
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.23288635140148,
+      "learning_rate": 1.931892238963601e-06,
+      "loss": 0.3949,
+      "step": 15458
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.009897063129984,
+      "learning_rate": 1.9308960124357167e-06,
+      "loss": 0.3791,
+      "step": 15459
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 35.84390328140876,
+      "learning_rate": 1.9299000153884274e-06,
+      "loss": 0.4538,
+      "step": 15460
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.902951677899033,
+      "learning_rate": 1.928904247850061e-06,
+      "loss": 0.4124,
+      "step": 15461
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 19.13498505469158,
+      "learning_rate": 1.927908709848938e-06,
+      "loss": 0.4538,
+      "step": 15462
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 19.210150469481583,
+      "learning_rate": 1.9269134014133706e-06,
+      "loss": 0.398,
+      "step": 15463
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.164378533589606,
+      "learning_rate": 1.9259183225716594e-06,
+      "loss": 0.5532,
+      "step": 15464
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.420596460686678,
+      "learning_rate": 1.924923473352108e-06,
+      "loss": 0.4582,
+      "step": 15465
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 25.620834669154583,
+      "learning_rate": 1.9239288537830136e-06,
+      "loss": 0.4289,
+      "step": 15466
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 32.40137726881236,
+      "learning_rate": 1.922934463892657e-06,
+      "loss": 0.5195,
+      "step": 15467
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 19.4995559923045,
+      "learning_rate": 1.9219403037093164e-06,
+      "loss": 0.5003,
+      "step": 15468
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 24.151146947461086,
+      "learning_rate": 1.9209463732612687e-06,
+      "loss": 0.4386,
+      "step": 15469
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 23.795229519280987,
+      "learning_rate": 1.919952672576779e-06,
+      "loss": 0.4837,
+      "step": 15470
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 19.702119020347897,
+      "learning_rate": 1.9189592016841154e-06,
+      "loss": 0.4508,
+      "step": 15471
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 23.393520058845333,
+      "learning_rate": 1.9179659606115187e-06,
+      "loss": 0.4745,
+      "step": 15472
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 27.81683536397295,
+      "learning_rate": 1.9169729493872423e-06,
+      "loss": 0.4712,
+      "step": 15473
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 25.68612940872227,
+      "learning_rate": 1.9159801680395264e-06,
+      "loss": 0.4178,
+      "step": 15474
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 36.52499344882476,
+      "learning_rate": 1.9149876165966088e-06,
+      "loss": 0.5706,
+      "step": 15475
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.428117968256183,
+      "learning_rate": 1.9139952950867135e-06,
+      "loss": 0.3901,
+      "step": 15476
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 28.240323277700302,
+      "learning_rate": 1.9130032035380595e-06,
+      "loss": 0.5087,
+      "step": 15477
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 18.345983534783485,
+      "learning_rate": 1.9120113419788635e-06,
+      "loss": 0.4124,
+      "step": 15478
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 28.73539116635506,
+      "learning_rate": 1.911019710437335e-06,
+      "loss": 0.4251,
+      "step": 15479
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.389330319163786,
+      "learning_rate": 1.9100283089416725e-06,
+      "loss": 0.4845,
+      "step": 15480
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.833135539329074,
+      "learning_rate": 1.9090371375200755e-06,
+      "loss": 0.4288,
+      "step": 15481
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 28.73584550535108,
+      "learning_rate": 1.9080461962007257e-06,
+      "loss": 0.4462,
+      "step": 15482
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 18.444229159599242,
+      "learning_rate": 1.9070554850118095e-06,
+      "loss": 0.3995,
+      "step": 15483
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 27.146113728480138,
+      "learning_rate": 1.906065003981503e-06,
+      "loss": 0.4411,
+      "step": 15484
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.86290123241126,
+      "learning_rate": 1.9050747531379698e-06,
+      "loss": 0.4998,
+      "step": 15485
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 27.4611787131556,
+      "learning_rate": 1.9040847325093791e-06,
+      "loss": 0.4423,
+      "step": 15486
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.461417487866154,
+      "learning_rate": 1.9030949421238787e-06,
+      "loss": 0.4245,
+      "step": 15487
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.842269138694503,
+      "learning_rate": 1.9021053820096246e-06,
+      "loss": 0.5377,
+      "step": 15488
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.282042748691268,
+      "learning_rate": 1.9011160521947548e-06,
+      "loss": 0.5069,
+      "step": 15489
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.012690382599313,
+      "learning_rate": 1.900126952707405e-06,
+      "loss": 0.4664,
+      "step": 15490
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.660097409166543,
+      "learning_rate": 1.89913808357571e-06,
+      "loss": 0.4061,
+      "step": 15491
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.227096898792517,
+      "learning_rate": 1.8981494448277893e-06,
+      "loss": 0.4035,
+      "step": 15492
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.70250947145214,
+      "learning_rate": 1.897161036491756e-06,
+      "loss": 0.4435,
+      "step": 15493
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.803582263133466,
+      "learning_rate": 1.8961728585957229e-06,
+      "loss": 0.4321,
+      "step": 15494
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 24.03617323495913,
+      "learning_rate": 1.8951849111677922e-06,
+      "loss": 0.5375,
+      "step": 15495
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 22.04168502338729,
+      "learning_rate": 1.894197194236065e-06,
+      "loss": 0.4519,
+      "step": 15496
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.570695008463197,
+      "learning_rate": 1.8932097078286283e-06,
+      "loss": 0.4613,
+      "step": 15497
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 23.397042540170354,
+      "learning_rate": 1.8922224519735611e-06,
+      "loss": 0.4647,
+      "step": 15498
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.23721492971823,
+      "learning_rate": 1.8912354266989453e-06,
+      "loss": 0.4537,
+      "step": 15499
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.640733786113206,
+      "learning_rate": 1.890248632032854e-06,
+      "loss": 0.4088,
+      "step": 15500
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.344779684642898,
+      "learning_rate": 1.8892620680033457e-06,
+      "loss": 0.4612,
+      "step": 15501
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 22.393334753011086,
+      "learning_rate": 1.8882757346384785e-06,
+      "loss": 0.4664,
+      "step": 15502
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 25.639841912752278,
+      "learning_rate": 1.8872896319663038e-06,
+      "loss": 0.4648,
+      "step": 15503
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.715601949015337,
+      "learning_rate": 1.886303760014868e-06,
+      "loss": 0.3739,
+      "step": 15504
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.487007607727623,
+      "learning_rate": 1.885318118812205e-06,
+      "loss": 0.4284,
+      "step": 15505
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 18.665020401948865,
+      "learning_rate": 1.8843327083863495e-06,
+      "loss": 0.3773,
+      "step": 15506
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.011841090803404,
+      "learning_rate": 1.8833475287653225e-06,
+      "loss": 0.4468,
+      "step": 15507
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 25.123861286468784,
+      "learning_rate": 1.8823625799771428e-06,
+      "loss": 0.4441,
+      "step": 15508
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.689392883258854,
+      "learning_rate": 1.8813778620498268e-06,
+      "loss": 0.4761,
+      "step": 15509
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 24.485711425186466,
+      "learning_rate": 1.8803933750113712e-06,
+      "loss": 0.3874,
+      "step": 15510
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 26.08969638941055,
+      "learning_rate": 1.8794091188897812e-06,
+      "loss": 0.4997,
+      "step": 15511
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 22.463916629092914,
+      "learning_rate": 1.8784250937130433e-06,
+      "loss": 0.4722,
+      "step": 15512
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 29.346236041874445,
+      "learning_rate": 1.8774412995091462e-06,
+      "loss": 0.5104,
+      "step": 15513
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.387853635175233,
+      "learning_rate": 1.8764577363060654e-06,
+      "loss": 0.525,
+      "step": 15514
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 26.565157835886573,
+      "learning_rate": 1.8754744041317763e-06,
+      "loss": 0.4754,
+      "step": 15515
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 24.772762045167756,
+      "learning_rate": 1.8744913030142409e-06,
+      "loss": 0.4753,
+      "step": 15516
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.084553604251518,
+      "learning_rate": 1.8735084329814213e-06,
+      "loss": 0.435,
+      "step": 15517
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.82897151171001,
+      "learning_rate": 1.872525794061265e-06,
+      "loss": 0.4226,
+      "step": 15518
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.005539103803844,
+      "learning_rate": 1.8715433862817224e-06,
+      "loss": 0.3689,
+      "step": 15519
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.70696593736945,
+      "learning_rate": 1.8705612096707314e-06,
+      "loss": 0.3947,
+      "step": 15520
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.32795276819128,
+      "learning_rate": 1.8695792642562226e-06,
+      "loss": 0.4634,
+      "step": 15521
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.592953810116423,
+      "learning_rate": 1.8685975500661257e-06,
+      "loss": 0.4318,
+      "step": 15522
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 32.822369493560196,
+      "learning_rate": 1.867616067128355e-06,
+      "loss": 0.4851,
+      "step": 15523
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.689784104714516,
+      "learning_rate": 1.8666348154708269e-06,
+      "loss": 0.429,
+      "step": 15524
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 21.81618102391986,
+      "learning_rate": 1.8656537951214482e-06,
+      "loss": 0.4085,
+      "step": 15525
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 23.91793502825176,
+      "learning_rate": 1.8646730061081174e-06,
+      "loss": 0.4567,
+      "step": 15526
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 22.960649132256282,
+      "learning_rate": 1.8636924484587248e-06,
+      "loss": 0.4243,
+      "step": 15527
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.55950211655765,
+      "learning_rate": 1.86271212220116e-06,
+      "loss": 0.4286,
+      "step": 15528
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.89256074221222,
+      "learning_rate": 1.8617320273633044e-06,
+      "loss": 0.4338,
+      "step": 15529
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.84907087098428,
+      "learning_rate": 1.8607521639730285e-06,
+      "loss": 0.3855,
+      "step": 15530
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 18.650400882214374,
+      "learning_rate": 1.8597725320581982e-06,
+      "loss": 0.4583,
+      "step": 15531
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 26.300842484516632,
+      "learning_rate": 1.8587931316466746e-06,
+      "loss": 0.4421,
+      "step": 15532
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 15.463793913339718,
+      "learning_rate": 1.8578139627663127e-06,
+      "loss": 0.4414,
+      "step": 15533
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 27.251963384737504,
+      "learning_rate": 1.8568350254449596e-06,
+      "loss": 0.5296,
+      "step": 15534
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 27.35784798210833,
+      "learning_rate": 1.8558563197104562e-06,
+      "loss": 0.4436,
+      "step": 15535
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.207881245193466,
+      "learning_rate": 1.8548778455906325e-06,
+      "loss": 0.4711,
+      "step": 15536
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 46.79857274066157,
+      "learning_rate": 1.8538996031133172e-06,
+      "loss": 0.4162,
+      "step": 15537
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.27897282744381,
+      "learning_rate": 1.8529215923063348e-06,
+      "loss": 0.4149,
+      "step": 15538
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 23.941187353591445,
+      "learning_rate": 1.8519438131974932e-06,
+      "loss": 0.52,
+      "step": 15539
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 21.009220502405334,
+      "learning_rate": 1.8509662658146067e-06,
+      "loss": 0.5146,
+      "step": 15540
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 15.457626560750542,
+      "learning_rate": 1.8499889501854685e-06,
+      "loss": 0.4807,
+      "step": 15541
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.223221916088125,
+      "learning_rate": 1.8490118663378809e-06,
+      "loss": 0.409,
+      "step": 15542
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.22119670766908,
+      "learning_rate": 1.848035014299625e-06,
+      "loss": 0.515,
+      "step": 15543
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 23.71409698187734,
+      "learning_rate": 1.8470583940984832e-06,
+      "loss": 0.434,
+      "step": 15544
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.1865642707045,
+      "learning_rate": 1.8460820057622353e-06,
+      "loss": 0.4485,
+      "step": 15545
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 28.92453661088341,
+      "learning_rate": 1.8451058493186424e-06,
+      "loss": 0.4738,
+      "step": 15546
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 18.645635838679198,
+      "learning_rate": 1.8441299247954713e-06,
+      "loss": 0.4146,
+      "step": 15547
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.40138518325208,
+      "learning_rate": 1.843154232220471e-06,
+      "loss": 0.4028,
+      "step": 15548
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.564279626909308,
+      "learning_rate": 1.8421787716213934e-06,
+      "loss": 0.4484,
+      "step": 15549
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.513939551734618,
+      "learning_rate": 1.8412035430259811e-06,
+      "loss": 0.4353,
+      "step": 15550
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.47360084561293,
+      "learning_rate": 1.8402285464619673e-06,
+      "loss": 0.4062,
+      "step": 15551
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 18.869193567127297,
+      "learning_rate": 1.839253781957079e-06,
+      "loss": 0.4228,
+      "step": 15552
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.74760828036703,
+      "learning_rate": 1.8382792495390378e-06,
+      "loss": 0.4246,
+      "step": 15553
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 26.291339030330214,
+      "learning_rate": 1.8373049492355633e-06,
+      "loss": 0.4226,
+      "step": 15554
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.175077504187207,
+      "learning_rate": 1.8363308810743619e-06,
+      "loss": 0.4958,
+      "step": 15555
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 21.893646382972268,
+      "learning_rate": 1.8353570450831303e-06,
+      "loss": 0.5396,
+      "step": 15556
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.594284514493253,
+      "learning_rate": 1.83438344128957e-06,
+      "loss": 0.4408,
+      "step": 15557
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.786191548393347,
+      "learning_rate": 1.8334100697213664e-06,
+      "loss": 0.4289,
+      "step": 15558
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.093224752731874,
+      "learning_rate": 1.8324369304062062e-06,
+      "loss": 0.4942,
+      "step": 15559
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 32.5159815503378,
+      "learning_rate": 1.8314640233717617e-06,
+      "loss": 0.4696,
+      "step": 15560
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.333487850142767,
+      "learning_rate": 1.8304913486456998e-06,
+      "loss": 0.4198,
+      "step": 15561
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 26.79393163425483,
+      "learning_rate": 1.8295189062556851e-06,
+      "loss": 0.5029,
+      "step": 15562
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 30.47662430370472,
+      "learning_rate": 1.8285466962293752e-06,
+      "loss": 0.456,
+      "step": 15563
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 18.83600360205312,
+      "learning_rate": 1.827574718594417e-06,
+      "loss": 0.4348,
+      "step": 15564
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.463346120194814,
+      "learning_rate": 1.8266029733784506e-06,
+      "loss": 0.4414,
+      "step": 15565
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 21.173209467846437,
+      "learning_rate": 1.8256314606091153e-06,
+      "loss": 0.4557,
+      "step": 15566
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.56113126268079,
+      "learning_rate": 1.8246601803140407e-06,
+      "loss": 0.4599,
+      "step": 15567
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.826642576888922,
+      "learning_rate": 1.823689132520846e-06,
+      "loss": 0.4432,
+      "step": 15568
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.693058125489113,
+      "learning_rate": 1.8227183172571516e-06,
+      "loss": 0.4092,
+      "step": 15569
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 15.781914976953203,
+      "learning_rate": 1.8217477345505619e-06,
+      "loss": 0.4502,
+      "step": 15570
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 27.233240033879568,
+      "learning_rate": 1.8207773844286835e-06,
+      "loss": 0.4368,
+      "step": 15571
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.52694403599269,
+      "learning_rate": 1.8198072669191136e-06,
+      "loss": 0.4958,
+      "step": 15572
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 22.219751880017537,
+      "learning_rate": 1.8188373820494376e-06,
+      "loss": 0.4521,
+      "step": 15573
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.045430620088553,
+      "learning_rate": 1.8178677298472425e-06,
+      "loss": 0.4735,
+      "step": 15574
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 22.332529526843867,
+      "learning_rate": 1.8168983103401006e-06,
+      "loss": 0.4764,
+      "step": 15575
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.68263924013169,
+      "learning_rate": 1.8159291235555864e-06,
+      "loss": 0.4603,
+      "step": 15576
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 26.4168065768868,
+      "learning_rate": 1.8149601695212571e-06,
+      "loss": 0.4721,
+      "step": 15577
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.256574396174575,
+      "learning_rate": 1.8139914482646737e-06,
+      "loss": 0.4618,
+      "step": 15578
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 35.16246848825398,
+      "learning_rate": 1.8130229598133865e-06,
+      "loss": 0.4945,
+      "step": 15579
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.050274288304607,
+      "learning_rate": 1.8120547041949366e-06,
+      "loss": 0.4849,
+      "step": 15580
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.73123641352344,
+      "learning_rate": 1.81108668143686e-06,
+      "loss": 0.4841,
+      "step": 15581
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.97760586319767,
+      "learning_rate": 1.8101188915666867e-06,
+      "loss": 0.4433,
+      "step": 15582
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 26.52608140765442,
+      "learning_rate": 1.8091513346119415e-06,
+      "loss": 0.4898,
+      "step": 15583
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.909912315841037,
+      "learning_rate": 1.8081840106001435e-06,
+      "loss": 0.4103,
+      "step": 15584
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 21.91392893215008,
+      "learning_rate": 1.8072169195587996e-06,
+      "loss": 0.4377,
+      "step": 15585
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.67713105815181,
+      "learning_rate": 1.8062500615154121e-06,
+      "loss": 0.5208,
+      "step": 15586
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.345071707390748,
+      "learning_rate": 1.8052834364974791e-06,
+      "loss": 0.4971,
+      "step": 15587
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.223992218249172,
+      "learning_rate": 1.8043170445324943e-06,
+      "loss": 0.401,
+      "step": 15588
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.01658956509949,
+      "learning_rate": 1.8033508856479387e-06,
+      "loss": 0.4579,
+      "step": 15589
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 16.816350958951464,
+      "learning_rate": 1.8023849598712861e-06,
+      "loss": 0.434,
+      "step": 15590
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.49204753221549,
+      "learning_rate": 1.801419267230009e-06,
+      "loss": 0.4251,
+      "step": 15591
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.476043720798017,
+      "learning_rate": 1.8004538077515755e-06,
+      "loss": 0.4256,
+      "step": 15592
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.009408024875126,
+      "learning_rate": 1.7994885814634366e-06,
+      "loss": 0.451,
+      "step": 15593
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.059859371806303,
+      "learning_rate": 1.7985235883930475e-06,
+      "loss": 0.4199,
+      "step": 15594
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.99359851057603,
+      "learning_rate": 1.797558828567847e-06,
+      "loss": 0.3981,
+      "step": 15595
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.7191569922826,
+      "learning_rate": 1.7965943020152753e-06,
+      "loss": 0.4991,
+      "step": 15596
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.35126964833335,
+      "learning_rate": 1.7956300087627655e-06,
+      "loss": 0.4206,
+      "step": 15597
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.147526625752768,
+      "learning_rate": 1.794665948837736e-06,
+      "loss": 0.421,
+      "step": 15598
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.842469635323038,
+      "learning_rate": 1.7937021222676098e-06,
+      "loss": 0.431,
+      "step": 15599
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.437492891273326,
+      "learning_rate": 1.7927385290797917e-06,
+      "loss": 0.4165,
+      "step": 15600
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 25.767849399350506,
+      "learning_rate": 1.7917751693016915e-06,
+      "loss": 0.4514,
+      "step": 15601
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.44545047461242,
+      "learning_rate": 1.7908120429607024e-06,
+      "loss": 0.421,
+      "step": 15602
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.466842943217578,
+      "learning_rate": 1.7898491500842152e-06,
+      "loss": 0.4603,
+      "step": 15603
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.735812061493073,
+      "learning_rate": 1.7888864906996183e-06,
+      "loss": 0.4109,
+      "step": 15604
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.371800362353035,
+      "learning_rate": 1.7879240648342854e-06,
+      "loss": 0.4994,
+      "step": 15605
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.85608643136716,
+      "learning_rate": 1.7869618725155868e-06,
+      "loss": 0.426,
+      "step": 15606
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.468068342478652,
+      "learning_rate": 1.7859999137708872e-06,
+      "loss": 0.4487,
+      "step": 15607
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.705747158243064,
+      "learning_rate": 1.7850381886275458e-06,
+      "loss": 0.497,
+      "step": 15608
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.666298408597527,
+      "learning_rate": 1.7840766971129153e-06,
+      "loss": 0.493,
+      "step": 15609
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.22067524485722,
+      "learning_rate": 1.783115439254337e-06,
+      "loss": 0.4137,
+      "step": 15610
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.730136175140682,
+      "learning_rate": 1.7821544150791459e-06,
+      "loss": 0.4131,
+      "step": 15611
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.197540813171795,
+      "learning_rate": 1.7811936246146765e-06,
+      "loss": 0.5077,
+      "step": 15612
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 27.45459468948094,
+      "learning_rate": 1.7802330678882552e-06,
+      "loss": 0.4683,
+      "step": 15613
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.172755624557027,
+      "learning_rate": 1.7792727449271962e-06,
+      "loss": 0.4453,
+      "step": 15614
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 28.151691070862654,
+      "learning_rate": 1.7783126557588092e-06,
+      "loss": 0.5051,
+      "step": 15615
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.603273924833005,
+      "learning_rate": 1.7773528004104012e-06,
+      "loss": 0.3779,
+      "step": 15616
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.866715365670732,
+      "learning_rate": 1.7763931789092715e-06,
+      "loss": 0.4988,
+      "step": 15617
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.680668224387993,
+      "learning_rate": 1.7754337912827092e-06,
+      "loss": 0.4164,
+      "step": 15618
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 24.94071613991043,
+      "learning_rate": 1.7744746375579968e-06,
+      "loss": 0.4503,
+      "step": 15619
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.952616256435636,
+      "learning_rate": 1.7735157177624129e-06,
+      "loss": 0.4294,
+      "step": 15620
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.878098029710657,
+      "learning_rate": 1.77255703192323e-06,
+      "loss": 0.4186,
+      "step": 15621
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.756194665738846,
+      "learning_rate": 1.7715985800677148e-06,
+      "loss": 0.4426,
+      "step": 15622
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.196275182305122,
+      "learning_rate": 1.7706403622231228e-06,
+      "loss": 0.4844,
+      "step": 15623
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.582590126210418,
+      "learning_rate": 1.7696823784167017e-06,
+      "loss": 0.4085,
+      "step": 15624
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 16.921045101593744,
+      "learning_rate": 1.7687246286756999e-06,
+      "loss": 0.4785,
+      "step": 15625
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 27.30614475842026,
+      "learning_rate": 1.7677671130273567e-06,
+      "loss": 0.4916,
+      "step": 15626
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.09746587820967,
+      "learning_rate": 1.766809831498898e-06,
+      "loss": 0.4793,
+      "step": 15627
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.991629341336196,
+      "learning_rate": 1.7658527841175543e-06,
+      "loss": 0.4679,
+      "step": 15628
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.524029075937946,
+      "learning_rate": 1.7648959709105384e-06,
+      "loss": 0.4695,
+      "step": 15629
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.787327944967217,
+      "learning_rate": 1.7639393919050662e-06,
+      "loss": 0.4756,
+      "step": 15630
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.943533902893822,
+      "learning_rate": 1.762983047128337e-06,
+      "loss": 0.4014,
+      "step": 15631
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.80827617634642,
+      "learning_rate": 1.7620269366075514e-06,
+      "loss": 0.4168,
+      "step": 15632
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.99525715500652,
+      "learning_rate": 1.7610710603699033e-06,
+      "loss": 0.4677,
+      "step": 15633
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.448211092185396,
+      "learning_rate": 1.7601154184425717e-06,
+      "loss": 0.4523,
+      "step": 15634
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.916203340889655,
+      "learning_rate": 1.7591600108527407e-06,
+      "loss": 0.4635,
+      "step": 15635
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 16.464627713633504,
+      "learning_rate": 1.758204837627575e-06,
+      "loss": 0.3869,
+      "step": 15636
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.811552922359514,
+      "learning_rate": 1.757249898794242e-06,
+      "loss": 0.4253,
+      "step": 15637
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.851804911960375,
+      "learning_rate": 1.7562951943799033e-06,
+      "loss": 0.4138,
+      "step": 15638
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.47046993264175,
+      "learning_rate": 1.755340724411707e-06,
+      "loss": 0.4075,
+      "step": 15639
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.026540968396244,
+      "learning_rate": 1.754386488916795e-06,
+      "loss": 0.5165,
+      "step": 15640
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.018244045936102,
+      "learning_rate": 1.7534324879223065e-06,
+      "loss": 0.3948,
+      "step": 15641
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.24360875182509,
+      "learning_rate": 1.7524787214553774e-06,
+      "loss": 0.5273,
+      "step": 15642
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 24.330801428254386,
+      "learning_rate": 1.7515251895431295e-06,
+      "loss": 0.486,
+      "step": 15643
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 30.040085542716724,
+      "learning_rate": 1.750571892212677e-06,
+      "loss": 0.5172,
+      "step": 15644
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.782636953037628,
+      "learning_rate": 1.7496188294911332e-06,
+      "loss": 0.5043,
+      "step": 15645
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.405097401911334,
+      "learning_rate": 1.7486660014056046e-06,
+      "loss": 0.4902,
+      "step": 15646
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.60867511151067,
+      "learning_rate": 1.7477134079831893e-06,
+      "loss": 0.4531,
+      "step": 15647
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.777383447618277,
+      "learning_rate": 1.746761049250979e-06,
+      "loss": 0.4798,
+      "step": 15648
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.04281763359204,
+      "learning_rate": 1.7458089252360522e-06,
+      "loss": 0.468,
+      "step": 15649
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.84701788420225,
+      "learning_rate": 1.7448570359654925e-06,
+      "loss": 0.4395,
+      "step": 15650
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.415634407676242,
+      "learning_rate": 1.743905381466371e-06,
+      "loss": 0.4606,
+      "step": 15651
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.031542974918363,
+      "learning_rate": 1.7429539617657487e-06,
+      "loss": 0.4578,
+      "step": 15652
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.683632227752598,
+      "learning_rate": 1.7420027768906877e-06,
+      "loss": 0.5084,
+      "step": 15653
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.666012341353326,
+      "learning_rate": 1.7410518268682353e-06,
+      "loss": 0.4138,
+      "step": 15654
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 24.73429182265147,
+      "learning_rate": 1.7401011117254397e-06,
+      "loss": 0.4406,
+      "step": 15655
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 23.55104984896507,
+      "learning_rate": 1.7391506314893337e-06,
+      "loss": 0.4558,
+      "step": 15656
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.390192437768842,
+      "learning_rate": 1.7382003861869512e-06,
+      "loss": 0.4631,
+      "step": 15657
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.926578607680682,
+      "learning_rate": 1.7372503758453196e-06,
+      "loss": 0.4201,
+      "step": 15658
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 37.3928337257797,
+      "learning_rate": 1.7363006004914507e-06,
+      "loss": 0.4767,
+      "step": 15659
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.056462213480682,
+      "learning_rate": 1.7353510601523616e-06,
+      "loss": 0.4543,
+      "step": 15660
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.124255974211724,
+      "learning_rate": 1.7344017548550506e-06,
+      "loss": 0.4279,
+      "step": 15661
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.64766689728801,
+      "learning_rate": 1.733452684626522e-06,
+      "loss": 0.4757,
+      "step": 15662
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.501590290138942,
+      "learning_rate": 1.7325038494937597e-06,
+      "loss": 0.418,
+      "step": 15663
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 40.71444599038201,
+      "learning_rate": 1.7315552494837552e-06,
+      "loss": 0.5379,
+      "step": 15664
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.89527322407755,
+      "learning_rate": 1.7306068846234791e-06,
+      "loss": 0.405,
+      "step": 15665
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 33.276344008462154,
+      "learning_rate": 1.7296587549399057e-06,
+      "loss": 0.4574,
+      "step": 15666
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.912322647377387,
+      "learning_rate": 1.7287108604600023e-06,
+      "loss": 0.4568,
+      "step": 15667
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.790070236593376,
+      "learning_rate": 1.7277632012107236e-06,
+      "loss": 0.4321,
+      "step": 15668
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.828375157511363,
+      "learning_rate": 1.7268157772190174e-06,
+      "loss": 0.4351,
+      "step": 15669
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 35.59040206179848,
+      "learning_rate": 1.725868588511831e-06,
+      "loss": 0.4765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.292652946057988,
+      "learning_rate": 1.7249216351161013e-06,
+      "loss": 0.4741,
+      "step": 15671
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 32.32808279308516,
+      "learning_rate": 1.7239749170587628e-06,
+      "loss": 0.5315,
+      "step": 15672
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 26.633858747457225,
+      "learning_rate": 1.723028434366736e-06,
+      "loss": 0.4776,
+      "step": 15673
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.156949570465212,
+      "learning_rate": 1.7220821870669358e-06,
+      "loss": 0.3494,
+      "step": 15674
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 30.33912482109138,
+      "learning_rate": 1.7211361751862754e-06,
+      "loss": 0.556,
+      "step": 15675
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.094025046356762,
+      "learning_rate": 1.720190398751662e-06,
+      "loss": 0.4877,
+      "step": 15676
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 33.400337224103104,
+      "learning_rate": 1.7192448577899912e-06,
+      "loss": 0.4672,
+      "step": 15677
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 19.280156895180415,
+      "learning_rate": 1.7182995523281488e-06,
+      "loss": 0.3835,
+      "step": 15678
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.90173532132101,
+      "learning_rate": 1.7173544823930232e-06,
+      "loss": 0.4142,
+      "step": 15679
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 22.076843355454923,
+      "learning_rate": 1.7164096480114934e-06,
+      "loss": 0.3858,
+      "step": 15680
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.93684716986153,
+      "learning_rate": 1.715465049210424e-06,
+      "loss": 0.4491,
+      "step": 15681
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.915089700065323,
+      "learning_rate": 1.7145206860166863e-06,
+      "loss": 0.4833,
+      "step": 15682
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.768664691547805,
+      "learning_rate": 1.7135765584571296e-06,
+      "loss": 0.4888,
+      "step": 15683
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 23.222977411601583,
+      "learning_rate": 1.712632666558609e-06,
+      "loss": 0.4562,
+      "step": 15684
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 29.22592705869698,
+      "learning_rate": 1.7116890103479688e-06,
+      "loss": 0.4759,
+      "step": 15685
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.455293674475193,
+      "learning_rate": 1.710745589852043e-06,
+      "loss": 0.4375,
+      "step": 15686
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 26.645101678992134,
+      "learning_rate": 1.7098024050976657e-06,
+      "loss": 0.43,
+      "step": 15687
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 23.064482267084994,
+      "learning_rate": 1.7088594561116567e-06,
+      "loss": 0.4928,
+      "step": 15688
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.91025739196887,
+      "learning_rate": 1.7079167429208366e-06,
+      "loss": 0.428,
+      "step": 15689
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 13.843329713869773,
+      "learning_rate": 1.7069742655520106e-06,
+      "loss": 0.4348,
+      "step": 15690
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 30.96254361674733,
+      "learning_rate": 1.706032024031985e-06,
+      "loss": 0.4761,
+      "step": 15691
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.25259133208445,
+      "learning_rate": 1.7050900183875608e-06,
+      "loss": 0.4165,
+      "step": 15692
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 24.25572723234725,
+      "learning_rate": 1.704148248645523e-06,
+      "loss": 0.4813,
+      "step": 15693
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.79142915029655,
+      "learning_rate": 1.7032067148326536e-06,
+      "loss": 0.4392,
+      "step": 15694
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.939948738332422,
+      "learning_rate": 1.7022654169757312e-06,
+      "loss": 0.4084,
+      "step": 15695
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 24.65556954882072,
+      "learning_rate": 1.701324355101528e-06,
+      "loss": 0.4667,
+      "step": 15696
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 19.71053279245022,
+      "learning_rate": 1.7003835292368064e-06,
+      "loss": 0.5133,
+      "step": 15697
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 22.391741919760776,
+      "learning_rate": 1.6994429394083235e-06,
+      "loss": 0.4073,
+      "step": 15698
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.278930824709033,
+      "learning_rate": 1.698502585642824e-06,
+      "loss": 0.4607,
+      "step": 15699
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.060065826381518,
+      "learning_rate": 1.6975624679670556e-06,
+      "loss": 0.4082,
+      "step": 15700
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 34.31509593921933,
+      "learning_rate": 1.6966225864077557e-06,
+      "loss": 0.5108,
+      "step": 15701
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.508322050031307,
+      "learning_rate": 1.695682940991652e-06,
+      "loss": 0.4909,
+      "step": 15702
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 22.985910287871675,
+      "learning_rate": 1.6947435317454653e-06,
+      "loss": 0.4292,
+      "step": 15703
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 19.47548698427316,
+      "learning_rate": 1.6938043586959141e-06,
+      "loss": 0.4284,
+      "step": 15704
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.050977083961271,
+      "learning_rate": 1.6928654218697105e-06,
+      "loss": 0.3717,
+      "step": 15705
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.90466366723827,
+      "learning_rate": 1.6919267212935531e-06,
+      "loss": 0.5338,
+      "step": 15706
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.04390598473832,
+      "learning_rate": 1.6909882569941405e-06,
+      "loss": 0.5185,
+      "step": 15707
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 19.311964475897952,
+      "learning_rate": 1.6900500289981603e-06,
+      "loss": 0.4271,
+      "step": 15708
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 19.922750902555027,
+      "learning_rate": 1.6891120373322955e-06,
+      "loss": 0.4614,
+      "step": 15709
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 22.186880507350157,
+      "learning_rate": 1.6881742820232249e-06,
+      "loss": 0.4484,
+      "step": 15710
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.79142881944137,
+      "learning_rate": 1.687236763097615e-06,
+      "loss": 0.5004,
+      "step": 15711
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 28.39079508785594,
+      "learning_rate": 1.6862994805821275e-06,
+      "loss": 0.4618,
+      "step": 15712
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 34.46493770176251,
+      "learning_rate": 1.685362434503418e-06,
+      "loss": 0.4565,
+      "step": 15713
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.168428165464427,
+      "learning_rate": 1.6844256248881408e-06,
+      "loss": 0.4587,
+      "step": 15714
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.442181943364776,
+      "learning_rate": 1.6834890517629309e-06,
+      "loss": 0.4486,
+      "step": 15715
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 25.33710046546341,
+      "learning_rate": 1.6825527151544307e-06,
+      "loss": 0.4138,
+      "step": 15716
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.24390139624741,
+      "learning_rate": 1.6816166150892622e-06,
+      "loss": 0.481,
+      "step": 15717
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.863243445189916,
+      "learning_rate": 1.6806807515940537e-06,
+      "loss": 0.4497,
+      "step": 15718
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.33179953520906,
+      "learning_rate": 1.6797451246954166e-06,
+      "loss": 0.4594,
+      "step": 15719
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.528177594521686,
+      "learning_rate": 1.6788097344199594e-06,
+      "loss": 0.4584,
+      "step": 15720
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.745148863171943,
+      "learning_rate": 1.6778745807942899e-06,
+      "loss": 0.3949,
+      "step": 15721
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.658972075059108,
+      "learning_rate": 1.6769396638449954e-06,
+      "loss": 0.438,
+      "step": 15722
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 29.867973959384695,
+      "learning_rate": 1.6760049835986703e-06,
+      "loss": 0.4756,
+      "step": 15723
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 31.91997725680295,
+      "learning_rate": 1.6750705400818913e-06,
+      "loss": 0.4711,
+      "step": 15724
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 25.30707162677131,
+      "learning_rate": 1.6741363333212368e-06,
+      "loss": 0.4578,
+      "step": 15725
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.47445156765777,
+      "learning_rate": 1.673202363343277e-06,
+      "loss": 0.3928,
+      "step": 15726
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.360269565740566,
+      "learning_rate": 1.6722686301745716e-06,
+      "loss": 0.4526,
+      "step": 15727
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.272234146009826,
+      "learning_rate": 1.6713351338416707e-06,
+      "loss": 0.5207,
+      "step": 15728
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.550500597329904,
+      "learning_rate": 1.670401874371128e-06,
+      "loss": 0.475,
+      "step": 15729
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.184570663904687,
+      "learning_rate": 1.6694688517894852e-06,
+      "loss": 0.4715,
+      "step": 15730
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.688760559540253,
+      "learning_rate": 1.6685360661232742e-06,
+      "loss": 0.5298,
+      "step": 15731
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.263396481997386,
+      "learning_rate": 1.6676035173990225e-06,
+      "loss": 0.488,
+      "step": 15732
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.907477818623446,
+      "learning_rate": 1.6666712056432522e-06,
+      "loss": 0.4376,
+      "step": 15733
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.06244295360002,
+      "learning_rate": 1.6657391308824777e-06,
+      "loss": 0.3825,
+      "step": 15734
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.560467777766524,
+      "learning_rate": 1.6648072931432091e-06,
+      "loss": 0.4598,
+      "step": 15735
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.348141967163514,
+      "learning_rate": 1.663875692451945e-06,
+      "loss": 0.4161,
+      "step": 15736
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 28.690267492226873,
+      "learning_rate": 1.6629443288351777e-06,
+      "loss": 0.4758,
+      "step": 15737
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.162360533755802,
+      "learning_rate": 1.6620132023193968e-06,
+      "loss": 0.4485,
+      "step": 15738
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.300505327792237,
+      "learning_rate": 1.661082312931085e-06,
+      "loss": 0.4433,
+      "step": 15739
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.127373860708108,
+      "learning_rate": 1.6601516606967127e-06,
+      "loss": 0.4189,
+      "step": 15740
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.10631513778579,
+      "learning_rate": 1.6592212456427492e-06,
+      "loss": 0.4226,
+      "step": 15741
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.828958180957322,
+      "learning_rate": 1.6582910677956532e-06,
+      "loss": 0.4769,
+      "step": 15742
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.875245256520007,
+      "learning_rate": 1.6573611271818812e-06,
+      "loss": 0.4423,
+      "step": 15743
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.888399507501056,
+      "learning_rate": 1.6564314238278755e-06,
+      "loss": 0.4677,
+      "step": 15744
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.235006550144908,
+      "learning_rate": 1.6555019577600806e-06,
+      "loss": 0.4204,
+      "step": 15745
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 26.225437003227533,
+      "learning_rate": 1.6545727290049307e-06,
+      "loss": 0.5065,
+      "step": 15746
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 27.23459102798812,
+      "learning_rate": 1.653643737588847e-06,
+      "loss": 0.4503,
+      "step": 15747
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.500837972027806,
+      "learning_rate": 1.652714983538256e-06,
+      "loss": 0.402,
+      "step": 15748
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.26094469070454,
+      "learning_rate": 1.6517864668795658e-06,
+      "loss": 0.3832,
+      "step": 15749
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.15137077129673,
+      "learning_rate": 1.650858187639185e-06,
+      "loss": 0.457,
+      "step": 15750
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.12292444306921,
+      "learning_rate": 1.6499301458435146e-06,
+      "loss": 0.4477,
+      "step": 15751
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 34.511774226922874,
+      "learning_rate": 1.6490023415189472e-06,
+      "loss": 0.5406,
+      "step": 15752
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.108051259073665,
+      "learning_rate": 1.6480747746918657e-06,
+      "loss": 0.3869,
+      "step": 15753
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.535380076997066,
+      "learning_rate": 1.6471474453886504e-06,
+      "loss": 0.4209,
+      "step": 15754
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.087204325168425,
+      "learning_rate": 1.6462203536356769e-06,
+      "loss": 0.406,
+      "step": 15755
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.541237341642887,
+      "learning_rate": 1.645293499459314e-06,
+      "loss": 0.4353,
+      "step": 15756
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.850660523697552,
+      "learning_rate": 1.6443668828859126e-06,
+      "loss": 0.5183,
+      "step": 15757
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.681784007261435,
+      "learning_rate": 1.6434405039418277e-06,
+      "loss": 0.4485,
+      "step": 15758
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 24.705386693449253,
+      "learning_rate": 1.642514362653408e-06,
+      "loss": 0.4631,
+      "step": 15759
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.775212829585673,
+      "learning_rate": 1.6415884590469922e-06,
+      "loss": 0.437,
+      "step": 15760
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.09529900212891,
+      "learning_rate": 1.6406627931489116e-06,
+      "loss": 0.4687,
+      "step": 15761
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.013072223904953,
+      "learning_rate": 1.6397373649854874e-06,
+      "loss": 0.4063,
+      "step": 15762
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.860996041540066,
+      "learning_rate": 1.6388121745830433e-06,
+      "loss": 0.4664,
+      "step": 15763
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 26.62977977728416,
+      "learning_rate": 1.637887221967892e-06,
+      "loss": 0.4672,
+      "step": 15764
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 30.710837432205327,
+      "learning_rate": 1.6369625071663365e-06,
+      "loss": 0.4251,
+      "step": 15765
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.91126554833314,
+      "learning_rate": 1.636038030204672e-06,
+      "loss": 0.3985,
+      "step": 15766
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 25.44609519397724,
+      "learning_rate": 1.6351137911091942e-06,
+      "loss": 0.4975,
+      "step": 15767
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 26.950533963484457,
+      "learning_rate": 1.6341897899061887e-06,
+      "loss": 0.4428,
+      "step": 15768
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.44905099303504,
+      "learning_rate": 1.6332660266219292e-06,
+      "loss": 0.4662,
+      "step": 15769
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.980274385446613,
+      "learning_rate": 1.632342501282692e-06,
+      "loss": 0.5126,
+      "step": 15770
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.486799757696925,
+      "learning_rate": 1.6314192139147366e-06,
+      "loss": 0.4376,
+      "step": 15771
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.123059281939863,
+      "learning_rate": 1.630496164544323e-06,
+      "loss": 0.4476,
+      "step": 15772
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.677437329063675,
+      "learning_rate": 1.6295733531977054e-06,
+      "loss": 0.4235,
+      "step": 15773
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.671952821584764,
+      "learning_rate": 1.628650779901122e-06,
+      "loss": 0.4668,
+      "step": 15774
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.2288892900679,
+      "learning_rate": 1.6277284446808162e-06,
+      "loss": 0.3926,
+      "step": 15775
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 28.175068523002945,
+      "learning_rate": 1.6268063475630136e-06,
+      "loss": 0.4747,
+      "step": 15776
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.578656915825086,
+      "learning_rate": 1.6258844885739434e-06,
+      "loss": 0.4324,
+      "step": 15777
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.973499852119023,
+      "learning_rate": 1.6249628677398155e-06,
+      "loss": 0.492,
+      "step": 15778
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.485454415223497,
+      "learning_rate": 1.6240414850868457e-06,
+      "loss": 0.4259,
+      "step": 15779
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.160058083135137,
+      "learning_rate": 1.6231203406412377e-06,
+      "loss": 0.3976,
+      "step": 15780
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 32.365971471735506,
+      "learning_rate": 1.622199434429188e-06,
+      "loss": 0.4728,
+      "step": 15781
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.908113643517492,
+      "learning_rate": 1.6212787664768826e-06,
+      "loss": 0.3815,
+      "step": 15782
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.140819016076644,
+      "learning_rate": 1.620358336810507e-06,
+      "loss": 0.4187,
+      "step": 15783
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.876185456811204,
+      "learning_rate": 1.6194381454562392e-06,
+      "loss": 0.4372,
+      "step": 15784
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.691695485320647,
+      "learning_rate": 1.6185181924402504e-06,
+      "loss": 0.4669,
+      "step": 15785
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.645488250443414,
+      "learning_rate": 1.617598477788701e-06,
+      "loss": 0.3984,
+      "step": 15786
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.609772265441492,
+      "learning_rate": 1.6166790015277446e-06,
+      "loss": 0.4795,
+      "step": 15787
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 26.187348424522906,
+      "learning_rate": 1.6157597636835332e-06,
+      "loss": 0.5007,
+      "step": 15788
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.776798204782633,
+      "learning_rate": 1.6148407642822128e-06,
+      "loss": 0.4749,
+      "step": 15789
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 34.4999099260343,
+      "learning_rate": 1.6139220033499148e-06,
+      "loss": 0.4462,
+      "step": 15790
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 25.093255654545267,
+      "learning_rate": 1.6130034809127682e-06,
+      "loss": 0.4726,
+      "step": 15791
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.5607394258983,
+      "learning_rate": 1.6120851969968954e-06,
+      "loss": 0.4774,
+      "step": 15792
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.460412827336775,
+      "learning_rate": 1.6111671516284155e-06,
+      "loss": 0.5472,
+      "step": 15793
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.65970534525912,
+      "learning_rate": 1.6102493448334312e-06,
+      "loss": 0.4788,
+      "step": 15794
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.722224742957692,
+      "learning_rate": 1.6093317766380511e-06,
+      "loss": 0.4785,
+      "step": 15795
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 27.16006378801096,
+      "learning_rate": 1.608414447068365e-06,
+      "loss": 0.4683,
+      "step": 15796
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.990680896939896,
+      "learning_rate": 1.6074973561504614e-06,
+      "loss": 0.4504,
+      "step": 15797
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.260492871672495,
+      "learning_rate": 1.6065805039104275e-06,
+      "loss": 0.4479,
+      "step": 15798
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.023663228475808,
+      "learning_rate": 1.6056638903743306e-06,
+      "loss": 0.4239,
+      "step": 15799
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.1769134753493,
+      "learning_rate": 1.6047475155682445e-06,
+      "loss": 0.4174,
+      "step": 15800
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.46304274868134,
+      "learning_rate": 1.603831379518227e-06,
+      "loss": 0.352,
+      "step": 15801
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.308352918001425,
+      "learning_rate": 1.6029154822503346e-06,
+      "loss": 0.4251,
+      "step": 15802
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.25301756145544,
+      "learning_rate": 1.601999823790611e-06,
+      "loss": 0.4419,
+      "step": 15803
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.106829267536867,
+      "learning_rate": 1.601084404165103e-06,
+      "loss": 0.4498,
+      "step": 15804
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.17004120742532,
+      "learning_rate": 1.600169223399839e-06,
+      "loss": 0.4301,
+      "step": 15805
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.01652959883348,
+      "learning_rate": 1.59925428152085e-06,
+      "loss": 0.474,
+      "step": 15806
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.756729835713678,
+      "learning_rate": 1.5983395785541533e-06,
+      "loss": 0.3706,
+      "step": 15807
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 27.04470500967914,
+      "learning_rate": 1.5974251145257635e-06,
+      "loss": 0.4483,
+      "step": 15808
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.001271127260578,
+      "learning_rate": 1.596510889461691e-06,
+      "loss": 0.4275,
+      "step": 15809
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 24.1429701360449,
+      "learning_rate": 1.5955969033879304e-06,
+      "loss": 0.4208,
+      "step": 15810
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.11069303382256,
+      "learning_rate": 1.59468315633048e-06,
+      "loss": 0.435,
+      "step": 15811
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.037680370690516,
+      "learning_rate": 1.5937696483153208e-06,
+      "loss": 0.4321,
+      "step": 15812
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.222345161909207,
+      "learning_rate": 1.592856379368436e-06,
+      "loss": 0.45,
+      "step": 15813
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.399253440500903,
+      "learning_rate": 1.5919433495157998e-06,
+      "loss": 0.4899,
+      "step": 15814
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.246569874262015,
+      "learning_rate": 1.5910305587833763e-06,
+      "loss": 0.4882,
+      "step": 15815
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.49344197240153,
+      "learning_rate": 1.5901180071971224e-06,
+      "loss": 0.4686,
+      "step": 15816
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.122848832315956,
+      "learning_rate": 1.5892056947829914e-06,
+      "loss": 0.3949,
+      "step": 15817
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.435141032438963,
+      "learning_rate": 1.588293621566932e-06,
+      "loss": 0.4591,
+      "step": 15818
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.318812911775456,
+      "learning_rate": 1.5873817875748854e-06,
+      "loss": 0.4251,
+      "step": 15819
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.49486935927179,
+      "learning_rate": 1.586470192832774e-06,
+      "loss": 0.4878,
+      "step": 15820
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 26.086181328734426,
+      "learning_rate": 1.5855588373665298e-06,
+      "loss": 0.4475,
+      "step": 15821
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 21.735985332209054,
+      "learning_rate": 1.5846477212020695e-06,
+      "loss": 0.5028,
+      "step": 15822
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.885797790478811,
+      "learning_rate": 1.5837368443653067e-06,
+      "loss": 0.4181,
+      "step": 15823
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 30.79569302694979,
+      "learning_rate": 1.5828262068821453e-06,
+      "loss": 0.4493,
+      "step": 15824
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.925579086135874,
+      "learning_rate": 1.5819158087784802e-06,
+      "loss": 0.4112,
+      "step": 15825
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.919531799959184,
+      "learning_rate": 1.5810056500802063e-06,
+      "loss": 0.4401,
+      "step": 15826
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 25.579587426495788,
+      "learning_rate": 1.580095730813208e-06,
+      "loss": 0.4446,
+      "step": 15827
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.015958006751642,
+      "learning_rate": 1.5791860510033584e-06,
+      "loss": 0.4614,
+      "step": 15828
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.109761882756516,
+      "learning_rate": 1.5782766106765357e-06,
+      "loss": 0.4462,
+      "step": 15829
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 21.728774618941227,
+      "learning_rate": 1.5773674098585955e-06,
+      "loss": 0.4592,
+      "step": 15830
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.566478365320659,
+      "learning_rate": 1.5764584485754031e-06,
+      "loss": 0.4157,
+      "step": 15831
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.097230370856966,
+      "learning_rate": 1.5755497268528008e-06,
+      "loss": 0.4192,
+      "step": 15832
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.77031149116484,
+      "learning_rate": 1.5746412447166371e-06,
+      "loss": 0.4167,
+      "step": 15833
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.808085527581685,
+      "learning_rate": 1.57373300219275e-06,
+      "loss": 0.4761,
+      "step": 15834
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 27.709133620494452,
+      "learning_rate": 1.5728249993069645e-06,
+      "loss": 0.4376,
+      "step": 15835
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 24.312625844232592,
+      "learning_rate": 1.5719172360851088e-06,
+      "loss": 0.4547,
+      "step": 15836
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 28.498933016103503,
+      "learning_rate": 1.5710097125529943e-06,
+      "loss": 0.4462,
+      "step": 15837
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 27.262975946077834,
+      "learning_rate": 1.570102428736432e-06,
+      "loss": 0.4534,
+      "step": 15838
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.098926065143413,
+      "learning_rate": 1.5691953846612284e-06,
+      "loss": 0.4624,
+      "step": 15839
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.155902649449267,
+      "learning_rate": 1.5682885803531767e-06,
+      "loss": 0.3913,
+      "step": 15840
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 21.10024052475499,
+      "learning_rate": 1.5673820158380615e-06,
+      "loss": 0.4716,
+      "step": 15841
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 27.76756608335823,
+      "learning_rate": 1.56647569114167e-06,
+      "loss": 0.4665,
+      "step": 15842
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.606101682044013,
+      "learning_rate": 1.5655696062897763e-06,
+      "loss": 0.422,
+      "step": 15843
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.78048032088841,
+      "learning_rate": 1.5646637613081529e-06,
+      "loss": 0.4435,
+      "step": 15844
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.583449949431344,
+      "learning_rate": 1.5637581562225535e-06,
+      "loss": 0.4397,
+      "step": 15845
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.0700346068123,
+      "learning_rate": 1.5628527910587376e-06,
+      "loss": 0.4828,
+      "step": 15846
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.261357479128073,
+      "learning_rate": 1.5619476658424516e-06,
+      "loss": 0.4376,
+      "step": 15847
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.56570500358424,
+      "learning_rate": 1.5610427805994411e-06,
+      "loss": 0.4662,
+      "step": 15848
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 25.245674551249213,
+      "learning_rate": 1.5601381353554369e-06,
+      "loss": 0.3799,
+      "step": 15849
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.070036385265002,
+      "learning_rate": 1.5592337301361648e-06,
+      "loss": 0.4301,
+      "step": 15850
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.830985084672424,
+      "learning_rate": 1.5583295649673468e-06,
+      "loss": 0.4381,
+      "step": 15851
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.979017601429273,
+      "learning_rate": 1.5574256398747011e-06,
+      "loss": 0.3895,
+      "step": 15852
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.893138392445845,
+      "learning_rate": 1.5565219548839317e-06,
+      "loss": 0.3965,
+      "step": 15853
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 28.620914134829555,
+      "learning_rate": 1.5556185100207356e-06,
+      "loss": 0.512,
+      "step": 15854
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.01075433303274,
+      "learning_rate": 1.5547153053108088e-06,
+      "loss": 0.3863,
+      "step": 15855
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.012544056937227,
+      "learning_rate": 1.5538123407798422e-06,
+      "loss": 0.4644,
+      "step": 15856
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.287770911495855,
+      "learning_rate": 1.5529096164535084e-06,
+      "loss": 0.4781,
+      "step": 15857
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.251015557401683,
+      "learning_rate": 1.552007132357486e-06,
+      "loss": 0.3952,
+      "step": 15858
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.990501520770504,
+      "learning_rate": 1.5511048885174361e-06,
+      "loss": 0.4279,
+      "step": 15859
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.696254944934083,
+      "learning_rate": 1.5502028849590212e-06,
+      "loss": 0.5025,
+      "step": 15860
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.408892287630163,
+      "learning_rate": 1.5493011217078958e-06,
+      "loss": 0.4544,
+      "step": 15861
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.323693465516758,
+      "learning_rate": 1.5483995987897006e-06,
+      "loss": 0.4783,
+      "step": 15862
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 25.099920832368543,
+      "learning_rate": 1.5474983162300794e-06,
+      "loss": 0.3937,
+      "step": 15863
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.983222073991183,
+      "learning_rate": 1.5465972740546586e-06,
+      "loss": 0.4702,
+      "step": 15864
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.119569879496662,
+      "learning_rate": 1.5456964722890688e-06,
+      "loss": 0.4174,
+      "step": 15865
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 28.876207339861036,
+      "learning_rate": 1.5447959109589228e-06,
+      "loss": 0.5151,
+      "step": 15866
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 26.473184638557015,
+      "learning_rate": 1.5438955900898355e-06,
+      "loss": 0.4941,
+      "step": 15867
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 28.065103913462607,
+      "learning_rate": 1.5429955097074133e-06,
+      "loss": 0.4472,
+      "step": 15868
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 23.769763028245873,
+      "learning_rate": 1.5420956698372513e-06,
+      "loss": 0.4528,
+      "step": 15869
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.888035565871476,
+      "learning_rate": 1.5411960705049389e-06,
+      "loss": 0.4603,
+      "step": 15870
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 33.09202352888714,
+      "learning_rate": 1.5402967117360622e-06,
+      "loss": 0.4594,
+      "step": 15871
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 31.206891998675562,
+      "learning_rate": 1.5393975935561977e-06,
+      "loss": 0.4702,
+      "step": 15872
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.277404367015603,
+      "learning_rate": 1.538498715990918e-06,
+      "loss": 0.4106,
+      "step": 15873
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.20349427934717,
+      "learning_rate": 1.5376000790657864e-06,
+      "loss": 0.4515,
+      "step": 15874
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.52893499769126,
+      "learning_rate": 1.5367016828063553e-06,
+      "loss": 0.4192,
+      "step": 15875
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.38506387302692,
+      "learning_rate": 1.5358035272381765e-06,
+      "loss": 0.4357,
+      "step": 15876
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 33.91025539726895,
+      "learning_rate": 1.5349056123867977e-06,
+      "loss": 0.4813,
+      "step": 15877
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 25.84013416937306,
+      "learning_rate": 1.5340079382777506e-06,
+      "loss": 0.5086,
+      "step": 15878
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.89609320740819,
+      "learning_rate": 1.5331105049365624e-06,
+      "loss": 0.4099,
+      "step": 15879
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.336404787758743,
+      "learning_rate": 1.5322133123887595e-06,
+      "loss": 0.4687,
+      "step": 15880
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.779022821362457,
+      "learning_rate": 1.5313163606598591e-06,
+      "loss": 0.4265,
+      "step": 15881
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.83796914897114,
+      "learning_rate": 1.5304196497753642e-06,
+      "loss": 0.4438,
+      "step": 15882
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.019366033482452,
+      "learning_rate": 1.5295231797607824e-06,
+      "loss": 0.4395,
+      "step": 15883
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 25.441464013342028,
+      "learning_rate": 1.528626950641604e-06,
+      "loss": 0.4454,
+      "step": 15884
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.39590464994623,
+      "learning_rate": 1.527730962443319e-06,
+      "loss": 0.5054,
+      "step": 15885
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.341402663888733,
+      "learning_rate": 1.5268352151914124e-06,
+      "loss": 0.4597,
+      "step": 15886
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.32514305865262,
+      "learning_rate": 1.5259397089113525e-06,
+      "loss": 0.4985,
+      "step": 15887
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.520970019557804,
+      "learning_rate": 1.5250444436286138e-06,
+      "loss": 0.423,
+      "step": 15888
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.786852633997505,
+      "learning_rate": 1.52414941936865e-06,
+      "loss": 0.4889,
+      "step": 15889
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.687739286765403,
+      "learning_rate": 1.5232546361569212e-06,
+      "loss": 0.448,
+      "step": 15890
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.65784604741263,
+      "learning_rate": 1.5223600940188688e-06,
+      "loss": 0.5193,
+      "step": 15891
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 25.98049628967927,
+      "learning_rate": 1.521465792979937e-06,
+      "loss": 0.4497,
+      "step": 15892
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.327504345313567,
+      "learning_rate": 1.52057173306556e-06,
+      "loss": 0.4631,
+      "step": 15893
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.638703404546824,
+      "learning_rate": 1.5196779143011619e-06,
+      "loss": 0.4216,
+      "step": 15894
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.099849734228446,
+      "learning_rate": 1.5187843367121623e-06,
+      "loss": 0.4455,
+      "step": 15895
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 26.04739663689291,
+      "learning_rate": 1.5178910003239732e-06,
+      "loss": 0.4662,
+      "step": 15896
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.37733372815419,
+      "learning_rate": 1.5169979051620033e-06,
+      "loss": 0.4447,
+      "step": 15897
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.22351636174334,
+      "learning_rate": 1.5161050512516528e-06,
+      "loss": 0.4236,
+      "step": 15898
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.24288336920094,
+      "learning_rate": 1.515212438618312e-06,
+      "loss": 0.4299,
+      "step": 15899
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.47311512797288,
+      "learning_rate": 1.5143200672873626e-06,
+      "loss": 0.4924,
+      "step": 15900
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.937756796998947,
+      "learning_rate": 1.5134279372841876e-06,
+      "loss": 0.4817,
+      "step": 15901
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.87034107733147,
+      "learning_rate": 1.512536048634159e-06,
+      "loss": 0.4097,
+      "step": 15902
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 23.113704871569478,
+      "learning_rate": 1.5116444013626407e-06,
+      "loss": 0.4651,
+      "step": 15903
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.308026687591433,
+      "learning_rate": 1.5107529954949873e-06,
+      "loss": 0.4509,
+      "step": 15904
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 23.147841699189456,
+      "learning_rate": 1.5098618310565527e-06,
+      "loss": 0.434,
+      "step": 15905
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.195651860447384,
+      "learning_rate": 1.5089709080726811e-06,
+      "loss": 0.4381,
+      "step": 15906
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 23.298331607871916,
+      "learning_rate": 1.5080802265687145e-06,
+      "loss": 0.4811,
+      "step": 15907
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.20899329786817,
+      "learning_rate": 1.5071897865699748e-06,
+      "loss": 0.4278,
+      "step": 15908
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.29002141344829,
+      "learning_rate": 1.5062995881017882e-06,
+      "loss": 0.4126,
+      "step": 15909
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 31.470961504396215,
+      "learning_rate": 1.5054096311894727e-06,
+      "loss": 0.4309,
+      "step": 15910
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.255734776730474,
+      "learning_rate": 1.5045199158583411e-06,
+      "loss": 0.4699,
+      "step": 15911
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.366989873052283,
+      "learning_rate": 1.5036304421336933e-06,
+      "loss": 0.5092,
+      "step": 15912
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 25.176274394080387,
+      "learning_rate": 1.5027412100408234e-06,
+      "loss": 0.4551,
+      "step": 15913
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 25.33427567491345,
+      "learning_rate": 1.5018522196050223e-06,
+      "loss": 0.4993,
+      "step": 15914
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 22.33923602624124,
+      "learning_rate": 1.5009634708515753e-06,
+      "loss": 0.3885,
+      "step": 15915
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 23.750658122782152,
+      "learning_rate": 1.5000749638057521e-06,
+      "loss": 0.4336,
+      "step": 15916
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.47256696674842,
+      "learning_rate": 1.4991866984928283e-06,
+      "loss": 0.3814,
+      "step": 15917
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.77747302466293,
+      "learning_rate": 1.498298674938059e-06,
+      "loss": 0.3898,
+      "step": 15918
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.55067618627532,
+      "learning_rate": 1.4974108931667042e-06,
+      "loss": 0.5037,
+      "step": 15919
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 26.458333694590667,
+      "learning_rate": 1.496523353204008e-06,
+      "loss": 0.4397,
+      "step": 15920
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.780742321410045,
+      "learning_rate": 1.4956360550752125e-06,
+      "loss": 0.3747,
+      "step": 15921
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.674386250382053,
+      "learning_rate": 1.4947489988055552e-06,
+      "loss": 0.4839,
+      "step": 15922
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.032490595906165,
+      "learning_rate": 1.4938621844202595e-06,
+      "loss": 0.4719,
+      "step": 15923
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.980267139579293,
+      "learning_rate": 1.4929756119445481e-06,
+      "loss": 0.4887,
+      "step": 15924
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.9562957245774237,
+      "learning_rate": 1.4920892814036324e-06,
+      "loss": 0.4581,
+      "step": 15925
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.830082211894204,
+      "learning_rate": 1.4912031928227211e-06,
+      "loss": 0.434,
+      "step": 15926
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.20785756003081,
+      "learning_rate": 1.4903173462270148e-06,
+      "loss": 0.5017,
+      "step": 15927
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.31120153494008,
+      "learning_rate": 1.489431741641706e-06,
+      "loss": 0.4552,
+      "step": 15928
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.68809942622622,
+      "learning_rate": 1.4885463790919774e-06,
+      "loss": 0.386,
+      "step": 15929
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 22.93904741690915,
+      "learning_rate": 1.4876612586030104e-06,
+      "loss": 0.4694,
+      "step": 15930
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.773813326665838,
+      "learning_rate": 1.4867763801999768e-06,
+      "loss": 0.513,
+      "step": 15931
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 25.49439518002952,
+      "learning_rate": 1.485891743908049e-06,
+      "loss": 0.4459,
+      "step": 15932
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.820630716911978,
+      "learning_rate": 1.4850073497523743e-06,
+      "loss": 0.4739,
+      "step": 15933
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.739568880549893,
+      "learning_rate": 1.4841231977581094e-06,
+      "loss": 0.4289,
+      "step": 15934
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 22.17817060679014,
+      "learning_rate": 1.4832392879503986e-06,
+      "loss": 0.4394,
+      "step": 15935
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 35.1045746040507,
+      "learning_rate": 1.4823556203543833e-06,
+      "loss": 0.4507,
+      "step": 15936
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.855989124231517,
+      "learning_rate": 1.4814721949951916e-06,
+      "loss": 0.4217,
+      "step": 15937
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.25297416587063,
+      "learning_rate": 1.4805890118979448e-06,
+      "loss": 0.3914,
+      "step": 15938
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 25.95979217932501,
+      "learning_rate": 1.479706071087762e-06,
+      "loss": 0.469,
+      "step": 15939
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.43190194668352,
+      "learning_rate": 1.4788233725897583e-06,
+      "loss": 0.4084,
+      "step": 15940
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.498548627155984,
+      "learning_rate": 1.4779409164290292e-06,
+      "loss": 0.4377,
+      "step": 15941
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 28.901810638691483,
+      "learning_rate": 1.4770587026306782e-06,
+      "loss": 0.4569,
+      "step": 15942
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.361813167127167,
+      "learning_rate": 1.4761767312197906e-06,
+      "loss": 0.3807,
+      "step": 15943
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 25.39141460423376,
+      "learning_rate": 1.4752950022214519e-06,
+      "loss": 0.4308,
+      "step": 15944
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.600865196398455,
+      "learning_rate": 1.4744135156607332e-06,
+      "loss": 0.4703,
+      "step": 15945
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.636290009763123,
+      "learning_rate": 1.4735322715627075e-06,
+      "loss": 0.3893,
+      "step": 15946
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.618136628260608,
+      "learning_rate": 1.4726512699524399e-06,
+      "loss": 0.4129,
+      "step": 15947
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 35.03591338670897,
+      "learning_rate": 1.4717705108549773e-06,
+      "loss": 0.4514,
+      "step": 15948
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 29.705063591574238,
+      "learning_rate": 1.4708899942953757e-06,
+      "loss": 0.5014,
+      "step": 15949
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 31.817246844554766,
+      "learning_rate": 1.4700097202986719e-06,
+      "loss": 0.4789,
+      "step": 15950
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 17.620312435153192,
+      "learning_rate": 1.4691296888899031e-06,
+      "loss": 0.4129,
+      "step": 15951
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.37884239331442,
+      "learning_rate": 1.468249900094093e-06,
+      "loss": 0.5041,
+      "step": 15952
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.99973168714669,
+      "learning_rate": 1.4673703539362683e-06,
+      "loss": 0.439,
+      "step": 15953
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 23.148361787855663,
+      "learning_rate": 1.4664910504414355e-06,
+      "loss": 0.4214,
+      "step": 15954
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.214204773318375,
+      "learning_rate": 1.4656119896346055e-06,
+      "loss": 0.427,
+      "step": 15955
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.789550828908649,
+      "learning_rate": 1.4647331715407809e-06,
+      "loss": 0.4045,
+      "step": 15956
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.866609413254544,
+      "learning_rate": 1.4638545961849525e-06,
+      "loss": 0.4367,
+      "step": 15957
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 34.8019548818339,
+      "learning_rate": 1.462976263592103e-06,
+      "loss": 0.4679,
+      "step": 15958
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.96090897683653,
+      "learning_rate": 1.4620981737872142e-06,
+      "loss": 0.48,
+      "step": 15959
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 23.801750441507693,
+      "learning_rate": 1.4612203267952584e-06,
+      "loss": 0.4926,
+      "step": 15960
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.57347303746456,
+      "learning_rate": 1.4603427226412048e-06,
+      "loss": 0.4394,
+      "step": 15961
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 32.975855135385835,
+      "learning_rate": 1.4594653613500086e-06,
+      "loss": 0.4786,
+      "step": 15962
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 30.657353516450904,
+      "learning_rate": 1.458588242946618e-06,
+      "loss": 0.4305,
+      "step": 15963
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 17.54057945369242,
+      "learning_rate": 1.457711367455983e-06,
+      "loss": 0.4019,
+      "step": 15964
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.056506158792367,
+      "learning_rate": 1.4568347349030398e-06,
+      "loss": 0.4478,
+      "step": 15965
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.068239892757056,
+      "learning_rate": 1.4559583453127213e-06,
+      "loss": 0.4297,
+      "step": 15966
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 23.80717821489896,
+      "learning_rate": 1.455082198709945e-06,
+      "loss": 0.4576,
+      "step": 15967
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 14.432028433405515,
+      "learning_rate": 1.4542062951196335e-06,
+      "loss": 0.4196,
+      "step": 15968
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.264111352038498,
+      "learning_rate": 1.4533306345666954e-06,
+      "loss": 0.4649,
+      "step": 15969
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.010694326691546,
+      "learning_rate": 1.4524552170760374e-06,
+      "loss": 0.456,
+      "step": 15970
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.893013837435078,
+      "learning_rate": 1.4515800426725524e-06,
+      "loss": 0.4185,
+      "step": 15971
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.58392009760701,
+      "learning_rate": 1.4507051113811277e-06,
+      "loss": 0.45,
+      "step": 15972
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 24.573590410816404,
+      "learning_rate": 1.449830423226649e-06,
+      "loss": 0.4549,
+      "step": 15973
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.79215989695097,
+      "learning_rate": 1.4489559782339935e-06,
+      "loss": 0.4456,
+      "step": 15974
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.806146162440406,
+      "learning_rate": 1.4480817764280254e-06,
+      "loss": 0.4965,
+      "step": 15975
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 30.914455157782747,
+      "learning_rate": 1.4472078178336102e-06,
+      "loss": 0.4982,
+      "step": 15976
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 16.522800470251514,
+      "learning_rate": 1.4463341024756006e-06,
+      "loss": 0.4306,
+      "step": 15977
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 24.581338942711422,
+      "learning_rate": 1.4454606303788466e-06,
+      "loss": 0.4073,
+      "step": 15978
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.85094390578264,
+      "learning_rate": 1.4445874015681872e-06,
+      "loss": 0.4205,
+      "step": 15979
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.8059160010933,
+      "learning_rate": 1.4437144160684558e-06,
+      "loss": 0.4377,
+      "step": 15980
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 16.376876506956027,
+      "learning_rate": 1.4428416739044836e-06,
+      "loss": 0.4604,
+      "step": 15981
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.56499205433905,
+      "learning_rate": 1.441969175101089e-06,
+      "loss": 0.4461,
+      "step": 15982
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.94847217455842,
+      "learning_rate": 1.4410969196830827e-06,
+      "loss": 0.4406,
+      "step": 15983
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 38.01468302307978,
+      "learning_rate": 1.4402249076752716e-06,
+      "loss": 0.4775,
+      "step": 15984
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 29.2364358975806,
+      "learning_rate": 1.4393531391024585e-06,
+      "loss": 0.4504,
+      "step": 15985
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.376873624890948,
+      "learning_rate": 1.4384816139894365e-06,
+      "loss": 0.4475,
+      "step": 15986
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.843045444340923,
+      "learning_rate": 1.437610332360988e-06,
+      "loss": 0.3982,
+      "step": 15987
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.12486526406602,
+      "learning_rate": 1.4367392942418912e-06,
+      "loss": 0.46,
+      "step": 15988
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.313938280588555,
+      "learning_rate": 1.4358684996569194e-06,
+      "loss": 0.4364,
+      "step": 15989
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.570817531998458,
+      "learning_rate": 1.4349979486308396e-06,
+      "loss": 0.4876,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.717144820880137,
+      "learning_rate": 1.4341276411884076e-06,
+      "loss": 0.4115,
+      "step": 15991
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.42005007933213,
+      "learning_rate": 1.4332575773543732e-06,
+      "loss": 0.442,
+      "step": 15992
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.655672717787574,
+      "learning_rate": 1.4323877571534817e-06,
+      "loss": 0.4908,
+      "step": 15993
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.220542468255783,
+      "learning_rate": 1.4315181806104694e-06,
+      "loss": 0.43,
+      "step": 15994
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 16.95881503331552,
+      "learning_rate": 1.4306488477500714e-06,
+      "loss": 0.394,
+      "step": 15995
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.022329283281767,
+      "learning_rate": 1.4297797585970063e-06,
+      "loss": 0.4465,
+      "step": 15996
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.07103298678883,
+      "learning_rate": 1.4289109131759894e-06,
+      "loss": 0.3926,
+      "step": 15997
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.532406228239893,
+      "learning_rate": 1.4280423115117316e-06,
+      "loss": 0.411,
+      "step": 15998
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.49068499277464,
+      "learning_rate": 1.4271739536289386e-06,
+      "loss": 0.4644,
+      "step": 15999
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 24.123598302629787,
+      "learning_rate": 1.4263058395523044e-06,
+      "loss": 0.4253,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.33500406266326,
+      "learning_rate": 1.4254379693065134e-06,
+      "loss": 0.4545,
+      "step": 16001
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 25.72876902326394,
+      "learning_rate": 1.4245703429162504e-06,
+      "loss": 0.3899,
+      "step": 16002
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.910769522184466,
+      "learning_rate": 1.4237029604061936e-06,
+      "loss": 0.4761,
+      "step": 16003
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.265467622710162,
+      "learning_rate": 1.422835821801004e-06,
+      "loss": 0.4419,
+      "step": 16004
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.359258228902299,
+      "learning_rate": 1.42196892712535e-06,
+      "loss": 0.426,
+      "step": 16005
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 16.92926850568379,
+      "learning_rate": 1.4211022764038785e-06,
+      "loss": 0.4148,
+      "step": 16006
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 32.90000095425404,
+      "learning_rate": 1.4202358696612417e-06,
+      "loss": 0.3752,
+      "step": 16007
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 24.18432842020453,
+      "learning_rate": 1.4193697069220758e-06,
+      "loss": 0.425,
+      "step": 16008
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 22.54681177586437,
+      "learning_rate": 1.4185037882110165e-06,
+      "loss": 0.4678,
+      "step": 16009
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 31.959970348180875,
+      "learning_rate": 1.4176381135526906e-06,
+      "loss": 0.4374,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 24.57881377540959,
+      "learning_rate": 1.4167726829717155e-06,
+      "loss": 0.4897,
+      "step": 16011
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.11600947514142,
+      "learning_rate": 1.4159074964927055e-06,
+      "loss": 0.4423,
+      "step": 16012
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.540866472246,
+      "learning_rate": 1.4150425541402623e-06,
+      "loss": 0.4107,
+      "step": 16013
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.356355147104633,
+      "learning_rate": 1.4141778559389875e-06,
+      "loss": 0.4353,
+      "step": 16014
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.35814947218924,
+      "learning_rate": 1.413313401913473e-06,
+      "loss": 0.5225,
+      "step": 16015
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 32.787733076047395,
+      "learning_rate": 1.4124491920883031e-06,
+      "loss": 0.4293,
+      "step": 16016
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.90812415991012,
+      "learning_rate": 1.4115852264880525e-06,
+      "loss": 0.4075,
+      "step": 16017
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.233990539348376,
+      "learning_rate": 1.4107215051372935e-06,
+      "loss": 0.4186,
+      "step": 16018
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.485973633662397,
+      "learning_rate": 1.409858028060591e-06,
+      "loss": 0.4672,
+      "step": 16019
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.039306624189297,
+      "learning_rate": 1.4089947952825057e-06,
+      "loss": 0.4286,
+      "step": 16020
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 22.8893474463208,
+      "learning_rate": 1.4081318068275773e-06,
+      "loss": 0.4361,
+      "step": 16021
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 24.24454182777104,
+      "learning_rate": 1.4072690627203544e-06,
+      "loss": 0.4408,
+      "step": 16022
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.432240352625374,
+      "learning_rate": 1.406406562985373e-06,
+      "loss": 0.4632,
+      "step": 16023
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 22.903756152927027,
+      "learning_rate": 1.4055443076471641e-06,
+      "loss": 0.465,
+      "step": 16024
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.215776742143245,
+      "learning_rate": 1.4046822967302477e-06,
+      "loss": 0.4997,
+      "step": 16025
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.620105898580562,
+      "learning_rate": 1.4038205302591368e-06,
+      "loss": 0.4179,
+      "step": 16026
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.436592743890895,
+      "learning_rate": 1.4029590082583411e-06,
+      "loss": 0.4302,
+      "step": 16027
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 29.291359342036372,
+      "learning_rate": 1.4020977307523642e-06,
+      "loss": 0.4827,
+      "step": 16028
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 30.086728616306587,
+      "learning_rate": 1.4012366977656954e-06,
+      "loss": 0.442,
+      "step": 16029
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.25214179815325,
+      "learning_rate": 1.4003759093228275e-06,
+      "loss": 0.3874,
+      "step": 16030
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.431748455690133,
+      "learning_rate": 1.3995153654482363e-06,
+      "loss": 0.4182,
+      "step": 16031
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.557807866195223,
+      "learning_rate": 1.3986550661663978e-06,
+      "loss": 0.393,
+      "step": 16032
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.632800985802522,
+      "learning_rate": 1.3977950115017768e-06,
+      "loss": 0.4477,
+      "step": 16033
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 28.932862035770018,
+      "learning_rate": 1.3969352014788328e-06,
+      "loss": 0.457,
+      "step": 16034
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.31484045201061,
+      "learning_rate": 1.3960756361220219e-06,
+      "loss": 0.5123,
+      "step": 16035
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 30.49663488495039,
+      "learning_rate": 1.395216315455784e-06,
+      "loss": 0.4648,
+      "step": 16036
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.023862308532454,
+      "learning_rate": 1.3943572395045624e-06,
+      "loss": 0.4263,
+      "step": 16037
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.209689269179492,
+      "learning_rate": 1.393498408292785e-06,
+      "loss": 0.4343,
+      "step": 16038
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 32.665241953493314,
+      "learning_rate": 1.3926398218448778e-06,
+      "loss": 0.4288,
+      "step": 16039
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.577035718088354,
+      "learning_rate": 1.3917814801852615e-06,
+      "loss": 0.4358,
+      "step": 16040
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.79365517657097,
+      "learning_rate": 1.390923383338345e-06,
+      "loss": 0.4464,
+      "step": 16041
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 31.17829810535406,
+      "learning_rate": 1.3900655313285283e-06,
+      "loss": 0.416,
+      "step": 16042
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 24.73465749021419,
+      "learning_rate": 1.3892079241802104e-06,
+      "loss": 0.3867,
+      "step": 16043
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 25.43623042951149,
+      "learning_rate": 1.3883505619177829e-06,
+      "loss": 0.4216,
+      "step": 16044
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.731861722553425,
+      "learning_rate": 1.3874934445656319e-06,
+      "loss": 0.4715,
+      "step": 16045
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 20.884758223208838,
+      "learning_rate": 1.3866365721481256e-06,
+      "loss": 0.3903,
+      "step": 16046
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 28.662180169975024,
+      "learning_rate": 1.3857799446896358e-06,
+      "loss": 0.4619,
+      "step": 16047
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 27.883706859547214,
+      "learning_rate": 1.3849235622145253e-06,
+      "loss": 0.4339,
+      "step": 16048
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.12828886108539,
+      "learning_rate": 1.3840674247471508e-06,
+      "loss": 0.4407,
+      "step": 16049
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.025920230127273,
+      "learning_rate": 1.383211532311859e-06,
+      "loss": 0.4243,
+      "step": 16050
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.96181464946443,
+      "learning_rate": 1.3823558849329877e-06,
+      "loss": 0.4347,
+      "step": 16051
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.193194252977401,
+      "learning_rate": 1.3815004826348744e-06,
+      "loss": 0.3938,
+      "step": 16052
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.72867822493261,
+      "learning_rate": 1.3806453254418483e-06,
+      "loss": 0.4712,
+      "step": 16053
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.049578108965338,
+      "learning_rate": 1.3797904133782269e-06,
+      "loss": 0.4461,
+      "step": 16054
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.052904414915574,
+      "learning_rate": 1.3789357464683206e-06,
+      "loss": 0.4673,
+      "step": 16055
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 27.994096920293632,
+      "learning_rate": 1.3780813247364388e-06,
+      "loss": 0.4564,
+      "step": 16056
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.532372325701015,
+      "learning_rate": 1.3772271482068799e-06,
+      "loss": 0.4643,
+      "step": 16057
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.587454237994585,
+      "learning_rate": 1.3763732169039401e-06,
+      "loss": 0.6139,
+      "step": 16058
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.64526821914078,
+      "learning_rate": 1.375519530851901e-06,
+      "loss": 0.4037,
+      "step": 16059
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.01109019400029,
+      "learning_rate": 1.3746660900750386e-06,
+      "loss": 0.4552,
+      "step": 16060
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 20.013676850076404,
+      "learning_rate": 1.373812894597627e-06,
+      "loss": 0.4199,
+      "step": 16061
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.5749172445807,
+      "learning_rate": 1.3729599444439323e-06,
+      "loss": 0.4345,
+      "step": 16062
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.31902425804104,
+      "learning_rate": 1.3721072396382085e-06,
+      "loss": 0.4092,
+      "step": 16063
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 29.792735227792747,
+      "learning_rate": 1.3712547802047094e-06,
+      "loss": 0.5029,
+      "step": 16064
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.350584615520887,
+      "learning_rate": 1.3704025661676757e-06,
+      "loss": 0.4988,
+      "step": 16065
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 24.355428596836692,
+      "learning_rate": 1.3695505975513456e-06,
+      "loss": 0.4713,
+      "step": 16066
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.577649434470043,
+      "learning_rate": 1.3686988743799466e-06,
+      "loss": 0.4024,
+      "step": 16067
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.396153702389274,
+      "learning_rate": 1.3678473966777018e-06,
+      "loss": 0.4521,
+      "step": 16068
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.607760226309008,
+      "learning_rate": 1.36699616446883e-06,
+      "loss": 0.4152,
+      "step": 16069
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 29.978763376248963,
+      "learning_rate": 1.3661451777775369e-06,
+      "loss": 0.4588,
+      "step": 16070
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.434229630867517,
+      "learning_rate": 1.3652944366280218e-06,
+      "loss": 0.4732,
+      "step": 16071
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.066913181196533,
+      "learning_rate": 1.364443941044482e-06,
+      "loss": 0.4508,
+      "step": 16072
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 20.84857807119181,
+      "learning_rate": 1.363593691051105e-06,
+      "loss": 0.4867,
+      "step": 16073
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.880398707938138,
+      "learning_rate": 1.3627436866720734e-06,
+      "loss": 0.4079,
+      "step": 16074
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.449364661795336,
+      "learning_rate": 1.3618939279315591e-06,
+      "loss": 0.4162,
+      "step": 16075
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 28.60231956932494,
+      "learning_rate": 1.3610444148537261e-06,
+      "loss": 0.4687,
+      "step": 16076
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.10890978011818,
+      "learning_rate": 1.3601951474627362e-06,
+      "loss": 0.4715,
+      "step": 16077
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.36689264581611,
+      "learning_rate": 1.3593461257827433e-06,
+      "loss": 0.4134,
+      "step": 16078
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.24007244469165,
+      "learning_rate": 1.3584973498378928e-06,
+      "loss": 0.4382,
+      "step": 16079
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.53252686380978,
+      "learning_rate": 1.3576488196523207e-06,
+      "loss": 0.3798,
+      "step": 16080
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 24.835521443719877,
+      "learning_rate": 1.3568005352501602e-06,
+      "loss": 0.4876,
+      "step": 16081
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.419010135196174,
+      "learning_rate": 1.355952496655536e-06,
+      "loss": 0.4499,
+      "step": 16082
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.79683973428614,
+      "learning_rate": 1.3551047038925692e-06,
+      "loss": 0.475,
+      "step": 16083
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.234364634410497,
+      "learning_rate": 1.3542571569853669e-06,
+      "loss": 0.3969,
+      "step": 16084
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 29.244811420088652,
+      "learning_rate": 1.353409855958031e-06,
+      "loss": 0.4857,
+      "step": 16085
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.346327759358495,
+      "learning_rate": 1.3525628008346613e-06,
+      "loss": 0.4477,
+      "step": 16086
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.592895253216835,
+      "learning_rate": 1.3517159916393485e-06,
+      "loss": 0.3904,
+      "step": 16087
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.77317982561453,
+      "learning_rate": 1.350869428396172e-06,
+      "loss": 0.4769,
+      "step": 16088
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.477146218228754,
+      "learning_rate": 1.3500231111292118e-06,
+      "loss": 0.4318,
+      "step": 16089
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.350442247701558,
+      "learning_rate": 1.3491770398625315e-06,
+      "loss": 0.4135,
+      "step": 16090
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 23.693940265991415,
+      "learning_rate": 1.3483312146201988e-06,
+      "loss": 0.4339,
+      "step": 16091
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.730008705654967,
+      "learning_rate": 1.3474856354262622e-06,
+      "loss": 0.366,
+      "step": 16092
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 33.185076490374136,
+      "learning_rate": 1.3466403023047724e-06,
+      "loss": 0.4819,
+      "step": 16093
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.95929009220646,
+      "learning_rate": 1.3457952152797727e-06,
+      "loss": 0.4776,
+      "step": 16094
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 24.565996670989747,
+      "learning_rate": 1.3449503743752945e-06,
+      "loss": 0.4369,
+      "step": 16095
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 27.02435263467979,
+      "learning_rate": 1.3441057796153634e-06,
+      "loss": 0.4302,
+      "step": 16096
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.766718204264784,
+      "learning_rate": 1.3432614310239989e-06,
+      "loss": 0.45,
+      "step": 16097
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.88300408973077,
+      "learning_rate": 1.3424173286252185e-06,
+      "loss": 0.4317,
+      "step": 16098
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.461529160615942,
+      "learning_rate": 1.3415734724430219e-06,
+      "loss": 0.4074,
+      "step": 16099
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.915799160713203,
+      "learning_rate": 1.340729862501413e-06,
+      "loss": 0.412,
+      "step": 16100
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 37.88191126842757,
+      "learning_rate": 1.3398864988243786e-06,
+      "loss": 0.453,
+      "step": 16101
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.244575570803324,
+      "learning_rate": 1.3390433814359072e-06,
+      "loss": 0.4578,
+      "step": 16102
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.751859441621695,
+      "learning_rate": 1.338200510359977e-06,
+      "loss": 0.4327,
+      "step": 16103
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.355720114478938,
+      "learning_rate": 1.3373578856205583e-06,
+      "loss": 0.4698,
+      "step": 16104
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 23.373897884773598,
+      "learning_rate": 1.3365155072416103e-06,
+      "loss": 0.42,
+      "step": 16105
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.9143363783638,
+      "learning_rate": 1.335673375247094e-06,
+      "loss": 0.4645,
+      "step": 16106
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.02268170599414,
+      "learning_rate": 1.3348314896609581e-06,
+      "loss": 0.4158,
+      "step": 16107
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.200770929030377,
+      "learning_rate": 1.3339898505071503e-06,
+      "loss": 0.454,
+      "step": 16108
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.129009123358525,
+      "learning_rate": 1.3331484578095976e-06,
+      "loss": 0.3922,
+      "step": 16109
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 24.081902440972662,
+      "learning_rate": 1.3323073115922324e-06,
+      "loss": 0.3776,
+      "step": 16110
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.908004810139264,
+      "learning_rate": 1.3314664118789766e-06,
+      "loss": 0.4503,
+      "step": 16111
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 29.913964095791915,
+      "learning_rate": 1.3306257586937476e-06,
+      "loss": 0.4645,
+      "step": 16112
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.913729501256434,
+      "learning_rate": 1.3297853520604497e-06,
+      "loss": 0.4864,
+      "step": 16113
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.73875133979761,
+      "learning_rate": 1.3289451920029838e-06,
+      "loss": 0.4436,
+      "step": 16114
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.27661197764153,
+      "learning_rate": 1.328105278545243e-06,
+      "loss": 0.3886,
+      "step": 16115
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.03122522778843,
+      "learning_rate": 1.3272656117111183e-06,
+      "loss": 0.42,
+      "step": 16116
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 23.480596286577722,
+      "learning_rate": 1.3264261915244836e-06,
+      "loss": 0.4274,
+      "step": 16117
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 24.36375371864161,
+      "learning_rate": 1.3255870180092157e-06,
+      "loss": 0.4866,
+      "step": 16118
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.549188702571065,
+      "learning_rate": 1.3247480911891775e-06,
+      "loss": 0.4201,
+      "step": 16119
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.961146512215674,
+      "learning_rate": 1.3239094110882279e-06,
+      "loss": 0.399,
+      "step": 16120
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 32.790279165894724,
+      "learning_rate": 1.323070977730222e-06,
+      "loss": 0.4622,
+      "step": 16121
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 23.84024557993538,
+      "learning_rate": 1.3222327911389988e-06,
+      "loss": 0.4507,
+      "step": 16122
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.435744356362395,
+      "learning_rate": 1.3213948513384011e-06,
+      "loss": 0.3852,
+      "step": 16123
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.939901484215818,
+      "learning_rate": 1.320557158352256e-06,
+      "loss": 0.4418,
+      "step": 16124
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 27.66030546562115,
+      "learning_rate": 1.3197197122043892e-06,
+      "loss": 0.3918,
+      "step": 16125
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 34.52606753646948,
+      "learning_rate": 1.3188825129186134e-06,
+      "loss": 0.5061,
+      "step": 16126
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.435270130179433,
+      "learning_rate": 1.3180455605187404e-06,
+      "loss": 0.5326,
+      "step": 16127
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 21.34916310763949,
+      "learning_rate": 1.3172088550285756e-06,
+      "loss": 0.4382,
+      "step": 16128
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.489164564853382,
+      "learning_rate": 1.3163723964719122e-06,
+      "loss": 0.4377,
+      "step": 16129
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.49358801901327,
+      "learning_rate": 1.3155361848725357e-06,
+      "loss": 0.4037,
+      "step": 16130
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.431644727348466,
+      "learning_rate": 1.3147002202542291e-06,
+      "loss": 0.4086,
+      "step": 16131
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.87720014982173,
+      "learning_rate": 1.3138645026407682e-06,
+      "loss": 0.5487,
+      "step": 16132
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 28.105182428339685,
+      "learning_rate": 1.3130290320559235e-06,
+      "loss": 0.4736,
+      "step": 16133
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.121196676710916,
+      "learning_rate": 1.3121938085234466e-06,
+      "loss": 0.3873,
+      "step": 16134
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 25.141077182181313,
+      "learning_rate": 1.3113588320670967e-06,
+      "loss": 0.521,
+      "step": 16135
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 25.3080304067366,
+      "learning_rate": 1.3105241027106187e-06,
+      "loss": 0.4905,
+      "step": 16136
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 14.552060328852011,
+      "learning_rate": 1.3096896204777532e-06,
+      "loss": 0.4588,
+      "step": 16137
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.95056523741904,
+      "learning_rate": 1.3088553853922325e-06,
+      "loss": 0.4285,
+      "step": 16138
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.930672067267473,
+      "learning_rate": 1.3080213974777766e-06,
+      "loss": 0.4071,
+      "step": 16139
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 21.45939903035727,
+      "learning_rate": 1.3071876567581087e-06,
+      "loss": 0.4511,
+      "step": 16140
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 28.289480477010375,
+      "learning_rate": 1.306354163256941e-06,
+      "loss": 0.4329,
+      "step": 16141
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.187050965291913,
+      "learning_rate": 1.3055209169979754e-06,
+      "loss": 0.4594,
+      "step": 16142
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 20.203624089294813,
+      "learning_rate": 1.3046879180049054e-06,
+      "loss": 0.4578,
+      "step": 16143
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 39.72745619587268,
+      "learning_rate": 1.3038551663014255e-06,
+      "loss": 0.4512,
+      "step": 16144
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.021933170998004,
+      "learning_rate": 1.3030226619112175e-06,
+      "loss": 0.4401,
+      "step": 16145
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 24.099133587682196,
+      "learning_rate": 1.3021904048579592e-06,
+      "loss": 0.4342,
+      "step": 16146
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.128758235745906,
+      "learning_rate": 1.3013583951653185e-06,
+      "loss": 0.4436,
+      "step": 16147
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.877803676924344,
+      "learning_rate": 1.3005266328569533e-06,
+      "loss": 0.4927,
+      "step": 16148
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 21.34779687540237,
+      "learning_rate": 1.2996951179565221e-06,
+      "loss": 0.4455,
+      "step": 16149
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 21.475657447125872,
+      "learning_rate": 1.2988638504876738e-06,
+      "loss": 0.4777,
+      "step": 16150
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.98429490863332,
+      "learning_rate": 1.2980328304740464e-06,
+      "loss": 0.4661,
+      "step": 16151
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.879735638306125,
+      "learning_rate": 1.2972020579392763e-06,
+      "loss": 0.4237,
+      "step": 16152
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 23.275216031121918,
+      "learning_rate": 1.296371532906987e-06,
+      "loss": 0.4969,
+      "step": 16153
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 27.978751454050144,
+      "learning_rate": 1.2955412554008018e-06,
+      "loss": 0.5051,
+      "step": 16154
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.76163886252795,
+      "learning_rate": 1.2947112254443283e-06,
+      "loss": 0.4076,
+      "step": 16155
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.30297299157287,
+      "learning_rate": 1.2938814430611756e-06,
+      "loss": 0.4316,
+      "step": 16156
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.578792343866596,
+      "learning_rate": 1.2930519082749448e-06,
+      "loss": 0.4022,
+      "step": 16157
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.995794345629422,
+      "learning_rate": 1.2922226211092237e-06,
+      "loss": 0.4509,
+      "step": 16158
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.25341360948246,
+      "learning_rate": 1.2913935815875945e-06,
+      "loss": 0.4199,
+      "step": 16159
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.742479203684088,
+      "learning_rate": 1.2905647897336382e-06,
+      "loss": 0.4515,
+      "step": 16160
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.099668360077146,
+      "learning_rate": 1.289736245570925e-06,
+      "loss": 0.4401,
+      "step": 16161
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.951648367037468,
+      "learning_rate": 1.2889079491230184e-06,
+      "loss": 0.4362,
+      "step": 16162
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 30.03526686559841,
+      "learning_rate": 1.2880799004134748e-06,
+      "loss": 0.4427,
+      "step": 16163
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.127564521880164,
+      "learning_rate": 1.2872520994658399e-06,
+      "loss": 0.3623,
+      "step": 16164
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 21.367385864731812,
+      "learning_rate": 1.2864245463036585e-06,
+      "loss": 0.4744,
+      "step": 16165
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.521943041635705,
+      "learning_rate": 1.285597240950468e-06,
+      "loss": 0.4562,
+      "step": 16166
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 25.701737878931475,
+      "learning_rate": 1.2847701834297944e-06,
+      "loss": 0.4992,
+      "step": 16167
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.597040625228257,
+      "learning_rate": 1.2839433737651564e-06,
+      "loss": 0.4109,
+      "step": 16168
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.400038246444073,
+      "learning_rate": 1.2831168119800696e-06,
+      "loss": 0.4778,
+      "step": 16169
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.384701177462002,
+      "learning_rate": 1.2822904980980422e-06,
+      "loss": 0.4188,
+      "step": 16170
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 28.6465904502806,
+      "learning_rate": 1.2814644321425752e-06,
+      "loss": 0.4454,
+      "step": 16171
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.54095085520488,
+      "learning_rate": 1.2806386141371596e-06,
+      "loss": 0.4357,
+      "step": 16172
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.882080351152075,
+      "learning_rate": 1.279813044105278e-06,
+      "loss": 0.4967,
+      "step": 16173
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.645148332403927,
+      "learning_rate": 1.2789877220704127e-06,
+      "loss": 0.4162,
+      "step": 16174
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.405512592262003,
+      "learning_rate": 1.2781626480560384e-06,
+      "loss": 0.4495,
+      "step": 16175
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 21.214584773953103,
+      "learning_rate": 1.2773378220856126e-06,
+      "loss": 0.4208,
+      "step": 16176
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.417531345061835,
+      "learning_rate": 1.2765132441825989e-06,
+      "loss": 0.3994,
+      "step": 16177
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 16.558808744906163,
+      "learning_rate": 1.2756889143704443e-06,
+      "loss": 0.4769,
+      "step": 16178
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 29.28959923896237,
+      "learning_rate": 1.2748648326725943e-06,
+      "loss": 0.4233,
+      "step": 16179
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 21.396533118104976,
+      "learning_rate": 1.2740409991124826e-06,
+      "loss": 0.3972,
+      "step": 16180
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.996897906889256,
+      "learning_rate": 1.273217413713541e-06,
+      "loss": 0.5071,
+      "step": 16181
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 29.145206892492695,
+      "learning_rate": 1.2723940764991927e-06,
+      "loss": 0.5035,
+      "step": 16182
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.378472820613535,
+      "learning_rate": 1.2715709874928518e-06,
+      "loss": 0.4704,
+      "step": 16183
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.606376671089954,
+      "learning_rate": 1.2707481467179228e-06,
+      "loss": 0.4685,
+      "step": 16184
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.684463760661522,
+      "learning_rate": 1.2699255541978117e-06,
+      "loss": 0.4176,
+      "step": 16185
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 32.92323274487656,
+      "learning_rate": 1.2691032099559097e-06,
+      "loss": 0.5049,
+      "step": 16186
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.893253018705774,
+      "learning_rate": 1.268281114015607e-06,
+      "loss": 0.4758,
+      "step": 16187
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 29.18216352260454,
+      "learning_rate": 1.2674592664002828e-06,
+      "loss": 0.4583,
+      "step": 16188
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 22.359304695898565,
+      "learning_rate": 1.2666376671333048e-06,
+      "loss": 0.4736,
+      "step": 16189
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.29364880022195,
+      "learning_rate": 1.2658163162380443e-06,
+      "loss": 0.4767,
+      "step": 16190
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.124525445447688,
+      "learning_rate": 1.2649952137378596e-06,
+      "loss": 0.5043,
+      "step": 16191
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.6021441932411,
+      "learning_rate": 1.2641743596561007e-06,
+      "loss": 0.4006,
+      "step": 16192
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.132631134151765,
+      "learning_rate": 1.2633537540161123e-06,
+      "loss": 0.4597,
+      "step": 16193
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.867707415623116,
+      "learning_rate": 1.2625333968412312e-06,
+      "loss": 0.4097,
+      "step": 16194
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.063351136889246,
+      "learning_rate": 1.261713288154789e-06,
+      "loss": 0.4489,
+      "step": 16195
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.23634535290724,
+      "learning_rate": 1.2608934279801133e-06,
+      "loss": 0.4439,
+      "step": 16196
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.721883468887953,
+      "learning_rate": 1.2600738163405124e-06,
+      "loss": 0.3929,
+      "step": 16197
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 22.324421003465098,
+      "learning_rate": 1.2592544532592988e-06,
+      "loss": 0.4098,
+      "step": 16198
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.71749577334569,
+      "learning_rate": 1.2584353387597758e-06,
+      "loss": 0.4543,
+      "step": 16199
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 30.93114246878487,
+      "learning_rate": 1.2576164728652406e-06,
+      "loss": 0.4372,
+      "step": 16200
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 24.22293906007539,
+      "learning_rate": 1.256797855598978e-06,
+      "loss": 0.525,
+      "step": 16201
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.434392992707878,
+      "learning_rate": 1.255979486984269e-06,
+      "loss": 0.418,
+      "step": 16202
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.143526084750306,
+      "learning_rate": 1.2551613670443862e-06,
+      "loss": 0.4043,
+      "step": 16203
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 16.07366426422336,
+      "learning_rate": 1.2543434958026024e-06,
+      "loss": 0.4313,
+      "step": 16204
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.119309566747837,
+      "learning_rate": 1.253525873282171e-06,
+      "loss": 0.3971,
+      "step": 16205
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.379038246841734,
+      "learning_rate": 1.2527084995063488e-06,
+      "loss": 0.391,
+      "step": 16206
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.502078468770403,
+      "learning_rate": 1.2518913744983786e-06,
+      "loss": 0.4458,
+      "step": 16207
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.802937243285648,
+      "learning_rate": 1.2510744982814993e-06,
+      "loss": 0.4392,
+      "step": 16208
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.102730585342382,
+      "learning_rate": 1.250257870878946e-06,
+      "loss": 0.4205,
+      "step": 16209
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 21.19327764976099,
+      "learning_rate": 1.2494414923139375e-06,
+      "loss": 0.3977,
+      "step": 16210
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 22.61116185472731,
+      "learning_rate": 1.2486253626096978e-06,
+      "loss": 0.3901,
+      "step": 16211
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 27.417180792549605,
+      "learning_rate": 1.2478094817894293e-06,
+      "loss": 0.513,
+      "step": 16212
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 24.06952805681748,
+      "learning_rate": 1.2469938498763423e-06,
+      "loss": 0.4251,
+      "step": 16213
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.828894784814825,
+      "learning_rate": 1.2461784668936283e-06,
+      "loss": 0.4059,
+      "step": 16214
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.310972599807217,
+      "learning_rate": 1.2453633328644765e-06,
+      "loss": 0.4425,
+      "step": 16215
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 22.515909237493734,
+      "learning_rate": 1.244548447812074e-06,
+      "loss": 0.4189,
+      "step": 16216
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 16.304867524568387,
+      "learning_rate": 1.243733811759591e-06,
+      "loss": 0.421,
+      "step": 16217
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 29.198554136947916,
+      "learning_rate": 1.2429194247301934e-06,
+      "loss": 0.4726,
+      "step": 16218
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.18794052924772,
+      "learning_rate": 1.2421052867470452e-06,
+      "loss": 0.4199,
+      "step": 16219
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.717766903570451,
+      "learning_rate": 1.2412913978332997e-06,
+      "loss": 0.4405,
+      "step": 16220
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 16.601425967833137,
+      "learning_rate": 1.2404777580121075e-06,
+      "loss": 0.4061,
+      "step": 16221
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 31.139884177170483,
+      "learning_rate": 1.2396643673065988e-06,
+      "loss": 0.5397,
+      "step": 16222
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.946683831331104,
+      "learning_rate": 1.238851225739911e-06,
+      "loss": 0.476,
+      "step": 16223
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 32.56812494059177,
+      "learning_rate": 1.2380383333351687e-06,
+      "loss": 0.4861,
+      "step": 16224
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 28.42693705199055,
+      "learning_rate": 1.2372256901154934e-06,
+      "loss": 0.467,
+      "step": 16225
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 27.79674657853102,
+      "learning_rate": 1.2364132961039933e-06,
+      "loss": 0.4302,
+      "step": 16226
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 33.61215200084452,
+      "learning_rate": 1.2356011513237708e-06,
+      "loss": 0.5053,
+      "step": 16227
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.108951646213594,
+      "learning_rate": 1.2347892557979236e-06,
+      "loss": 0.4456,
+      "step": 16228
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 22.40774533005232,
+      "learning_rate": 1.233977609549546e-06,
+      "loss": 0.4224,
+      "step": 16229
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 27.591776182143025,
+      "learning_rate": 1.2331662126017141e-06,
+      "loss": 0.4777,
+      "step": 16230
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 23.88816294511219,
+      "learning_rate": 1.2323550649775085e-06,
+      "loss": 0.4592,
+      "step": 16231
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.915892017501292,
+      "learning_rate": 1.2315441666999939e-06,
+      "loss": 0.4355,
+      "step": 16232
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.90625801864774,
+      "learning_rate": 1.2307335177922342e-06,
+      "loss": 0.4418,
+      "step": 16233
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.950949541329113,
+      "learning_rate": 1.2299231182772852e-06,
+      "loss": 0.3922,
+      "step": 16234
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.943142660519563,
+      "learning_rate": 1.229112968178191e-06,
+      "loss": 0.4171,
+      "step": 16235
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 26.25359470413586,
+      "learning_rate": 1.2283030675179951e-06,
+      "loss": 0.4693,
+      "step": 16236
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.474521105256585,
+      "learning_rate": 1.227493416319726e-06,
+      "loss": 0.3891,
+      "step": 16237
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.272140795207816,
+      "learning_rate": 1.226684014606414e-06,
+      "loss": 0.413,
+      "step": 16238
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 28.175248185835784,
+      "learning_rate": 1.2258748624010752e-06,
+      "loss": 0.415,
+      "step": 16239
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 24.11331563107011,
+      "learning_rate": 1.2250659597267244e-06,
+      "loss": 0.4373,
+      "step": 16240
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.56547091595823,
+      "learning_rate": 1.2242573066063623e-06,
+      "loss": 0.3961,
+      "step": 16241
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.297519435494202,
+      "learning_rate": 1.2234489030629916e-06,
+      "loss": 0.4004,
+      "step": 16242
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.961800212842434,
+      "learning_rate": 1.2226407491195969e-06,
+      "loss": 0.4261,
+      "step": 16243
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.35944365916944,
+      "learning_rate": 1.2218328447991657e-06,
+      "loss": 0.4462,
+      "step": 16244
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.408130405496266,
+      "learning_rate": 1.2210251901246739e-06,
+      "loss": 0.4898,
+      "step": 16245
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 20.053793525757946,
+      "learning_rate": 1.2202177851190912e-06,
+      "loss": 0.4534,
+      "step": 16246
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 28.486410025399437,
+      "learning_rate": 1.2194106298053765e-06,
+      "loss": 0.4737,
+      "step": 16247
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.557755326887538,
+      "learning_rate": 1.2186037242064863e-06,
+      "loss": 0.4554,
+      "step": 16248
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.909335613609805,
+      "learning_rate": 1.21779706834537e-06,
+      "loss": 0.4745,
+      "step": 16249
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.707214554806292,
+      "learning_rate": 1.2169906622449701e-06,
+      "loss": 0.4348,
+      "step": 16250
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 20.95286784700535,
+      "learning_rate": 1.2161845059282174e-06,
+      "loss": 0.494,
+      "step": 16251
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.850565022757994,
+      "learning_rate": 1.2153785994180366e-06,
+      "loss": 0.4144,
+      "step": 16252
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.879570797585448,
+      "learning_rate": 1.2145729427373499e-06,
+      "loss": 0.4222,
+      "step": 16253
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.964010260370724,
+      "learning_rate": 1.2137675359090705e-06,
+      "loss": 0.4409,
+      "step": 16254
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.876305411875467,
+      "learning_rate": 1.212962378956104e-06,
+      "loss": 0.4295,
+      "step": 16255
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.384379962252396,
+      "learning_rate": 1.212157471901344e-06,
+      "loss": 0.4495,
+      "step": 16256
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.731816360848157,
+      "learning_rate": 1.2113528147676855e-06,
+      "loss": 0.3938,
+      "step": 16257
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 39.87602008363829,
+      "learning_rate": 1.2105484075780117e-06,
+      "loss": 0.5317,
+      "step": 16258
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.149658652239655,
+      "learning_rate": 1.2097442503552003e-06,
+      "loss": 0.4414,
+      "step": 16259
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.093697265390233,
+      "learning_rate": 1.2089403431221213e-06,
+      "loss": 0.4942,
+      "step": 16260
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 23.624666810209124,
+      "learning_rate": 1.2081366859016353e-06,
+      "loss": 0.5039,
+      "step": 16261
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.74907721754449,
+      "learning_rate": 1.207333278716597e-06,
+      "loss": 0.47,
+      "step": 16262
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.915220300859985,
+      "learning_rate": 1.2065301215898595e-06,
+      "loss": 0.4499,
+      "step": 16263
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.250262511773297,
+      "learning_rate": 1.20572721454426e-06,
+      "loss": 0.4474,
+      "step": 16264
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.62519518378732,
+      "learning_rate": 1.2049245576026346e-06,
+      "loss": 0.444,
+      "step": 16265
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.28177501991546,
+      "learning_rate": 1.2041221507878087e-06,
+      "loss": 0.4567,
+      "step": 16266
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.670112972037764,
+      "learning_rate": 1.203319994122606e-06,
+      "loss": 0.4854,
+      "step": 16267
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.713256090085531,
+      "learning_rate": 1.2025180876298338e-06,
+      "loss": 0.4029,
+      "step": 16268
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 34.608469280851885,
+      "learning_rate": 1.2017164313323004e-06,
+      "loss": 0.4417,
+      "step": 16269
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.796480723855371,
+      "learning_rate": 1.200915025252808e-06,
+      "loss": 0.371,
+      "step": 16270
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 20.9708927860443,
+      "learning_rate": 1.2001138694141423e-06,
+      "loss": 0.4312,
+      "step": 16271
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.769344572411853,
+      "learning_rate": 1.199312963839092e-06,
+      "loss": 0.4483,
+      "step": 16272
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.40493711783794,
+      "learning_rate": 1.1985123085504312e-06,
+      "loss": 0.4976,
+      "step": 16273
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.644188912463495,
+      "learning_rate": 1.1977119035709329e-06,
+      "loss": 0.4464,
+      "step": 16274
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.74301314517832,
+      "learning_rate": 1.1969117489233595e-06,
+      "loss": 0.4299,
+      "step": 16275
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 20.396701071155285,
+      "learning_rate": 1.1961118446304666e-06,
+      "loss": 0.4506,
+      "step": 16276
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.053488097304097,
+      "learning_rate": 1.195312190715e-06,
+      "loss": 0.4286,
+      "step": 16277
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.954936002378727,
+      "learning_rate": 1.1945127871997042e-06,
+      "loss": 0.4265,
+      "step": 16278
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.260374305349496,
+      "learning_rate": 1.1937136341073163e-06,
+      "loss": 0.3861,
+      "step": 16279
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.288779962533507,
+      "learning_rate": 1.1929147314605617e-06,
+      "loss": 0.4704,
+      "step": 16280
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.91514803069053,
+      "learning_rate": 1.1921160792821572e-06,
+      "loss": 0.4546,
+      "step": 16281
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.176547670587869,
+      "learning_rate": 1.1913176775948187e-06,
+      "loss": 0.4447,
+      "step": 16282
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.289080741813354,
+      "learning_rate": 1.1905195264212532e-06,
+      "loss": 0.4224,
+      "step": 16283
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 24.3583317619512,
+      "learning_rate": 1.1897216257841605e-06,
+      "loss": 0.455,
+      "step": 16284
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 24.582762408631325,
+      "learning_rate": 1.1889239757062309e-06,
+      "loss": 0.435,
+      "step": 16285
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 32.851864820851475,
+      "learning_rate": 1.188126576210148e-06,
+      "loss": 0.4154,
+      "step": 16286
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 20.140068313593815,
+      "learning_rate": 1.1873294273185898e-06,
+      "loss": 0.4222,
+      "step": 16287
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.573977639807598,
+      "learning_rate": 1.1865325290542295e-06,
+      "loss": 0.4762,
+      "step": 16288
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.67043129313243,
+      "learning_rate": 1.185735881439728e-06,
+      "loss": 0.5002,
+      "step": 16289
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.47103234439085,
+      "learning_rate": 1.1849394844977402e-06,
+      "loss": 0.4952,
+      "step": 16290
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.97538011588581,
+      "learning_rate": 1.184143338250917e-06,
+      "loss": 0.4181,
+      "step": 16291
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.072978286731447,
+      "learning_rate": 1.1833474427219015e-06,
+      "loss": 0.4126,
+      "step": 16292
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.14310812340901,
+      "learning_rate": 1.1825517979333256e-06,
+      "loss": 0.4642,
+      "step": 16293
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.274444614032154,
+      "learning_rate": 1.18175640390782e-06,
+      "loss": 0.4687,
+      "step": 16294
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 29.038151591014906,
+      "learning_rate": 1.180961260668002e-06,
+      "loss": 0.4576,
+      "step": 16295
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 25.431581194050647,
+      "learning_rate": 1.1801663682364873e-06,
+      "loss": 0.4689,
+      "step": 16296
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.64219839869337,
+      "learning_rate": 1.179371726635883e-06,
+      "loss": 0.4392,
+      "step": 16297
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.215664515886605,
+      "learning_rate": 1.1785773358887854e-06,
+      "loss": 0.4256,
+      "step": 16298
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 21.038373586691502,
+      "learning_rate": 1.1777831960177898e-06,
+      "loss": 0.4521,
+      "step": 16299
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.7798821311084,
+      "learning_rate": 1.1769893070454774e-06,
+      "loss": 0.4299,
+      "step": 16300
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 15.645600610730098,
+      "learning_rate": 1.1761956689944288e-06,
+      "loss": 0.4111,
+      "step": 16301
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 22.118569230200322,
+      "learning_rate": 1.1754022818872123e-06,
+      "loss": 0.3853,
+      "step": 16302
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 21.478384515151088,
+      "learning_rate": 1.1746091457463927e-06,
+      "loss": 0.5403,
+      "step": 16303
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 26.74090856179101,
+      "learning_rate": 1.173816260594529e-06,
+      "loss": 0.4452,
+      "step": 16304
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.75884874746568,
+      "learning_rate": 1.1730236264541661e-06,
+      "loss": 0.3696,
+      "step": 16305
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.003414411604094,
+      "learning_rate": 1.1722312433478467e-06,
+      "loss": 0.4272,
+      "step": 16306
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.056585662918472,
+      "learning_rate": 1.1714391112981071e-06,
+      "loss": 0.4952,
+      "step": 16307
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.832641771191643,
+      "learning_rate": 1.170647230327473e-06,
+      "loss": 0.4021,
+      "step": 16308
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.17824702966873,
+      "learning_rate": 1.1698556004584728e-06,
+      "loss": 0.4867,
+      "step": 16309
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 25.214565479448012,
+      "learning_rate": 1.1690642217136084e-06,
+      "loss": 0.4566,
+      "step": 16310
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 16.151760605568352,
+      "learning_rate": 1.1682730941153918e-06,
+      "loss": 0.3906,
+      "step": 16311
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.242825475319567,
+      "learning_rate": 1.167482217686322e-06,
+      "loss": 0.4192,
+      "step": 16312
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 28.212979516681244,
+      "learning_rate": 1.1666915924488931e-06,
+      "loss": 0.4565,
+      "step": 16313
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.35953668076768,
+      "learning_rate": 1.165901218425588e-06,
+      "loss": 0.4163,
+      "step": 16314
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 22.42526052880745,
+      "learning_rate": 1.1651110956388822e-06,
+      "loss": 0.4955,
+      "step": 16315
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.786270793508038,
+      "learning_rate": 1.164321224111249e-06,
+      "loss": 0.4752,
+      "step": 16316
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.603192277741512,
+      "learning_rate": 1.1635316038651524e-06,
+      "loss": 0.4186,
+      "step": 16317
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 16.976893413424033,
+      "learning_rate": 1.1627422349230465e-06,
+      "loss": 0.3975,
+      "step": 16318
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.2559601350352265,
+      "learning_rate": 1.161953117307385e-06,
+      "loss": 0.512,
+      "step": 16319
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.43850911033743,
+      "learning_rate": 1.161164251040603e-06,
+      "loss": 0.4227,
+      "step": 16320
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 31.428002993401954,
+      "learning_rate": 1.1603756361451402e-06,
+      "loss": 0.4544,
+      "step": 16321
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.824160259324255,
+      "learning_rate": 1.1595872726434243e-06,
+      "loss": 0.4135,
+      "step": 16322
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.397785517018377,
+      "learning_rate": 1.158799160557874e-06,
+      "loss": 0.4647,
+      "step": 16323
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 26.32288204931034,
+      "learning_rate": 1.158011299910905e-06,
+      "loss": 0.4082,
+      "step": 16324
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.529266931174835,
+      "learning_rate": 1.15722369072492e-06,
+      "loss": 0.4936,
+      "step": 16325
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.12718292792,
+      "learning_rate": 1.1564363330223227e-06,
+      "loss": 0.4463,
+      "step": 16326
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.26092624867748,
+      "learning_rate": 1.1556492268255004e-06,
+      "loss": 0.4761,
+      "step": 16327
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.714218955119456,
+      "learning_rate": 1.1548623721568409e-06,
+      "loss": 0.4228,
+      "step": 16328
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 27.73140780412827,
+      "learning_rate": 1.1540757690387227e-06,
+      "loss": 0.4478,
+      "step": 16329
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.243511439384427,
+      "learning_rate": 1.153289417493515e-06,
+      "loss": 0.4369,
+      "step": 16330
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 31.92530178660795,
+      "learning_rate": 1.1525033175435796e-06,
+      "loss": 0.4552,
+      "step": 16331
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 27.64650107558058,
+      "learning_rate": 1.1517174692112742e-06,
+      "loss": 0.4551,
+      "step": 16332
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.307559452401318,
+      "learning_rate": 1.1509318725189477e-06,
+      "loss": 0.4117,
+      "step": 16333
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.166770202347617,
+      "learning_rate": 1.1501465274889457e-06,
+      "loss": 0.4535,
+      "step": 16334
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.961670130533356,
+      "learning_rate": 1.1493614341435954e-06,
+      "loss": 0.4072,
+      "step": 16335
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.85060751398322,
+      "learning_rate": 1.1485765925052294e-06,
+      "loss": 0.4467,
+      "step": 16336
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.073122976941274,
+      "learning_rate": 1.1477920025961664e-06,
+      "loss": 0.4546,
+      "step": 16337
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.775591125620586,
+      "learning_rate": 1.1470076644387229e-06,
+      "loss": 0.4211,
+      "step": 16338
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.727395348309006,
+      "learning_rate": 1.1462235780552023e-06,
+      "loss": 0.3557,
+      "step": 16339
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.4260286182358,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.3497,
+      "step": 16340
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.70163296055469,
+      "learning_rate": 1.1446561606991158e-06,
+      "loss": 0.4274,
+      "step": 16341
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.421664148602872,
+      "learning_rate": 1.1438728297711288e-06,
+      "loss": 0.3865,
+      "step": 16342
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 22.885467101816943,
+      "learning_rate": 1.14308975070622e-06,
+      "loss": 0.4025,
+      "step": 16343
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 21.706163560990117,
+      "learning_rate": 1.1423069235266538e-06,
+      "loss": 0.396,
+      "step": 16344
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.88431262782147,
+      "learning_rate": 1.1415243482546977e-06,
+      "loss": 0.4234,
+      "step": 16345
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 25.39930066883431,
+      "learning_rate": 1.1407420249126068e-06,
+      "loss": 0.4089,
+      "step": 16346
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.919922317626167,
+      "learning_rate": 1.1399599535226324e-06,
+      "loss": 0.5246,
+      "step": 16347
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.213375302989913,
+      "learning_rate": 1.139178134107014e-06,
+      "loss": 0.4129,
+      "step": 16348
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.411083641510903,
+      "learning_rate": 1.1383965666879847e-06,
+      "loss": 0.4604,
+      "step": 16349
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.891414150234976,
+      "learning_rate": 1.1376152512877725e-06,
+      "loss": 0.4733,
+      "step": 16350
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 28.763862318604023,
+      "learning_rate": 1.1368341879286004e-06,
+      "loss": 0.5224,
+      "step": 16351
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.95521647029882,
+      "learning_rate": 1.1360533766326765e-06,
+      "loss": 0.4916,
+      "step": 16352
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.685922021809194,
+      "learning_rate": 1.1352728174222128e-06,
+      "loss": 0.5037,
+      "step": 16353
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 22.051910685021223,
+      "learning_rate": 1.1344925103194005e-06,
+      "loss": 0.4671,
+      "step": 16354
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.850159651312126,
+      "learning_rate": 1.1337124553464384e-06,
+      "loss": 0.4175,
+      "step": 16355
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 28.218984533203432,
+      "learning_rate": 1.1329326525255046e-06,
+      "loss": 0.4321,
+      "step": 16356
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 27.66665776750027,
+      "learning_rate": 1.1321531018787801e-06,
+      "loss": 0.4387,
+      "step": 16357
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.757202231934077,
+      "learning_rate": 1.131373803428435e-06,
+      "loss": 0.4483,
+      "step": 16358
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.520247251566996,
+      "learning_rate": 1.1305947571966291e-06,
+      "loss": 0.4666,
+      "step": 16359
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.31917600609384,
+      "learning_rate": 1.1298159632055228e-06,
+      "loss": 0.403,
+      "step": 16360
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 32.47964946381058,
+      "learning_rate": 1.1290374214772582e-06,
+      "loss": 0.4483,
+      "step": 16361
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.91319271816273,
+      "learning_rate": 1.1282591320339809e-06,
+      "loss": 0.4727,
+      "step": 16362
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.226594818197338,
+      "learning_rate": 1.1274810948978255e-06,
+      "loss": 0.4471,
+      "step": 16363
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 24.34232487993892,
+      "learning_rate": 1.1267033100909174e-06,
+      "loss": 0.3947,
+      "step": 16364
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.37973193621713,
+      "learning_rate": 1.125925777635375e-06,
+      "loss": 0.4139,
+      "step": 16365
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 27.15655636806649,
+      "learning_rate": 1.1251484975533123e-06,
+      "loss": 0.4633,
+      "step": 16366
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.152954780469305,
+      "learning_rate": 1.1243714698668363e-06,
+      "loss": 0.4225,
+      "step": 16367
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 31.86852340582149,
+      "learning_rate": 1.1235946945980435e-06,
+      "loss": 0.4758,
+      "step": 16368
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 28.23079905957375,
+      "learning_rate": 1.1228181717690234e-06,
+      "loss": 0.4735,
+      "step": 16369
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.843039559854052,
+      "learning_rate": 1.1220419014018613e-06,
+      "loss": 0.4771,
+      "step": 16370
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 29.149476975734725,
+      "learning_rate": 1.121265883518634e-06,
+      "loss": 0.4881,
+      "step": 16371
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.689965827791,
+      "learning_rate": 1.1204901181414141e-06,
+      "loss": 0.3978,
+      "step": 16372
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.492457200085124,
+      "learning_rate": 1.1197146052922592e-06,
+      "loss": 0.4384,
+      "step": 16373
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.207400283378668,
+      "learning_rate": 1.118939344993225e-06,
+      "loss": 0.4618,
+      "step": 16374
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.166570634928924,
+      "learning_rate": 1.1181643372663608e-06,
+      "loss": 0.448,
+      "step": 16375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.849882378177295,
+      "learning_rate": 1.1173895821337088e-06,
+      "loss": 0.4165,
+      "step": 16376
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.187464382828306,
+      "learning_rate": 1.1166150796172981e-06,
+      "loss": 0.3819,
+      "step": 16377
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 28.87890983405053,
+      "learning_rate": 1.115840829739161e-06,
+      "loss": 0.3964,
+      "step": 16378
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.05034665574483,
+      "learning_rate": 1.115066832521311e-06,
+      "loss": 0.4707,
+      "step": 16379
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.667140587899357,
+      "learning_rate": 1.114293087985766e-06,
+      "loss": 0.4569,
+      "step": 16380
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 24.16680677033514,
+      "learning_rate": 1.1135195961545242e-06,
+      "loss": 0.5102,
+      "step": 16381
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.046267554607923,
+      "learning_rate": 1.1127463570495867e-06,
+      "loss": 0.4581,
+      "step": 16382
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.73502799543351,
+      "learning_rate": 1.111973370692947e-06,
+      "loss": 0.376,
+      "step": 16383
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 29.466802734973157,
+      "learning_rate": 1.111200637106582e-06,
+      "loss": 0.45,
+      "step": 16384
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.16809120994578,
+      "learning_rate": 1.1104281563124741e-06,
+      "loss": 0.4051,
+      "step": 16385
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 30.422738142851834,
+      "learning_rate": 1.109655928332587e-06,
+      "loss": 0.531,
+      "step": 16386
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.69068650540431,
+      "learning_rate": 1.1088839531888862e-06,
+      "loss": 0.4703,
+      "step": 16387
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.187114970826515,
+      "learning_rate": 1.108112230903322e-06,
+      "loss": 0.4329,
+      "step": 16388
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.41895960636258,
+      "learning_rate": 1.1073407614978471e-06,
+      "loss": 0.3945,
+      "step": 16389
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.135008072362435,
+      "learning_rate": 1.106569544994397e-06,
+      "loss": 0.3688,
+      "step": 16390
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.880747776876863,
+      "learning_rate": 1.1057985814149063e-06,
+      "loss": 0.3895,
+      "step": 16391
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.17307082745536,
+      "learning_rate": 1.1050278707813033e-06,
+      "loss": 0.4243,
+      "step": 16392
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 24.12439138133963,
+      "learning_rate": 1.1042574131155048e-06,
+      "loss": 0.4753,
+      "step": 16393
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.372917637979565,
+      "learning_rate": 1.1034872084394187e-06,
+      "loss": 0.3861,
+      "step": 16394
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 31.82668770310337,
+      "learning_rate": 1.1027172567749523e-06,
+      "loss": 0.47,
+      "step": 16395
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 24.784327551633144,
+      "learning_rate": 1.101947558144002e-06,
+      "loss": 0.4832,
+      "step": 16396
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 23.05370878380459,
+      "learning_rate": 1.1011781125684618e-06,
+      "loss": 0.43,
+      "step": 16397
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 23.402403041692537,
+      "learning_rate": 1.1004089200702072e-06,
+      "loss": 0.4124,
+      "step": 16398
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.81428395768574,
+      "learning_rate": 1.0996399806711167e-06,
+      "loss": 0.4103,
+      "step": 16399
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.168087967249285,
+      "learning_rate": 1.098871294393058e-06,
+      "loss": 0.4625,
+      "step": 16400
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.887338680312844,
+      "learning_rate": 1.0981028612578949e-06,
+      "loss": 0.478,
+      "step": 16401
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.29427435026181,
+      "learning_rate": 1.0973346812874796e-06,
+      "loss": 0.4044,
+      "step": 16402
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 28.09919110479566,
+      "learning_rate": 1.0965667545036552e-06,
+      "loss": 0.5247,
+      "step": 16403
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.68692232232192,
+      "learning_rate": 1.0957990809282649e-06,
+      "loss": 0.4263,
+      "step": 16404
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.735905248330916,
+      "learning_rate": 1.0950316605831413e-06,
+      "loss": 0.5449,
+      "step": 16405
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.55569018693021,
+      "learning_rate": 1.0942644934901059e-06,
+      "loss": 0.4351,
+      "step": 16406
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 24.362020268177055,
+      "learning_rate": 1.0934975796709801e-06,
+      "loss": 0.4947,
+      "step": 16407
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 24.814027515230823,
+      "learning_rate": 1.0927309191475722e-06,
+      "loss": 0.4882,
+      "step": 16408
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.472303019451868,
+      "learning_rate": 1.0919645119416855e-06,
+      "loss": 0.4447,
+      "step": 16409
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 32.39421401356696,
+      "learning_rate": 1.0911983580751195e-06,
+      "loss": 0.5532,
+      "step": 16410
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.9577344643695,
+      "learning_rate": 1.090432457569659e-06,
+      "loss": 0.456,
+      "step": 16411
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.53984790669875,
+      "learning_rate": 1.0896668104470886e-06,
+      "loss": 0.419,
+      "step": 16412
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.15302915325032,
+      "learning_rate": 1.08890141672918e-06,
+      "loss": 0.4957,
+      "step": 16413
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.084723896009816,
+      "learning_rate": 1.0881362764377046e-06,
+      "loss": 0.4884,
+      "step": 16414
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.732870058300346,
+      "learning_rate": 1.087371389594417e-06,
+      "loss": 0.3784,
+      "step": 16415
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.86533736246151,
+      "learning_rate": 1.0866067562210748e-06,
+      "loss": 0.4726,
+      "step": 16416
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.70157432773767,
+      "learning_rate": 1.0858423763394243e-06,
+      "loss": 0.4598,
+      "step": 16417
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.265653735508337,
+      "learning_rate": 1.085078249971201e-06,
+      "loss": 0.4704,
+      "step": 16418
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.638974170667638,
+      "learning_rate": 1.084314377138136e-06,
+      "loss": 0.459,
+      "step": 16419
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 27.043482959470058,
+      "learning_rate": 1.0835507578619542e-06,
+      "loss": 0.4217,
+      "step": 16420
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 30.753017869132258,
+      "learning_rate": 1.0827873921643727e-06,
+      "loss": 0.4773,
+      "step": 16421
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 20.176999774484784,
+      "learning_rate": 1.0820242800671032e-06,
+      "loss": 0.433,
+      "step": 16422
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.42590269907103,
+      "learning_rate": 1.0812614215918472e-06,
+      "loss": 0.4405,
+      "step": 16423
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.14324710765302,
+      "learning_rate": 1.080498816760296e-06,
+      "loss": 0.4595,
+      "step": 16424
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 22.12809486888121,
+      "learning_rate": 1.0797364655941411e-06,
+      "loss": 0.4784,
+      "step": 16425
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.407784876578418,
+      "learning_rate": 1.0789743681150656e-06,
+      "loss": 0.476,
+      "step": 16426
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.517716857248157,
+      "learning_rate": 1.0782125243447395e-06,
+      "loss": 0.4441,
+      "step": 16427
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.70820166669645,
+      "learning_rate": 1.077450934304829e-06,
+      "loss": 0.4642,
+      "step": 16428
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.020234554545574,
+      "learning_rate": 1.0766895980169933e-06,
+      "loss": 0.3908,
+      "step": 16429
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 29.052248154618663,
+      "learning_rate": 1.0759285155028887e-06,
+      "loss": 0.4746,
+      "step": 16430
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 22.39138326472355,
+      "learning_rate": 1.0751676867841553e-06,
+      "loss": 0.4037,
+      "step": 16431
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 26.086537281988726,
+      "learning_rate": 1.0744071118824306e-06,
+      "loss": 0.459,
+      "step": 16432
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.999004045840632,
+      "learning_rate": 1.0736467908193471e-06,
+      "loss": 0.4116,
+      "step": 16433
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 20.219300575071028,
+      "learning_rate": 1.0728867236165264e-06,
+      "loss": 0.5159,
+      "step": 16434
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 25.054370111489764,
+      "learning_rate": 1.0721269102955866e-06,
+      "loss": 0.4311,
+      "step": 16435
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 29.298196747019396,
+      "learning_rate": 1.0713673508781353e-06,
+      "loss": 0.4409,
+      "step": 16436
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 30.486264198143818,
+      "learning_rate": 1.0706080453857714e-06,
+      "loss": 0.4789,
+      "step": 16437
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 24.735902269193858,
+      "learning_rate": 1.0698489938400914e-06,
+      "loss": 0.5981,
+      "step": 16438
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.34813898558111,
+      "learning_rate": 1.0690901962626843e-06,
+      "loss": 0.4232,
+      "step": 16439
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 29.209659925968698,
+      "learning_rate": 1.0683316526751253e-06,
+      "loss": 0.4309,
+      "step": 16440
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 24.62095876998977,
+      "learning_rate": 1.0675733630989904e-06,
+      "loss": 0.4082,
+      "step": 16441
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.230446956732127,
+      "learning_rate": 1.0668153275558424e-06,
+      "loss": 0.4662,
+      "step": 16442
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.593381128739416,
+      "learning_rate": 1.066057546067243e-06,
+      "loss": 0.383,
+      "step": 16443
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.658941749451742,
+      "learning_rate": 1.0653000186547379e-06,
+      "loss": 0.3876,
+      "step": 16444
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.735885907644683,
+      "learning_rate": 1.0645427453398748e-06,
+      "loss": 0.4485,
+      "step": 16445
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 20.46667836858727,
+      "learning_rate": 1.0637857261441898e-06,
+      "loss": 0.4401,
+      "step": 16446
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.609907039136365,
+      "learning_rate": 1.0630289610892097e-06,
+      "loss": 0.4144,
+      "step": 16447
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 28.078244097058516,
+      "learning_rate": 1.06227245019646e-06,
+      "loss": 0.5035,
+      "step": 16448
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.47437042310223,
+      "learning_rate": 1.061516193487452e-06,
+      "loss": 0.4938,
+      "step": 16449
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.657574786604457,
+      "learning_rate": 1.060760190983694e-06,
+      "loss": 0.4026,
+      "step": 16450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.167211752626248,
+      "learning_rate": 1.060004442706688e-06,
+      "loss": 0.4985,
+      "step": 16451
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.32445025377652,
+      "learning_rate": 1.0592489486779267e-06,
+      "loss": 0.5047,
+      "step": 16452
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.524915949851813,
+      "learning_rate": 1.0584937089188941e-06,
+      "loss": 0.4068,
+      "step": 16453
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.75082880311266,
+      "learning_rate": 1.0577387234510684e-06,
+      "loss": 0.4077,
+      "step": 16454
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.761977327557,
+      "learning_rate": 1.0569839922959247e-06,
+      "loss": 0.4417,
+      "step": 16455
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 26.623101283704802,
+      "learning_rate": 1.0562295154749248e-06,
+      "loss": 0.4138,
+      "step": 16456
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.67506943946318,
+      "learning_rate": 1.0554752930095236e-06,
+      "loss": 0.4246,
+      "step": 16457
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.054041908445058,
+      "learning_rate": 1.054721324921173e-06,
+      "loss": 0.4126,
+      "step": 16458
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.155209200175637,
+      "learning_rate": 1.0539676112313147e-06,
+      "loss": 0.4115,
+      "step": 16459
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.782498422136245,
+      "learning_rate": 1.053214151961386e-06,
+      "loss": 0.4144,
+      "step": 16460
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.62627294384664,
+      "learning_rate": 1.052460947132814e-06,
+      "loss": 0.4104,
+      "step": 16461
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.60887920028195,
+      "learning_rate": 1.0517079967670152e-06,
+      "loss": 0.433,
+      "step": 16462
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 20.079174148297014,
+      "learning_rate": 1.0509553008854067e-06,
+      "loss": 0.4878,
+      "step": 16463
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.125783898461613,
+      "learning_rate": 1.0502028595093972e-06,
+      "loss": 0.4447,
+      "step": 16464
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.811373283223823,
+      "learning_rate": 1.0494506726603804e-06,
+      "loss": 0.4465,
+      "step": 16465
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.78558384809646,
+      "learning_rate": 1.0486987403597526e-06,
+      "loss": 0.4971,
+      "step": 16466
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.334436267807632,
+      "learning_rate": 1.0479470626288946e-06,
+      "loss": 0.4017,
+      "step": 16467
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.196108079437995,
+      "learning_rate": 1.0471956394891868e-06,
+      "loss": 0.4238,
+      "step": 16468
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 20.874548661612288,
+      "learning_rate": 1.0464444709619959e-06,
+      "loss": 0.432,
+      "step": 16469
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 22.229850659234515,
+      "learning_rate": 1.0456935570686866e-06,
+      "loss": 0.4903,
+      "step": 16470
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.426667251866668,
+      "learning_rate": 1.0449428978306164e-06,
+      "loss": 0.4599,
+      "step": 16471
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.81085391456848,
+      "learning_rate": 1.0441924932691293e-06,
+      "loss": 0.4118,
+      "step": 16472
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.729226681851035,
+      "learning_rate": 1.0434423434055719e-06,
+      "loss": 0.4039,
+      "step": 16473
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.3170232443151,
+      "learning_rate": 1.042692448261272e-06,
+      "loss": 0.4412,
+      "step": 16474
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 25.12976591168797,
+      "learning_rate": 1.041942807857559e-06,
+      "loss": 0.4329,
+      "step": 16475
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 23.713110554898023,
+      "learning_rate": 1.0411934222157538e-06,
+      "loss": 0.4795,
+      "step": 16476
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 25.977543597516572,
+      "learning_rate": 1.0404442913571678e-06,
+      "loss": 0.4456,
+      "step": 16477
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 22.726915575093056,
+      "learning_rate": 1.0396954153031024e-06,
+      "loss": 0.4091,
+      "step": 16478
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.9383974673595,
+      "learning_rate": 1.0389467940748576e-06,
+      "loss": 0.4055,
+      "step": 16479
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.640588296215252,
+      "learning_rate": 1.0381984276937263e-06,
+      "loss": 0.4168,
+      "step": 16480
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 25.45203231458933,
+      "learning_rate": 1.037450316180989e-06,
+      "loss": 0.4831,
+      "step": 16481
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.595715689208223,
+      "learning_rate": 1.03670245955792e-06,
+      "loss": 0.4195,
+      "step": 16482
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.879232477025074,
+      "learning_rate": 1.03595485784579e-06,
+      "loss": 0.4238,
+      "step": 16483
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.84826947146234,
+      "learning_rate": 1.0352075110658588e-06,
+      "loss": 0.4502,
+      "step": 16484
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.93688077949329,
+      "learning_rate": 1.034460419239387e-06,
+      "loss": 0.4514,
+      "step": 16485
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 25.199733733303642,
+      "learning_rate": 1.033713582387611e-06,
+      "loss": 0.4522,
+      "step": 16486
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 15.935862619774106,
+      "learning_rate": 1.032967000531777e-06,
+      "loss": 0.372,
+      "step": 16487
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.415324910507577,
+      "learning_rate": 1.0322206736931152e-06,
+      "loss": 0.404,
+      "step": 16488
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.555851951187535,
+      "learning_rate": 1.0314746018928535e-06,
+      "loss": 0.3731,
+      "step": 16489
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.2027583405399,
+      "learning_rate": 1.0307287851522074e-06,
+      "loss": 0.4799,
+      "step": 16490
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 32.87259338276433,
+      "learning_rate": 1.0299832234923857e-06,
+      "loss": 0.485,
+      "step": 16491
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.95576761204422,
+      "learning_rate": 1.0292379169345945e-06,
+      "loss": 0.495,
+      "step": 16492
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 24.967863538615823,
+      "learning_rate": 1.0284928655000303e-06,
+      "loss": 0.4787,
+      "step": 16493
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.18830731282949,
+      "learning_rate": 1.0277480692098796e-06,
+      "loss": 0.4514,
+      "step": 16494
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 28.02049827103553,
+      "learning_rate": 1.0270035280853275e-06,
+      "loss": 0.4805,
+      "step": 16495
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.312320815030127,
+      "learning_rate": 1.0262592421475436e-06,
+      "loss": 0.4213,
+      "step": 16496
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.471464950329693,
+      "learning_rate": 1.025515211417697e-06,
+      "loss": 0.3936,
+      "step": 16497
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 22.155199736894907,
+      "learning_rate": 1.0247714359169502e-06,
+      "loss": 0.4295,
+      "step": 16498
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 25.742556656432157,
+      "learning_rate": 1.0240279156664512e-06,
+      "loss": 0.4644,
+      "step": 16499
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 29.23617973343494,
+      "learning_rate": 1.0232846506873495e-06,
+      "loss": 0.3922,
+      "step": 16500
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.179230597439556,
+      "learning_rate": 1.0225416410007794e-06,
+      "loss": 0.4816,
+      "step": 16501
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.564784694933927,
+      "learning_rate": 1.021798886627875e-06,
+      "loss": 0.4516,
+      "step": 16502
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 15.216237414953463,
+      "learning_rate": 1.0210563875897561e-06,
+      "loss": 0.386,
+      "step": 16503
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 15.690356470878475,
+      "learning_rate": 1.0203141439075415e-06,
+      "loss": 0.4463,
+      "step": 16504
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 26.02801413877146,
+      "learning_rate": 1.0195721556023409e-06,
+      "loss": 0.4337,
+      "step": 16505
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 30.748108313299866,
+      "learning_rate": 1.0188304226952562e-06,
+      "loss": 0.4857,
+      "step": 16506
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.634515379408477,
+      "learning_rate": 1.0180889452073772e-06,
+      "loss": 0.4818,
+      "step": 16507
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.562205802827227,
+      "learning_rate": 1.017347723159795e-06,
+      "loss": 0.3984,
+      "step": 16508
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 22.995305402438024,
+      "learning_rate": 1.0166067565735881e-06,
+      "loss": 0.3964,
+      "step": 16509
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.219990642106335,
+      "learning_rate": 1.015866045469832e-06,
+      "loss": 0.4352,
+      "step": 16510
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.647476794914745,
+      "learning_rate": 1.0151255898695911e-06,
+      "loss": 0.4233,
+      "step": 16511
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 30.48425150796893,
+      "learning_rate": 1.0143853897939193e-06,
+      "loss": 0.4566,
+      "step": 16512
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.278968117401885,
+      "learning_rate": 1.013645445263871e-06,
+      "loss": 0.5062,
+      "step": 16513
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.820874956646325,
+      "learning_rate": 1.012905756300492e-06,
+      "loss": 0.4455,
+      "step": 16514
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.608701171827537,
+      "learning_rate": 1.0121663229248145e-06,
+      "loss": 0.4525,
+      "step": 16515
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 24.581410107235467,
+      "learning_rate": 1.0114271451578684e-06,
+      "loss": 0.3984,
+      "step": 16516
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 31.58511180775014,
+      "learning_rate": 1.0106882230206749e-06,
+      "loss": 0.5215,
+      "step": 16517
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 33.65416585690738,
+      "learning_rate": 1.0099495565342532e-06,
+      "loss": 0.4343,
+      "step": 16518
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.51411665074277,
+      "learning_rate": 1.0092111457196041e-06,
+      "loss": 0.4125,
+      "step": 16519
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.70205503767291,
+      "learning_rate": 1.0084729905977332e-06,
+      "loss": 0.4557,
+      "step": 16520
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.814874729169205,
+      "learning_rate": 1.0077350911896278e-06,
+      "loss": 0.4395,
+      "step": 16521
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 25.97192915817965,
+      "learning_rate": 1.006997447516276e-06,
+      "loss": 0.4404,
+      "step": 16522
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.230180421208296,
+      "learning_rate": 1.0062600595986582e-06,
+      "loss": 0.4285,
+      "step": 16523
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.677279918512994,
+      "learning_rate": 1.0055229274577417e-06,
+      "loss": 0.4765,
+      "step": 16524
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.578095830390765,
+      "learning_rate": 1.0047860511144937e-06,
+      "loss": 0.4114,
+      "step": 16525
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.861821633638986,
+      "learning_rate": 1.004049430589865e-06,
+      "loss": 0.507,
+      "step": 16526
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.770514644063674,
+      "learning_rate": 1.0033130659048119e-06,
+      "loss": 0.3952,
+      "step": 16527
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.791023707120715,
+      "learning_rate": 1.002576957080269e-06,
+      "loss": 0.4247,
+      "step": 16528
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.604076591004727,
+      "learning_rate": 1.0018411041371756e-06,
+      "loss": 0.4734,
+      "step": 16529
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 28.33773838349109,
+      "learning_rate": 1.001105507096457e-06,
+      "loss": 0.4376,
+      "step": 16530
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.510467834047187,
+      "learning_rate": 1.0003701659790344e-06,
+      "loss": 0.4145,
+      "step": 16531
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 22.457611751966294,
+      "learning_rate": 9.996350808058175e-07,
+      "loss": 0.3987,
+      "step": 16532
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 28.852800145939398,
+      "learning_rate": 9.989002515977154e-07,
+      "loss": 0.4239,
+      "step": 16533
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.30892513875,
+      "learning_rate": 9.981656783756255e-07,
+      "loss": 0.4467,
+      "step": 16534
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.044757461402007,
+      "learning_rate": 9.974313611604358e-07,
+      "loss": 0.4423,
+      "step": 16535
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.533637408461708,
+      "learning_rate": 9.96697299973034e-07,
+      "loss": 0.4163,
+      "step": 16536
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.661578496027914,
+      "learning_rate": 9.959634948342923e-07,
+      "loss": 0.4391,
+      "step": 16537
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.70728507022078,
+      "learning_rate": 9.952299457650805e-07,
+      "loss": 0.4423,
+      "step": 16538
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.358198417946898,
+      "learning_rate": 9.944966527862644e-07,
+      "loss": 0.432,
+      "step": 16539
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.589319925929807,
+      "learning_rate": 9.93763615918696e-07,
+      "loss": 0.4515,
+      "step": 16540
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.80563146201027,
+      "learning_rate": 9.930308351832185e-07,
+      "loss": 0.4338,
+      "step": 16541
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.925555980125676,
+      "learning_rate": 9.922983106006766e-07,
+      "loss": 0.4576,
+      "step": 16542
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.65480177265285,
+      "learning_rate": 9.915660421919027e-07,
+      "loss": 0.4294,
+      "step": 16543
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.348788224983505,
+      "learning_rate": 9.908340299777208e-07,
+      "loss": 0.4625,
+      "step": 16544
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.468527006326607,
+      "learning_rate": 9.901022739789468e-07,
+      "loss": 0.4469,
+      "step": 16545
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 24.437181231061157,
+      "learning_rate": 9.893707742163926e-07,
+      "loss": 0.4391,
+      "step": 16546
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.26344859691854,
+      "learning_rate": 9.886395307108643e-07,
+      "loss": 0.3868,
+      "step": 16547
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.340548640366055,
+      "learning_rate": 9.87908543483157e-07,
+      "loss": 0.425,
+      "step": 16548
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.25846971871695,
+      "learning_rate": 9.871778125540587e-07,
+      "loss": 0.4199,
+      "step": 16549
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.34363216250013,
+      "learning_rate": 9.864473379443495e-07,
+      "loss": 0.4126,
+      "step": 16550
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 22.553022229353086,
+      "learning_rate": 9.85717119674806e-07,
+      "loss": 0.4906,
+      "step": 16551
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 28.276992042104844,
+      "learning_rate": 9.84987157766195e-07,
+      "loss": 0.4179,
+      "step": 16552
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 23.807934996806832,
+      "learning_rate": 9.842574522392744e-07,
+      "loss": 0.4719,
+      "step": 16553
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.923307514886028,
+      "learning_rate": 9.835280031147999e-07,
+      "loss": 0.3939,
+      "step": 16554
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.91793874800877,
+      "learning_rate": 9.827988104135122e-07,
+      "loss": 0.4043,
+      "step": 16555
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.212522157107077,
+      "learning_rate": 9.82069874156154e-07,
+      "loss": 0.4205,
+      "step": 16556
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 29.75233433735346,
+      "learning_rate": 9.81341194363451e-07,
+      "loss": 0.4767,
+      "step": 16557
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 24.046657248955565,
+      "learning_rate": 9.806127710561274e-07,
+      "loss": 0.459,
+      "step": 16558
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 22.186223392967094,
+      "learning_rate": 9.798846042549037e-07,
+      "loss": 0.419,
+      "step": 16559
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.39623698390906,
+      "learning_rate": 9.79156693980483e-07,
+      "loss": 0.4234,
+      "step": 16560
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 24.044912794215403,
+      "learning_rate": 9.784290402535713e-07,
+      "loss": 0.4487,
+      "step": 16561
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.668414027335913,
+      "learning_rate": 9.777016430948572e-07,
+      "loss": 0.4626,
+      "step": 16562
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 29.21101457331106,
+      "learning_rate": 9.76974502525031e-07,
+      "loss": 0.4468,
+      "step": 16563
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 16.63610310433114,
+      "learning_rate": 9.76247618564774e-07,
+      "loss": 0.4276,
+      "step": 16564
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 16.28588605642223,
+      "learning_rate": 9.755209912347552e-07,
+      "loss": 0.4021,
+      "step": 16565
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 21.22000810491783,
+      "learning_rate": 9.747946205556391e-07,
+      "loss": 0.4279,
+      "step": 16566
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 38.557370057164384,
+      "learning_rate": 9.740685065480837e-07,
+      "loss": 0.4808,
+      "step": 16567
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.66066391805108,
+      "learning_rate": 9.733426492327425e-07,
+      "loss": 0.4066,
+      "step": 16568
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.276100867352753,
+      "learning_rate": 9.726170486302554e-07,
+      "loss": 0.4234,
+      "step": 16569
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 22.07153241997382,
+      "learning_rate": 9.718917047612575e-07,
+      "loss": 0.4925,
+      "step": 16570
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 21.499364029133467,
+      "learning_rate": 9.711666176463775e-07,
+      "loss": 0.4306,
+      "step": 16571
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 30.003989369660047,
+      "learning_rate": 9.704417873062366e-07,
+      "loss": 0.4829,
+      "step": 16572
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.47555246960147,
+      "learning_rate": 9.697172137614518e-07,
+      "loss": 0.4287,
+      "step": 16573
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 24.756652362469854,
+      "learning_rate": 9.689928970326268e-07,
+      "loss": 0.4669,
+      "step": 16574
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.175715192197819,
+      "learning_rate": 9.682688371403593e-07,
+      "loss": 0.4344,
+      "step": 16575
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.885279892447244,
+      "learning_rate": 9.675450341052427e-07,
+      "loss": 0.4255,
+      "step": 16576
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 23.888077227686253,
+      "learning_rate": 9.668214879478622e-07,
+      "loss": 0.5191,
+      "step": 16577
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 29.036166649908623,
+      "learning_rate": 9.66098198688795e-07,
+      "loss": 0.3896,
+      "step": 16578
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 25.648444942899722,
+      "learning_rate": 9.65375166348609e-07,
+      "loss": 0.4077,
+      "step": 16579
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 32.028053514972335,
+      "learning_rate": 9.646523909478677e-07,
+      "loss": 0.4307,
+      "step": 16580
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 21.693729862502174,
+      "learning_rate": 9.63929872507129e-07,
+      "loss": 0.4196,
+      "step": 16581
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 24.124346025032906,
+      "learning_rate": 9.632076110469368e-07,
+      "loss": 0.4165,
+      "step": 16582
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.81676246507038,
+      "learning_rate": 9.624856065878351e-07,
+      "loss": 0.4586,
+      "step": 16583
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 28.342226783993983,
+      "learning_rate": 9.617638591503542e-07,
+      "loss": 0.4957,
+      "step": 16584
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.93790422533783,
+      "learning_rate": 9.61042368755023e-07,
+      "loss": 0.4046,
+      "step": 16585
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.6169104061244,
+      "learning_rate": 9.603211354223597e-07,
+      "loss": 0.385,
+      "step": 16586
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 21.741579402639957,
+      "learning_rate": 9.596001591728743e-07,
+      "loss": 0.4956,
+      "step": 16587
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 30.306177006793096,
+      "learning_rate": 9.58879440027074e-07,
+      "loss": 0.4563,
+      "step": 16588
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.765973896916027,
+      "learning_rate": 9.5815897800545e-07,
+      "loss": 0.3972,
+      "step": 16589
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 23.11424219462224,
+      "learning_rate": 9.574387731284984e-07,
+      "loss": 0.4215,
+      "step": 16590
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.610770729820203,
+      "learning_rate": 9.56718825416697e-07,
+      "loss": 0.3933,
+      "step": 16591
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 33.816859833278606,
+      "learning_rate": 9.559991348905218e-07,
+      "loss": 0.468,
+      "step": 16592
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.852515207095763,
+      "learning_rate": 9.55279701570442e-07,
+      "loss": 0.4243,
+      "step": 16593
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 29.141677976509175,
+      "learning_rate": 9.545605254769164e-07,
+      "loss": 0.526,
+      "step": 16594
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.216068097417466,
+      "learning_rate": 9.538416066303956e-07,
+      "loss": 0.4218,
+      "step": 16595
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.26951078096083,
+      "learning_rate": 9.531229450513291e-07,
+      "loss": 0.4137,
+      "step": 16596
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.919263688774954,
+      "learning_rate": 9.524045407601534e-07,
+      "loss": 0.4682,
+      "step": 16597
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 27.113522541581116,
+      "learning_rate": 9.516863937773013e-07,
+      "loss": 0.4943,
+      "step": 16598
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 25.228212863150578,
+      "learning_rate": 9.509685041231953e-07,
+      "loss": 0.4431,
+      "step": 16599
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.440371437438177,
+      "learning_rate": 9.502508718182491e-07,
+      "loss": 0.4354,
+      "step": 16600
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 27.165188017538007,
+      "learning_rate": 9.49533496882874e-07,
+      "loss": 0.5733,
+      "step": 16601
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 26.408273239894697,
+      "learning_rate": 9.488163793374749e-07,
+      "loss": 0.4491,
+      "step": 16602
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.26673032118206,
+      "learning_rate": 9.480995192024423e-07,
+      "loss": 0.4764,
+      "step": 16603
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.712476225159275,
+      "learning_rate": 9.473829164981629e-07,
+      "loss": 0.4284,
+      "step": 16604
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.149277999189202,
+      "learning_rate": 9.466665712450174e-07,
+      "loss": 0.3953,
+      "step": 16605
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.17356222773973,
+      "learning_rate": 9.459504834633804e-07,
+      "loss": 0.3883,
+      "step": 16606
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 43.14834145489166,
+      "learning_rate": 9.452346531736134e-07,
+      "loss": 0.4145,
+      "step": 16607
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 23.048702268416047,
+      "learning_rate": 9.44519080396078e-07,
+      "loss": 0.4732,
+      "step": 16608
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.85824593299342,
+      "learning_rate": 9.438037651511201e-07,
+      "loss": 0.4093,
+      "step": 16609
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 25.90883123655177,
+      "learning_rate": 9.430887074590855e-07,
+      "loss": 0.4416,
+      "step": 16610
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.54133176457306,
+      "learning_rate": 9.423739073403126e-07,
+      "loss": 0.4391,
+      "step": 16611
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.732579868080226,
+      "learning_rate": 9.416593648151251e-07,
+      "loss": 0.4338,
+      "step": 16612
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.647606183870503,
+      "learning_rate": 9.409450799038478e-07,
+      "loss": 0.3797,
+      "step": 16613
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 30.206813752770014,
+      "learning_rate": 9.40231052626791e-07,
+      "loss": 0.4625,
+      "step": 16614
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 32.31348099962588,
+      "learning_rate": 9.395172830042653e-07,
+      "loss": 0.5482,
+      "step": 16615
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.6672429337497,
+      "learning_rate": 9.388037710565667e-07,
+      "loss": 0.5197,
+      "step": 16616
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.326256056476627,
+      "learning_rate": 9.380905168039878e-07,
+      "loss": 0.4341,
+      "step": 16617
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 23.368618083106085,
+      "learning_rate": 9.373775202668156e-07,
+      "loss": 0.4431,
+      "step": 16618
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 30.806549561635077,
+      "learning_rate": 9.36664781465324e-07,
+      "loss": 0.6125,
+      "step": 16619
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.9371326523099,
+      "learning_rate": 9.359523004197835e-07,
+      "loss": 0.414,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.144667803148963,
+      "learning_rate": 9.352400771504566e-07,
+      "loss": 0.3924,
+      "step": 16621
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.327806282623662,
+      "learning_rate": 9.345281116775995e-07,
+      "loss": 0.4246,
+      "step": 16622
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.115795528223863,
+      "learning_rate": 9.338164040214614e-07,
+      "loss": 0.4323,
+      "step": 16623
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.947981052218893,
+      "learning_rate": 9.331049542022818e-07,
+      "loss": 0.4235,
+      "step": 16624
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.063613919127054,
+      "learning_rate": 9.323937622402912e-07,
+      "loss": 0.4602,
+      "step": 16625
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 28.645485867676015,
+      "learning_rate": 9.316828281557178e-07,
+      "loss": 0.4192,
+      "step": 16626
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.99592090255853,
+      "learning_rate": 9.309721519687809e-07,
+      "loss": 0.4456,
+      "step": 16627
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.749788272377494,
+      "learning_rate": 9.30261733699691e-07,
+      "loss": 0.4229,
+      "step": 16628
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.854005343297715,
+      "learning_rate": 9.295515733686511e-07,
+      "loss": 0.456,
+      "step": 16629
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.565194515023208,
+      "learning_rate": 9.28841670995857e-07,
+      "loss": 0.451,
+      "step": 16630
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.278853611314297,
+      "learning_rate": 9.281320266015015e-07,
+      "loss": 0.4267,
+      "step": 16631
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 26.674816431988763,
+      "learning_rate": 9.274226402057651e-07,
+      "loss": 0.4463,
+      "step": 16632
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.554797704469767,
+      "learning_rate": 9.267135118288184e-07,
+      "loss": 0.3773,
+      "step": 16633
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.98594642991937,
+      "learning_rate": 9.26004641490833e-07,
+      "loss": 0.4135,
+      "step": 16634
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.150973173064106,
+      "learning_rate": 9.252960292119661e-07,
+      "loss": 0.4367,
+      "step": 16635
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 25.336592091477296,
+      "learning_rate": 9.24587675012375e-07,
+      "loss": 0.4626,
+      "step": 16636
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 29.170722339989087,
+      "learning_rate": 9.238795789122002e-07,
+      "loss": 0.4563,
+      "step": 16637
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.588816872444657,
+      "learning_rate": 9.231717409315788e-07,
+      "loss": 0.4561,
+      "step": 16638
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.91705659845594,
+      "learning_rate": 9.224641610906437e-07,
+      "loss": 0.4929,
+      "step": 16639
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 19.008658256639237,
+      "learning_rate": 9.2175683940952e-07,
+      "loss": 0.4021,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 25.727204710112144,
+      "learning_rate": 9.210497759083193e-07,
+      "loss": 0.4594,
+      "step": 16641
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 35.764165740201804,
+      "learning_rate": 9.203429706071531e-07,
+      "loss": 0.5151,
+      "step": 16642
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.089536312398872,
+      "learning_rate": 9.196364235261202e-07,
+      "loss": 0.4105,
+      "step": 16643
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.135423963964545,
+      "learning_rate": 9.189301346853163e-07,
+      "loss": 0.4699,
+      "step": 16644
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.064086434663675,
+      "learning_rate": 9.182241041048267e-07,
+      "loss": 0.5064,
+      "step": 16645
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 19.879351289909327,
+      "learning_rate": 9.175183318047298e-07,
+      "loss": 0.3794,
+      "step": 16646
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.104982016410478,
+      "learning_rate": 9.168128178051005e-07,
+      "loss": 0.4872,
+      "step": 16647
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.443144078723233,
+      "learning_rate": 9.161075621259997e-07,
+      "loss": 0.434,
+      "step": 16648
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 25.691266606296317,
+      "learning_rate": 9.154025647874875e-07,
+      "loss": 0.4629,
+      "step": 16649
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.55391299331185,
+      "learning_rate": 9.146978258096107e-07,
+      "loss": 0.4259,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 27.076821679404723,
+      "learning_rate": 9.139933452124117e-07,
+      "loss": 0.4911,
+      "step": 16651
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.259026280708042,
+      "learning_rate": 9.132891230159302e-07,
+      "loss": 0.4465,
+      "step": 16652
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.937368933457027,
+      "learning_rate": 9.125851592401891e-07,
+      "loss": 0.4729,
+      "step": 16653
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.395639017717112,
+      "learning_rate": 9.11881453905209e-07,
+      "loss": 0.4109,
+      "step": 16654
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.980382389952663,
+      "learning_rate": 9.111780070310038e-07,
+      "loss": 0.3972,
+      "step": 16655
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.66773722962771,
+      "learning_rate": 9.104748186375811e-07,
+      "loss": 0.4461,
+      "step": 16656
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.951502970467896,
+      "learning_rate": 9.097718887449381e-07,
+      "loss": 0.5012,
+      "step": 16657
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 23.418857307534378,
+      "learning_rate": 9.090692173730619e-07,
+      "loss": 0.455,
+      "step": 16658
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 24.002348904105794,
+      "learning_rate": 9.083668045419391e-07,
+      "loss": 0.4388,
+      "step": 16659
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 24.089457186991297,
+      "learning_rate": 9.076646502715469e-07,
+      "loss": 0.4302,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.07289727410144,
+      "learning_rate": 9.069627545818549e-07,
+      "loss": 0.4516,
+      "step": 16661
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 26.221881132746827,
+      "learning_rate": 9.062611174928226e-07,
+      "loss": 0.4544,
+      "step": 16662
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.094976478807144,
+      "learning_rate": 9.05559739024402e-07,
+      "loss": 0.3801,
+      "step": 16663
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 25.400617592166768,
+      "learning_rate": 9.048586191965437e-07,
+      "loss": 0.4404,
+      "step": 16664
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.691117029543463,
+      "learning_rate": 9.04157758029186e-07,
+      "loss": 0.4639,
+      "step": 16665
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 27.536180260436744,
+      "learning_rate": 9.0345715554226e-07,
+      "loss": 0.4018,
+      "step": 16666
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.50435428209808,
+      "learning_rate": 9.027568117556928e-07,
+      "loss": 0.4154,
+      "step": 16667
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 24.243344416720603,
+      "learning_rate": 9.020567266893976e-07,
+      "loss": 0.485,
+      "step": 16668
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.560779535954246,
+      "learning_rate": 9.013569003632894e-07,
+      "loss": 0.4223,
+      "step": 16669
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.521239161875705,
+      "learning_rate": 9.006573327972667e-07,
+      "loss": 0.4801,
+      "step": 16670
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.434193194990357,
+      "learning_rate": 8.99958024011226e-07,
+      "loss": 0.4201,
+      "step": 16671
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 23.114078630467418,
+      "learning_rate": 8.992589740250579e-07,
+      "loss": 0.4696,
+      "step": 16672
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 23.288510534533547,
+      "learning_rate": 8.985601828586399e-07,
+      "loss": 0.4432,
+      "step": 16673
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.627993637626343,
+      "learning_rate": 8.978616505318461e-07,
+      "loss": 0.4122,
+      "step": 16674
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.652127299895884,
+      "learning_rate": 8.971633770645416e-07,
+      "loss": 0.3973,
+      "step": 16675
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 32.2956595079074,
+      "learning_rate": 8.964653624765874e-07,
+      "loss": 0.5038,
+      "step": 16676
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 29.932339219471015,
+      "learning_rate": 8.957676067878307e-07,
+      "loss": 0.5604,
+      "step": 16677
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.154867314072508,
+      "learning_rate": 8.950701100181203e-07,
+      "loss": 0.4561,
+      "step": 16678
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.638338017033707,
+      "learning_rate": 8.943728721872868e-07,
+      "loss": 0.4534,
+      "step": 16679
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.70356290678026,
+      "learning_rate": 8.936758933151623e-07,
+      "loss": 0.4547,
+      "step": 16680
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.97695713721677,
+      "learning_rate": 8.92979173421571e-07,
+      "loss": 0.3834,
+      "step": 16681
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.435849945476544,
+      "learning_rate": 8.922827125263234e-07,
+      "loss": 0.3932,
+      "step": 16682
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 27.428948725196182,
+      "learning_rate": 8.91586510649226e-07,
+      "loss": 0.473,
+      "step": 16683
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.52376543878768,
+      "learning_rate": 8.90890567810081e-07,
+      "loss": 0.5012,
+      "step": 16684
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 27.61751614527499,
+      "learning_rate": 8.901948840286789e-07,
+      "loss": 0.3716,
+      "step": 16685
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.86593807010374,
+      "learning_rate": 8.894994593248063e-07,
+      "loss": 0.3881,
+      "step": 16686
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.410364220851308,
+      "learning_rate": 8.888042937182406e-07,
+      "loss": 0.4258,
+      "step": 16687
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.115935050366904,
+      "learning_rate": 8.881093872287483e-07,
+      "loss": 0.4096,
+      "step": 16688
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.652318385530542,
+      "learning_rate": 8.874147398760957e-07,
+      "loss": 0.4372,
+      "step": 16689
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 30.22928360546812,
+      "learning_rate": 8.867203516800382e-07,
+      "loss": 0.4408,
+      "step": 16690
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 22.39774081527456,
+      "learning_rate": 8.860262226603234e-07,
+      "loss": 0.3935,
+      "step": 16691
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 22.919712059188633,
+      "learning_rate": 8.853323528366886e-07,
+      "loss": 0.4538,
+      "step": 16692
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 35.905719633642555,
+      "learning_rate": 8.846387422288704e-07,
+      "loss": 0.4302,
+      "step": 16693
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.033654368842818,
+      "learning_rate": 8.839453908565965e-07,
+      "loss": 0.4794,
+      "step": 16694
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 38.43862449892962,
+      "learning_rate": 8.832522987395798e-07,
+      "loss": 0.4356,
+      "step": 16695
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.464374741397606,
+      "learning_rate": 8.825594658975367e-07,
+      "loss": 0.4589,
+      "step": 16696
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.422209357572804,
+      "learning_rate": 8.818668923501683e-07,
+      "loss": 0.3923,
+      "step": 16697
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 22.09562721937399,
+      "learning_rate": 8.811745781171699e-07,
+      "loss": 0.4206,
+      "step": 16698
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.442060279077392,
+      "learning_rate": 8.804825232182345e-07,
+      "loss": 0.4029,
+      "step": 16699
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.58317273293,
+      "learning_rate": 8.7979072767304e-07,
+      "loss": 0.474,
+      "step": 16700
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 34.61315114537455,
+      "learning_rate": 8.790991915012636e-07,
+      "loss": 0.5113,
+      "step": 16701
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.743257077126724,
+      "learning_rate": 8.784079147225688e-07,
+      "loss": 0.5002,
+      "step": 16702
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.50267092235498,
+      "learning_rate": 8.777168973566186e-07,
+      "loss": 0.353,
+      "step": 16703
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.95125831094712,
+      "learning_rate": 8.770261394230617e-07,
+      "loss": 0.4314,
+      "step": 16704
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.74905084830219,
+      "learning_rate": 8.763356409415447e-07,
+      "loss": 0.4074,
+      "step": 16705
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.579036870571148,
+      "learning_rate": 8.756454019317063e-07,
+      "loss": 0.4278,
+      "step": 16706
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 31.87071763962313,
+      "learning_rate": 8.749554224131751e-07,
+      "loss": 0.553,
+      "step": 16707
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.60790887329473,
+      "learning_rate": 8.742657024055723e-07,
+      "loss": 0.4269,
+      "step": 16708
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 26.285916786115394,
+      "learning_rate": 8.735762419285143e-07,
+      "loss": 0.4468,
+      "step": 16709
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.82882622284602,
+      "learning_rate": 8.728870410016099e-07,
+      "loss": 0.4227,
+      "step": 16710
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 44.476794032022404,
+      "learning_rate": 8.721980996444601e-07,
+      "loss": 0.4647,
+      "step": 16711
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.05022180011697,
+      "learning_rate": 8.715094178766559e-07,
+      "loss": 0.4309,
+      "step": 16712
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 24.75239388612179,
+      "learning_rate": 8.708209957177826e-07,
+      "loss": 0.4265,
+      "step": 16713
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 34.79433308889338,
+      "learning_rate": 8.701328331874204e-07,
+      "loss": 0.5209,
+      "step": 16714
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.99271583940879,
+      "learning_rate": 8.694449303051411e-07,
+      "loss": 0.4841,
+      "step": 16715
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.185274258323,
+      "learning_rate": 8.68757287090507e-07,
+      "loss": 0.4563,
+      "step": 16716
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.957059233477462,
+      "learning_rate": 8.680699035630713e-07,
+      "loss": 0.4054,
+      "step": 16717
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.60780446819741,
+      "learning_rate": 8.673827797423862e-07,
+      "loss": 0.4715,
+      "step": 16718
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 27.01343319652089,
+      "learning_rate": 8.666959156479938e-07,
+      "loss": 0.4525,
+      "step": 16719
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.50782784891804,
+      "learning_rate": 8.660093112994261e-07,
+      "loss": 0.4836,
+      "step": 16720
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.246002297155318,
+      "learning_rate": 8.653229667162111e-07,
+      "loss": 0.4106,
+      "step": 16721
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.868173611141453,
+      "learning_rate": 8.646368819178652e-07,
+      "loss": 0.3736,
+      "step": 16722
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 24.618808935065324,
+      "learning_rate": 8.639510569239029e-07,
+      "loss": 0.4321,
+      "step": 16723
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 28.00095475433834,
+      "learning_rate": 8.632654917538297e-07,
+      "loss": 0.4507,
+      "step": 16724
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.469426191949704,
+      "learning_rate": 8.625801864271411e-07,
+      "loss": 0.4288,
+      "step": 16725
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 26.974057184806597,
+      "learning_rate": 8.618951409633258e-07,
+      "loss": 0.4268,
+      "step": 16726
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 22.822644507968878,
+      "learning_rate": 8.612103553818663e-07,
+      "loss": 0.5234,
+      "step": 16727
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.76676574909621,
+      "learning_rate": 8.605258297022401e-07,
+      "loss": 0.3717,
+      "step": 16728
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.899604249361133,
+      "learning_rate": 8.598415639439107e-07,
+      "loss": 0.4623,
+      "step": 16729
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.332693949128902,
+      "learning_rate": 8.591575581263422e-07,
+      "loss": 0.4472,
+      "step": 16730
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 24.60243323073229,
+      "learning_rate": 8.58473812268984e-07,
+      "loss": 0.4391,
+      "step": 16731
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.63328755101761,
+      "learning_rate": 8.577903263912846e-07,
+      "loss": 0.3988,
+      "step": 16732
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 32.29013818102348,
+      "learning_rate": 8.571071005126785e-07,
+      "loss": 0.4678,
+      "step": 16733
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.242449249701234,
+      "learning_rate": 8.564241346525992e-07,
+      "loss": 0.4137,
+      "step": 16734
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.87398343052293,
+      "learning_rate": 8.557414288304689e-07,
+      "loss": 0.4265,
+      "step": 16735
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 25.34155911196266,
+      "learning_rate": 8.550589830657019e-07,
+      "loss": 0.3786,
+      "step": 16736
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.371625624494065,
+      "learning_rate": 8.543767973777095e-07,
+      "loss": 0.5357,
+      "step": 16737
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 22.74892713301613,
+      "learning_rate": 8.536948717858895e-07,
+      "loss": 0.4312,
+      "step": 16738
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.31856884009419,
+      "learning_rate": 8.530132063096364e-07,
+      "loss": 0.4346,
+      "step": 16739
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 25.60380375364114,
+      "learning_rate": 8.52331800968339e-07,
+      "loss": 0.426,
+      "step": 16740
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 24.597455123096058,
+      "learning_rate": 8.516506557813742e-07,
+      "loss": 0.4364,
+      "step": 16741
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 28.896660653956385,
+      "learning_rate": 8.509697707681108e-07,
+      "loss": 0.4143,
+      "step": 16742
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.727868880270037,
+      "learning_rate": 8.502891459479145e-07,
+      "loss": 0.4241,
+      "step": 16743
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.96661380279549,
+      "learning_rate": 8.496087813401454e-07,
+      "loss": 0.433,
+      "step": 16744
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.645807605775385,
+      "learning_rate": 8.489286769641492e-07,
+      "loss": 0.417,
+      "step": 16745
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 37.30492073505449,
+      "learning_rate": 8.482488328392668e-07,
+      "loss": 0.4635,
+      "step": 16746
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 24.241161234969553,
+      "learning_rate": 8.475692489848342e-07,
+      "loss": 0.4698,
+      "step": 16747
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 26.97141750841817,
+      "learning_rate": 8.468899254201768e-07,
+      "loss": 0.4484,
+      "step": 16748
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.310213648074953,
+      "learning_rate": 8.462108621646182e-07,
+      "loss": 0.3882,
+      "step": 16749
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.321206551973784,
+      "learning_rate": 8.455320592374683e-07,
+      "loss": 0.5108,
+      "step": 16750
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.945814612197484,
+      "learning_rate": 8.448535166580286e-07,
+      "loss": 0.4398,
+      "step": 16751
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.481744167599437,
+      "learning_rate": 8.441752344456001e-07,
+      "loss": 0.4991,
+      "step": 16752
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.533260033633923,
+      "learning_rate": 8.43497212619474e-07,
+      "loss": 0.4354,
+      "step": 16753
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 24.105916417968604,
+      "learning_rate": 8.428194511989285e-07,
+      "loss": 0.4483,
+      "step": 16754
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.228632958799906,
+      "learning_rate": 8.421419502032425e-07,
+      "loss": 0.406,
+      "step": 16755
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.048982944080215,
+      "learning_rate": 8.414647096516804e-07,
+      "loss": 0.4343,
+      "step": 16756
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 25.892150579569723,
+      "learning_rate": 8.40787729563507e-07,
+      "loss": 0.3986,
+      "step": 16757
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 15.535536964975222,
+      "learning_rate": 8.401110099579702e-07,
+      "loss": 0.4378,
+      "step": 16758
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.418707928177778,
+      "learning_rate": 8.394345508543178e-07,
+      "loss": 0.3899,
+      "step": 16759
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 24.74179145609242,
+      "learning_rate": 8.387583522717901e-07,
+      "loss": 0.4278,
+      "step": 16760
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 36.1969580306127,
+      "learning_rate": 8.38082414229614e-07,
+      "loss": 0.4397,
+      "step": 16761
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.140718870484307,
+      "learning_rate": 8.374067367470151e-07,
+      "loss": 0.432,
+      "step": 16762
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.99467742622639,
+      "learning_rate": 8.367313198432081e-07,
+      "loss": 0.4388,
+      "step": 16763
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.80601838524848,
+      "learning_rate": 8.36056163537402e-07,
+      "loss": 0.4025,
+      "step": 16764
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.8081271366024,
+      "learning_rate": 8.353812678487993e-07,
+      "loss": 0.4205,
+      "step": 16765
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.926565856637104,
+      "learning_rate": 8.347066327965925e-07,
+      "loss": 0.4961,
+      "step": 16766
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 35.660807793839375,
+      "learning_rate": 8.340322583999649e-07,
+      "loss": 0.3895,
+      "step": 16767
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.030089446166176,
+      "learning_rate": 8.333581446780981e-07,
+      "loss": 0.4632,
+      "step": 16768
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 27.060058253986636,
+      "learning_rate": 8.326842916501654e-07,
+      "loss": 0.5217,
+      "step": 16769
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.039772982655416,
+      "learning_rate": 8.320106993353294e-07,
+      "loss": 0.3913,
+      "step": 16770
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 25.421402582750858,
+      "learning_rate": 8.313373677527437e-07,
+      "loss": 0.4352,
+      "step": 16771
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 30.13481000243911,
+      "learning_rate": 8.306642969215595e-07,
+      "loss": 0.4793,
+      "step": 16772
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.837973011066698,
+      "learning_rate": 8.299914868609193e-07,
+      "loss": 0.423,
+      "step": 16773
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 31.082366852376566,
+      "learning_rate": 8.293189375899579e-07,
+      "loss": 0.4193,
+      "step": 16774
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 24.50445860318181,
+      "learning_rate": 8.286466491278023e-07,
+      "loss": 0.4479,
+      "step": 16775
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 27.43710492182028,
+      "learning_rate": 8.27974621493568e-07,
+      "loss": 0.4794,
+      "step": 16776
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.1983498746172,
+      "learning_rate": 8.27302854706371e-07,
+      "loss": 0.3988,
+      "step": 16777
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.101546037506935,
+      "learning_rate": 8.266313487853162e-07,
+      "loss": 0.4968,
+      "step": 16778
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.62009349238312,
+      "learning_rate": 8.259601037494991e-07,
+      "loss": 0.4279,
+      "step": 16779
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.489964474939928,
+      "learning_rate": 8.252891196180091e-07,
+      "loss": 0.3905,
+      "step": 16780
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.66843271886234,
+      "learning_rate": 8.246183964099286e-07,
+      "loss": 0.4721,
+      "step": 16781
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 27.24275700854788,
+      "learning_rate": 8.239479341443357e-07,
+      "loss": 0.4734,
+      "step": 16782
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 25.523912614546024,
+      "learning_rate": 8.232777328402941e-07,
+      "loss": 0.4371,
+      "step": 16783
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.74135799249883,
+      "learning_rate": 8.226077925168674e-07,
+      "loss": 0.4146,
+      "step": 16784
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 37.497727513598754,
+      "learning_rate": 8.219381131931048e-07,
+      "loss": 0.3987,
+      "step": 16785
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 26.18147060306356,
+      "learning_rate": 8.212686948880521e-07,
+      "loss": 0.4137,
+      "step": 16786
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.95040533477703,
+      "learning_rate": 8.20599537620751e-07,
+      "loss": 0.4516,
+      "step": 16787
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 33.21193701012433,
+      "learning_rate": 8.199306414102282e-07,
+      "loss": 0.5055,
+      "step": 16788
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 25.343243771446673,
+      "learning_rate": 8.192620062755085e-07,
+      "loss": 0.4725,
+      "step": 16789
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.176099939035836,
+      "learning_rate": 8.185936322356047e-07,
+      "loss": 0.4315,
+      "step": 16790
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 33.648030290338376,
+      "learning_rate": 8.179255193095292e-07,
+      "loss": 0.4847,
+      "step": 16791
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 26.493735355662018,
+      "learning_rate": 8.172576675162791e-07,
+      "loss": 0.4058,
+      "step": 16792
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.61509767974897,
+      "learning_rate": 8.165900768748491e-07,
+      "loss": 0.422,
+      "step": 16793
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.645227434823752,
+      "learning_rate": 8.159227474042276e-07,
+      "loss": 0.4167,
+      "step": 16794
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.955256101404252,
+      "learning_rate": 8.15255679123389e-07,
+      "loss": 0.4241,
+      "step": 16795
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.46999967207054,
+      "learning_rate": 8.145888720513051e-07,
+      "loss": 0.4071,
+      "step": 16796
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 22.830333613319574,
+      "learning_rate": 8.139223262069407e-07,
+      "loss": 0.4656,
+      "step": 16797
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.8787181911867,
+      "learning_rate": 8.132560416092516e-07,
+      "loss": 0.4108,
+      "step": 16798
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.966142990180433,
+      "learning_rate": 8.125900182771874e-07,
+      "loss": 0.514,
+      "step": 16799
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.276508767152627,
+      "learning_rate": 8.119242562296892e-07,
+      "loss": 0.3818,
+      "step": 16800
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.75664585271627,
+      "learning_rate": 8.112587554856888e-07,
+      "loss": 0.3909,
+      "step": 16801
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.617668005466477,
+      "learning_rate": 8.105935160641143e-07,
+      "loss": 0.4209,
+      "step": 16802
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.143456110065678,
+      "learning_rate": 8.099285379838862e-07,
+      "loss": 0.4811,
+      "step": 16803
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.81945785979983,
+      "learning_rate": 8.092638212639136e-07,
+      "loss": 0.4048,
+      "step": 16804
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 29.83116819485186,
+      "learning_rate": 8.085993659231006e-07,
+      "loss": 0.4267,
+      "step": 16805
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.959123716419516,
+      "learning_rate": 8.079351719803441e-07,
+      "loss": 0.4186,
+      "step": 16806
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.705781315240756,
+      "learning_rate": 8.072712394545368e-07,
+      "loss": 0.4208,
+      "step": 16807
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 30.147405975955014,
+      "learning_rate": 8.066075683645557e-07,
+      "loss": 0.5079,
+      "step": 16808
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.853072253400462,
+      "learning_rate": 8.059441587292782e-07,
+      "loss": 0.3982,
+      "step": 16809
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 20.264731188925893,
+      "learning_rate": 8.052810105675702e-07,
+      "loss": 0.4114,
+      "step": 16810
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 20.17411913086202,
+      "learning_rate": 8.046181238982908e-07,
+      "loss": 0.4288,
+      "step": 16811
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 63.012065382080834,
+      "learning_rate": 8.039554987402942e-07,
+      "loss": 0.45,
+      "step": 16812
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.451686381382414,
+      "learning_rate": 8.03293135112423e-07,
+      "loss": 0.4377,
+      "step": 16813
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 25.465290990800092,
+      "learning_rate": 8.026310330335163e-07,
+      "loss": 0.4364,
+      "step": 16814
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 27.945110257392457,
+      "learning_rate": 8.019691925224004e-07,
+      "loss": 0.4653,
+      "step": 16815
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 36.84208957556049,
+      "learning_rate": 8.013076135979025e-07,
+      "loss": 0.453,
+      "step": 16816
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.14770967550887,
+      "learning_rate": 8.006462962788331e-07,
+      "loss": 0.3888,
+      "step": 16817
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.031860844561525,
+      "learning_rate": 7.999852405840025e-07,
+      "loss": 0.4422,
+      "step": 16818
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 32.92649912212816,
+      "learning_rate": 7.993244465322092e-07,
+      "loss": 0.5207,
+      "step": 16819
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.903131962600781,
+      "learning_rate": 7.986639141422469e-07,
+      "loss": 0.3809,
+      "step": 16820
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 27.86737348966611,
+      "learning_rate": 7.980036434328997e-07,
+      "loss": 0.4695,
+      "step": 16821
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 30.168104516958334,
+      "learning_rate": 7.973436344229458e-07,
+      "loss": 0.4653,
+      "step": 16822
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.473699372465013,
+      "learning_rate": 7.966838871311566e-07,
+      "loss": 0.3886,
+      "step": 16823
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.798365960033053,
+      "learning_rate": 7.960244015762919e-07,
+      "loss": 0.4322,
+      "step": 16824
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 38.22287298032886,
+      "learning_rate": 7.953651777771121e-07,
+      "loss": 0.4198,
+      "step": 16825
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 29.210312917499845,
+      "learning_rate": 7.947062157523589e-07,
+      "loss": 0.4228,
+      "step": 16826
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 20.062279305625704,
+      "learning_rate": 7.940475155207772e-07,
+      "loss": 0.4341,
+      "step": 16827
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.993112556899609,
+      "learning_rate": 7.933890771010999e-07,
+      "loss": 0.4629,
+      "step": 16828
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.010568034084475,
+      "learning_rate": 7.927309005120521e-07,
+      "loss": 0.4348,
+      "step": 16829
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.577993284998563,
+      "learning_rate": 7.920729857723486e-07,
+      "loss": 0.4326,
+      "step": 16830
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.987880536647573,
+      "learning_rate": 7.914153329007035e-07,
+      "loss": 0.4189,
+      "step": 16831
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.49306578711967,
+      "learning_rate": 7.907579419158196e-07,
+      "loss": 0.411,
+      "step": 16832
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.166915742350334,
+      "learning_rate": 7.901008128363963e-07,
+      "loss": 0.4271,
+      "step": 16833
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.84249898585503,
+      "learning_rate": 7.894439456811143e-07,
+      "loss": 0.4914,
+      "step": 16834
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.18164285128788,
+      "learning_rate": 7.887873404686586e-07,
+      "loss": 0.4232,
+      "step": 16835
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 30.780090636964363,
+      "learning_rate": 7.88130997217702e-07,
+      "loss": 0.4636,
+      "step": 16836
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.54647295481414,
+      "learning_rate": 7.874749159469131e-07,
+      "loss": 0.4376,
+      "step": 16837
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.70252201298119,
+      "learning_rate": 7.868190966749489e-07,
+      "loss": 0.4249,
+      "step": 16838
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 24.30281482411287,
+      "learning_rate": 7.861635394204581e-07,
+      "loss": 0.4623,
+      "step": 16839
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.627583553703232,
+      "learning_rate": 7.855082442020867e-07,
+      "loss": 0.3949,
+      "step": 16840
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.371049234059328,
+      "learning_rate": 7.84853211038471e-07,
+      "loss": 0.4823,
+      "step": 16841
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.317755842555908,
+      "learning_rate": 7.841984399482383e-07,
+      "loss": 0.4953,
+      "step": 16842
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.099669392705323,
+      "learning_rate": 7.835439309500126e-07,
+      "loss": 0.4315,
+      "step": 16843
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.893913050184157,
+      "learning_rate": 7.828896840624045e-07,
+      "loss": 0.427,
+      "step": 16844
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.702461904718188,
+      "learning_rate": 7.822356993040236e-07,
+      "loss": 0.4448,
+      "step": 16845
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.718610975794803,
+      "learning_rate": 7.815819766934651e-07,
+      "loss": 0.5642,
+      "step": 16846
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.398796786558435,
+      "learning_rate": 7.80928516249323e-07,
+      "loss": 0.3895,
+      "step": 16847
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.030846737854098,
+      "learning_rate": 7.802753179901823e-07,
+      "loss": 0.426,
+      "step": 16848
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.90180827535106,
+      "learning_rate": 7.796223819346171e-07,
+      "loss": 0.3751,
+      "step": 16849
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 34.96958743022472,
+      "learning_rate": 7.789697081011982e-07,
+      "loss": 0.3893,
+      "step": 16850
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.01562875515033,
+      "learning_rate": 7.783172965084851e-07,
+      "loss": 0.4309,
+      "step": 16851
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.948257573877996,
+      "learning_rate": 7.77665147175034e-07,
+      "loss": 0.4026,
+      "step": 16852
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.418381484613125,
+      "learning_rate": 7.770132601193936e-07,
+      "loss": 0.46,
+      "step": 16853
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.844320560823682,
+      "learning_rate": 7.763616353601e-07,
+      "loss": 0.4746,
+      "step": 16854
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.733132272031753,
+      "learning_rate": 7.75710272915684e-07,
+      "loss": 0.4695,
+      "step": 16855
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 28.000478809596892,
+      "learning_rate": 7.750591728046719e-07,
+      "loss": 0.4036,
+      "step": 16856
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 24.820126679601113,
+      "learning_rate": 7.744083350455811e-07,
+      "loss": 0.4649,
+      "step": 16857
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 20.32344043127804,
+      "learning_rate": 7.737577596569223e-07,
+      "loss": 0.4875,
+      "step": 16858
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.400681451454044,
+      "learning_rate": 7.731074466571942e-07,
+      "loss": 0.398,
+      "step": 16859
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.71669893634057,
+      "learning_rate": 7.724573960648907e-07,
+      "loss": 0.3693,
+      "step": 16860
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.352677457157014,
+      "learning_rate": 7.718076078985026e-07,
+      "loss": 0.3971,
+      "step": 16861
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.02533477471114,
+      "learning_rate": 7.711580821765085e-07,
+      "loss": 0.432,
+      "step": 16862
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.217546166501688,
+      "learning_rate": 7.705088189173804e-07,
+      "loss": 0.4335,
+      "step": 16863
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.160675082907265,
+      "learning_rate": 7.6985981813958e-07,
+      "loss": 0.4206,
+      "step": 16864
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.202533562825728,
+      "learning_rate": 7.692110798615682e-07,
+      "loss": 0.4083,
+      "step": 16865
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.503359812365275,
+      "learning_rate": 7.685626041017935e-07,
+      "loss": 0.4502,
+      "step": 16866
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.27997414275035,
+      "learning_rate": 7.67914390878699e-07,
+      "loss": 0.4318,
+      "step": 16867
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 25.835566482267733,
+      "learning_rate": 7.672664402107166e-07,
+      "loss": 0.4624,
+      "step": 16868
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.61740463190695,
+      "learning_rate": 7.66618752116276e-07,
+      "loss": 0.3958,
+      "step": 16869
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 22.85098183309853,
+      "learning_rate": 7.659713266137992e-07,
+      "loss": 0.4409,
+      "step": 16870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.75568043175452,
+      "learning_rate": 7.653241637216924e-07,
+      "loss": 0.3987,
+      "step": 16871
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 38.46801953044423,
+      "learning_rate": 7.646772634583677e-07,
+      "loss": 0.4885,
+      "step": 16872
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.697771136275506,
+      "learning_rate": 7.64030625842217e-07,
+      "loss": 0.4978,
+      "step": 16873
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.03607930650342,
+      "learning_rate": 7.633842508916323e-07,
+      "loss": 0.4302,
+      "step": 16874
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.131389335085494,
+      "learning_rate": 7.627381386249976e-07,
+      "loss": 0.4964,
+      "step": 16875
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.221972021076663,
+      "learning_rate": 7.620922890606852e-07,
+      "loss": 0.4262,
+      "step": 16876
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.861658705586958,
+      "learning_rate": 7.614467022170658e-07,
+      "loss": 0.4186,
+      "step": 16877
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 26.86455808652885,
+      "learning_rate": 7.608013781124956e-07,
+      "loss": 0.5342,
+      "step": 16878
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 23.439050092300334,
+      "learning_rate": 7.601563167653314e-07,
+      "loss": 0.4012,
+      "step": 16879
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 22.456579716420062,
+      "learning_rate": 7.59511518193915e-07,
+      "loss": 0.4512,
+      "step": 16880
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.554704605630015,
+      "learning_rate": 7.588669824165851e-07,
+      "loss": 0.4681,
+      "step": 16881
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 34.10344989705391,
+      "learning_rate": 7.582227094516747e-07,
+      "loss": 0.493,
+      "step": 16882
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.93347241115998,
+      "learning_rate": 7.575786993175028e-07,
+      "loss": 0.4627,
+      "step": 16883
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.393700149294325,
+      "learning_rate": 7.569349520323854e-07,
+      "loss": 0.4639,
+      "step": 16884
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.614876370363987,
+      "learning_rate": 7.562914676146304e-07,
+      "loss": 0.4296,
+      "step": 16885
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 16.77048501804122,
+      "learning_rate": 7.556482460825399e-07,
+      "loss": 0.3855,
+      "step": 16886
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.079864852779185,
+      "learning_rate": 7.550052874544056e-07,
+      "loss": 0.4153,
+      "step": 16887
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 27.814226345533076,
+      "learning_rate": 7.543625917485142e-07,
+      "loss": 0.3975,
+      "step": 16888
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.274484261906014,
+      "learning_rate": 7.537201589831389e-07,
+      "loss": 0.4465,
+      "step": 16889
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.42591955955083,
+      "learning_rate": 7.530779891765549e-07,
+      "loss": 0.3628,
+      "step": 16890
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.478408078269176,
+      "learning_rate": 7.524360823470244e-07,
+      "loss": 0.4413,
+      "step": 16891
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 23.506952992044383,
+      "learning_rate": 7.517944385128018e-07,
+      "loss": 0.4642,
+      "step": 16892
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.75456311728819,
+      "learning_rate": 7.511530576921344e-07,
+      "loss": 0.4452,
+      "step": 16893
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 22.437404387968996,
+      "learning_rate": 7.505119399032623e-07,
+      "loss": 0.4101,
+      "step": 16894
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.54236982975728,
+      "learning_rate": 7.498710851644231e-07,
+      "loss": 0.3821,
+      "step": 16895
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.62335673484736,
+      "learning_rate": 7.492304934938365e-07,
+      "loss": 0.4359,
+      "step": 16896
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 28.060790699130283,
+      "learning_rate": 7.48590164909725e-07,
+      "loss": 0.4738,
+      "step": 16897
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.914088653631087,
+      "learning_rate": 7.479500994302957e-07,
+      "loss": 0.3831,
+      "step": 16898
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 16.568575890819545,
+      "learning_rate": 7.473102970737534e-07,
+      "loss": 0.4251,
+      "step": 16899
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.149919860230582,
+      "learning_rate": 7.466707578582954e-07,
+      "loss": 0.5003,
+      "step": 16900
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.38762558108232,
+      "learning_rate": 7.460314818021053e-07,
+      "loss": 0.3956,
+      "step": 16901
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.40323957311481,
+      "learning_rate": 7.453924689233693e-07,
+      "loss": 0.4099,
+      "step": 16902
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.58806714822402,
+      "learning_rate": 7.447537192402554e-07,
+      "loss": 0.4006,
+      "step": 16903
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 33.796221153153,
+      "learning_rate": 7.441152327709334e-07,
+      "loss": 0.407,
+      "step": 16904
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 23.945559204473557,
+      "learning_rate": 7.434770095335575e-07,
+      "loss": 0.441,
+      "step": 16905
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 28.343743056156608,
+      "learning_rate": 7.428390495462812e-07,
+      "loss": 0.4714,
+      "step": 16906
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 33.61964477292253,
+      "learning_rate": 7.422013528272487e-07,
+      "loss": 0.4491,
+      "step": 16907
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.092504680060543,
+      "learning_rate": 7.415639193945945e-07,
+      "loss": 0.4407,
+      "step": 16908
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.522835893585672,
+      "learning_rate": 7.409267492664441e-07,
+      "loss": 0.5003,
+      "step": 16909
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 26.416019802692528,
+      "learning_rate": 7.402898424609206e-07,
+      "loss": 0.4278,
+      "step": 16910
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.506691327371218,
+      "learning_rate": 7.396531989961364e-07,
+      "loss": 0.4247,
+      "step": 16911
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 34.72702402638485,
+      "learning_rate": 7.390168188902014e-07,
+      "loss": 0.4221,
+      "step": 16912
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 27.332931770317877,
+      "learning_rate": 7.383807021612089e-07,
+      "loss": 0.4089,
+      "step": 16913
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.784822813677792,
+      "learning_rate": 7.377448488272509e-07,
+      "loss": 0.4775,
+      "step": 16914
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 23.164310796863965,
+      "learning_rate": 7.3710925890641e-07,
+      "loss": 0.4712,
+      "step": 16915
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.609184651001883,
+      "learning_rate": 7.364739324167658e-07,
+      "loss": 0.3752,
+      "step": 16916
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.25547887539159,
+      "learning_rate": 7.35838869376384e-07,
+      "loss": 0.4694,
+      "step": 16917
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 46.19113616541803,
+      "learning_rate": 7.352040698033236e-07,
+      "loss": 0.4421,
+      "step": 16918
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 37.32707569440677,
+      "learning_rate": 7.345695337156389e-07,
+      "loss": 0.5207,
+      "step": 16919
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 27.358150016859156,
+      "learning_rate": 7.339352611313777e-07,
+      "loss": 0.4851,
+      "step": 16920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.542554103288225,
+      "learning_rate": 7.33301252068579e-07,
+      "loss": 0.5202,
+      "step": 16921
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.700291455294751,
+      "learning_rate": 7.326675065452693e-07,
+      "loss": 0.3829,
+      "step": 16922
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 39.02120483625192,
+      "learning_rate": 7.320340245794755e-07,
+      "loss": 0.4012,
+      "step": 16923
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.32930446624821,
+      "learning_rate": 7.314008061892108e-07,
+      "loss": 0.4506,
+      "step": 16924
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.33022061773795,
+      "learning_rate": 7.307678513924877e-07,
+      "loss": 0.3949,
+      "step": 16925
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.4365060809573,
+      "learning_rate": 7.301351602073048e-07,
+      "loss": 0.4611,
+      "step": 16926
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.600283221111766,
+      "learning_rate": 7.295027326516535e-07,
+      "loss": 0.4199,
+      "step": 16927
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.78572774344092,
+      "learning_rate": 7.288705687435204e-07,
+      "loss": 0.435,
+      "step": 16928
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.812328181439334,
+      "learning_rate": 7.28238668500888e-07,
+      "loss": 0.5015,
+      "step": 16929
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 22.96636934009789,
+      "learning_rate": 7.276070319417216e-07,
+      "loss": 0.4346,
+      "step": 16930
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.218819628644813,
+      "learning_rate": 7.269756590839883e-07,
+      "loss": 0.4315,
+      "step": 16931
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 23.283954821665816,
+      "learning_rate": 7.263445499456412e-07,
+      "loss": 0.4576,
+      "step": 16932
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.04858804417328,
+      "learning_rate": 7.257137045446327e-07,
+      "loss": 0.4329,
+      "step": 16933
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.926081830243813,
+      "learning_rate": 7.250831228988986e-07,
+      "loss": 0.4557,
+      "step": 16934
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.174457614433024,
+      "learning_rate": 7.244528050263744e-07,
+      "loss": 0.454,
+      "step": 16935
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.867504037493568,
+      "learning_rate": 7.238227509449891e-07,
+      "loss": 0.3953,
+      "step": 16936
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 34.6054061172561,
+      "learning_rate": 7.231929606726562e-07,
+      "loss": 0.4575,
+      "step": 16937
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.417321541570763,
+      "learning_rate": 7.225634342272903e-07,
+      "loss": 0.3648,
+      "step": 16938
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.1481978231408,
+      "learning_rate": 7.219341716267925e-07,
+      "loss": 0.4202,
+      "step": 16939
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 24.39213821181831,
+      "learning_rate": 7.213051728890586e-07,
+      "loss": 0.4107,
+      "step": 16940
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 29.51286279517155,
+      "learning_rate": 7.206764380319786e-07,
+      "loss": 0.4013,
+      "step": 16941
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 16.480239657708285,
+      "learning_rate": 7.20047967073434e-07,
+      "loss": 0.4018,
+      "step": 16942
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.85465106748411,
+      "learning_rate": 7.194197600312936e-07,
+      "loss": 0.4261,
+      "step": 16943
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.240310128614013,
+      "learning_rate": 7.187918169234265e-07,
+      "loss": 0.4741,
+      "step": 16944
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.877992688804472,
+      "learning_rate": 7.181641377676918e-07,
+      "loss": 0.4236,
+      "step": 16945
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.54637360025032,
+      "learning_rate": 7.175367225819418e-07,
+      "loss": 0.4088,
+      "step": 16946
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 28.202189692577342,
+      "learning_rate": 7.169095713840135e-07,
+      "loss": 0.4307,
+      "step": 16947
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.242653932280582,
+      "learning_rate": 7.16282684191747e-07,
+      "loss": 0.4118,
+      "step": 16948
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 25.825725198157475,
+      "learning_rate": 7.156560610229701e-07,
+      "loss": 0.4823,
+      "step": 16949
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 30.023298817616297,
+      "learning_rate": 7.150297018955055e-07,
+      "loss": 0.4348,
+      "step": 16950
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 24.26751415712457,
+      "learning_rate": 7.144036068271654e-07,
+      "loss": 0.4398,
+      "step": 16951
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.18227637387653,
+      "learning_rate": 7.137777758357523e-07,
+      "loss": 0.4926,
+      "step": 16952
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.60419619903871,
+      "learning_rate": 7.131522089390663e-07,
+      "loss": 0.4225,
+      "step": 16953
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.218780213293343,
+      "learning_rate": 7.125269061549012e-07,
+      "loss": 0.4487,
+      "step": 16954
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 24.474617092675352,
+      "learning_rate": 7.119018675010369e-07,
+      "loss": 0.4443,
+      "step": 16955
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.064154574624544,
+      "learning_rate": 7.112770929952506e-07,
+      "loss": 0.4581,
+      "step": 16956
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 24.531868571178812,
+      "learning_rate": 7.10652582655309e-07,
+      "loss": 0.4319,
+      "step": 16957
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 24.806323178645442,
+      "learning_rate": 7.100283364989757e-07,
+      "loss": 0.4148,
+      "step": 16958
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 28.25120667395479,
+      "learning_rate": 7.094043545439999e-07,
+      "loss": 0.4339,
+      "step": 16959
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 25.00788101906701,
+      "learning_rate": 7.087806368081296e-07,
+      "loss": 0.481,
+      "step": 16960
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.9109394047861,
+      "learning_rate": 7.081571833091039e-07,
+      "loss": 0.4802,
+      "step": 16961
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.740040837849502,
+      "learning_rate": 7.075339940646508e-07,
+      "loss": 0.4054,
+      "step": 16962
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.74466591663702,
+      "learning_rate": 7.069110690924974e-07,
+      "loss": 0.3905,
+      "step": 16963
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 22.29085204531449,
+      "learning_rate": 7.06288408410355e-07,
+      "loss": 0.385,
+      "step": 16964
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.844893795619,
+      "learning_rate": 7.056660120359349e-07,
+      "loss": 0.4143,
+      "step": 16965
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 25.0207018040152,
+      "learning_rate": 7.050438799869352e-07,
+      "loss": 0.4699,
+      "step": 16966
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.597550155580713,
+      "learning_rate": 7.044220122810508e-07,
+      "loss": 0.3672,
+      "step": 16967
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.35701236577994,
+      "learning_rate": 7.038004089359663e-07,
+      "loss": 0.4449,
+      "step": 16968
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 22.91420574904314,
+      "learning_rate": 7.031790699693586e-07,
+      "loss": 0.4751,
+      "step": 16969
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.909948800782285,
+      "learning_rate": 7.025579953989026e-07,
+      "loss": 0.4267,
+      "step": 16970
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.633419328048156,
+      "learning_rate": 7.019371852422574e-07,
+      "loss": 0.4522,
+      "step": 16971
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 40.40897275317225,
+      "learning_rate": 7.01316639517079e-07,
+      "loss": 0.4922,
+      "step": 16972
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.68186955093115,
+      "learning_rate": 7.006963582410153e-07,
+      "loss": 0.3753,
+      "step": 16973
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 32.352136896557326,
+      "learning_rate": 7.000763414317069e-07,
+      "loss": 0.4366,
+      "step": 16974
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.033830856347995,
+      "learning_rate": 6.994565891067884e-07,
+      "loss": 0.3926,
+      "step": 16975
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.2580439478215,
+      "learning_rate": 6.988371012838846e-07,
+      "loss": 0.4956,
+      "step": 16976
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 26.155731691673708,
+      "learning_rate": 6.982178779806104e-07,
+      "loss": 0.4942,
+      "step": 16977
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.71712777113979,
+      "learning_rate": 6.975989192145782e-07,
+      "loss": 0.4181,
+      "step": 16978
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 27.454886283890847,
+      "learning_rate": 6.96980225003393e-07,
+      "loss": 0.4593,
+      "step": 16979
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.569902573897206,
+      "learning_rate": 6.963617953646484e-07,
+      "loss": 0.4401,
+      "step": 16980
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 34.14101564546121,
+      "learning_rate": 6.957436303159304e-07,
+      "loss": 0.4662,
+      "step": 16981
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.973733968815406,
+      "learning_rate": 6.951257298748204e-07,
+      "loss": 0.4327,
+      "step": 16982
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 28.526577083748773,
+      "learning_rate": 6.94508094058891e-07,
+      "loss": 0.4512,
+      "step": 16983
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.720721383234114,
+      "learning_rate": 6.938907228857095e-07,
+      "loss": 0.4055,
+      "step": 16984
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.930534144588915,
+      "learning_rate": 6.932736163728327e-07,
+      "loss": 0.4427,
+      "step": 16985
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.495941110873215,
+      "learning_rate": 6.926567745378076e-07,
+      "loss": 0.3663,
+      "step": 16986
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 32.750840610351986,
+      "learning_rate": 6.920401973981794e-07,
+      "loss": 0.4481,
+      "step": 16987
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.216017385605664,
+      "learning_rate": 6.914238849714849e-07,
+      "loss": 0.4434,
+      "step": 16988
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.142577046394596,
+      "learning_rate": 6.90807837275248e-07,
+      "loss": 0.4575,
+      "step": 16989
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.24530196288164,
+      "learning_rate": 6.901920543269925e-07,
+      "loss": 0.4538,
+      "step": 16990
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 16.196424638366462,
+      "learning_rate": 6.895765361442264e-07,
+      "loss": 0.4212,
+      "step": 16991
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.46682812361794,
+      "learning_rate": 6.889612827444592e-07,
+      "loss": 0.4293,
+      "step": 16992
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.018479384910545,
+      "learning_rate": 6.883462941451846e-07,
+      "loss": 0.4197,
+      "step": 16993
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.42966825185873,
+      "learning_rate": 6.877315703638943e-07,
+      "loss": 0.4504,
+      "step": 16994
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.598967359315626,
+      "learning_rate": 6.871171114180719e-07,
+      "loss": 0.4458,
+      "step": 16995
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.857768877455179,
+      "learning_rate": 6.865029173251914e-07,
+      "loss": 0.4249,
+      "step": 16996
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 22.479247635223278,
+      "learning_rate": 6.858889881027187e-07,
+      "loss": 0.4384,
+      "step": 16997
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 28.270730075820964,
+      "learning_rate": 6.852753237681131e-07,
+      "loss": 0.4507,
+      "step": 16998
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 31.32222442805024,
+      "learning_rate": 6.846619243388297e-07,
+      "loss": 0.4584,
+      "step": 16999
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 24.638024602031646,
+      "learning_rate": 6.840487898323122e-07,
+      "loss": 0.4003,
+      "step": 17000
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.438656737815304,
+      "learning_rate": 6.83435920265999e-07,
+      "loss": 0.4725,
+      "step": 17001
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.78180766489473,
+      "learning_rate": 6.82823315657315e-07,
+      "loss": 0.4358,
+      "step": 17002
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 27.469788973800213,
+      "learning_rate": 6.822109760236861e-07,
+      "loss": 0.4364,
+      "step": 17003
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.724507139790685,
+      "learning_rate": 6.815989013825285e-07,
+      "loss": 0.3935,
+      "step": 17004
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.300029688664686,
+      "learning_rate": 6.809870917512462e-07,
+      "loss": 0.4098,
+      "step": 17005
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.41557284557138,
+      "learning_rate": 6.803755471472384e-07,
+      "loss": 0.4639,
+      "step": 17006
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.21578261332035,
+      "learning_rate": 6.79764267587898e-07,
+      "loss": 0.4053,
+      "step": 17007
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.20306508758734,
+      "learning_rate": 6.791532530906098e-07,
+      "loss": 0.4261,
+      "step": 17008
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.602186768497322,
+      "learning_rate": 6.785425036727522e-07,
+      "loss": 0.401,
+      "step": 17009
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.02387156686479,
+      "learning_rate": 6.779320193516925e-07,
+      "loss": 0.408,
+      "step": 17010
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.59933059258428,
+      "learning_rate": 6.773218001447923e-07,
+      "loss": 0.4902,
+      "step": 17011
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.781701508115393,
+      "learning_rate": 6.767118460694056e-07,
+      "loss": 0.4027,
+      "step": 17012
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.55440458234907,
+      "learning_rate": 6.761021571428816e-07,
+      "loss": 0.4106,
+      "step": 17013
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 39.882061410794,
+      "learning_rate": 6.75492733382559e-07,
+      "loss": 0.4822,
+      "step": 17014
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 31.0322179436863,
+      "learning_rate": 6.748835748057658e-07,
+      "loss": 0.4131,
+      "step": 17015
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.6244217059585,
+      "learning_rate": 6.742746814298285e-07,
+      "loss": 0.4519,
+      "step": 17016
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 33.92483251954795,
+      "learning_rate": 6.736660532720662e-07,
+      "loss": 0.4562,
+      "step": 17017
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 29.13638939002006,
+      "learning_rate": 6.730576903497832e-07,
+      "loss": 0.4267,
+      "step": 17018
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.769779602741252,
+      "learning_rate": 6.724495926802854e-07,
+      "loss": 0.4623,
+      "step": 17019
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.370774772676448,
+      "learning_rate": 6.718417602808636e-07,
+      "loss": 0.4245,
+      "step": 17020
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.943558489430227,
+      "learning_rate": 6.71234193168806e-07,
+      "loss": 0.3762,
+      "step": 17021
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.346078157515993,
+      "learning_rate": 6.7062689136139e-07,
+      "loss": 0.4377,
+      "step": 17022
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.105330236805308,
+      "learning_rate": 6.700198548758874e-07,
+      "loss": 0.3829,
+      "step": 17023
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 24.84923823693445,
+      "learning_rate": 6.694130837295632e-07,
+      "loss": 0.4291,
+      "step": 17024
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 27.275788335089313,
+      "learning_rate": 6.688065779396713e-07,
+      "loss": 0.485,
+      "step": 17025
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.590290079460623,
+      "learning_rate": 6.682003375234636e-07,
+      "loss": 0.4434,
+      "step": 17026
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.593048153860142,
+      "learning_rate": 6.675943624981785e-07,
+      "loss": 0.4093,
+      "step": 17027
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.285089568057483,
+      "learning_rate": 6.669886528810498e-07,
+      "loss": 0.3994,
+      "step": 17028
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.830597707345294,
+      "learning_rate": 6.663832086893052e-07,
+      "loss": 0.4229,
+      "step": 17029
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 32.45391485311872,
+      "learning_rate": 6.657780299401628e-07,
+      "loss": 0.4441,
+      "step": 17030
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.40082117049648,
+      "learning_rate": 6.651731166508313e-07,
+      "loss": 0.4093,
+      "step": 17031
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 28.322825016510862,
+      "learning_rate": 6.645684688385156e-07,
+      "loss": 0.3998,
+      "step": 17032
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.069347035224137,
+      "learning_rate": 6.639640865204111e-07,
+      "loss": 0.3896,
+      "step": 17033
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 25.01354521374338,
+      "learning_rate": 6.633599697137105e-07,
+      "loss": 0.4234,
+      "step": 17034
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.49225597143289,
+      "learning_rate": 6.627561184355869e-07,
+      "loss": 0.4791,
+      "step": 17035
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.835923343331707,
+      "learning_rate": 6.621525327032174e-07,
+      "loss": 0.3775,
+      "step": 17036
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.424137567780104,
+      "learning_rate": 6.615492125337663e-07,
+      "loss": 0.4314,
+      "step": 17037
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.96515394052094,
+      "learning_rate": 6.609461579443954e-07,
+      "loss": 0.4201,
+      "step": 17038
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 25.84355037628256,
+      "learning_rate": 6.603433689522531e-07,
+      "loss": 0.4409,
+      "step": 17039
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 29.039379297744702,
+      "learning_rate": 6.597408455744792e-07,
+      "loss": 0.4797,
+      "step": 17040
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.028265306948867,
+      "learning_rate": 6.59138587828212e-07,
+      "loss": 0.5161,
+      "step": 17041
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 29.6197833989168,
+      "learning_rate": 6.585365957305812e-07,
+      "loss": 0.489,
+      "step": 17042
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.549998363375323,
+      "learning_rate": 6.57934869298703e-07,
+      "loss": 0.4531,
+      "step": 17043
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.46886577563109,
+      "learning_rate": 6.573334085496941e-07,
+      "loss": 0.4155,
+      "step": 17044
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.211848286248678,
+      "learning_rate": 6.567322135006559e-07,
+      "loss": 0.4408,
+      "step": 17045
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.880414287276913,
+      "learning_rate": 6.561312841686895e-07,
+      "loss": 0.4522,
+      "step": 17046
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.521092950427791,
+      "learning_rate": 6.555306205708812e-07,
+      "loss": 0.4371,
+      "step": 17047
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.563371343641823,
+      "learning_rate": 6.54930222724317e-07,
+      "loss": 0.4079,
+      "step": 17048
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.989714800296422,
+      "learning_rate": 6.543300906460714e-07,
+      "loss": 0.4755,
+      "step": 17049
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 27.088749978304143,
+      "learning_rate": 6.537302243532095e-07,
+      "loss": 0.4328,
+      "step": 17050
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 23.439463036027046,
+      "learning_rate": 6.531306238627932e-07,
+      "loss": 0.4756,
+      "step": 17051
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 27.9192020244739,
+      "learning_rate": 6.525312891918745e-07,
+      "loss": 0.4623,
+      "step": 17052
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 25.48685072819762,
+      "learning_rate": 6.519322203574962e-07,
+      "loss": 0.4263,
+      "step": 17053
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.352761920106488,
+      "learning_rate": 6.513334173766994e-07,
+      "loss": 0.4176,
+      "step": 17054
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.283399535065746,
+      "learning_rate": 6.507348802665115e-07,
+      "loss": 0.4183,
+      "step": 17055
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.49803511140803,
+      "learning_rate": 6.501366090439521e-07,
+      "loss": 0.3975,
+      "step": 17056
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.426900729752138,
+      "learning_rate": 6.495386037260387e-07,
+      "loss": 0.3751,
+      "step": 17057
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.43626412452176,
+      "learning_rate": 6.489408643297778e-07,
+      "loss": 0.4494,
+      "step": 17058
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.855540566078112,
+      "learning_rate": 6.483433908721715e-07,
+      "loss": 0.4005,
+      "step": 17059
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.73181552104461,
+      "learning_rate": 6.477461833702047e-07,
+      "loss": 0.4296,
+      "step": 17060
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.51348699700499,
+      "learning_rate": 6.471492418408665e-07,
+      "loss": 0.3751,
+      "step": 17061
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.553816710355985,
+      "learning_rate": 6.465525663011318e-07,
+      "loss": 0.4578,
+      "step": 17062
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.542725772818944,
+      "learning_rate": 6.459561567679728e-07,
+      "loss": 0.4782,
+      "step": 17063
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 29.15203652187422,
+      "learning_rate": 6.453600132583482e-07,
+      "loss": 0.5064,
+      "step": 17064
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.260137969017435,
+      "learning_rate": 6.44764135789211e-07,
+      "loss": 0.4755,
+      "step": 17065
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.9456889478899,
+      "learning_rate": 6.441685243775097e-07,
+      "loss": 0.4129,
+      "step": 17066
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 27.729086738170775,
+      "learning_rate": 6.435731790401833e-07,
+      "loss": 0.5379,
+      "step": 17067
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.976227029806033,
+      "learning_rate": 6.429780997941626e-07,
+      "loss": 0.447,
+      "step": 17068
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 22.96528416714328,
+      "learning_rate": 6.423832866563696e-07,
+      "loss": 0.4098,
+      "step": 17069
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 30.1824875218266,
+      "learning_rate": 6.417887396437217e-07,
+      "loss": 0.4455,
+      "step": 17070
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 25.392324768098185,
+      "learning_rate": 6.411944587731279e-07,
+      "loss": 0.3924,
+      "step": 17071
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.80517515827337,
+      "learning_rate": 6.406004440614911e-07,
+      "loss": 0.409,
+      "step": 17072
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.29175992762999,
+      "learning_rate": 6.400066955257023e-07,
+      "loss": 0.4157,
+      "step": 17073
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.612353141023707,
+      "learning_rate": 6.394132131826447e-07,
+      "loss": 0.4413,
+      "step": 17074
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 24.05933694114306,
+      "learning_rate": 6.388199970492015e-07,
+      "loss": 0.4136,
+      "step": 17075
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 30.137079820429193,
+      "learning_rate": 6.382270471422413e-07,
+      "loss": 0.4603,
+      "step": 17076
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 26.955829232600486,
+      "learning_rate": 6.376343634786264e-07,
+      "loss": 0.4748,
+      "step": 17077
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 34.27518691882161,
+      "learning_rate": 6.370419460752153e-07,
+      "loss": 0.4157,
+      "step": 17078
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 27.31960892582676,
+      "learning_rate": 6.364497949488524e-07,
+      "loss": 0.4197,
+      "step": 17079
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 25.238343928804962,
+      "learning_rate": 6.35857910116382e-07,
+      "loss": 0.4026,
+      "step": 17080
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.92394860997047,
+      "learning_rate": 6.352662915946317e-07,
+      "loss": 0.3977,
+      "step": 17081
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 35.61218212490607,
+      "learning_rate": 6.346749394004314e-07,
+      "loss": 0.445,
+      "step": 17082
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.427349684647595,
+      "learning_rate": 6.340838535505978e-07,
+      "loss": 0.3971,
+      "step": 17083
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.6933223110442,
+      "learning_rate": 6.334930340619405e-07,
+      "loss": 0.4132,
+      "step": 17084
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.313131267683865,
+      "learning_rate": 6.329024809512607e-07,
+      "loss": 0.3959,
+      "step": 17085
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 31.41987764705529,
+      "learning_rate": 6.323121942353549e-07,
+      "loss": 0.469,
+      "step": 17086
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.329605457661362,
+      "learning_rate": 6.317221739310109e-07,
+      "loss": 0.4416,
+      "step": 17087
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 30.76519342568286,
+      "learning_rate": 6.311324200550084e-07,
+      "loss": 0.3996,
+      "step": 17088
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.547293628811232,
+      "learning_rate": 6.305429326241208e-07,
+      "loss": 0.4028,
+      "step": 17089
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.297110116154766,
+      "learning_rate": 6.29953711655108e-07,
+      "loss": 0.3779,
+      "step": 17090
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.02544492277749,
+      "learning_rate": 6.29364757164731e-07,
+      "loss": 0.4378,
+      "step": 17091
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 31.638128218862505,
+      "learning_rate": 6.287760691697409e-07,
+      "loss": 0.3771,
+      "step": 17092
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 28.34857753829879,
+      "learning_rate": 6.281876476868764e-07,
+      "loss": 0.4398,
+      "step": 17093
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.48143995246555,
+      "learning_rate": 6.27599492732871e-07,
+      "loss": 0.3982,
+      "step": 17094
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 34.10834487900786,
+      "learning_rate": 6.270116043244545e-07,
+      "loss": 0.4615,
+      "step": 17095
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.37226666944277,
+      "learning_rate": 6.264239824783447e-07,
+      "loss": 0.4009,
+      "step": 17096
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.788523795163485,
+      "learning_rate": 6.258366272112537e-07,
+      "loss": 0.4299,
+      "step": 17097
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 34.367802364904556,
+      "learning_rate": 6.252495385398871e-07,
+      "loss": 0.4405,
+      "step": 17098
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.54238221033468,
+      "learning_rate": 6.246627164809371e-07,
+      "loss": 0.4143,
+      "step": 17099
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.206419770341999,
+      "learning_rate": 6.240761610510948e-07,
+      "loss": 0.4049,
+      "step": 17100
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 25.004906530504943,
+      "learning_rate": 6.234898722670435e-07,
+      "loss": 0.4295,
+      "step": 17101
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.665223184318045,
+      "learning_rate": 6.229038501454532e-07,
+      "loss": 0.4223,
+      "step": 17102
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 27.555947482896414,
+      "learning_rate": 6.223180947029939e-07,
+      "loss": 0.4118,
+      "step": 17103
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 36.97469000283389,
+      "learning_rate": 6.217326059563211e-07,
+      "loss": 0.5406,
+      "step": 17104
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.676037536469888,
+      "learning_rate": 6.211473839220883e-07,
+      "loss": 0.4801,
+      "step": 17105
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 16.2837532172914,
+      "learning_rate": 6.205624286169354e-07,
+      "loss": 0.4184,
+      "step": 17106
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 22.598559802165607,
+      "learning_rate": 6.199777400575013e-07,
+      "loss": 0.4292,
+      "step": 17107
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.6410971124788,
+      "learning_rate": 6.193933182604128e-07,
+      "loss": 0.4302,
+      "step": 17108
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 22.660236184239434,
+      "learning_rate": 6.188091632422921e-07,
+      "loss": 0.4313,
+      "step": 17109
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.236551590030146,
+      "learning_rate": 6.182252750197493e-07,
+      "loss": 0.4422,
+      "step": 17110
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.96506350085633,
+      "learning_rate": 6.17641653609391e-07,
+      "loss": 0.3998,
+      "step": 17111
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 22.552410013389125,
+      "learning_rate": 6.170582990278173e-07,
+      "loss": 0.4117,
+      "step": 17112
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.934543313498008,
+      "learning_rate": 6.16475211291615e-07,
+      "loss": 0.4282,
+      "step": 17113
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.85941118320791,
+      "learning_rate": 6.158923904173709e-07,
+      "loss": 0.4317,
+      "step": 17114
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.408329756931018,
+      "learning_rate": 6.153098364216548e-07,
+      "loss": 0.4434,
+      "step": 17115
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 30.182374382344552,
+      "learning_rate": 6.147275493210381e-07,
+      "loss": 0.4152,
+      "step": 17116
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.10646288741191,
+      "learning_rate": 6.141455291320808e-07,
+      "loss": 0.4617,
+      "step": 17117
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.288819508480705,
+      "learning_rate": 6.135637758713342e-07,
+      "loss": 0.3848,
+      "step": 17118
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 27.006153999233906,
+      "learning_rate": 6.129822895553417e-07,
+      "loss": 0.4295,
+      "step": 17119
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.434662383401378,
+      "learning_rate": 6.124010702006411e-07,
+      "loss": 0.3879,
+      "step": 17120
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 16.224438716109823,
+      "learning_rate": 6.118201178237626e-07,
+      "loss": 0.4242,
+      "step": 17121
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.356076925460892,
+      "learning_rate": 6.112394324412308e-07,
+      "loss": 0.4398,
+      "step": 17122
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 28.671184323590666,
+      "learning_rate": 6.106590140695545e-07,
+      "loss": 0.4515,
+      "step": 17123
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 30.10050385512265,
+      "learning_rate": 6.100788627252441e-07,
+      "loss": 0.443,
+      "step": 17124
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.642741107107778,
+      "learning_rate": 6.094989784247973e-07,
+      "loss": 0.4382,
+      "step": 17125
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.19994401930214,
+      "learning_rate": 6.089193611847066e-07,
+      "loss": 0.405,
+      "step": 17126
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.401677884194676,
+      "learning_rate": 6.083400110214577e-07,
+      "loss": 0.4165,
+      "step": 17127
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.75780475667396,
+      "learning_rate": 6.077609279515217e-07,
+      "loss": 0.4232,
+      "step": 17128
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 25.733366327489033,
+      "learning_rate": 6.071821119913713e-07,
+      "loss": 0.4643,
+      "step": 17129
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 18.175979693518443,
+      "learning_rate": 6.066035631574685e-07,
+      "loss": 0.4021,
+      "step": 17130
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 26.234837500134557,
+      "learning_rate": 6.060252814662637e-07,
+      "loss": 0.404,
+      "step": 17131
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.97044382449291,
+      "learning_rate": 6.05447266934206e-07,
+      "loss": 0.4089,
+      "step": 17132
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 23.005256356697373,
+      "learning_rate": 6.04869519577731e-07,
+      "loss": 0.4979,
+      "step": 17133
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 32.22426743259202,
+      "learning_rate": 6.042920394132712e-07,
+      "loss": 0.4754,
+      "step": 17134
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 31.007453232056136,
+      "learning_rate": 6.037148264572512e-07,
+      "loss": 0.4417,
+      "step": 17135
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.632163141931105,
+      "learning_rate": 6.031378807260823e-07,
+      "loss": 0.4203,
+      "step": 17136
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 15.153728765022663,
+      "learning_rate": 6.025612022361771e-07,
+      "loss": 0.3838,
+      "step": 17137
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.110892726310215,
+      "learning_rate": 6.019847910039334e-07,
+      "loss": 0.4138,
+      "step": 17138
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.509416667543377,
+      "learning_rate": 6.014086470457448e-07,
+      "loss": 0.3619,
+      "step": 17139
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.1188525284285,
+      "learning_rate": 6.008327703779948e-07,
+      "loss": 0.5014,
+      "step": 17140
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 27.94531645440214,
+      "learning_rate": 6.002571610170627e-07,
+      "loss": 0.5094,
+      "step": 17141
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 28.45396947262181,
+      "learning_rate": 5.996818189793207e-07,
+      "loss": 0.4211,
+      "step": 17142
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.533463448691474,
+      "learning_rate": 5.991067442811272e-07,
+      "loss": 0.4518,
+      "step": 17143
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.161471037321725,
+      "learning_rate": 5.985319369388376e-07,
+      "loss": 0.4032,
+      "step": 17144
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.837735811080933,
+      "learning_rate": 5.979573969688001e-07,
+      "loss": 0.3626,
+      "step": 17145
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 21.65626426244929,
+      "learning_rate": 5.973831243873551e-07,
+      "loss": 0.4867,
+      "step": 17146
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 15.804789619662726,
+      "learning_rate": 5.96809119210835e-07,
+      "loss": 0.3739,
+      "step": 17147
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 25.26088088300529,
+      "learning_rate": 5.96235381455561e-07,
+      "loss": 0.4652,
+      "step": 17148
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 32.06862106091011,
+      "learning_rate": 5.956619111378514e-07,
+      "loss": 0.468,
+      "step": 17149
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 18.802936646778313,
+      "learning_rate": 5.950887082740153e-07,
+      "loss": 0.4661,
+      "step": 17150
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 31.468231054836213,
+      "learning_rate": 5.945157728803563e-07,
+      "loss": 0.4134,
+      "step": 17151
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 31.13163359167715,
+      "learning_rate": 5.93943104973167e-07,
+      "loss": 0.5141,
+      "step": 17152
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 17.594273805020006,
+      "learning_rate": 5.93370704568732e-07,
+      "loss": 0.4046,
+      "step": 17153
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.04453531877571,
+      "learning_rate": 5.927985716833317e-07,
+      "loss": 0.4324,
+      "step": 17154
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 29.099627139363037,
+      "learning_rate": 5.922267063332376e-07,
+      "loss": 0.454,
+      "step": 17155
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 18.01097826282531,
+      "learning_rate": 5.916551085347134e-07,
+      "loss": 0.4073,
+      "step": 17156
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.356552635662474,
+      "learning_rate": 5.910837783040113e-07,
+      "loss": 0.4202,
+      "step": 17157
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 26.287884465615402,
+      "learning_rate": 5.905127156573842e-07,
+      "loss": 0.381,
+      "step": 17158
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.636182347838588,
+      "learning_rate": 5.899419206110702e-07,
+      "loss": 0.4357,
+      "step": 17159
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.067440509782113,
+      "learning_rate": 5.89371393181305e-07,
+      "loss": 0.4253,
+      "step": 17160
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.297473449668768,
+      "learning_rate": 5.888011333843113e-07,
+      "loss": 0.4758,
+      "step": 17161
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.258932992607946,
+      "learning_rate": 5.882311412363073e-07,
+      "loss": 0.4643,
+      "step": 17162
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 18.8381388852069,
+      "learning_rate": 5.876614167535044e-07,
+      "loss": 0.504,
+      "step": 17163
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.209615633281366,
+      "learning_rate": 5.870919599521052e-07,
+      "loss": 0.4723,
+      "step": 17164
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 21.734926843390635,
+      "learning_rate": 5.865227708483034e-07,
+      "loss": 0.4168,
+      "step": 17165
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.323562420114825,
+      "learning_rate": 5.859538494582895e-07,
+      "loss": 0.4534,
+      "step": 17166
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 26.53938933118751,
+      "learning_rate": 5.853851957982381e-07,
+      "loss": 0.4255,
+      "step": 17167
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.755254837116833,
+      "learning_rate": 5.848168098843265e-07,
+      "loss": 0.3997,
+      "step": 17168
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.590963999854182,
+      "learning_rate": 5.842486917327162e-07,
+      "loss": 0.4001,
+      "step": 17169
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 27.10782022277645,
+      "learning_rate": 5.836808413595641e-07,
+      "loss": 0.4644,
+      "step": 17170
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 29.15008667856092,
+      "learning_rate": 5.831132587810228e-07,
+      "loss": 0.4219,
+      "step": 17171
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 27.068763426568403,
+      "learning_rate": 5.825459440132308e-07,
+      "loss": 0.4718,
+      "step": 17172
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.258163998574254,
+      "learning_rate": 5.819788970723217e-07,
+      "loss": 0.4259,
+      "step": 17173
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 25.124331967762906,
+      "learning_rate": 5.814121179744248e-07,
+      "loss": 0.4675,
+      "step": 17174
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 17.22273095093739,
+      "learning_rate": 5.80845606735656e-07,
+      "loss": 0.3965,
+      "step": 17175
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.882530648626766,
+      "learning_rate": 5.802793633721304e-07,
+      "loss": 0.5005,
+      "step": 17176
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.156578297933518,
+      "learning_rate": 5.797133878999484e-07,
+      "loss": 0.5105,
+      "step": 17177
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.17856966759336,
+      "learning_rate": 5.791476803352058e-07,
+      "loss": 0.3771,
+      "step": 17178
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 23.08959148490535,
+      "learning_rate": 5.785822406939934e-07,
+      "loss": 0.4779,
+      "step": 17179
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 33.8620337184854,
+      "learning_rate": 5.780170689923902e-07,
+      "loss": 0.477,
+      "step": 17180
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 21.712296244686986,
+      "learning_rate": 5.774521652464715e-07,
+      "loss": 0.4023,
+      "step": 17181
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.410673496091725,
+      "learning_rate": 5.768875294722987e-07,
+      "loss": 0.4226,
+      "step": 17182
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 30.004939134636484,
+      "learning_rate": 5.763231616859333e-07,
+      "loss": 0.4388,
+      "step": 17183
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.166109351432183,
+      "learning_rate": 5.757590619034236e-07,
+      "loss": 0.4416,
+      "step": 17184
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.43622743482468,
+      "learning_rate": 5.751952301408148e-07,
+      "loss": 0.4666,
+      "step": 17185
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.429942193469923,
+      "learning_rate": 5.746316664141394e-07,
+      "loss": 0.3997,
+      "step": 17186
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.667216781736396,
+      "learning_rate": 5.740683707394256e-07,
+      "loss": 0.3777,
+      "step": 17187
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.43765278814437,
+      "learning_rate": 5.735053431326931e-07,
+      "loss": 0.4392,
+      "step": 17188
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 28.492888375320824,
+      "learning_rate": 5.729425836099556e-07,
+      "loss": 0.4162,
+      "step": 17189
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 22.973017063949513,
+      "learning_rate": 5.723800921872147e-07,
+      "loss": 0.3937,
+      "step": 17190
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.098068979821022,
+      "learning_rate": 5.718178688804699e-07,
+      "loss": 0.4462,
+      "step": 17191
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.438251553087387,
+      "learning_rate": 5.712559137057094e-07,
+      "loss": 0.4125,
+      "step": 17192
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.594294559109997,
+      "learning_rate": 5.706942266789162e-07,
+      "loss": 0.4147,
+      "step": 17193
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.007051574218718,
+      "learning_rate": 5.701328078160606e-07,
+      "loss": 0.4217,
+      "step": 17194
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 28.82472530363959,
+      "learning_rate": 5.695716571331134e-07,
+      "loss": 0.4465,
+      "step": 17195
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.328374989759716,
+      "learning_rate": 5.690107746460316e-07,
+      "loss": 0.4311,
+      "step": 17196
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 28.769635109024442,
+      "learning_rate": 5.684501603707671e-07,
+      "loss": 0.4654,
+      "step": 17197
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 33.43446302437116,
+      "learning_rate": 5.678898143232614e-07,
+      "loss": 0.4013,
+      "step": 17198
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.164500736159326,
+      "learning_rate": 5.673297365194508e-07,
+      "loss": 0.4557,
+      "step": 17199
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.908793173259525,
+      "learning_rate": 5.667699269752658e-07,
+      "loss": 0.5235,
+      "step": 17200
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 17.83536273800507,
+      "learning_rate": 5.66210385706627e-07,
+      "loss": 0.3905,
+      "step": 17201
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.226073741441912,
+      "learning_rate": 5.65651112729445e-07,
+      "loss": 0.4059,
+      "step": 17202
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.18215302873196,
+      "learning_rate": 5.650921080596261e-07,
+      "loss": 0.4531,
+      "step": 17203
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.992592838627985,
+      "learning_rate": 5.645333717130685e-07,
+      "loss": 0.3939,
+      "step": 17204
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.218793059280017,
+      "learning_rate": 5.63974903705663e-07,
+      "loss": 0.4742,
+      "step": 17205
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.178349945725383,
+      "learning_rate": 5.634167040532922e-07,
+      "loss": 0.4189,
+      "step": 17206
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 38.621292027323015,
+      "learning_rate": 5.628587727718282e-07,
+      "loss": 0.4737,
+      "step": 17207
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.099672147540957,
+      "learning_rate": 5.623011098771391e-07,
+      "loss": 0.4318,
+      "step": 17208
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 22.830996172230464,
+      "learning_rate": 5.617437153850868e-07,
+      "loss": 0.4269,
+      "step": 17209
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 25.3327741915449,
+      "learning_rate": 5.611865893115243e-07,
+      "loss": 0.4992,
+      "step": 17210
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 17.42511808120721,
+      "learning_rate": 5.60629731672292e-07,
+      "loss": 0.4356,
+      "step": 17211
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 30.328067335322576,
+      "learning_rate": 5.600731424832273e-07,
+      "loss": 0.4427,
+      "step": 17212
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.915881625417073,
+      "learning_rate": 5.595168217601599e-07,
+      "loss": 0.4896,
+      "step": 17213
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.41446837647619,
+      "learning_rate": 5.589607695189136e-07,
+      "loss": 0.401,
+      "step": 17214
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.635645349931963,
+      "learning_rate": 5.584049857752993e-07,
+      "loss": 0.4354,
+      "step": 17215
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 68.99571866595082,
+      "learning_rate": 5.578494705451232e-07,
+      "loss": 0.5034,
+      "step": 17216
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.323278459515997,
+      "learning_rate": 5.572942238441847e-07,
+      "loss": 0.3747,
+      "step": 17217
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 22.21319699409457,
+      "learning_rate": 5.567392456882758e-07,
+      "loss": 0.4258,
+      "step": 17218
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.219894182453054,
+      "learning_rate": 5.561845360931784e-07,
+      "loss": 0.4587,
+      "step": 17219
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.46678305962246,
+      "learning_rate": 5.556300950746684e-07,
+      "loss": 0.409,
+      "step": 17220
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.733542125577035,
+      "learning_rate": 5.550759226485126e-07,
+      "loss": 0.508,
+      "step": 17221
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.443719254239705,
+      "learning_rate": 5.545220188304723e-07,
+      "loss": 0.3717,
+      "step": 17222
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.11651088669281,
+      "learning_rate": 5.539683836363019e-07,
+      "loss": 0.4335,
+      "step": 17223
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.367746253091465,
+      "learning_rate": 5.534150170817431e-07,
+      "loss": 0.3919,
+      "step": 17224
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.107033985613825,
+      "learning_rate": 5.52861919182538e-07,
+      "loss": 0.5067,
+      "step": 17225
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 25.81298360590927,
+      "learning_rate": 5.523090899544104e-07,
+      "loss": 0.3868,
+      "step": 17226
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 22.31718210774328,
+      "learning_rate": 5.517565294130877e-07,
+      "loss": 0.5,
+      "step": 17227
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 36.27488346861624,
+      "learning_rate": 5.51204237574281e-07,
+      "loss": 0.448,
+      "step": 17228
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.9085475282073,
+      "learning_rate": 5.506522144536975e-07,
+      "loss": 0.4176,
+      "step": 17229
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 27.731048607476463,
+      "learning_rate": 5.501004600670401e-07,
+      "loss": 0.4536,
+      "step": 17230
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.953519242620448,
+      "learning_rate": 5.495489744299965e-07,
+      "loss": 0.4309,
+      "step": 17231
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.110863016909242,
+      "learning_rate": 5.489977575582506e-07,
+      "loss": 0.3961,
+      "step": 17232
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.042106379043073,
+      "learning_rate": 5.4844680946748e-07,
+      "loss": 0.3987,
+      "step": 17233
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.635019648119307,
+      "learning_rate": 5.478961301733531e-07,
+      "loss": 0.4655,
+      "step": 17234
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.573123024646776,
+      "learning_rate": 5.473457196915332e-07,
+      "loss": 0.4032,
+      "step": 17235
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.71383649078869,
+      "learning_rate": 5.467955780376688e-07,
+      "loss": 0.424,
+      "step": 17236
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 32.21400051926023,
+      "learning_rate": 5.462457052274084e-07,
+      "loss": 0.4757,
+      "step": 17237
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.520777585629546,
+      "learning_rate": 5.456961012763906e-07,
+      "loss": 0.4792,
+      "step": 17238
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.843145934995647,
+      "learning_rate": 5.451467662002452e-07,
+      "loss": 0.4313,
+      "step": 17239
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 30.24670527999096,
+      "learning_rate": 5.445977000145952e-07,
+      "loss": 0.4449,
+      "step": 17240
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 20.887309504627083,
+      "learning_rate": 5.440489027350548e-07,
+      "loss": 0.4383,
+      "step": 17241
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.896255175288672,
+      "learning_rate": 5.435003743772305e-07,
+      "loss": 0.4563,
+      "step": 17242
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 15.937445373067547,
+      "learning_rate": 5.429521149567263e-07,
+      "loss": 0.4015,
+      "step": 17243
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 17.625888676085836,
+      "learning_rate": 5.424041244891298e-07,
+      "loss": 0.397,
+      "step": 17244
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 25.574446228644778,
+      "learning_rate": 5.418564029900286e-07,
+      "loss": 0.5172,
+      "step": 17245
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.00128933528401,
+      "learning_rate": 5.413089504749979e-07,
+      "loss": 0.4048,
+      "step": 17246
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.8865913729184,
+      "learning_rate": 5.407617669596066e-07,
+      "loss": 0.4687,
+      "step": 17247
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.791511789683064,
+      "learning_rate": 5.402148524594198e-07,
+      "loss": 0.4571,
+      "step": 17248
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.402168720117135,
+      "learning_rate": 5.396682069899861e-07,
+      "loss": 0.3675,
+      "step": 17249
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.038101778678637,
+      "learning_rate": 5.391218305668566e-07,
+      "loss": 0.4146,
+      "step": 17250
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.849226415182248,
+      "learning_rate": 5.385757232055655e-07,
+      "loss": 0.4134,
+      "step": 17251
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 20.840673754751307,
+      "learning_rate": 5.380298849216481e-07,
+      "loss": 0.4323,
+      "step": 17252
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.10852311346565,
+      "learning_rate": 5.374843157306253e-07,
+      "loss": 0.392,
+      "step": 17253
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.352512086178507,
+      "learning_rate": 5.369390156480126e-07,
+      "loss": 0.4611,
+      "step": 17254
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.725599810451797,
+      "learning_rate": 5.363939846893184e-07,
+      "loss": 0.4011,
+      "step": 17255
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.855961721961,
+      "learning_rate": 5.35849222870044e-07,
+      "loss": 0.465,
+      "step": 17256
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.07508804166514,
+      "learning_rate": 5.353047302056802e-07,
+      "loss": 0.4646,
+      "step": 17257
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.49336396481214,
+      "learning_rate": 5.347605067117134e-07,
+      "loss": 0.3989,
+      "step": 17258
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.49373117952393,
+      "learning_rate": 5.342165524036224e-07,
+      "loss": 0.4778,
+      "step": 17259
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.8572480771274,
+      "learning_rate": 5.33672867296875e-07,
+      "loss": 0.4625,
+      "step": 17260
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.662540146654848,
+      "learning_rate": 5.331294514069318e-07,
+      "loss": 0.4573,
+      "step": 17261
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 27.067351887518708,
+      "learning_rate": 5.325863047492496e-07,
+      "loss": 0.4378,
+      "step": 17262
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 23.64252606732429,
+      "learning_rate": 5.320434273392738e-07,
+      "loss": 0.4271,
+      "step": 17263
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.61442315262329,
+      "learning_rate": 5.315008191924464e-07,
+      "loss": 0.4566,
+      "step": 17264
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.27401069011807,
+      "learning_rate": 5.309584803241973e-07,
+      "loss": 0.4084,
+      "step": 17265
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 27.305309391170642,
+      "learning_rate": 5.304164107499477e-07,
+      "loss": 0.4578,
+      "step": 17266
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.952895964459437,
+      "learning_rate": 5.29874610485116e-07,
+      "loss": 0.379,
+      "step": 17267
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.262228147157717,
+      "learning_rate": 5.293330795451112e-07,
+      "loss": 0.4231,
+      "step": 17268
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 25.623663939559407,
+      "learning_rate": 5.287918179453344e-07,
+      "loss": 0.5229,
+      "step": 17269
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.013308274250857,
+      "learning_rate": 5.282508257011764e-07,
+      "loss": 0.4515,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.646344360743818,
+      "learning_rate": 5.27710102828024e-07,
+      "loss": 0.3797,
+      "step": 17271
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 26.91870644510296,
+      "learning_rate": 5.271696493412548e-07,
+      "loss": 0.4468,
+      "step": 17272
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.81347788521876,
+      "learning_rate": 5.266294652562409e-07,
+      "loss": 0.4902,
+      "step": 17273
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.685796437551218,
+      "learning_rate": 5.260895505883423e-07,
+      "loss": 0.4316,
+      "step": 17274
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.414744642470925,
+      "learning_rate": 5.255499053529145e-07,
+      "loss": 0.4286,
+      "step": 17275
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 34.59554654223095,
+      "learning_rate": 5.250105295653052e-07,
+      "loss": 0.3948,
+      "step": 17276
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 31.880595986161005,
+      "learning_rate": 5.244714232408544e-07,
+      "loss": 0.463,
+      "step": 17277
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.55392232655942,
+      "learning_rate": 5.23932586394893e-07,
+      "loss": 0.4426,
+      "step": 17278
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.202779860189478,
+      "learning_rate": 5.233940190427456e-07,
+      "loss": 0.4516,
+      "step": 17279
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.452537788284626,
+      "learning_rate": 5.228557211997276e-07,
+      "loss": 0.3993,
+      "step": 17280
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.07642177092423,
+      "learning_rate": 5.223176928811502e-07,
+      "loss": 0.4256,
+      "step": 17281
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.108226470057977,
+      "learning_rate": 5.217799341023122e-07,
+      "loss": 0.3946,
+      "step": 17282
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.841880061435862,
+      "learning_rate": 5.212424448785092e-07,
+      "loss": 0.4574,
+      "step": 17283
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 27.884641124032992,
+      "learning_rate": 5.207052252250266e-07,
+      "loss": 0.4703,
+      "step": 17284
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.883343293380833,
+      "learning_rate": 5.201682751571402e-07,
+      "loss": 0.4807,
+      "step": 17285
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.402956927065283,
+      "learning_rate": 5.196315946901254e-07,
+      "loss": 0.4299,
+      "step": 17286
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.7772194234793,
+      "learning_rate": 5.1909518383924e-07,
+      "loss": 0.4514,
+      "step": 17287
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.206078323145508,
+      "learning_rate": 5.185590426197406e-07,
+      "loss": 0.4958,
+      "step": 17288
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.493932784104928,
+      "learning_rate": 5.180231710468775e-07,
+      "loss": 0.4336,
+      "step": 17289
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.3163464168877,
+      "learning_rate": 5.174875691358894e-07,
+      "loss": 0.4442,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.3136152806508,
+      "learning_rate": 5.169522369020052e-07,
+      "loss": 0.4246,
+      "step": 17291
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 23.089451561457448,
+      "learning_rate": 5.164171743604529e-07,
+      "loss": 0.5384,
+      "step": 17292
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.51344816114946,
+      "learning_rate": 5.158823815264491e-07,
+      "loss": 0.445,
+      "step": 17293
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 30.97963196097292,
+      "learning_rate": 5.153478584152028e-07,
+      "loss": 0.4123,
+      "step": 17294
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.122142329044262,
+      "learning_rate": 5.148136050419139e-07,
+      "loss": 0.4176,
+      "step": 17295
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 17.935988362513783,
+      "learning_rate": 5.14279621421776e-07,
+      "loss": 0.3834,
+      "step": 17296
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.94540360780038,
+      "learning_rate": 5.13745907569978e-07,
+      "loss": 0.5192,
+      "step": 17297
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.993483268059627,
+      "learning_rate": 5.132124635016977e-07,
+      "loss": 0.4088,
+      "step": 17298
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 33.0552618562427,
+      "learning_rate": 5.126792892321064e-07,
+      "loss": 0.4924,
+      "step": 17299
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.015626259569817,
+      "learning_rate": 5.121463847763641e-07,
+      "loss": 0.442,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.004528866612098,
+      "learning_rate": 5.116137501496288e-07,
+      "loss": 0.4004,
+      "step": 17301
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 30.554486900496265,
+      "learning_rate": 5.110813853670482e-07,
+      "loss": 0.451,
+      "step": 17302
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.652322624921526,
+      "learning_rate": 5.105492904437636e-07,
+      "loss": 0.4929,
+      "step": 17303
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.202941979260043,
+      "learning_rate": 5.100174653949031e-07,
+      "loss": 0.4504,
+      "step": 17304
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.288963161211036,
+      "learning_rate": 5.094859102355953e-07,
+      "loss": 0.3748,
+      "step": 17305
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.11259556259227,
+      "learning_rate": 5.089546249809584e-07,
+      "loss": 0.4413,
+      "step": 17306
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 39.29029262048367,
+      "learning_rate": 5.08423609646097e-07,
+      "loss": 0.4246,
+      "step": 17307
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.690272692919102,
+      "learning_rate": 5.078928642461178e-07,
+      "loss": 0.4498,
+      "step": 17308
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.085286966064466,
+      "learning_rate": 5.073623887961121e-07,
+      "loss": 0.4678,
+      "step": 17309
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.842060203629657,
+      "learning_rate": 5.068321833111667e-07,
+      "loss": 0.4252,
+      "step": 17310
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.359827984218143,
+      "learning_rate": 5.063022478063617e-07,
+      "loss": 0.3983,
+      "step": 17311
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.35831420844723,
+      "learning_rate": 5.057725822967663e-07,
+      "loss": 0.4112,
+      "step": 17312
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.621102548928857,
+      "learning_rate": 5.05243186797445e-07,
+      "loss": 0.4945,
+      "step": 17313
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.109947128184082,
+      "learning_rate": 5.047140613234524e-07,
+      "loss": 0.3994,
+      "step": 17314
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.81475669737337,
+      "learning_rate": 5.041852058898389e-07,
+      "loss": 0.4711,
+      "step": 17315
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.809144760978814,
+      "learning_rate": 5.036566205116422e-07,
+      "loss": 0.4196,
+      "step": 17316
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.800104107147042,
+      "learning_rate": 5.031283052038949e-07,
+      "loss": 0.4705,
+      "step": 17317
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.571628230613605,
+      "learning_rate": 5.026002599816248e-07,
+      "loss": 0.4439,
+      "step": 17318
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.640442659638463,
+      "learning_rate": 5.02072484859848e-07,
+      "loss": 0.4091,
+      "step": 17319
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.543029136953702,
+      "learning_rate": 5.015449798535721e-07,
+      "loss": 0.4187,
+      "step": 17320
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 26.946093871726127,
+      "learning_rate": 5.010177449778009e-07,
+      "loss": 0.5137,
+      "step": 17321
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 28.47027117143808,
+      "learning_rate": 5.004907802475278e-07,
+      "loss": 0.4929,
+      "step": 17322
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 20.634373193810383,
+      "learning_rate": 4.999640856777421e-07,
+      "loss": 0.4675,
+      "step": 17323
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.77781522887984,
+      "learning_rate": 4.994376612834185e-07,
+      "loss": 0.4162,
+      "step": 17324
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.850471273597794,
+      "learning_rate": 4.989115070795303e-07,
+      "loss": 0.4495,
+      "step": 17325
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.957795243371162,
+      "learning_rate": 4.983856230810402e-07,
+      "loss": 0.5059,
+      "step": 17326
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.66063598569354,
+      "learning_rate": 4.978600093029062e-07,
+      "loss": 0.4327,
+      "step": 17327
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 26.61425020913945,
+      "learning_rate": 4.973346657600752e-07,
+      "loss": 0.4447,
+      "step": 17328
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.46283718401621,
+      "learning_rate": 4.968095924674854e-07,
+      "loss": 0.4474,
+      "step": 17329
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 25.523382559309375,
+      "learning_rate": 4.962847894400724e-07,
+      "loss": 0.4229,
+      "step": 17330
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 39.79228124626586,
+      "learning_rate": 4.95760256692761e-07,
+      "loss": 0.5159,
+      "step": 17331
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 25.669889569488955,
+      "learning_rate": 4.952359942404672e-07,
+      "loss": 0.4275,
+      "step": 17332
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.177408284016053,
+      "learning_rate": 4.947120020981034e-07,
+      "loss": 0.3933,
+      "step": 17333
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.664859572605161,
+      "learning_rate": 4.941882802805675e-07,
+      "loss": 0.4255,
+      "step": 17334
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 24.90021194727825,
+      "learning_rate": 4.936648288027568e-07,
+      "loss": 0.4106,
+      "step": 17335
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.637261371792675,
+      "learning_rate": 4.931416476795592e-07,
+      "loss": 0.4176,
+      "step": 17336
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 20.438985157442556,
+      "learning_rate": 4.926187369258495e-07,
+      "loss": 0.4423,
+      "step": 17337
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.940822931403154,
+      "learning_rate": 4.920960965565036e-07,
+      "loss": 0.4107,
+      "step": 17338
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.160953824999524,
+      "learning_rate": 4.915737265863807e-07,
+      "loss": 0.4597,
+      "step": 17339
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.406864437603977,
+      "learning_rate": 4.9105162703034e-07,
+      "loss": 0.3596,
+      "step": 17340
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.127785650216396,
+      "learning_rate": 4.905297979032264e-07,
+      "loss": 0.4405,
+      "step": 17341
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.541078316342178,
+      "learning_rate": 4.900082392198835e-07,
+      "loss": 0.4054,
+      "step": 17342
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.943470575163786,
+      "learning_rate": 4.894869509951449e-07,
+      "loss": 0.3533,
+      "step": 17343
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.57018770795412,
+      "learning_rate": 4.889659332438334e-07,
+      "loss": 0.4153,
+      "step": 17344
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 24.409024522917367,
+      "learning_rate": 4.884451859807648e-07,
+      "loss": 0.4465,
+      "step": 17345
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 16.92758619534769,
+      "learning_rate": 4.879247092207518e-07,
+      "loss": 0.4631,
+      "step": 17346
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 24.209903256734332,
+      "learning_rate": 4.874045029785957e-07,
+      "loss": 0.41,
+      "step": 17347
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.104014462954876,
+      "learning_rate": 4.868845672690937e-07,
+      "loss": 0.4574,
+      "step": 17348
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 25.669795151158045,
+      "learning_rate": 4.863649021070261e-07,
+      "loss": 0.4538,
+      "step": 17349
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 17.90745147532213,
+      "learning_rate": 4.858455075071766e-07,
+      "loss": 0.4072,
+      "step": 17350
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.577945008540926,
+      "learning_rate": 4.853263834843136e-07,
+      "loss": 0.3928,
+      "step": 17351
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.459120343549202,
+      "learning_rate": 4.848075300532051e-07,
+      "loss": 0.4506,
+      "step": 17352
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 16.599486700094594,
+      "learning_rate": 4.842889472286039e-07,
+      "loss": 0.4168,
+      "step": 17353
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 20.776638771370983,
+      "learning_rate": 4.837706350252569e-07,
+      "loss": 0.4259,
+      "step": 17354
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.100395334157533,
+      "learning_rate": 4.832525934579058e-07,
+      "loss": 0.399,
+      "step": 17355
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 33.67534898065224,
+      "learning_rate": 4.827348225412864e-07,
+      "loss": 0.4403,
+      "step": 17356
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 30.387288894532016,
+      "learning_rate": 4.822173222901194e-07,
+      "loss": 0.4431,
+      "step": 17357
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.940554654277092,
+      "learning_rate": 4.81700092719124e-07,
+      "loss": 0.4197,
+      "step": 17358
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 26.50262252296206,
+      "learning_rate": 4.811831338430095e-07,
+      "loss": 0.3843,
+      "step": 17359
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.92135082594988,
+      "learning_rate": 4.806664456764787e-07,
+      "loss": 0.4721,
+      "step": 17360
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 28.6315027965235,
+      "learning_rate": 4.801500282342264e-07,
+      "loss": 0.4465,
+      "step": 17361
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 27.82184890108278,
+      "learning_rate": 4.796338815309387e-07,
+      "loss": 0.4856,
+      "step": 17362
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.897452871933826,
+      "learning_rate": 4.791180055812928e-07,
+      "loss": 0.4594,
+      "step": 17363
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 24.95841888144139,
+      "learning_rate": 4.786024003999612e-07,
+      "loss": 0.4399,
+      "step": 17364
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 19.214807093508366,
+      "learning_rate": 4.780870660016091e-07,
+      "loss": 0.4338,
+      "step": 17365
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 24.18263758656307,
+      "learning_rate": 4.77572002400889e-07,
+      "loss": 0.4697,
+      "step": 17366
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.136460204437313,
+      "learning_rate": 4.770572096124515e-07,
+      "loss": 0.421,
+      "step": 17367
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.01549275061047,
+      "learning_rate": 4.7654268765093604e-07,
+      "loss": 0.4369,
+      "step": 17368
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.06620930056725,
+      "learning_rate": 4.760284365309753e-07,
+      "loss": 0.4101,
+      "step": 17369
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 29.334126203470845,
+      "learning_rate": 4.755144562671943e-07,
+      "loss": 0.4209,
+      "step": 17370
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 31.54646969006299,
+      "learning_rate": 4.7500074687421017e-07,
+      "loss": 0.4521,
+      "step": 17371
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 39.968513085421456,
+      "learning_rate": 4.744873083666346e-07,
+      "loss": 0.5315,
+      "step": 17372
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 24.247894269115044,
+      "learning_rate": 4.739741407590659e-07,
+      "loss": 0.4935,
+      "step": 17373
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 25.56898532853315,
+      "learning_rate": 4.7346124406610125e-07,
+      "loss": 0.4109,
+      "step": 17374
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 27.765093978651173,
+      "learning_rate": 4.729486183023246e-07,
+      "loss": 0.4593,
+      "step": 17375
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.762721227724175,
+      "learning_rate": 4.724362634823165e-07,
+      "loss": 0.4055,
+      "step": 17376
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 28.816990680308916,
+      "learning_rate": 4.7192417962064865e-07,
+      "loss": 0.439,
+      "step": 17377
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.274613292861563,
+      "learning_rate": 4.714123667318837e-07,
+      "loss": 0.4647,
+      "step": 17378
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.5723381890236676,
+      "learning_rate": 4.7090082483057577e-07,
+      "loss": 0.4093,
+      "step": 17379
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 22.15896494174935,
+      "learning_rate": 4.7038955393127306e-07,
+      "loss": 0.4018,
+      "step": 17380
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.399819966163685,
+      "learning_rate": 4.6987855404851955e-07,
+      "loss": 0.4273,
+      "step": 17381
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.085045005489427,
+      "learning_rate": 4.6936782519684365e-07,
+      "loss": 0.4396,
+      "step": 17382
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.70013423262256,
+      "learning_rate": 4.6885736739077146e-07,
+      "loss": 0.4026,
+      "step": 17383
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 30.19754895395075,
+      "learning_rate": 4.683471806448192e-07,
+      "loss": 0.4879,
+      "step": 17384
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 27.371877122015096,
+      "learning_rate": 4.6783726497349747e-07,
+      "loss": 0.5188,
+      "step": 17385
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.821238949350676,
+      "learning_rate": 4.673276203913091e-07,
+      "loss": 0.4776,
+      "step": 17386
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 23.906197476796812,
+      "learning_rate": 4.668182469127469e-07,
+      "loss": 0.4316,
+      "step": 17387
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.890558692781248,
+      "learning_rate": 4.6630914455229493e-07,
+      "loss": 0.4068,
+      "step": 17388
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.79089253736976,
+      "learning_rate": 4.6580031332443487e-07,
+      "loss": 0.4102,
+      "step": 17389
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 27.523150481244254,
+      "learning_rate": 4.652917532436374e-07,
+      "loss": 0.4516,
+      "step": 17390
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.306461668556436,
+      "learning_rate": 4.6478346432436426e-07,
+      "loss": 0.3934,
+      "step": 17391
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.29821462189279,
+      "learning_rate": 4.642754465810717e-07,
+      "loss": 0.4605,
+      "step": 17392
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 31.493702984748477,
+      "learning_rate": 4.6376770002820593e-07,
+      "loss": 0.4108,
+      "step": 17393
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 26.095407241458357,
+      "learning_rate": 4.632602246802109e-07,
+      "loss": 0.4612,
+      "step": 17394
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.646601453076475,
+      "learning_rate": 4.6275302055151293e-07,
+      "loss": 0.3693,
+      "step": 17395
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.10464786368163,
+      "learning_rate": 4.6224608765654153e-07,
+      "loss": 0.454,
+      "step": 17396
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.524280938816876,
+      "learning_rate": 4.61739426009713e-07,
+      "loss": 0.464,
+      "step": 17397
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.6426630919563,
+      "learning_rate": 4.612330356254335e-07,
+      "loss": 0.458,
+      "step": 17398
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 26.19173882235598,
+      "learning_rate": 4.607269165181083e-07,
+      "loss": 0.4576,
+      "step": 17399
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 22.28478750981967,
+      "learning_rate": 4.6022106870212804e-07,
+      "loss": 0.4246,
+      "step": 17400
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 28.18063883475103,
+      "learning_rate": 4.597154921918812e-07,
+      "loss": 0.4302,
+      "step": 17401
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.225910386650213,
+      "learning_rate": 4.592101870017429e-07,
+      "loss": 0.4514,
+      "step": 17402
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.98886841096316,
+      "learning_rate": 4.587051531460873e-07,
+      "loss": 0.4558,
+      "step": 17403
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 36.02204642984345,
+      "learning_rate": 4.5820039063927514e-07,
+      "loss": 0.4453,
+      "step": 17404
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 44.551800977650416,
+      "learning_rate": 4.5769589949566153e-07,
+      "loss": 0.4731,
+      "step": 17405
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.904202026760487,
+      "learning_rate": 4.57191679729595e-07,
+      "loss": 0.4218,
+      "step": 17406
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.949185903881208,
+      "learning_rate": 4.566877313554152e-07,
+      "loss": 0.4611,
+      "step": 17407
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 23.986283792343404,
+      "learning_rate": 4.561840543874529e-07,
+      "loss": 0.4625,
+      "step": 17408
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.171183132603076,
+      "learning_rate": 4.5568064884003337e-07,
+      "loss": 0.4917,
+      "step": 17409
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.441739249149823,
+      "learning_rate": 4.551775147274717e-07,
+      "loss": 0.4176,
+      "step": 17410
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 28.75684373970585,
+      "learning_rate": 4.5467465206408103e-07,
+      "loss": 0.4153,
+      "step": 17411
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.389540213874831,
+      "learning_rate": 4.541720608641575e-07,
+      "loss": 0.3785,
+      "step": 17412
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.016112998412776,
+      "learning_rate": 4.5366974114199546e-07,
+      "loss": 0.4106,
+      "step": 17413
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.931950525588608,
+      "learning_rate": 4.5316769291188223e-07,
+      "loss": 0.4221,
+      "step": 17414
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.240826921865555,
+      "learning_rate": 4.526659161880964e-07,
+      "loss": 0.434,
+      "step": 17415
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 31.13718517903119,
+      "learning_rate": 4.521644109849066e-07,
+      "loss": 0.4298,
+      "step": 17416
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.48098421548587,
+      "learning_rate": 4.5166317731657363e-07,
+      "loss": 0.4454,
+      "step": 17417
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.434729211114256,
+      "learning_rate": 4.5116221519735493e-07,
+      "loss": 0.4622,
+      "step": 17418
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.51366634119941,
+      "learning_rate": 4.5066152464149914e-07,
+      "loss": 0.444,
+      "step": 17419
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.712331832276472,
+      "learning_rate": 4.5016110566324044e-07,
+      "loss": 0.3636,
+      "step": 17420
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.2683599113328,
+      "learning_rate": 4.4966095827681524e-07,
+      "loss": 0.4288,
+      "step": 17421
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.892715665588616,
+      "learning_rate": 4.491610824964454e-07,
+      "loss": 0.4047,
+      "step": 17422
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.20558238929761,
+      "learning_rate": 4.486614783363463e-07,
+      "loss": 0.408,
+      "step": 17423
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.223760404762384,
+      "learning_rate": 4.481621458107288e-07,
+      "loss": 0.5298,
+      "step": 17424
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.68640664998534,
+      "learning_rate": 4.476630849337904e-07,
+      "loss": 0.4707,
+      "step": 17425
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 26.581314671462636,
+      "learning_rate": 4.471642957197275e-07,
+      "loss": 0.4379,
+      "step": 17426
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.671837457419493,
+      "learning_rate": 4.4666577818272327e-07,
+      "loss": 0.5048,
+      "step": 17427
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 28.38455317499184,
+      "learning_rate": 4.461675323369563e-07,
+      "loss": 0.4704,
+      "step": 17428
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.656114082300245,
+      "learning_rate": 4.456695581965942e-07,
+      "loss": 0.3824,
+      "step": 17429
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.952204471225592,
+      "learning_rate": 4.4517185577580226e-07,
+      "loss": 0.3957,
+      "step": 17430
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.677067259212578,
+      "learning_rate": 4.446744250887336e-07,
+      "loss": 0.4821,
+      "step": 17431
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.978744868491923,
+      "learning_rate": 4.441772661495347e-07,
+      "loss": 0.4703,
+      "step": 17432
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.496449254990207,
+      "learning_rate": 4.4368037897234317e-07,
+      "loss": 0.4496,
+      "step": 17433
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.877871280257406,
+      "learning_rate": 4.4318376357129103e-07,
+      "loss": 0.409,
+      "step": 17434
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.01844481682169,
+      "learning_rate": 4.4268741996050244e-07,
+      "loss": 0.5087,
+      "step": 17435
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.261647695163436,
+      "learning_rate": 4.421913481540929e-07,
+      "loss": 0.4911,
+      "step": 17436
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.478901469878732,
+      "learning_rate": 4.4169554816617224e-07,
+      "loss": 0.3723,
+      "step": 17437
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.70580459692987,
+      "learning_rate": 4.4120002001083575e-07,
+      "loss": 0.4246,
+      "step": 17438
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 22.589175481408834,
+      "learning_rate": 4.4070476370218e-07,
+      "loss": 0.4291,
+      "step": 17439
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.679987959219805,
+      "learning_rate": 4.402097792542892e-07,
+      "loss": 0.3547,
+      "step": 17440
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.396166383333188,
+      "learning_rate": 4.397150666812411e-07,
+      "loss": 0.4053,
+      "step": 17441
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.025061248631832,
+      "learning_rate": 4.392206259971021e-07,
+      "loss": 0.4068,
+      "step": 17442
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.413034781841418,
+      "learning_rate": 4.3872645721593556e-07,
+      "loss": 0.4552,
+      "step": 17443
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 30.036685707204484,
+      "learning_rate": 4.382325603517956e-07,
+      "loss": 0.4983,
+      "step": 17444
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.63665511456837,
+      "learning_rate": 4.3773893541873005e-07,
+      "loss": 0.4048,
+      "step": 17445
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 25.225865457520595,
+      "learning_rate": 4.372455824307731e-07,
+      "loss": 0.4295,
+      "step": 17446
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 24.630164364921853,
+      "learning_rate": 4.36752501401958e-07,
+      "loss": 0.4304,
+      "step": 17447
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 25.856865290300433,
+      "learning_rate": 4.3625969234630694e-07,
+      "loss": 0.4567,
+      "step": 17448
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.071757241203127,
+      "learning_rate": 4.3576715527783755e-07,
+      "loss": 0.3787,
+      "step": 17449
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.501434511668156,
+      "learning_rate": 4.3527489021055414e-07,
+      "loss": 0.3805,
+      "step": 17450
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.617238735253185,
+      "learning_rate": 4.347828971584578e-07,
+      "loss": 0.4372,
+      "step": 17451
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.29021742708971,
+      "learning_rate": 4.342911761355395e-07,
+      "loss": 0.3763,
+      "step": 17452
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 19.58187449331832,
+      "learning_rate": 4.3379972715578587e-07,
+      "loss": 0.4352,
+      "step": 17453
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 27.83295509223901,
+      "learning_rate": 4.333085502331713e-07,
+      "loss": 0.4433,
+      "step": 17454
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.9373031023361,
+      "learning_rate": 4.328176453816657e-07,
+      "loss": 0.4278,
+      "step": 17455
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.0500319547027,
+      "learning_rate": 4.32327012615229e-07,
+      "loss": 0.4205,
+      "step": 17456
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.327597823533026,
+      "learning_rate": 4.3183665194781564e-07,
+      "loss": 0.4277,
+      "step": 17457
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 32.74199777947312,
+      "learning_rate": 4.3134656339337e-07,
+      "loss": 0.4478,
+      "step": 17458
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 30.692695535246823,
+      "learning_rate": 4.308567469658298e-07,
+      "loss": 0.4943,
+      "step": 17459
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.9940120098935,
+      "learning_rate": 4.3036720267912833e-07,
+      "loss": 0.4214,
+      "step": 17460
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.34305745650129,
+      "learning_rate": 4.2987793054718343e-07,
+      "loss": 0.4019,
+      "step": 17461
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.68511982658362,
+      "learning_rate": 4.2938893058391385e-07,
+      "loss": 0.3996,
+      "step": 17462
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 24.150543381469717,
+      "learning_rate": 4.289002028032219e-07,
+      "loss": 0.4458,
+      "step": 17463
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 28.56440195766985,
+      "learning_rate": 4.284117472190108e-07,
+      "loss": 0.4547,
+      "step": 17464
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.686787437429523,
+      "learning_rate": 4.2792356384517063e-07,
+      "loss": 0.3967,
+      "step": 17465
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.997359866756085,
+      "learning_rate": 4.2743565269558475e-07,
+      "loss": 0.4226,
+      "step": 17466
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.35440114600762,
+      "learning_rate": 4.2694801378412867e-07,
+      "loss": 0.412,
+      "step": 17467
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.008676945964883,
+      "learning_rate": 4.2646064712467016e-07,
+      "loss": 0.485,
+      "step": 17468
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.297305301680044,
+      "learning_rate": 4.2597355273107267e-07,
+      "loss": 0.4124,
+      "step": 17469
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.817850589987756,
+      "learning_rate": 4.2548673061718614e-07,
+      "loss": 0.4001,
+      "step": 17470
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.62385161757375,
+      "learning_rate": 4.25000180796854e-07,
+      "loss": 0.4426,
+      "step": 17471
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.607778978955473,
+      "learning_rate": 4.245139032839163e-07,
+      "loss": 0.4312,
+      "step": 17472
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 26.70832361383802,
+      "learning_rate": 4.240278980922019e-07,
+      "loss": 0.4486,
+      "step": 17473
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.118358524866036,
+      "learning_rate": 4.2354216523553314e-07,
+      "loss": 0.4001,
+      "step": 17474
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 29.531044915915153,
+      "learning_rate": 4.230567047277234e-07,
+      "loss": 0.3881,
+      "step": 17475
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.110147238950233,
+      "learning_rate": 4.2257151658257724e-07,
+      "loss": 0.4405,
+      "step": 17476
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 27.63439163061701,
+      "learning_rate": 4.2208660081389463e-07,
+      "loss": 0.4667,
+      "step": 17477
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 25.894108281843828,
+      "learning_rate": 4.2160195743546797e-07,
+      "loss": 0.3913,
+      "step": 17478
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 24.97732492786066,
+      "learning_rate": 4.2111758646107617e-07,
+      "loss": 0.4572,
+      "step": 17479
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 14.659188849784474,
+      "learning_rate": 4.2063348790449823e-07,
+      "loss": 0.375,
+      "step": 17480
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.860107567251355,
+      "learning_rate": 4.201496617794998e-07,
+      "loss": 0.4538,
+      "step": 17481
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.708013347464217,
+      "learning_rate": 4.19666108099841e-07,
+      "loss": 0.4527,
+      "step": 17482
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 24.922707518062698,
+      "learning_rate": 4.1918282687927304e-07,
+      "loss": 0.4257,
+      "step": 17483
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.934500199316048,
+      "learning_rate": 4.186998181315405e-07,
+      "loss": 0.4286,
+      "step": 17484
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.812547378621478,
+      "learning_rate": 4.1821708187038236e-07,
+      "loss": 0.4758,
+      "step": 17485
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 25.638804827331093,
+      "learning_rate": 4.177346181095232e-07,
+      "loss": 0.42,
+      "step": 17486
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.398371528714584,
+      "learning_rate": 4.172524268626876e-07,
+      "loss": 0.4493,
+      "step": 17487
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 28.63009658880959,
+      "learning_rate": 4.167705081435858e-07,
+      "loss": 0.4378,
+      "step": 17488
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 27.343390325633028,
+      "learning_rate": 4.162888619659244e-07,
+      "loss": 0.4357,
+      "step": 17489
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 27.446397579123143,
+      "learning_rate": 4.158074883434038e-07,
+      "loss": 0.5091,
+      "step": 17490
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.738836955756252,
+      "learning_rate": 4.1532638728971065e-07,
+      "loss": 0.3896,
+      "step": 17491
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.552671287215457,
+      "learning_rate": 4.148455588185274e-07,
+      "loss": 0.3835,
+      "step": 17492
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 17.479494448313787,
+      "learning_rate": 4.143650029435287e-07,
+      "loss": 0.5104,
+      "step": 17493
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.20213456466183,
+      "learning_rate": 4.138847196783835e-07,
+      "loss": 0.3914,
+      "step": 17494
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 36.61430820435879,
+      "learning_rate": 4.134047090367488e-07,
+      "loss": 0.4465,
+      "step": 17495
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 32.111421649333316,
+      "learning_rate": 4.129249710322758e-07,
+      "loss": 0.4553,
+      "step": 17496
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 29.861711339807677,
+      "learning_rate": 4.124455056786081e-07,
+      "loss": 0.4718,
+      "step": 17497
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 21.267011851593708,
+      "learning_rate": 4.119663129893814e-07,
+      "loss": 0.4152,
+      "step": 17498
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.956345237191314,
+      "learning_rate": 4.114873929782259e-07,
+      "loss": 0.4485,
+      "step": 17499
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.606313338359346,
+      "learning_rate": 4.1100874565875745e-07,
+      "loss": 0.4006,
+      "step": 17500
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.98487525168327,
+      "learning_rate": 4.105303710445918e-07,
+      "loss": 0.4085,
+      "step": 17501
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 19.179010733572685,
+      "learning_rate": 4.100522691493325e-07,
+      "loss": 0.3811,
+      "step": 17502
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 34.54975317185062,
+      "learning_rate": 4.0957443998657754e-07,
+      "loss": 0.4092,
+      "step": 17503
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.810297646989035,
+      "learning_rate": 4.0909688356991495e-07,
+      "loss": 0.4638,
+      "step": 17504
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 17.72673533655055,
+      "learning_rate": 4.086195999129261e-07,
+      "loss": 0.3704,
+      "step": 17505
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 21.70174804765831,
+      "learning_rate": 4.081425890291846e-07,
+      "loss": 0.4318,
+      "step": 17506
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.040318933655207,
+      "learning_rate": 4.076658509322573e-07,
+      "loss": 0.3723,
+      "step": 17507
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 27.408368450702692,
+      "learning_rate": 4.0718938563570233e-07,
+      "loss": 0.4915,
+      "step": 17508
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 31.30847192503798,
+      "learning_rate": 4.067131931530699e-07,
+      "loss": 0.4231,
+      "step": 17509
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 39.510686933138025,
+      "learning_rate": 4.0623727349790034e-07,
+      "loss": 0.4788,
+      "step": 17510
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.3988233895468,
+      "learning_rate": 4.0576162668373164e-07,
+      "loss": 0.4353,
+      "step": 17511
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.015188683340405,
+      "learning_rate": 4.0528625272409083e-07,
+      "loss": 0.5014,
+      "step": 17512
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 27.904477455087267,
+      "learning_rate": 4.0481115163249483e-07,
+      "loss": 0.4241,
+      "step": 17513
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.779880772563885,
+      "learning_rate": 4.0433632342245735e-07,
+      "loss": 0.4098,
+      "step": 17514
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.437373304966645,
+      "learning_rate": 4.038617681074808e-07,
+      "loss": 0.3925,
+      "step": 17515
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 29.490459909934003,
+      "learning_rate": 4.0338748570106333e-07,
+      "loss": 0.4817,
+      "step": 17516
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 26.63319283121807,
+      "learning_rate": 4.0291347621669084e-07,
+      "loss": 0.4922,
+      "step": 17517
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.398392111616047,
+      "learning_rate": 4.0243973966784477e-07,
+      "loss": 0.3934,
+      "step": 17518
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.90392494787278,
+      "learning_rate": 4.019662760679988e-07,
+      "loss": 0.412,
+      "step": 17519
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.04840720045855,
+      "learning_rate": 4.014930854306176e-07,
+      "loss": 0.47,
+      "step": 17520
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.888840216364056,
+      "learning_rate": 4.010201677691572e-07,
+      "loss": 0.4637,
+      "step": 17521
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 40.307585703550416,
+      "learning_rate": 4.005475230970679e-07,
+      "loss": 0.4092,
+      "step": 17522
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.070339697621073,
+      "learning_rate": 4.000751514277912e-07,
+      "loss": 0.4502,
+      "step": 17523
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 25.372528754479347,
+      "learning_rate": 3.9960305277476294e-07,
+      "loss": 0.4244,
+      "step": 17524
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.54098039265196,
+      "learning_rate": 3.9913122715140804e-07,
+      "loss": 0.5074,
+      "step": 17525
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 42.24266727359193,
+      "learning_rate": 3.986596745711424e-07,
+      "loss": 0.53,
+      "step": 17526
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.467002523587144,
+      "learning_rate": 3.9818839504737974e-07,
+      "loss": 0.4493,
+      "step": 17527
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.461490860852557,
+      "learning_rate": 3.977173885935237e-07,
+      "loss": 0.3873,
+      "step": 17528
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.942109732590932,
+      "learning_rate": 3.972466552229681e-07,
+      "loss": 0.4736,
+      "step": 17529
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.932327548759254,
+      "learning_rate": 3.967761949490978e-07,
+      "loss": 0.4134,
+      "step": 17530
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.789274291083245,
+      "learning_rate": 3.963060077852965e-07,
+      "loss": 0.422,
+      "step": 17531
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 32.12679830575649,
+      "learning_rate": 3.958360937449335e-07,
+      "loss": 0.4283,
+      "step": 17532
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.195025399161796,
+      "learning_rate": 3.953664528413737e-07,
+      "loss": 0.4202,
+      "step": 17533
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.35008528858892,
+      "learning_rate": 3.948970850879752e-07,
+      "loss": 0.5014,
+      "step": 17534
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.336119845520958,
+      "learning_rate": 3.94427990498083e-07,
+      "loss": 0.3942,
+      "step": 17535
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 26.739048433966083,
+      "learning_rate": 3.9395916908503863e-07,
+      "loss": 0.4088,
+      "step": 17536
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 37.68300077363878,
+      "learning_rate": 3.9349062086217915e-07,
+      "loss": 0.4558,
+      "step": 17537
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 32.05418614695762,
+      "learning_rate": 3.930223458428239e-07,
+      "loss": 0.4647,
+      "step": 17538
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.312230954455963,
+      "learning_rate": 3.9255434404029567e-07,
+      "loss": 0.4275,
+      "step": 17539
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.36721689766602,
+      "learning_rate": 3.9208661546789927e-07,
+      "loss": 0.4145,
+      "step": 17540
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.389367887074666,
+      "learning_rate": 3.9161916013894186e-07,
+      "loss": 0.4328,
+      "step": 17541
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 32.79270567615219,
+      "learning_rate": 3.9115197806671277e-07,
+      "loss": 0.4913,
+      "step": 17542
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.452098591277245,
+      "learning_rate": 3.9068506926450146e-07,
+      "loss": 0.4522,
+      "step": 17543
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.44429547840225,
+      "learning_rate": 3.9021843374558385e-07,
+      "loss": 0.4402,
+      "step": 17544
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.951166938881396,
+      "learning_rate": 3.897520715232339e-07,
+      "loss": 0.4318,
+      "step": 17545
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.28365225407174,
+      "learning_rate": 3.8928598261071313e-07,
+      "loss": 0.3915,
+      "step": 17546
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.111141844508477,
+      "learning_rate": 3.8882016702127544e-07,
+      "loss": 0.4442,
+      "step": 17547
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.799639862152425,
+      "learning_rate": 3.883546247681713e-07,
+      "loss": 0.4386,
+      "step": 17548
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.933889535364973,
+      "learning_rate": 3.8788935586463906e-07,
+      "loss": 0.3988,
+      "step": 17549
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.307817436950778,
+      "learning_rate": 3.8742436032391027e-07,
+      "loss": 0.4015,
+      "step": 17550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.48341164372044,
+      "learning_rate": 3.8695963815920887e-07,
+      "loss": 0.4712,
+      "step": 17551
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.99712163074532,
+      "learning_rate": 3.864951893837521e-07,
+      "loss": 0.4125,
+      "step": 17552
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.410124987344634,
+      "learning_rate": 3.8603101401074927e-07,
+      "loss": 0.4214,
+      "step": 17553
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.453021551439846,
+      "learning_rate": 3.855671120533999e-07,
+      "loss": 0.4249,
+      "step": 17554
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.90595372189574,
+      "learning_rate": 3.851034835248979e-07,
+      "loss": 0.4297,
+      "step": 17555
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.302956417423683,
+      "learning_rate": 3.8464012843842714e-07,
+      "loss": 0.4296,
+      "step": 17556
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.57739614835241,
+      "learning_rate": 3.8417704680716704e-07,
+      "loss": 0.4344,
+      "step": 17557
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.784546681606066,
+      "learning_rate": 3.8371423864428826e-07,
+      "loss": 0.4267,
+      "step": 17558
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 24.219278412253438,
+      "learning_rate": 3.8325170396294907e-07,
+      "loss": 0.4416,
+      "step": 17559
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.641724036574267,
+      "learning_rate": 3.827894427763057e-07,
+      "loss": 0.4565,
+      "step": 17560
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.717063972618014,
+      "learning_rate": 3.8232745509750423e-07,
+      "loss": 0.4422,
+      "step": 17561
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.222039680477796,
+      "learning_rate": 3.818657409396853e-07,
+      "loss": 0.4581,
+      "step": 17562
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.941673798151543,
+      "learning_rate": 3.8140430031597844e-07,
+      "loss": 0.4145,
+      "step": 17563
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.126597698868775,
+      "learning_rate": 3.809431332395053e-07,
+      "loss": 0.4306,
+      "step": 17564
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.492216500980064,
+      "learning_rate": 3.804822397233832e-07,
+      "loss": 0.3688,
+      "step": 17565
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 23.945840111002425,
+      "learning_rate": 3.800216197807194e-07,
+      "loss": 0.4492,
+      "step": 17566
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.556779334890756,
+      "learning_rate": 3.7956127342461127e-07,
+      "loss": 0.4067,
+      "step": 17567
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.31972029118095,
+      "learning_rate": 3.791012006681549e-07,
+      "loss": 0.4254,
+      "step": 17568
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 27.829320157970898,
+      "learning_rate": 3.786414015244311e-07,
+      "loss": 0.4306,
+      "step": 17569
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 27.47234722090023,
+      "learning_rate": 3.781818760065181e-07,
+      "loss": 0.4251,
+      "step": 17570
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.16782867231646,
+      "learning_rate": 3.777226241274834e-07,
+      "loss": 0.4364,
+      "step": 17571
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.465201876310815,
+      "learning_rate": 3.772636459003887e-07,
+      "loss": 0.434,
+      "step": 17572
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.965029729306384,
+      "learning_rate": 3.768049413382868e-07,
+      "loss": 0.3755,
+      "step": 17573
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.79009514268903,
+      "learning_rate": 3.76346510454223e-07,
+      "loss": 0.4047,
+      "step": 17574
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 35.86451489298629,
+      "learning_rate": 3.758883532612356e-07,
+      "loss": 0.4105,
+      "step": 17575
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 30.348565907862902,
+      "learning_rate": 3.754304697723521e-07,
+      "loss": 0.4163,
+      "step": 17576
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 27.512065754730056,
+      "learning_rate": 3.749728600005953e-07,
+      "loss": 0.3804,
+      "step": 17577
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 24.34524124876427,
+      "learning_rate": 3.745155239589815e-07,
+      "loss": 0.3923,
+      "step": 17578
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.767678273590956,
+      "learning_rate": 3.7405846166051585e-07,
+      "loss": 0.4289,
+      "step": 17579
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 23.07189445613623,
+      "learning_rate": 3.736016731181946e-07,
+      "loss": 0.4136,
+      "step": 17580
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.27953030308091,
+      "learning_rate": 3.7314515834501075e-07,
+      "loss": 0.4527,
+      "step": 17581
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.27630390969021,
+      "learning_rate": 3.726889173539483e-07,
+      "loss": 0.4432,
+      "step": 17582
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 29.0326694145637,
+      "learning_rate": 3.7223295015798133e-07,
+      "loss": 0.5179,
+      "step": 17583
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.664408589420702,
+      "learning_rate": 3.7177725677007615e-07,
+      "loss": 0.3659,
+      "step": 17584
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 25.147560411856237,
+      "learning_rate": 3.713218372031935e-07,
+      "loss": 0.5466,
+      "step": 17585
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.922662537900592,
+      "learning_rate": 3.708666914702852e-07,
+      "loss": 0.4031,
+      "step": 17586
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 23.310389091427545,
+      "learning_rate": 3.704118195842965e-07,
+      "loss": 0.4602,
+      "step": 17587
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.95408639254217,
+      "learning_rate": 3.699572215581615e-07,
+      "loss": 0.4711,
+      "step": 17588
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 29.79057878706563,
+      "learning_rate": 3.695028974048098e-07,
+      "loss": 0.5159,
+      "step": 17589
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.45883003666137,
+      "learning_rate": 3.6904884713716114e-07,
+      "loss": 0.3995,
+      "step": 17590
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 29.49774046908421,
+      "learning_rate": 3.6859507076813073e-07,
+      "loss": 0.4374,
+      "step": 17591
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 38.878012121898465,
+      "learning_rate": 3.6814156831062264e-07,
+      "loss": 0.4094,
+      "step": 17592
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.39699222270177,
+      "learning_rate": 3.6768833977753214e-07,
+      "loss": 0.4804,
+      "step": 17593
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 14.367121750254363,
+      "learning_rate": 3.672353851817512e-07,
+      "loss": 0.4046,
+      "step": 17594
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 23.538314317472818,
+      "learning_rate": 3.667827045361616e-07,
+      "loss": 0.512,
+      "step": 17595
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 28.10235191596009,
+      "learning_rate": 3.663302978536354e-07,
+      "loss": 0.3986,
+      "step": 17596
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 27.12716722345551,
+      "learning_rate": 3.658781651470422e-07,
+      "loss": 0.4979,
+      "step": 17597
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 20.902512735201658,
+      "learning_rate": 3.654263064292363e-07,
+      "loss": 0.4153,
+      "step": 17598
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.508206423505467,
+      "learning_rate": 3.649747217130695e-07,
+      "loss": 0.4354,
+      "step": 17599
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.13389008572231,
+      "learning_rate": 3.645234110113871e-07,
+      "loss": 0.3795,
+      "step": 17600
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.074589900520056,
+      "learning_rate": 3.6407237433702116e-07,
+      "loss": 0.4382,
+      "step": 17601
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 31.822803042074202,
+      "learning_rate": 3.636216117028013e-07,
+      "loss": 0.4385,
+      "step": 17602
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.846443769012538,
+      "learning_rate": 3.63171123121544e-07,
+      "loss": 0.3833,
+      "step": 17603
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 24.389533195032417,
+      "learning_rate": 3.627209086060635e-07,
+      "loss": 0.3963,
+      "step": 17604
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 20.93648974130658,
+      "learning_rate": 3.6227096816916274e-07,
+      "loss": 0.4219,
+      "step": 17605
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.453932798052776,
+      "learning_rate": 3.6182130182363716e-07,
+      "loss": 0.449,
+      "step": 17606
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.16350660663446,
+      "learning_rate": 3.6137190958227655e-07,
+      "loss": 0.3822,
+      "step": 17607
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 24.11960044964342,
+      "learning_rate": 3.609227914578606e-07,
+      "loss": 0.4434,
+      "step": 17608
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 35.27148795923943,
+      "learning_rate": 3.6047394746316023e-07,
+      "loss": 0.4816,
+      "step": 17609
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 28.587634090214188,
+      "learning_rate": 3.6002537761094193e-07,
+      "loss": 0.4546,
+      "step": 17610
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 29.720156458838176,
+      "learning_rate": 3.595770819139632e-07,
+      "loss": 0.4599,
+      "step": 17611
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 24.607773131052568,
+      "learning_rate": 3.5912906038497287e-07,
+      "loss": 0.5316,
+      "step": 17612
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 20.001750958060477,
+      "learning_rate": 3.5868131303671393e-07,
+      "loss": 0.4304,
+      "step": 17613
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.677063960201608,
+      "learning_rate": 3.582338398819163e-07,
+      "loss": 0.3819,
+      "step": 17614
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.54366094423086,
+      "learning_rate": 3.577866409333075e-07,
+      "loss": 0.3947,
+      "step": 17615
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.872411805571907,
+      "learning_rate": 3.5733971620360853e-07,
+      "loss": 0.4644,
+      "step": 17616
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 27.67797548079895,
+      "learning_rate": 3.5689306570552584e-07,
+      "loss": 0.4206,
+      "step": 17617
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.201829078776536,
+      "learning_rate": 3.5644668945176264e-07,
+      "loss": 0.4066,
+      "step": 17618
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 25.251090222725587,
+      "learning_rate": 3.5600058745501543e-07,
+      "loss": 0.4283,
+      "step": 17619
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 26.384651325590177,
+      "learning_rate": 3.555547597279696e-07,
+      "loss": 0.4019,
+      "step": 17620
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 24.472285705221083,
+      "learning_rate": 3.551092062833039e-07,
+      "loss": 0.4387,
+      "step": 17621
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.641588680221716,
+      "learning_rate": 3.546639271336916e-07,
+      "loss": 0.4783,
+      "step": 17622
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 19.627109028541238,
+      "learning_rate": 3.5421892229179354e-07,
+      "loss": 0.4309,
+      "step": 17623
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 23.303779076656976,
+      "learning_rate": 3.537741917702664e-07,
+      "loss": 0.3577,
+      "step": 17624
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.95909705822597,
+      "learning_rate": 3.5332973558175996e-07,
+      "loss": 0.4548,
+      "step": 17625
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 36.23098984089105,
+      "learning_rate": 3.52885553738912e-07,
+      "loss": 0.4801,
+      "step": 17626
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 17.545105013485518,
+      "learning_rate": 3.524416462543556e-07,
+      "loss": 0.3947,
+      "step": 17627
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 17.51014834664943,
+      "learning_rate": 3.519980131407152e-07,
+      "loss": 0.4006,
+      "step": 17628
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 41.87052157353007,
+      "learning_rate": 3.515546544106074e-07,
+      "loss": 0.3838,
+      "step": 17629
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.070821691892426,
+      "learning_rate": 3.5111157007663986e-07,
+      "loss": 0.3933,
+      "step": 17630
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 23.179131830676027,
+      "learning_rate": 3.506687601514158e-07,
+      "loss": 0.3897,
+      "step": 17631
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.031478653390828,
+      "learning_rate": 3.502262246475285e-07,
+      "loss": 0.4245,
+      "step": 17632
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 20.278895357513118,
+      "learning_rate": 3.497839635775613e-07,
+      "loss": 0.3874,
+      "step": 17633
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 24.700946334179584,
+      "learning_rate": 3.493419769540929e-07,
+      "loss": 0.4042,
+      "step": 17634
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 26.533459846565975,
+      "learning_rate": 3.489002647896933e-07,
+      "loss": 0.3964,
+      "step": 17635
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 20.071327751024754,
+      "learning_rate": 3.4845882709692356e-07,
+      "loss": 0.4315,
+      "step": 17636
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.20121481290997,
+      "learning_rate": 3.4801766388834037e-07,
+      "loss": 0.424,
+      "step": 17637
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.653808124428775,
+      "learning_rate": 3.4757677517648916e-07,
+      "loss": 0.438,
+      "step": 17638
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 17.261791980119334,
+      "learning_rate": 3.471361609739055e-07,
+      "loss": 0.4069,
+      "step": 17639
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 13.676234698956899,
+      "learning_rate": 3.4669582129312373e-07,
+      "loss": 0.4399,
+      "step": 17640
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.76643994187275,
+      "learning_rate": 3.462557561466662e-07,
+      "loss": 0.3931,
+      "step": 17641
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 39.28658178867971,
+      "learning_rate": 3.458159655470483e-07,
+      "loss": 0.4604,
+      "step": 17642
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 17.26384231158706,
+      "learning_rate": 3.4537644950677576e-07,
+      "loss": 0.4436,
+      "step": 17643
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.873303179177018,
+      "learning_rate": 3.449372080383484e-07,
+      "loss": 0.4575,
+      "step": 17644
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.13398775509945,
+      "learning_rate": 3.4449824115426076e-07,
+      "loss": 0.4997,
+      "step": 17645
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.411864926831946,
+      "learning_rate": 3.440595488669951e-07,
+      "loss": 0.3692,
+      "step": 17646
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 22.910887637094188,
+      "learning_rate": 3.4362113118902475e-07,
+      "loss": 0.4282,
+      "step": 17647
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.869032780226853,
+      "learning_rate": 3.4318298813282193e-07,
+      "loss": 0.4278,
+      "step": 17648
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 28.135475465928945,
+      "learning_rate": 3.4274511971084556e-07,
+      "loss": 0.4167,
+      "step": 17649
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.30215450484469,
+      "learning_rate": 3.423075259355491e-07,
+      "loss": 0.3607,
+      "step": 17650
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 40.711977948129025,
+      "learning_rate": 3.41870206819378e-07,
+      "loss": 0.5404,
+      "step": 17651
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.00839845113648,
+      "learning_rate": 3.4143316237476686e-07,
+      "loss": 0.3767,
+      "step": 17652
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.35476609541522,
+      "learning_rate": 3.409963926141457e-07,
+      "loss": 0.4266,
+      "step": 17653
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.065449786604272,
+      "learning_rate": 3.4055989754993913e-07,
+      "loss": 0.385,
+      "step": 17654
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.312340432053816,
+      "learning_rate": 3.4012367719455595e-07,
+      "loss": 0.4318,
+      "step": 17655
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 26.78982832485094,
+      "learning_rate": 3.3968773156040745e-07,
+      "loss": 0.4785,
+      "step": 17656
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 24.784937148331927,
+      "learning_rate": 3.392520606598859e-07,
+      "loss": 0.403,
+      "step": 17657
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 27.538531470623646,
+      "learning_rate": 3.3881666450538695e-07,
+      "loss": 0.4635,
+      "step": 17658
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 23.1561216262008,
+      "learning_rate": 3.383815431092885e-07,
+      "loss": 0.3897,
+      "step": 17659
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.400811593688012,
+      "learning_rate": 3.3794669648396727e-07,
+      "loss": 0.387,
+      "step": 17660
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.950143112675182,
+      "learning_rate": 3.3751212464179007e-07,
+      "loss": 0.5017,
+      "step": 17661
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 16.941965160337542,
+      "learning_rate": 3.370778275951159e-07,
+      "loss": 0.4648,
+      "step": 17662
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 27.33928110279793,
+      "learning_rate": 3.3664380535629595e-07,
+      "loss": 0.4601,
+      "step": 17663
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.795253560765687,
+      "learning_rate": 3.362100579376726e-07,
+      "loss": 0.4437,
+      "step": 17664
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 37.16993754871001,
+      "learning_rate": 3.357765853515815e-07,
+      "loss": 0.4154,
+      "step": 17665
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.075425902714972,
+      "learning_rate": 3.353433876103529e-07,
+      "loss": 0.4378,
+      "step": 17666
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 16.344221736489434,
+      "learning_rate": 3.349104647263046e-07,
+      "loss": 0.4249,
+      "step": 17667
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 22.077663558881213,
+      "learning_rate": 3.3447781671174684e-07,
+      "loss": 0.3781,
+      "step": 17668
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 22.78743554873619,
+      "learning_rate": 3.340454435789864e-07,
+      "loss": 0.4619,
+      "step": 17669
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 17.302493432926585,
+      "learning_rate": 3.336133453403201e-07,
+      "loss": 0.4299,
+      "step": 17670
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.881522257396004,
+      "learning_rate": 3.331815220080359e-07,
+      "loss": 0.4149,
+      "step": 17671
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.621173031454553,
+      "learning_rate": 3.3274997359441174e-07,
+      "loss": 0.4128,
+      "step": 17672
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.368764571547246,
+      "learning_rate": 3.323187001117245e-07,
+      "loss": 0.4402,
+      "step": 17673
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 27.149741829475385,
+      "learning_rate": 3.318877015722377e-07,
+      "loss": 0.5089,
+      "step": 17674
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.583278227643845,
+      "learning_rate": 3.3145697798821044e-07,
+      "loss": 0.4237,
+      "step": 17675
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 42.922950104870736,
+      "learning_rate": 3.3102652937189064e-07,
+      "loss": 0.5172,
+      "step": 17676
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 26.244669910682696,
+      "learning_rate": 3.3059635573551854e-07,
+      "loss": 0.4231,
+      "step": 17677
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 26.483313130159356,
+      "learning_rate": 3.301664570913299e-07,
+      "loss": 0.4778,
+      "step": 17678
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 28.538593450242786,
+      "learning_rate": 3.297368334515516e-07,
+      "loss": 0.4695,
+      "step": 17679
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 38.310937172454736,
+      "learning_rate": 3.2930748482840056e-07,
+      "loss": 0.4796,
+      "step": 17680
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 15.126211821269354,
+      "learning_rate": 3.2887841123408817e-07,
+      "loss": 0.3916,
+      "step": 17681
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 25.02724307084497,
+      "learning_rate": 3.2844961268081457e-07,
+      "loss": 0.4375,
+      "step": 17682
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.57181444122784,
+      "learning_rate": 3.2802108918077777e-07,
+      "loss": 0.3595,
+      "step": 17683
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 32.14063490115,
+      "learning_rate": 3.275928407461615e-07,
+      "loss": 0.4169,
+      "step": 17684
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 23.790765076268357,
+      "learning_rate": 3.271648673891481e-07,
+      "loss": 0.4441,
+      "step": 17685
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 24.983642147497704,
+      "learning_rate": 3.267371691219068e-07,
+      "loss": 0.4119,
+      "step": 17686
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 30.6282771940744,
+      "learning_rate": 3.2630974595660226e-07,
+      "loss": 0.4309,
+      "step": 17687
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.755118661930588,
+      "learning_rate": 3.2588259790538923e-07,
+      "loss": 0.4263,
+      "step": 17688
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 28.05257720361429,
+      "learning_rate": 3.2545572498041576e-07,
+      "loss": 0.4567,
+      "step": 17689
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 24.982562950701627,
+      "learning_rate": 3.2502912719382327e-07,
+      "loss": 0.4216,
+      "step": 17690
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 17.59691952127479,
+      "learning_rate": 3.2460280455774096e-07,
+      "loss": 0.3996,
+      "step": 17691
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.05470851463165,
+      "learning_rate": 3.2417675708429574e-07,
+      "loss": 0.4324,
+      "step": 17692
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 22.1273676903876,
+      "learning_rate": 3.237509847856035e-07,
+      "loss": 0.4311,
+      "step": 17693
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 23.049775757003967,
+      "learning_rate": 3.2332548767377237e-07,
+      "loss": 0.3871,
+      "step": 17694
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 37.63828354627713,
+      "learning_rate": 3.229002657609037e-07,
+      "loss": 0.403,
+      "step": 17695
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.827930429193863,
+      "learning_rate": 3.224753190590912e-07,
+      "loss": 0.3944,
+      "step": 17696
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.935924212600288,
+      "learning_rate": 3.220506475804186e-07,
+      "loss": 0.3809,
+      "step": 17697
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 45.98717967743818,
+      "learning_rate": 3.21626251336965e-07,
+      "loss": 0.4497,
+      "step": 17698
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.917288850400528,
+      "learning_rate": 3.212021303407975e-07,
+      "loss": 0.4398,
+      "step": 17699
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 27.201496306671956,
+      "learning_rate": 3.20778284603982e-07,
+      "loss": 0.5111,
+      "step": 17700
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.294124470951655,
+      "learning_rate": 3.203547141385688e-07,
+      "loss": 0.4211,
+      "step": 17701
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 25.9780568712668,
+      "learning_rate": 3.1993141895660385e-07,
+      "loss": 0.4014,
+      "step": 17702
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.963547719935852,
+      "learning_rate": 3.1950839907012755e-07,
+      "loss": 0.3566,
+      "step": 17703
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.675189528343374,
+      "learning_rate": 3.1908565449117026e-07,
+      "loss": 0.3755,
+      "step": 17704
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 27.725663897973195,
+      "learning_rate": 3.1866318523175344e-07,
+      "loss": 0.4299,
+      "step": 17705
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 33.222484769384224,
+      "learning_rate": 3.18240991303892e-07,
+      "loss": 0.5161,
+      "step": 17706
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 31.508710530122755,
+      "learning_rate": 3.178190727195918e-07,
+      "loss": 0.3962,
+      "step": 17707
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 23.482210403450914,
+      "learning_rate": 3.173974294908555e-07,
+      "loss": 0.4355,
+      "step": 17708
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.52061158129383,
+      "learning_rate": 3.169760616296702e-07,
+      "loss": 0.4382,
+      "step": 17709
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 24.220009148382967,
+      "learning_rate": 3.1655496914802296e-07,
+      "loss": 0.4598,
+      "step": 17710
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 27.42700087521264,
+      "learning_rate": 3.1613415205788643e-07,
+      "loss": 0.5331,
+      "step": 17711
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 37.06494454046371,
+      "learning_rate": 3.1571361037122996e-07,
+      "loss": 0.4245,
+      "step": 17712
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 19.8939808698795,
+      "learning_rate": 3.152933441000139e-07,
+      "loss": 0.4783,
+      "step": 17713
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 24.52823995905179,
+      "learning_rate": 3.148733532561887e-07,
+      "loss": 0.4167,
+      "step": 17714
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 29.28182344170294,
+      "learning_rate": 3.1445363785170155e-07,
+      "loss": 0.4548,
+      "step": 17715
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 40.379343304478844,
+      "learning_rate": 3.140341978984851e-07,
+      "loss": 0.5007,
+      "step": 17716
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 25.06055781813353,
+      "learning_rate": 3.13615033408472e-07,
+      "loss": 0.4108,
+      "step": 17717
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 17.125701212101035,
+      "learning_rate": 3.1319614439357827e-07,
+      "loss": 0.3849,
+      "step": 17718
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 22.820831586764452,
+      "learning_rate": 3.1277753086572105e-07,
+      "loss": 0.4135,
+      "step": 17719
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 22.55893056345587,
+      "learning_rate": 3.123591928368042e-07,
+      "loss": 0.4514,
+      "step": 17720
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 24.306792548790323,
+      "learning_rate": 3.1194113031872584e-07,
+      "loss": 0.4446,
+      "step": 17721
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.925897571174666,
+      "learning_rate": 3.115233433233722e-07,
+      "loss": 0.4256,
+      "step": 17722
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.986215810505279,
+      "learning_rate": 3.111058318626281e-07,
+      "loss": 0.3723,
+      "step": 17723
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 25.33960368579462,
+      "learning_rate": 3.106885959483652e-07,
+      "loss": 0.3996,
+      "step": 17724
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 22.430890990748615,
+      "learning_rate": 3.102716355924518e-07,
+      "loss": 0.4419,
+      "step": 17725
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 34.25280472790416,
+      "learning_rate": 3.09854950806745e-07,
+      "loss": 0.4201,
+      "step": 17726
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 23.198934483621148,
+      "learning_rate": 3.094385416030943e-07,
+      "loss": 0.4546,
+      "step": 17727
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.194276718604623,
+      "learning_rate": 3.0902240799334237e-07,
+      "loss": 0.4304,
+      "step": 17728
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 27.617083893275396,
+      "learning_rate": 3.0860654998932536e-07,
+      "loss": 0.4229,
+      "step": 17729
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.292763872018554,
+      "learning_rate": 3.0819096760286826e-07,
+      "loss": 0.4271,
+      "step": 17730
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.192417339603097,
+      "learning_rate": 3.0777566084578934e-07,
+      "loss": 0.3836,
+      "step": 17731
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 27.763047490997057,
+      "learning_rate": 3.0736062972990145e-07,
+      "loss": 0.4651,
+      "step": 17732
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.461177569174026,
+      "learning_rate": 3.069458742670073e-07,
+      "loss": 0.4566,
+      "step": 17733
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.637765681173875,
+      "learning_rate": 3.065313944689019e-07,
+      "loss": 0.4697,
+      "step": 17734
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.065984463969304,
+      "learning_rate": 3.0611719034737363e-07,
+      "loss": 0.3595,
+      "step": 17735
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 20.578260994811572,
+      "learning_rate": 3.05703261914202e-07,
+      "loss": 0.4632,
+      "step": 17736
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 22.242904327948846,
+      "learning_rate": 3.0528960918115744e-07,
+      "loss": 0.4301,
+      "step": 17737
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 29.37043488924438,
+      "learning_rate": 3.048762321600063e-07,
+      "loss": 0.4664,
+      "step": 17738
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 20.084968414431373,
+      "learning_rate": 3.0446313086250346e-07,
+      "loss": 0.4456,
+      "step": 17739
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 31.489069310879728,
+      "learning_rate": 3.0405030530039734e-07,
+      "loss": 0.4943,
+      "step": 17740
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.18330621369408,
+      "learning_rate": 3.0363775548542753e-07,
+      "loss": 0.5134,
+      "step": 17741
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.600441304473748,
+      "learning_rate": 3.0322548142932893e-07,
+      "loss": 0.4052,
+      "step": 17742
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.945506302469074,
+      "learning_rate": 3.0281348314382453e-07,
+      "loss": 0.4587,
+      "step": 17743
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.23798674940019,
+      "learning_rate": 3.0240176064063266e-07,
+      "loss": 0.4016,
+      "step": 17744
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 23.22841361426061,
+      "learning_rate": 3.0199031393146174e-07,
+      "loss": 0.5186,
+      "step": 17745
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.262510650425778,
+      "learning_rate": 3.015791430280135e-07,
+      "loss": 0.4451,
+      "step": 17746
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 34.379464918082775,
+      "learning_rate": 3.0116824794197977e-07,
+      "loss": 0.4826,
+      "step": 17747
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.097081391445617,
+      "learning_rate": 3.0075762868504777e-07,
+      "loss": 0.438,
+      "step": 17748
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 22.355576222488185,
+      "learning_rate": 3.003472852688949e-07,
+      "loss": 0.4465,
+      "step": 17749
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 17.42190627407466,
+      "learning_rate": 2.999372177051918e-07,
+      "loss": 0.3745,
+      "step": 17750
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.05724652493624,
+      "learning_rate": 2.9952742600559917e-07,
+      "loss": 0.4138,
+      "step": 17751
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 24.001924959113907,
+      "learning_rate": 2.9911791018177204e-07,
+      "loss": 0.4154,
+      "step": 17752
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.837550494465944,
+      "learning_rate": 2.9870867024535675e-07,
+      "loss": 0.4127,
+      "step": 17753
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 20.721680934436208,
+      "learning_rate": 2.9829970620799176e-07,
+      "loss": 0.448,
+      "step": 17754
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.171433216848833,
+      "learning_rate": 2.978910180813088e-07,
+      "loss": 0.3914,
+      "step": 17755
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.16145196297292,
+      "learning_rate": 2.9748260587692867e-07,
+      "loss": 0.4608,
+      "step": 17756
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 32.4257215006004,
+      "learning_rate": 2.9707446960646644e-07,
+      "loss": 0.4329,
+      "step": 17757
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 25.96223306307443,
+      "learning_rate": 2.966666092815318e-07,
+      "loss": 0.4311,
+      "step": 17758
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 19.596092187560284,
+      "learning_rate": 2.962590249137232e-07,
+      "loss": 0.4524,
+      "step": 17759
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.09075659731838,
+      "learning_rate": 2.958517165146302e-07,
+      "loss": 0.3782,
+      "step": 17760
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 32.97059382439807,
+      "learning_rate": 2.954446840958369e-07,
+      "loss": 0.4049,
+      "step": 17761
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 20.238962076232934,
+      "learning_rate": 2.950379276689197e-07,
+      "loss": 0.4492,
+      "step": 17762
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.979882544482278,
+      "learning_rate": 2.946314472454481e-07,
+      "loss": 0.4367,
+      "step": 17763
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 23.16373111494247,
+      "learning_rate": 2.9422524283698074e-07,
+      "loss": 0.479,
+      "step": 17764
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.527442167625203,
+      "learning_rate": 2.9381931445506826e-07,
+      "loss": 0.434,
+      "step": 17765
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 22.600064304142172,
+      "learning_rate": 2.934136621112571e-07,
+      "loss": 0.4528,
+      "step": 17766
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.506270650759074,
+      "learning_rate": 2.9300828581708464e-07,
+      "loss": 0.4098,
+      "step": 17767
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.06691320673563,
+      "learning_rate": 2.926031855840761e-07,
+      "loss": 0.4246,
+      "step": 17768
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 23.005656413037922,
+      "learning_rate": 2.9219836142375557e-07,
+      "loss": 0.3901,
+      "step": 17769
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.414776448735328,
+      "learning_rate": 2.917938133476339e-07,
+      "loss": 0.498,
+      "step": 17770
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 19.923262375038078,
+      "learning_rate": 2.913895413672174e-07,
+      "loss": 0.4637,
+      "step": 17771
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.83951212911085,
+      "learning_rate": 2.909855454940025e-07,
+      "loss": 0.431,
+      "step": 17772
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 32.079504482302205,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.4811,
+      "step": 17773
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.802675499478198,
+      "learning_rate": 2.901783821151305e-07,
+      "loss": 0.3864,
+      "step": 17774
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.217788936758698,
+      "learning_rate": 2.897752146324262e-07,
+      "loss": 0.4397,
+      "step": 17775
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 29.758415823459107,
+      "learning_rate": 2.893723233028367e-07,
+      "loss": 0.429,
+      "step": 17776
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.32663103816911,
+      "learning_rate": 2.889697081378162e-07,
+      "loss": 0.4532,
+      "step": 17777
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.447919822576537,
+      "learning_rate": 2.885673691488167e-07,
+      "loss": 0.4094,
+      "step": 17778
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 22.50888376324859,
+      "learning_rate": 2.8816530634728e-07,
+      "loss": 0.3904,
+      "step": 17779
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 28.601438018269526,
+      "learning_rate": 2.877635197446427e-07,
+      "loss": 0.4367,
+      "step": 17780
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 11.398677674365361,
+      "learning_rate": 2.873620093523266e-07,
+      "loss": 0.4287,
+      "step": 17781
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 19.87314811662009,
+      "learning_rate": 2.869607751817538e-07,
+      "loss": 0.482,
+      "step": 17782
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.94801116452939,
+      "learning_rate": 2.8655981724433625e-07,
+      "loss": 0.418,
+      "step": 17783
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.651201468766368,
+      "learning_rate": 2.861591355514748e-07,
+      "loss": 0.4093,
+      "step": 17784
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.469020688108923,
+      "learning_rate": 2.857587301145637e-07,
+      "loss": 0.4403,
+      "step": 17785
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.939070353022245,
+      "learning_rate": 2.853586009449927e-07,
+      "loss": 0.4085,
+      "step": 17786
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.162492029552457,
+      "learning_rate": 2.8495874805413937e-07,
+      "loss": 0.3995,
+      "step": 17787
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.75535694908689,
+      "learning_rate": 2.845591714533769e-07,
+      "loss": 0.4161,
+      "step": 17788
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.344400072171297,
+      "learning_rate": 2.841598711540694e-07,
+      "loss": 0.4065,
+      "step": 17789
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 31.825420460429424,
+      "learning_rate": 2.83760847167569e-07,
+      "loss": 0.4998,
+      "step": 17790
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.79331486690112,
+      "learning_rate": 2.8336209950522774e-07,
+      "loss": 0.4992,
+      "step": 17791
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 29.61541954485294,
+      "learning_rate": 2.829636281783843e-07,
+      "loss": 0.4473,
+      "step": 17792
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 27.110733862248356,
+      "learning_rate": 2.825654331983707e-07,
+      "loss": 0.4556,
+      "step": 17793
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 17.257772404834387,
+      "learning_rate": 2.8216751457651124e-07,
+      "loss": 0.4037,
+      "step": 17794
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 30.776153927814182,
+      "learning_rate": 2.8176987232412354e-07,
+      "loss": 0.4971,
+      "step": 17795
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 19.981997399943353,
+      "learning_rate": 2.8137250645251636e-07,
+      "loss": 0.4503,
+      "step": 17796
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.78206100896924,
+      "learning_rate": 2.8097541697298835e-07,
+      "loss": 0.4174,
+      "step": 17797
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.219222166950722,
+      "learning_rate": 2.80578603896835e-07,
+      "loss": 0.3872,
+      "step": 17798
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 13.712528351656207,
+      "learning_rate": 2.8018206723533947e-07,
+      "loss": 0.3565,
+      "step": 17799
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.620171348536875,
+      "learning_rate": 2.7978580699978054e-07,
+      "loss": 0.3975,
+      "step": 17800
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 28.43347515820931,
+      "learning_rate": 2.7938982320142693e-07,
+      "loss": 0.4835,
+      "step": 17801
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 33.41241235295184,
+      "learning_rate": 2.789941158515408e-07,
+      "loss": 0.4707,
+      "step": 17802
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 34.165840575894414,
+      "learning_rate": 2.785986849613753e-07,
+      "loss": 0.4095,
+      "step": 17803
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 22.33324719318697,
+      "learning_rate": 2.78203530542176e-07,
+      "loss": 0.4473,
+      "step": 17804
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.86358169649824,
+      "learning_rate": 2.7780865260518265e-07,
+      "loss": 0.4104,
+      "step": 17805
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 17.247536409893662,
+      "learning_rate": 2.774140511616219e-07,
+      "loss": 0.3936,
+      "step": 17806
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.528570963119602,
+      "learning_rate": 2.770197262227181e-07,
+      "loss": 0.4658,
+      "step": 17807
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.644543764248844,
+      "learning_rate": 2.7662567779968677e-07,
+      "loss": 0.4331,
+      "step": 17808
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.046297085607907,
+      "learning_rate": 2.762319059037333e-07,
+      "loss": 0.446,
+      "step": 17809
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.52821308456191,
+      "learning_rate": 2.758384105460554e-07,
+      "loss": 0.3938,
+      "step": 17810
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.39181994345637,
+      "learning_rate": 2.7544519173784423e-07,
+      "loss": 0.4082,
+      "step": 17811
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.79687175558414,
+      "learning_rate": 2.7505224949028297e-07,
+      "loss": 0.4646,
+      "step": 17812
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 30.372453472941118,
+      "learning_rate": 2.7465958381454714e-07,
+      "loss": 0.4564,
+      "step": 17813
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.662787697018608,
+      "learning_rate": 2.7426719472180454e-07,
+      "loss": 0.4665,
+      "step": 17814
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.80120540974446,
+      "learning_rate": 2.738750822232128e-07,
+      "loss": 0.4285,
+      "step": 17815
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 19.817918329884925,
+      "learning_rate": 2.7348324632992416e-07,
+      "loss": 0.3917,
+      "step": 17816
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.154300239838921,
+      "learning_rate": 2.7309168705308196e-07,
+      "loss": 0.3714,
+      "step": 17817
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 24.161499256870428,
+      "learning_rate": 2.7270040440382283e-07,
+      "loss": 0.4113,
+      "step": 17818
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.196009073951082,
+      "learning_rate": 2.723093983932734e-07,
+      "loss": 0.3644,
+      "step": 17819
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 35.6433101626748,
+      "learning_rate": 2.719186690325537e-07,
+      "loss": 0.4416,
+      "step": 17820
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 19.85555892314469,
+      "learning_rate": 2.7152821633277705e-07,
+      "loss": 0.418,
+      "step": 17821
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.78496606706317,
+      "learning_rate": 2.7113804030504564e-07,
+      "loss": 0.4408,
+      "step": 17822
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.062839569058323,
+      "learning_rate": 2.7074814096045954e-07,
+      "loss": 0.3768,
+      "step": 17823
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 14.511767594472179,
+      "learning_rate": 2.7035851831010206e-07,
+      "loss": 0.3782,
+      "step": 17824
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 17.075831909355863,
+      "learning_rate": 2.699691723650577e-07,
+      "loss": 0.4239,
+      "step": 17825
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 22.544019319984162,
+      "learning_rate": 2.6958010313639873e-07,
+      "loss": 0.4414,
+      "step": 17826
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.612221742317452,
+      "learning_rate": 2.6919131063518956e-07,
+      "loss": 0.4288,
+      "step": 17827
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 31.06784850113173,
+      "learning_rate": 2.6880279487248697e-07,
+      "loss": 0.4189,
+      "step": 17828
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.37054766468191,
+      "learning_rate": 2.684145558593398e-07,
+      "loss": 0.4404,
+      "step": 17829
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 21.728114558601035,
+      "learning_rate": 2.680265936067905e-07,
+      "loss": 0.42,
+      "step": 17830
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.826270009886528,
+      "learning_rate": 2.6763890812587126e-07,
+      "loss": 0.4683,
+      "step": 17831
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.725095256460804,
+      "learning_rate": 2.6725149942760875e-07,
+      "loss": 0.4641,
+      "step": 17832
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.005758651779562,
+      "learning_rate": 2.668643675230198e-07,
+      "loss": 0.4046,
+      "step": 17833
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.67878561715757,
+      "learning_rate": 2.664775124231156e-07,
+      "loss": 0.4316,
+      "step": 17834
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.55525916574125,
+      "learning_rate": 2.660909341388951e-07,
+      "loss": 0.3749,
+      "step": 17835
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 27.402857852172595,
+      "learning_rate": 2.65704632681355e-07,
+      "loss": 0.4265,
+      "step": 17836
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.315326874458894,
+      "learning_rate": 2.653186080614811e-07,
+      "loss": 0.3921,
+      "step": 17837
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.873260227104232,
+      "learning_rate": 2.649328602902501e-07,
+      "loss": 0.4263,
+      "step": 17838
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 21.511431394077448,
+      "learning_rate": 2.6454738937863545e-07,
+      "loss": 0.3848,
+      "step": 17839
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.758257753225937,
+      "learning_rate": 2.6416219533759724e-07,
+      "loss": 0.3854,
+      "step": 17840
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.036648640242426,
+      "learning_rate": 2.637772781780901e-07,
+      "loss": 0.4291,
+      "step": 17841
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.628179702317524,
+      "learning_rate": 2.6339263791106294e-07,
+      "loss": 0.4086,
+      "step": 17842
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 22.163010623028203,
+      "learning_rate": 2.630082745474538e-07,
+      "loss": 0.4376,
+      "step": 17843
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.723038656749363,
+      "learning_rate": 2.6262418809819166e-07,
+      "loss": 0.3908,
+      "step": 17844
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.039714017393614,
+      "learning_rate": 2.6224037857420225e-07,
+      "loss": 0.4845,
+      "step": 17845
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 24.176386528678915,
+      "learning_rate": 2.6185684598640013e-07,
+      "loss": 0.3959,
+      "step": 17846
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.52618035637394,
+      "learning_rate": 2.614735903456933e-07,
+      "loss": 0.4044,
+      "step": 17847
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 21.43696592738896,
+      "learning_rate": 2.610906116629797e-07,
+      "loss": 0.4376,
+      "step": 17848
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 22.564452098105797,
+      "learning_rate": 2.607079099491516e-07,
+      "loss": 0.4248,
+      "step": 17849
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.37316254670309,
+      "learning_rate": 2.6032548521509384e-07,
+      "loss": 0.4104,
+      "step": 17850
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.671465737161004,
+      "learning_rate": 2.59943337471682e-07,
+      "loss": 0.3953,
+      "step": 17851
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 27.93382537280599,
+      "learning_rate": 2.595614667297841e-07,
+      "loss": 0.4228,
+      "step": 17852
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 27.247811925166463,
+      "learning_rate": 2.591798730002593e-07,
+      "loss": 0.4455,
+      "step": 17853
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.55483761390252,
+      "learning_rate": 2.587985562939599e-07,
+      "loss": 0.4557,
+      "step": 17854
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 33.14607875360679,
+      "learning_rate": 2.5841751662173287e-07,
+      "loss": 0.5067,
+      "step": 17855
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.296160658338522,
+      "learning_rate": 2.5803675399441173e-07,
+      "loss": 0.5111,
+      "step": 17856
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 19.43695436414049,
+      "learning_rate": 2.576562684228279e-07,
+      "loss": 0.406,
+      "step": 17857
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 27.328284299679527,
+      "learning_rate": 2.5727605991779925e-07,
+      "loss": 0.4766,
+      "step": 17858
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 27.19147341620862,
+      "learning_rate": 2.5689612849014166e-07,
+      "loss": 0.4341,
+      "step": 17859
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.58358957947211,
+      "learning_rate": 2.565164741506576e-07,
+      "loss": 0.4133,
+      "step": 17860
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 17.142896297440632,
+      "learning_rate": 2.56137096910144e-07,
+      "loss": 0.4456,
+      "step": 17861
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.02469362052001,
+      "learning_rate": 2.557579967793933e-07,
+      "loss": 0.467,
+      "step": 17862
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 22.910104164526125,
+      "learning_rate": 2.5537917376918466e-07,
+      "loss": 0.4748,
+      "step": 17863
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 30.37279936389306,
+      "learning_rate": 2.5500062789029165e-07,
+      "loss": 0.3871,
+      "step": 17864
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.313061759308482,
+      "learning_rate": 2.546223591534802e-07,
+      "loss": 0.3915,
+      "step": 17865
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 19.73817612321626,
+      "learning_rate": 2.542443675695072e-07,
+      "loss": 0.4208,
+      "step": 17866
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.748565011865182,
+      "learning_rate": 2.538666531491241e-07,
+      "loss": 0.4057,
+      "step": 17867
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 17.54794355741235,
+      "learning_rate": 2.5348921590307216e-07,
+      "loss": 0.4045,
+      "step": 17868
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 37.00632794176606,
+      "learning_rate": 2.5311205584208523e-07,
+      "loss": 0.4751,
+      "step": 17869
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 37.13500278433084,
+      "learning_rate": 2.52735172976889e-07,
+      "loss": 0.4981,
+      "step": 17870
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.010762255813322,
+      "learning_rate": 2.5235856731820276e-07,
+      "loss": 0.4664,
+      "step": 17871
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.37970622446912,
+      "learning_rate": 2.519822388767379e-07,
+      "loss": 0.4873,
+      "step": 17872
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 48.70518842419169,
+      "learning_rate": 2.5160618766319477e-07,
+      "loss": 0.4545,
+      "step": 17873
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 32.41233336258351,
+      "learning_rate": 2.512304136882682e-07,
+      "loss": 0.4451,
+      "step": 17874
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 19.278552234495788,
+      "learning_rate": 2.508549169626462e-07,
+      "loss": 0.4012,
+      "step": 17875
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 17.404091322349313,
+      "learning_rate": 2.504796974970081e-07,
+      "loss": 0.3592,
+      "step": 17876
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.55736350355255,
+      "learning_rate": 2.5010475530202414e-07,
+      "loss": 0.3977,
+      "step": 17877
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.432892991775566,
+      "learning_rate": 2.4973009038835593e-07,
+      "loss": 0.3709,
+      "step": 17878
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.600486098199184,
+      "learning_rate": 2.4935570276666157e-07,
+      "loss": 0.4114,
+      "step": 17879
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 43.655283837355256,
+      "learning_rate": 2.4898159244758693e-07,
+      "loss": 0.4455,
+      "step": 17880
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 31.267924611682588,
+      "learning_rate": 2.486077594417724e-07,
+      "loss": 0.3964,
+      "step": 17881
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.25477956399121,
+      "learning_rate": 2.482342037598473e-07,
+      "loss": 0.5159,
+      "step": 17882
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 21.82578694458567,
+      "learning_rate": 2.4786092541243645e-07,
+      "loss": 0.4631,
+      "step": 17883
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 34.87979369989602,
+      "learning_rate": 2.474879244101569e-07,
+      "loss": 0.4869,
+      "step": 17884
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 17.024314767607805,
+      "learning_rate": 2.471152007636157e-07,
+      "loss": 0.4116,
+      "step": 17885
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 17.135663809825058,
+      "learning_rate": 2.4674275448341333e-07,
+      "loss": 0.4106,
+      "step": 17886
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 17.26663923434738,
+      "learning_rate": 2.463705855801413e-07,
+      "loss": 0.3772,
+      "step": 17887
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 29.899714734935777,
+      "learning_rate": 2.4599869406438327e-07,
+      "loss": 0.4664,
+      "step": 17888
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.476963347470463,
+      "learning_rate": 2.456270799467175e-07,
+      "loss": 0.4235,
+      "step": 17889
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.803210952907094,
+      "learning_rate": 2.452557432377101e-07,
+      "loss": 0.4169,
+      "step": 17890
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.70008974908491,
+      "learning_rate": 2.4488468394792463e-07,
+      "loss": 0.4275,
+      "step": 17891
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 27.93936782231887,
+      "learning_rate": 2.4451390208791173e-07,
+      "loss": 0.4253,
+      "step": 17892
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 22.847078125282508,
+      "learning_rate": 2.4414339766821614e-07,
+      "loss": 0.4504,
+      "step": 17893
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 28.535500378005608,
+      "learning_rate": 2.437731706993751e-07,
+      "loss": 0.4524,
+      "step": 17894
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.802235770219895,
+      "learning_rate": 2.434032211919179e-07,
+      "loss": 0.4161,
+      "step": 17895
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 23.109124842586482,
+      "learning_rate": 2.4303354915636735e-07,
+      "loss": 0.3741,
+      "step": 17896
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.01378010572684,
+      "learning_rate": 2.426641546032338e-07,
+      "loss": 0.4588,
+      "step": 17897
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.192167988932617,
+      "learning_rate": 2.422950375430233e-07,
+      "loss": 0.3965,
+      "step": 17898
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 35.99056494918064,
+      "learning_rate": 2.419261979862342e-07,
+      "loss": 0.4083,
+      "step": 17899
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.44663932228759,
+      "learning_rate": 2.415576359433558e-07,
+      "loss": 0.3789,
+      "step": 17900
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 21.106621868369555,
+      "learning_rate": 2.4118935142486975e-07,
+      "loss": 0.4439,
+      "step": 17901
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.056615825907002,
+      "learning_rate": 2.408213444412499e-07,
+      "loss": 0.4741,
+      "step": 17902
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.947655725701033,
+      "learning_rate": 2.404536150029624e-07,
+      "loss": 0.4008,
+      "step": 17903
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.189831583301878,
+      "learning_rate": 2.4008616312046315e-07,
+      "loss": 0.4449,
+      "step": 17904
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 18.155705264877064,
+      "learning_rate": 2.397189888042062e-07,
+      "loss": 0.4005,
+      "step": 17905
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.218961599102407,
+      "learning_rate": 2.3935209206463195e-07,
+      "loss": 0.4187,
+      "step": 17906
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.703775366458647,
+      "learning_rate": 2.389854729121721e-07,
+      "loss": 0.4173,
+      "step": 17907
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.050152176026867,
+      "learning_rate": 2.3861913135725613e-07,
+      "loss": 0.378,
+      "step": 17908
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.44977727219458,
+      "learning_rate": 2.382530674103023e-07,
+      "loss": 0.4788,
+      "step": 17909
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 28.367481589992543,
+      "learning_rate": 2.3788728108172121e-07,
+      "loss": 0.4312,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 20.97728748311856,
+      "learning_rate": 2.3752177238191455e-07,
+      "loss": 0.4681,
+      "step": 17911
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 21.009056153800827,
+      "learning_rate": 2.371565413212773e-07,
+      "loss": 0.435,
+      "step": 17912
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.329011798063423,
+      "learning_rate": 2.3679158791019673e-07,
+      "loss": 0.3975,
+      "step": 17913
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 26.850000225587483,
+      "learning_rate": 2.3642691215905344e-07,
+      "loss": 0.4575,
+      "step": 17914
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 25.12512308536575,
+      "learning_rate": 2.360625140782169e-07,
+      "loss": 0.4266,
+      "step": 17915
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.387586907688945,
+      "learning_rate": 2.3569839367805103e-07,
+      "loss": 0.4544,
+      "step": 17916
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 31.57983804281556,
+      "learning_rate": 2.3533455096890978e-07,
+      "loss": 0.4846,
+      "step": 17917
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.851156143276796,
+      "learning_rate": 2.3497098596114265e-07,
+      "loss": 0.4673,
+      "step": 17918
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.525718790305167,
+      "learning_rate": 2.3460769866508693e-07,
+      "loss": 0.4218,
+      "step": 17919
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 18.95308114193766,
+      "learning_rate": 2.3424468909107656e-07,
+      "loss": 0.4226,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.58590657467277,
+      "learning_rate": 2.338819572494355e-07,
+      "loss": 0.4016,
+      "step": 17921
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.077446403818065,
+      "learning_rate": 2.3351950315047777e-07,
+      "loss": 0.4855,
+      "step": 17922
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 34.780151764564344,
+      "learning_rate": 2.3315732680451175e-07,
+      "loss": 0.4779,
+      "step": 17923
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 23.17808452343965,
+      "learning_rate": 2.3279542822183698e-07,
+      "loss": 0.4154,
+      "step": 17924
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.08709453612449,
+      "learning_rate": 2.3243380741274745e-07,
+      "loss": 0.4062,
+      "step": 17925
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.56946141756476,
+      "learning_rate": 2.3207246438752719e-07,
+      "loss": 0.3592,
+      "step": 17926
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 30.321893269192223,
+      "learning_rate": 2.3171139915645235e-07,
+      "loss": 0.4537,
+      "step": 17927
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 21.366456171662513,
+      "learning_rate": 2.3135061172979034e-07,
+      "loss": 0.468,
+      "step": 17928
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 23.441128012555836,
+      "learning_rate": 2.3099010211780183e-07,
+      "loss": 0.4628,
+      "step": 17929
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 26.74679484617763,
+      "learning_rate": 2.306298703307408e-07,
+      "loss": 0.4472,
+      "step": 17930
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.431240642995363,
+      "learning_rate": 2.3026991637885242e-07,
+      "loss": 0.445,
+      "step": 17931
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.38865373158325,
+      "learning_rate": 2.2991024027237075e-07,
+      "loss": 0.4536,
+      "step": 17932
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.054190229574676,
+      "learning_rate": 2.2955084202152644e-07,
+      "loss": 0.3965,
+      "step": 17933
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.367665986616785,
+      "learning_rate": 2.2919172163654134e-07,
+      "loss": 0.405,
+      "step": 17934
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.21667680189889,
+      "learning_rate": 2.2883287912762842e-07,
+      "loss": 0.3957,
+      "step": 17935
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.55229995509514,
+      "learning_rate": 2.2847431450499169e-07,
+      "loss": 0.4233,
+      "step": 17936
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 31.149448112394793,
+      "learning_rate": 2.281160277788297e-07,
+      "loss": 0.4639,
+      "step": 17937
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.998319068664873,
+      "learning_rate": 2.277580189593298e-07,
+      "loss": 0.4088,
+      "step": 17938
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.619317761211487,
+      "learning_rate": 2.2740028805667725e-07,
+      "loss": 0.4614,
+      "step": 17939
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.025920572975707,
+      "learning_rate": 2.2704283508104386e-07,
+      "loss": 0.4051,
+      "step": 17940
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 26.720652305075667,
+      "learning_rate": 2.2668566004259486e-07,
+      "loss": 0.4812,
+      "step": 17941
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.52308261524274,
+      "learning_rate": 2.2632876295148765e-07,
+      "loss": 0.392,
+      "step": 17942
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.196821214205656,
+      "learning_rate": 2.2597214381787301e-07,
+      "loss": 0.4463,
+      "step": 17943
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.770853842847917,
+      "learning_rate": 2.2561580265189397e-07,
+      "loss": 0.4212,
+      "step": 17944
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 20.035732236463883,
+      "learning_rate": 2.2525973946368352e-07,
+      "loss": 0.4327,
+      "step": 17945
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.25822230189805,
+      "learning_rate": 2.2490395426336687e-07,
+      "loss": 0.4215,
+      "step": 17946
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.332611294960905,
+      "learning_rate": 2.2454844706106482e-07,
+      "loss": 0.4411,
+      "step": 17947
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 25.437310616739577,
+      "learning_rate": 2.2419321786688485e-07,
+      "loss": 0.4239,
+      "step": 17948
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.242596139103917,
+      "learning_rate": 2.2383826669093222e-07,
+      "loss": 0.4182,
+      "step": 17949
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 31.37208079629567,
+      "learning_rate": 2.2348359354330106e-07,
+      "loss": 0.4515,
+      "step": 17950
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.475944026074135,
+      "learning_rate": 2.2312919843407555e-07,
+      "loss": 0.4151,
+      "step": 17951
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 23.45641486178327,
+      "learning_rate": 2.227750813733376e-07,
+      "loss": 0.4315,
+      "step": 17952
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 18.957619987855367,
+      "learning_rate": 2.2242124237115693e-07,
+      "loss": 0.4343,
+      "step": 17953
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.50887749060113,
+      "learning_rate": 2.220676814375955e-07,
+      "loss": 0.4303,
+      "step": 17954
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 18.942143100024747,
+      "learning_rate": 2.217143985827097e-07,
+      "loss": 0.4136,
+      "step": 17955
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 33.555725354274415,
+      "learning_rate": 2.2136139381654708e-07,
+      "loss": 0.4696,
+      "step": 17956
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 20.660537463771654,
+      "learning_rate": 2.2100866714914514e-07,
+      "loss": 0.3986,
+      "step": 17957
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.221334921705278,
+      "learning_rate": 2.2065621859053587e-07,
+      "loss": 0.3785,
+      "step": 17958
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.54857315566461,
+      "learning_rate": 2.2030404815074346e-07,
+      "loss": 0.4516,
+      "step": 17959
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.719385980183056,
+      "learning_rate": 2.1995215583978435e-07,
+      "loss": 0.4367,
+      "step": 17960
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.032189383986218,
+      "learning_rate": 2.1960054166766386e-07,
+      "loss": 0.4147,
+      "step": 17961
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.170908254158707,
+      "learning_rate": 2.1924920564438178e-07,
+      "loss": 0.3889,
+      "step": 17962
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.646906441494428,
+      "learning_rate": 2.1889814777993013e-07,
+      "loss": 0.394,
+      "step": 17963
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 20.925506812705247,
+      "learning_rate": 2.1854736808429533e-07,
+      "loss": 0.4245,
+      "step": 17964
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.66466775180062,
+      "learning_rate": 2.1819686656745053e-07,
+      "loss": 0.3924,
+      "step": 17965
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.6367657252809,
+      "learning_rate": 2.1784664323936445e-07,
+      "loss": 0.4043,
+      "step": 17966
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.429948439189545,
+      "learning_rate": 2.1749669810999686e-07,
+      "loss": 0.4044,
+      "step": 17967
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 34.47296548072508,
+      "learning_rate": 2.1714703118930203e-07,
+      "loss": 0.4403,
+      "step": 17968
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.426145397725204,
+      "learning_rate": 2.16797642487222e-07,
+      "loss": 0.4375,
+      "step": 17969
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.97156176744616,
+      "learning_rate": 2.1644853201369444e-07,
+      "loss": 0.4265,
+      "step": 17970
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 20.83134378700535,
+      "learning_rate": 2.1609969977864687e-07,
+      "loss": 0.4478,
+      "step": 17971
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.90028893533822,
+      "learning_rate": 2.157511457920014e-07,
+      "loss": 0.3927,
+      "step": 17972
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 15.405559098728263,
+      "learning_rate": 2.1540287006366788e-07,
+      "loss": 0.3829,
+      "step": 17973
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.163114747719167,
+      "learning_rate": 2.1505487260355396e-07,
+      "loss": 0.346,
+      "step": 17974
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.13482991280381,
+      "learning_rate": 2.147071534215561e-07,
+      "loss": 0.4346,
+      "step": 17975
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 20.43033444490758,
+      "learning_rate": 2.1435971252756093e-07,
+      "loss": 0.4628,
+      "step": 17976
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.91660346603199,
+      "learning_rate": 2.140125499314527e-07,
+      "loss": 0.4146,
+      "step": 17977
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.263103447713288,
+      "learning_rate": 2.1366566564310244e-07,
+      "loss": 0.4877,
+      "step": 17978
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 22.674337070683304,
+      "learning_rate": 2.1331905967237553e-07,
+      "loss": 0.4124,
+      "step": 17979
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.106870838529687,
+      "learning_rate": 2.1297273202912972e-07,
+      "loss": 0.4446,
+      "step": 17980
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 31.21133850374774,
+      "learning_rate": 2.1262668272321486e-07,
+      "loss": 0.4912,
+      "step": 17981
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 22.936691970261247,
+      "learning_rate": 2.1228091176447086e-07,
+      "loss": 0.4554,
+      "step": 17982
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 29.217553275381192,
+      "learning_rate": 2.1193541916273097e-07,
+      "loss": 0.474,
+      "step": 17983
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.885164960164687,
+      "learning_rate": 2.11590204927824e-07,
+      "loss": 0.4252,
+      "step": 17984
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.9163463589218,
+      "learning_rate": 2.1124526906956545e-07,
+      "loss": 0.4332,
+      "step": 17985
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.514917725054374,
+      "learning_rate": 2.109006115977641e-07,
+      "loss": 0.4102,
+      "step": 17986
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.945164940878776,
+      "learning_rate": 2.1055623252222211e-07,
+      "loss": 0.4305,
+      "step": 17987
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.402236483872583,
+      "learning_rate": 2.1021213185273504e-07,
+      "loss": 0.3995,
+      "step": 17988
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.70476495208655,
+      "learning_rate": 2.0986830959908943e-07,
+      "loss": 0.4086,
+      "step": 17989
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.44522549748909,
+      "learning_rate": 2.0952476577106085e-07,
+      "loss": 0.3842,
+      "step": 17990
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.175598975506993,
+      "learning_rate": 2.0918150037842032e-07,
+      "loss": 0.3839,
+      "step": 17991
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.379939230450287,
+      "learning_rate": 2.088385134309312e-07,
+      "loss": 0.4379,
+      "step": 17992
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.57784531265054,
+      "learning_rate": 2.0849580493834675e-07,
+      "loss": 0.4463,
+      "step": 17993
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 26.167006690036793,
+      "learning_rate": 2.0815337491041476e-07,
+      "loss": 0.4224,
+      "step": 17994
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.843951997334646,
+      "learning_rate": 2.0781122335687075e-07,
+      "loss": 0.4467,
+      "step": 17995
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.981891019743163,
+      "learning_rate": 2.0746935028744808e-07,
+      "loss": 0.3986,
+      "step": 17996
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 26.026015872279658,
+      "learning_rate": 2.0712775571186895e-07,
+      "loss": 0.4376,
+      "step": 17997
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.838699151246583,
+      "learning_rate": 2.067864396398478e-07,
+      "loss": 0.3984,
+      "step": 17998
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.99544139480847,
+      "learning_rate": 2.0644540208109133e-07,
+      "loss": 0.4548,
+      "step": 17999
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.062223881244527,
+      "learning_rate": 2.0610464304529843e-07,
+      "loss": 0.4533,
+      "step": 18000
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.443422817345073,
+      "learning_rate": 2.0576416254216024e-07,
+      "loss": 0.4411,
+      "step": 18001
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 24.47054582219941,
+      "learning_rate": 2.0542396058136015e-07,
+      "loss": 0.462,
+      "step": 18002
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 23.04319876947105,
+      "learning_rate": 2.0508403717257263e-07,
+      "loss": 0.4435,
+      "step": 18003
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.397809130271202,
+      "learning_rate": 2.047443923254655e-07,
+      "loss": 0.3962,
+      "step": 18004
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.505013511246501,
+      "learning_rate": 2.0440502604969659e-07,
+      "loss": 0.3772,
+      "step": 18005
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.704697973212328,
+      "learning_rate": 2.040659383549204e-07,
+      "loss": 0.4494,
+      "step": 18006
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.39206411195344,
+      "learning_rate": 2.0372712925077697e-07,
+      "loss": 0.4095,
+      "step": 18007
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 26.479218452469503,
+      "learning_rate": 2.0338859874690308e-07,
+      "loss": 0.4731,
+      "step": 18008
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 36.94376860534872,
+      "learning_rate": 2.030503468529288e-07,
+      "loss": 0.4564,
+      "step": 18009
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.20613844833497,
+      "learning_rate": 2.0271237357847085e-07,
+      "loss": 0.3896,
+      "step": 18010
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 25.56121906787919,
+      "learning_rate": 2.0237467893314044e-07,
+      "loss": 0.423,
+      "step": 18011
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.851176906520124,
+      "learning_rate": 2.0203726292654324e-07,
+      "loss": 0.4429,
+      "step": 18012
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.81612367579466,
+      "learning_rate": 2.0170012556827379e-07,
+      "loss": 0.4194,
+      "step": 18013
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.072584658421118,
+      "learning_rate": 2.0136326686792217e-07,
+      "loss": 0.4459,
+      "step": 18014
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.352500836323383,
+      "learning_rate": 2.0102668683506743e-07,
+      "loss": 0.4194,
+      "step": 18015
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 29.785433168218354,
+      "learning_rate": 2.0069038547928078e-07,
+      "loss": 0.4289,
+      "step": 18016
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.201255488696873,
+      "learning_rate": 2.0035436281012678e-07,
+      "loss": 0.4176,
+      "step": 18017
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.335751808431972,
+      "learning_rate": 2.0001861883716335e-07,
+      "loss": 0.512,
+      "step": 18018
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.189362042739145,
+      "learning_rate": 1.996831535699373e-07,
+      "loss": 0.4082,
+      "step": 18019
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 24.687079651791045,
+      "learning_rate": 1.9934796701798876e-07,
+      "loss": 0.4497,
+      "step": 18020
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 28.828919896484148,
+      "learning_rate": 1.9901305919085123e-07,
+      "loss": 0.4248,
+      "step": 18021
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 27.216542038342144,
+      "learning_rate": 1.9867843009804822e-07,
+      "loss": 0.4025,
+      "step": 18022
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 26.15781855807853,
+      "learning_rate": 1.9834407974909764e-07,
+      "loss": 0.3836,
+      "step": 18023
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.04948801610312,
+      "learning_rate": 1.9801000815350745e-07,
+      "loss": 0.4651,
+      "step": 18024
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.298808560093953,
+      "learning_rate": 1.9767621532077896e-07,
+      "loss": 0.4397,
+      "step": 18025
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.37913525058086,
+      "learning_rate": 1.9734270126040344e-07,
+      "loss": 0.432,
+      "step": 18026
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.940686963333558,
+      "learning_rate": 1.9700946598186887e-07,
+      "loss": 0.4241,
+      "step": 18027
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.604745911987667,
+      "learning_rate": 1.966765094946499e-07,
+      "loss": 0.4077,
+      "step": 18028
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.332987679916233,
+      "learning_rate": 1.963438318082156e-07,
+      "loss": 0.4453,
+      "step": 18029
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.15408249229552,
+      "learning_rate": 1.960114329320284e-07,
+      "loss": 0.4226,
+      "step": 18030
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.414135281127027,
+      "learning_rate": 1.9567931287554075e-07,
+      "loss": 0.444,
+      "step": 18031
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.12404062488127,
+      "learning_rate": 1.9534747164819733e-07,
+      "loss": 0.4192,
+      "step": 18032
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 26.36948730159487,
+      "learning_rate": 1.9501590925943726e-07,
+      "loss": 0.4181,
+      "step": 18033
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 29.864448556483914,
+      "learning_rate": 1.9468462571868852e-07,
+      "loss": 0.4646,
+      "step": 18034
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.638843784390666,
+      "learning_rate": 1.9435362103537357e-07,
+      "loss": 0.3883,
+      "step": 18035
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.180132010990018,
+      "learning_rate": 1.9402289521890495e-07,
+      "loss": 0.4655,
+      "step": 18036
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.873556200103817,
+      "learning_rate": 1.936924482786884e-07,
+      "loss": 0.4052,
+      "step": 18037
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.944717047681387,
+      "learning_rate": 1.9336228022412306e-07,
+      "loss": 0.4075,
+      "step": 18038
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 15.459358404554317,
+      "learning_rate": 1.9303239106459703e-07,
+      "loss": 0.4124,
+      "step": 18039
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.725490016865255,
+      "learning_rate": 1.927027808094939e-07,
+      "loss": 0.402,
+      "step": 18040
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 17.46549335330472,
+      "learning_rate": 1.9237344946818614e-07,
+      "loss": 0.3899,
+      "step": 18041
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.593249390398487,
+      "learning_rate": 1.9204439705003964e-07,
+      "loss": 0.4879,
+      "step": 18042
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.53584317932088,
+      "learning_rate": 1.9171562356441353e-07,
+      "loss": 0.4065,
+      "step": 18043
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.074040117900097,
+      "learning_rate": 1.9138712902065703e-07,
+      "loss": 0.4057,
+      "step": 18044
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 23.156110949105237,
+      "learning_rate": 1.9105891342811268e-07,
+      "loss": 0.4093,
+      "step": 18045
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 17.38387574962378,
+      "learning_rate": 1.9073097679611518e-07,
+      "loss": 0.4512,
+      "step": 18046
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 23.113693351611204,
+      "learning_rate": 1.9040331913398934e-07,
+      "loss": 0.4384,
+      "step": 18047
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.82495614169634,
+      "learning_rate": 1.900759404510577e-07,
+      "loss": 0.4256,
+      "step": 18048
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.005652331541793,
+      "learning_rate": 1.8974884075662503e-07,
+      "loss": 0.4273,
+      "step": 18049
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 13.767564499450945,
+      "learning_rate": 1.894220200599961e-07,
+      "loss": 0.4957,
+      "step": 18050
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.785842358146194,
+      "learning_rate": 1.8909547837046683e-07,
+      "loss": 0.3231,
+      "step": 18051
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.597745308570083,
+      "learning_rate": 1.887692156973231e-07,
+      "loss": 0.457,
+      "step": 18052
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 32.00777855411114,
+      "learning_rate": 1.8844323204984417e-07,
+      "loss": 0.4225,
+      "step": 18053
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 26.357248174116904,
+      "learning_rate": 1.8811752743729817e-07,
+      "loss": 0.4255,
+      "step": 18054
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.549965623221635,
+      "learning_rate": 1.8779210186895102e-07,
+      "loss": 0.4693,
+      "step": 18055
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 22.033894527946728,
+      "learning_rate": 1.8746695535405645e-07,
+      "loss": 0.4267,
+      "step": 18056
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 26.834830938454907,
+      "learning_rate": 1.871420879018615e-07,
+      "loss": 0.4254,
+      "step": 18057
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.04780078453937,
+      "learning_rate": 1.8681749952160544e-07,
+      "loss": 0.4107,
+      "step": 18058
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.944762309132198,
+      "learning_rate": 1.8649319022251865e-07,
+      "loss": 0.4351,
+      "step": 18059
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 27.39598569404578,
+      "learning_rate": 1.8616916001382602e-07,
+      "loss": 0.4731,
+      "step": 18060
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.46669486757505,
+      "learning_rate": 1.8584540890474012e-07,
+      "loss": 0.329,
+      "step": 18061
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.708401359306325,
+      "learning_rate": 1.855219369044703e-07,
+      "loss": 0.4504,
+      "step": 18062
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.80060555401762,
+      "learning_rate": 1.8519874402221472e-07,
+      "loss": 0.4368,
+      "step": 18063
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.55700747070408,
+      "learning_rate": 1.8487583026716604e-07,
+      "loss": 0.4529,
+      "step": 18064
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.17969897434173,
+      "learning_rate": 1.8455319564850694e-07,
+      "loss": 0.3771,
+      "step": 18065
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 15.298909719272224,
+      "learning_rate": 1.8423084017541227e-07,
+      "loss": 0.4362,
+      "step": 18066
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.20105157117016,
+      "learning_rate": 1.8390876385705136e-07,
+      "loss": 0.4054,
+      "step": 18067
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.69374549427703,
+      "learning_rate": 1.8358696670258357e-07,
+      "loss": 0.4537,
+      "step": 18068
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.454268990957402,
+      "learning_rate": 1.8326544872115936e-07,
+      "loss": 0.4051,
+      "step": 18069
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.01332041744621,
+      "learning_rate": 1.829442099219225e-07,
+      "loss": 0.466,
+      "step": 18070
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.520798793850297,
+      "learning_rate": 1.826232503140102e-07,
+      "loss": 0.4633,
+      "step": 18071
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.453714394736885,
+      "learning_rate": 1.823025699065495e-07,
+      "loss": 0.4518,
+      "step": 18072
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 22.71209743885157,
+      "learning_rate": 1.8198216870866093e-07,
+      "loss": 0.4082,
+      "step": 18073
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.074791542603535,
+      "learning_rate": 1.8166204672945496e-07,
+      "loss": 0.5501,
+      "step": 18074
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.001064209887243,
+      "learning_rate": 1.813422039780377e-07,
+      "loss": 0.3711,
+      "step": 18075
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 22.77712935005126,
+      "learning_rate": 1.8102264046350404e-07,
+      "loss": 0.4231,
+      "step": 18076
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.90221294152296,
+      "learning_rate": 1.8070335619494227e-07,
+      "loss": 0.4831,
+      "step": 18077
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.483302189049436,
+      "learning_rate": 1.8038435118143405e-07,
+      "loss": 0.4072,
+      "step": 18078
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.329837099582967,
+      "learning_rate": 1.8006562543204986e-07,
+      "loss": 0.4118,
+      "step": 18079
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.900038995032247,
+      "learning_rate": 1.797471789558547e-07,
+      "loss": 0.4752,
+      "step": 18080
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.903649596175914,
+      "learning_rate": 1.7942901176190574e-07,
+      "loss": 0.4263,
+      "step": 18081
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 27.36413278391528,
+      "learning_rate": 1.7911112385925022e-07,
+      "loss": 0.4231,
+      "step": 18082
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 15.371269739348032,
+      "learning_rate": 1.7879351525692978e-07,
+      "loss": 0.4034,
+      "step": 18083
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 34.24500229798534,
+      "learning_rate": 1.7847618596397498e-07,
+      "loss": 0.4887,
+      "step": 18084
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.28627134059287,
+      "learning_rate": 1.7815913598941303e-07,
+      "loss": 0.4218,
+      "step": 18085
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.181616275644974,
+      "learning_rate": 1.7784236534226006e-07,
+      "loss": 0.4187,
+      "step": 18086
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.870334229454777,
+      "learning_rate": 1.7752587403152444e-07,
+      "loss": 0.3943,
+      "step": 18087
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.01649455262894,
+      "learning_rate": 1.7720966206620672e-07,
+      "loss": 0.41,
+      "step": 18088
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.797464520004223,
+      "learning_rate": 1.768937294552997e-07,
+      "loss": 0.4171,
+      "step": 18089
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.43060913708382,
+      "learning_rate": 1.7657807620778956e-07,
+      "loss": 0.4399,
+      "step": 18090
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.210114085611043,
+      "learning_rate": 1.762627023326513e-07,
+      "loss": 0.4687,
+      "step": 18091
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 28.032104896981732,
+      "learning_rate": 1.7594760783885557e-07,
+      "loss": 0.4251,
+      "step": 18092
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 28.886526744480392,
+      "learning_rate": 1.7563279273536294e-07,
+      "loss": 0.4783,
+      "step": 18093
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.713493722860782,
+      "learning_rate": 1.7531825703112736e-07,
+      "loss": 0.3652,
+      "step": 18094
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.35630160621016,
+      "learning_rate": 1.7500400073509172e-07,
+      "loss": 0.3439,
+      "step": 18095
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 46.848033351011836,
+      "learning_rate": 1.7469002385619548e-07,
+      "loss": 0.4755,
+      "step": 18096
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 27.098689946154227,
+      "learning_rate": 1.7437632640336822e-07,
+      "loss": 0.475,
+      "step": 18097
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.826411482122195,
+      "learning_rate": 1.7406290838552943e-07,
+      "loss": 0.4656,
+      "step": 18098
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.272662359847125,
+      "learning_rate": 1.737497698115942e-07,
+      "loss": 0.4917,
+      "step": 18099
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 23.69031228781328,
+      "learning_rate": 1.7343691069046654e-07,
+      "loss": 0.4148,
+      "step": 18100
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 22.14307886994061,
+      "learning_rate": 1.7312433103104486e-07,
+      "loss": 0.4642,
+      "step": 18101
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.091150067789329,
+      "learning_rate": 1.7281203084221876e-07,
+      "loss": 0.4693,
+      "step": 18102
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 24.46290276015051,
+      "learning_rate": 1.725000101328711e-07,
+      "loss": 0.3895,
+      "step": 18103
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.00665190591452,
+      "learning_rate": 1.7218826891187258e-07,
+      "loss": 0.3895,
+      "step": 18104
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.228924223964846,
+      "learning_rate": 1.7187680718809053e-07,
+      "loss": 0.4256,
+      "step": 18105
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 26.48815626534871,
+      "learning_rate": 1.7156562497038342e-07,
+      "loss": 0.4698,
+      "step": 18106
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.5080570814964,
+      "learning_rate": 1.7125472226760087e-07,
+      "loss": 0.3958,
+      "step": 18107
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.23897477723079,
+      "learning_rate": 1.7094409908858357e-07,
+      "loss": 0.4444,
+      "step": 18108
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.72896190740729,
+      "learning_rate": 1.7063375544216666e-07,
+      "loss": 0.4255,
+      "step": 18109
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 23.006760234765697,
+      "learning_rate": 1.703236913371753e-07,
+      "loss": 0.43,
+      "step": 18110
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.394987076413656,
+      "learning_rate": 1.7001390678242803e-07,
+      "loss": 0.4356,
+      "step": 18111
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.898448192981473,
+      "learning_rate": 1.6970440178673552e-07,
+      "loss": 0.4079,
+      "step": 18112
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.822246959766153,
+      "learning_rate": 1.6939517635889857e-07,
+      "loss": 0.3617,
+      "step": 18113
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.644290886983896,
+      "learning_rate": 1.6908623050771234e-07,
+      "loss": 0.4395,
+      "step": 18114
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.721337680432228,
+      "learning_rate": 1.6877756424196312e-07,
+      "loss": 0.4423,
+      "step": 18115
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 29.035468263154407,
+      "learning_rate": 1.6846917757042836e-07,
+      "loss": 0.4062,
+      "step": 18116
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.028969375069227,
+      "learning_rate": 1.6816107050187992e-07,
+      "loss": 0.4754,
+      "step": 18117
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 22.97966827731043,
+      "learning_rate": 1.6785324304507745e-07,
+      "loss": 0.4466,
+      "step": 18118
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.103186958032786,
+      "learning_rate": 1.675456952087795e-07,
+      "loss": 0.3774,
+      "step": 18119
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 28.04438139793631,
+      "learning_rate": 1.67238427001728e-07,
+      "loss": 0.4336,
+      "step": 18120
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.868874365569596,
+      "learning_rate": 1.669314384326648e-07,
+      "loss": 0.4362,
+      "step": 18121
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.185261840727215,
+      "learning_rate": 1.666247295103185e-07,
+      "loss": 0.4189,
+      "step": 18122
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 22.618388579897058,
+      "learning_rate": 1.6631830024341323e-07,
+      "loss": 0.4924,
+      "step": 18123
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 36.765768636932776,
+      "learning_rate": 1.66012150640662e-07,
+      "loss": 0.4616,
+      "step": 18124
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 14.82754277894135,
+      "learning_rate": 1.6570628071077343e-07,
+      "loss": 0.3961,
+      "step": 18125
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.195369655447088,
+      "learning_rate": 1.65400690462445e-07,
+      "loss": 0.4697,
+      "step": 18126
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.645068167612955,
+      "learning_rate": 1.6509537990436753e-07,
+      "loss": 0.4393,
+      "step": 18127
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.288813188068776,
+      "learning_rate": 1.6479034904522518e-07,
+      "loss": 0.4724,
+      "step": 18128
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.54076714896541,
+      "learning_rate": 1.6448559789369102e-07,
+      "loss": 0.4038,
+      "step": 18129
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.671157344290176,
+      "learning_rate": 1.6418112645843253e-07,
+      "loss": 0.3625,
+      "step": 18130
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.851430332127304,
+      "learning_rate": 1.638769347481095e-07,
+      "loss": 0.3647,
+      "step": 18131
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.464256768237657,
+      "learning_rate": 1.6357302277137388e-07,
+      "loss": 0.4647,
+      "step": 18132
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.426634853449738,
+      "learning_rate": 1.632693905368654e-07,
+      "loss": 0.4356,
+      "step": 18133
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 28.152442060722105,
+      "learning_rate": 1.6296603805322163e-07,
+      "loss": 0.4448,
+      "step": 18134
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.292625044420976,
+      "learning_rate": 1.62662965329069e-07,
+      "loss": 0.4173,
+      "step": 18135
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 16.925419858819957,
+      "learning_rate": 1.6236017237302838e-07,
+      "loss": 0.4472,
+      "step": 18136
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 16.481163865588723,
+      "learning_rate": 1.6205765919370843e-07,
+      "loss": 0.4739,
+      "step": 18137
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 16.99978433938535,
+      "learning_rate": 1.6175542579971336e-07,
+      "loss": 0.4241,
+      "step": 18138
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.73395309030394,
+      "learning_rate": 1.6145347219963857e-07,
+      "loss": 0.362,
+      "step": 18139
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 22.04112731607826,
+      "learning_rate": 1.6115179840207274e-07,
+      "loss": 0.4428,
+      "step": 18140
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.14290410309478,
+      "learning_rate": 1.608504044155934e-07,
+      "loss": 0.4003,
+      "step": 18141
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 16.596535828683223,
+      "learning_rate": 1.6054929024877375e-07,
+      "loss": 0.3913,
+      "step": 18142
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 27.040931804578047,
+      "learning_rate": 1.6024845591017468e-07,
+      "loss": 0.4133,
+      "step": 18143
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 29.86997905191291,
+      "learning_rate": 1.599479014083549e-07,
+      "loss": 0.3703,
+      "step": 18144
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 23.98163823595663,
+      "learning_rate": 1.5964762675185984e-07,
+      "loss": 0.4183,
+      "step": 18145
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 34.122474402284546,
+      "learning_rate": 1.5934763194923043e-07,
+      "loss": 0.4737,
+      "step": 18146
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 16.175138456593668,
+      "learning_rate": 1.5904791700899758e-07,
+      "loss": 0.4222,
+      "step": 18147
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.617639431288524,
+      "learning_rate": 1.587484819396856e-07,
+      "loss": 0.3616,
+      "step": 18148
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.52473055491459,
+      "learning_rate": 1.5844932674980885e-07,
+      "loss": 0.4336,
+      "step": 18149
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 30.533524061043977,
+      "learning_rate": 1.58150451447876e-07,
+      "loss": 0.3729,
+      "step": 18150
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 22.754961925460908,
+      "learning_rate": 1.5785185604238805e-07,
+      "loss": 0.4209,
+      "step": 18151
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.098454730627093,
+      "learning_rate": 1.5755354054183491e-07,
+      "loss": 0.3867,
+      "step": 18152
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 29.90687654450791,
+      "learning_rate": 1.57255504954702e-07,
+      "loss": 0.4125,
+      "step": 18153
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.95612312828468,
+      "learning_rate": 1.5695774928946473e-07,
+      "loss": 0.461,
+      "step": 18154
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 22.89656441354108,
+      "learning_rate": 1.5666027355458967e-07,
+      "loss": 0.417,
+      "step": 18155
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 25.8180123070717,
+      "learning_rate": 1.563630777585401e-07,
+      "loss": 0.3954,
+      "step": 18156
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 31.728573400762432,
+      "learning_rate": 1.5606616190976587e-07,
+      "loss": 0.3966,
+      "step": 18157
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 23.59168916790271,
+      "learning_rate": 1.5576952601671135e-07,
+      "loss": 0.4001,
+      "step": 18158
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 20.335210898568214,
+      "learning_rate": 1.5547317008781205e-07,
+      "loss": 0.4492,
+      "step": 18159
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 32.75339082643325,
+      "learning_rate": 1.5517709413149674e-07,
+      "loss": 0.4202,
+      "step": 18160
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 34.61698646202934,
+      "learning_rate": 1.5488129815618757e-07,
+      "loss": 0.4497,
+      "step": 18161
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 16.712190585082542,
+      "learning_rate": 1.5458578217029342e-07,
+      "loss": 0.4043,
+      "step": 18162
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 26.23191589203769,
+      "learning_rate": 1.5429054618222083e-07,
+      "loss": 0.4396,
+      "step": 18163
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 26.617654282955858,
+      "learning_rate": 1.5399559020036537e-07,
+      "loss": 0.4442,
+      "step": 18164
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 24.92596551157176,
+      "learning_rate": 1.5370091423311584e-07,
+      "loss": 0.4267,
+      "step": 18165
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 20.889899187026224,
+      "learning_rate": 1.5340651828885223e-07,
+      "loss": 0.4176,
+      "step": 18166
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.74141980706441,
+      "learning_rate": 1.531124023759467e-07,
+      "loss": 0.4263,
+      "step": 18167
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.603786962738102,
+      "learning_rate": 1.5281856650276482e-07,
+      "loss": 0.4044,
+      "step": 18168
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 26.053305121127813,
+      "learning_rate": 1.5252501067766324e-07,
+      "loss": 0.4292,
+      "step": 18169
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 13.951804333518846,
+      "learning_rate": 1.5223173490898856e-07,
+      "loss": 0.4112,
+      "step": 18170
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 35.875249590985746,
+      "learning_rate": 1.5193873920508306e-07,
+      "loss": 0.5211,
+      "step": 18171
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 15.136251961678564,
+      "learning_rate": 1.5164602357427781e-07,
+      "loss": 0.4034,
+      "step": 18172
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 19.237586526090553,
+      "learning_rate": 1.5135358802489952e-07,
+      "loss": 0.4031,
+      "step": 18173
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 20.087554054982746,
+      "learning_rate": 1.5106143256526374e-07,
+      "loss": 0.424,
+      "step": 18174
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 31.098079034513844,
+      "learning_rate": 1.5076955720367937e-07,
+      "loss": 0.4257,
+      "step": 18175
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 26.387130861713654,
+      "learning_rate": 1.5047796194844754e-07,
+      "loss": 0.4193,
+      "step": 18176
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 17.34765911209643,
+      "learning_rate": 1.5018664680785943e-07,
+      "loss": 0.4429,
+      "step": 18177
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 27.522948187908025,
+      "learning_rate": 1.4989561179020283e-07,
+      "loss": 0.5117,
+      "step": 18178
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 23.37303788000141,
+      "learning_rate": 1.4960485690375115e-07,
+      "loss": 0.4123,
+      "step": 18179
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 46.58080469463978,
+      "learning_rate": 1.4931438215677663e-07,
+      "loss": 0.4591,
+      "step": 18180
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 24.04519566527334,
+      "learning_rate": 1.490241875575371e-07,
+      "loss": 0.398,
+      "step": 18181
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 20.34164301483965,
+      "learning_rate": 1.4873427311428822e-07,
+      "loss": 0.3969,
+      "step": 18182
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 15.675626770634809,
+      "learning_rate": 1.484446388352734e-07,
+      "loss": 0.4051,
+      "step": 18183
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 17.31448853000177,
+      "learning_rate": 1.481552847287293e-07,
+      "loss": 0.4227,
+      "step": 18184
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.04627522730495,
+      "learning_rate": 1.478662108028872e-07,
+      "loss": 0.4129,
+      "step": 18185
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 33.439381932838494,
+      "learning_rate": 1.4757741706596605e-07,
+      "loss": 0.4465,
+      "step": 18186
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 28.171230519307425,
+      "learning_rate": 1.472889035261793e-07,
+      "loss": 0.435,
+      "step": 18187
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.740528923769265,
+      "learning_rate": 1.4700067019173258e-07,
+      "loss": 0.3903,
+      "step": 18188
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 16.00891777270265,
+      "learning_rate": 1.4671271707082268e-07,
+      "loss": 0.452,
+      "step": 18189
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 43.32758584806517,
+      "learning_rate": 1.4642504417163973e-07,
+      "loss": 0.4603,
+      "step": 18190
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 25.222767669989338,
+      "learning_rate": 1.4613765150236382e-07,
+      "loss": 0.4583,
+      "step": 18191
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 44.02331776166877,
+      "learning_rate": 1.4585053907116953e-07,
+      "loss": 0.4935,
+      "step": 18192
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 22.404369818038525,
+      "learning_rate": 1.4556370688622036e-07,
+      "loss": 0.3643,
+      "step": 18193
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 34.20547161922587,
+      "learning_rate": 1.4527715495567529e-07,
+      "loss": 0.4229,
+      "step": 18194
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.17256618589119,
+      "learning_rate": 1.4499088328768341e-07,
+      "loss": 0.391,
+      "step": 18195
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 23.934183261958513,
+      "learning_rate": 1.4470489189038595e-07,
+      "loss": 0.5173,
+      "step": 18196
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.611196711468622,
+      "learning_rate": 1.444191807719153e-07,
+      "loss": 0.3712,
+      "step": 18197
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 19.25601794255633,
+      "learning_rate": 1.4413374994039718e-07,
+      "loss": 0.4396,
+      "step": 18198
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 25.81377415364099,
+      "learning_rate": 1.438485994039518e-07,
+      "loss": 0.4755,
+      "step": 18199
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.55632520525946,
+      "learning_rate": 1.4356372917068595e-07,
+      "loss": 0.4468,
+      "step": 18200
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 19.516011784785015,
+      "learning_rate": 1.4327913924870097e-07,
+      "loss": 0.4856,
+      "step": 18201
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 20.55362997955764,
+      "learning_rate": 1.4299482964609035e-07,
+      "loss": 0.4636,
+      "step": 18202
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 25.437518246337195,
+      "learning_rate": 1.427108003709421e-07,
+      "loss": 0.4259,
+      "step": 18203
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 28.093469669801973,
+      "learning_rate": 1.4242705143133195e-07,
+      "loss": 0.4142,
+      "step": 18204
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 30.588926237977518,
+      "learning_rate": 1.4214358283533124e-07,
+      "loss": 0.3983,
+      "step": 18205
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.730443331926935,
+      "learning_rate": 1.41860394590998e-07,
+      "loss": 0.4127,
+      "step": 18206
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 32.90276969082344,
+      "learning_rate": 1.4157748670639015e-07,
+      "loss": 0.4189,
+      "step": 18207
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 24.323852056015433,
+      "learning_rate": 1.412948591895502e-07,
+      "loss": 0.4323,
+      "step": 18208
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 22.991900443859063,
+      "learning_rate": 1.4101251204851841e-07,
+      "loss": 0.4128,
+      "step": 18209
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.864021626420843,
+      "learning_rate": 1.4073044529132273e-07,
+      "loss": 0.4514,
+      "step": 18210
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 24.75328245576418,
+      "learning_rate": 1.4044865892598681e-07,
+      "loss": 0.4128,
+      "step": 18211
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 27.72153820456351,
+      "learning_rate": 1.4016715296052198e-07,
+      "loss": 0.4743,
+      "step": 18212
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.796791619506735,
+      "learning_rate": 1.3988592740293629e-07,
+      "loss": 0.4505,
+      "step": 18213
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 22.097491542734378,
+      "learning_rate": 1.3960498226122554e-07,
+      "loss": 0.3998,
+      "step": 18214
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 31.23495737136862,
+      "learning_rate": 1.3932431754338227e-07,
+      "loss": 0.4552,
+      "step": 18215
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.836687720006733,
+      "learning_rate": 1.3904393325738785e-07,
+      "loss": 0.5238,
+      "step": 18216
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 28.404939685266307,
+      "learning_rate": 1.387638294112137e-07,
+      "loss": 0.4581,
+      "step": 18217
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 22.696002639152205,
+      "learning_rate": 1.3848400601282784e-07,
+      "loss": 0.396,
+      "step": 18218
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 15.952713981624163,
+      "learning_rate": 1.3820446307018843e-07,
+      "loss": 0.3807,
+      "step": 18219
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 27.57750819329659,
+      "learning_rate": 1.379252005912457e-07,
+      "loss": 0.4514,
+      "step": 18220
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.08277278108093,
+      "learning_rate": 1.3764621858394e-07,
+      "loss": 0.444,
+      "step": 18221
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 28.012148032345465,
+      "learning_rate": 1.373675170562072e-07,
+      "loss": 0.4614,
+      "step": 18222
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 29.257655526568993,
+      "learning_rate": 1.370890960159721e-07,
+      "loss": 0.405,
+      "step": 18223
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.971149642466557,
+      "learning_rate": 1.3681095547115497e-07,
+      "loss": 0.4054,
+      "step": 18224
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 19.95551134139376,
+      "learning_rate": 1.3653309542966286e-07,
+      "loss": 0.4338,
+      "step": 18225
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 22.394114875216385,
+      "learning_rate": 1.3625551589940056e-07,
+      "loss": 0.4221,
+      "step": 18226
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.77124156926482,
+      "learning_rate": 1.3597821688826063e-07,
+      "loss": 0.4406,
+      "step": 18227
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 26.27295173470462,
+      "learning_rate": 1.357011984041301e-07,
+      "loss": 0.3948,
+      "step": 18228
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 19.550298932935227,
+      "learning_rate": 1.3542446045488712e-07,
+      "loss": 0.4221,
+      "step": 18229
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.261717487237515,
+      "learning_rate": 1.3514800304840203e-07,
+      "loss": 0.4513,
+      "step": 18230
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.378416580288324,
+      "learning_rate": 1.3487182619253636e-07,
+      "loss": 0.4522,
+      "step": 18231
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.04620413916343,
+      "learning_rate": 1.3459592989514604e-07,
+      "loss": 0.3865,
+      "step": 18232
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.92382540828731,
+      "learning_rate": 1.3432031416407476e-07,
+      "loss": 0.4419,
+      "step": 18233
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 25.065576182623328,
+      "learning_rate": 1.3404497900716406e-07,
+      "loss": 0.4844,
+      "step": 18234
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 31.668818941299143,
+      "learning_rate": 1.3376992443224213e-07,
+      "loss": 0.4946,
+      "step": 18235
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 16.565309548950104,
+      "learning_rate": 1.3349515044713157e-07,
+      "loss": 0.4246,
+      "step": 18236
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 22.682159837032373,
+      "learning_rate": 1.3322065705964838e-07,
+      "loss": 0.4108,
+      "step": 18237
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.122144044725097,
+      "learning_rate": 1.3294644427759628e-07,
+      "loss": 0.4337,
+      "step": 18238
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 23.58910743561029,
+      "learning_rate": 1.3267251210877686e-07,
+      "loss": 0.4263,
+      "step": 18239
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 16.794766502225077,
+      "learning_rate": 1.3239886056097718e-07,
+      "loss": 0.3911,
+      "step": 18240
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 20.03918227444374,
+      "learning_rate": 1.3212548964198324e-07,
+      "loss": 0.3791,
+      "step": 18241
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 26.254945583740252,
+      "learning_rate": 1.3185239935956662e-07,
+      "loss": 0.4239,
+      "step": 18242
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.113176742126637,
+      "learning_rate": 1.3157958972149553e-07,
+      "loss": 0.3723,
+      "step": 18243
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 25.40501717418585,
+      "learning_rate": 1.3130706073552824e-07,
+      "loss": 0.4449,
+      "step": 18244
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 30.65019077825835,
+      "learning_rate": 1.310348124094152e-07,
+      "loss": 0.4439,
+      "step": 18245
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.28012598840872,
+      "learning_rate": 1.3076284475089906e-07,
+      "loss": 0.4163,
+      "step": 18246
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.461155361982932,
+      "learning_rate": 1.304911577677137e-07,
+      "loss": 0.4483,
+      "step": 18247
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.86781911680022,
+      "learning_rate": 1.3021975146758514e-07,
+      "loss": 0.4491,
+      "step": 18248
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.215532946643265,
+      "learning_rate": 1.2994862585823608e-07,
+      "loss": 0.4408,
+      "step": 18249
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 27.474143368909964,
+      "learning_rate": 1.296777809473726e-07,
+      "loss": 0.4995,
+      "step": 18250
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 16.86276445619728,
+      "learning_rate": 1.294072167426985e-07,
+      "loss": 0.4199,
+      "step": 18251
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 32.63909716063981,
+      "learning_rate": 1.291369332519099e-07,
+      "loss": 0.4275,
+      "step": 18252
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 32.4958815936282,
+      "learning_rate": 1.2886693048269284e-07,
+      "loss": 0.4332,
+      "step": 18253
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 29.899361495708675,
+      "learning_rate": 1.285972084427256e-07,
+      "loss": 0.4823,
+      "step": 18254
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 20.185397513270228,
+      "learning_rate": 1.2832776713967876e-07,
+      "loss": 0.3765,
+      "step": 18255
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 23.96969018911251,
+      "learning_rate": 1.2805860658121505e-07,
+      "loss": 0.4326,
+      "step": 18256
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 20.157667397778972,
+      "learning_rate": 1.2778972677499057e-07,
+      "loss": 0.4084,
+      "step": 18257
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 26.398294822711463,
+      "learning_rate": 1.2752112772865144e-07,
+      "loss": 0.4793,
+      "step": 18258
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.819168584249628,
+      "learning_rate": 1.27252809449836e-07,
+      "loss": 0.4424,
+      "step": 18259
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 34.710535464812274,
+      "learning_rate": 1.2698477194617475e-07,
+      "loss": 0.4617,
+      "step": 18260
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.050015686515742,
+      "learning_rate": 1.2671701522529055e-07,
+      "loss": 0.3938,
+      "step": 18261
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 23.491837843033025,
+      "learning_rate": 1.2644953929479954e-07,
+      "loss": 0.4553,
+      "step": 18262
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 32.02693140364885,
+      "learning_rate": 1.2618234416230778e-07,
+      "loss": 0.4523,
+      "step": 18263
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.53436103032286,
+      "learning_rate": 1.2591542983541371e-07,
+      "loss": 0.4631,
+      "step": 18264
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 15.125403133844381,
+      "learning_rate": 1.25648796321709e-07,
+      "loss": 0.4357,
+      "step": 18265
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 30.400829232860094,
+      "learning_rate": 1.253824436287765e-07,
+      "loss": 0.4937,
+      "step": 18266
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 27.894364971225666,
+      "learning_rate": 1.2511637176419012e-07,
+      "loss": 0.4206,
+      "step": 18267
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 35.16153862190187,
+      "learning_rate": 1.2485058073551825e-07,
+      "loss": 0.4851,
+      "step": 18268
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 19.18343405947422,
+      "learning_rate": 1.2458507055031822e-07,
+      "loss": 0.4282,
+      "step": 18269
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 15.941899986257978,
+      "learning_rate": 1.2431984121614282e-07,
+      "loss": 0.4293,
+      "step": 18270
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 19.170185557965027,
+      "learning_rate": 1.2405489274053273e-07,
+      "loss": 0.4209,
+      "step": 18271
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.26559713335416,
+      "learning_rate": 1.2379022513102524e-07,
+      "loss": 0.4748,
+      "step": 18272
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.58268633357884,
+      "learning_rate": 1.2352583839514655e-07,
+      "loss": 0.3817,
+      "step": 18273
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 31.294829727891404,
+      "learning_rate": 1.2326173254041506e-07,
+      "loss": 0.4544,
+      "step": 18274
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 19.57193461527387,
+      "learning_rate": 1.229979075743415e-07,
+      "loss": 0.4164,
+      "step": 18275
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 27.49040521870134,
+      "learning_rate": 1.227343635044298e-07,
+      "loss": 0.4649,
+      "step": 18276
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 30.295074280515394,
+      "learning_rate": 1.224711003381751e-07,
+      "loss": 0.4698,
+      "step": 18277
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.15709087191234,
+      "learning_rate": 1.2220811808306477e-07,
+      "loss": 0.4272,
+      "step": 18278
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.525911835352204,
+      "learning_rate": 1.219454167465761e-07,
+      "loss": 0.4171,
+      "step": 18279
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.250762091936767,
+      "learning_rate": 1.2168299633618209e-07,
+      "loss": 0.4116,
+      "step": 18280
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.535413923673268,
+      "learning_rate": 1.2142085685934446e-07,
+      "loss": 0.446,
+      "step": 18281
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.227910429830516,
+      "learning_rate": 1.2115899832351952e-07,
+      "loss": 0.4088,
+      "step": 18282
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.945430437984616,
+      "learning_rate": 1.2089742073615352e-07,
+      "loss": 0.4462,
+      "step": 18283
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.46003956677633,
+      "learning_rate": 1.2063612410468496e-07,
+      "loss": 0.4724,
+      "step": 18284
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.531547658743047,
+      "learning_rate": 1.2037510843654565e-07,
+      "loss": 0.3948,
+      "step": 18285
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.28698440908482,
+      "learning_rate": 1.201143737391597e-07,
+      "loss": 0.4691,
+      "step": 18286
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.219199387627988,
+      "learning_rate": 1.1985392001994112e-07,
+      "loss": 0.424,
+      "step": 18287
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.407352152367334,
+      "learning_rate": 1.1959374728629847e-07,
+      "loss": 0.4237,
+      "step": 18288
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 28.14379353294335,
+      "learning_rate": 1.1933385554562803e-07,
+      "loss": 0.4131,
+      "step": 18289
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.762368673543325,
+      "learning_rate": 1.1907424480532282e-07,
+      "loss": 0.4638,
+      "step": 18290
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.293099082992246,
+      "learning_rate": 1.1881491507276688e-07,
+      "loss": 0.4623,
+      "step": 18291
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.96405503489841,
+      "learning_rate": 1.1855586635533434e-07,
+      "loss": 0.4386,
+      "step": 18292
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.452285267790614,
+      "learning_rate": 1.1829709866039263e-07,
+      "loss": 0.4069,
+      "step": 18293
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.581660493226238,
+      "learning_rate": 1.180386119952992e-07,
+      "loss": 0.3631,
+      "step": 18294
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.5751306664675,
+      "learning_rate": 1.1778040636740928e-07,
+      "loss": 0.3806,
+      "step": 18295
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.562729975952855,
+      "learning_rate": 1.1752248178406144e-07,
+      "loss": 0.4119,
+      "step": 18296
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.991797096186204,
+      "learning_rate": 1.1726483825259427e-07,
+      "loss": 0.4657,
+      "step": 18297
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.926652440305812,
+      "learning_rate": 1.1700747578033412e-07,
+      "loss": 0.4166,
+      "step": 18298
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.856143375593142,
+      "learning_rate": 1.1675039437459956e-07,
+      "loss": 0.4117,
+      "step": 18299
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 25.064544546774627,
+      "learning_rate": 1.1649359404270144e-07,
+      "loss": 0.5282,
+      "step": 18300
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.1012338751548,
+      "learning_rate": 1.1623707479194501e-07,
+      "loss": 0.4263,
+      "step": 18301
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.484279506911268,
+      "learning_rate": 1.1598083662962335e-07,
+      "loss": 0.4213,
+      "step": 18302
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.43540317695954,
+      "learning_rate": 1.1572487956302613e-07,
+      "loss": 0.4002,
+      "step": 18303
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 18.724048546019766,
+      "learning_rate": 1.1546920359943092e-07,
+      "loss": 0.4233,
+      "step": 18304
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.13283798311038,
+      "learning_rate": 1.1521380874610854e-07,
+      "loss": 0.5084,
+      "step": 18305
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.356874741693716,
+      "learning_rate": 1.1495869501032209e-07,
+      "loss": 0.4074,
+      "step": 18306
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 30.088977923583656,
+      "learning_rate": 1.1470386239932907e-07,
+      "loss": 0.4498,
+      "step": 18307
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.527666461746104,
+      "learning_rate": 1.1444931092037592e-07,
+      "loss": 0.4372,
+      "step": 18308
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.48198879070703,
+      "learning_rate": 1.141950405807013e-07,
+      "loss": 0.4184,
+      "step": 18309
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.287432520406412,
+      "learning_rate": 1.1394105138753498e-07,
+      "loss": 0.394,
+      "step": 18310
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.894798388320066,
+      "learning_rate": 1.1368734334810339e-07,
+      "loss": 0.4483,
+      "step": 18311
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.3576798947693,
+      "learning_rate": 1.1343391646962076e-07,
+      "loss": 0.4501,
+      "step": 18312
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.284140015930348,
+      "learning_rate": 1.1318077075929357e-07,
+      "loss": 0.437,
+      "step": 18313
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.4743317261089,
+      "learning_rate": 1.129279062243216e-07,
+      "loss": 0.4076,
+      "step": 18314
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 26.494946311682842,
+      "learning_rate": 1.1267532287189576e-07,
+      "loss": 0.4364,
+      "step": 18315
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.778848902942965,
+      "learning_rate": 1.1242302070920141e-07,
+      "loss": 0.3701,
+      "step": 18316
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 32.33168403832644,
+      "learning_rate": 1.1217099974341173e-07,
+      "loss": 0.5231,
+      "step": 18317
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.193657443112794,
+      "learning_rate": 1.1191925998169317e-07,
+      "loss": 0.4437,
+      "step": 18318
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.553740629145125,
+      "learning_rate": 1.1166780143120781e-07,
+      "loss": 0.376,
+      "step": 18319
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.83741249538844,
+      "learning_rate": 1.1141662409910547e-07,
+      "loss": 0.3954,
+      "step": 18320
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.31835728310645,
+      "learning_rate": 1.1116572799252934e-07,
+      "loss": 0.4073,
+      "step": 18321
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.507013174424475,
+      "learning_rate": 1.109151131186148e-07,
+      "loss": 0.4684,
+      "step": 18322
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 18.420351265117326,
+      "learning_rate": 1.106647794844895e-07,
+      "loss": 0.446,
+      "step": 18323
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 30.284290627564843,
+      "learning_rate": 1.104147270972733e-07,
+      "loss": 0.4706,
+      "step": 18324
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.72598521107956,
+      "learning_rate": 1.1016495596407717e-07,
+      "loss": 0.3406,
+      "step": 18325
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 30.89734352383938,
+      "learning_rate": 1.0991546609200432e-07,
+      "loss": 0.4189,
+      "step": 18326
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 32.40921275179439,
+      "learning_rate": 1.0966625748815019e-07,
+      "loss": 0.4185,
+      "step": 18327
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 25.249523932033956,
+      "learning_rate": 1.0941733015960132e-07,
+      "loss": 0.3936,
+      "step": 18328
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.365700176310437,
+      "learning_rate": 1.0916868411343872e-07,
+      "loss": 0.41,
+      "step": 18329
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.20717487894361,
+      "learning_rate": 1.089203193567312e-07,
+      "loss": 0.3419,
+      "step": 18330
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 18.1012528787738,
+      "learning_rate": 1.0867223589654418e-07,
+      "loss": 0.4043,
+      "step": 18331
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.8934833337288,
+      "learning_rate": 1.0842443373993427e-07,
+      "loss": 0.4408,
+      "step": 18332
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.506846019694269,
+      "learning_rate": 1.0817691289394583e-07,
+      "loss": 0.4368,
+      "step": 18333
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 25.07982944626364,
+      "learning_rate": 1.0792967336561877e-07,
+      "loss": 0.4592,
+      "step": 18334
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.647404359982936,
+      "learning_rate": 1.0768271516198525e-07,
+      "loss": 0.3941,
+      "step": 18335
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.869458228862086,
+      "learning_rate": 1.0743603829006854e-07,
+      "loss": 0.4166,
+      "step": 18336
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.32477108863656,
+      "learning_rate": 1.0718964275688526e-07,
+      "loss": 0.3814,
+      "step": 18337
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.72555269787382,
+      "learning_rate": 1.069435285694409e-07,
+      "loss": 0.4282,
+      "step": 18338
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.716896747793303,
+      "learning_rate": 1.0669769573473432e-07,
+      "loss": 0.4296,
+      "step": 18339
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.428220041937145,
+      "learning_rate": 1.064521442597577e-07,
+      "loss": 0.4553,
+      "step": 18340
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.108314938081648,
+      "learning_rate": 1.0620687415149655e-07,
+      "loss": 0.4194,
+      "step": 18341
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 34.529464409693254,
+      "learning_rate": 1.0596188541692309e-07,
+      "loss": 0.471,
+      "step": 18342
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.951167182996517,
+      "learning_rate": 1.0571717806300508e-07,
+      "loss": 0.4234,
+      "step": 18343
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.66355459530119,
+      "learning_rate": 1.0547275209670249e-07,
+      "loss": 0.3926,
+      "step": 18344
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 30.87482004653751,
+      "learning_rate": 1.0522860752496755e-07,
+      "loss": 0.4628,
+      "step": 18345
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 28.40349049640971,
+      "learning_rate": 1.0498474435474249e-07,
+      "loss": 0.413,
+      "step": 18346
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.8221611035089,
+      "learning_rate": 1.0474116259296285e-07,
+      "loss": 0.3641,
+      "step": 18347
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 17.732197671568045,
+      "learning_rate": 1.0449786224655644e-07,
+      "loss": 0.4381,
+      "step": 18348
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.640845109168804,
+      "learning_rate": 1.0425484332244107e-07,
+      "loss": 0.4326,
+      "step": 18349
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.5540106365669,
+      "learning_rate": 1.0401210582753007e-07,
+      "loss": 0.3778,
+      "step": 18350
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.943376669544964,
+      "learning_rate": 1.0376964976872461e-07,
+      "loss": 0.4012,
+      "step": 18351
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.3970843567236,
+      "learning_rate": 1.035274751529225e-07,
+      "loss": 0.4325,
+      "step": 18352
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.53049528656263,
+      "learning_rate": 1.0328558198700933e-07,
+      "loss": 0.4194,
+      "step": 18353
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 28.183961027611318,
+      "learning_rate": 1.0304397027786516e-07,
+      "loss": 0.3995,
+      "step": 18354
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 32.494961688259316,
+      "learning_rate": 1.0280264003236006e-07,
+      "loss": 0.4576,
+      "step": 18355
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.538597885877973,
+      "learning_rate": 1.0256159125735742e-07,
+      "loss": 0.4198,
+      "step": 18356
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.949595634207135,
+      "learning_rate": 1.023208239597151e-07,
+      "loss": 0.4104,
+      "step": 18357
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.961369053442983,
+      "learning_rate": 1.0208033814627872e-07,
+      "loss": 0.4271,
+      "step": 18358
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.07025856054119,
+      "learning_rate": 1.0184013382388613e-07,
+      "loss": 0.4391,
+      "step": 18359
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 28.14149565993997,
+      "learning_rate": 1.0160021099936967e-07,
+      "loss": 0.4248,
+      "step": 18360
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 27.49864984687522,
+      "learning_rate": 1.0136056967955388e-07,
+      "loss": 0.4734,
+      "step": 18361
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.68384317890662,
+      "learning_rate": 1.0112120987125329e-07,
+      "loss": 0.4223,
+      "step": 18362
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.888227326931965,
+      "learning_rate": 1.0088213158127358e-07,
+      "loss": 0.4164,
+      "step": 18363
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.193933531366707,
+      "learning_rate": 1.0064333481641597e-07,
+      "loss": 0.4623,
+      "step": 18364
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.71110605223992,
+      "learning_rate": 1.0040481958346948e-07,
+      "loss": 0.4058,
+      "step": 18365
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 25.67617787563051,
+      "learning_rate": 1.0016658588922091e-07,
+      "loss": 0.4615,
+      "step": 18366
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 26.667018233654044,
+      "learning_rate": 9.99286337404426e-08,
+      "loss": 0.4211,
+      "step": 18367
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 15.43700196258174,
+      "learning_rate": 9.969096314390248e-08,
+      "loss": 0.4526,
+      "step": 18368
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 25.116893618274897,
+      "learning_rate": 9.945357410635959e-08,
+      "loss": 0.478,
+      "step": 18369
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 17.175931624933508,
+      "learning_rate": 9.92164666345663e-08,
+      "loss": 0.4395,
+      "step": 18370
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 22.993850233103725,
+      "learning_rate": 9.8979640735265e-08,
+      "loss": 0.4018,
+      "step": 18371
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.71850141414339,
+      "learning_rate": 9.87430964151892e-08,
+      "loss": 0.4075,
+      "step": 18372
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 19.911529531824726,
+      "learning_rate": 9.850683368106794e-08,
+      "loss": 0.3908,
+      "step": 18373
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 17.664318398786268,
+      "learning_rate": 9.827085253962144e-08,
+      "loss": 0.3525,
+      "step": 18374
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.89809913931761,
+      "learning_rate": 9.803515299755872e-08,
+      "loss": 0.3946,
+      "step": 18375
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 25.24531524817793,
+      "learning_rate": 9.779973506158446e-08,
+      "loss": 0.4048,
+      "step": 18376
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.046205461876834,
+      "learning_rate": 9.75645987383922e-08,
+      "loss": 0.4428,
+      "step": 18377
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 23.8457090734149,
+      "learning_rate": 9.732974403466988e-08,
+      "loss": 0.4265,
+      "step": 18378
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.375499224514808,
+      "learning_rate": 9.709517095709775e-08,
+      "loss": 0.3892,
+      "step": 18379
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 30.318810043126096,
+      "learning_rate": 9.68608795123449e-08,
+      "loss": 0.4579,
+      "step": 18380
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 23.540822286499488,
+      "learning_rate": 9.662686970707602e-08,
+      "loss": 0.4104,
+      "step": 18381
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.78546797700525,
+      "learning_rate": 9.639314154794576e-08,
+      "loss": 0.45,
+      "step": 18382
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 29.403820297245723,
+      "learning_rate": 9.615969504160106e-08,
+      "loss": 0.4565,
+      "step": 18383
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.36859784779545,
+      "learning_rate": 9.5926530194681e-08,
+      "loss": 0.4095,
+      "step": 18384
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 27.500862742025546,
+      "learning_rate": 9.569364701381589e-08,
+      "loss": 0.425,
+      "step": 18385
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.634175997254733,
+      "learning_rate": 9.546104550563152e-08,
+      "loss": 0.3868,
+      "step": 18386
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 28.924624283014616,
+      "learning_rate": 9.522872567673924e-08,
+      "loss": 0.4375,
+      "step": 18387
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 25.80431594257756,
+      "learning_rate": 9.499668753374936e-08,
+      "loss": 0.4815,
+      "step": 18388
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.989342069424193,
+      "learning_rate": 9.476493108325768e-08,
+      "loss": 0.48,
+      "step": 18389
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 17.54848776409242,
+      "learning_rate": 9.453345633185784e-08,
+      "loss": 0.4332,
+      "step": 18390
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 28.07947207060383,
+      "learning_rate": 9.430226328613234e-08,
+      "loss": 0.4644,
+      "step": 18391
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.896055242361882,
+      "learning_rate": 9.407135195265593e-08,
+      "loss": 0.4584,
+      "step": 18392
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.43880150489254,
+      "learning_rate": 9.384072233799447e-08,
+      "loss": 0.44,
+      "step": 18393
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 38.68869021191799,
+      "learning_rate": 9.361037444870824e-08,
+      "loss": 0.3873,
+      "step": 18394
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.748485184951647,
+      "learning_rate": 9.33803082913487e-08,
+      "loss": 0.4395,
+      "step": 18395
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 22.035705717796795,
+      "learning_rate": 9.315052387245726e-08,
+      "loss": 0.4298,
+      "step": 18396
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.570742163152755,
+      "learning_rate": 9.292102119856983e-08,
+      "loss": 0.4361,
+      "step": 18397
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.56563215726859,
+      "learning_rate": 9.269180027621228e-08,
+      "loss": 0.4201,
+      "step": 18398
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.876684476042943,
+      "learning_rate": 9.246286111190495e-08,
+      "loss": 0.4364,
+      "step": 18399
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 22.931858661805176,
+      "learning_rate": 9.223420371215818e-08,
+      "loss": 0.3939,
+      "step": 18400
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.19178338097972,
+      "learning_rate": 9.200582808347569e-08,
+      "loss": 0.3766,
+      "step": 18401
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 26.112399496229294,
+      "learning_rate": 9.177773423235004e-08,
+      "loss": 0.4233,
+      "step": 18402
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.344523069302543,
+      "learning_rate": 9.154992216526937e-08,
+      "loss": 0.3983,
+      "step": 18403
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 35.93790694628309,
+      "learning_rate": 9.132239188871405e-08,
+      "loss": 0.431,
+      "step": 18404
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 20.423853945781914,
+      "learning_rate": 9.109514340915226e-08,
+      "loss": 0.4403,
+      "step": 18405
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 32.61197658236122,
+      "learning_rate": 9.086817673304882e-08,
+      "loss": 0.4037,
+      "step": 18406
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 22.570937209101274,
+      "learning_rate": 9.064149186685744e-08,
+      "loss": 0.4733,
+      "step": 18407
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.297242912460888,
+      "learning_rate": 9.041508881702521e-08,
+      "loss": 0.3966,
+      "step": 18408
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.985048350014203,
+      "learning_rate": 9.018896758999029e-08,
+      "loss": 0.4334,
+      "step": 18409
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 22.705161815201944,
+      "learning_rate": 8.996312819218533e-08,
+      "loss": 0.4409,
+      "step": 18410
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 17.741281472827755,
+      "learning_rate": 8.973757063003075e-08,
+      "loss": 0.371,
+      "step": 18411
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 14.623725437191284,
+      "learning_rate": 8.95122949099425e-08,
+      "loss": 0.4228,
+      "step": 18412
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.906957812172738,
+      "learning_rate": 8.928730103832773e-08,
+      "loss": 0.4596,
+      "step": 18413
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 17.49742682987387,
+      "learning_rate": 8.90625890215846e-08,
+      "loss": 0.381,
+      "step": 18414
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 39.01680808513573,
+      "learning_rate": 8.883815886610358e-08,
+      "loss": 0.496,
+      "step": 18415
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 22.809953867555414,
+      "learning_rate": 8.861401057826624e-08,
+      "loss": 0.4104,
+      "step": 18416
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 27.70306184334395,
+      "learning_rate": 8.839014416444969e-08,
+      "loss": 0.4811,
+      "step": 18417
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.58237640362796,
+      "learning_rate": 8.816655963101883e-08,
+      "loss": 0.4681,
+      "step": 18418
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 28.904941402404614,
+      "learning_rate": 8.794325698433193e-08,
+      "loss": 0.4524,
+      "step": 18419
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.799944844158663,
+      "learning_rate": 8.772023623074166e-08,
+      "loss": 0.4036,
+      "step": 18420
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 26.88421682233436,
+      "learning_rate": 8.749749737658964e-08,
+      "loss": 0.4637,
+      "step": 18421
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 34.75290987983933,
+      "learning_rate": 8.727504042820966e-08,
+      "loss": 0.4365,
+      "step": 18422
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 25.53116679907588,
+      "learning_rate": 8.705286539192781e-08,
+      "loss": 0.4023,
+      "step": 18423
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.29214367331964,
+      "learning_rate": 8.683097227406456e-08,
+      "loss": 0.4595,
+      "step": 18424
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.424526102433298,
+      "learning_rate": 8.660936108093043e-08,
+      "loss": 0.4449,
+      "step": 18425
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.800541174967424,
+      "learning_rate": 8.638803181882483e-08,
+      "loss": 0.4619,
+      "step": 18426
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.868383997538675,
+      "learning_rate": 8.616698449404492e-08,
+      "loss": 0.4536,
+      "step": 18427
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 27.182119321572173,
+      "learning_rate": 8.594621911287682e-08,
+      "loss": 0.4578,
+      "step": 18428
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.349822477599126,
+      "learning_rate": 8.572573568159881e-08,
+      "loss": 0.4155,
+      "step": 18429
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.235199743034283,
+      "learning_rate": 8.550553420648145e-08,
+      "loss": 0.426,
+      "step": 18430
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 29.155371849240236,
+      "learning_rate": 8.528561469378637e-08,
+      "loss": 0.4482,
+      "step": 18431
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.000468394635895,
+      "learning_rate": 8.50659771497686e-08,
+      "loss": 0.4205,
+      "step": 18432
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 27.272111640289037,
+      "learning_rate": 8.484662158067425e-08,
+      "loss": 0.5392,
+      "step": 18433
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.60949134149706,
+      "learning_rate": 8.462754799274275e-08,
+      "loss": 0.3962,
+      "step": 18434
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.73631820588152,
+      "learning_rate": 8.440875639220247e-08,
+      "loss": 0.3951,
+      "step": 18435
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 32.32679850814688,
+      "learning_rate": 8.41902467852762e-08,
+      "loss": 0.4657,
+      "step": 18436
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.361122131769324,
+      "learning_rate": 8.397201917817899e-08,
+      "loss": 0.4396,
+      "step": 18437
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.761292744870147,
+      "learning_rate": 8.375407357711806e-08,
+      "loss": 0.4351,
+      "step": 18438
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.202548007847533,
+      "learning_rate": 8.35364099882896e-08,
+      "loss": 0.4313,
+      "step": 18439
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.581560612045358,
+      "learning_rate": 8.331902841788531e-08,
+      "loss": 0.3866,
+      "step": 18440
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 25.526838452592482,
+      "learning_rate": 8.310192887208579e-08,
+      "loss": 0.4333,
+      "step": 18441
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.345680104239268,
+      "learning_rate": 8.288511135706611e-08,
+      "loss": 0.4141,
+      "step": 18442
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.12467925970497,
+      "learning_rate": 8.266857587899357e-08,
+      "loss": 0.4855,
+      "step": 18443
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.38043819880168,
+      "learning_rate": 8.245232244402435e-08,
+      "loss": 0.4931,
+      "step": 18444
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 26.510071820912223,
+      "learning_rate": 8.223635105831018e-08,
+      "loss": 0.4557,
+      "step": 18445
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 30.774422568064423,
+      "learning_rate": 8.202066172799172e-08,
+      "loss": 0.4404,
+      "step": 18446
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.156260721810025,
+      "learning_rate": 8.180525445920407e-08,
+      "loss": 0.4852,
+      "step": 18447
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.6878259796467,
+      "learning_rate": 8.159012925807341e-08,
+      "loss": 0.3993,
+      "step": 18448
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 39.4023906835687,
+      "learning_rate": 8.1375286130716e-08,
+      "loss": 0.4522,
+      "step": 18449
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.398988439646708,
+      "learning_rate": 8.116072508324579e-08,
+      "loss": 0.3749,
+      "step": 18450
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 35.94074062165958,
+      "learning_rate": 8.094644612176017e-08,
+      "loss": 0.4712,
+      "step": 18451
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.645613179361575,
+      "learning_rate": 8.073244925235535e-08,
+      "loss": 0.4866,
+      "step": 18452
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.49746228232012,
+      "learning_rate": 8.051873448111758e-08,
+      "loss": 0.4786,
+      "step": 18453
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.18045609404505,
+      "learning_rate": 8.030530181412422e-08,
+      "loss": 0.4345,
+      "step": 18454
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.963552300654136,
+      "learning_rate": 8.009215125744486e-08,
+      "loss": 0.4205,
+      "step": 18455
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.741096115634896,
+      "learning_rate": 7.987928281714241e-08,
+      "loss": 0.4457,
+      "step": 18456
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 17.735347494693396,
+      "learning_rate": 7.966669649926872e-08,
+      "loss": 0.4812,
+      "step": 18457
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.551751869945793,
+      "learning_rate": 7.945439230987228e-08,
+      "loss": 0.3879,
+      "step": 18458
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.915078012297402,
+      "learning_rate": 7.924237025498938e-08,
+      "loss": 0.3645,
+      "step": 18459
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 26.16597690965238,
+      "learning_rate": 7.903063034064962e-08,
+      "loss": 0.4434,
+      "step": 18460
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 22.936425450228953,
+      "learning_rate": 7.881917257287485e-08,
+      "loss": 0.4915,
+      "step": 18461
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.1757377045868,
+      "learning_rate": 7.860799695767918e-08,
+      "loss": 0.4435,
+      "step": 18462
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.778394975136973,
+      "learning_rate": 7.839710350106888e-08,
+      "loss": 0.4262,
+      "step": 18463
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.386251898154999,
+      "learning_rate": 7.81864922090414e-08,
+      "loss": 0.3686,
+      "step": 18464
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.32013498595952,
+      "learning_rate": 7.797616308758526e-08,
+      "loss": 0.4027,
+      "step": 18465
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.44788166374201,
+      "learning_rate": 7.776611614268236e-08,
+      "loss": 0.4164,
+      "step": 18466
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.394630355713797,
+      "learning_rate": 7.755635138030682e-08,
+      "loss": 0.4065,
+      "step": 18467
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 37.130097902971805,
+      "learning_rate": 7.734686880642495e-08,
+      "loss": 0.548,
+      "step": 18468
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 26.742833770222312,
+      "learning_rate": 7.71376684269931e-08,
+      "loss": 0.3763,
+      "step": 18469
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 30.812196427333543,
+      "learning_rate": 7.692875024796099e-08,
+      "loss": 0.4334,
+      "step": 18470
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 29.26335271271613,
+      "learning_rate": 7.672011427527159e-08,
+      "loss": 0.4692,
+      "step": 18471
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.540203146146805,
+      "learning_rate": 7.651176051485576e-08,
+      "loss": 0.4594,
+      "step": 18472
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 33.6817246831706,
+      "learning_rate": 7.630368897264096e-08,
+      "loss": 0.4307,
+      "step": 18473
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.05279806573189,
+      "learning_rate": 7.609589965454356e-08,
+      "loss": 0.3759,
+      "step": 18474
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.47577836004555,
+      "learning_rate": 7.58883925664744e-08,
+      "loss": 0.4349,
+      "step": 18475
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.057102090107257,
+      "learning_rate": 7.56811677143332e-08,
+      "loss": 0.4606,
+      "step": 18476
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.720885014053273,
+      "learning_rate": 7.547422510401414e-08,
+      "loss": 0.4472,
+      "step": 18477
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.19976653725924,
+      "learning_rate": 7.526756474140252e-08,
+      "loss": 0.4428,
+      "step": 18478
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 33.325196305159665,
+      "learning_rate": 7.506118663237583e-08,
+      "loss": 0.4944,
+      "step": 18479
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.749362944541655,
+      "learning_rate": 7.485509078280384e-08,
+      "loss": 0.4245,
+      "step": 18480
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.145792908361546,
+      "learning_rate": 7.46492771985452e-08,
+      "loss": 0.3999,
+      "step": 18481
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.13890644840671,
+      "learning_rate": 7.444374588545522e-08,
+      "loss": 0.4817,
+      "step": 18482
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.025304870086725,
+      "learning_rate": 7.423849684938033e-08,
+      "loss": 0.4045,
+      "step": 18483
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 30.917424138972326,
+      "learning_rate": 7.403353009615588e-08,
+      "loss": 0.4678,
+      "step": 18484
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.667072842114589,
+      "learning_rate": 7.382884563161052e-08,
+      "loss": 0.4348,
+      "step": 18485
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.697520489006024,
+      "learning_rate": 7.362444346156517e-08,
+      "loss": 0.4321,
+      "step": 18486
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.577317795299162,
+      "learning_rate": 7.342032359183515e-08,
+      "loss": 0.4574,
+      "step": 18487
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 25.125105402141795,
+      "learning_rate": 7.321648602822473e-08,
+      "loss": 0.4092,
+      "step": 18488
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 26.371183509091896,
+      "learning_rate": 7.301293077653038e-08,
+      "loss": 0.3988,
+      "step": 18489
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.46116850182559,
+      "learning_rate": 7.280965784253968e-08,
+      "loss": 0.4658,
+      "step": 18490
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.9599841667247,
+      "learning_rate": 7.26066672320358e-08,
+      "loss": 0.4495,
+      "step": 18491
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.241037532658318,
+      "learning_rate": 7.240395895079189e-08,
+      "loss": 0.4071,
+      "step": 18492
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.04750890978004,
+      "learning_rate": 7.22015330045711e-08,
+      "loss": 0.38,
+      "step": 18493
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.986960141430107,
+      "learning_rate": 7.199938939913109e-08,
+      "loss": 0.4363,
+      "step": 18494
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.181144282223094,
+      "learning_rate": 7.179752814022056e-08,
+      "loss": 0.399,
+      "step": 18495
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.774148673717654,
+      "learning_rate": 7.159594923358159e-08,
+      "loss": 0.3851,
+      "step": 18496
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.196637559690757,
+      "learning_rate": 7.139465268494405e-08,
+      "loss": 0.4522,
+      "step": 18497
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.357463917609813,
+      "learning_rate": 7.119363850003558e-08,
+      "loss": 0.4343,
+      "step": 18498
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.695675280088306,
+      "learning_rate": 7.09929066845727e-08,
+      "loss": 0.3993,
+      "step": 18499
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.779029934285226,
+      "learning_rate": 7.079245724426198e-08,
+      "loss": 0.4216,
+      "step": 18500
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.167377265631004,
+      "learning_rate": 7.059229018480551e-08,
+      "loss": 0.4308,
+      "step": 18501
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 30.58525938254278,
+      "learning_rate": 7.039240551189653e-08,
+      "loss": 0.4907,
+      "step": 18502
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 31.406738420740492,
+      "learning_rate": 7.019280323121714e-08,
+      "loss": 0.484,
+      "step": 18503
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 24.438240513838203,
+      "learning_rate": 6.999348334844724e-08,
+      "loss": 0.4921,
+      "step": 18504
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 24.756662620740393,
+      "learning_rate": 6.979444586925455e-08,
+      "loss": 0.4635,
+      "step": 18505
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.424191867938323,
+      "learning_rate": 6.959569079929674e-08,
+      "loss": 0.45,
+      "step": 18506
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.418451151997708,
+      "learning_rate": 6.939721814422928e-08,
+      "loss": 0.3599,
+      "step": 18507
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 26.611592926667754,
+      "learning_rate": 6.919902790969657e-08,
+      "loss": 0.5605,
+      "step": 18508
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 27.175031345401116,
+      "learning_rate": 6.900112010133297e-08,
+      "loss": 0.4164,
+      "step": 18509
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.911385013371248,
+      "learning_rate": 6.880349472476844e-08,
+      "loss": 0.409,
+      "step": 18510
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 25.36275611826147,
+      "learning_rate": 6.86061517856229e-08,
+      "loss": 0.4153,
+      "step": 18511
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 24.641146904686117,
+      "learning_rate": 6.840909128950857e-08,
+      "loss": 0.4182,
+      "step": 18512
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.634173210946592,
+      "learning_rate": 6.821231324203093e-08,
+      "loss": 0.4154,
+      "step": 18513
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.822013163736386,
+      "learning_rate": 6.801581764878329e-08,
+      "loss": 0.4476,
+      "step": 18514
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.616144023306047,
+      "learning_rate": 6.781960451535674e-08,
+      "loss": 0.3498,
+      "step": 18515
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.768925716139016,
+      "learning_rate": 6.762367384733015e-08,
+      "loss": 0.43,
+      "step": 18516
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.002053399482897,
+      "learning_rate": 6.742802565027684e-08,
+      "loss": 0.4196,
+      "step": 18517
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 27.32091563766875,
+      "learning_rate": 6.723265992975903e-08,
+      "loss": 0.4213,
+      "step": 18518
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 25.195549789612453,
+      "learning_rate": 6.703757669133448e-08,
+      "loss": 0.4309,
+      "step": 18519
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.58093273626427,
+      "learning_rate": 6.6842775940551e-08,
+      "loss": 0.4127,
+      "step": 18520
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 25.544681947236025,
+      "learning_rate": 6.664825768294747e-08,
+      "loss": 0.4339,
+      "step": 18521
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.074768979841554,
+      "learning_rate": 6.645402192405726e-08,
+      "loss": 0.4174,
+      "step": 18522
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 30.15982142829394,
+      "learning_rate": 6.626006866940371e-08,
+      "loss": 0.4876,
+      "step": 18523
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.21421075210531,
+      "learning_rate": 6.606639792450131e-08,
+      "loss": 0.427,
+      "step": 18524
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.43151432509714,
+      "learning_rate": 6.587300969486122e-08,
+      "loss": 0.4401,
+      "step": 18525
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.738934036978918,
+      "learning_rate": 6.567990398598012e-08,
+      "loss": 0.4326,
+      "step": 18526
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 28.927823425160025,
+      "learning_rate": 6.548708080335143e-08,
+      "loss": 0.415,
+      "step": 18527
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 23.358991507504868,
+      "learning_rate": 6.529454015245962e-08,
+      "loss": 0.4195,
+      "step": 18528
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 26.89922264475566,
+      "learning_rate": 6.510228203877811e-08,
+      "loss": 0.4366,
+      "step": 18529
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 29.8925049106383,
+      "learning_rate": 6.491030646777696e-08,
+      "loss": 0.4165,
+      "step": 18530
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.89408575877051,
+      "learning_rate": 6.471861344491514e-08,
+      "loss": 0.4147,
+      "step": 18531
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.746844744221828,
+      "learning_rate": 6.452720297564275e-08,
+      "loss": 0.4141,
+      "step": 18532
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.393911793127998,
+      "learning_rate": 6.43360750654054e-08,
+      "loss": 0.4115,
+      "step": 18533
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 24.086794263540646,
+      "learning_rate": 6.414522971963766e-08,
+      "loss": 0.4074,
+      "step": 18534
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 33.061442889204315,
+      "learning_rate": 6.395466694376739e-08,
+      "loss": 0.417,
+      "step": 18535
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 24.032101190098853,
+      "learning_rate": 6.376438674321361e-08,
+      "loss": 0.4324,
+      "step": 18536
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.81695936901427,
+      "learning_rate": 6.357438912338864e-08,
+      "loss": 0.4226,
+      "step": 18537
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 28.27477320593617,
+      "learning_rate": 6.338467408969484e-08,
+      "loss": 0.4412,
+      "step": 18538
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 23.402800398745548,
+      "learning_rate": 6.319524164752899e-08,
+      "loss": 0.4642,
+      "step": 18539
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 27.7534454951741,
+      "learning_rate": 6.300609180227679e-08,
+      "loss": 0.4372,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.486044208739152,
+      "learning_rate": 6.281722455931839e-08,
+      "loss": 0.3834,
+      "step": 18541
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.74008368530438,
+      "learning_rate": 6.262863992402612e-08,
+      "loss": 0.4021,
+      "step": 18542
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 27.894347682381458,
+      "learning_rate": 6.24403379017613e-08,
+      "loss": 0.4531,
+      "step": 18543
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 30.498933595634526,
+      "learning_rate": 6.225231849787961e-08,
+      "loss": 0.4429,
+      "step": 18544
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 24.38677491513527,
+      "learning_rate": 6.2064581717729e-08,
+      "loss": 0.4493,
+      "step": 18545
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 32.29721347417395,
+      "learning_rate": 6.187712756664854e-08,
+      "loss": 0.4592,
+      "step": 18546
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 36.42578139317899,
+      "learning_rate": 6.168995604996841e-08,
+      "loss": 0.4573,
+      "step": 18547
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.88275092978021,
+      "learning_rate": 6.150306717301213e-08,
+      "loss": 0.4129,
+      "step": 18548
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 26.442839142550525,
+      "learning_rate": 6.131646094109545e-08,
+      "loss": 0.4592,
+      "step": 18549
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 23.199971906040194,
+      "learning_rate": 6.1130137359523e-08,
+      "loss": 0.4071,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 41.83557046894883,
+      "learning_rate": 6.094409643359722e-08,
+      "loss": 0.5329,
+      "step": 18551
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.64986434035781,
+      "learning_rate": 6.07583381686061e-08,
+      "loss": 0.4239,
+      "step": 18552
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 28.978336676913905,
+      "learning_rate": 6.05728625698343e-08,
+      "loss": 0.4062,
+      "step": 18553
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 20.85896528963201,
+      "learning_rate": 6.038766964255538e-08,
+      "loss": 0.3951,
+      "step": 18554
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.77472125437093,
+      "learning_rate": 6.020275939203734e-08,
+      "loss": 0.4777,
+      "step": 18555
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 31.66742725057573,
+      "learning_rate": 6.00181318235371e-08,
+      "loss": 0.4444,
+      "step": 18556
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 29.886319099080577,
+      "learning_rate": 5.983378694230713e-08,
+      "loss": 0.4193,
+      "step": 18557
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.01866139274335,
+      "learning_rate": 5.964972475358876e-08,
+      "loss": 0.4237,
+      "step": 18558
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.787489964367737,
+      "learning_rate": 5.946594526261895e-08,
+      "loss": 0.4168,
+      "step": 18559
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 23.870874983506553,
+      "learning_rate": 5.928244847462128e-08,
+      "loss": 0.4386,
+      "step": 18560
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 24.5904435391408,
+      "learning_rate": 5.909923439481491e-08,
+      "loss": 0.4326,
+      "step": 18561
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.30948369156938,
+      "learning_rate": 5.8916303028412334e-08,
+      "loss": 0.4061,
+      "step": 18562
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 12.587707451907432,
+      "learning_rate": 5.873365438061385e-08,
+      "loss": 0.3877,
+      "step": 18563
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 20.80746495569008,
+      "learning_rate": 5.8551288456616395e-08,
+      "loss": 0.3906,
+      "step": 18564
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 25.238816500492025,
+      "learning_rate": 5.83692052616025e-08,
+      "loss": 0.4238,
+      "step": 18565
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.389540932413485,
+      "learning_rate": 5.8187404800753575e-08,
+      "loss": 0.4388,
+      "step": 18566
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.91401453385809,
+      "learning_rate": 5.800588707923993e-08,
+      "loss": 0.3984,
+      "step": 18567
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 28.492520582732723,
+      "learning_rate": 5.782465210222299e-08,
+      "loss": 0.4057,
+      "step": 18568
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 24.379133578247433,
+      "learning_rate": 5.7643699874855296e-08,
+      "loss": 0.4185,
+      "step": 18569
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.013488443175785,
+      "learning_rate": 5.746303040228607e-08,
+      "loss": 0.4521,
+      "step": 18570
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.820375565830574,
+      "learning_rate": 5.7282643689652306e-08,
+      "loss": 0.4176,
+      "step": 18571
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 28.053879731802162,
+      "learning_rate": 5.710253974208324e-08,
+      "loss": 0.4198,
+      "step": 18572
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.947676484006685,
+      "learning_rate": 5.692271856470144e-08,
+      "loss": 0.4331,
+      "step": 18573
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 19.998587581561058,
+      "learning_rate": 5.674318016262059e-08,
+      "loss": 0.4298,
+      "step": 18574
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.38055579174918,
+      "learning_rate": 5.656392454094661e-08,
+      "loss": 0.4834,
+      "step": 18575
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.908862691443375,
+      "learning_rate": 5.638495170477876e-08,
+      "loss": 0.4222,
+      "step": 18576
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 25.870998559218247,
+      "learning_rate": 5.6206261659206284e-08,
+      "loss": 0.4519,
+      "step": 18577
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 19.67904293092041,
+      "learning_rate": 5.602785440931069e-08,
+      "loss": 0.4067,
+      "step": 18578
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 29.013571404097277,
+      "learning_rate": 5.584972996016569e-08,
+      "loss": 0.4515,
+      "step": 18579
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 34.651736640621614,
+      "learning_rate": 5.567188831683723e-08,
+      "loss": 0.5154,
+      "step": 18580
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.13680798425253,
+      "learning_rate": 5.549432948438238e-08,
+      "loss": 0.4097,
+      "step": 18581
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.473877677655718,
+      "learning_rate": 5.531705346785265e-08,
+      "loss": 0.3921,
+      "step": 18582
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.161040284591447,
+      "learning_rate": 5.514006027228735e-08,
+      "loss": 0.4131,
+      "step": 18583
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.229827713016393,
+      "learning_rate": 5.4963349902722454e-08,
+      "loss": 0.3707,
+      "step": 18584
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 20.378601134818286,
+      "learning_rate": 5.478692236418059e-08,
+      "loss": 0.4258,
+      "step": 18585
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.933541126465354,
+      "learning_rate": 5.4610777661681105e-08,
+      "loss": 0.4573,
+      "step": 18586
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 21.56154401014442,
+      "learning_rate": 5.443491580023441e-08,
+      "loss": 0.4066,
+      "step": 18587
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.57596237400272,
+      "learning_rate": 5.4259336784839855e-08,
+      "loss": 0.4055,
+      "step": 18588
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 26.93826397190838,
+      "learning_rate": 5.408404062049233e-08,
+      "loss": 0.3974,
+      "step": 18589
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.51858298798443,
+      "learning_rate": 5.390902731217562e-08,
+      "loss": 0.3898,
+      "step": 18590
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.660792571553728,
+      "learning_rate": 5.373429686486797e-08,
+      "loss": 0.4059,
+      "step": 18591
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 25.495204372916653,
+      "learning_rate": 5.355984928353986e-08,
+      "loss": 0.468,
+      "step": 18592
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 25.989328184130365,
+      "learning_rate": 5.338568457314952e-08,
+      "loss": 0.4397,
+      "step": 18593
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 20.465690008419305,
+      "learning_rate": 5.321180273865301e-08,
+      "loss": 0.4257,
+      "step": 18594
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 20.193174553750463,
+      "learning_rate": 5.303820378499303e-08,
+      "loss": 0.4375,
+      "step": 18595
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 26.67642790736344,
+      "learning_rate": 5.286488771710785e-08,
+      "loss": 0.4524,
+      "step": 18596
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 13.756035368829533,
+      "learning_rate": 5.2691854539926865e-08,
+      "loss": 0.4082,
+      "step": 18597
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 21.197258258517074,
+      "learning_rate": 5.251910425837059e-08,
+      "loss": 0.4361,
+      "step": 18598
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.33317766935585,
+      "learning_rate": 5.234663687735064e-08,
+      "loss": 0.4043,
+      "step": 18599
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.62280763299338,
+      "learning_rate": 5.217445240177421e-08,
+      "loss": 0.5052,
+      "step": 18600
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.138903199289413,
+      "learning_rate": 5.200255083653738e-08,
+      "loss": 0.4137,
+      "step": 18601
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 22.868304257937098,
+      "learning_rate": 5.183093218652846e-08,
+      "loss": 0.3824,
+      "step": 18602
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 24.447223790989273,
+      "learning_rate": 5.1659596456627995e-08,
+      "loss": 0.4383,
+      "step": 18603
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 21.43291422741993,
+      "learning_rate": 5.1488543651708746e-08,
+      "loss": 0.4885,
+      "step": 18604
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 21.090242826865616,
+      "learning_rate": 5.1317773776635715e-08,
+      "loss": 0.4,
+      "step": 18605
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 22.938900327943976,
+      "learning_rate": 5.114728683626613e-08,
+      "loss": 0.4606,
+      "step": 18606
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 23.022326562030177,
+      "learning_rate": 5.0977082835447224e-08,
+      "loss": 0.446,
+      "step": 18607
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 15.911164662251947,
+      "learning_rate": 5.0807161779019565e-08,
+      "loss": 0.3709,
+      "step": 18608
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 20.155628051329376,
+      "learning_rate": 5.0637523671817065e-08,
+      "loss": 0.4125,
+      "step": 18609
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 16.328969247337927,
+      "learning_rate": 5.046816851866254e-08,
+      "loss": 0.3728,
+      "step": 18610
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 28.27759517110512,
+      "learning_rate": 5.029909632437324e-08,
+      "loss": 0.4255,
+      "step": 18611
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 23.982083425487772,
+      "learning_rate": 5.013030709375644e-08,
+      "loss": 0.4397,
+      "step": 18612
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 20.643691713699532,
+      "learning_rate": 4.996180083161384e-08,
+      "loss": 0.469,
+      "step": 18613
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 20.25380596551742,
+      "learning_rate": 4.979357754273717e-08,
+      "loss": 0.3876,
+      "step": 18614
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 24.61539945340483,
+      "learning_rate": 4.9625637231910385e-08,
+      "loss": 0.4598,
+      "step": 18615
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 28.110118856654594,
+      "learning_rate": 4.9457979903909656e-08,
+      "loss": 0.4055,
+      "step": 18616
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 24.847831118561604,
+      "learning_rate": 4.92906055635034e-08,
+      "loss": 0.4431,
+      "step": 18617
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.283652785266383,
+      "learning_rate": 4.912351421545114e-08,
+      "loss": 0.46,
+      "step": 18618
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 24.557359963271853,
+      "learning_rate": 4.8956705864504625e-08,
+      "loss": 0.4171,
+      "step": 18619
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 19.77325504479544,
+      "learning_rate": 4.879018051540785e-08,
+      "loss": 0.4968,
+      "step": 18620
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.937921242507,
+      "learning_rate": 4.8623938172898125e-08,
+      "loss": 0.4832,
+      "step": 18621
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.63163128508053,
+      "learning_rate": 4.8457978841702777e-08,
+      "loss": 0.4852,
+      "step": 18622
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 25.216370902881867,
+      "learning_rate": 4.8292302526539156e-08,
+      "loss": 0.4614,
+      "step": 18623
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 16.058903222800982,
+      "learning_rate": 4.8126909232121265e-08,
+      "loss": 0.4089,
+      "step": 18624
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 28.55656914493002,
+      "learning_rate": 4.7961798963153116e-08,
+      "loss": 0.3658,
+      "step": 18625
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 19.335173828765612,
+      "learning_rate": 4.7796971724329844e-08,
+      "loss": 0.4176,
+      "step": 18626
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 29.92343023157761,
+      "learning_rate": 4.763242752033881e-08,
+      "loss": 0.4596,
+      "step": 18627
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.754862113961618,
+      "learning_rate": 4.746816635585849e-08,
+      "loss": 0.4642,
+      "step": 18628
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 21.106521012563128,
+      "learning_rate": 4.730418823556182e-08,
+      "loss": 0.4528,
+      "step": 18629
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 22.291964721395384,
+      "learning_rate": 4.7140493164112844e-08,
+      "loss": 0.3749,
+      "step": 18630
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 23.228612524459315,
+      "learning_rate": 4.6977081146165614e-08,
+      "loss": 0.3591,
+      "step": 18631
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.39592857343386,
+      "learning_rate": 4.6813952186366415e-08,
+      "loss": 0.4128,
+      "step": 18632
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.597083295235763,
+      "learning_rate": 4.6651106289357095e-08,
+      "loss": 0.4548,
+      "step": 18633
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.806309333640822,
+      "learning_rate": 4.648854345976839e-08,
+      "loss": 0.3841,
+      "step": 18634
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 15.220720784791276,
+      "learning_rate": 4.632626370222215e-08,
+      "loss": 0.4147,
+      "step": 18635
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 25.83988797709514,
+      "learning_rate": 4.6164267021334693e-08,
+      "loss": 0.4305,
+      "step": 18636
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 22.425065938667647,
+      "learning_rate": 4.6002553421711226e-08,
+      "loss": 0.3922,
+      "step": 18637
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 30.214374762004176,
+      "learning_rate": 4.5841122907953614e-08,
+      "loss": 0.4502,
+      "step": 18638
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 22.165399191736235,
+      "learning_rate": 4.567997548465153e-08,
+      "loss": 0.4719,
+      "step": 18639
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 19.626846342036412,
+      "learning_rate": 4.551911115638685e-08,
+      "loss": 0.3702,
+      "step": 18640
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.0712803203373,
+      "learning_rate": 4.5358529927735925e-08,
+      "loss": 0.4592,
+      "step": 18641
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 29.73346874401523,
+      "learning_rate": 4.5198231803265103e-08,
+      "loss": 0.3759,
+      "step": 18642
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.92594201247352,
+      "learning_rate": 4.503821678753406e-08,
+      "loss": 0.4157,
+      "step": 18643
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.94960621579408,
+      "learning_rate": 4.487848488509139e-08,
+      "loss": 0.3873,
+      "step": 18644
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 27.034417033027776,
+      "learning_rate": 4.4719036100481225e-08,
+      "loss": 0.4707,
+      "step": 18645
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 20.6283469836735,
+      "learning_rate": 4.455987043823884e-08,
+      "loss": 0.4497,
+      "step": 18646
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.516282058220437,
+      "learning_rate": 4.440098790288949e-08,
+      "loss": 0.4272,
+      "step": 18647
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 42.846011660680006,
+      "learning_rate": 4.4242388498951797e-08,
+      "loss": 0.4764,
+      "step": 18648
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.041086199032765,
+      "learning_rate": 4.408407223093658e-08,
+      "loss": 0.4858,
+      "step": 18649
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 27.69422095747818,
+      "learning_rate": 4.3926039103346915e-08,
+      "loss": 0.4583,
+      "step": 18650
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.84931033681021,
+      "learning_rate": 4.376828912067699e-08,
+      "loss": 0.4087,
+      "step": 18651
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.845552992305535,
+      "learning_rate": 4.361082228741209e-08,
+      "loss": 0.4178,
+      "step": 18652
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 25.577283857073823,
+      "learning_rate": 4.3453638608030865e-08,
+      "loss": 0.481,
+      "step": 18653
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 25.728272857018755,
+      "learning_rate": 4.329673808700308e-08,
+      "loss": 0.5032,
+      "step": 18654
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 34.84744185745317,
+      "learning_rate": 4.3140120728792925e-08,
+      "loss": 0.4678,
+      "step": 18655
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 25.495810458498816,
+      "learning_rate": 4.29837865378524e-08,
+      "loss": 0.4702,
+      "step": 18656
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.704265429051265,
+      "learning_rate": 4.2827735518629065e-08,
+      "loss": 0.3311,
+      "step": 18657
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.398259129198157,
+      "learning_rate": 4.2671967675559365e-08,
+      "loss": 0.3712,
+      "step": 18658
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.14479882031989,
+      "learning_rate": 4.2516483013074205e-08,
+      "loss": 0.4725,
+      "step": 18659
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.658998022762546,
+      "learning_rate": 4.23612815355956e-08,
+      "loss": 0.4169,
+      "step": 18660
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 39.56902023118311,
+      "learning_rate": 4.2206363247536684e-08,
+      "loss": 0.3783,
+      "step": 18661
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 15.908342868291848,
+      "learning_rate": 4.205172815330394e-08,
+      "loss": 0.3516,
+      "step": 18662
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.428531260360174,
+      "learning_rate": 4.189737625729384e-08,
+      "loss": 0.411,
+      "step": 18663
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 43.42183158051474,
+      "learning_rate": 4.174330756389844e-08,
+      "loss": 0.5289,
+      "step": 18664
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.131858629523386,
+      "learning_rate": 4.1589522077497556e-08,
+      "loss": 0.3714,
+      "step": 18665
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 32.079054858466264,
+      "learning_rate": 4.143601980246437e-08,
+      "loss": 0.4846,
+      "step": 18666
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 25.2285427377757,
+      "learning_rate": 4.128280074316649e-08,
+      "loss": 0.4167,
+      "step": 18667
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 16.831271742485498,
+      "learning_rate": 4.112986490395931e-08,
+      "loss": 0.3816,
+      "step": 18668
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 26.932611297880097,
+      "learning_rate": 4.0977212289192717e-08,
+      "loss": 0.4682,
+      "step": 18669
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.87775375318171,
+      "learning_rate": 4.082484290320876e-08,
+      "loss": 0.4161,
+      "step": 18670
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 32.37367054334205,
+      "learning_rate": 4.067275675034066e-08,
+      "loss": 0.5205,
+      "step": 18671
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 22.177344558226117,
+      "learning_rate": 4.0520953834912724e-08,
+      "loss": 0.435,
+      "step": 18672
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.2196649354697,
+      "learning_rate": 4.0369434161242616e-08,
+      "loss": 0.4382,
+      "step": 18673
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.14482200298816,
+      "learning_rate": 4.021819773364022e-08,
+      "loss": 0.4702,
+      "step": 18674
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 23.06524709813954,
+      "learning_rate": 4.0067244556405424e-08,
+      "loss": 0.4338,
+      "step": 18675
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 16.27565151927008,
+      "learning_rate": 3.9916574633832586e-08,
+      "loss": 0.4441,
+      "step": 18676
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 26.532480593809655,
+      "learning_rate": 3.976618797020493e-08,
+      "loss": 0.4556,
+      "step": 18677
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.46870527326587,
+      "learning_rate": 3.961608456980126e-08,
+      "loss": 0.4861,
+      "step": 18678
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 19.92178228169725,
+      "learning_rate": 3.946626443688817e-08,
+      "loss": 0.4586,
+      "step": 18679
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 32.09177854234582,
+      "learning_rate": 3.9316727575728906e-08,
+      "loss": 0.4041,
+      "step": 18680
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 31.647061054597792,
+      "learning_rate": 3.9167473990575635e-08,
+      "loss": 0.5062,
+      "step": 18681
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.565054413074023,
+      "learning_rate": 3.901850368567161e-08,
+      "loss": 0.4193,
+      "step": 18682
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.209451963656534,
+      "learning_rate": 3.886981666525347e-08,
+      "loss": 0.381,
+      "step": 18683
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 23.65321754964633,
+      "learning_rate": 3.8721412933552246e-08,
+      "loss": 0.4466,
+      "step": 18684
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 22.789543695374963,
+      "learning_rate": 3.85732924947857e-08,
+      "loss": 0.4341,
+      "step": 18685
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.63322895789619,
+      "learning_rate": 3.8425455353168215e-08,
+      "loss": 0.4401,
+      "step": 18686
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 24.262060936557013,
+      "learning_rate": 3.827790151290312e-08,
+      "loss": 0.4497,
+      "step": 18687
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.735849732075785,
+      "learning_rate": 3.813063097818703e-08,
+      "loss": 0.4099,
+      "step": 18688
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 21.165894683579662,
+      "learning_rate": 3.798364375320773e-08,
+      "loss": 0.4682,
+      "step": 18689
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 21.28106580562114,
+      "learning_rate": 3.783693984214743e-08,
+      "loss": 0.386,
+      "step": 18690
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.474656576320363,
+      "learning_rate": 3.7690519249174996e-08,
+      "loss": 0.4668,
+      "step": 18691
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 28.814855488587384,
+      "learning_rate": 3.754438197845822e-08,
+      "loss": 0.4856,
+      "step": 18692
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 30.833775246627976,
+      "learning_rate": 3.739852803415045e-08,
+      "loss": 0.456,
+      "step": 18693
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.52357529838724,
+      "learning_rate": 3.725295742040058e-08,
+      "loss": 0.4246,
+      "step": 18694
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 22.3044660233226,
+      "learning_rate": 3.710767014134864e-08,
+      "loss": 0.4162,
+      "step": 18695
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.671897384199188,
+      "learning_rate": 3.696266620112576e-08,
+      "loss": 0.3841,
+      "step": 18696
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 21.70616058229042,
+      "learning_rate": 3.681794560385754e-08,
+      "loss": 0.4908,
+      "step": 18697
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.202102105211527,
+      "learning_rate": 3.667350835365846e-08,
+      "loss": 0.4185,
+      "step": 18698
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 23.94815043423386,
+      "learning_rate": 3.6529354454635236e-08,
+      "loss": 0.4304,
+      "step": 18699
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 31.318605188511896,
+      "learning_rate": 3.638548391089014e-08,
+      "loss": 0.3747,
+      "step": 18700
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 16.999345002458742,
+      "learning_rate": 3.6241896726513236e-08,
+      "loss": 0.4546,
+      "step": 18701
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 28.485159415881544,
+      "learning_rate": 3.6098592905587925e-08,
+      "loss": 0.4317,
+      "step": 18702
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.923521386622834,
+      "learning_rate": 3.595557245218983e-08,
+      "loss": 0.3923,
+      "step": 18703
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 25.324056303276155,
+      "learning_rate": 3.581283537038571e-08,
+      "loss": 0.4537,
+      "step": 18704
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.395555892423726,
+      "learning_rate": 3.567038166423675e-08,
+      "loss": 0.3949,
+      "step": 18705
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 29.604980440563356,
+      "learning_rate": 3.552821133779305e-08,
+      "loss": 0.5119,
+      "step": 18706
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.632383607149897,
+      "learning_rate": 3.5386324395096924e-08,
+      "loss": 0.5184,
+      "step": 18707
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 22.740222758894927,
+      "learning_rate": 3.5244720840186266e-08,
+      "loss": 0.4184,
+      "step": 18708
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 17.288677795653232,
+      "learning_rate": 3.510340067708562e-08,
+      "loss": 0.3519,
+      "step": 18709
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 24.570040017513822,
+      "learning_rate": 3.4962363909815114e-08,
+      "loss": 0.4102,
+      "step": 18710
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 25.039475040340207,
+      "learning_rate": 3.482161054238486e-08,
+      "loss": 0.4078,
+      "step": 18711
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 35.83570232526256,
+      "learning_rate": 3.4681140578799453e-08,
+      "loss": 0.4508,
+      "step": 18712
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 14.077287785597578,
+      "learning_rate": 3.4540954023052356e-08,
+      "loss": 0.373,
+      "step": 18713
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 28.38416494438566,
+      "learning_rate": 3.440105087913148e-08,
+      "loss": 0.5014,
+      "step": 18714
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 26.44152072786472,
+      "learning_rate": 3.426143115101477e-08,
+      "loss": 0.4134,
+      "step": 18715
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 23.044596320506358,
+      "learning_rate": 3.412209484267237e-08,
+      "loss": 0.3973,
+      "step": 18716
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.570489149932435,
+      "learning_rate": 3.39830419580689e-08,
+      "loss": 0.4287,
+      "step": 18717
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 17.726730926742437,
+      "learning_rate": 3.384427250115674e-08,
+      "loss": 0.3965,
+      "step": 18718
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.45951085234899,
+      "learning_rate": 3.370578647588496e-08,
+      "loss": 0.4517,
+      "step": 18719
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 27.034624249653447,
+      "learning_rate": 3.3567583886189304e-08,
+      "loss": 0.4584,
+      "step": 18720
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 33.022644246421905,
+      "learning_rate": 3.3429664736001064e-08,
+      "loss": 0.3956,
+      "step": 18721
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.400692286618987,
+      "learning_rate": 3.3292029029243777e-08,
+      "loss": 0.4209,
+      "step": 18722
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 24.62312803559912,
+      "learning_rate": 3.315467676982986e-08,
+      "loss": 0.419,
+      "step": 18723
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 19.84010695785895,
+      "learning_rate": 3.301760796166731e-08,
+      "loss": 0.4454,
+      "step": 18724
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 20.53350933336121,
+      "learning_rate": 3.2880822608653e-08,
+      "loss": 0.3966,
+      "step": 18725
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 21.667395581232263,
+      "learning_rate": 3.274432071467826e-08,
+      "loss": 0.3987,
+      "step": 18726
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 18.842186840818783,
+      "learning_rate": 3.260810228362332e-08,
+      "loss": 0.4979,
+      "step": 18727
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 22.34221306429955,
+      "learning_rate": 3.247216731936398e-08,
+      "loss": 0.4167,
+      "step": 18728
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.175157848862305,
+      "learning_rate": 3.233651582576491e-08,
+      "loss": 0.4058,
+      "step": 18729
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 21.698653936717903,
+      "learning_rate": 3.220114780668415e-08,
+      "loss": 0.4214,
+      "step": 18730
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 25.789606687807048,
+      "learning_rate": 3.206606326597306e-08,
+      "loss": 0.4594,
+      "step": 18731
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.24767071004808,
+      "learning_rate": 3.193126220747078e-08,
+      "loss": 0.4763,
+      "step": 18732
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 16.19364381521254,
+      "learning_rate": 3.1796744635013144e-08,
+      "loss": 0.4126,
+      "step": 18733
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 27.44508202963948,
+      "learning_rate": 3.166251055242375e-08,
+      "loss": 0.4675,
+      "step": 18734
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 23.68752461812091,
+      "learning_rate": 3.1528559963522886e-08,
+      "loss": 0.4252,
+      "step": 18735
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.882091399878316,
+      "learning_rate": 3.139489287211639e-08,
+      "loss": 0.3822,
+      "step": 18736
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 18.122581247374395,
+      "learning_rate": 3.1261509282009e-08,
+      "loss": 0.474,
+      "step": 18737
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.869086343954912,
+      "learning_rate": 3.112840919699212e-08,
+      "loss": 0.4094,
+      "step": 18738
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.61216789833138,
+      "learning_rate": 3.099559262085272e-08,
+      "loss": 0.3968,
+      "step": 18739
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 29.14144736758251,
+      "learning_rate": 3.086305955736557e-08,
+      "loss": 0.4646,
+      "step": 18740
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.577402657819984,
+      "learning_rate": 3.073081001030209e-08,
+      "loss": 0.4229,
+      "step": 18741
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 31.779288952076058,
+      "learning_rate": 3.05988439834215e-08,
+      "loss": 0.445,
+      "step": 18742
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 24.00281304771441,
+      "learning_rate": 3.046716148047968e-08,
+      "loss": 0.3836,
+      "step": 18743
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 28.742786190747925,
+      "learning_rate": 3.033576250521919e-08,
+      "loss": 0.4648,
+      "step": 18744
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 22.357514310985454,
+      "learning_rate": 3.0204647061375936e-08,
+      "loss": 0.3959,
+      "step": 18745
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.758723977431917,
+      "learning_rate": 3.0073815152681374e-08,
+      "loss": 0.3909,
+      "step": 18746
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 29.566795283058966,
+      "learning_rate": 2.994326678285586e-08,
+      "loss": 0.4181,
+      "step": 18747
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 18.89971685474977,
+      "learning_rate": 2.981300195561088e-08,
+      "loss": 0.4496,
+      "step": 18748
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.142760209003928,
+      "learning_rate": 2.968302067465234e-08,
+      "loss": 0.3766,
+      "step": 18749
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.50725785704199,
+      "learning_rate": 2.955332294367508e-08,
+      "loss": 0.3809,
+      "step": 18750
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.53937405425937,
+      "learning_rate": 2.942390876636947e-08,
+      "loss": 0.4405,
+      "step": 18751
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 30.427025853152482,
+      "learning_rate": 2.9294778146415902e-08,
+      "loss": 0.4646,
+      "step": 18752
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 24.40538684873113,
+      "learning_rate": 2.916593108748589e-08,
+      "loss": 0.4403,
+      "step": 18753
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 36.357294153424114,
+      "learning_rate": 2.903736759324316e-08,
+      "loss": 0.4464,
+      "step": 18754
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 37.04455475316805,
+      "learning_rate": 2.8909087667345905e-08,
+      "loss": 0.5209,
+      "step": 18755
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.993564961493274,
+      "learning_rate": 2.878109131344009e-08,
+      "loss": 0.4659,
+      "step": 18756
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.53477503590143,
+      "learning_rate": 2.8653378535168363e-08,
+      "loss": 0.3906,
+      "step": 18757
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 25.053004733778717,
+      "learning_rate": 2.852594933616004e-08,
+      "loss": 0.3998,
+      "step": 18758
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 25.938913936739894,
+      "learning_rate": 2.8398803720042223e-08,
+      "loss": 0.4132,
+      "step": 18759
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 24.73032173157626,
+      "learning_rate": 2.8271941690427574e-08,
+      "loss": 0.4392,
+      "step": 18760
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 15.357475297163742,
+      "learning_rate": 2.8145363250926537e-08,
+      "loss": 0.466,
+      "step": 18761
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 26.91957492822058,
+      "learning_rate": 2.801906840513735e-08,
+      "loss": 0.455,
+      "step": 18762
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 16.342303117196654,
+      "learning_rate": 2.7893057156653802e-08,
+      "loss": 0.3851,
+      "step": 18763
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.9469692244357,
+      "learning_rate": 2.776732950905636e-08,
+      "loss": 0.4209,
+      "step": 18764
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.63479290076147,
+      "learning_rate": 2.764188546592439e-08,
+      "loss": 0.4102,
+      "step": 18765
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.009678048630057,
+      "learning_rate": 2.7516725030821702e-08,
+      "loss": 0.4475,
+      "step": 18766
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 30.39524087934214,
+      "learning_rate": 2.7391848207311e-08,
+      "loss": 0.4804,
+      "step": 18767
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 16.279164359612306,
+      "learning_rate": 2.7267254998941673e-08,
+      "loss": 0.4328,
+      "step": 18768
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.07723564946217,
+      "learning_rate": 2.714294540925866e-08,
+      "loss": 0.4298,
+      "step": 18769
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 21.173648797225805,
+      "learning_rate": 2.701891944179469e-08,
+      "loss": 0.4188,
+      "step": 18770
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 33.013802930793744,
+      "learning_rate": 2.6895177100079163e-08,
+      "loss": 0.4474,
+      "step": 18771
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 21.179250839302412,
+      "learning_rate": 2.6771718387631485e-08,
+      "loss": 0.4239,
+      "step": 18772
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 36.06192174477639,
+      "learning_rate": 2.6648543307962183e-08,
+      "loss": 0.413,
+      "step": 18773
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 22.85214303021008,
+      "learning_rate": 2.65256518645729e-08,
+      "loss": 0.4541,
+      "step": 18774
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 31.70182055726893,
+      "learning_rate": 2.6403044060959726e-08,
+      "loss": 0.5998,
+      "step": 18775
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 24.550304011863258,
+      "learning_rate": 2.6280719900609874e-08,
+      "loss": 0.4539,
+      "step": 18776
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 21.817987495704166,
+      "learning_rate": 2.6158679387002782e-08,
+      "loss": 0.4139,
+      "step": 18777
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 33.64175094761346,
+      "learning_rate": 2.60369225236079e-08,
+      "loss": 0.4318,
+      "step": 18778
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 26.811499794842092,
+      "learning_rate": 2.591544931388801e-08,
+      "loss": 0.4256,
+      "step": 18779
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 26.57258492936507,
+      "learning_rate": 2.5794259761298124e-08,
+      "loss": 0.4902,
+      "step": 18780
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 32.15790848305961,
+      "learning_rate": 2.5673353869284378e-08,
+      "loss": 0.4376,
+      "step": 18781
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 18.634583213528288,
+      "learning_rate": 2.5552731641286243e-08,
+      "loss": 0.3896,
+      "step": 18782
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 27.702546031343232,
+      "learning_rate": 2.5432393080733197e-08,
+      "loss": 0.4426,
+      "step": 18783
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.141080917750035,
+      "learning_rate": 2.5312338191048057e-08,
+      "loss": 0.3945,
+      "step": 18784
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.313118559148904,
+      "learning_rate": 2.5192566975644762e-08,
+      "loss": 0.4106,
+      "step": 18785
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 26.031776390844172,
+      "learning_rate": 2.507307943793058e-08,
+      "loss": 0.443,
+      "step": 18786
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 24.179225148724576,
+      "learning_rate": 2.495387558130169e-08,
+      "loss": 0.4772,
+      "step": 18787
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.72467751902362,
+      "learning_rate": 2.4834955409149818e-08,
+      "loss": 0.3954,
+      "step": 18788
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.96829298177997,
+      "learning_rate": 2.471631892485671e-08,
+      "loss": 0.405,
+      "step": 18789
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 19.188812197011092,
+      "learning_rate": 2.4597966131796324e-08,
+      "loss": 0.4266,
+      "step": 18790
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.624193024996938,
+      "learning_rate": 2.4479897033333756e-08,
+      "loss": 0.3841,
+      "step": 18791
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 19.53607711040371,
+      "learning_rate": 2.4362111632827423e-08,
+      "loss": 0.374,
+      "step": 18792
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 20.580840083730862,
+      "learning_rate": 2.4244609933627982e-08,
+      "loss": 0.4027,
+      "step": 18793
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 30.237717700965412,
+      "learning_rate": 2.4127391939076094e-08,
+      "loss": 0.4185,
+      "step": 18794
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.415351902388192,
+      "learning_rate": 2.401045765250465e-08,
+      "loss": 0.3951,
+      "step": 18795
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 21.193501235139802,
+      "learning_rate": 2.3893807077239872e-08,
+      "loss": 0.4041,
+      "step": 18796
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 21.20760443617767,
+      "learning_rate": 2.3777440216600224e-08,
+      "loss": 0.4118,
+      "step": 18797
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 26.4211039557542,
+      "learning_rate": 2.3661357073894166e-08,
+      "loss": 0.5376,
+      "step": 18798
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 29.610038295955878,
+      "learning_rate": 2.3545557652422392e-08,
+      "loss": 0.4614,
+      "step": 18799
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 18.333718131824373,
+      "learning_rate": 2.343004195547893e-08,
+      "loss": 0.4835,
+      "step": 18800
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.04216062168671,
+      "learning_rate": 2.3314809986348942e-08,
+      "loss": 0.4097,
+      "step": 18801
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 27.712977368429215,
+      "learning_rate": 2.3199861748309793e-08,
+      "loss": 0.3999,
+      "step": 18802
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.79517534651258,
+      "learning_rate": 2.3085197244631097e-08,
+      "loss": 0.4006,
+      "step": 18803
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 30.395588764128835,
+      "learning_rate": 2.2970816478572466e-08,
+      "loss": 0.4285,
+      "step": 18804
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.41010286146284,
+      "learning_rate": 2.2856719453386856e-08,
+      "loss": 0.3604,
+      "step": 18805
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 19.367444586422824,
+      "learning_rate": 2.2742906172320555e-08,
+      "loss": 0.4169,
+      "step": 18806
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 16.25920666257959,
+      "learning_rate": 2.262937663860876e-08,
+      "loss": 0.4082,
+      "step": 18807
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 35.683292052732334,
+      "learning_rate": 2.25161308554811e-08,
+      "loss": 0.4507,
+      "step": 18808
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 33.708381624898436,
+      "learning_rate": 2.240316882615834e-08,
+      "loss": 0.5187,
+      "step": 18809
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.625063948136773,
+      "learning_rate": 2.2290490553852352e-08,
+      "loss": 0.449,
+      "step": 18810
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.675553624388154,
+      "learning_rate": 2.217809604176835e-08,
+      "loss": 0.3801,
+      "step": 18811
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 26.2654836391024,
+      "learning_rate": 2.2065985293102664e-08,
+      "loss": 0.423,
+      "step": 18812
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 34.24203353227539,
+      "learning_rate": 2.195415831104275e-08,
+      "loss": 0.4576,
+      "step": 18813
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 37.33191503746383,
+      "learning_rate": 2.1842615098769394e-08,
+      "loss": 0.4552,
+      "step": 18814
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 21.282757662276865,
+      "learning_rate": 2.1731355659456722e-08,
+      "loss": 0.4664,
+      "step": 18815
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 21.31259725987568,
+      "learning_rate": 2.162037999626554e-08,
+      "loss": 0.3783,
+      "step": 18816
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 38.773255630057406,
+      "learning_rate": 2.1509688112354432e-08,
+      "loss": 0.4386,
+      "step": 18817
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 20.559423254153288,
+      "learning_rate": 2.139928001086977e-08,
+      "loss": 0.4554,
+      "step": 18818
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 29.073919519978006,
+      "learning_rate": 2.128915569495238e-08,
+      "loss": 0.4391,
+      "step": 18819
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 18.948614969190782,
+      "learning_rate": 2.1179315167734194e-08,
+      "loss": 0.3962,
+      "step": 18820
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 20.99502684539474,
+      "learning_rate": 2.1069758432339382e-08,
+      "loss": 0.3903,
+      "step": 18821
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 15.779475235051981,
+      "learning_rate": 2.096048549188212e-08,
+      "loss": 0.3892,
+      "step": 18822
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 17.94539963331978,
+      "learning_rate": 2.0851496349472144e-08,
+      "loss": 0.4245,
+      "step": 18823
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 20.91316604901738,
+      "learning_rate": 2.0742791008206974e-08,
+      "loss": 0.4323,
+      "step": 18824
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 28.108933797854498,
+      "learning_rate": 2.0634369471179692e-08,
+      "loss": 0.5228,
+      "step": 18825
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.348109574930174,
+      "learning_rate": 2.052623174147339e-08,
+      "loss": 0.4347,
+      "step": 18826
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.69923514183554,
+      "learning_rate": 2.0418377822162272e-08,
+      "loss": 0.3567,
+      "step": 18827
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 16.375016200625435,
+      "learning_rate": 2.0310807716316107e-08,
+      "loss": 0.407,
+      "step": 18828
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.110251296777143,
+      "learning_rate": 2.0203521426991333e-08,
+      "loss": 0.442,
+      "step": 18829
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 23.680601394345754,
+      "learning_rate": 2.0096518957241072e-08,
+      "loss": 0.3985,
+      "step": 18830
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 19.201817616219056,
+      "learning_rate": 1.9989800310107333e-08,
+      "loss": 0.449,
+      "step": 18831
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 18.49303878118887,
+      "learning_rate": 1.9883365488625462e-08,
+      "loss": 0.418,
+      "step": 18832
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 34.593306186244746,
+      "learning_rate": 1.977721449582304e-08,
+      "loss": 0.4808,
+      "step": 18833
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.258314419282225,
+      "learning_rate": 1.9671347334717648e-08,
+      "loss": 0.4271,
+      "step": 18834
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 19.64951328495019,
+      "learning_rate": 1.956576400832133e-08,
+      "loss": 0.4314,
+      "step": 18835
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 25.061544618923524,
+      "learning_rate": 1.9460464519636124e-08,
+      "loss": 0.4569,
+      "step": 18836
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 26.005278786779574,
+      "learning_rate": 1.9355448871657413e-08,
+      "loss": 0.4879,
+      "step": 18837
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 23.59853772602622,
+      "learning_rate": 1.9250717067370583e-08,
+      "loss": 0.4838,
+      "step": 18838
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.536518005080442,
+      "learning_rate": 1.9146269109755477e-08,
+      "loss": 0.5077,
+      "step": 18839
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.854227551529316,
+      "learning_rate": 1.904210500178083e-08,
+      "loss": 0.4185,
+      "step": 18840
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.634960533997567,
+      "learning_rate": 1.8938224746410938e-08,
+      "loss": 0.384,
+      "step": 18841
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 25.961780777768983,
+      "learning_rate": 1.8834628346598993e-08,
+      "loss": 0.4577,
+      "step": 18842
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 18.768583418684553,
+      "learning_rate": 1.8731315805290418e-08,
+      "loss": 0.4575,
+      "step": 18843
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 33.07991543925951,
+      "learning_rate": 1.862828712542508e-08,
+      "loss": 0.4274,
+      "step": 18844
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 16.91244361565658,
+      "learning_rate": 1.8525542309932865e-08,
+      "loss": 0.4454,
+      "step": 18845
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 21.393036626288726,
+      "learning_rate": 1.8423081361734762e-08,
+      "loss": 0.4224,
+      "step": 18846
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 16.805750169499632,
+      "learning_rate": 1.832090428374511e-08,
+      "loss": 0.4225,
+      "step": 18847
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 17.846242157756087,
+      "learning_rate": 1.8219011078869365e-08,
+      "loss": 0.4427,
+      "step": 18848
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 22.013429965478007,
+      "learning_rate": 1.8117401750006314e-08,
+      "loss": 0.4592,
+      "step": 18849
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 31.881296388191885,
+      "learning_rate": 1.801607630004476e-08,
+      "loss": 0.4904,
+      "step": 18850
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 25.435922524561278,
+      "learning_rate": 1.7915034731867952e-08,
+      "loss": 0.4484,
+      "step": 18851
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 17.4465189340236,
+      "learning_rate": 1.781427704834693e-08,
+      "loss": 0.418,
+      "step": 18852
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 19.433702504222726,
+      "learning_rate": 1.7713803252348283e-08,
+      "loss": 0.4352,
+      "step": 18853
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 27.7159977608798,
+      "learning_rate": 1.761361334673084e-08,
+      "loss": 0.4113,
+      "step": 18854
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 14.719987480874398,
+      "learning_rate": 1.7513707334341212e-08,
+      "loss": 0.4255,
+      "step": 18855
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 24.99010997292553,
+      "learning_rate": 1.741408521802379e-08,
+      "loss": 0.3937,
+      "step": 18856
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 38.3508837187754,
+      "learning_rate": 1.7314747000608532e-08,
+      "loss": 0.4187,
+      "step": 18857
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 25.896340583041276,
+      "learning_rate": 1.721569268492318e-08,
+      "loss": 0.4719,
+      "step": 18858
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 24.16564306250834,
+      "learning_rate": 1.7116922273783255e-08,
+      "loss": 0.4503,
+      "step": 18859
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 27.82957492523101,
+      "learning_rate": 1.7018435769998732e-08,
+      "loss": 0.4227,
+      "step": 18860
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 18.948515972378544,
+      "learning_rate": 1.6920233176369594e-08,
+      "loss": 0.4106,
+      "step": 18861
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 23.739760804248764,
+      "learning_rate": 1.6822314495689164e-08,
+      "loss": 0.4203,
+      "step": 18862
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 16.041260531989856,
+      "learning_rate": 1.6724679730742986e-08,
+      "loss": 0.3665,
+      "step": 18863
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 24.729139696897622,
+      "learning_rate": 1.6627328884305516e-08,
+      "loss": 0.4103,
+      "step": 18864
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 27.878113369639426,
+      "learning_rate": 1.6530261959147863e-08,
+      "loss": 0.4326,
+      "step": 18865
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 25.231729775072946,
+      "learning_rate": 1.6433478958028938e-08,
+      "loss": 0.4551,
+      "step": 18866
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 20.282995790276836,
+      "learning_rate": 1.6336979883700976e-08,
+      "loss": 0.4121,
+      "step": 18867
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 32.43296778811208,
+      "learning_rate": 1.6240764738909566e-08,
+      "loss": 0.478,
+      "step": 18868
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 22.033986177707625,
+      "learning_rate": 1.6144833526390292e-08,
+      "loss": 0.4184,
+      "step": 18869
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.658520664140898,
+      "learning_rate": 1.6049186248872084e-08,
+      "loss": 0.4372,
+      "step": 18870
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 22.776506783762112,
+      "learning_rate": 1.595382290907388e-08,
+      "loss": 0.4233,
+      "step": 18871
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 16.954463981453696,
+      "learning_rate": 1.585874350970906e-08,
+      "loss": 0.3926,
+      "step": 18872
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.216906088205832,
+      "learning_rate": 1.5763948053481026e-08,
+      "loss": 0.3413,
+      "step": 18873
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 19.106083866466513,
+      "learning_rate": 1.566943654308539e-08,
+      "loss": 0.3729,
+      "step": 18874
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.707289195317248,
+      "learning_rate": 1.557520898121001e-08,
+      "loss": 0.3961,
+      "step": 18875
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.36446340655278,
+      "learning_rate": 1.5481265370536068e-08,
+      "loss": 0.4516,
+      "step": 18876
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 31.912753232155914,
+      "learning_rate": 1.5387605713732545e-08,
+      "loss": 0.4214,
+      "step": 18877
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 17.271925813908354,
+      "learning_rate": 1.5294230013466194e-08,
+      "loss": 0.3983,
+      "step": 18878
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 26.304181007673105,
+      "learning_rate": 1.520113827239045e-08,
+      "loss": 0.4787,
+      "step": 18879
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 17.73109102470334,
+      "learning_rate": 1.510833049315319e-08,
+      "loss": 0.468,
+      "step": 18880
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 28.12424159722876,
+      "learning_rate": 1.501580667839453e-08,
+      "loss": 0.4406,
+      "step": 18881
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 35.95885184847225,
+      "learning_rate": 1.4923566830744586e-08,
+      "loss": 0.3809,
+      "step": 18882
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 32.0822114380872,
+      "learning_rate": 1.4831610952827925e-08,
+      "loss": 0.4653,
+      "step": 18883
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 14.792258107059261,
+      "learning_rate": 1.4739939047259122e-08,
+      "loss": 0.4393,
+      "step": 18884
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 19.699509350934477,
+      "learning_rate": 1.4648551116644982e-08,
+      "loss": 0.408,
+      "step": 18885
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 17.44364783821956,
+      "learning_rate": 1.4557447163584538e-08,
+      "loss": 0.4689,
+      "step": 18886
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 26.215502723706983,
+      "learning_rate": 1.4466627190669047e-08,
+      "loss": 0.4321,
+      "step": 18887
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.272045555064263,
+      "learning_rate": 1.4376091200482002e-08,
+      "loss": 0.4471,
+      "step": 18888
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 34.28184718943699,
+      "learning_rate": 1.4285839195596896e-08,
+      "loss": 0.4598,
+      "step": 18889
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 27.576272715320606,
+      "learning_rate": 1.4195871178580567e-08,
+      "loss": 0.4785,
+      "step": 18890
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 19.07893117830633,
+      "learning_rate": 1.410618715199319e-08,
+      "loss": 0.4771,
+      "step": 18891
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 31.61028932803755,
+      "learning_rate": 1.4016787118383835e-08,
+      "loss": 0.4269,
+      "step": 18892
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 14.811103497941325,
+      "learning_rate": 1.3927671080294913e-08,
+      "loss": 0.4185,
+      "step": 18893
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 28.70594859420759,
+      "learning_rate": 1.3838839040262175e-08,
+      "loss": 0.4399,
+      "step": 18894
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 15.81090350158245,
+      "learning_rate": 1.3750291000811377e-08,
+      "loss": 0.4383,
+      "step": 18895
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 17.86401355509483,
+      "learning_rate": 1.3662026964459396e-08,
+      "loss": 0.3881,
+      "step": 18896
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 17.640748512639835,
+      "learning_rate": 1.3574046933717556e-08,
+      "loss": 0.3694,
+      "step": 18897
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.11328455583545,
+      "learning_rate": 1.3486350911089407e-08,
+      "loss": 0.3511,
+      "step": 18898
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 18.487984189649875,
+      "learning_rate": 1.3398938899066294e-08,
+      "loss": 0.401,
+      "step": 18899
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 17.72068262240934,
+      "learning_rate": 1.3311810900135113e-08,
+      "loss": 0.4295,
+      "step": 18900
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 18.793717919091723,
+      "learning_rate": 1.3224966916774995e-08,
+      "loss": 0.4255,
+      "step": 18901
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 25.787911773959113,
+      "learning_rate": 1.3138406951453963e-08,
+      "loss": 0.4231,
+      "step": 18902
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 23.04411393259211,
+      "learning_rate": 1.3052131006634494e-08,
+      "loss": 0.4177,
+      "step": 18903
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 20.70476230121021,
+      "learning_rate": 1.296613908477018e-08,
+      "loss": 0.4449,
+      "step": 18904
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 18.640186251891148,
+      "learning_rate": 1.2880431188306841e-08,
+      "loss": 0.429,
+      "step": 18905
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 47.296971730264694,
+      "learning_rate": 1.2795007319681418e-08,
+      "loss": 0.4389,
+      "step": 18906
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 20.84662707402527,
+      "learning_rate": 1.270986748132419e-08,
+      "loss": 0.4326,
+      "step": 18907
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 32.03667124438352,
+      "learning_rate": 1.262501167565655e-08,
+      "loss": 0.4631,
+      "step": 18908
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 24.41483813343104,
+      "learning_rate": 1.2540439905089907e-08,
+      "loss": 0.4659,
+      "step": 18909
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 26.103207194699998,
+      "learning_rate": 1.2456152172031222e-08,
+      "loss": 0.4136,
+      "step": 18910
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 22.227457055994396,
+      "learning_rate": 1.2372148478876356e-08,
+      "loss": 0.3986,
+      "step": 18911
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 27.45741520807344,
+      "learning_rate": 1.228842882801562e-08,
+      "loss": 0.419,
+      "step": 18912
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 30.814716971225845,
+      "learning_rate": 1.2204993221829331e-08,
+      "loss": 0.412,
+      "step": 18913
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 18.874521113630074,
+      "learning_rate": 1.2121841662690037e-08,
+      "loss": 0.4038,
+      "step": 18914
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 21.780510409590452,
+      "learning_rate": 1.2038974152963623e-08,
+      "loss": 0.4103,
+      "step": 18915
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 37.31944332104858,
+      "learning_rate": 1.1956390695004871e-08,
+      "loss": 0.4795,
+      "step": 18916
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 15.891607859925616,
+      "learning_rate": 1.1874091291164125e-08,
+      "loss": 0.4385,
+      "step": 18917
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 26.307133652937484,
+      "learning_rate": 1.1792075943781733e-08,
+      "loss": 0.431,
+      "step": 18918
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.552838197163407,
+      "learning_rate": 1.1710344655189166e-08,
+      "loss": 0.422,
+      "step": 18919
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 17.897930163882112,
+      "learning_rate": 1.1628897427711227e-08,
+      "loss": 0.4156,
+      "step": 18920
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 20.86070074197505,
+      "learning_rate": 1.1547734263664957e-08,
+      "loss": 0.3833,
+      "step": 18921
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 19.380461471426464,
+      "learning_rate": 1.1466855165357393e-08,
+      "loss": 0.4023,
+      "step": 18922
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 17.83093619687225,
+      "learning_rate": 1.1386260135090033e-08,
+      "loss": 0.4547,
+      "step": 18923
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 16.225807193943098,
+      "learning_rate": 1.1305949175154374e-08,
+      "loss": 0.4514,
+      "step": 18924
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 25.397365373777898,
+      "learning_rate": 1.1225922287834146e-08,
+      "loss": 0.4352,
+      "step": 18925
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 28.502076288768954,
+      "learning_rate": 1.1146179475404195e-08,
+      "loss": 0.5079,
+      "step": 18926
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.354008033831885,
+      "learning_rate": 1.1066720740134928e-08,
+      "loss": 0.4392,
+      "step": 18927
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 26.87022337714897,
+      "learning_rate": 1.098754608428454e-08,
+      "loss": 0.4118,
+      "step": 18928
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 23.40075480046642,
+      "learning_rate": 1.090865551010456e-08,
+      "loss": 0.4529,
+      "step": 18929
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 17.644945994855508,
+      "learning_rate": 1.0830049019839862e-08,
+      "loss": 0.3868,
+      "step": 18930
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 28.6227328337586,
+      "learning_rate": 1.0751726615724212e-08,
+      "loss": 0.4741,
+      "step": 18931
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 17.12361973400038,
+      "learning_rate": 1.0673688299985829e-08,
+      "loss": 0.4264,
+      "step": 18932
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 31.25673761405072,
+      "learning_rate": 1.059593407484516e-08,
+      "loss": 0.4484,
+      "step": 18933
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 17.018648184647915,
+      "learning_rate": 1.0518463942511547e-08,
+      "loss": 0.4339,
+      "step": 18934
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 20.877510915580245,
+      "learning_rate": 1.0441277905188785e-08,
+      "loss": 0.4812,
+      "step": 18935
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.089566961833093,
+      "learning_rate": 1.0364375965074003e-08,
+      "loss": 0.4252,
+      "step": 18936
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.504862675545432,
+      "learning_rate": 1.0287758124351011e-08,
+      "loss": 0.3679,
+      "step": 18937
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 32.761000417391585,
+      "learning_rate": 1.0211424385201396e-08,
+      "loss": 0.4554,
+      "step": 18938
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 24.312658174583717,
+      "learning_rate": 1.0135374749794536e-08,
+      "loss": 0.4521,
+      "step": 18939
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 15.280678559951753,
+      "learning_rate": 1.0059609220293142e-08,
+      "loss": 0.4338,
+      "step": 18940
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 17.874495016480942,
+      "learning_rate": 9.984127798853271e-09,
+      "loss": 0.3965,
+      "step": 18941
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 24.805881613276036,
+      "learning_rate": 9.908930487620983e-09,
+      "loss": 0.3996,
+      "step": 18942
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 31.750902270559063,
+      "learning_rate": 9.834017288734565e-09,
+      "loss": 0.4024,
+      "step": 18943
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.997023547056152,
+      "learning_rate": 9.759388204323428e-09,
+      "loss": 0.4377,
+      "step": 18944
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 18.637025379256198,
+      "learning_rate": 9.685043236512537e-09,
+      "loss": 0.3674,
+      "step": 18945
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 25.995770388912867,
+      "learning_rate": 9.610982387414647e-09,
+      "loss": 0.4357,
+      "step": 18946
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 32.8932047350986,
+      "learning_rate": 9.53720565913585e-09,
+      "loss": 0.4433,
+      "step": 18947
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 24.27065160741977,
+      "learning_rate": 9.46371305377447e-09,
+      "loss": 0.4437,
+      "step": 18948
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.935168933779824,
+      "learning_rate": 9.390504573422166e-09,
+      "loss": 0.4225,
+      "step": 18949
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 16.20277267742511,
+      "learning_rate": 9.317580220158385e-09,
+      "loss": 0.3808,
+      "step": 18950
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 24.955198566677257,
+      "learning_rate": 9.244939996058134e-09,
+      "loss": 0.4451,
+      "step": 18951
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 19.38059984314208,
+      "learning_rate": 9.172583903187537e-09,
+      "loss": 0.4397,
+      "step": 18952
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 14.94150959851793,
+      "learning_rate": 9.10051194360495e-09,
+      "loss": 0.3739,
+      "step": 18953
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 23.468157869988666,
+      "learning_rate": 9.028724119358734e-09,
+      "loss": 0.4121,
+      "step": 18954
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 27.10669632905505,
+      "learning_rate": 8.957220432490587e-09,
+      "loss": 0.4185,
+      "step": 18955
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 24.47403983111849,
+      "learning_rate": 8.886000885034441e-09,
+      "loss": 0.4557,
+      "step": 18956
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 16.38599176442435,
+      "learning_rate": 8.815065479016449e-09,
+      "loss": 0.4632,
+      "step": 18957
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.903400396116684,
+      "learning_rate": 8.74441421645278e-09,
+      "loss": 0.4089,
+      "step": 18958
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 24.430226858935495,
+      "learning_rate": 8.674047099354044e-09,
+      "loss": 0.4088,
+      "step": 18959
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.816443375817986,
+      "learning_rate": 8.603964129719755e-09,
+      "loss": 0.4056,
+      "step": 18960
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 29.52230697530616,
+      "learning_rate": 8.534165309544985e-09,
+      "loss": 0.422,
+      "step": 18961
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 18.756022535125073,
+      "learning_rate": 8.464650640812588e-09,
+      "loss": 0.4325,
+      "step": 18962
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.615046005459465,
+      "learning_rate": 8.395420125500986e-09,
+      "loss": 0.4399,
+      "step": 18963
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 37.68698940714089,
+      "learning_rate": 8.326473765579713e-09,
+      "loss": 0.4363,
+      "step": 18964
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 19.227217201578128,
+      "learning_rate": 8.2578115630072e-09,
+      "loss": 0.4266,
+      "step": 18965
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 29.493925372324,
+      "learning_rate": 8.189433519738555e-09,
+      "loss": 0.4452,
+      "step": 18966
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 19.750361215933907,
+      "learning_rate": 8.121339637716663e-09,
+      "loss": 0.4659,
+      "step": 18967
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.051687452921847,
+      "learning_rate": 8.053529918878867e-09,
+      "loss": 0.4494,
+      "step": 18968
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 27.385109292214995,
+      "learning_rate": 7.986004365153621e-09,
+      "loss": 0.5378,
+      "step": 18969
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.2300064321709,
+      "learning_rate": 7.918762978461613e-09,
+      "loss": 0.4553,
+      "step": 18970
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 20.503338150930492,
+      "learning_rate": 7.851805760714647e-09,
+      "loss": 0.4762,
+      "step": 18971
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 20.896910256170234,
+      "learning_rate": 7.785132713816756e-09,
+      "loss": 0.389,
+      "step": 18972
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 20.76167807373196,
+      "learning_rate": 7.7187438396642e-09,
+      "loss": 0.4352,
+      "step": 18973
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 16.797145053999174,
+      "learning_rate": 7.652639140146579e-09,
+      "loss": 0.4545,
+      "step": 18974
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 14.602992906713823,
+      "learning_rate": 7.586818617141279e-09,
+      "loss": 0.462,
+      "step": 18975
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 21.221055704374898,
+      "learning_rate": 7.521282272521246e-09,
+      "loss": 0.4316,
+      "step": 18976
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 26.49754136753965,
+      "learning_rate": 7.456030108151657e-09,
+      "loss": 0.4585,
+      "step": 18977
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 31.44579497447003,
+      "learning_rate": 7.391062125886583e-09,
+      "loss": 0.4044,
+      "step": 18978
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 27.599035151131698,
+      "learning_rate": 7.326378327574546e-09,
+      "loss": 0.3935,
+      "step": 18979
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 18.12478616851425,
+      "learning_rate": 7.261978715054074e-09,
+      "loss": 0.4299,
+      "step": 18980
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 23.206302658995273,
+      "learning_rate": 7.197863290157037e-09,
+      "loss": 0.4947,
+      "step": 18981
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 31.141215870998966,
+      "learning_rate": 7.134032054707529e-09,
+      "loss": 0.4481,
+      "step": 18982
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 31.301658485732723,
+      "learning_rate": 7.070485010520767e-09,
+      "loss": 0.4303,
+      "step": 18983
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 19.564166917408766,
+      "learning_rate": 7.007222159404192e-09,
+      "loss": 0.4189,
+      "step": 18984
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 16.430214103897324,
+      "learning_rate": 6.944243503155257e-09,
+      "loss": 0.4579,
+      "step": 18985
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 22.445940848722614,
+      "learning_rate": 6.881549043568081e-09,
+      "loss": 0.4568,
+      "step": 18986
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 22.023330939915716,
+      "learning_rate": 6.819138782422352e-09,
+      "loss": 0.4271,
+      "step": 18987
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.884309167028974,
+      "learning_rate": 6.757012721494427e-09,
+      "loss": 0.4144,
+      "step": 18988
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 24.605609217544227,
+      "learning_rate": 6.6951708625517805e-09,
+      "loss": 0.4263,
+      "step": 18989
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 45.597811336967645,
+      "learning_rate": 6.633613207351897e-09,
+      "loss": 0.4614,
+      "step": 18990
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 18.01267228656309,
+      "learning_rate": 6.5723397576467066e-09,
+      "loss": 0.4038,
+      "step": 18991
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 27.522545004457985,
+      "learning_rate": 6.51135051517815e-09,
+      "loss": 0.4144,
+      "step": 18992
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 22.75316349110637,
+      "learning_rate": 6.450645481681506e-09,
+      "loss": 0.3934,
+      "step": 18993
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 15.503631236625589,
+      "learning_rate": 6.39022465888095e-09,
+      "loss": 0.3806,
+      "step": 18994
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.616927052392633,
+      "learning_rate": 6.330088048497329e-09,
+      "loss": 0.4745,
+      "step": 18995
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 24.368009699544025,
+      "learning_rate": 6.270235652239276e-09,
+      "loss": 0.4141,
+      "step": 18996
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 17.230089111227763,
+      "learning_rate": 6.2106674718098725e-09,
+      "loss": 0.3998,
+      "step": 18997
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.98243307727471,
+      "learning_rate": 6.1513835089033194e-09,
+      "loss": 0.4365,
+      "step": 18998
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 18.82280379797896,
+      "learning_rate": 6.0923837652038245e-09,
+      "loss": 0.3938,
+      "step": 18999
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 25.494579999640557,
+      "learning_rate": 6.033668242392266e-09,
+      "loss": 0.4302,
+      "step": 19000
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 22.20595285419225,
+      "learning_rate": 5.975236942136197e-09,
+      "loss": 0.3808,
+      "step": 19001
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 30.59397621041931,
+      "learning_rate": 5.917089866097625e-09,
+      "loss": 0.4022,
+      "step": 19002
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 31.912210014301774,
+      "learning_rate": 5.85922701593078e-09,
+      "loss": 0.4656,
+      "step": 19003
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 27.910059455826534,
+      "learning_rate": 5.801648393282122e-09,
+      "loss": 0.4577,
+      "step": 19004
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 18.43865478584657,
+      "learning_rate": 5.744353999787011e-09,
+      "loss": 0.4491,
+      "step": 19005
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.1279621980003,
+      "learning_rate": 5.6873438370763645e-09,
+      "loss": 0.4439,
+      "step": 19006
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 15.487279225054342,
+      "learning_rate": 5.630617906771108e-09,
+      "loss": 0.4437,
+      "step": 19007
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 36.774827812953504,
+      "learning_rate": 5.574176210484394e-09,
+      "loss": 0.4529,
+      "step": 19008
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 35.57147601809623,
+      "learning_rate": 5.518018749821607e-09,
+      "loss": 0.4119,
+      "step": 19009
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 19.362953589589303,
+      "learning_rate": 5.462145526380358e-09,
+      "loss": 0.3986,
+      "step": 19010
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 16.30089443632787,
+      "learning_rate": 5.406556541748265e-09,
+      "loss": 0.3757,
+      "step": 19011
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 26.287451505684306,
+      "learning_rate": 5.351251797507395e-09,
+      "loss": 0.5059,
+      "step": 19012
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 27.050379262202735,
+      "learning_rate": 5.296231295229826e-09,
+      "loss": 0.4001,
+      "step": 19013
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 24.09386004708536,
+      "learning_rate": 5.2414950364809706e-09,
+      "loss": 0.4343,
+      "step": 19014
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 19.7126359657613,
+      "learning_rate": 5.187043022817362e-09,
+      "loss": 0.398,
+      "step": 19015
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 22.009488539272215,
+      "learning_rate": 5.1328752557877615e-09,
+      "loss": 0.4432,
+      "step": 19016
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 18.44981720515287,
+      "learning_rate": 5.078991736932049e-09,
+      "loss": 0.4513,
+      "step": 19017
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.877136236803327,
+      "learning_rate": 5.025392467783441e-09,
+      "loss": 0.427,
+      "step": 19018
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 27.957027941121694,
+      "learning_rate": 4.972077449865165e-09,
+      "loss": 0.43,
+      "step": 19019
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 17.118143936799132,
+      "learning_rate": 4.919046684693785e-09,
+      "loss": 0.4203,
+      "step": 19020
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 21.99521631893356,
+      "learning_rate": 4.8663001737780934e-09,
+      "loss": 0.4202,
+      "step": 19021
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 17.424257384974354,
+      "learning_rate": 4.813837918618003e-09,
+      "loss": 0.4745,
+      "step": 19022
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.060107964969642,
+      "learning_rate": 4.761659920705652e-09,
+      "loss": 0.4413,
+      "step": 19023
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 19.837777820509434,
+      "learning_rate": 4.7097661815243e-09,
+      "loss": 0.4484,
+      "step": 19024
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 13.61512186914786,
+      "learning_rate": 4.6581567025494324e-09,
+      "loss": 0.3757,
+      "step": 19025
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 13.75563206851141,
+      "learning_rate": 4.6068314852498745e-09,
+      "loss": 0.386,
+      "step": 19026
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 21.740098357364158,
+      "learning_rate": 4.555790531085569e-09,
+      "loss": 0.4401,
+      "step": 19027
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 33.30524951882078,
+      "learning_rate": 4.505033841506468e-09,
+      "loss": 0.4379,
+      "step": 19028
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 28.644074719748602,
+      "learning_rate": 4.454561417958081e-09,
+      "loss": 0.4253,
+      "step": 19029
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 26.08593718125171,
+      "learning_rate": 4.404373261873707e-09,
+      "loss": 0.4307,
+      "step": 19030
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 29.210141061411946,
+      "learning_rate": 4.354469374682202e-09,
+      "loss": 0.4129,
+      "step": 19031
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 27.10043342670859,
+      "learning_rate": 4.304849757802432e-09,
+      "loss": 0.4332,
+      "step": 19032
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.35246038649922,
+      "learning_rate": 4.25551441264549e-09,
+      "loss": 0.4806,
+      "step": 19033
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 23.600085213828606,
+      "learning_rate": 4.206463340614697e-09,
+      "loss": 0.4317,
+      "step": 19034
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 23.15291259566251,
+      "learning_rate": 4.157696543103384e-09,
+      "loss": 0.3772,
+      "step": 19035
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 27.431742348623246,
+      "learning_rate": 4.109214021500441e-09,
+      "loss": 0.5211,
+      "step": 19036
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 28.368095301477446,
+      "learning_rate": 4.061015777184762e-09,
+      "loss": 0.4635,
+      "step": 19037
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 23.357818429175374,
+      "learning_rate": 4.013101811525255e-09,
+      "loss": 0.4779,
+      "step": 19038
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 22.078226363265927,
+      "learning_rate": 3.9654721258863826e-09,
+      "loss": 0.4067,
+      "step": 19039
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 24.644678762069717,
+      "learning_rate": 3.918126721621507e-09,
+      "loss": 0.4163,
+      "step": 19040
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 25.96143597960283,
+      "learning_rate": 3.871065600078439e-09,
+      "loss": 0.4695,
+      "step": 19041
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 14.725658046460605,
+      "learning_rate": 3.824288762593886e-09,
+      "loss": 0.3881,
+      "step": 19042
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 15.693133629704562,
+      "learning_rate": 3.777796210499008e-09,
+      "loss": 0.421,
+      "step": 19043
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 16.176605285607106,
+      "learning_rate": 3.731587945116078e-09,
+      "loss": 0.4373,
+      "step": 19044
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 26.68550790715355,
+      "learning_rate": 3.6856639677584906e-09,
+      "loss": 0.4516,
+      "step": 19045
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 22.38465464336222,
+      "learning_rate": 3.640024279734089e-09,
+      "loss": 0.4024,
+      "step": 19046
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 20.297508200502715,
+      "learning_rate": 3.5946688823396137e-09,
+      "loss": 0.4067,
+      "step": 19047
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.32188164123176,
+      "learning_rate": 3.5495977768640332e-09,
+      "loss": 0.5031,
+      "step": 19048
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 21.24744857870061,
+      "learning_rate": 3.5048109645907657e-09,
+      "loss": 0.4013,
+      "step": 19049
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 27.54048088587474,
+      "learning_rate": 3.460308446793237e-09,
+      "loss": 0.4876,
+      "step": 19050
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 18.753353146495563,
+      "learning_rate": 3.4160902247359904e-09,
+      "loss": 0.4494,
+      "step": 19051
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 30.05542718618167,
+      "learning_rate": 3.37215629967802e-09,
+      "loss": 0.4343,
+      "step": 19052
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 23.280379620866015,
+      "learning_rate": 3.3285066728672156e-09,
+      "loss": 0.473,
+      "step": 19053
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 15.254602517955334,
+      "learning_rate": 3.285141345547027e-09,
+      "loss": 0.3855,
+      "step": 19054
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 17.312123829063434,
+      "learning_rate": 3.242060318948692e-09,
+      "loss": 0.4466,
+      "step": 19055
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 22.590060792896477,
+      "learning_rate": 3.1992635942978965e-09,
+      "loss": 0.3997,
+      "step": 19056
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 30.826135758257536,
+      "learning_rate": 3.1567511728125555e-09,
+      "loss": 0.3865,
+      "step": 19057
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 17.03789455743406,
+      "learning_rate": 3.114523055700591e-09,
+      "loss": 0.3902,
+      "step": 19058
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 25.533856886818658,
+      "learning_rate": 3.0725792441643755e-09,
+      "loss": 0.3939,
+      "step": 19059
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 24.253246856882708,
+      "learning_rate": 3.0309197393962874e-09,
+      "loss": 0.3718,
+      "step": 19060
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 25.898124468538473,
+      "learning_rate": 2.9895445425798253e-09,
+      "loss": 0.4364,
+      "step": 19061
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 20.483069362247637,
+      "learning_rate": 2.9484536548929355e-09,
+      "loss": 0.4733,
+      "step": 19062
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 24.538813588830145,
+      "learning_rate": 2.907647077504683e-09,
+      "loss": 0.4989,
+      "step": 19063
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 16.220303074303274,
+      "learning_rate": 2.86712481157414e-09,
+      "loss": 0.4245,
+      "step": 19064
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 18.952393335275733,
+      "learning_rate": 2.826886858253719e-09,
+      "loss": 0.411,
+      "step": 19065
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 25.645906260133913,
+      "learning_rate": 2.7869332186891697e-09,
+      "loss": 0.4338,
+      "step": 19066
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 21.984469154633462,
+      "learning_rate": 2.7472638940162498e-09,
+      "loss": 0.4415,
+      "step": 19067
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 27.13863583892154,
+      "learning_rate": 2.707878885362947e-09,
+      "loss": 0.455,
+      "step": 19068
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 21.20115647423043,
+      "learning_rate": 2.6687781938483647e-09,
+      "loss": 0.389,
+      "step": 19069
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 37.3323743692061,
+      "learning_rate": 2.629961820587168e-09,
+      "loss": 0.4341,
+      "step": 19070
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 17.45510550013736,
+      "learning_rate": 2.5914297666806975e-09,
+      "loss": 0.46,
+      "step": 19071
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 16.590035204122856,
+      "learning_rate": 2.5531820332247436e-09,
+      "loss": 0.4174,
+      "step": 19072
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.41161283175662,
+      "learning_rate": 2.515218621309545e-09,
+      "loss": 0.4657,
+      "step": 19073
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 18.175893233524274,
+      "learning_rate": 2.4775395320120187e-09,
+      "loss": 0.4031,
+      "step": 19074
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 15.164030899003132,
+      "learning_rate": 2.4401447664046398e-09,
+      "loss": 0.3686,
+      "step": 19075
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 19.19627788989915,
+      "learning_rate": 2.4030343255521115e-09,
+      "loss": 0.3659,
+      "step": 19076
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 17.318520219648978,
+      "learning_rate": 2.3662082105080365e-09,
+      "loss": 0.436,
+      "step": 19077
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 19.873689941949166,
+      "learning_rate": 2.3296664223215747e-09,
+      "loss": 0.3835,
+      "step": 19078
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 25.057739024403343,
+      "learning_rate": 2.2934089620307852e-09,
+      "loss": 0.4465,
+      "step": 19079
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.07033242626989,
+      "learning_rate": 2.257435830665955e-09,
+      "loss": 0.426,
+      "step": 19080
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 18.889132040771294,
+      "learning_rate": 2.2217470292518196e-09,
+      "loss": 0.4156,
+      "step": 19081
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 33.06543095474694,
+      "learning_rate": 2.1863425588020125e-09,
+      "loss": 0.4709,
+      "step": 19082
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 22.210637933547137,
+      "learning_rate": 2.1512224203246167e-09,
+      "loss": 0.4326,
+      "step": 19083
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 19.713890987628858,
+      "learning_rate": 2.116386614818833e-09,
+      "loss": 0.3843,
+      "step": 19084
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 18.52326693027949,
+      "learning_rate": 2.081835143272759e-09,
+      "loss": 0.5101,
+      "step": 19085
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 32.725982118124115,
+      "learning_rate": 2.0475680066711632e-09,
+      "loss": 0.4518,
+      "step": 19086
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 21.844984637720195,
+      "learning_rate": 2.0135852059888215e-09,
+      "loss": 0.4331,
+      "step": 19087
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 23.3869577545039,
+      "learning_rate": 1.9798867421905176e-09,
+      "loss": 0.4327,
+      "step": 19088
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 22.035357402173492,
+      "learning_rate": 1.9464726162365944e-09,
+      "loss": 0.4743,
+      "step": 19089
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 31.04858755985672,
+      "learning_rate": 1.9133428290751822e-09,
+      "loss": 0.4268,
+      "step": 19090
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 29.413243408754294,
+      "learning_rate": 1.880497381651081e-09,
+      "loss": 0.4537,
+      "step": 19091
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 21.444012303361635,
+      "learning_rate": 1.8479362748957674e-09,
+      "loss": 0.4457,
+      "step": 19092
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 24.13550402604632,
+      "learning_rate": 1.8156595097362783e-09,
+      "loss": 0.43,
+      "step": 19093
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 14.194137590752707,
+      "learning_rate": 1.783667087090768e-09,
+      "loss": 0.3687,
+      "step": 19094
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 13.481841177483908,
+      "learning_rate": 1.7519590078696192e-09,
+      "loss": 0.4039,
+      "step": 19095
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 25.700649642595774,
+      "learning_rate": 1.7205352729732227e-09,
+      "loss": 0.4307,
+      "step": 19096
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 24.809557723470284,
+      "learning_rate": 1.6893958832964186e-09,
+      "loss": 0.4146,
+      "step": 19097
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.156701973914807,
+      "learning_rate": 1.6585408397240544e-09,
+      "loss": 0.4533,
+      "step": 19098
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 19.22753737224552,
+      "learning_rate": 1.6279701431343165e-09,
+      "loss": 0.4442,
+      "step": 19099
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.227705283745607,
+      "learning_rate": 1.5976837943953993e-09,
+      "loss": 0.4315,
+      "step": 19100
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 22.538953271135078,
+      "learning_rate": 1.5676817943699462e-09,
+      "loss": 0.3854,
+      "step": 19101
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 22.77952149718791,
+      "learning_rate": 1.5379641439106086e-09,
+      "loss": 0.4179,
+      "step": 19102
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 17.82095976092635,
+      "learning_rate": 1.5085308438633762e-09,
+      "loss": 0.4657,
+      "step": 19103
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.558446492963586,
+      "learning_rate": 1.4793818950642469e-09,
+      "loss": 0.4205,
+      "step": 19104
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 29.937010316586935,
+      "learning_rate": 1.4505172983425576e-09,
+      "loss": 0.4217,
+      "step": 19105
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.99159863846815,
+      "learning_rate": 1.4219370545187628e-09,
+      "loss": 0.4263,
+      "step": 19106
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 26.506005887529,
+      "learning_rate": 1.393641164405546e-09,
+      "loss": 0.4655,
+      "step": 19107
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 25.773095289087827,
+      "learning_rate": 1.3656296288089288e-09,
+      "loss": 0.4774,
+      "step": 19108
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 16.211623797309347,
+      "learning_rate": 1.3379024485249414e-09,
+      "loss": 0.4244,
+      "step": 19109
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 21.072265779157753,
+      "learning_rate": 1.310459624341842e-09,
+      "loss": 0.3614,
+      "step": 19110
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.334449939676414,
+      "learning_rate": 1.2833011570401177e-09,
+      "loss": 0.4664,
+      "step": 19111
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.51298214015917,
+      "learning_rate": 1.2564270473924834e-09,
+      "loss": 0.4528,
+      "step": 19112
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.299673938899996,
+      "learning_rate": 1.2298372961627726e-09,
+      "loss": 0.393,
+      "step": 19113
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 19.474701590143233,
+      "learning_rate": 1.203531904107047e-09,
+      "loss": 0.4197,
+      "step": 19114
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 14.48891243957197,
+      "learning_rate": 1.1775108719735973e-09,
+      "loss": 0.4491,
+      "step": 19115
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.8203287279255,
+      "learning_rate": 1.1517742005029419e-09,
+      "loss": 0.4563,
+      "step": 19116
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 24.33407077518308,
+      "learning_rate": 1.1263218904267181e-09,
+      "loss": 0.4634,
+      "step": 19117
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.749761730805847,
+      "learning_rate": 1.101153942468791e-09,
+      "loss": 0.4136,
+      "step": 19118
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 19.71688178895801,
+      "learning_rate": 1.0762703573452548e-09,
+      "loss": 0.4013,
+      "step": 19119
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.78527307129287,
+      "learning_rate": 1.051671135762211e-09,
+      "loss": 0.4017,
+      "step": 19120
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.272434161483535,
+      "learning_rate": 1.0273562784213209e-09,
+      "loss": 0.45,
+      "step": 19121
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 31.26487638863152,
+      "learning_rate": 1.0033257860131428e-09,
+      "loss": 0.4873,
+      "step": 19122
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 28.505163860036593,
+      "learning_rate": 9.795796592204643e-10,
+      "loss": 0.4282,
+      "step": 19123
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.571739471281944,
+      "learning_rate": 9.561178987205211e-10,
+      "loss": 0.4026,
+      "step": 19124
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 15.528194327419424,
+      "learning_rate": 9.32940505178337e-10,
+      "loss": 0.4177,
+      "step": 19125
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 31.210650128317532,
+      "learning_rate": 9.100474792544944e-10,
+      "loss": 0.4726,
+      "step": 19126
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 33.317345870618894,
+      "learning_rate": 8.874388215995844e-10,
+      "loss": 0.4645,
+      "step": 19127
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 20.593244654948567,
+      "learning_rate": 8.651145328564259e-10,
+      "loss": 0.4757,
+      "step": 19128
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 21.82547565416537,
+      "learning_rate": 8.430746136600665e-10,
+      "loss": 0.4405,
+      "step": 19129
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.637793451431314,
+      "learning_rate": 8.21319064636672e-10,
+      "loss": 0.4178,
+      "step": 19130
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 28.063395681827846,
+      "learning_rate": 7.998478864068571e-10,
+      "loss": 0.4538,
+      "step": 19131
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 22.113518286915223,
+      "learning_rate": 7.786610795790239e-10,
+      "loss": 0.3897,
+      "step": 19132
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.41134479960007,
+      "learning_rate": 7.577586447571339e-10,
+      "loss": 0.4371,
+      "step": 19133
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 20.617627286972553,
+      "learning_rate": 7.371405825351563e-10,
+      "loss": 0.388,
+      "step": 19134
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 33.88056126214513,
+      "learning_rate": 7.168068934992889e-10,
+      "loss": 0.5159,
+      "step": 19135
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 22.62335761705785,
+      "learning_rate": 6.967575782279579e-10,
+      "loss": 0.3901,
+      "step": 19136
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.51772474035553,
+      "learning_rate": 6.769926372918178e-10,
+      "loss": 0.3448,
+      "step": 19137
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 26.582755048821255,
+      "learning_rate": 6.575120712526417e-10,
+      "loss": 0.4743,
+      "step": 19138
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 25.05416605983395,
+      "learning_rate": 6.383158806644307e-10,
+      "loss": 0.4437,
+      "step": 19139
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 27.444541828032182,
+      "learning_rate": 6.194040660723044e-10,
+      "loss": 0.3962,
+      "step": 19140
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 14.257702662423737,
+      "learning_rate": 6.007766280158312e-10,
+      "loss": 0.3867,
+      "step": 19141
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 27.69013418579101,
+      "learning_rate": 5.824335670234771e-10,
+      "loss": 0.4224,
+      "step": 19142
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 20.327416924438346,
+      "learning_rate": 5.643748836170471e-10,
+      "loss": 0.3976,
+      "step": 19143
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 21.44905646809337,
+      "learning_rate": 5.466005783105743e-10,
+      "loss": 0.4313,
+      "step": 19144
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.317397559300634,
+      "learning_rate": 5.291106516092103e-10,
+      "loss": 0.4331,
+      "step": 19145
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 26.386520590556934,
+      "learning_rate": 5.119051040092249e-10,
+      "loss": 0.3946,
+      "step": 19146
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.22320143247084,
+      "learning_rate": 4.949839360024467e-10,
+      "loss": 0.4299,
+      "step": 19147
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.268333595438364,
+      "learning_rate": 4.78347148068492e-10,
+      "loss": 0.4069,
+      "step": 19148
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 14.636166646912105,
+      "learning_rate": 4.6199474068031603e-10,
+      "loss": 0.4332,
+      "step": 19149
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 37.888977622175,
+      "learning_rate": 4.459267143042123e-10,
+      "loss": 0.538,
+      "step": 19150
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 18.815274491028674,
+      "learning_rate": 4.301430693964825e-10,
+      "loss": 0.4236,
+      "step": 19151
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 19.45068957085827,
+      "learning_rate": 4.1464380640565683e-10,
+      "loss": 0.3914,
+      "step": 19152
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 25.674100075925043,
+      "learning_rate": 3.994289257724937e-10,
+      "loss": 0.3994,
+      "step": 19153
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 19.993605566397438,
+      "learning_rate": 3.8449842792998016e-10,
+      "loss": 0.4182,
+      "step": 19154
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 16.46876706298002,
+      "learning_rate": 3.698523133022214e-10,
+      "loss": 0.4588,
+      "step": 19155
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.780245249741547,
+      "learning_rate": 3.5549058230777143e-10,
+      "loss": 0.3533,
+      "step": 19156
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 20.89951575376462,
+      "learning_rate": 3.4141323535186175e-10,
+      "loss": 0.39,
+      "step": 19157
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 15.106894907918312,
+      "learning_rate": 3.276202728375033e-10,
+      "loss": 0.4479,
+      "step": 19158
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.51762769795409,
+      "learning_rate": 3.1411169515549457e-10,
+      "loss": 0.3296,
+      "step": 19159
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 14.507648721817132,
+      "learning_rate": 3.0088750269108293e-10,
+      "loss": 0.413,
+      "step": 19160
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 22.44856247441304,
+      "learning_rate": 2.8794769581952376e-10,
+      "loss": 0.4144,
+      "step": 19161
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 26.166302820269554,
+      "learning_rate": 2.7529227490941115e-10,
+      "loss": 0.4075,
+      "step": 19162
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 20.290964703556106,
+      "learning_rate": 2.629212403193471e-10,
+      "loss": 0.4244,
+      "step": 19163
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 26.950469954927932,
+      "learning_rate": 2.508345924023825e-10,
+      "loss": 0.4177,
+      "step": 19164
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 26.53453770813269,
+      "learning_rate": 2.390323315026866e-10,
+      "loss": 0.4609,
+      "step": 19165
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 25.715568517200367,
+      "learning_rate": 2.2751445795443638e-10,
+      "loss": 0.493,
+      "step": 19166
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 25.548141875046,
+      "learning_rate": 2.162809720873682e-10,
+      "loss": 0.4288,
+      "step": 19167
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 15.212979452481388,
+      "learning_rate": 2.0533187421789557e-10,
+      "loss": 0.392,
+      "step": 19168
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 24.534495238661925,
+      "learning_rate": 1.9466716466021162e-10,
+      "loss": 0.3738,
+      "step": 19169
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 14.67929715322247,
+      "learning_rate": 1.84286843716297e-10,
+      "loss": 0.4557,
+      "step": 19170
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 20.541738157376844,
+      "learning_rate": 1.7419091168036085e-10,
+      "loss": 0.3638,
+      "step": 19171
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 18.0859665938405,
+      "learning_rate": 1.6437936884217132e-10,
+      "loss": 0.4544,
+      "step": 19172
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 23.817494495637863,
+      "learning_rate": 1.5485221547817396e-10,
+      "loss": 0.4007,
+      "step": 19173
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.901047592987165,
+      "learning_rate": 1.4560945186148366e-10,
+      "loss": 0.3542,
+      "step": 19174
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 18.6928571618243,
+      "learning_rate": 1.3665107825300284e-10,
+      "loss": 0.4492,
+      "step": 19175
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 29.834554139735523,
+      "learning_rate": 1.2797709490919296e-10,
+      "loss": 0.4039,
+      "step": 19176
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 22.011335605512496,
+      "learning_rate": 1.1958750207541337e-10,
+      "loss": 0.3808,
+      "step": 19177
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 22.482615369557795,
+      "learning_rate": 1.114822999914722e-10,
+      "loss": 0.4353,
+      "step": 19178
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 23.451599940045742,
+      "learning_rate": 1.0366148888607541e-10,
+      "loss": 0.4419,
+      "step": 19179
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 18.674575316721064,
+      "learning_rate": 9.612506898348805e-11,
+      "loss": 0.475,
+      "step": 19180
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 20.415924631877793,
+      "learning_rate": 8.887304049798317e-11,
+      "loss": 0.432,
+      "step": 19181
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 24.395616023162404,
+      "learning_rate": 8.190540363384181e-11,
+      "loss": 0.4367,
+      "step": 19182
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 20.142571665203263,
+      "learning_rate": 7.522215859090409e-11,
+      "loss": 0.4019,
+      "step": 19183
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 30.525650095070272,
+      "learning_rate": 6.882330556012839e-11,
+      "loss": 0.3953,
+      "step": 19184
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.500252992394028,
+      "learning_rate": 6.270884472026062e-11,
+      "loss": 0.4905,
+      "step": 19185
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.868225348266858,
+      "learning_rate": 5.687877624893645e-11,
+      "loss": 0.5009,
+      "step": 19186
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 16.7091206779695,
+      "learning_rate": 5.133310030935867e-11,
+      "loss": 0.4345,
+      "step": 19187
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 19.444532200969107,
+      "learning_rate": 4.607181705917896e-11,
+      "loss": 0.4049,
+      "step": 19188
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 31.746449589275194,
+      "learning_rate": 4.1094926650497855e-11,
+      "loss": 0.4715,
+      "step": 19189
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.348807097256906,
+      "learning_rate": 3.640242922209325e-11,
+      "loss": 0.4398,
+      "step": 19190
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 27.03494853284541,
+      "learning_rate": 3.199432490941234e-11,
+      "loss": 0.4329,
+      "step": 19191
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 24.461395597782875,
+      "learning_rate": 2.787061383680012e-11,
+      "loss": 0.4204,
+      "step": 19192
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 23.763644367276402,
+      "learning_rate": 2.4031296120829993e-11,
+      "loss": 0.4877,
+      "step": 19193
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 14.940773575296761,
+      "learning_rate": 2.047637187252427e-11,
+      "loss": 0.4944,
+      "step": 19194
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 30.485828040456667,
+      "learning_rate": 1.720584119180302e-11,
+      "loss": 0.4837,
+      "step": 19195
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.933422551153875,
+      "learning_rate": 1.4219704171924976e-11,
+      "loss": 0.4275,
+      "step": 19196
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 22.252548970525478,
+      "learning_rate": 1.1517960898377312e-11,
+      "loss": 0.4125,
+      "step": 19197
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 22.012802411460225,
+      "learning_rate": 9.100611446655194e-12,
+      "loss": 0.4283,
+      "step": 19198
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 19.45199348039055,
+      "learning_rate": 6.967655886702674e-12,
+      "loss": 0.3943,
+      "step": 19199
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 30.535730943838594,
+      "learning_rate": 5.119094278471792e-12,
+      "loss": 0.4638,
+      "step": 19200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.369485826707194,
+      "learning_rate": 3.5549266752532564e-12,
+      "loss": 0.4573,
+      "step": 19201
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 19.86926674322362,
+      "learning_rate": 2.275153120345763e-12,
+      "loss": 0.4187,
+      "step": 19202
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 18.529909896802643,
+      "learning_rate": 1.2797736514968962e-12,
+      "loss": 0.4601,
+      "step": 19203
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 20.335206130392084,
+      "learning_rate": 5.687882964622305e-13,
+      "loss": 0.4302,
+      "step": 19204
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 21.343195540094797,
+      "learning_rate": 1.4219707522578064e-13,
+      "loss": 0.3673,
+      "step": 19205
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 26.62800484030171,
+      "learning_rate": 0.0,
+      "loss": 0.4807,
+      "step": 19206
+    },
+    {
+      "epoch": 3.0,
+      "step": 19206,
+      "total_flos": 8.250659187302859e+18,
+      "train_loss": 0.6410556696475601,
+      "train_runtime": 20419.4789,
+      "train_samples_per_second": 120.393,
+      "train_steps_per_second": 0.941
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 19206,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50000,
+  "total_flos": 8.250659187302859e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}