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": 404.0940464199526,
+      "learning_rate": 3.466204506065858e-08,
+      "loss": 4.6712,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 468.96919471413077,
+      "learning_rate": 6.932409012131716e-08,
+      "loss": 4.8715,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 359.6808924037411,
+      "learning_rate": 1.0398613518197575e-07,
+      "loss": 5.1036,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 341.11279910232884,
+      "learning_rate": 1.386481802426343e-07,
+      "loss": 4.3693,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 403.0153830647367,
+      "learning_rate": 1.733102253032929e-07,
+      "loss": 4.707,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 383.17443114570926,
+      "learning_rate": 2.079722703639515e-07,
+      "loss": 4.765,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 379.5486393694711,
+      "learning_rate": 2.426343154246101e-07,
+      "loss": 4.7541,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 362.6404948982381,
+      "learning_rate": 2.772963604852686e-07,
+      "loss": 4.6183,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 370.2003726604941,
+      "learning_rate": 3.119584055459272e-07,
+      "loss": 4.7244,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 425.32293595986056,
+      "learning_rate": 3.466204506065858e-07,
+      "loss": 4.7455,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 365.35194798774165,
+      "learning_rate": 3.8128249566724436e-07,
+      "loss": 4.9468,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 385.3259000974223,
+      "learning_rate": 4.15944540727903e-07,
+      "loss": 4.5773,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 389.17691849486675,
+      "learning_rate": 4.5060658578856156e-07,
+      "loss": 4.9241,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 292.23759879240254,
+      "learning_rate": 4.852686308492202e-07,
+      "loss": 4.1968,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 408.6507313801951,
+      "learning_rate": 5.199306759098788e-07,
+      "loss": 4.6545,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 341.0328372571276,
+      "learning_rate": 5.545927209705372e-07,
+      "loss": 4.3244,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 316.3921247811569,
+      "learning_rate": 5.89254766031196e-07,
+      "loss": 4.2739,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 368.98894316692383,
+      "learning_rate": 6.239168110918544e-07,
+      "loss": 4.619,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 391.3551961931799,
+      "learning_rate": 6.58578856152513e-07,
+      "loss": 4.2722,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 324.16069828086273,
+      "learning_rate": 6.932409012131716e-07,
+      "loss": 4.0374,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 346.63615937123836,
+      "learning_rate": 7.279029462738301e-07,
+      "loss": 4.4495,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 349.70747820499923,
+      "learning_rate": 7.625649913344887e-07,
+      "loss": 4.2368,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 324.4425462783377,
+      "learning_rate": 7.972270363951473e-07,
+      "loss": 3.8778,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 355.4208938424768,
+      "learning_rate": 8.31889081455806e-07,
+      "loss": 3.689,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 287.2878685965508,
+      "learning_rate": 8.665511265164646e-07,
+      "loss": 3.3905,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 296.2215910991722,
+      "learning_rate": 9.012131715771231e-07,
+      "loss": 3.3905,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 297.13335540796504,
+      "learning_rate": 9.358752166377817e-07,
+      "loss": 3.2381,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 260.27900160433126,
+      "learning_rate": 9.705372616984403e-07,
+      "loss": 3.067,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 252.7902433842504,
+      "learning_rate": 1.0051993067590988e-06,
+      "loss": 3.2098,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 251.94387271044693,
+      "learning_rate": 1.0398613518197575e-06,
+      "loss": 2.8078,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 259.93646338353125,
+      "learning_rate": 1.074523396880416e-06,
+      "loss": 2.7944,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 232.413526602955,
+      "learning_rate": 1.1091854419410745e-06,
+      "loss": 2.6545,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 208.38564706751654,
+      "learning_rate": 1.1438474870017332e-06,
+      "loss": 2.7643,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 265.57456848345043,
+      "learning_rate": 1.178509532062392e-06,
+      "loss": 2.8018,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 215.4814469035693,
+      "learning_rate": 1.2131715771230504e-06,
+      "loss": 2.7896,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 230.06549904341952,
+      "learning_rate": 1.2478336221837089e-06,
+      "loss": 2.5891,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 219.1181457457231,
+      "learning_rate": 1.2824956672443676e-06,
+      "loss": 2.6911,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 185.43037205937995,
+      "learning_rate": 1.317157712305026e-06,
+      "loss": 2.5892,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 187.6794994680262,
+      "learning_rate": 1.3518197573656846e-06,
+      "loss": 2.4185,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 162.47805663798107,
+      "learning_rate": 1.3864818024263433e-06,
+      "loss": 2.1959,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 161.39439264930397,
+      "learning_rate": 1.4211438474870018e-06,
+      "loss": 2.3471,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 159.29627737519704,
+      "learning_rate": 1.4558058925476603e-06,
+      "loss": 2.4133,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 164.20193479561988,
+      "learning_rate": 1.490467937608319e-06,
+      "loss": 2.2179,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 154.96959210809524,
+      "learning_rate": 1.5251299826689774e-06,
+      "loss": 2.2498,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 170.34193926954364,
+      "learning_rate": 1.5597920277296362e-06,
+      "loss": 2.499,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 154.95462454843133,
+      "learning_rate": 1.5944540727902946e-06,
+      "loss": 2.3535,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 142.67736791126654,
+      "learning_rate": 1.6291161178509536e-06,
+      "loss": 2.1833,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 160.6522866668997,
+      "learning_rate": 1.663778162911612e-06,
+      "loss": 2.185,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 135.5072566549118,
+      "learning_rate": 1.6984402079722705e-06,
+      "loss": 2.1275,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 131.61379134082398,
+      "learning_rate": 1.7331022530329292e-06,
+      "loss": 2.1528,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 136.90740248824787,
+      "learning_rate": 1.7677642980935877e-06,
+      "loss": 1.9863,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 127.57276105940032,
+      "learning_rate": 1.8024263431542462e-06,
+      "loss": 1.941,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 125.8039279221426,
+      "learning_rate": 1.837088388214905e-06,
+      "loss": 2.2142,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 124.27249475324061,
+      "learning_rate": 1.8717504332755634e-06,
+      "loss": 2.0604,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 118.9750226358455,
+      "learning_rate": 1.906412478336222e-06,
+      "loss": 2.0378,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 106.44983872615984,
+      "learning_rate": 1.9410745233968806e-06,
+      "loss": 1.8921,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 124.8947448241858,
+      "learning_rate": 1.9757365684575393e-06,
+      "loss": 2.1142,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 167.3484240879138,
+      "learning_rate": 2.0103986135181976e-06,
+      "loss": 1.835,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 109.70099925904884,
+      "learning_rate": 2.0450606585788563e-06,
+      "loss": 1.9961,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 114.9557432514537,
+      "learning_rate": 2.079722703639515e-06,
+      "loss": 1.8055,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 115.33180474681399,
+      "learning_rate": 2.1143847487001733e-06,
+      "loss": 2.1272,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 111.62692082237943,
+      "learning_rate": 2.149046793760832e-06,
+      "loss": 1.9824,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 112.7239811177589,
+      "learning_rate": 2.1837088388214907e-06,
+      "loss": 1.8635,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 97.5961206290538,
+      "learning_rate": 2.218370883882149e-06,
+      "loss": 1.7146,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 108.63945918100269,
+      "learning_rate": 2.2530329289428077e-06,
+      "loss": 1.9992,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 87.3367736869193,
+      "learning_rate": 2.2876949740034664e-06,
+      "loss": 1.7425,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 95.43797650870907,
+      "learning_rate": 2.322357019064125e-06,
+      "loss": 1.9344,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 94.94115456534219,
+      "learning_rate": 2.357019064124784e-06,
+      "loss": 1.8151,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 99.95097686763462,
+      "learning_rate": 2.391681109185442e-06,
+      "loss": 1.7317,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 92.65154395672366,
+      "learning_rate": 2.4263431542461008e-06,
+      "loss": 1.9249,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 90.46520054101683,
+      "learning_rate": 2.4610051993067595e-06,
+      "loss": 1.7785,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 101.6665522611214,
+      "learning_rate": 2.4956672443674178e-06,
+      "loss": 1.6704,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.92677421532774,
+      "learning_rate": 2.530329289428076e-06,
+      "loss": 1.8488,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.97563239307323,
+      "learning_rate": 2.564991334488735e-06,
+      "loss": 1.5999,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 91.01428362074849,
+      "learning_rate": 2.599653379549394e-06,
+      "loss": 1.6263,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 91.0252122536573,
+      "learning_rate": 2.634315424610052e-06,
+      "loss": 1.7976,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 109.82334489415369,
+      "learning_rate": 2.668977469670711e-06,
+      "loss": 1.8451,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 79.42541959493845,
+      "learning_rate": 2.703639514731369e-06,
+      "loss": 1.6034,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 98.83109221938268,
+      "learning_rate": 2.7383015597920283e-06,
+      "loss": 1.828,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 101.91010871139507,
+      "learning_rate": 2.7729636048526865e-06,
+      "loss": 1.7744,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 75.03496062204442,
+      "learning_rate": 2.8076256499133452e-06,
+      "loss": 1.5633,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 94.35167528129774,
+      "learning_rate": 2.8422876949740035e-06,
+      "loss": 1.7147,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 95.57066047908835,
+      "learning_rate": 2.8769497400346622e-06,
+      "loss": 1.7321,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 102.31826831214212,
+      "learning_rate": 2.9116117850953205e-06,
+      "loss": 1.6867,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 111.04018339749344,
+      "learning_rate": 2.9462738301559796e-06,
+      "loss": 1.568,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 76.59675617769855,
+      "learning_rate": 2.980935875216638e-06,
+      "loss": 1.5591,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 79.96790642720121,
+      "learning_rate": 3.0155979202772966e-06,
+      "loss": 1.5541,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.29283786058718,
+      "learning_rate": 3.050259965337955e-06,
+      "loss": 1.5211,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 95.00391205211656,
+      "learning_rate": 3.084922010398614e-06,
+      "loss": 1.6392,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 100.4529262107013,
+      "learning_rate": 3.1195840554592723e-06,
+      "loss": 1.5106,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 77.51873073034794,
+      "learning_rate": 3.154246100519931e-06,
+      "loss": 1.5892,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 84.06513971771699,
+      "learning_rate": 3.1889081455805893e-06,
+      "loss": 1.5065,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 73.74296727673983,
+      "learning_rate": 3.223570190641248e-06,
+      "loss": 1.5483,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 100.30736048952058,
+      "learning_rate": 3.258232235701907e-06,
+      "loss": 1.6447,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.31648827324219,
+      "learning_rate": 3.2928942807625654e-06,
+      "loss": 1.4618,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 101.49758138734676,
+      "learning_rate": 3.327556325823224e-06,
+      "loss": 1.5554,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 86.77888675118942,
+      "learning_rate": 3.3622183708838824e-06,
+      "loss": 1.5308,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 90.01539358312242,
+      "learning_rate": 3.396880415944541e-06,
+      "loss": 1.4422,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 95.36489411655344,
+      "learning_rate": 3.4315424610051994e-06,
+      "loss": 1.5629,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 98.17004763918068,
+      "learning_rate": 3.4662045060658585e-06,
+      "loss": 1.686,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 84.97771960690547,
+      "learning_rate": 3.5008665511265168e-06,
+      "loss": 1.3106,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 85.5491097828104,
+      "learning_rate": 3.5355285961871755e-06,
+      "loss": 1.5609,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 88.67665712342344,
+      "learning_rate": 3.5701906412478338e-06,
+      "loss": 1.6042,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 71.97949606290564,
+      "learning_rate": 3.6048526863084925e-06,
+      "loss": 1.3444,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 87.21499113964947,
+      "learning_rate": 3.6395147313691507e-06,
+      "loss": 1.4628,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 101.5656154870711,
+      "learning_rate": 3.67417677642981e-06,
+      "loss": 1.4604,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 89.81157378566616,
+      "learning_rate": 3.708838821490468e-06,
+      "loss": 1.5364,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 76.76703483332035,
+      "learning_rate": 3.743500866551127e-06,
+      "loss": 1.3857,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 96.5341018349352,
+      "learning_rate": 3.778162911611785e-06,
+      "loss": 1.5067,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 84.75106910998973,
+      "learning_rate": 3.812824956672444e-06,
+      "loss": 1.3686,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.52757354812553,
+      "learning_rate": 3.8474870017331025e-06,
+      "loss": 1.4072,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 80.23185860050212,
+      "learning_rate": 3.882149046793761e-06,
+      "loss": 1.4697,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 107.68565747881608,
+      "learning_rate": 3.91681109185442e-06,
+      "loss": 1.546,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.76222339815915,
+      "learning_rate": 3.951473136915079e-06,
+      "loss": 1.4746,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 80.57627327126316,
+      "learning_rate": 3.986135181975737e-06,
+      "loss": 1.29,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 74.2386373180671,
+      "learning_rate": 4.020797227036395e-06,
+      "loss": 1.3129,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.01371573601023,
+      "learning_rate": 4.055459272097054e-06,
+      "loss": 1.2939,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 78.08226861163364,
+      "learning_rate": 4.090121317157713e-06,
+      "loss": 1.3189,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 84.33331723897828,
+      "learning_rate": 4.124783362218371e-06,
+      "loss": 1.4596,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 77.38301817680343,
+      "learning_rate": 4.15944540727903e-06,
+      "loss": 1.4615,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 77.4481977042508,
+      "learning_rate": 4.194107452339689e-06,
+      "loss": 1.3382,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 74.1087535146045,
+      "learning_rate": 4.228769497400347e-06,
+      "loss": 1.3371,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 104.99142435567022,
+      "learning_rate": 4.263431542461005e-06,
+      "loss": 1.3598,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 82.88597066541594,
+      "learning_rate": 4.298093587521664e-06,
+      "loss": 1.3445,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 89.69105224486349,
+      "learning_rate": 4.332755632582323e-06,
+      "loss": 1.2581,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.15837188900747,
+      "learning_rate": 4.367417677642981e-06,
+      "loss": 1.4423,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 94.32233799813578,
+      "learning_rate": 4.40207972270364e-06,
+      "loss": 1.4572,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.28615486631894,
+      "learning_rate": 4.436741767764298e-06,
+      "loss": 1.3149,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 100.00153248566345,
+      "learning_rate": 4.471403812824957e-06,
+      "loss": 1.6833,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 81.66875837592688,
+      "learning_rate": 4.506065857885615e-06,
+      "loss": 1.418,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 74.32158586735912,
+      "learning_rate": 4.540727902946274e-06,
+      "loss": 1.326,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 85.54334482094389,
+      "learning_rate": 4.575389948006933e-06,
+      "loss": 1.4848,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 96.97171742271972,
+      "learning_rate": 4.6100519930675915e-06,
+      "loss": 1.5587,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 83.55619372793188,
+      "learning_rate": 4.64471403812825e-06,
+      "loss": 1.3478,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.78435631340496,
+      "learning_rate": 4.679376083188908e-06,
+      "loss": 1.4974,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 81.52643143731635,
+      "learning_rate": 4.714038128249568e-06,
+      "loss": 1.3749,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 83.39424103446007,
+      "learning_rate": 4.7487001733102254e-06,
+      "loss": 1.4502,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 95.73381432564736,
+      "learning_rate": 4.783362218370884e-06,
+      "loss": 1.5991,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 76.30204787599611,
+      "learning_rate": 4.818024263431543e-06,
+      "loss": 1.4635,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 73.38713590089472,
+      "learning_rate": 4.8526863084922016e-06,
+      "loss": 1.3703,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 65.5665709883524,
+      "learning_rate": 4.88734835355286e-06,
+      "loss": 1.3076,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 70.47605962968214,
+      "learning_rate": 4.922010398613519e-06,
+      "loss": 1.3047,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 77.7575761961434,
+      "learning_rate": 4.956672443674177e-06,
+      "loss": 1.363,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.18746845501549,
+      "learning_rate": 4.9913344887348355e-06,
+      "loss": 1.3,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 77.6199801383697,
+      "learning_rate": 5.025996533795494e-06,
+      "loss": 1.446,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 82.223414475443,
+      "learning_rate": 5.060658578856152e-06,
+      "loss": 1.5178,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 77.62438005086186,
+      "learning_rate": 5.095320623916812e-06,
+      "loss": 1.295,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 81.07619617672512,
+      "learning_rate": 5.12998266897747e-06,
+      "loss": 1.3572,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 76.10072223907095,
+      "learning_rate": 5.164644714038128e-06,
+      "loss": 1.3493,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 65.288093095014,
+      "learning_rate": 5.199306759098788e-06,
+      "loss": 1.3008,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.11591228289403,
+      "learning_rate": 5.2339688041594464e-06,
+      "loss": 1.3107,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 67.03511024866081,
+      "learning_rate": 5.268630849220104e-06,
+      "loss": 1.3956,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 72.68706964314558,
+      "learning_rate": 5.303292894280763e-06,
+      "loss": 1.3863,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 105.51713476198213,
+      "learning_rate": 5.337954939341422e-06,
+      "loss": 1.3117,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.50619242800786,
+      "learning_rate": 5.37261698440208e-06,
+      "loss": 1.4298,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.99314665049353,
+      "learning_rate": 5.407279029462738e-06,
+      "loss": 1.413,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.30252093291182,
+      "learning_rate": 5.441941074523397e-06,
+      "loss": 1.3212,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 87.88455132008714,
+      "learning_rate": 5.4766031195840565e-06,
+      "loss": 1.443,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 89.72171856397968,
+      "learning_rate": 5.511265164644714e-06,
+      "loss": 1.5778,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 93.61447295224879,
+      "learning_rate": 5.545927209705373e-06,
+      "loss": 1.4377,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 97.8933019370986,
+      "learning_rate": 5.580589254766031e-06,
+      "loss": 1.4244,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 74.09261784840938,
+      "learning_rate": 5.6152512998266905e-06,
+      "loss": 1.3485,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 68.91482938956673,
+      "learning_rate": 5.649913344887349e-06,
+      "loss": 1.2798,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 88.66105793181275,
+      "learning_rate": 5.684575389948007e-06,
+      "loss": 1.3902,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 70.96146292588008,
+      "learning_rate": 5.719237435008666e-06,
+      "loss": 1.3223,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.44857165827753,
+      "learning_rate": 5.7538994800693245e-06,
+      "loss": 1.2052,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 71.87809456175768,
+      "learning_rate": 5.788561525129983e-06,
+      "loss": 1.181,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 75.35527850965437,
+      "learning_rate": 5.823223570190641e-06,
+      "loss": 1.2732,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 58.223028274570936,
+      "learning_rate": 5.8578856152513006e-06,
+      "loss": 1.2246,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 91.15871235971257,
+      "learning_rate": 5.892547660311959e-06,
+      "loss": 1.291,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 81.9137631041113,
+      "learning_rate": 5.927209705372617e-06,
+      "loss": 1.4234,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 79.95032564756528,
+      "learning_rate": 5.961871750433276e-06,
+      "loss": 1.1723,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 69.5735335737958,
+      "learning_rate": 5.996533795493935e-06,
+      "loss": 1.3356,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 95.28540228193931,
+      "learning_rate": 6.031195840554593e-06,
+      "loss": 1.3032,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 69.87978360993628,
+      "learning_rate": 6.065857885615252e-06,
+      "loss": 1.1991,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 94.37093126229402,
+      "learning_rate": 6.10051993067591e-06,
+      "loss": 1.2384,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 93.96625321892587,
+      "learning_rate": 6.135181975736569e-06,
+      "loss": 1.4056,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 87.16217826194142,
+      "learning_rate": 6.169844020797228e-06,
+      "loss": 1.4524,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 93.57736796329085,
+      "learning_rate": 6.204506065857886e-06,
+      "loss": 1.4169,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 87.01804519824903,
+      "learning_rate": 6.239168110918545e-06,
+      "loss": 1.4716,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.94306322873956,
+      "learning_rate": 6.273830155979203e-06,
+      "loss": 1.29,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 80.92353788745528,
+      "learning_rate": 6.308492201039862e-06,
+      "loss": 1.1274,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 71.6250127023416,
+      "learning_rate": 6.34315424610052e-06,
+      "loss": 1.1417,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.07605848962,
+      "learning_rate": 6.377816291161179e-06,
+      "loss": 1.214,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 84.33838089499427,
+      "learning_rate": 6.412478336221838e-06,
+      "loss": 1.5608,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 78.73777173484888,
+      "learning_rate": 6.447140381282496e-06,
+      "loss": 1.2967,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 76.31499718306733,
+      "learning_rate": 6.481802426343155e-06,
+      "loss": 1.3586,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 82.21064279879197,
+      "learning_rate": 6.516464471403814e-06,
+      "loss": 1.4171,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 92.28633457622406,
+      "learning_rate": 6.551126516464472e-06,
+      "loss": 1.3439,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 68.90417207452326,
+      "learning_rate": 6.585788561525131e-06,
+      "loss": 1.2615,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.74827681714025,
+      "learning_rate": 6.620450606585789e-06,
+      "loss": 1.3458,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 71.90478599016213,
+      "learning_rate": 6.655112651646448e-06,
+      "loss": 1.3289,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 82.91009337729507,
+      "learning_rate": 6.689774696707106e-06,
+      "loss": 1.3056,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 62.62662593839958,
+      "learning_rate": 6.724436741767765e-06,
+      "loss": 1.3549,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 80.88839984476397,
+      "learning_rate": 6.759098786828423e-06,
+      "loss": 1.2851,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 80.18929521271168,
+      "learning_rate": 6.793760831889082e-06,
+      "loss": 1.343,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 70.53815394978403,
+      "learning_rate": 6.828422876949741e-06,
+      "loss": 1.2328,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.69264541735608,
+      "learning_rate": 6.863084922010399e-06,
+      "loss": 1.323,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.13368114750419,
+      "learning_rate": 6.8977469670710574e-06,
+      "loss": 1.2999,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 87.86871694185278,
+      "learning_rate": 6.932409012131717e-06,
+      "loss": 1.2731,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 68.51962309669658,
+      "learning_rate": 6.967071057192375e-06,
+      "loss": 1.2992,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 76.88283378570544,
+      "learning_rate": 7.0017331022530336e-06,
+      "loss": 1.2616,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 82.85656695541597,
+      "learning_rate": 7.036395147313691e-06,
+      "loss": 1.3263,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.88185445292964,
+      "learning_rate": 7.071057192374351e-06,
+      "loss": 1.3667,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 61.85915605005731,
+      "learning_rate": 7.10571923743501e-06,
+      "loss": 1.2455,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 65.55487751538267,
+      "learning_rate": 7.1403812824956675e-06,
+      "loss": 1.3052,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 58.57049641116647,
+      "learning_rate": 7.175043327556327e-06,
+      "loss": 1.2226,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 78.58818768030056,
+      "learning_rate": 7.209705372616985e-06,
+      "loss": 1.267,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 80.17001336448646,
+      "learning_rate": 7.244367417677644e-06,
+      "loss": 1.2535,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 69.09225256585592,
+      "learning_rate": 7.2790294627383015e-06,
+      "loss": 1.2368,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 78.71091447929254,
+      "learning_rate": 7.313691507798961e-06,
+      "loss": 1.4124,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 78.34738885974916,
+      "learning_rate": 7.34835355285962e-06,
+      "loss": 1.4138,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 72.51440749557011,
+      "learning_rate": 7.383015597920278e-06,
+      "loss": 1.2799,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 78.08911108043738,
+      "learning_rate": 7.417677642980936e-06,
+      "loss": 1.1688,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 62.77898940274033,
+      "learning_rate": 7.452339688041596e-06,
+      "loss": 1.1464,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 81.68976664007008,
+      "learning_rate": 7.487001733102254e-06,
+      "loss": 1.3159,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 68.28945912611569,
+      "learning_rate": 7.521663778162912e-06,
+      "loss": 1.2301,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 64.43879762040434,
+      "learning_rate": 7.55632582322357e-06,
+      "loss": 1.3778,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 75.42466541326074,
+      "learning_rate": 7.59098786828423e-06,
+      "loss": 1.3486,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 67.29466349854258,
+      "learning_rate": 7.625649913344888e-06,
+      "loss": 1.2353,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 65.43458944149177,
+      "learning_rate": 7.660311958405546e-06,
+      "loss": 1.2286,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 64.19574837584821,
+      "learning_rate": 7.694974003466205e-06,
+      "loss": 1.2867,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 84.67562008121887,
+      "learning_rate": 7.729636048526865e-06,
+      "loss": 1.3781,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 89.39433592661847,
+      "learning_rate": 7.764298093587522e-06,
+      "loss": 1.276,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 65.65204638497036,
+      "learning_rate": 7.79896013864818e-06,
+      "loss": 1.1359,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 66.47622758406872,
+      "learning_rate": 7.83362218370884e-06,
+      "loss": 1.1241,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 69.4787789680458,
+      "learning_rate": 7.868284228769498e-06,
+      "loss": 1.4059,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 73.50699488123999,
+      "learning_rate": 7.902946273830157e-06,
+      "loss": 1.2805,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 75.38566173601308,
+      "learning_rate": 7.937608318890815e-06,
+      "loss": 1.124,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 67.73156317096121,
+      "learning_rate": 7.972270363951475e-06,
+      "loss": 1.3136,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 64.23053603971192,
+      "learning_rate": 8.006932409012133e-06,
+      "loss": 1.2891,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 62.30371623367409,
+      "learning_rate": 8.04159445407279e-06,
+      "loss": 1.3063,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 60.982274004873595,
+      "learning_rate": 8.076256499133448e-06,
+      "loss": 1.2015,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 77.36486524478585,
+      "learning_rate": 8.110918544194108e-06,
+      "loss": 1.2691,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.58735483576407,
+      "learning_rate": 8.145580589254767e-06,
+      "loss": 1.3073,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.24962572844737,
+      "learning_rate": 8.180242634315425e-06,
+      "loss": 1.2662,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 77.79631434857455,
+      "learning_rate": 8.214904679376083e-06,
+      "loss": 1.1677,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 77.73689597484474,
+      "learning_rate": 8.249566724436743e-06,
+      "loss": 1.1177,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 89.11243854188291,
+      "learning_rate": 8.2842287694974e-06,
+      "loss": 1.3025,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 67.14021879609939,
+      "learning_rate": 8.31889081455806e-06,
+      "loss": 1.4252,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 61.52338545539713,
+      "learning_rate": 8.353552859618718e-06,
+      "loss": 1.2461,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 73.11863654072667,
+      "learning_rate": 8.388214904679377e-06,
+      "loss": 1.3165,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 76.21772807344504,
+      "learning_rate": 8.422876949740035e-06,
+      "loss": 1.3325,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 70.18232593165159,
+      "learning_rate": 8.457538994800693e-06,
+      "loss": 1.2266,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 84.79396516900984,
+      "learning_rate": 8.492201039861353e-06,
+      "loss": 1.1572,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 80.0635223442515,
+      "learning_rate": 8.52686308492201e-06,
+      "loss": 1.3424,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 65.89280063510057,
+      "learning_rate": 8.56152512998267e-06,
+      "loss": 1.3302,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 70.65968432509531,
+      "learning_rate": 8.596187175043328e-06,
+      "loss": 1.1999,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 72.55020605882214,
+      "learning_rate": 8.630849220103988e-06,
+      "loss": 1.2625,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.36523835050055,
+      "learning_rate": 8.665511265164645e-06,
+      "loss": 1.2979,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 65.97286967213022,
+      "learning_rate": 8.700173310225303e-06,
+      "loss": 1.2179,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 68.82645408516005,
+      "learning_rate": 8.734835355285963e-06,
+      "loss": 1.1572,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 75.11394215604057,
+      "learning_rate": 8.769497400346622e-06,
+      "loss": 1.2247,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 62.42230827171627,
+      "learning_rate": 8.80415944540728e-06,
+      "loss": 1.2971,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 63.40664724617534,
+      "learning_rate": 8.838821490467938e-06,
+      "loss": 1.2295,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 75.10096757471212,
+      "learning_rate": 8.873483535528596e-06,
+      "loss": 1.3005,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 76.41822925108471,
+      "learning_rate": 8.908145580589255e-06,
+      "loss": 1.2472,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 82.55947807493577,
+      "learning_rate": 8.942807625649913e-06,
+      "loss": 1.2824,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 71.89019229379794,
+      "learning_rate": 8.977469670710573e-06,
+      "loss": 1.165,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 70.59994876794875,
+      "learning_rate": 9.01213171577123e-06,
+      "loss": 1.2805,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 63.76536700966057,
+      "learning_rate": 9.04679376083189e-06,
+      "loss": 1.2512,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 73.17475755427851,
+      "learning_rate": 9.081455805892548e-06,
+      "loss": 1.0849,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 64.38020067410628,
+      "learning_rate": 9.116117850953206e-06,
+      "loss": 1.1304,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 69.80888207058923,
+      "learning_rate": 9.150779896013866e-06,
+      "loss": 1.224,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.09562579692238,
+      "learning_rate": 9.185441941074525e-06,
+      "loss": 1.2356,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 60.42932296080759,
+      "learning_rate": 9.220103986135183e-06,
+      "loss": 1.2895,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 78.92199398281785,
+      "learning_rate": 9.25476603119584e-06,
+      "loss": 1.1719,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 63.7496298386542,
+      "learning_rate": 9.2894280762565e-06,
+      "loss": 1.2104,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.69661493585275,
+      "learning_rate": 9.324090121317158e-06,
+      "loss": 1.2585,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 62.73472058876782,
+      "learning_rate": 9.358752166377816e-06,
+      "loss": 1.2323,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 63.10154469650261,
+      "learning_rate": 9.393414211438476e-06,
+      "loss": 1.1616,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 68.15482457212605,
+      "learning_rate": 9.428076256499135e-06,
+      "loss": 1.2739,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 80.91560719271482,
+      "learning_rate": 9.462738301559793e-06,
+      "loss": 1.309,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 65.1295896585928,
+      "learning_rate": 9.497400346620451e-06,
+      "loss": 1.1954,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 68.82844038684304,
+      "learning_rate": 9.532062391681109e-06,
+      "loss": 1.3214,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.26397526676573,
+      "learning_rate": 9.566724436741768e-06,
+      "loss": 1.4152,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 65.09316639814858,
+      "learning_rate": 9.601386481802428e-06,
+      "loss": 1.1335,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 71.19376600920194,
+      "learning_rate": 9.636048526863086e-06,
+      "loss": 1.2977,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 50.52090685127913,
+      "learning_rate": 9.670710571923744e-06,
+      "loss": 1.0319,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 66.37389322568637,
+      "learning_rate": 9.705372616984403e-06,
+      "loss": 1.2205,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 58.91509883400627,
+      "learning_rate": 9.740034662045061e-06,
+      "loss": 1.1621,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 64.12478707558898,
+      "learning_rate": 9.77469670710572e-06,
+      "loss": 1.2108,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 71.92159263063138,
+      "learning_rate": 9.809358752166378e-06,
+      "loss": 1.2372,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 60.4547514806354,
+      "learning_rate": 9.844020797227038e-06,
+      "loss": 1.1497,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 58.30009812662494,
+      "learning_rate": 9.878682842287696e-06,
+      "loss": 1.2004,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 71.349371402487,
+      "learning_rate": 9.913344887348354e-06,
+      "loss": 1.1825,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 87.66660868590823,
+      "learning_rate": 9.948006932409013e-06,
+      "loss": 1.1576,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 62.97275167391369,
+      "learning_rate": 9.982668977469671e-06,
+      "loss": 1.1288,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 73.92494489108506,
+      "learning_rate": 1.001733102253033e-05,
+      "loss": 1.288,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 67.6683912065295,
+      "learning_rate": 1.0051993067590988e-05,
+      "loss": 1.1967,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 75.3569102808644,
+      "learning_rate": 1.0086655112651646e-05,
+      "loss": 1.3508,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 83.8219628393352,
+      "learning_rate": 1.0121317157712304e-05,
+      "loss": 1.24,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 70.03856616974602,
+      "learning_rate": 1.0155979202772965e-05,
+      "loss": 1.2625,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 66.36984413941386,
+      "learning_rate": 1.0190641247833623e-05,
+      "loss": 1.2037,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 78.38786448833612,
+      "learning_rate": 1.0225303292894281e-05,
+      "loss": 1.1828,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 86.28527762944036,
+      "learning_rate": 1.025996533795494e-05,
+      "loss": 1.2004,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.94417661795292,
+      "learning_rate": 1.0294627383015599e-05,
+      "loss": 1.2249,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 62.55630324520849,
+      "learning_rate": 1.0329289428076256e-05,
+      "loss": 1.2524,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 66.0771802544681,
+      "learning_rate": 1.0363951473136914e-05,
+      "loss": 1.0969,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 71.01405610385807,
+      "learning_rate": 1.0398613518197575e-05,
+      "loss": 1.2197,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 63.76545554159353,
+      "learning_rate": 1.0433275563258233e-05,
+      "loss": 1.1933,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 93.93077955743048,
+      "learning_rate": 1.0467937608318893e-05,
+      "loss": 1.2229,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 73.96581262155901,
+      "learning_rate": 1.050259965337955e-05,
+      "loss": 1.268,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 49.223061702727335,
+      "learning_rate": 1.0537261698440209e-05,
+      "loss": 1.099,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 67.25899597183113,
+      "learning_rate": 1.0571923743500866e-05,
+      "loss": 1.263,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 65.60229239558048,
+      "learning_rate": 1.0606585788561526e-05,
+      "loss": 1.2032,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 51.566702936135485,
+      "learning_rate": 1.0641247833622184e-05,
+      "loss": 1.1452,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 68.11182572956385,
+      "learning_rate": 1.0675909878682843e-05,
+      "loss": 1.2211,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 67.78959206249743,
+      "learning_rate": 1.0710571923743503e-05,
+      "loss": 1.1892,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 81.02616413248847,
+      "learning_rate": 1.074523396880416e-05,
+      "loss": 1.1845,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 80.75786006690277,
+      "learning_rate": 1.0779896013864819e-05,
+      "loss": 1.2253,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 65.43325629405486,
+      "learning_rate": 1.0814558058925477e-05,
+      "loss": 1.1961,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 67.25023566704621,
+      "learning_rate": 1.0849220103986136e-05,
+      "loss": 1.1948,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 58.609349746312205,
+      "learning_rate": 1.0883882149046794e-05,
+      "loss": 1.2042,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 63.385707845988144,
+      "learning_rate": 1.0918544194107452e-05,
+      "loss": 1.1536,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 70.68019264557626,
+      "learning_rate": 1.0953206239168113e-05,
+      "loss": 1.2477,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 65.80991708409144,
+      "learning_rate": 1.0987868284228771e-05,
+      "loss": 1.1846,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 67.74653719396343,
+      "learning_rate": 1.1022530329289429e-05,
+      "loss": 1.1521,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 61.2948307160295,
+      "learning_rate": 1.1057192374350088e-05,
+      "loss": 1.168,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 57.13640744356658,
+      "learning_rate": 1.1091854419410746e-05,
+      "loss": 1.2183,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 63.03090801155694,
+      "learning_rate": 1.1126516464471404e-05,
+      "loss": 1.3613,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 65.91954267507542,
+      "learning_rate": 1.1161178509532062e-05,
+      "loss": 1.1695,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 73.0382326809134,
+      "learning_rate": 1.1195840554592723e-05,
+      "loss": 1.211,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 59.32469368657422,
+      "learning_rate": 1.1230502599653381e-05,
+      "loss": 1.1417,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 58.64989713837042,
+      "learning_rate": 1.1265164644714039e-05,
+      "loss": 1.082,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 58.49607800109346,
+      "learning_rate": 1.1299826689774698e-05,
+      "loss": 1.2012,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 58.701229378598,
+      "learning_rate": 1.1334488734835356e-05,
+      "loss": 1.104,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 62.903573446851354,
+      "learning_rate": 1.1369150779896014e-05,
+      "loss": 1.1482,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.31490107170087,
+      "learning_rate": 1.1403812824956672e-05,
+      "loss": 1.1785,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 55.78536348108741,
+      "learning_rate": 1.1438474870017332e-05,
+      "loss": 1.1983,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 57.72512312124677,
+      "learning_rate": 1.1473136915077991e-05,
+      "loss": 1.112,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 74.24071131188715,
+      "learning_rate": 1.1507798960138649e-05,
+      "loss": 1.1472,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 71.13702183711722,
+      "learning_rate": 1.1542461005199308e-05,
+      "loss": 1.1831,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 71.22769819874947,
+      "learning_rate": 1.1577123050259966e-05,
+      "loss": 1.1927,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 67.81029153207348,
+      "learning_rate": 1.1611785095320624e-05,
+      "loss": 1.1728,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.13990087065547,
+      "learning_rate": 1.1646447140381282e-05,
+      "loss": 1.1377,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 65.74844902442979,
+      "learning_rate": 1.1681109185441942e-05,
+      "loss": 1.1422,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 85.3650019810954,
+      "learning_rate": 1.1715771230502601e-05,
+      "loss": 1.1299,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.7260160912315,
+      "learning_rate": 1.175043327556326e-05,
+      "loss": 1.2441,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 75.1354249331658,
+      "learning_rate": 1.1785095320623919e-05,
+      "loss": 1.0657,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 69.54765149010646,
+      "learning_rate": 1.1819757365684576e-05,
+      "loss": 1.2289,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 71.47371946203792,
+      "learning_rate": 1.1854419410745234e-05,
+      "loss": 1.3144,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 79.95203344491958,
+      "learning_rate": 1.1889081455805894e-05,
+      "loss": 1.2666,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 60.53547014089961,
+      "learning_rate": 1.1923743500866552e-05,
+      "loss": 1.2814,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 63.36572762849882,
+      "learning_rate": 1.195840554592721e-05,
+      "loss": 0.9879,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 79.8861767108486,
+      "learning_rate": 1.199306759098787e-05,
+      "loss": 1.1528,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 70.90433690267237,
+      "learning_rate": 1.2027729636048529e-05,
+      "loss": 1.2205,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.59792898655348,
+      "learning_rate": 1.2062391681109186e-05,
+      "loss": 1.2038,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 63.45088364306725,
+      "learning_rate": 1.2097053726169844e-05,
+      "loss": 1.1477,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 80.55826494732206,
+      "learning_rate": 1.2131715771230504e-05,
+      "loss": 1.2393,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 54.70517768484438,
+      "learning_rate": 1.2166377816291162e-05,
+      "loss": 1.1038,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 63.89259651409343,
+      "learning_rate": 1.220103986135182e-05,
+      "loss": 1.0892,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 60.157544303940824,
+      "learning_rate": 1.2235701906412477e-05,
+      "loss": 1.1897,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 56.50610050248727,
+      "learning_rate": 1.2270363951473139e-05,
+      "loss": 1.1902,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 74.20016789764368,
+      "learning_rate": 1.2305025996533797e-05,
+      "loss": 1.0565,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 58.000982901038604,
+      "learning_rate": 1.2339688041594456e-05,
+      "loss": 1.2204,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 71.83579902075508,
+      "learning_rate": 1.2374350086655114e-05,
+      "loss": 1.1661,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 61.480850118686355,
+      "learning_rate": 1.2409012131715772e-05,
+      "loss": 1.099,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 57.86124714030461,
+      "learning_rate": 1.244367417677643e-05,
+      "loss": 1.1887,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 84.13906210011879,
+      "learning_rate": 1.247833622183709e-05,
+      "loss": 1.3047,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 68.7304530151037,
+      "learning_rate": 1.2512998266897749e-05,
+      "loss": 1.2676,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 72.10378700501013,
+      "learning_rate": 1.2547660311958407e-05,
+      "loss": 1.2424,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 79.15764619399386,
+      "learning_rate": 1.2582322357019066e-05,
+      "loss": 1.2426,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 68.13793265401824,
+      "learning_rate": 1.2616984402079724e-05,
+      "loss": 1.3042,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 56.827794776525536,
+      "learning_rate": 1.2651646447140382e-05,
+      "loss": 1.1288,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 62.104838256207636,
+      "learning_rate": 1.268630849220104e-05,
+      "loss": 1.2537,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 61.83964711554935,
+      "learning_rate": 1.27209705372617e-05,
+      "loss": 1.1719,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 59.06921430095909,
+      "learning_rate": 1.2755632582322357e-05,
+      "loss": 1.145,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 59.14989526776468,
+      "learning_rate": 1.2790294627383017e-05,
+      "loss": 1.1483,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 63.29215216267334,
+      "learning_rate": 1.2824956672443676e-05,
+      "loss": 1.1508,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 54.20104531339387,
+      "learning_rate": 1.2859618717504334e-05,
+      "loss": 1.1036,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 74.4088322665211,
+      "learning_rate": 1.2894280762564992e-05,
+      "loss": 1.1305,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 82.69548773442922,
+      "learning_rate": 1.2928942807625652e-05,
+      "loss": 1.0137,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 58.47610773127586,
+      "learning_rate": 1.296360485268631e-05,
+      "loss": 1.1698,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 67.01826290774318,
+      "learning_rate": 1.2998266897746967e-05,
+      "loss": 1.152,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 79.25480838833575,
+      "learning_rate": 1.3032928942807628e-05,
+      "loss": 1.0871,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 61.750457182824995,
+      "learning_rate": 1.3067590987868286e-05,
+      "loss": 1.2561,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 59.47473034278443,
+      "learning_rate": 1.3102253032928944e-05,
+      "loss": 1.1967,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 54.85825595567278,
+      "learning_rate": 1.3136915077989602e-05,
+      "loss": 1.0799,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 67.71089697185539,
+      "learning_rate": 1.3171577123050262e-05,
+      "loss": 1.3144,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 54.79755397927458,
+      "learning_rate": 1.320623916811092e-05,
+      "loss": 1.2351,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 79.82527401354717,
+      "learning_rate": 1.3240901213171577e-05,
+      "loss": 1.2384,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 63.23489568942272,
+      "learning_rate": 1.3275563258232235e-05,
+      "loss": 1.0403,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 65.6372113882769,
+      "learning_rate": 1.3310225303292896e-05,
+      "loss": 1.0802,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 74.34288040428639,
+      "learning_rate": 1.3344887348353554e-05,
+      "loss": 1.1387,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 70.93585861256454,
+      "learning_rate": 1.3379549393414212e-05,
+      "loss": 1.136,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 81.64973385166537,
+      "learning_rate": 1.3414211438474872e-05,
+      "loss": 1.2067,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 76.61602473006299,
+      "learning_rate": 1.344887348353553e-05,
+      "loss": 1.0373,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 69.50594836789516,
+      "learning_rate": 1.3483535528596187e-05,
+      "loss": 1.1393,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 58.30404901135747,
+      "learning_rate": 1.3518197573656845e-05,
+      "loss": 1.3005,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 63.1057979498411,
+      "learning_rate": 1.3552859618717506e-05,
+      "loss": 1.0849,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 57.3034890515657,
+      "learning_rate": 1.3587521663778164e-05,
+      "loss": 1.1536,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 71.24803954656451,
+      "learning_rate": 1.3622183708838824e-05,
+      "loss": 1.1657,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 74.67389672546372,
+      "learning_rate": 1.3656845753899482e-05,
+      "loss": 1.2172,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 62.772450355881766,
+      "learning_rate": 1.369150779896014e-05,
+      "loss": 1.1514,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 72.43855461294235,
+      "learning_rate": 1.3726169844020797e-05,
+      "loss": 1.2246,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 58.645114668874605,
+      "learning_rate": 1.3760831889081457e-05,
+      "loss": 1.1424,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 63.135209932939695,
+      "learning_rate": 1.3795493934142115e-05,
+      "loss": 1.116,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 58.99421106645925,
+      "learning_rate": 1.3830155979202774e-05,
+      "loss": 1.0929,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 51.08519152901567,
+      "learning_rate": 1.3864818024263434e-05,
+      "loss": 1.0516,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 53.06996186790728,
+      "learning_rate": 1.3899480069324092e-05,
+      "loss": 0.9497,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 64.06055074564,
+      "learning_rate": 1.393414211438475e-05,
+      "loss": 1.1981,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 63.86788225921808,
+      "learning_rate": 1.3968804159445408e-05,
+      "loss": 1.0479,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 52.835603506771875,
+      "learning_rate": 1.4003466204506067e-05,
+      "loss": 1.2545,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 93.6489366355929,
+      "learning_rate": 1.4038128249566725e-05,
+      "loss": 1.0769,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 65.65745398803303,
+      "learning_rate": 1.4072790294627383e-05,
+      "loss": 1.2111,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 60.243606837593575,
+      "learning_rate": 1.4107452339688044e-05,
+      "loss": 1.1864,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 53.969864701784786,
+      "learning_rate": 1.4142114384748702e-05,
+      "loss": 1.1016,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 61.94301067977638,
+      "learning_rate": 1.417677642980936e-05,
+      "loss": 1.2142,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 53.66868018091035,
+      "learning_rate": 1.421143847487002e-05,
+      "loss": 1.2838,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 59.38272058986461,
+      "learning_rate": 1.4246100519930677e-05,
+      "loss": 1.0981,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 60.95718659708407,
+      "learning_rate": 1.4280762564991335e-05,
+      "loss": 1.1298,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 87.93604989337199,
+      "learning_rate": 1.4315424610051993e-05,
+      "loss": 1.0849,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 52.94528788149057,
+      "learning_rate": 1.4350086655112654e-05,
+      "loss": 1.0133,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 76.67072036088314,
+      "learning_rate": 1.4384748700173312e-05,
+      "loss": 0.9958,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 64.23578508073244,
+      "learning_rate": 1.441941074523397e-05,
+      "loss": 1.1518,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 66.9117196852638,
+      "learning_rate": 1.445407279029463e-05,
+      "loss": 1.088,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 53.972249834814164,
+      "learning_rate": 1.4488734835355287e-05,
+      "loss": 1.1206,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.86533311795346,
+      "learning_rate": 1.4523396880415945e-05,
+      "loss": 1.0437,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 75.2435856743609,
+      "learning_rate": 1.4558058925476603e-05,
+      "loss": 1.3,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 62.63796945006593,
+      "learning_rate": 1.4592720970537263e-05,
+      "loss": 1.0384,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 57.12562726383858,
+      "learning_rate": 1.4627383015597922e-05,
+      "loss": 1.2567,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 65.71783695502725,
+      "learning_rate": 1.466204506065858e-05,
+      "loss": 1.1368,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 65.03464232674808,
+      "learning_rate": 1.469670710571924e-05,
+      "loss": 1.0667,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 99.7031743561085,
+      "learning_rate": 1.4731369150779897e-05,
+      "loss": 1.0991,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 52.26865163572845,
+      "learning_rate": 1.4766031195840555e-05,
+      "loss": 1.1365,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 65.26190192827704,
+      "learning_rate": 1.4800693240901213e-05,
+      "loss": 1.116,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.7455163142336,
+      "learning_rate": 1.4835355285961873e-05,
+      "loss": 1.0812,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 54.22442644729594,
+      "learning_rate": 1.4870017331022532e-05,
+      "loss": 1.0409,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 46.98669148282551,
+      "learning_rate": 1.4904679376083192e-05,
+      "loss": 1.1024,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 62.19206714824058,
+      "learning_rate": 1.493934142114385e-05,
+      "loss": 1.1478,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 68.98061233229923,
+      "learning_rate": 1.4974003466204507e-05,
+      "loss": 1.0617,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 67.07023612958395,
+      "learning_rate": 1.5008665511265165e-05,
+      "loss": 1.0752,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.03186027819428,
+      "learning_rate": 1.5043327556325825e-05,
+      "loss": 1.0074,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 63.39181656971643,
+      "learning_rate": 1.5077989601386483e-05,
+      "loss": 1.1737,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 62.50658345453454,
+      "learning_rate": 1.511265164644714e-05,
+      "loss": 1.1648,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 68.33028306743597,
+      "learning_rate": 1.5147313691507802e-05,
+      "loss": 1.1315,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 54.39571897898176,
+      "learning_rate": 1.518197573656846e-05,
+      "loss": 1.1112,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 59.952835869707094,
+      "learning_rate": 1.5216637781629117e-05,
+      "loss": 1.1112,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 62.96802893289651,
+      "learning_rate": 1.5251299826689775e-05,
+      "loss": 1.087,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.93282575714194,
+      "learning_rate": 1.5285961871750435e-05,
+      "loss": 1.1854,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 51.81839486744537,
+      "learning_rate": 1.532062391681109e-05,
+      "loss": 1.0663,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 77.02845347437089,
+      "learning_rate": 1.535528596187175e-05,
+      "loss": 1.181,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 51.34726779655865,
+      "learning_rate": 1.538994800693241e-05,
+      "loss": 1.0873,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 58.686183398736155,
+      "learning_rate": 1.542461005199307e-05,
+      "loss": 1.0679,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 51.41774986131819,
+      "learning_rate": 1.545927209705373e-05,
+      "loss": 1.2115,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.74194147644112,
+      "learning_rate": 1.5493934142114385e-05,
+      "loss": 1.1091,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 58.31134262653624,
+      "learning_rate": 1.5528596187175045e-05,
+      "loss": 1.0671,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 53.87842327668372,
+      "learning_rate": 1.5563258232235705e-05,
+      "loss": 1.0835,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 60.901140729747,
+      "learning_rate": 1.559792027729636e-05,
+      "loss": 1.1129,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 54.889982932635924,
+      "learning_rate": 1.563258232235702e-05,
+      "loss": 1.0839,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 47.60344202615584,
+      "learning_rate": 1.566724436741768e-05,
+      "loss": 0.9668,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 67.21208588478083,
+      "learning_rate": 1.570190641247834e-05,
+      "loss": 1.1727,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 64.99863745655128,
+      "learning_rate": 1.5736568457538996e-05,
+      "loss": 1.2614,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 52.756781661906004,
+      "learning_rate": 1.5771230502599655e-05,
+      "loss": 1.0732,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 60.54090987028067,
+      "learning_rate": 1.5805892547660315e-05,
+      "loss": 1.1314,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 51.529545937996694,
+      "learning_rate": 1.584055459272097e-05,
+      "loss": 1.0455,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 50.36091645971097,
+      "learning_rate": 1.587521663778163e-05,
+      "loss": 1.0717,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 60.23629395511878,
+      "learning_rate": 1.5909878682842286e-05,
+      "loss": 1.1121,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.4531206922321,
+      "learning_rate": 1.594454072790295e-05,
+      "loss": 1.177,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.54051444856762,
+      "learning_rate": 1.5979202772963606e-05,
+      "loss": 1.073,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 52.288249001643166,
+      "learning_rate": 1.6013864818024265e-05,
+      "loss": 1.0638,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 54.80172913357167,
+      "learning_rate": 1.6048526863084925e-05,
+      "loss": 1.1067,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.20176035900222,
+      "learning_rate": 1.608318890814558e-05,
+      "loss": 1.0353,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 73.39779826195995,
+      "learning_rate": 1.611785095320624e-05,
+      "loss": 1.1815,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 68.45858492798254,
+      "learning_rate": 1.6152512998266897e-05,
+      "loss": 1.1328,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.1967002568119,
+      "learning_rate": 1.618717504332756e-05,
+      "loss": 1.2506,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.8320414693709,
+      "learning_rate": 1.6221837088388216e-05,
+      "loss": 1.0804,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 47.002327891792966,
+      "learning_rate": 1.6256499133448875e-05,
+      "loss": 1.0963,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 63.094641737283176,
+      "learning_rate": 1.6291161178509535e-05,
+      "loss": 1.1115,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 56.747609336005226,
+      "learning_rate": 1.632582322357019e-05,
+      "loss": 1.112,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 51.723842229556936,
+      "learning_rate": 1.636048526863085e-05,
+      "loss": 1.0193,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.054021373587965,
+      "learning_rate": 1.639514731369151e-05,
+      "loss": 1.109,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 63.24030930838565,
+      "learning_rate": 1.6429809358752166e-05,
+      "loss": 1.1094,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 52.06202073809703,
+      "learning_rate": 1.6464471403812826e-05,
+      "loss": 1.1136,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 58.98064682654012,
+      "learning_rate": 1.6499133448873485e-05,
+      "loss": 1.0803,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 64.27605146084422,
+      "learning_rate": 1.6533795493934145e-05,
+      "loss": 1.0317,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 77.13972193525666,
+      "learning_rate": 1.65684575389948e-05,
+      "loss": 1.064,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 77.54698462462186,
+      "learning_rate": 1.660311958405546e-05,
+      "loss": 1.1975,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 70.87072898469366,
+      "learning_rate": 1.663778162911612e-05,
+      "loss": 1.2299,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 69.44794147316117,
+      "learning_rate": 1.6672443674176776e-05,
+      "loss": 1.1362,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 48.59338932210147,
+      "learning_rate": 1.6707105719237436e-05,
+      "loss": 1.1652,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 66.3923783406331,
+      "learning_rate": 1.6741767764298095e-05,
+      "loss": 1.0673,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 45.97978170917194,
+      "learning_rate": 1.6776429809358755e-05,
+      "loss": 1.0986,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 122.66647180424292,
+      "learning_rate": 1.681109185441941e-05,
+      "loss": 1.094,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 57.25808322679577,
+      "learning_rate": 1.684575389948007e-05,
+      "loss": 1.0637,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 51.31527695941718,
+      "learning_rate": 1.688041594454073e-05,
+      "loss": 1.2241,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 47.93772629009287,
+      "learning_rate": 1.6915077989601386e-05,
+      "loss": 1.2574,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 59.125490396252275,
+      "learning_rate": 1.6949740034662046e-05,
+      "loss": 1.3108,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 74.42804261415363,
+      "learning_rate": 1.6984402079722705e-05,
+      "loss": 1.1507,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 55.406919291370215,
+      "learning_rate": 1.7019064124783365e-05,
+      "loss": 1.1584,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 55.69597607464922,
+      "learning_rate": 1.705372616984402e-05,
+      "loss": 1.2148,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 69.75228003053383,
+      "learning_rate": 1.708838821490468e-05,
+      "loss": 1.1521,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 63.08016773211252,
+      "learning_rate": 1.712305025996534e-05,
+      "loss": 1.1564,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 52.53254386765321,
+      "learning_rate": 1.7157712305025996e-05,
+      "loss": 1.1221,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 54.9087911939422,
+      "learning_rate": 1.7192374350086656e-05,
+      "loss": 1.0352,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 71.68029331876355,
+      "learning_rate": 1.7227036395147316e-05,
+      "loss": 1.192,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 55.27508106389651,
+      "learning_rate": 1.7261698440207975e-05,
+      "loss": 1.151,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 56.534749670813014,
+      "learning_rate": 1.729636048526863e-05,
+      "loss": 1.189,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 55.33066436016416,
+      "learning_rate": 1.733102253032929e-05,
+      "loss": 1.1011,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 61.02580433041579,
+      "learning_rate": 1.736568457538995e-05,
+      "loss": 1.091,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 53.64756437107474,
+      "learning_rate": 1.7400346620450606e-05,
+      "loss": 1.1407,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 91.81030516990582,
+      "learning_rate": 1.7435008665511266e-05,
+      "loss": 1.2099,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 65.02907800965572,
+      "learning_rate": 1.7469670710571926e-05,
+      "loss": 1.0422,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 54.50449419302549,
+      "learning_rate": 1.7504332755632585e-05,
+      "loss": 1.22,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 61.283138854909815,
+      "learning_rate": 1.7538994800693245e-05,
+      "loss": 1.1386,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 50.347231658700814,
+      "learning_rate": 1.75736568457539e-05,
+      "loss": 1.0809,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 58.84205037486989,
+      "learning_rate": 1.760831889081456e-05,
+      "loss": 1.1181,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 58.84022372837956,
+      "learning_rate": 1.7642980935875217e-05,
+      "loss": 1.0518,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 70.16454526800784,
+      "learning_rate": 1.7677642980935876e-05,
+      "loss": 1.0832,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 51.420614333160756,
+      "learning_rate": 1.7712305025996536e-05,
+      "loss": 1.1343,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 61.23445202198913,
+      "learning_rate": 1.7746967071057192e-05,
+      "loss": 0.9714,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 52.32590597236863,
+      "learning_rate": 1.7781629116117855e-05,
+      "loss": 1.1535,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 63.19009071117808,
+      "learning_rate": 1.781629116117851e-05,
+      "loss": 1.2021,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 61.11862826535637,
+      "learning_rate": 1.785095320623917e-05,
+      "loss": 1.1743,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 46.391135850549624,
+      "learning_rate": 1.7885615251299827e-05,
+      "loss": 0.949,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 52.33712511641282,
+      "learning_rate": 1.7920277296360486e-05,
+      "loss": 1.0821,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 58.93357980904598,
+      "learning_rate": 1.7954939341421146e-05,
+      "loss": 1.1201,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 67.19939103758868,
+      "learning_rate": 1.7989601386481802e-05,
+      "loss": 1.08,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 58.44516818892173,
+      "learning_rate": 1.802426343154246e-05,
+      "loss": 1.1087,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 57.28093852173341,
+      "learning_rate": 1.805892547660312e-05,
+      "loss": 1.117,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 57.85086545865576,
+      "learning_rate": 1.809358752166378e-05,
+      "loss": 1.0491,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 63.14967697797904,
+      "learning_rate": 1.812824956672444e-05,
+      "loss": 1.178,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 59.52868360058764,
+      "learning_rate": 1.8162911611785096e-05,
+      "loss": 1.0969,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 48.253970704228244,
+      "learning_rate": 1.8197573656845756e-05,
+      "loss": 1.1448,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 75.96647159826999,
+      "learning_rate": 1.8232235701906412e-05,
+      "loss": 1.1161,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 60.29439514865257,
+      "learning_rate": 1.826689774696707e-05,
+      "loss": 1.0875,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 52.68853727068561,
+      "learning_rate": 1.830155979202773e-05,
+      "loss": 0.9589,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 49.441299249087656,
+      "learning_rate": 1.833622183708839e-05,
+      "loss": 1.1368,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 63.5675404218378,
+      "learning_rate": 1.837088388214905e-05,
+      "loss": 1.0286,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 63.85545863907957,
+      "learning_rate": 1.8405545927209706e-05,
+      "loss": 1.1007,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 57.45718326298227,
+      "learning_rate": 1.8440207972270366e-05,
+      "loss": 1.0269,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 45.09918090304873,
+      "learning_rate": 1.8474870017331022e-05,
+      "loss": 0.9757,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 98.72950199668871,
+      "learning_rate": 1.850953206239168e-05,
+      "loss": 1.0449,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 65.59801920336815,
+      "learning_rate": 1.854419410745234e-05,
+      "loss": 1.1194,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 53.75383752829987,
+      "learning_rate": 1.8578856152513e-05,
+      "loss": 1.1504,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 58.1738722399688,
+      "learning_rate": 1.861351819757366e-05,
+      "loss": 1.1342,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 53.5150507674769,
+      "learning_rate": 1.8648180242634316e-05,
+      "loss": 1.0258,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 54.044732948241275,
+      "learning_rate": 1.8682842287694976e-05,
+      "loss": 1.1848,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 53.4871396089391,
+      "learning_rate": 1.8717504332755632e-05,
+      "loss": 1.0197,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 80.96432941225731,
+      "learning_rate": 1.8752166377816292e-05,
+      "loss": 1.0242,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 61.25385032540811,
+      "learning_rate": 1.878682842287695e-05,
+      "loss": 1.2019,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 131.25973844457675,
+      "learning_rate": 1.882149046793761e-05,
+      "loss": 1.1536,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 47.381649028837366,
+      "learning_rate": 1.885615251299827e-05,
+      "loss": 1.0039,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.874961195896,
+      "learning_rate": 1.8890814558058927e-05,
+      "loss": 1.1824,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 48.70245452899101,
+      "learning_rate": 1.8925476603119586e-05,
+      "loss": 1.0284,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 50.13903965855429,
+      "learning_rate": 1.8960138648180246e-05,
+      "loss": 0.9775,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 82.47604817732976,
+      "learning_rate": 1.8994800693240902e-05,
+      "loss": 1.1361,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.17112071543417,
+      "learning_rate": 1.902946273830156e-05,
+      "loss": 1.1038,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 54.73929380660311,
+      "learning_rate": 1.9064124783362217e-05,
+      "loss": 1.1647,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 61.9813562787518,
+      "learning_rate": 1.909878682842288e-05,
+      "loss": 1.1791,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 60.58720033609694,
+      "learning_rate": 1.9133448873483537e-05,
+      "loss": 1.1097,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 102.62834003138241,
+      "learning_rate": 1.9168110918544196e-05,
+      "loss": 0.9908,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.79383628601807,
+      "learning_rate": 1.9202772963604856e-05,
+      "loss": 1.0576,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 46.69858631836959,
+      "learning_rate": 1.9237435008665512e-05,
+      "loss": 0.9884,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 54.780819971167,
+      "learning_rate": 1.927209705372617e-05,
+      "loss": 0.9957,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 53.076683269658574,
+      "learning_rate": 1.9306759098786828e-05,
+      "loss": 1.0947,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 53.74629941665709,
+      "learning_rate": 1.9341421143847487e-05,
+      "loss": 1.2104,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 61.33187157333344,
+      "learning_rate": 1.9376083188908147e-05,
+      "loss": 1.1281,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 87.40997614196097,
+      "learning_rate": 1.9410745233968806e-05,
+      "loss": 1.1147,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 56.03507474492482,
+      "learning_rate": 1.9445407279029466e-05,
+      "loss": 1.0697,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 59.97289696329996,
+      "learning_rate": 1.9480069324090122e-05,
+      "loss": 1.07,
+      "step": 562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.91022911858592,
+      "learning_rate": 1.951473136915078e-05,
+      "loss": 1.1342,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 67.1174814570591,
+      "learning_rate": 1.954939341421144e-05,
+      "loss": 1.2072,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 66.61034951871841,
+      "learning_rate": 1.9584055459272097e-05,
+      "loss": 1.2416,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 54.00239030969475,
+      "learning_rate": 1.9618717504332757e-05,
+      "loss": 1.0829,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.53352868208733,
+      "learning_rate": 1.9653379549393416e-05,
+      "loss": 1.1581,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 66.56007893059021,
+      "learning_rate": 1.9688041594454076e-05,
+      "loss": 1.3105,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 77.9410725794194,
+      "learning_rate": 1.9722703639514732e-05,
+      "loss": 1.1361,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 64.42815015357264,
+      "learning_rate": 1.975736568457539e-05,
+      "loss": 1.1388,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.33433724932763,
+      "learning_rate": 1.979202772963605e-05,
+      "loss": 1.1196,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 87.81772280328438,
+      "learning_rate": 1.9826689774696707e-05,
+      "loss": 1.3534,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 47.58649236956498,
+      "learning_rate": 1.9861351819757367e-05,
+      "loss": 0.9848,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 63.85717963058837,
+      "learning_rate": 1.9896013864818026e-05,
+      "loss": 1.1686,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 46.758520323989586,
+      "learning_rate": 1.9930675909878686e-05,
+      "loss": 0.9264,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 58.92162762986776,
+      "learning_rate": 1.9965337954939342e-05,
+      "loss": 1.0197,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 49.744388575360745,
+      "learning_rate": 2e-05,
+      "loss": 1.1176,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 42.217362703674,
+      "learning_rate": 1.9999999857802926e-05,
+      "loss": 1.0283,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 71.21309949780895,
+      "learning_rate": 1.9999999431211706e-05,
+      "loss": 1.1918,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 50.69351685509998,
+      "learning_rate": 1.999999872022635e-05,
+      "loss": 1.0243,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 62.15486273121246,
+      "learning_rate": 1.9999997724846883e-05,
+      "loss": 1.0769,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 58.740661962511425,
+      "learning_rate": 1.9999996445073327e-05,
+      "loss": 1.1035,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 46.4017035726521,
+      "learning_rate": 1.9999994880905726e-05,
+      "loss": 1.0023,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 47.50697587137097,
+      "learning_rate": 1.9999993032344115e-05,
+      "loss": 1.0715,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 54.17867912118686,
+      "learning_rate": 1.9999990899388556e-05,
+      "loss": 0.9323,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 46.148569370673556,
+      "learning_rate": 1.9999988482039104e-05,
+      "loss": 1.1269,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 71.89146346905163,
+      "learning_rate": 1.999998578029583e-05,
+      "loss": 1.1846,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 59.11374436196252,
+      "learning_rate": 1.999998279415881e-05,
+      "loss": 1.124,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 49.27416492782256,
+      "learning_rate": 1.999997952362813e-05,
+      "loss": 1.0561,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 51.30880117617682,
+      "learning_rate": 1.999997596870388e-05,
+      "loss": 1.0388,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 48.64586351163638,
+      "learning_rate": 1.9999972129386165e-05,
+      "loss": 1.0892,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 66.59416416993541,
+      "learning_rate": 1.999996800567509e-05,
+      "loss": 1.0751,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 43.65759426005191,
+      "learning_rate": 1.999996359757078e-05,
+      "loss": 1.1354,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 47.93917610988283,
+      "learning_rate": 1.9999958905073352e-05,
+      "loss": 1.1209,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 45.74898717624224,
+      "learning_rate": 1.9999953928182942e-05,
+      "loss": 1.0866,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 49.822837561017444,
+      "learning_rate": 1.9999948666899695e-05,
+      "loss": 1.1233,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 75.58021693620238,
+      "learning_rate": 1.9999943121223753e-05,
+      "loss": 1.0774,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 54.56295185404521,
+      "learning_rate": 1.999993729115528e-05,
+      "loss": 1.0404,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 98.88310605373482,
+      "learning_rate": 1.999993117669444e-05,
+      "loss": 1.1002,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 50.86188642216582,
+      "learning_rate": 1.999992477784141e-05,
+      "loss": 1.2364,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 47.820575864827276,
+      "learning_rate": 1.999991809459637e-05,
+      "loss": 1.0348,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 49.95043562927293,
+      "learning_rate": 1.9999911126959503e-05,
+      "loss": 1.1337,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 57.35904654229834,
+      "learning_rate": 1.9999903874931017e-05,
+      "loss": 1.0447,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.92453609678296,
+      "learning_rate": 1.9999896338511117e-05,
+      "loss": 1.0645,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 55.43837618646121,
+      "learning_rate": 1.999988851770001e-05,
+      "loss": 1.1486,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 54.62743359170688,
+      "learning_rate": 1.9999880412497927e-05,
+      "loss": 1.0157,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 48.133752670975525,
+      "learning_rate": 1.9999872022905094e-05,
+      "loss": 1.1341,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 67.89514242052658,
+      "learning_rate": 1.9999863348921748e-05,
+      "loss": 1.1169,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 61.060705254221254,
+      "learning_rate": 1.999985439054814e-05,
+      "loss": 1.0205,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 47.48796156869539,
+      "learning_rate": 1.9999845147784526e-05,
+      "loss": 0.9331,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 56.75049681593087,
+      "learning_rate": 1.999983562063116e-05,
+      "loss": 0.9312,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 62.85088458504673,
+      "learning_rate": 1.999982580908832e-05,
+      "loss": 1.0815,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 100.86787752440355,
+      "learning_rate": 1.9999815713156285e-05,
+      "loss": 1.0571,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 61.37045153139247,
+      "learning_rate": 1.9999805332835344e-05,
+      "loss": 1.0698,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 50.78197195625619,
+      "learning_rate": 1.9999794668125784e-05,
+      "loss": 1.1001,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 60.26491957061978,
+      "learning_rate": 1.9999783719027913e-05,
+      "loss": 1.1052,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 54.284428509475084,
+      "learning_rate": 1.9999772485542048e-05,
+      "loss": 1.0877,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.61541639695681,
+      "learning_rate": 1.99997609676685e-05,
+      "loss": 0.9592,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 48.64106426536958,
+      "learning_rate": 1.99997491654076e-05,
+      "loss": 1.0144,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 48.91810439378172,
+      "learning_rate": 1.9999737078759684e-05,
+      "loss": 1.1304,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 43.143812095689135,
+      "learning_rate": 1.999972470772509e-05,
+      "loss": 1.1728,
+      "step": 621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 50.68297269760392,
+      "learning_rate": 1.999971205230418e-05,
+      "loss": 1.0645,
+      "step": 622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 64.7163805193997,
+      "learning_rate": 1.999969911249731e-05,
+      "loss": 1.0508,
+      "step": 623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.99453584682664,
+      "learning_rate": 1.9999685888304844e-05,
+      "loss": 1.149,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 49.29045603413668,
+      "learning_rate": 1.9999672379727165e-05,
+      "loss": 1.0843,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 71.60508864024432,
+      "learning_rate": 1.999965858676465e-05,
+      "loss": 1.1005,
+      "step": 626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.71331742442617,
+      "learning_rate": 1.9999644509417694e-05,
+      "loss": 1.1306,
+      "step": 627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 44.400099712981564,
+      "learning_rate": 1.99996301476867e-05,
+      "loss": 1.0973,
+      "step": 628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 52.943206254375525,
+      "learning_rate": 1.9999615501572073e-05,
+      "loss": 1.1409,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 59.22576981861789,
+      "learning_rate": 1.999960057107423e-05,
+      "loss": 1.1789,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 45.17573221074537,
+      "learning_rate": 1.9999585356193597e-05,
+      "loss": 1.0023,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 44.826109034088795,
+      "learning_rate": 1.9999569856930604e-05,
+      "loss": 0.9659,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 73.71165845525469,
+      "learning_rate": 1.9999554073285695e-05,
+      "loss": 1.0679,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 55.43520458419372,
+      "learning_rate": 1.999953800525932e-05,
+      "loss": 1.1424,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 59.38139340963854,
+      "learning_rate": 1.9999521652851933e-05,
+      "loss": 1.0445,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 40.409876667307444,
+      "learning_rate": 1.9999505016063998e-05,
+      "loss": 1.0902,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 74.17836981030409,
+      "learning_rate": 1.9999488094895992e-05,
+      "loss": 1.1416,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 56.317509562783286,
+      "learning_rate": 1.9999470889348394e-05,
+      "loss": 1.2164,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 54.34339906426737,
+      "learning_rate": 1.9999453399421692e-05,
+      "loss": 1.0834,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 56.052493594098515,
+      "learning_rate": 1.9999435625116383e-05,
+      "loss": 1.0671,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 56.9444154419347,
+      "learning_rate": 1.999941756643298e-05,
+      "loss": 1.0352,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 48.32263292000576,
+      "learning_rate": 1.9999399223371988e-05,
+      "loss": 1.105,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 50.76798935747381,
+      "learning_rate": 1.999938059593393e-05,
+      "loss": 0.9312,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 48.41877872086541,
+      "learning_rate": 1.999936168411934e-05,
+      "loss": 1.1202,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 52.80356447369172,
+      "learning_rate": 1.999934248792875e-05,
+      "loss": 1.0671,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 43.43779075580135,
+      "learning_rate": 1.9999323007362708e-05,
+      "loss": 1.0992,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 55.049440634342986,
+      "learning_rate": 1.9999303242421773e-05,
+      "loss": 1.0443,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 52.40580164263194,
+      "learning_rate": 1.9999283193106504e-05,
+      "loss": 1.1782,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 57.50535978961775,
+      "learning_rate": 1.9999262859417466e-05,
+      "loss": 1.0435,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 66.89207123670141,
+      "learning_rate": 1.9999242241355247e-05,
+      "loss": 1.0089,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 62.11301218959004,
+      "learning_rate": 1.9999221338920424e-05,
+      "loss": 0.9694,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.63193811249696,
+      "learning_rate": 1.9999200152113595e-05,
+      "loss": 1.1497,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 55.08835327408598,
+      "learning_rate": 1.9999178680935366e-05,
+      "loss": 0.9823,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 97.78889757767074,
+      "learning_rate": 1.9999156925386344e-05,
+      "loss": 1.1106,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 68.65371696650708,
+      "learning_rate": 1.9999134885467148e-05,
+      "loss": 1.1113,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 61.242885425521315,
+      "learning_rate": 1.9999112561178402e-05,
+      "loss": 1.1231,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 56.79316380692482,
+      "learning_rate": 1.9999089952520746e-05,
+      "loss": 1.0548,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 64.8775715400972,
+      "learning_rate": 1.9999067059494824e-05,
+      "loss": 1.1202,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 59.74505186142053,
+      "learning_rate": 1.999904388210128e-05,
+      "loss": 1.1407,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 62.481827398030326,
+      "learning_rate": 1.9999020420340782e-05,
+      "loss": 1.1449,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.627043639736414,
+      "learning_rate": 1.999899667421399e-05,
+      "loss": 1.1385,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 57.78447397912724,
+      "learning_rate": 1.999897264372158e-05,
+      "loss": 1.0621,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 131.60369523115784,
+      "learning_rate": 1.999894832886424e-05,
+      "loss": 0.9783,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 50.21075260881517,
+      "learning_rate": 1.9998923729642657e-05,
+      "loss": 1.0613,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 45.580057076405446,
+      "learning_rate": 1.999889884605753e-05,
+      "loss": 0.9602,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 61.816623916604335,
+      "learning_rate": 1.9998873678109575e-05,
+      "loss": 1.1549,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 56.00650039478951,
+      "learning_rate": 1.99988482257995e-05,
+      "loss": 1.1319,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 50.73152115493126,
+      "learning_rate": 1.9998822489128028e-05,
+      "loss": 1.017,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 50.19608902742463,
+      "learning_rate": 1.9998796468095897e-05,
+      "loss": 1.031,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 59.804773571069575,
+      "learning_rate": 1.999877016270384e-05,
+      "loss": 1.1238,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 51.74200473518699,
+      "learning_rate": 1.9998743572952608e-05,
+      "loss": 1.1626,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 52.19232178301816,
+      "learning_rate": 1.999871669884296e-05,
+      "loss": 1.1516,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.49294498877882,
+      "learning_rate": 1.999868954037566e-05,
+      "loss": 1.1215,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.470632725960876,
+      "learning_rate": 1.9998662097551475e-05,
+      "loss": 1.1799,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 45.00047891679867,
+      "learning_rate": 1.9998634370371192e-05,
+      "loss": 1.0268,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 65.41468059740274,
+      "learning_rate": 1.9998606358835596e-05,
+      "loss": 1.0754,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.74724039837633,
+      "learning_rate": 1.9998578062945483e-05,
+      "loss": 1.0079,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 57.7142718301956,
+      "learning_rate": 1.999854948270166e-05,
+      "loss": 1.0388,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 50.010796377907944,
+      "learning_rate": 1.999852061810494e-05,
+      "loss": 1.2304,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 44.60804617217047,
+      "learning_rate": 1.9998491469156137e-05,
+      "loss": 1.0375,
+      "step": 680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 73.72861125298857,
+      "learning_rate": 1.999846203585609e-05,
+      "loss": 1.1611,
+      "step": 681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 54.180358152082334,
+      "learning_rate": 1.9998432318205632e-05,
+      "loss": 1.0537,
+      "step": 682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 47.1919785603725,
+      "learning_rate": 1.9998402316205606e-05,
+      "loss": 1.1537,
+      "step": 683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 58.60782457913101,
+      "learning_rate": 1.9998372029856866e-05,
+      "loss": 1.0171,
+      "step": 684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 49.95952076756143,
+      "learning_rate": 1.9998341459160277e-05,
+      "loss": 0.9724,
+      "step": 685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.27533148923914,
+      "learning_rate": 1.9998310604116704e-05,
+      "loss": 1.16,
+      "step": 686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 40.44054476847412,
+      "learning_rate": 1.999827946472703e-05,
+      "loss": 0.8918,
+      "step": 687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.09491315445135,
+      "learning_rate": 1.999824804099213e-05,
+      "loss": 1.0876,
+      "step": 688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 49.78652140742441,
+      "learning_rate": 1.9998216332912908e-05,
+      "loss": 1.0696,
+      "step": 689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 57.809859888777694,
+      "learning_rate": 1.9998184340490264e-05,
+      "loss": 1.1261,
+      "step": 690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 44.611173432503115,
+      "learning_rate": 1.9998152063725107e-05,
+      "loss": 1.033,
+      "step": 691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 53.16297484328663,
+      "learning_rate": 1.9998119502618353e-05,
+      "loss": 1.2342,
+      "step": 692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 43.9677882465517,
+      "learning_rate": 1.9998086657170925e-05,
+      "loss": 0.9216,
+      "step": 693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 54.83568701963368,
+      "learning_rate": 1.9998053527383766e-05,
+      "loss": 1.0804,
+      "step": 694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.52328569264612,
+      "learning_rate": 1.999802011325781e-05,
+      "loss": 1.1232,
+      "step": 695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 45.85794454549921,
+      "learning_rate": 1.9997986414794012e-05,
+      "loss": 1.0023,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 53.336959862594654,
+      "learning_rate": 1.999795243199333e-05,
+      "loss": 1.0411,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.88742439670702,
+      "learning_rate": 1.9997918164856728e-05,
+      "loss": 1.0718,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 58.46267207008015,
+      "learning_rate": 1.9997883613385184e-05,
+      "loss": 0.9944,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 56.604563635748164,
+      "learning_rate": 1.999784877757968e-05,
+      "loss": 1.1053,
+      "step": 700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 85.70988417393619,
+      "learning_rate": 1.99978136574412e-05,
+      "loss": 1.0121,
+      "step": 701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 54.95901070303401,
+      "learning_rate": 1.999777825297075e-05,
+      "loss": 1.2434,
+      "step": 702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 53.558427938533356,
+      "learning_rate": 1.9997742564169335e-05,
+      "loss": 1.0634,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 50.24475477371556,
+      "learning_rate": 1.999770659103797e-05,
+      "loss": 1.0232,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 60.65317505201505,
+      "learning_rate": 1.999767033357768e-05,
+      "loss": 1.1675,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 54.56604878091885,
+      "learning_rate": 1.9997633791789496e-05,
+      "loss": 1.1019,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.32255035273793,
+      "learning_rate": 1.999759696567445e-05,
+      "loss": 1.0834,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 60.256424945141696,
+      "learning_rate": 1.9997559855233596e-05,
+      "loss": 0.9966,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 48.42338583781208,
+      "learning_rate": 1.999752246046799e-05,
+      "loss": 1.0484,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.36502982245758,
+      "learning_rate": 1.9997484781378694e-05,
+      "loss": 1.124,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 61.31694291142357,
+      "learning_rate": 1.9997446817966776e-05,
+      "loss": 1.2816,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 57.09464931353948,
+      "learning_rate": 1.9997408570233322e-05,
+      "loss": 1.0708,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 50.76834658967656,
+      "learning_rate": 1.9997370038179414e-05,
+      "loss": 1.1124,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 57.31933865779926,
+      "learning_rate": 1.9997331221806152e-05,
+      "loss": 1.1026,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 42.29738802928118,
+      "learning_rate": 1.999729212111464e-05,
+      "loss": 1.0083,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 55.31380748604328,
+      "learning_rate": 1.9997252736105985e-05,
+      "loss": 1.0817,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 50.83680390668594,
+      "learning_rate": 1.9997213066781312e-05,
+      "loss": 0.9712,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.855582276802366,
+      "learning_rate": 1.9997173113141747e-05,
+      "loss": 1.0761,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.18014404170258,
+      "learning_rate": 1.9997132875188427e-05,
+      "loss": 1.1169,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 49.98949882858944,
+      "learning_rate": 1.9997092352922495e-05,
+      "loss": 1.0751,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 45.42246244645769,
+      "learning_rate": 1.999705154634511e-05,
+      "loss": 1.0002,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 56.92035926226957,
+      "learning_rate": 1.999701045545742e-05,
+      "loss": 1.2117,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 88.25348788276938,
+      "learning_rate": 1.9996969080260604e-05,
+      "loss": 1.2515,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 42.86090001735319,
+      "learning_rate": 1.999692742075584e-05,
+      "loss": 0.8993,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 45.05055304895923,
+      "learning_rate": 1.99968854769443e-05,
+      "loss": 1.0214,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.78925993124397,
+      "learning_rate": 1.999684324882719e-05,
+      "loss": 1.0741,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 49.6400527032577,
+      "learning_rate": 1.9996800736405702e-05,
+      "loss": 1.0193,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 55.25248064539535,
+      "learning_rate": 1.9996757939681052e-05,
+      "loss": 1.0555,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 62.24699964531423,
+      "learning_rate": 1.9996714858654456e-05,
+      "loss": 1.072,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 65.12508857999258,
+      "learning_rate": 1.9996671493327135e-05,
+      "loss": 1.0807,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 46.72772834266097,
+      "learning_rate": 1.9996627843700325e-05,
+      "loss": 0.9566,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 42.24151342131756,
+      "learning_rate": 1.9996583909775267e-05,
+      "loss": 1.0945,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 53.07124056229958,
+      "learning_rate": 1.999653969155321e-05,
+      "loss": 1.1053,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.48518029810589,
+      "learning_rate": 1.999649518903541e-05,
+      "loss": 1.0469,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 49.2598685190194,
+      "learning_rate": 1.9996450402223137e-05,
+      "loss": 1.0564,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 45.77019861041953,
+      "learning_rate": 1.9996405331117662e-05,
+      "loss": 0.9993,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.141358149794996,
+      "learning_rate": 1.999635997572027e-05,
+      "loss": 1.0058,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 49.994847117977216,
+      "learning_rate": 1.9996314336032243e-05,
+      "loss": 1.0809,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 48.68114165483692,
+      "learning_rate": 1.9996268412054887e-05,
+      "loss": 0.9446,
+      "step": 739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 45.536679092786045,
+      "learning_rate": 1.9996222203789504e-05,
+      "loss": 1.0578,
+      "step": 740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 51.55057347012913,
+      "learning_rate": 1.9996175711237406e-05,
+      "loss": 1.0068,
+      "step": 741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 44.23702373544687,
+      "learning_rate": 1.9996128934399923e-05,
+      "loss": 1.0332,
+      "step": 742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 52.547859684035146,
+      "learning_rate": 1.999608187327838e-05,
+      "loss": 0.9175,
+      "step": 743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 56.95853833503732,
+      "learning_rate": 1.9996034527874117e-05,
+      "loss": 1.0215,
+      "step": 744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 41.52533884513077,
+      "learning_rate": 1.9995986898188477e-05,
+      "loss": 0.9547,
+      "step": 745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.91923997754541,
+      "learning_rate": 1.9995938984222815e-05,
+      "loss": 1.0344,
+      "step": 746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 48.83820521438902,
+      "learning_rate": 1.99958907859785e-05,
+      "loss": 1.0562,
+      "step": 747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 53.59327859921458,
+      "learning_rate": 1.99958423034569e-05,
+      "loss": 0.987,
+      "step": 748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.6299317643339,
+      "learning_rate": 1.9995793536659388e-05,
+      "loss": 1.0349,
+      "step": 749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.355119870561346,
+      "learning_rate": 1.9995744485587356e-05,
+      "loss": 0.992,
+      "step": 750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 56.4057235862282,
+      "learning_rate": 1.99956951502422e-05,
+      "loss": 0.9747,
+      "step": 751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 43.8748683221042,
+      "learning_rate": 1.999564553062532e-05,
+      "loss": 1.0944,
+      "step": 752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 44.340766741868656,
+      "learning_rate": 1.9995595626738128e-05,
+      "loss": 0.9452,
+      "step": 753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 59.12313436585067,
+      "learning_rate": 1.9995545438582044e-05,
+      "loss": 0.9269,
+      "step": 754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 52.959167856063544,
+      "learning_rate": 1.9995494966158494e-05,
+      "loss": 1.1396,
+      "step": 755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 52.43981466538466,
+      "learning_rate": 1.9995444209468916e-05,
+      "loss": 1.1972,
+      "step": 756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 42.08219057678129,
+      "learning_rate": 1.999539316851475e-05,
+      "loss": 1.1093,
+      "step": 757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 44.03398046968544,
+      "learning_rate": 1.999534184329745e-05,
+      "loss": 1.0752,
+      "step": 758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 97.27493577384935,
+      "learning_rate": 1.9995290233818475e-05,
+      "loss": 1.0183,
+      "step": 759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 45.97042945267882,
+      "learning_rate": 1.9995238340079295e-05,
+      "loss": 0.9786,
+      "step": 760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 45.184408485742644,
+      "learning_rate": 1.9995186162081384e-05,
+      "loss": 1.0223,
+      "step": 761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 54.74564433952792,
+      "learning_rate": 1.9995133699826222e-05,
+      "loss": 1.0523,
+      "step": 762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 55.51063246734853,
+      "learning_rate": 1.999508095331531e-05,
+      "loss": 1.0252,
+      "step": 763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 53.790094400895,
+      "learning_rate": 1.9995027922550137e-05,
+      "loss": 1.0411,
+      "step": 764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 56.801628549829275,
+      "learning_rate": 1.999497460753222e-05,
+      "loss": 1.1387,
+      "step": 765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 49.423448921683004,
+      "learning_rate": 1.9994921008263072e-05,
+      "loss": 0.9896,
+      "step": 766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 53.07634833424985,
+      "learning_rate": 1.9994867124744216e-05,
+      "loss": 0.9209,
+      "step": 767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 51.91497049678569,
+      "learning_rate": 1.9994812956977183e-05,
+      "loss": 1.1652,
+      "step": 768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 69.22657645866512,
+      "learning_rate": 1.9994758504963522e-05,
+      "loss": 1.0012,
+      "step": 769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 52.985000642656075,
+      "learning_rate": 1.9994703768704773e-05,
+      "loss": 1.0373,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 56.12732607602791,
+      "learning_rate": 1.9994648748202493e-05,
+      "loss": 1.0026,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 73.36050310171413,
+      "learning_rate": 1.9994593443458252e-05,
+      "loss": 1.2432,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.426070879785854,
+      "learning_rate": 1.999453785447362e-05,
+      "loss": 0.9994,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 70.6292674843033,
+      "learning_rate": 1.999448198125018e-05,
+      "loss": 1.0381,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 55.22229142806187,
+      "learning_rate": 1.9994425823789517e-05,
+      "loss": 0.9702,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 53.905213069064764,
+      "learning_rate": 1.999436938209323e-05,
+      "loss": 0.9368,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 49.12292078097634,
+      "learning_rate": 1.9994312656162928e-05,
+      "loss": 1.0411,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 49.48068421563035,
+      "learning_rate": 1.9994255646000217e-05,
+      "loss": 0.9903,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 55.72997715535938,
+      "learning_rate": 1.999419835160672e-05,
+      "loss": 1.1691,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 54.4143409168709,
+      "learning_rate": 1.999414077298407e-05,
+      "loss": 0.9732,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 53.58300322949585,
+      "learning_rate": 1.9994082910133903e-05,
+      "loss": 1.1937,
+      "step": 781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 48.848846945727324,
+      "learning_rate": 1.9994024763057865e-05,
+      "loss": 0.9588,
+      "step": 782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 62.38941281442987,
+      "learning_rate": 1.9993966331757607e-05,
+      "loss": 1.0121,
+      "step": 783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 50.473928099686745,
+      "learning_rate": 1.9993907616234796e-05,
+      "loss": 1.0922,
+      "step": 784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 47.3407901254139,
+      "learning_rate": 1.9993848616491097e-05,
+      "loss": 1.065,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 55.233678305628146,
+      "learning_rate": 1.9993789332528193e-05,
+      "loss": 1.2466,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 49.535528254535365,
+      "learning_rate": 1.9993729764347763e-05,
+      "loss": 1.0845,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 43.59179400359375,
+      "learning_rate": 1.9993669911951504e-05,
+      "loss": 0.9879,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 51.12202028950768,
+      "learning_rate": 1.999360977534112e-05,
+      "loss": 1.0698,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.83792673552346,
+      "learning_rate": 1.999354935451832e-05,
+      "loss": 1.0365,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 42.64568867990435,
+      "learning_rate": 1.999348864948482e-05,
+      "loss": 0.9681,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 39.34019349910234,
+      "learning_rate": 1.9993427660242356e-05,
+      "loss": 0.947,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 46.12978304122491,
+      "learning_rate": 1.999336638679265e-05,
+      "loss": 1.1253,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 45.15348888786267,
+      "learning_rate": 1.999330482913745e-05,
+      "loss": 0.9507,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 52.733120391615095,
+      "learning_rate": 1.9993242987278508e-05,
+      "loss": 1.043,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 45.61452793496336,
+      "learning_rate": 1.999318086121758e-05,
+      "loss": 1.0698,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 58.24089227946547,
+      "learning_rate": 1.9993118450956434e-05,
+      "loss": 1.1031,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 48.48422077435949,
+      "learning_rate": 1.9993055756496845e-05,
+      "loss": 1.1262,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 54.634223072067904,
+      "learning_rate": 1.9992992777840596e-05,
+      "loss": 1.169,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 58.96262343965718,
+      "learning_rate": 1.999292951498948e-05,
+      "loss": 1.0665,
+      "step": 800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 57.06008366193231,
+      "learning_rate": 1.9992865967945295e-05,
+      "loss": 0.9819,
+      "step": 801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 44.723957230431054,
+      "learning_rate": 1.9992802136709842e-05,
+      "loss": 0.9565,
+      "step": 802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 50.890028067236436,
+      "learning_rate": 1.999273802128495e-05,
+      "loss": 1.0403,
+      "step": 803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 52.32342137280578,
+      "learning_rate": 1.9992673621672427e-05,
+      "loss": 1.0842,
+      "step": 804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 66.76724888964642,
+      "learning_rate": 1.9992608937874115e-05,
+      "loss": 1.1647,
+      "step": 805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 48.18292523350588,
+      "learning_rate": 1.999254396989185e-05,
+      "loss": 0.9446,
+      "step": 806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 51.20844959352403,
+      "learning_rate": 1.9992478717727478e-05,
+      "loss": 0.9596,
+      "step": 807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.4648884919621,
+      "learning_rate": 1.999241318138286e-05,
+      "loss": 1.0731,
+      "step": 808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 69.19878945945595,
+      "learning_rate": 1.9992347360859858e-05,
+      "loss": 1.026,
+      "step": 809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.7705128860658,
+      "learning_rate": 1.9992281256160337e-05,
+      "loss": 1.1192,
+      "step": 810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 69.02203692411481,
+      "learning_rate": 1.9992214867286182e-05,
+      "loss": 1.1161,
+      "step": 811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 50.212537612692024,
+      "learning_rate": 1.9992148194239287e-05,
+      "loss": 0.9234,
+      "step": 812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 57.13641050639885,
+      "learning_rate": 1.999208123702154e-05,
+      "loss": 1.1298,
+      "step": 813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 78.05817182039625,
+      "learning_rate": 1.999201399563485e-05,
+      "loss": 1.0392,
+      "step": 814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 69.65140409457796,
+      "learning_rate": 1.9991946470081124e-05,
+      "loss": 1.2363,
+      "step": 815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 60.277294268429266,
+      "learning_rate": 1.9991878660362285e-05,
+      "loss": 1.0996,
+      "step": 816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 52.0042850993777,
+      "learning_rate": 1.9991810566480264e-05,
+      "loss": 1.108,
+      "step": 817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 76.14006980100199,
+      "learning_rate": 1.9991742188436992e-05,
+      "loss": 0.9796,
+      "step": 818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 63.76323048050353,
+      "learning_rate": 1.999167352623442e-05,
+      "loss": 1.0674,
+      "step": 819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 52.03342154155507,
+      "learning_rate": 1.99916045798745e-05,
+      "loss": 1.09,
+      "step": 820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 51.49288438971617,
+      "learning_rate": 1.999153534935919e-05,
+      "loss": 1.041,
+      "step": 821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 54.994166095146674,
+      "learning_rate": 1.999146583469046e-05,
+      "loss": 1.1679,
+      "step": 822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 52.31997221673292,
+      "learning_rate": 1.9991396035870282e-05,
+      "loss": 1.1245,
+      "step": 823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 63.23923708446479,
+      "learning_rate": 1.999132595290065e-05,
+      "loss": 0.9847,
+      "step": 824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 51.134191297762875,
+      "learning_rate": 1.9991255585783547e-05,
+      "loss": 1.1322,
+      "step": 825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 48.44142059702477,
+      "learning_rate": 1.9991184934520987e-05,
+      "loss": 1.0283,
+      "step": 826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 42.23819147239972,
+      "learning_rate": 1.9991113999114966e-05,
+      "loss": 1.0212,
+      "step": 827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 47.0703553765446,
+      "learning_rate": 1.9991042779567512e-05,
+      "loss": 1.0209,
+      "step": 828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 55.195627851725966,
+      "learning_rate": 1.9990971275880643e-05,
+      "loss": 1.0853,
+      "step": 829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 41.59535033072864,
+      "learning_rate": 1.9990899488056396e-05,
+      "loss": 1.0403,
+      "step": 830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 48.449475910907076,
+      "learning_rate": 1.9990827416096813e-05,
+      "loss": 1.1135,
+      "step": 831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 53.19852512545227,
+      "learning_rate": 1.9990755060003945e-05,
+      "loss": 1.0807,
+      "step": 832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 61.11915146945207,
+      "learning_rate": 1.9990682419779844e-05,
+      "loss": 1.1779,
+      "step": 833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 46.476215273402794,
+      "learning_rate": 1.9990609495426582e-05,
+      "loss": 0.9788,
+      "step": 834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 62.22171973321687,
+      "learning_rate": 1.9990536286946227e-05,
+      "loss": 1.0188,
+      "step": 835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 50.573588877596414,
+      "learning_rate": 1.9990462794340864e-05,
+      "loss": 0.9366,
+      "step": 836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 48.94710001805791,
+      "learning_rate": 1.9990389017612587e-05,
+      "loss": 0.9967,
+      "step": 837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 43.55896824075458,
+      "learning_rate": 1.999031495676349e-05,
+      "loss": 1.1024,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 46.36679143310661,
+      "learning_rate": 1.999024061179568e-05,
+      "loss": 1.166,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.98453208282114,
+      "learning_rate": 1.9990165982711266e-05,
+      "loss": 1.1018,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.224257085717596,
+      "learning_rate": 1.999009106951238e-05,
+      "loss": 0.9423,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 48.16283375208339,
+      "learning_rate": 1.9990015872201147e-05,
+      "loss": 1.0838,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 50.04969479051453,
+      "learning_rate": 1.998994039077971e-05,
+      "loss": 0.9898,
+      "step": 843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 60.73763284270273,
+      "learning_rate": 1.9989864625250206e-05,
+      "loss": 1.1159,
+      "step": 844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 68.88442418446658,
+      "learning_rate": 1.99897885756148e-05,
+      "loss": 1.058,
+      "step": 845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 49.32433798761895,
+      "learning_rate": 1.998971224187565e-05,
+      "loss": 0.973,
+      "step": 846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 50.69798206173334,
+      "learning_rate": 1.998963562403493e-05,
+      "loss": 1.0132,
+      "step": 847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 77.04813048701959,
+      "learning_rate": 1.9989558722094813e-05,
+      "loss": 0.9006,
+      "step": 848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.263268332714354,
+      "learning_rate": 1.998948153605749e-05,
+      "loss": 0.9541,
+      "step": 849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 61.55474435198691,
+      "learning_rate": 1.9989404065925157e-05,
+      "loss": 1.1146,
+      "step": 850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 51.21208388196974,
+      "learning_rate": 1.9989326311700016e-05,
+      "loss": 1.1428,
+      "step": 851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 43.05719932077738,
+      "learning_rate": 1.9989248273384278e-05,
+      "loss": 1.0161,
+      "step": 852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 52.91484119490579,
+      "learning_rate": 1.998916995098016e-05,
+      "loss": 0.952,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 42.34960366078261,
+      "learning_rate": 1.9989091344489896e-05,
+      "loss": 0.9822,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 50.888671739132064,
+      "learning_rate": 1.9989012453915718e-05,
+      "loss": 1.1415,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 45.65762887866027,
+      "learning_rate": 1.998893327925987e-05,
+      "loss": 1.0797,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 57.042479170963354,
+      "learning_rate": 1.9988853820524596e-05,
+      "loss": 1.0806,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 51.45173809500342,
+      "learning_rate": 1.998877407771217e-05,
+      "loss": 1.1554,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 63.23317064137997,
+      "learning_rate": 1.9988694050824847e-05,
+      "loss": 1.1479,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 61.1414369821188,
+      "learning_rate": 1.998861373986491e-05,
+      "loss": 0.9716,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 48.693645329412476,
+      "learning_rate": 1.9988533144834642e-05,
+      "loss": 1.0532,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 59.11252263940518,
+      "learning_rate": 1.9988452265736335e-05,
+      "loss": 1.1781,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 44.85691259648733,
+      "learning_rate": 1.998837110257229e-05,
+      "loss": 1.0763,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 46.98880056096952,
+      "learning_rate": 1.9988289655344814e-05,
+      "loss": 0.9822,
+      "step": 864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 38.4876570361094,
+      "learning_rate": 1.998820792405622e-05,
+      "loss": 0.9233,
+      "step": 865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 50.59968597103575,
+      "learning_rate": 1.998812590870884e-05,
+      "loss": 1.0074,
+      "step": 866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.00120746022949,
+      "learning_rate": 1.9988043609304995e-05,
+      "loss": 0.9882,
+      "step": 867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 52.66280517978647,
+      "learning_rate": 1.9987961025847037e-05,
+      "loss": 1.0588,
+      "step": 868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 60.46127883532832,
+      "learning_rate": 1.998787815833731e-05,
+      "loss": 1.2107,
+      "step": 869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 48.72768626982206,
+      "learning_rate": 1.9987795006778173e-05,
+      "loss": 0.9928,
+      "step": 870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 58.07984174233936,
+      "learning_rate": 1.9987711571171987e-05,
+      "loss": 1.0639,
+      "step": 871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 51.23054097759872,
+      "learning_rate": 1.9987627851521126e-05,
+      "loss": 1.1361,
+      "step": 872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 55.339919196243216,
+      "learning_rate": 1.998754384782797e-05,
+      "loss": 1.0397,
+      "step": 873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 49.17839556899151,
+      "learning_rate": 1.998745956009491e-05,
+      "loss": 1.1328,
+      "step": 874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 50.5874006588514,
+      "learning_rate": 1.9987374988324345e-05,
+      "loss": 1.1287,
+      "step": 875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 49.776326407171325,
+      "learning_rate": 1.9987290132518677e-05,
+      "loss": 1.0719,
+      "step": 876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 56.04033350251024,
+      "learning_rate": 1.9987204992680322e-05,
+      "loss": 1.041,
+      "step": 877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 53.34224695234613,
+      "learning_rate": 1.9987119568811693e-05,
+      "loss": 1.1436,
+      "step": 878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.459777129907685,
+      "learning_rate": 1.998703386091523e-05,
+      "loss": 1.1862,
+      "step": 879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 40.732053285960056,
+      "learning_rate": 1.9986947868993366e-05,
+      "loss": 0.8659,
+      "step": 880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.77942713547777,
+      "learning_rate": 1.998686159304855e-05,
+      "loss": 0.9167,
+      "step": 881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 71.42219379481587,
+      "learning_rate": 1.9986775033083225e-05,
+      "loss": 1.1671,
+      "step": 882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 51.21035120531687,
+      "learning_rate": 1.9986688189099865e-05,
+      "loss": 1.063,
+      "step": 883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.53905379093711,
+      "learning_rate": 1.9986601061100934e-05,
+      "loss": 0.9976,
+      "step": 884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 45.72338666369144,
+      "learning_rate": 1.9986513649088915e-05,
+      "loss": 1.0267,
+      "step": 885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 40.61034822783893,
+      "learning_rate": 1.9986425953066284e-05,
+      "loss": 0.9512,
+      "step": 886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 49.150058737183215,
+      "learning_rate": 1.9986337973035542e-05,
+      "loss": 1.1123,
+      "step": 887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 45.01273629831772,
+      "learning_rate": 1.9986249708999194e-05,
+      "loss": 1.0221,
+      "step": 888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 52.22449206154985,
+      "learning_rate": 1.998616116095974e-05,
+      "loss": 1.039,
+      "step": 889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 53.20238250796886,
+      "learning_rate": 1.9986072328919707e-05,
+      "loss": 1.048,
+      "step": 890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 55.89150457673888,
+      "learning_rate": 1.9985983212881618e-05,
+      "loss": 1.0224,
+      "step": 891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 103.56394875525463,
+      "learning_rate": 1.9985893812848007e-05,
+      "loss": 0.9321,
+      "step": 892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 47.7284679897305,
+      "learning_rate": 1.998580412882142e-05,
+      "loss": 1.0732,
+      "step": 893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 67.12566688675167,
+      "learning_rate": 1.9985714160804406e-05,
+      "loss": 1.0596,
+      "step": 894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 57.59990704764045,
+      "learning_rate": 1.9985623908799518e-05,
+      "loss": 0.9747,
+      "step": 895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.699846884580616,
+      "learning_rate": 1.9985533372809335e-05,
+      "loss": 0.9969,
+      "step": 896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 58.01884174136258,
+      "learning_rate": 1.9985442552836417e-05,
+      "loss": 1.0444,
+      "step": 897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.8368467877389,
+      "learning_rate": 1.9985351448883358e-05,
+      "loss": 1.0567,
+      "step": 898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 47.927092918301625,
+      "learning_rate": 1.9985260060952742e-05,
+      "loss": 1.0539,
+      "step": 899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 47.458516733732296,
+      "learning_rate": 1.9985168389047175e-05,
+      "loss": 1.0127,
+      "step": 900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 56.18593143739028,
+      "learning_rate": 1.9985076433169258e-05,
+      "loss": 1.0726,
+      "step": 901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 65.66761506391391,
+      "learning_rate": 1.9984984193321608e-05,
+      "loss": 1.0147,
+      "step": 902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 41.173909227182676,
+      "learning_rate": 1.9984891669506847e-05,
+      "loss": 1.0252,
+      "step": 903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 49.435384640004145,
+      "learning_rate": 1.9984798861727614e-05,
+      "loss": 1.0651,
+      "step": 904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 53.48862799881216,
+      "learning_rate": 1.9984705769986538e-05,
+      "loss": 1.1711,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 59.84414761164673,
+      "learning_rate": 1.998461239428627e-05,
+      "loss": 1.1806,
+      "step": 906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 50.73778779746358,
+      "learning_rate": 1.9984518734629467e-05,
+      "loss": 0.9994,
+      "step": 907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 55.637121268607245,
+      "learning_rate": 1.998442479101879e-05,
+      "loss": 0.9514,
+      "step": 908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 44.297412525932195,
+      "learning_rate": 1.9984330563456918e-05,
+      "loss": 1.1765,
+      "step": 909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 47.296710706865916,
+      "learning_rate": 1.998423605194652e-05,
+      "loss": 1.086,
+      "step": 910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 42.207697214276926,
+      "learning_rate": 1.9984141256490294e-05,
+      "loss": 1.0336,
+      "step": 911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 50.83771565834528,
+      "learning_rate": 1.9984046177090926e-05,
+      "loss": 1.1062,
+      "step": 912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 48.37957390150111,
+      "learning_rate": 1.998395081375113e-05,
+      "loss": 1.0073,
+      "step": 913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 38.33110918696844,
+      "learning_rate": 1.9983855166473613e-05,
+      "loss": 1.0092,
+      "step": 914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 37.883023951171026,
+      "learning_rate": 1.998375923526109e-05,
+      "loss": 0.8824,
+      "step": 915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 43.340358389660764,
+      "learning_rate": 1.99836630201163e-05,
+      "loss": 0.951,
+      "step": 916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 41.6218608325459,
+      "learning_rate": 1.9983566521041973e-05,
+      "loss": 1.0903,
+      "step": 917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 65.13005219117588,
+      "learning_rate": 1.9983469738040852e-05,
+      "loss": 1.1635,
+      "step": 918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 42.29638667647359,
+      "learning_rate": 1.9983372671115696e-05,
+      "loss": 1.0074,
+      "step": 919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 46.30113888590107,
+      "learning_rate": 1.998327532026926e-05,
+      "loss": 1.034,
+      "step": 920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 53.10481898986034,
+      "learning_rate": 1.9983177685504312e-05,
+      "loss": 1.01,
+      "step": 921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 51.79756002534152,
+      "learning_rate": 1.998307976682363e-05,
+      "loss": 1.019,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 40.354138269777145,
+      "learning_rate": 1.998298156423e-05,
+      "loss": 0.9117,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 47.64774098960683,
+      "learning_rate": 1.9982883077726218e-05,
+      "loss": 0.937,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 52.14458192895296,
+      "learning_rate": 1.998278430731508e-05,
+      "loss": 1.1073,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 55.53771175985514,
+      "learning_rate": 1.9982685252999393e-05,
+      "loss": 0.9779,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 64.01581435842444,
+      "learning_rate": 1.998258591478198e-05,
+      "loss": 0.9773,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 60.758319096748224,
+      "learning_rate": 1.998248629266566e-05,
+      "loss": 1.0596,
+      "step": 928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 42.37622664935177,
+      "learning_rate": 1.9982386386653273e-05,
+      "loss": 1.0293,
+      "step": 929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 39.19900536053808,
+      "learning_rate": 1.9982286196747653e-05,
+      "loss": 0.9833,
+      "step": 930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 50.45827562983537,
+      "learning_rate": 1.9982185722951655e-05,
+      "loss": 0.9013,
+      "step": 931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 57.27341338065955,
+      "learning_rate": 1.9982084965268134e-05,
+      "loss": 0.9686,
+      "step": 932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.041567677819565,
+      "learning_rate": 1.9981983923699956e-05,
+      "loss": 1.0776,
+      "step": 933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 47.27274645916677,
+      "learning_rate": 1.9981882598249996e-05,
+      "loss": 1.2512,
+      "step": 934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.209416936603496,
+      "learning_rate": 1.998178098892113e-05,
+      "loss": 0.9825,
+      "step": 935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 57.46795520990007,
+      "learning_rate": 1.998167909571626e-05,
+      "loss": 1.1167,
+      "step": 936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 60.69330585652307,
+      "learning_rate": 1.998157691863827e-05,
+      "loss": 1.2467,
+      "step": 937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 52.57752672056269,
+      "learning_rate": 1.998147445769007e-05,
+      "loss": 0.9625,
+      "step": 938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.94957465428888,
+      "learning_rate": 1.9981371712874576e-05,
+      "loss": 1.1186,
+      "step": 939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.789586028132035,
+      "learning_rate": 1.9981268684194712e-05,
+      "loss": 0.9599,
+      "step": 940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 50.28807519426832,
+      "learning_rate": 1.9981165371653404e-05,
+      "loss": 1.0795,
+      "step": 941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 40.94922456065359,
+      "learning_rate": 1.998106177525359e-05,
+      "loss": 0.905,
+      "step": 942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 46.11732910692852,
+      "learning_rate": 1.998095789499822e-05,
+      "loss": 1.0997,
+      "step": 943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 35.59181894719201,
+      "learning_rate": 1.9980853730890248e-05,
+      "loss": 0.9456,
+      "step": 944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 43.69865738881624,
+      "learning_rate": 1.998074928293263e-05,
+      "loss": 0.9524,
+      "step": 945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 56.7877346822435,
+      "learning_rate": 1.9980644551128347e-05,
+      "loss": 1.0517,
+      "step": 946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 42.707036603764855,
+      "learning_rate": 1.9980539535480365e-05,
+      "loss": 1.0364,
+      "step": 947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 52.93071473648603,
+      "learning_rate": 1.998043423599168e-05,
+      "loss": 1.1098,
+      "step": 948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 51.14164820975276,
+      "learning_rate": 1.9980328652665285e-05,
+      "loss": 0.935,
+      "step": 949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 59.47987387972161,
+      "learning_rate": 1.9980222785504178e-05,
+      "loss": 1.038,
+      "step": 950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 56.236222444815965,
+      "learning_rate": 1.9980116634511377e-05,
+      "loss": 0.9936,
+      "step": 951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 57.53147762984995,
+      "learning_rate": 1.9980010199689893e-05,
+      "loss": 0.966,
+      "step": 952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 75.12805351741689,
+      "learning_rate": 1.997990348104276e-05,
+      "loss": 1.0727,
+      "step": 953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 45.60124209828804,
+      "learning_rate": 1.997979647857301e-05,
+      "loss": 0.9146,
+      "step": 954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 78.55434145934201,
+      "learning_rate": 1.9979689192283683e-05,
+      "loss": 1.0897,
+      "step": 955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 66.20002811004872,
+      "learning_rate": 1.997958162217784e-05,
+      "loss": 0.961,
+      "step": 956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 43.29700393685093,
+      "learning_rate": 1.9979473768258528e-05,
+      "loss": 0.9956,
+      "step": 957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 55.84589293498393,
+      "learning_rate": 1.9979365630528824e-05,
+      "loss": 0.9424,
+      "step": 958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 44.47624946991993,
+      "learning_rate": 1.9979257208991796e-05,
+      "loss": 0.9458,
+      "step": 959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 57.18571784989538,
+      "learning_rate": 1.9979148503650528e-05,
+      "loss": 0.9424,
+      "step": 960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 41.41040983213291,
+      "learning_rate": 1.997903951450812e-05,
+      "loss": 0.9566,
+      "step": 961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 60.60246143016136,
+      "learning_rate": 1.9978930241567662e-05,
+      "loss": 1.0425,
+      "step": 962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 50.632959460689875,
+      "learning_rate": 1.9978820684832268e-05,
+      "loss": 1.1282,
+      "step": 963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 53.537572929338,
+      "learning_rate": 1.997871084430505e-05,
+      "loss": 1.0258,
+      "step": 964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.61769140101144,
+      "learning_rate": 1.997860071998913e-05,
+      "loss": 0.8946,
+      "step": 965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 48.89987204404289,
+      "learning_rate": 1.997849031188765e-05,
+      "loss": 1.1147,
+      "step": 966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 46.462396573456274,
+      "learning_rate": 1.9978379620003737e-05,
+      "loss": 1.0214,
+      "step": 967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 46.10162991168277,
+      "learning_rate": 1.9978268644340545e-05,
+      "loss": 0.9521,
+      "step": 968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 34.74051504568104,
+      "learning_rate": 1.997815738490123e-05,
+      "loss": 0.8519,
+      "step": 969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 48.31901461955983,
+      "learning_rate": 1.997804584168896e-05,
+      "loss": 1.0389,
+      "step": 970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 46.336647369138895,
+      "learning_rate": 1.9977934014706898e-05,
+      "loss": 1.1637,
+      "step": 971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 52.056370212884666,
+      "learning_rate": 1.997782190395823e-05,
+      "loss": 1.1299,
+      "step": 972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 68.20705969648061,
+      "learning_rate": 1.9977709509446148e-05,
+      "loss": 1.0359,
+      "step": 973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 48.10377515939441,
+      "learning_rate": 1.9977596831173843e-05,
+      "loss": 1.0663,
+      "step": 974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 48.751950136940096,
+      "learning_rate": 1.9977483869144522e-05,
+      "loss": 1.0326,
+      "step": 975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 70.0772225176176,
+      "learning_rate": 1.9977370623361393e-05,
+      "loss": 1.0457,
+      "step": 976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 52.64841483484027,
+      "learning_rate": 1.997725709382768e-05,
+      "loss": 0.9518,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 44.74333593627468,
+      "learning_rate": 1.9977143280546616e-05,
+      "loss": 0.9626,
+      "step": 978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 41.96900382572316,
+      "learning_rate": 1.997702918352143e-05,
+      "loss": 1.0082,
+      "step": 979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 65.6887811160175,
+      "learning_rate": 1.997691480275537e-05,
+      "loss": 1.059,
+      "step": 980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 48.66880147423477,
+      "learning_rate": 1.997680013825169e-05,
+      "loss": 1.1688,
+      "step": 981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.801383129518356,
+      "learning_rate": 1.997668519001365e-05,
+      "loss": 0.9513,
+      "step": 982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 82.98621361918129,
+      "learning_rate": 1.9976569958044523e-05,
+      "loss": 0.99,
+      "step": 983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 44.64914875949154,
+      "learning_rate": 1.997645444234758e-05,
+      "loss": 0.9886,
+      "step": 984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 44.59129565193975,
+      "learning_rate": 1.997633864292611e-05,
+      "loss": 1.0099,
+      "step": 985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 54.46690212317148,
+      "learning_rate": 1.99762225597834e-05,
+      "loss": 1.1061,
+      "step": 986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 48.49792752311351,
+      "learning_rate": 1.997610619292276e-05,
+      "loss": 1.0872,
+      "step": 987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.46184511847755,
+      "learning_rate": 1.9975989542347495e-05,
+      "loss": 1.1565,
+      "step": 988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 56.19301900607758,
+      "learning_rate": 1.9975872608060927e-05,
+      "loss": 0.9392,
+      "step": 989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 44.89566710663342,
+      "learning_rate": 1.9975755390066373e-05,
+      "loss": 1.0791,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 46.691065068714984,
+      "learning_rate": 1.9975637888367173e-05,
+      "loss": 1.0049,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.64569744624319,
+      "learning_rate": 1.9975520102966667e-05,
+      "loss": 0.9819,
+      "step": 992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 48.43480783179659,
+      "learning_rate": 1.9975402033868207e-05,
+      "loss": 0.9299,
+      "step": 993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.36925177728487,
+      "learning_rate": 1.9975283681075142e-05,
+      "loss": 0.9699,
+      "step": 994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 159.99797409682375,
+      "learning_rate": 1.9975165044590853e-05,
+      "loss": 1.1648,
+      "step": 995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.81855658054716,
+      "learning_rate": 1.99750461244187e-05,
+      "loss": 0.9943,
+      "step": 996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 47.221770668987425,
+      "learning_rate": 1.997492692056207e-05,
+      "loss": 0.9228,
+      "step": 997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 94.99277851364188,
+      "learning_rate": 1.9974807433024357e-05,
+      "loss": 1.1267,
+      "step": 998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 54.49525119495747,
+      "learning_rate": 1.9974687661808954e-05,
+      "loss": 1.0564,
+      "step": 999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 49.22182744169017,
+      "learning_rate": 1.997456760691927e-05,
+      "loss": 0.9788,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 58.137521781416986,
+      "learning_rate": 1.9974447268358715e-05,
+      "loss": 0.945,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.87169932586954,
+      "learning_rate": 1.9974326646130716e-05,
+      "loss": 1.079,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 79.6170164264963,
+      "learning_rate": 1.9974205740238704e-05,
+      "loss": 1.0654,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 55.740189950781534,
+      "learning_rate": 1.9974084550686116e-05,
+      "loss": 1.0469,
+      "step": 1004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 48.97500454531187,
+      "learning_rate": 1.9973963077476394e-05,
+      "loss": 1.0308,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 55.043350730600935,
+      "learning_rate": 1.9973841320613e-05,
+      "loss": 1.1378,
+      "step": 1006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 44.51338956608709,
+      "learning_rate": 1.997371928009939e-05,
+      "loss": 0.9583,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.597000625523556,
+      "learning_rate": 1.9973596955939043e-05,
+      "loss": 1.0681,
+      "step": 1008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 54.77413485116718,
+      "learning_rate": 1.997347434813543e-05,
+      "loss": 1.0507,
+      "step": 1009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 64.73614031064832,
+      "learning_rate": 1.997335145669204e-05,
+      "loss": 1.0639,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 41.46808250597207,
+      "learning_rate": 1.997322828161237e-05,
+      "loss": 1.0312,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 41.94668562643936,
+      "learning_rate": 1.9973104822899923e-05,
+      "loss": 1.0885,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.8089214392629,
+      "learning_rate": 1.9972981080558208e-05,
+      "loss": 0.9579,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.471401306515574,
+      "learning_rate": 1.9972857054590744e-05,
+      "loss": 0.9885,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 57.079044596743444,
+      "learning_rate": 1.997273274500106e-05,
+      "loss": 1.1345,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 45.04125441869237,
+      "learning_rate": 1.997260815179269e-05,
+      "loss": 0.9919,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 83.83111617554651,
+      "learning_rate": 1.997248327496918e-05,
+      "loss": 1.1985,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 49.255555657666136,
+      "learning_rate": 1.9972358114534074e-05,
+      "loss": 1.0054,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 44.21768911052843,
+      "learning_rate": 1.9972232670490945e-05,
+      "loss": 0.9714,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 62.06329389757777,
+      "learning_rate": 1.997210694284335e-05,
+      "loss": 1.0212,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.02715191971834,
+      "learning_rate": 1.9971980931594865e-05,
+      "loss": 0.9579,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 58.90635171168119,
+      "learning_rate": 1.9971854636749075e-05,
+      "loss": 1.1057,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 54.811151085570124,
+      "learning_rate": 1.9971728058309572e-05,
+      "loss": 1.058,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 56.319516057121675,
+      "learning_rate": 1.9971601196279963e-05,
+      "loss": 0.9917,
+      "step": 1024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 55.94234567890532,
+      "learning_rate": 1.997147405066384e-05,
+      "loss": 1.078,
+      "step": 1025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 57.177930012408154,
+      "learning_rate": 1.9971346621464833e-05,
+      "loss": 1.1365,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 41.726692146652674,
+      "learning_rate": 1.9971218908686563e-05,
+      "loss": 0.9188,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 39.612457032088585,
+      "learning_rate": 1.9971090912332657e-05,
+      "loss": 0.9516,
+      "step": 1028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 37.70696069878167,
+      "learning_rate": 1.9970962632406757e-05,
+      "loss": 0.9549,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 49.97970235817626,
+      "learning_rate": 1.9970834068912515e-05,
+      "loss": 1.082,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 56.31455393780753,
+      "learning_rate": 1.9970705221853584e-05,
+      "loss": 1.0901,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 58.34608174074355,
+      "learning_rate": 1.9970576091233633e-05,
+      "loss": 1.0547,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.654492152009674,
+      "learning_rate": 1.9970446677056325e-05,
+      "loss": 1.1523,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 39.14083199512927,
+      "learning_rate": 1.997031697932535e-05,
+      "loss": 0.9386,
+      "step": 1034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 48.95650910029277,
+      "learning_rate": 1.9970186998044392e-05,
+      "loss": 0.9639,
+      "step": 1035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 45.69436885063518,
+      "learning_rate": 1.9970056733217147e-05,
+      "loss": 1.0111,
+      "step": 1036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 51.239074151826465,
+      "learning_rate": 1.996992618484732e-05,
+      "loss": 1.0132,
+      "step": 1037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.45037963161402,
+      "learning_rate": 1.9969795352938625e-05,
+      "loss": 1.0508,
+      "step": 1038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 59.413569822657585,
+      "learning_rate": 1.9969664237494785e-05,
+      "loss": 1.1118,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 52.06836663721922,
+      "learning_rate": 1.9969532838519524e-05,
+      "loss": 1.0625,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 44.23774628237862,
+      "learning_rate": 1.996940115601658e-05,
+      "loss": 0.8648,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 49.006782728894194,
+      "learning_rate": 1.99692691899897e-05,
+      "loss": 0.9439,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 40.49868699288389,
+      "learning_rate": 1.9969136940442636e-05,
+      "loss": 1.0013,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 45.77594925850887,
+      "learning_rate": 1.996900440737915e-05,
+      "loss": 1.0866,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 53.42410072405221,
+      "learning_rate": 1.996887159080301e-05,
+      "loss": 1.09,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 51.28670527088599,
+      "learning_rate": 1.9968738490717992e-05,
+      "loss": 1.036,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 43.15468826064668,
+      "learning_rate": 1.9968605107127886e-05,
+      "loss": 0.9358,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 45.168212821345215,
+      "learning_rate": 1.996847144003648e-05,
+      "loss": 0.9922,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 46.901764214891415,
+      "learning_rate": 1.9968337489447578e-05,
+      "loss": 1.0128,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 50.673851094633086,
+      "learning_rate": 1.996820325536499e-05,
+      "loss": 1.0743,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 49.661170874091916,
+      "learning_rate": 1.996806873779253e-05,
+      "loss": 1.1139,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 45.34856050135084,
+      "learning_rate": 1.9967933936734027e-05,
+      "loss": 1.011,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 42.76148206318273,
+      "learning_rate": 1.996779885219332e-05,
+      "loss": 0.9984,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 41.63498361244763,
+      "learning_rate": 1.9967663484174235e-05,
+      "loss": 0.9717,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 42.47043342425756,
+      "learning_rate": 1.9967527832680637e-05,
+      "loss": 1.0115,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 33.31616497097787,
+      "learning_rate": 1.996739189771638e-05,
+      "loss": 0.9004,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 48.331239143033926,
+      "learning_rate": 1.9967255679285324e-05,
+      "loss": 1.0282,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.29985256797121,
+      "learning_rate": 1.9967119177391348e-05,
+      "loss": 1.105,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 46.86552071663719,
+      "learning_rate": 1.9966982392038335e-05,
+      "loss": 1.0688,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 41.45474851659215,
+      "learning_rate": 1.996684532323017e-05,
+      "loss": 0.9381,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 41.79700765713469,
+      "learning_rate": 1.9966707970970756e-05,
+      "loss": 1.0517,
+      "step": 1061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 64.27144911196451,
+      "learning_rate": 1.9966570335264e-05,
+      "loss": 0.9866,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 49.50747080500346,
+      "learning_rate": 1.9966432416113812e-05,
+      "loss": 0.9907,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 37.20478856173655,
+      "learning_rate": 1.9966294213524117e-05,
+      "loss": 0.773,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.70357399512713,
+      "learning_rate": 1.9966155727498843e-05,
+      "loss": 1.1108,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 50.94572064113822,
+      "learning_rate": 1.9966016958041933e-05,
+      "loss": 1.1498,
+      "step": 1066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 38.21070425289566,
+      "learning_rate": 1.996587790515733e-05,
+      "loss": 0.8655,
+      "step": 1067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 63.406373250887576,
+      "learning_rate": 1.9965738568848986e-05,
+      "loss": 1.0695,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 51.625196401116256,
+      "learning_rate": 1.996559894912087e-05,
+      "loss": 1.0676,
+      "step": 1069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 41.19590882409231,
+      "learning_rate": 1.996545904597695e-05,
+      "loss": 1.0157,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 49.23370917283765,
+      "learning_rate": 1.9965318859421203e-05,
+      "loss": 1.0781,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 62.52696052415243,
+      "learning_rate": 1.9965178389457617e-05,
+      "loss": 1.0065,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 53.40466063820294,
+      "learning_rate": 1.9965037636090187e-05,
+      "loss": 1.0105,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 163.41185305734408,
+      "learning_rate": 1.9964896599322917e-05,
+      "loss": 0.918,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 39.43037529744758,
+      "learning_rate": 1.9964755279159816e-05,
+      "loss": 0.9538,
+      "step": 1075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 58.716609479331495,
+      "learning_rate": 1.9964613675604904e-05,
+      "loss": 1.0121,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 53.18135122669954,
+      "learning_rate": 1.996447178866221e-05,
+      "loss": 1.1668,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 45.75348634412362,
+      "learning_rate": 1.9964329618335766e-05,
+      "loss": 0.8998,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 42.80908632800003,
+      "learning_rate": 1.9964187164629617e-05,
+      "loss": 0.9189,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 83.85025619076447,
+      "learning_rate": 1.996404442754781e-05,
+      "loss": 1.0742,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 39.2673516187987,
+      "learning_rate": 1.9963901407094415e-05,
+      "loss": 0.9179,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 39.67873858364763,
+      "learning_rate": 1.996375810327349e-05,
+      "loss": 0.9519,
+      "step": 1082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 49.86885725641627,
+      "learning_rate": 1.9963614516089112e-05,
+      "loss": 1.2181,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 38.66874605101434,
+      "learning_rate": 1.9963470645545365e-05,
+      "loss": 0.9171,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 50.565916683911,
+      "learning_rate": 1.9963326491646346e-05,
+      "loss": 0.8943,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 92.0041978541769,
+      "learning_rate": 1.9963182054396144e-05,
+      "loss": 0.9634,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 48.247657007441966,
+      "learning_rate": 1.9963037333798876e-05,
+      "loss": 1.1396,
+      "step": 1087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 40.835842362346504,
+      "learning_rate": 1.9962892329858655e-05,
+      "loss": 1.0484,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 49.70728399965671,
+      "learning_rate": 1.99627470425796e-05,
+      "loss": 0.9392,
+      "step": 1089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 36.75808971920736,
+      "learning_rate": 1.9962601471965854e-05,
+      "loss": 1.0304,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 51.9059241719271,
+      "learning_rate": 1.9962455618021545e-05,
+      "loss": 1.0482,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.62835378177044,
+      "learning_rate": 1.9962309480750827e-05,
+      "loss": 1.0834,
+      "step": 1092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 63.74842777183401,
+      "learning_rate": 1.9962163060157854e-05,
+      "loss": 1.17,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.35468625702317,
+      "learning_rate": 1.9962016356246794e-05,
+      "loss": 0.9861,
+      "step": 1094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 43.033496551489776,
+      "learning_rate": 1.9961869369021814e-05,
+      "loss": 0.9312,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 50.64679430676675,
+      "learning_rate": 1.9961722098487098e-05,
+      "loss": 0.987,
+      "step": 1096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 40.11249140486442,
+      "learning_rate": 1.9961574544646835e-05,
+      "loss": 0.9508,
+      "step": 1097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 48.38904406114749,
+      "learning_rate": 1.9961426707505217e-05,
+      "loss": 0.9392,
+      "step": 1098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.12058574760634,
+      "learning_rate": 1.996127858706645e-05,
+      "loss": 0.9309,
+      "step": 1099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.30072598949605,
+      "learning_rate": 1.996113018333475e-05,
+      "loss": 1.0663,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 38.00579472955881,
+      "learning_rate": 1.996098149631433e-05,
+      "loss": 0.9169,
+      "step": 1101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 39.00971014234417,
+      "learning_rate": 1.9960832526009427e-05,
+      "loss": 1.0269,
+      "step": 1102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.82070988497923,
+      "learning_rate": 1.996068327242427e-05,
+      "loss": 0.9618,
+      "step": 1103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 57.65310206723226,
+      "learning_rate": 1.9960533735563113e-05,
+      "loss": 1.1099,
+      "step": 1104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 37.47073195483441,
+      "learning_rate": 1.99603839154302e-05,
+      "loss": 0.9502,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 40.006644990434474,
+      "learning_rate": 1.9960233812029798e-05,
+      "loss": 0.9209,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 64.27384518862331,
+      "learning_rate": 1.996008342536617e-05,
+      "loss": 0.9053,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 50.51133614954897,
+      "learning_rate": 1.9959932755443596e-05,
+      "loss": 1.0453,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 46.76107034830382,
+      "learning_rate": 1.995978180226636e-05,
+      "loss": 0.997,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 43.79964396079467,
+      "learning_rate": 1.9959630565838758e-05,
+      "loss": 0.8698,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.6197122705468,
+      "learning_rate": 1.9959479046165088e-05,
+      "loss": 1.022,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.22803829488919,
+      "learning_rate": 1.9959327243249662e-05,
+      "loss": 1.0052,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 67.95360105571007,
+      "learning_rate": 1.9959175157096794e-05,
+      "loss": 1.1459,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.817834598912064,
+      "learning_rate": 1.995902278771081e-05,
+      "loss": 1.0451,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 52.0799147157879,
+      "learning_rate": 1.9958870135096044e-05,
+      "loss": 0.9271,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 57.77337250309319,
+      "learning_rate": 1.9958717199256833e-05,
+      "loss": 0.9278,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 51.2169123070302,
+      "learning_rate": 1.9958563980197536e-05,
+      "loss": 1.0507,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 41.85858070386005,
+      "learning_rate": 1.9958410477922504e-05,
+      "loss": 1.0212,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 54.00232427504615,
+      "learning_rate": 1.9958256692436103e-05,
+      "loss": 1.0985,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 49.171597848795415,
+      "learning_rate": 1.995810262374271e-05,
+      "loss": 0.91,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.25841632948514,
+      "learning_rate": 1.99579482718467e-05,
+      "loss": 1.0888,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 50.1404486463185,
+      "learning_rate": 1.9957793636752466e-05,
+      "loss": 0.9886,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 57.24308545197199,
+      "learning_rate": 1.9957638718464405e-05,
+      "loss": 1.0527,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 46.210930380091106,
+      "learning_rate": 1.995748351698693e-05,
+      "loss": 1.0668,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 44.95743041358903,
+      "learning_rate": 1.995732803232444e-05,
+      "loss": 1.0227,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 55.480120036745475,
+      "learning_rate": 1.995717226448137e-05,
+      "loss": 1.0094,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.599940517035385,
+      "learning_rate": 1.995701621346215e-05,
+      "loss": 0.9754,
+      "step": 1127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 51.12471267507534,
+      "learning_rate": 1.9956859879271207e-05,
+      "loss": 0.9781,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.787547816887624,
+      "learning_rate": 1.9956703261913e-05,
+      "loss": 0.9478,
+      "step": 1129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 35.66267976952887,
+      "learning_rate": 1.995654636139197e-05,
+      "loss": 0.9195,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 44.702911163106634,
+      "learning_rate": 1.9956389177712592e-05,
+      "loss": 0.9012,
+      "step": 1131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 63.70318578811227,
+      "learning_rate": 1.9956231710879325e-05,
+      "loss": 0.9062,
+      "step": 1132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.96167560654539,
+      "learning_rate": 1.9956073960896652e-05,
+      "loss": 1.0013,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 44.16652304548028,
+      "learning_rate": 1.9955915927769066e-05,
+      "loss": 1.1142,
+      "step": 1134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 53.367510464438915,
+      "learning_rate": 1.995575761150105e-05,
+      "loss": 1.0526,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 41.30623465509024,
+      "learning_rate": 1.995559901209711e-05,
+      "loss": 0.9676,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 49.37295114153382,
+      "learning_rate": 1.995544012956176e-05,
+      "loss": 1.0915,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 40.17314883051957,
+      "learning_rate": 1.995528096389952e-05,
+      "loss": 0.9508,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 48.40758078650103,
+      "learning_rate": 1.995512151511491e-05,
+      "loss": 0.9279,
+      "step": 1139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 42.70452295130464,
+      "learning_rate": 1.995496178321247e-05,
+      "loss": 1.0611,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 50.70011194149052,
+      "learning_rate": 1.9954801768196738e-05,
+      "loss": 0.9834,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 48.361672598593074,
+      "learning_rate": 1.9954641470072267e-05,
+      "loss": 0.9343,
+      "step": 1142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.757915744857144,
+      "learning_rate": 1.9954480888843614e-05,
+      "loss": 1.0109,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.31061436379961,
+      "learning_rate": 1.995432002451535e-05,
+      "loss": 0.9841,
+      "step": 1144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 41.041336359673224,
+      "learning_rate": 1.995415887709205e-05,
+      "loss": 0.9745,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 58.31594203100104,
+      "learning_rate": 1.995399744657829e-05,
+      "loss": 1.1409,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.819601973466185,
+      "learning_rate": 1.9953835732978668e-05,
+      "loss": 1.0071,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 44.43360349416102,
+      "learning_rate": 1.995367373629778e-05,
+      "loss": 0.8967,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 39.07013234152445,
+      "learning_rate": 1.9953511456540233e-05,
+      "loss": 0.8965,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 45.911789063293654,
+      "learning_rate": 1.9953348893710645e-05,
+      "loss": 0.9594,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 49.16727143039645,
+      "learning_rate": 1.9953186047813637e-05,
+      "loss": 1.0672,
+      "step": 1151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 36.68163506778126,
+      "learning_rate": 1.9953022918853837e-05,
+      "loss": 0.9227,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 66.37353717679099,
+      "learning_rate": 1.9952859506835887e-05,
+      "loss": 1.17,
+      "step": 1153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 41.723693783326574,
+      "learning_rate": 1.995269581176444e-05,
+      "loss": 0.9656,
+      "step": 1154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 49.03610078097504,
+      "learning_rate": 1.9952531833644145e-05,
+      "loss": 1.1552,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 51.886310888535576,
+      "learning_rate": 1.9952367572479664e-05,
+      "loss": 0.9169,
+      "step": 1156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 48.18072044486321,
+      "learning_rate": 1.995220302827567e-05,
+      "loss": 1.1001,
+      "step": 1157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 46.97176629062073,
+      "learning_rate": 1.9952038201036848e-05,
+      "loss": 1.1011,
+      "step": 1158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 43.064188355233966,
+      "learning_rate": 1.995187309076788e-05,
+      "loss": 0.9492,
+      "step": 1159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 41.99103992118336,
+      "learning_rate": 1.9951707697473465e-05,
+      "loss": 0.8466,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 42.96798267581024,
+      "learning_rate": 1.99515420211583e-05,
+      "loss": 1.0662,
+      "step": 1161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 53.09466175645887,
+      "learning_rate": 1.9951376061827104e-05,
+      "loss": 1.0475,
+      "step": 1162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 48.10240414731775,
+      "learning_rate": 1.995120981948459e-05,
+      "loss": 0.8664,
+      "step": 1163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 55.16061895427785,
+      "learning_rate": 1.9951043294135498e-05,
+      "loss": 1.0784,
+      "step": 1164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 39.29747345243065,
+      "learning_rate": 1.995087648578455e-05,
+      "loss": 0.9923,
+      "step": 1165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 51.161208297562254,
+      "learning_rate": 1.9950709394436497e-05,
+      "loss": 1.0538,
+      "step": 1166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 53.2092956011854,
+      "learning_rate": 1.9950542020096094e-05,
+      "loss": 1.0275,
+      "step": 1167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.867818440667975,
+      "learning_rate": 1.995037436276809e-05,
+      "loss": 1.0301,
+      "step": 1168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 49.6895415264558,
+      "learning_rate": 1.9950206422457263e-05,
+      "loss": 1.0321,
+      "step": 1169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 52.728228012687,
+      "learning_rate": 1.9950038199168387e-05,
+      "loss": 0.9994,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 42.03979945250205,
+      "learning_rate": 1.9949869692906245e-05,
+      "loss": 1.0922,
+      "step": 1171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.552932960075864,
+      "learning_rate": 1.994970090367563e-05,
+      "loss": 1.0042,
+      "step": 1172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 40.950245674150906,
+      "learning_rate": 1.994953183148134e-05,
+      "loss": 1.0326,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 37.61437199752515,
+      "learning_rate": 1.9949362476328184e-05,
+      "loss": 1.0493,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.806837053219176,
+      "learning_rate": 1.9949192838220982e-05,
+      "loss": 1.0002,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 36.659401571953396,
+      "learning_rate": 1.9949022917164556e-05,
+      "loss": 0.9298,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.35552222810473,
+      "learning_rate": 1.9948852713163737e-05,
+      "loss": 0.9824,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 64.76893861988509,
+      "learning_rate": 1.9948682226223366e-05,
+      "loss": 1.0608,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 49.197022918244244,
+      "learning_rate": 1.9948511456348295e-05,
+      "loss": 1.0431,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 40.27461218170662,
+      "learning_rate": 1.9948340403543375e-05,
+      "loss": 0.9479,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 48.873936614739335,
+      "learning_rate": 1.9948169067813473e-05,
+      "loss": 0.9614,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 66.38365889857387,
+      "learning_rate": 1.9947997449163467e-05,
+      "loss": 1.1265,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 34.701071625633055,
+      "learning_rate": 1.9947825547598226e-05,
+      "loss": 0.9324,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 31.784284422432872,
+      "learning_rate": 1.994765336312265e-05,
+      "loss": 0.8232,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 56.64705561050729,
+      "learning_rate": 1.994748089574163e-05,
+      "loss": 1.2486,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 40.232732978572436,
+      "learning_rate": 1.9947308145460073e-05,
+      "loss": 0.8785,
+      "step": 1186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.37776785945451,
+      "learning_rate": 1.9947135112282894e-05,
+      "loss": 0.9691,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 54.72854043897317,
+      "learning_rate": 1.9946961796215006e-05,
+      "loss": 0.9773,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 40.66629509710492,
+      "learning_rate": 1.994678819726135e-05,
+      "loss": 0.9772,
+      "step": 1189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 64.85270366576871,
+      "learning_rate": 1.994661431542685e-05,
+      "loss": 1.0179,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.19752953659015,
+      "learning_rate": 1.994644015071646e-05,
+      "loss": 0.9652,
+      "step": 1191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 41.792774057723676,
+      "learning_rate": 1.994626570313513e-05,
+      "loss": 1.0052,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 42.485985781854126,
+      "learning_rate": 1.9946090972687827e-05,
+      "loss": 0.9498,
+      "step": 1193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 49.75744473043961,
+      "learning_rate": 1.994591595937951e-05,
+      "loss": 1.0941,
+      "step": 1194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 55.85904546040251,
+      "learning_rate": 1.994574066321516e-05,
+      "loss": 0.9484,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 38.51493938524472,
+      "learning_rate": 1.9945565084199766e-05,
+      "loss": 0.9416,
+      "step": 1196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 40.350839823688986,
+      "learning_rate": 1.994538922233832e-05,
+      "loss": 0.9748,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 41.14743540224266,
+      "learning_rate": 1.994521307763582e-05,
+      "loss": 0.9295,
+      "step": 1198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 66.4662736683396,
+      "learning_rate": 1.994503665009728e-05,
+      "loss": 1.0667,
+      "step": 1199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 39.70019451977713,
+      "learning_rate": 1.9944859939727714e-05,
+      "loss": 0.9033,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 35.616566004333436,
+      "learning_rate": 1.994468294653215e-05,
+      "loss": 0.8348,
+      "step": 1201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 51.5510012804131,
+      "learning_rate": 1.994450567051562e-05,
+      "loss": 1.0192,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 49.92363317315827,
+      "learning_rate": 1.9944328111683166e-05,
+      "loss": 0.9424,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 49.17486882194706,
+      "learning_rate": 1.9944150270039837e-05,
+      "loss": 1.0468,
+      "step": 1204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 38.16153266510411,
+      "learning_rate": 1.994397214559069e-05,
+      "loss": 0.9625,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 44.10582712669953,
+      "learning_rate": 1.9943793738340796e-05,
+      "loss": 1.1433,
+      "step": 1206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 50.40503894086571,
+      "learning_rate": 1.9943615048295224e-05,
+      "loss": 1.0328,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 45.30122025679955,
+      "learning_rate": 1.9943436075459056e-05,
+      "loss": 0.9257,
+      "step": 1208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 37.63562591343934,
+      "learning_rate": 1.994325681983738e-05,
+      "loss": 1.0369,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.46950387261908,
+      "learning_rate": 1.9943077281435302e-05,
+      "loss": 1.1726,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 41.49621046815287,
+      "learning_rate": 1.994289746025792e-05,
+      "loss": 0.9743,
+      "step": 1211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 43.99562648960964,
+      "learning_rate": 1.994271735631035e-05,
+      "loss": 0.8408,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 48.306858996761164,
+      "learning_rate": 1.9942536969597713e-05,
+      "loss": 1.1001,
+      "step": 1213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 59.772350791512935,
+      "learning_rate": 1.9942356300125146e-05,
+      "loss": 1.0028,
+      "step": 1214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.34521112988557,
+      "learning_rate": 1.994217534789778e-05,
+      "loss": 0.9909,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 57.51625236453027,
+      "learning_rate": 1.994199411292076e-05,
+      "loss": 1.0593,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 45.77384516490031,
+      "learning_rate": 1.9941812595199246e-05,
+      "loss": 1.0274,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 38.46357125100573,
+      "learning_rate": 1.99416307947384e-05,
+      "loss": 1.0478,
+      "step": 1218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 86.60543260955986,
+      "learning_rate": 1.9941448711543386e-05,
+      "loss": 1.0439,
+      "step": 1219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 34.93111624085498,
+      "learning_rate": 1.994126634561939e-05,
+      "loss": 1.0081,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 60.219999703236574,
+      "learning_rate": 1.994108369697159e-05,
+      "loss": 1.0405,
+      "step": 1221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 49.39930047435697,
+      "learning_rate": 1.9940900765605187e-05,
+      "loss": 0.9087,
+      "step": 1222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 59.09830857904731,
+      "learning_rate": 1.994071755152538e-05,
+      "loss": 0.9685,
+      "step": 1223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 45.9811547590905,
+      "learning_rate": 1.9940534054737383e-05,
+      "loss": 0.9626,
+      "step": 1224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 52.54878056270061,
+      "learning_rate": 1.9940350275246412e-05,
+      "loss": 0.9363,
+      "step": 1225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 53.490582867290605,
+      "learning_rate": 1.9940166213057693e-05,
+      "loss": 1.0062,
+      "step": 1226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 45.871897064999935,
+      "learning_rate": 1.9939981868176462e-05,
+      "loss": 1.0277,
+      "step": 1227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 70.88272629863687,
+      "learning_rate": 1.9939797240607966e-05,
+      "loss": 1.1251,
+      "step": 1228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 41.02840346596995,
+      "learning_rate": 1.9939612330357446e-05,
+      "loss": 0.8774,
+      "step": 1229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 71.06773462533843,
+      "learning_rate": 1.993942713743017e-05,
+      "loss": 1.0302,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 51.291033734967385,
+      "learning_rate": 1.9939241661831398e-05,
+      "loss": 1.0031,
+      "step": 1231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 49.774820797423764,
+      "learning_rate": 1.9939055903566403e-05,
+      "loss": 0.8954,
+      "step": 1232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 49.88597015447656,
+      "learning_rate": 1.9938869862640476e-05,
+      "loss": 1.031,
+      "step": 1233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.0168150715997,
+      "learning_rate": 1.9938683539058906e-05,
+      "loss": 0.9682,
+      "step": 1234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 41.64855461554947,
+      "learning_rate": 1.9938496932826988e-05,
+      "loss": 0.9156,
+      "step": 1235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 40.855645324984884,
+      "learning_rate": 1.993831004395003e-05,
+      "loss": 0.9961,
+      "step": 1236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.38004507110983,
+      "learning_rate": 1.9938122872433353e-05,
+      "loss": 1.0147,
+      "step": 1237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 52.22082611285226,
+      "learning_rate": 1.9937935418282274e-05,
+      "loss": 1.0139,
+      "step": 1238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 57.097380557663335,
+      "learning_rate": 1.9937747681502122e-05,
+      "loss": 0.9942,
+      "step": 1239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 39.45620531804434,
+      "learning_rate": 1.993755966209824e-05,
+      "loss": 1.0119,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 55.05050124027687,
+      "learning_rate": 1.9937371360075977e-05,
+      "loss": 1.0785,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 39.66571284931829,
+      "learning_rate": 1.9937182775440683e-05,
+      "loss": 0.9684,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.864106933525306,
+      "learning_rate": 1.9936993908197725e-05,
+      "loss": 0.8806,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 50.440739443718705,
+      "learning_rate": 1.9936804758352473e-05,
+      "loss": 1.0142,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 47.98346556556628,
+      "learning_rate": 1.9936615325910308e-05,
+      "loss": 1.06,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 41.742848863169115,
+      "learning_rate": 1.9936425610876614e-05,
+      "loss": 1.0276,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 43.49544670675453,
+      "learning_rate": 1.993623561325679e-05,
+      "loss": 0.9977,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 37.95306016422543,
+      "learning_rate": 1.9936045333056235e-05,
+      "loss": 0.886,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 48.50395673657192,
+      "learning_rate": 1.9935854770280362e-05,
+      "loss": 1.0366,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 58.02663895093265,
+      "learning_rate": 1.9935663924934596e-05,
+      "loss": 1.1392,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 52.156991591075425,
+      "learning_rate": 1.993547279702436e-05,
+      "loss": 0.9805,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 41.12735462115027,
+      "learning_rate": 1.9935281386555087e-05,
+      "loss": 1.0394,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 41.459809875392786,
+      "learning_rate": 1.9935089693532225e-05,
+      "loss": 0.8264,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 72.5139845021601,
+      "learning_rate": 1.9934897717961223e-05,
+      "loss": 1.003,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 43.183210775470634,
+      "learning_rate": 1.9934705459847544e-05,
+      "loss": 0.9795,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 44.46059332730973,
+      "learning_rate": 1.9934512919196647e-05,
+      "loss": 0.8255,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 41.83451292317708,
+      "learning_rate": 1.9934320096014022e-05,
+      "loss": 0.9362,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 42.54486178406279,
+      "learning_rate": 1.9934126990305142e-05,
+      "loss": 0.991,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 46.36293329327998,
+      "learning_rate": 1.99339336020755e-05,
+      "loss": 0.9917,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 39.11278666278151,
+      "learning_rate": 1.99337399313306e-05,
+      "loss": 0.909,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 44.180002679889505,
+      "learning_rate": 1.9933545978075942e-05,
+      "loss": 0.9769,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 75.04811598708564,
+      "learning_rate": 1.9933351742317053e-05,
+      "loss": 1.0607,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 46.84663974127035,
+      "learning_rate": 1.993315722405945e-05,
+      "loss": 1.0374,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 53.62729671868499,
+      "learning_rate": 1.9932962423308666e-05,
+      "loss": 1.0424,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 56.63367563274835,
+      "learning_rate": 1.9932767340070243e-05,
+      "loss": 1.1102,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 51.72588069197363,
+      "learning_rate": 1.9932571974349724e-05,
+      "loss": 1.0541,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 50.31859571125037,
+      "learning_rate": 1.9932376326152672e-05,
+      "loss": 1.0269,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 56.33950027967328,
+      "learning_rate": 1.9932180395484644e-05,
+      "loss": 1.0273,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 38.00738531109795,
+      "learning_rate": 1.993198418235122e-05,
+      "loss": 0.9546,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 47.82077360686535,
+      "learning_rate": 1.993178768675797e-05,
+      "loss": 1.0376,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 74.65468675294879,
+      "learning_rate": 1.9931590908710495e-05,
+      "loss": 0.9229,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 39.52092619573167,
+      "learning_rate": 1.9931393848214376e-05,
+      "loss": 0.92,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 47.035360391143875,
+      "learning_rate": 1.9931196505275233e-05,
+      "loss": 1.1145,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 46.191847089474386,
+      "learning_rate": 1.993099887989867e-05,
+      "loss": 0.9524,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 58.505741563072526,
+      "learning_rate": 1.9930800972090307e-05,
+      "loss": 1.0322,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 60.50322292350102,
+      "learning_rate": 1.9930602781855772e-05,
+      "loss": 0.966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 51.19964712121923,
+      "learning_rate": 1.9930404309200706e-05,
+      "loss": 0.9347,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 57.735750950585754,
+      "learning_rate": 1.9930205554130748e-05,
+      "loss": 0.9837,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 40.40100273644189,
+      "learning_rate": 1.9930006516651555e-05,
+      "loss": 0.9229,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 36.35531173185216,
+      "learning_rate": 1.9929807196768785e-05,
+      "loss": 0.8663,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 55.06257268270833,
+      "learning_rate": 1.9929607594488104e-05,
+      "loss": 1.0425,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 43.34226283662309,
+      "learning_rate": 1.9929407709815197e-05,
+      "loss": 1.0182,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 53.462685335278216,
+      "learning_rate": 1.992920754275574e-05,
+      "loss": 0.9286,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 44.77855508685496,
+      "learning_rate": 1.992900709331543e-05,
+      "loss": 0.9937,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 43.5849072966575,
+      "learning_rate": 1.9928806361499965e-05,
+      "loss": 0.9416,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 41.72436425082982,
+      "learning_rate": 1.9928605347315057e-05,
+      "loss": 0.8422,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 53.027216841684584,
+      "learning_rate": 1.992840405076642e-05,
+      "loss": 0.9801,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 40.8961652326727,
+      "learning_rate": 1.992820247185978e-05,
+      "loss": 0.9837,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 44.025678229320164,
+      "learning_rate": 1.992800061060087e-05,
+      "loss": 0.9377,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 41.804984302497274,
+      "learning_rate": 1.992779846699543e-05,
+      "loss": 0.8968,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 48.4958540776476,
+      "learning_rate": 1.992759604104921e-05,
+      "loss": 0.9845,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 64.41925542382768,
+      "learning_rate": 1.9927393332767965e-05,
+      "loss": 1.0592,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 44.50959436520186,
+      "learning_rate": 1.9927190342157462e-05,
+      "loss": 0.9178,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 54.31831516613704,
+      "learning_rate": 1.9926987069223475e-05,
+      "loss": 0.9906,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 40.95054300458234,
+      "learning_rate": 1.9926783513971777e-05,
+      "loss": 0.9288,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 49.5340378226441,
+      "learning_rate": 1.9926579676408168e-05,
+      "loss": 1.0065,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 40.293295320786,
+      "learning_rate": 1.992637555653844e-05,
+      "loss": 0.9897,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 42.06811748364076,
+      "learning_rate": 1.9926171154368393e-05,
+      "loss": 0.9821,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 39.46025299040469,
+      "learning_rate": 1.992596646990385e-05,
+      "loss": 1.0361,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 61.967645001430505,
+      "learning_rate": 1.992576150315062e-05,
+      "loss": 0.9558,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 40.816794801845546,
+      "learning_rate": 1.9925556254114547e-05,
+      "loss": 0.9513,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 56.22173436265846,
+      "learning_rate": 1.9925350722801458e-05,
+      "loss": 0.8551,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 42.5768529299477,
+      "learning_rate": 1.9925144909217197e-05,
+      "loss": 0.9741,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 39.37037948867338,
+      "learning_rate": 1.9924938813367626e-05,
+      "loss": 1.0031,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 76.77116755736712,
+      "learning_rate": 1.99247324352586e-05,
+      "loss": 1.0394,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 53.297425145620444,
+      "learning_rate": 1.9924525774895986e-05,
+      "loss": 1.0645,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 33.014108524618116,
+      "learning_rate": 1.9924318832285667e-05,
+      "loss": 0.9166,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 46.45979679978724,
+      "learning_rate": 1.9924111607433528e-05,
+      "loss": 1.0719,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 36.66172296979776,
+      "learning_rate": 1.992390410034546e-05,
+      "loss": 1.0042,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 44.1309482574412,
+      "learning_rate": 1.992369631102736e-05,
+      "loss": 1.0202,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 54.34472851916978,
+      "learning_rate": 1.9923488239485146e-05,
+      "loss": 1.027,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 36.2452743639924,
+      "learning_rate": 1.992327988572473e-05,
+      "loss": 0.9512,
+      "step": 1312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.539096804371994,
+      "learning_rate": 1.9923071249752038e-05,
+      "loss": 0.929,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 61.99402454637325,
+      "learning_rate": 1.9922862331573007e-05,
+      "loss": 1.0741,
+      "step": 1314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.42043511528046,
+      "learning_rate": 1.9922653131193575e-05,
+      "loss": 0.9703,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.43902070656226,
+      "learning_rate": 1.9922443648619696e-05,
+      "loss": 1.0854,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 41.80058373369901,
+      "learning_rate": 1.992223388385732e-05,
+      "loss": 0.9145,
+      "step": 1317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.12923178844933,
+      "learning_rate": 1.9922023836912418e-05,
+      "loss": 1.071,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 50.41323464537227,
+      "learning_rate": 1.992181350779096e-05,
+      "loss": 1.1742,
+      "step": 1319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 37.54408506520476,
+      "learning_rate": 1.9921602896498934e-05,
+      "loss": 0.8804,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 45.386337261597774,
+      "learning_rate": 1.9921392003042322e-05,
+      "loss": 1.0976,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.99040433981839,
+      "learning_rate": 1.9921180827427126e-05,
+      "loss": 0.9689,
+      "step": 1322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 45.670902055210064,
+      "learning_rate": 1.9920969369659354e-05,
+      "loss": 0.8969,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.643179560120444,
+      "learning_rate": 1.9920757629745015e-05,
+      "loss": 0.9771,
+      "step": 1324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 128.2001341755086,
+      "learning_rate": 1.992054560769013e-05,
+      "loss": 1.0254,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.6116473592451,
+      "learning_rate": 1.9920333303500734e-05,
+      "loss": 0.9852,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 48.195822554913306,
+      "learning_rate": 1.992012071718286e-05,
+      "loss": 1.0244,
+      "step": 1327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 46.17316703215017,
+      "learning_rate": 1.9919907848742556e-05,
+      "loss": 0.988,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 52.6664076156925,
+      "learning_rate": 1.991969469818588e-05,
+      "loss": 0.9566,
+      "step": 1329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 47.62997860054466,
+      "learning_rate": 1.9919481265518885e-05,
+      "loss": 0.9483,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 41.39078710919579,
+      "learning_rate": 1.9919267550747647e-05,
+      "loss": 0.9651,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 44.01853530248638,
+      "learning_rate": 1.991905355387824e-05,
+      "loss": 0.9193,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 33.04350578502141,
+      "learning_rate": 1.9918839274916757e-05,
+      "loss": 0.8305,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 34.371277156721575,
+      "learning_rate": 1.9918624713869287e-05,
+      "loss": 0.9007,
+      "step": 1334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.63405758849442,
+      "learning_rate": 1.9918409870741928e-05,
+      "loss": 1.0144,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 41.567859912107295,
+      "learning_rate": 1.9918194745540798e-05,
+      "loss": 1.0101,
+      "step": 1336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 35.74265308514639,
+      "learning_rate": 1.991797933827201e-05,
+      "loss": 0.9752,
+      "step": 1337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 47.638638778724456,
+      "learning_rate": 1.9917763648941692e-05,
+      "loss": 1.0426,
+      "step": 1338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.25469185795316,
+      "learning_rate": 1.991754767755598e-05,
+      "loss": 0.9061,
+      "step": 1339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 52.89732072784047,
+      "learning_rate": 1.991733142412101e-05,
+      "loss": 0.9692,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 49.262350286310955,
+      "learning_rate": 1.9917114888642937e-05,
+      "loss": 1.064,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 40.6988082266051,
+      "learning_rate": 1.9916898071127915e-05,
+      "loss": 1.0128,
+      "step": 1342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 58.78675832700314,
+      "learning_rate": 1.9916680971582116e-05,
+      "loss": 1.0679,
+      "step": 1343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 34.206130201149804,
+      "learning_rate": 1.9916463590011713e-05,
+      "loss": 0.9263,
+      "step": 1344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.346289342356414,
+      "learning_rate": 1.9916245926422883e-05,
+      "loss": 1.0049,
+      "step": 1345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 50.97089589160672,
+      "learning_rate": 1.9916027980821823e-05,
+      "loss": 1.1579,
+      "step": 1346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 54.26402521562447,
+      "learning_rate": 1.9915809753214725e-05,
+      "loss": 0.9748,
+      "step": 1347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 37.54807649452933,
+      "learning_rate": 1.99155912436078e-05,
+      "loss": 0.9361,
+      "step": 1348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 48.293198508827125,
+      "learning_rate": 1.991537245200726e-05,
+      "loss": 1.027,
+      "step": 1349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 47.5688798863802,
+      "learning_rate": 1.991515337841933e-05,
+      "loss": 1.0162,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 34.15477125800747,
+      "learning_rate": 1.9914934022850233e-05,
+      "loss": 1.0921,
+      "step": 1351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 41.910225107202386,
+      "learning_rate": 1.9914714385306217e-05,
+      "loss": 1.1166,
+      "step": 1352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 44.04450772390034,
+      "learning_rate": 1.9914494465793522e-05,
+      "loss": 0.9137,
+      "step": 1353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 45.04298500786563,
+      "learning_rate": 1.9914274264318402e-05,
+      "loss": 0.9045,
+      "step": 1354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 52.52610185802858,
+      "learning_rate": 1.9914053780887122e-05,
+      "loss": 0.9678,
+      "step": 1355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 49.98191381931664,
+      "learning_rate": 1.9913833015505957e-05,
+      "loss": 0.8697,
+      "step": 1356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 44.994237938837074,
+      "learning_rate": 1.9913611968181178e-05,
+      "loss": 0.9883,
+      "step": 1357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.25274562459209,
+      "learning_rate": 1.9913390638919074e-05,
+      "loss": 1.0003,
+      "step": 1358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 37.20398583311437,
+      "learning_rate": 1.9913169027725936e-05,
+      "loss": 0.875,
+      "step": 1359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.55310060082049,
+      "learning_rate": 1.9912947134608073e-05,
+      "loss": 0.9695,
+      "step": 1360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.36840254598057,
+      "learning_rate": 1.9912724959571793e-05,
+      "loss": 0.8381,
+      "step": 1361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 50.933668505021856,
+      "learning_rate": 1.9912502502623413e-05,
+      "loss": 1.0943,
+      "step": 1362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 41.708750602979855,
+      "learning_rate": 1.991227976376926e-05,
+      "loss": 0.9653,
+      "step": 1363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.60219502010079,
+      "learning_rate": 1.991205674301567e-05,
+      "loss": 1.0635,
+      "step": 1364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.3421573635008,
+      "learning_rate": 1.9911833440368982e-05,
+      "loss": 1.0063,
+      "step": 1365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.52439346559882,
+      "learning_rate": 1.9911609855835553e-05,
+      "loss": 0.9901,
+      "step": 1366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 57.88261401389174,
+      "learning_rate": 1.9911385989421736e-05,
+      "loss": 0.9508,
+      "step": 1367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.95236580495092,
+      "learning_rate": 1.9911161841133898e-05,
+      "loss": 0.9369,
+      "step": 1368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 36.01743148940177,
+      "learning_rate": 1.9910937410978418e-05,
+      "loss": 0.99,
+      "step": 1369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.89647107470846,
+      "learning_rate": 1.9910712698961673e-05,
+      "loss": 1.04,
+      "step": 1370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 37.909050268031265,
+      "learning_rate": 1.991048770509006e-05,
+      "loss": 1.0143,
+      "step": 1371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 37.176444408982896,
+      "learning_rate": 1.991026242936997e-05,
+      "loss": 0.967,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 45.98808187047675,
+      "learning_rate": 1.9910036871807817e-05,
+      "loss": 0.9696,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 45.77863627709507,
+      "learning_rate": 1.9909811032410012e-05,
+      "loss": 0.8696,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.02207687850281,
+      "learning_rate": 1.9909584911182977e-05,
+      "loss": 0.9732,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.49728567877626,
+      "learning_rate": 1.9909358508133145e-05,
+      "loss": 0.9244,
+      "step": 1376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.264973009498874,
+      "learning_rate": 1.990913182326695e-05,
+      "loss": 0.9332,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.47687042171626,
+      "learning_rate": 1.990890485659085e-05,
+      "loss": 0.8936,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.497978169816186,
+      "learning_rate": 1.9908677608111287e-05,
+      "loss": 1.0699,
+      "step": 1379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.10714618204177,
+      "learning_rate": 1.9908450077834732e-05,
+      "loss": 0.902,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 57.527740769328595,
+      "learning_rate": 1.9908222265767653e-05,
+      "loss": 1.0383,
+      "step": 1381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 63.04358043736723,
+      "learning_rate": 1.9907994171916524e-05,
+      "loss": 0.9364,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 64.73845747206309,
+      "learning_rate": 1.9907765796287844e-05,
+      "loss": 0.9945,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 54.92768713203686,
+      "learning_rate": 1.9907537138888097e-05,
+      "loss": 1.0392,
+      "step": 1384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 50.501524127962035,
+      "learning_rate": 1.990730819972379e-05,
+      "loss": 0.984,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 48.02387442016927,
+      "learning_rate": 1.9907078978801432e-05,
+      "loss": 0.9864,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 44.09694021783937,
+      "learning_rate": 1.9906849476127545e-05,
+      "loss": 1.0791,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 54.457715679169425,
+      "learning_rate": 1.9906619691708652e-05,
+      "loss": 0.9783,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 62.370497294491294,
+      "learning_rate": 1.9906389625551294e-05,
+      "loss": 1.0736,
+      "step": 1389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.86056276467184,
+      "learning_rate": 1.990615927766201e-05,
+      "loss": 0.9938,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 46.374357825714476,
+      "learning_rate": 1.9905928648047346e-05,
+      "loss": 0.9381,
+      "step": 1391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 40.92852406997147,
+      "learning_rate": 1.990569773671387e-05,
+      "loss": 0.9852,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 44.863405233302366,
+      "learning_rate": 1.9905466543668143e-05,
+      "loss": 0.9263,
+      "step": 1393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 38.865752061570376,
+      "learning_rate": 1.9905235068916745e-05,
+      "loss": 0.9573,
+      "step": 1394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 54.546339484246566,
+      "learning_rate": 1.9905003312466255e-05,
+      "loss": 0.9864,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 91.74621224965205,
+      "learning_rate": 1.9904771274323262e-05,
+      "loss": 0.9304,
+      "step": 1396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.38014113186885,
+      "learning_rate": 1.990453895449437e-05,
+      "loss": 0.8957,
+      "step": 1397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 33.36639063943591,
+      "learning_rate": 1.9904306352986183e-05,
+      "loss": 0.9615,
+      "step": 1398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.979542881847195,
+      "learning_rate": 1.990407346980532e-05,
+      "loss": 0.95,
+      "step": 1399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 42.22525625405212,
+      "learning_rate": 1.99038403049584e-05,
+      "loss": 0.9062,
+      "step": 1400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 49.342316619366834,
+      "learning_rate": 1.9903606858452056e-05,
+      "loss": 1.0786,
+      "step": 1401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 50.28142117506138,
+      "learning_rate": 1.9903373130292928e-05,
+      "loss": 0.999,
+      "step": 1402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 46.87292811074576,
+      "learning_rate": 1.9903139120487654e-05,
+      "loss": 1.0182,
+      "step": 1403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 45.11098570106772,
+      "learning_rate": 1.9902904829042902e-05,
+      "loss": 0.9481,
+      "step": 1404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 48.56047411292838,
+      "learning_rate": 1.9902670255965333e-05,
+      "loss": 1.0839,
+      "step": 1405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.878701348312525,
+      "learning_rate": 1.9902435401261608e-05,
+      "loss": 1.0437,
+      "step": 1406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 54.961526843980096,
+      "learning_rate": 1.990220026493842e-05,
+      "loss": 1.1243,
+      "step": 1407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 38.15897139108616,
+      "learning_rate": 1.9901964847002442e-05,
+      "loss": 0.9171,
+      "step": 1408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 49.148318645750166,
+      "learning_rate": 1.990172914746038e-05,
+      "loss": 1.013,
+      "step": 1409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 39.94874193647948,
+      "learning_rate": 1.9901493166318934e-05,
+      "loss": 0.9502,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 45.86057532330163,
+      "learning_rate": 1.9901256903584812e-05,
+      "loss": 1.0183,
+      "step": 1411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 44.69645909487475,
+      "learning_rate": 1.9901020359264738e-05,
+      "loss": 0.9299,
+      "step": 1412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 38.752461213735444,
+      "learning_rate": 1.9900783533365437e-05,
+      "loss": 0.871,
+      "step": 1413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 40.716268084661436,
+      "learning_rate": 1.9900546425893644e-05,
+      "loss": 1.0327,
+      "step": 1414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 51.42820844474866,
+      "learning_rate": 1.99003090368561e-05,
+      "loss": 1.0274,
+      "step": 1415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 46.411447407370126,
+      "learning_rate": 1.990007136625956e-05,
+      "loss": 1.0082,
+      "step": 1416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 49.42299661198805,
+      "learning_rate": 1.989983341411078e-05,
+      "loss": 0.9738,
+      "step": 1417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 46.75804047620579,
+      "learning_rate": 1.9899595180416532e-05,
+      "loss": 1.0323,
+      "step": 1418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 43.47086070768687,
+      "learning_rate": 1.9899356665183584e-05,
+      "loss": 0.9703,
+      "step": 1419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 50.25367414982048,
+      "learning_rate": 1.989911786841873e-05,
+      "loss": 1.0091,
+      "step": 1420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 52.410342125535664,
+      "learning_rate": 1.9898878790128752e-05,
+      "loss": 0.9457,
+      "step": 1421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 48.982992977133655,
+      "learning_rate": 1.9898639430320448e-05,
+      "loss": 0.8876,
+      "step": 1422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 49.924359517962856,
+      "learning_rate": 1.9898399789000633e-05,
+      "loss": 0.9026,
+      "step": 1423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 50.75924051656007,
+      "learning_rate": 1.9898159866176115e-05,
+      "loss": 1.0151,
+      "step": 1424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 51.46776032784672,
+      "learning_rate": 1.9897919661853723e-05,
+      "loss": 1.04,
+      "step": 1425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 44.297714144715876,
+      "learning_rate": 1.9897679176040284e-05,
+      "loss": 0.9703,
+      "step": 1426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 50.38701176900115,
+      "learning_rate": 1.9897438408742644e-05,
+      "loss": 1.0011,
+      "step": 1427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 44.76718091141513,
+      "learning_rate": 1.9897197359967642e-05,
+      "loss": 1.0624,
+      "step": 1428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.743402969051935,
+      "learning_rate": 1.9896956029722138e-05,
+      "loss": 1.0317,
+      "step": 1429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 38.126288295304015,
+      "learning_rate": 1.9896714418012993e-05,
+      "loss": 1.0111,
+      "step": 1430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 48.63694707369427,
+      "learning_rate": 1.9896472524847078e-05,
+      "loss": 0.9237,
+      "step": 1431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 35.40710098711899,
+      "learning_rate": 1.9896230350231274e-05,
+      "loss": 0.9511,
+      "step": 1432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 40.72330014970742,
+      "learning_rate": 1.9895987894172474e-05,
+      "loss": 0.9322,
+      "step": 1433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 44.43268560198322,
+      "learning_rate": 1.989574515667756e-05,
+      "loss": 1.0737,
+      "step": 1434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 42.07785873810461,
+      "learning_rate": 1.9895502137753448e-05,
+      "loss": 0.9463,
+      "step": 1435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 51.862861913750045,
+      "learning_rate": 1.989525883740704e-05,
+      "loss": 0.9289,
+      "step": 1436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 37.47139062238346,
+      "learning_rate": 1.989501525564526e-05,
+      "loss": 0.8472,
+      "step": 1437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 48.96326123157876,
+      "learning_rate": 1.9894771392475036e-05,
+      "loss": 0.9637,
+      "step": 1438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 58.12722861826904,
+      "learning_rate": 1.98945272479033e-05,
+      "loss": 1.1685,
+      "step": 1439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 37.98382856383177,
+      "learning_rate": 1.9894282821936995e-05,
+      "loss": 0.9691,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 40.83558800304924,
+      "learning_rate": 1.989403811458308e-05,
+      "loss": 0.9485,
+      "step": 1441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 47.21922173759873,
+      "learning_rate": 1.9893793125848506e-05,
+      "loss": 0.9795,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 35.35148859005784,
+      "learning_rate": 1.9893547855740243e-05,
+      "loss": 0.8705,
+      "step": 1443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 59.80424948045423,
+      "learning_rate": 1.9893302304265267e-05,
+      "loss": 0.8352,
+      "step": 1444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 44.611136716671936,
+      "learning_rate": 1.989305647143056e-05,
+      "loss": 1.0779,
+      "step": 1445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 43.37518251693978,
+      "learning_rate": 1.9892810357243116e-05,
+      "loss": 0.8901,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 44.10170210756247,
+      "learning_rate": 1.989256396170993e-05,
+      "loss": 0.9555,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 46.95523804338715,
+      "learning_rate": 1.9892317284838016e-05,
+      "loss": 0.9494,
+      "step": 1448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 47.634706294056315,
+      "learning_rate": 1.9892070326634383e-05,
+      "loss": 0.9486,
+      "step": 1449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.69918916151797,
+      "learning_rate": 1.9891823087106057e-05,
+      "loss": 1.0125,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 59.3707392960765,
+      "learning_rate": 1.9891575566260068e-05,
+      "loss": 0.9252,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 38.031165897900166,
+      "learning_rate": 1.9891327764103456e-05,
+      "loss": 1.0034,
+      "step": 1452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 47.16733436426394,
+      "learning_rate": 1.989107968064327e-05,
+      "loss": 0.8512,
+      "step": 1453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 47.34972322256208,
+      "learning_rate": 1.9890831315886566e-05,
+      "loss": 0.8756,
+      "step": 1454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 48.88718690667081,
+      "learning_rate": 1.98905826698404e-05,
+      "loss": 1.0336,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 49.81287907033775,
+      "learning_rate": 1.989033374251185e-05,
+      "loss": 1.0056,
+      "step": 1456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.78973840798161,
+      "learning_rate": 1.9890084533907998e-05,
+      "loss": 0.9713,
+      "step": 1457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 44.36090161047088,
+      "learning_rate": 1.9889835044035925e-05,
+      "loss": 0.9462,
+      "step": 1458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 45.22404683059152,
+      "learning_rate": 1.988958527290273e-05,
+      "loss": 0.9285,
+      "step": 1459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 49.8342680901897,
+      "learning_rate": 1.9889335220515512e-05,
+      "loss": 0.9419,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.9020079417016,
+      "learning_rate": 1.9889084886881387e-05,
+      "loss": 0.9434,
+      "step": 1461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 34.18444627168031,
+      "learning_rate": 1.9888834272007475e-05,
+      "loss": 0.9319,
+      "step": 1462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 54.96695104260945,
+      "learning_rate": 1.9888583375900897e-05,
+      "loss": 1.0158,
+      "step": 1463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 42.32039396275706,
+      "learning_rate": 1.9888332198568794e-05,
+      "loss": 0.9876,
+      "step": 1464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 49.5824249211077,
+      "learning_rate": 1.988808074001831e-05,
+      "loss": 0.9514,
+      "step": 1465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 43.31532297664154,
+      "learning_rate": 1.9887829000256592e-05,
+      "loss": 1.0372,
+      "step": 1466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 42.712965595274284,
+      "learning_rate": 1.9887576979290803e-05,
+      "loss": 0.8378,
+      "step": 1467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 50.1277649371346,
+      "learning_rate": 1.9887324677128107e-05,
+      "loss": 1.0534,
+      "step": 1468
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 37.198781281806795,
+      "learning_rate": 1.988707209377568e-05,
+      "loss": 0.9024,
+      "step": 1469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.52750176312307,
+      "learning_rate": 1.9886819229240708e-05,
+      "loss": 0.863,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.02937798976212,
+      "learning_rate": 1.9886566083530382e-05,
+      "loss": 0.9285,
+      "step": 1471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.288747929201044,
+      "learning_rate": 1.9886312656651897e-05,
+      "loss": 0.94,
+      "step": 1472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 49.61241184821722,
+      "learning_rate": 1.9886058948612466e-05,
+      "loss": 0.9098,
+      "step": 1473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.03440031802546,
+      "learning_rate": 1.98858049594193e-05,
+      "loss": 0.9953,
+      "step": 1474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 58.525803499437934,
+      "learning_rate": 1.9885550689079624e-05,
+      "loss": 0.9465,
+      "step": 1475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 49.239169580722475,
+      "learning_rate": 1.988529613760067e-05,
+      "loss": 1.0323,
+      "step": 1476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 48.39901322706652,
+      "learning_rate": 1.9885041304989682e-05,
+      "loss": 0.9519,
+      "step": 1477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 53.799986869782046,
+      "learning_rate": 1.9884786191253895e-05,
+      "loss": 0.9104,
+      "step": 1478
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 58.81073140110997,
+      "learning_rate": 1.988453079640057e-05,
+      "loss": 1.0068,
+      "step": 1479
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 45.24981398570569,
+      "learning_rate": 1.9884275120436976e-05,
+      "loss": 0.8453,
+      "step": 1480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 50.82932614755063,
+      "learning_rate": 1.988401916337038e-05,
+      "loss": 0.9668,
+      "step": 1481
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 45.44896365537967,
+      "learning_rate": 1.9883762925208058e-05,
+      "loss": 1.1528,
+      "step": 1482
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.01139101100413,
+      "learning_rate": 1.98835064059573e-05,
+      "loss": 0.9999,
+      "step": 1483
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 45.58756990754249,
+      "learning_rate": 1.9883249605625403e-05,
+      "loss": 1.0373,
+      "step": 1484
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 43.374739736826044,
+      "learning_rate": 1.988299252421967e-05,
+      "loss": 0.9107,
+      "step": 1485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 39.87572141371439,
+      "learning_rate": 1.9882735161747407e-05,
+      "loss": 0.931,
+      "step": 1486
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 67.81588757047821,
+      "learning_rate": 1.9882477518215938e-05,
+      "loss": 0.9907,
+      "step": 1487
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 50.561870552171904,
+      "learning_rate": 1.9882219593632595e-05,
+      "loss": 0.9668,
+      "step": 1488
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 42.34243406665085,
+      "learning_rate": 1.9881961388004702e-05,
+      "loss": 0.959,
+      "step": 1489
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 48.378596524543475,
+      "learning_rate": 1.988170290133961e-05,
+      "loss": 1.0464,
+      "step": 1490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 44.52519673213516,
+      "learning_rate": 1.9881444133644668e-05,
+      "loss": 1.0489,
+      "step": 1491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 120.93752730021798,
+      "learning_rate": 1.9881185084927234e-05,
+      "loss": 0.8467,
+      "step": 1492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 40.920817334932934,
+      "learning_rate": 1.9880925755194677e-05,
+      "loss": 0.9675,
+      "step": 1493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 47.639028680718006,
+      "learning_rate": 1.9880666144454372e-05,
+      "loss": 0.888,
+      "step": 1494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 130.51458784394825,
+      "learning_rate": 1.9880406252713703e-05,
+      "loss": 1.025,
+      "step": 1495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 38.49586410415256,
+      "learning_rate": 1.988014607998006e-05,
+      "loss": 0.8901,
+      "step": 1496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 43.78778999378509,
+      "learning_rate": 1.987988562626084e-05,
+      "loss": 1.0235,
+      "step": 1497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 32.83754994737466,
+      "learning_rate": 1.9879624891563456e-05,
+      "loss": 0.8785,
+      "step": 1498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.114777021948136,
+      "learning_rate": 1.9879363875895317e-05,
+      "loss": 0.9829,
+      "step": 1499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 53.07838782154737,
+      "learning_rate": 1.987910257926385e-05,
+      "loss": 0.9789,
+      "step": 1500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.430676147243375,
+      "learning_rate": 1.9878841001676484e-05,
+      "loss": 0.9287,
+      "step": 1501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.13959405971382,
+      "learning_rate": 1.9878579143140656e-05,
+      "loss": 0.9821,
+      "step": 1502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 64.17239633055372,
+      "learning_rate": 1.987831700366382e-05,
+      "loss": 1.0687,
+      "step": 1503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 49.430663102664674,
+      "learning_rate": 1.9878054583253426e-05,
+      "loss": 0.979,
+      "step": 1504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 55.8286730197302,
+      "learning_rate": 1.9877791881916935e-05,
+      "loss": 1.0092,
+      "step": 1505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 57.54270503989944,
+      "learning_rate": 1.9877528899661826e-05,
+      "loss": 0.9934,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 71.40052453924957,
+      "learning_rate": 1.9877265636495574e-05,
+      "loss": 1.1005,
+      "step": 1507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 45.862618150036425,
+      "learning_rate": 1.987700209242566e-05,
+      "loss": 1.0048,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 48.98731183704353,
+      "learning_rate": 1.9876738267459587e-05,
+      "loss": 0.9702,
+      "step": 1509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 34.14835567285652,
+      "learning_rate": 1.9876474161604858e-05,
+      "loss": 0.9243,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 69.51946986157616,
+      "learning_rate": 1.9876209774868975e-05,
+      "loss": 1.1579,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 46.79618677515238,
+      "learning_rate": 1.9875945107259467e-05,
+      "loss": 0.8646,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 47.31198305782374,
+      "learning_rate": 1.987568015878386e-05,
+      "loss": 0.8942,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 65.48381295278124,
+      "learning_rate": 1.9875414929449684e-05,
+      "loss": 0.9279,
+      "step": 1514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 37.361793876910475,
+      "learning_rate": 1.9875149419264483e-05,
+      "loss": 0.932,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 36.318932218000974,
+      "learning_rate": 1.987488362823581e-05,
+      "loss": 0.9295,
+      "step": 1516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 46.36442962646146,
+      "learning_rate": 1.9874617556371224e-05,
+      "loss": 0.9422,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 38.864046319901355,
+      "learning_rate": 1.987435120367829e-05,
+      "loss": 0.9004,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 79.42194101232185,
+      "learning_rate": 1.987408457016459e-05,
+      "loss": 1.0111,
+      "step": 1519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 54.271603418354815,
+      "learning_rate": 1.9873817655837695e-05,
+      "loss": 1.0575,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 50.655231444410234,
+      "learning_rate": 1.9873550460705203e-05,
+      "loss": 0.9574,
+      "step": 1521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 46.34373357853563,
+      "learning_rate": 1.987328298477471e-05,
+      "loss": 0.9583,
+      "step": 1522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 49.419503670364875,
+      "learning_rate": 1.9873015228053828e-05,
+      "loss": 0.9649,
+      "step": 1523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 40.617440201079084,
+      "learning_rate": 1.9872747190550167e-05,
+      "loss": 1.094,
+      "step": 1524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 39.654499499296456,
+      "learning_rate": 1.987247887227135e-05,
+      "loss": 0.9095,
+      "step": 1525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 53.649478118434395,
+      "learning_rate": 1.9872210273225012e-05,
+      "loss": 0.9826,
+      "step": 1526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 35.96008389603733,
+      "learning_rate": 1.9871941393418788e-05,
+      "loss": 0.9255,
+      "step": 1527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.739709875594976,
+      "learning_rate": 1.9871672232860323e-05,
+      "loss": 0.9706,
+      "step": 1528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.58525941262827,
+      "learning_rate": 1.9871402791557276e-05,
+      "loss": 0.9352,
+      "step": 1529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 34.02385866099389,
+      "learning_rate": 1.987113306951731e-05,
+      "loss": 0.9237,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 50.052586577628155,
+      "learning_rate": 1.9870863066748092e-05,
+      "loss": 1.0225,
+      "step": 1531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 36.81680513606029,
+      "learning_rate": 1.9870592783257303e-05,
+      "loss": 0.8339,
+      "step": 1532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 63.21392231344922,
+      "learning_rate": 1.987032221905263e-05,
+      "loss": 1.0942,
+      "step": 1533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.90156921197677,
+      "learning_rate": 1.9870051374141765e-05,
+      "loss": 0.9142,
+      "step": 1534
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 41.69971219702185,
+      "learning_rate": 1.9869780248532416e-05,
+      "loss": 1.0387,
+      "step": 1535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 48.365193048145336,
+      "learning_rate": 1.986950884223229e-05,
+      "loss": 0.9532,
+      "step": 1536
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 41.30130367946233,
+      "learning_rate": 1.9869237155249105e-05,
+      "loss": 0.9429,
+      "step": 1537
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 38.234024064948166,
+      "learning_rate": 1.9868965187590588e-05,
+      "loss": 0.8717,
+      "step": 1538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 48.62077327882241,
+      "learning_rate": 1.9868692939264474e-05,
+      "loss": 0.9267,
+      "step": 1539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 34.49081196618081,
+      "learning_rate": 1.9868420410278506e-05,
+      "loss": 0.8647,
+      "step": 1540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 47.96239974642992,
+      "learning_rate": 1.9868147600640433e-05,
+      "loss": 0.9443,
+      "step": 1541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 49.915393959064296,
+      "learning_rate": 1.9867874510358017e-05,
+      "loss": 0.9402,
+      "step": 1542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 35.47633986412676,
+      "learning_rate": 1.9867601139439023e-05,
+      "loss": 1.0271,
+      "step": 1543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 49.03381183902092,
+      "learning_rate": 1.9867327487891224e-05,
+      "loss": 1.0613,
+      "step": 1544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 47.68030853342073,
+      "learning_rate": 1.9867053555722406e-05,
+      "loss": 0.9875,
+      "step": 1545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 48.412636224194074,
+      "learning_rate": 1.9866779342940354e-05,
+      "loss": 0.8982,
+      "step": 1546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 51.353499505569296,
+      "learning_rate": 1.986650484955287e-05,
+      "loss": 1.0029,
+      "step": 1547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 42.77044098691608,
+      "learning_rate": 1.986623007556776e-05,
+      "loss": 1.061,
+      "step": 1548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 50.20538420926135,
+      "learning_rate": 1.9865955020992837e-05,
+      "loss": 1.0156,
+      "step": 1549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 40.21028123330804,
+      "learning_rate": 1.9865679685835924e-05,
+      "loss": 0.8706,
+      "step": 1550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.32752246898588,
+      "learning_rate": 1.9865404070104857e-05,
+      "loss": 1.0248,
+      "step": 1551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 56.58082885518034,
+      "learning_rate": 1.9865128173807465e-05,
+      "loss": 1.0125,
+      "step": 1552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 56.47147435855777,
+      "learning_rate": 1.98648519969516e-05,
+      "loss": 1.0016,
+      "step": 1553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 56.82293543641934,
+      "learning_rate": 1.9864575539545115e-05,
+      "loss": 1.0697,
+      "step": 1554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 45.935070118898864,
+      "learning_rate": 1.986429880159587e-05,
+      "loss": 0.8945,
+      "step": 1555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 60.1545120424169,
+      "learning_rate": 1.986402178311174e-05,
+      "loss": 0.9949,
+      "step": 1556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 35.305800475886656,
+      "learning_rate": 1.98637444841006e-05,
+      "loss": 0.8959,
+      "step": 1557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 63.84105186343969,
+      "learning_rate": 1.9863466904570338e-05,
+      "loss": 1.0252,
+      "step": 1558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 35.387219020492026,
+      "learning_rate": 1.9863189044528847e-05,
+      "loss": 1.0151,
+      "step": 1559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 44.979185090525775,
+      "learning_rate": 1.986291090398403e-05,
+      "loss": 0.884,
+      "step": 1560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 52.08417338749213,
+      "learning_rate": 1.9862632482943796e-05,
+      "loss": 0.9133,
+      "step": 1561
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.82705564775315,
+      "learning_rate": 1.9862353781416062e-05,
+      "loss": 0.9456,
+      "step": 1562
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 34.2293129734218,
+      "learning_rate": 1.9862074799408757e-05,
+      "loss": 0.9187,
+      "step": 1563
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 47.2487681293869,
+      "learning_rate": 1.9861795536929814e-05,
+      "loss": 0.9817,
+      "step": 1564
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 49.303787520621896,
+      "learning_rate": 1.9861515993987174e-05,
+      "loss": 0.9399,
+      "step": 1565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 44.20282820503862,
+      "learning_rate": 1.986123617058879e-05,
+      "loss": 1.046,
+      "step": 1566
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 48.77816986828779,
+      "learning_rate": 1.9860956066742613e-05,
+      "loss": 0.9504,
+      "step": 1567
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 41.76376508745093,
+      "learning_rate": 1.986067568245662e-05,
+      "loss": 0.9209,
+      "step": 1568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.71512810652026,
+      "learning_rate": 1.9860395017738778e-05,
+      "loss": 0.9873,
+      "step": 1569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 46.7858115643454,
+      "learning_rate": 1.9860114072597064e-05,
+      "loss": 0.9548,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 76.79700697987452,
+      "learning_rate": 1.985983284703948e-05,
+      "loss": 0.8808,
+      "step": 1571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.46698582166789,
+      "learning_rate": 1.9859551341074016e-05,
+      "loss": 0.9631,
+      "step": 1572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 43.04571554072886,
+      "learning_rate": 1.985926955470868e-05,
+      "loss": 1.0045,
+      "step": 1573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 42.3313656253269,
+      "learning_rate": 1.9858987487951485e-05,
+      "loss": 0.9546,
+      "step": 1574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.69057774640319,
+      "learning_rate": 1.9858705140810453e-05,
+      "loss": 1.0479,
+      "step": 1575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 39.98785555181527,
+      "learning_rate": 1.985842251329361e-05,
+      "loss": 0.9079,
+      "step": 1576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.082799932109744,
+      "learning_rate": 1.9858139605409005e-05,
+      "loss": 0.9266,
+      "step": 1577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 53.863150698612664,
+      "learning_rate": 1.985785641716467e-05,
+      "loss": 1.0497,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 42.43285357920016,
+      "learning_rate": 1.985757294856867e-05,
+      "loss": 0.943,
+      "step": 1579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.70193681684137,
+      "learning_rate": 1.985728919962906e-05,
+      "loss": 1.1179,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 51.194176849757824,
+      "learning_rate": 1.985700517035391e-05,
+      "loss": 1.0204,
+      "step": 1581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 44.06815693842645,
+      "learning_rate": 1.98567208607513e-05,
+      "loss": 1.048,
+      "step": 1582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 45.93143493162479,
+      "learning_rate": 1.9856436270829315e-05,
+      "loss": 0.9787,
+      "step": 1583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 42.284593448549685,
+      "learning_rate": 1.985615140059605e-05,
+      "loss": 1.0349,
+      "step": 1584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 64.27477889436537,
+      "learning_rate": 1.9855866250059604e-05,
+      "loss": 1.0386,
+      "step": 1585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 39.91953714173107,
+      "learning_rate": 1.9855580819228087e-05,
+      "loss": 0.8822,
+      "step": 1586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 33.0395337619398,
+      "learning_rate": 1.985529510810962e-05,
+      "loss": 0.8879,
+      "step": 1587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 67.91335269667198,
+      "learning_rate": 1.9855009116712318e-05,
+      "loss": 0.9487,
+      "step": 1588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 53.71651660477267,
+      "learning_rate": 1.985472284504433e-05,
+      "loss": 1.0841,
+      "step": 1589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 46.50872108754875,
+      "learning_rate": 1.9854436293113782e-05,
+      "loss": 1.0188,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 37.862052775550495,
+      "learning_rate": 1.9854149460928832e-05,
+      "loss": 0.9186,
+      "step": 1591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 35.31672575655992,
+      "learning_rate": 1.9853862348497638e-05,
+      "loss": 0.8512,
+      "step": 1592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 39.00929633629753,
+      "learning_rate": 1.9853574955828364e-05,
+      "loss": 0.9175,
+      "step": 1593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 44.92534208767181,
+      "learning_rate": 1.985328728292918e-05,
+      "loss": 0.9584,
+      "step": 1594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 53.650840144409564,
+      "learning_rate": 1.985299932980827e-05,
+      "loss": 1.1497,
+      "step": 1595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 76.53675081695366,
+      "learning_rate": 1.9852711096473824e-05,
+      "loss": 0.926,
+      "step": 1596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 45.89690539331183,
+      "learning_rate": 1.9852422582934037e-05,
+      "loss": 0.8576,
+      "step": 1597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 62.45459077712162,
+      "learning_rate": 1.9852133789197117e-05,
+      "loss": 1.1572,
+      "step": 1598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 43.381039432596616,
+      "learning_rate": 1.9851844715271272e-05,
+      "loss": 0.9243,
+      "step": 1599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 46.6260970756179,
+      "learning_rate": 1.985155536116473e-05,
+      "loss": 0.8822,
+      "step": 1600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 47.12438435240897,
+      "learning_rate": 1.9851265726885716e-05,
+      "loss": 1.0672,
+      "step": 1601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.46888835767444,
+      "learning_rate": 1.9850975812442465e-05,
+      "loss": 0.8307,
+      "step": 1602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 39.543084634024915,
+      "learning_rate": 1.9850685617843225e-05,
+      "loss": 0.9324,
+      "step": 1603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 41.3244339937056,
+      "learning_rate": 1.985039514309625e-05,
+      "loss": 0.8865,
+      "step": 1604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 56.14856059351642,
+      "learning_rate": 1.9850104388209802e-05,
+      "loss": 0.9803,
+      "step": 1605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 36.94695335288888,
+      "learning_rate": 1.9849813353192142e-05,
+      "loss": 1.0192,
+      "step": 1606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 35.51141844381091,
+      "learning_rate": 1.9849522038051553e-05,
+      "loss": 0.9273,
+      "step": 1607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 54.05464168171182,
+      "learning_rate": 1.9849230442796324e-05,
+      "loss": 1.1402,
+      "step": 1608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 53.93622382942878,
+      "learning_rate": 1.9848938567434737e-05,
+      "loss": 1.0275,
+      "step": 1609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 36.63390959994239,
+      "learning_rate": 1.98486464119751e-05,
+      "loss": 0.9379,
+      "step": 1610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 337.27342150709495,
+      "learning_rate": 1.9848353976425723e-05,
+      "loss": 0.9563,
+      "step": 1611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.28634266513039,
+      "learning_rate": 1.984806126079492e-05,
+      "loss": 0.8863,
+      "step": 1612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 43.039708690601394,
+      "learning_rate": 1.9847768265091015e-05,
+      "loss": 0.9027,
+      "step": 1613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 30.78351638893042,
+      "learning_rate": 1.984747498932234e-05,
+      "loss": 0.8793,
+      "step": 1614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 37.57767853977247,
+      "learning_rate": 1.984718143349724e-05,
+      "loss": 0.8628,
+      "step": 1615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 34.7760899050046,
+      "learning_rate": 1.9846887597624054e-05,
+      "loss": 0.8626,
+      "step": 1616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 54.0788234873899,
+      "learning_rate": 1.984659348171115e-05,
+      "loss": 0.9805,
+      "step": 1617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 45.1158798754889,
+      "learning_rate": 1.9846299085766888e-05,
+      "loss": 0.9557,
+      "step": 1618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 43.28400949983851,
+      "learning_rate": 1.9846004409799636e-05,
+      "loss": 1.0,
+      "step": 1619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 42.120136104700755,
+      "learning_rate": 1.984570945381778e-05,
+      "loss": 0.9598,
+      "step": 1620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 45.50481855637323,
+      "learning_rate": 1.9845414217829708e-05,
+      "loss": 0.9435,
+      "step": 1621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 34.23769745924722,
+      "learning_rate": 1.9845118701843816e-05,
+      "loss": 0.7992,
+      "step": 1622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.59779017996274,
+      "learning_rate": 1.9844822905868507e-05,
+      "loss": 1.0056,
+      "step": 1623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.178819361845285,
+      "learning_rate": 1.984452682991219e-05,
+      "loss": 1.0546,
+      "step": 1624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 40.50387471113468,
+      "learning_rate": 1.984423047398329e-05,
+      "loss": 1.0065,
+      "step": 1625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 37.28201272700663,
+      "learning_rate": 1.9843933838090236e-05,
+      "loss": 0.9884,
+      "step": 1626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 40.044586271029445,
+      "learning_rate": 1.9843636922241463e-05,
+      "loss": 0.9166,
+      "step": 1627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 61.25503182275618,
+      "learning_rate": 1.984333972644541e-05,
+      "loss": 0.9738,
+      "step": 1628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 40.316925489010934,
+      "learning_rate": 1.984304225071054e-05,
+      "loss": 0.9195,
+      "step": 1629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 40.87647862049393,
+      "learning_rate": 1.98427444950453e-05,
+      "loss": 0.8491,
+      "step": 1630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 36.378142204768864,
+      "learning_rate": 1.9842446459458166e-05,
+      "loss": 0.86,
+      "step": 1631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 72.99920382130836,
+      "learning_rate": 1.9842148143957612e-05,
+      "loss": 0.9717,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 39.47603881174527,
+      "learning_rate": 1.9841849548552124e-05,
+      "loss": 0.9365,
+      "step": 1633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 42.02151581880287,
+      "learning_rate": 1.9841550673250192e-05,
+      "loss": 0.8489,
+      "step": 1634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.395293767734266,
+      "learning_rate": 1.9841251518060314e-05,
+      "loss": 1.0261,
+      "step": 1635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 51.723140315848916,
+      "learning_rate": 1.9840952082991e-05,
+      "loss": 0.9132,
+      "step": 1636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 46.98132650763192,
+      "learning_rate": 1.984065236805077e-05,
+      "loss": 1.0148,
+      "step": 1637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 45.07428220239787,
+      "learning_rate": 1.984035237324814e-05,
+      "loss": 1.0556,
+      "step": 1638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.304241197489894,
+      "learning_rate": 1.9840052098591645e-05,
+      "loss": 0.9632,
+      "step": 1639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 64.68359170431413,
+      "learning_rate": 1.9839751544089827e-05,
+      "loss": 1.0403,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 38.50127527985833,
+      "learning_rate": 1.983945070975123e-05,
+      "loss": 0.9556,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 36.259159709985305,
+      "learning_rate": 1.983914959558441e-05,
+      "loss": 0.9528,
+      "step": 1642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 42.83892722778579,
+      "learning_rate": 1.983884820159793e-05,
+      "loss": 0.8979,
+      "step": 1643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.236356339982294,
+      "learning_rate": 1.9838546527800364e-05,
+      "loss": 0.9492,
+      "step": 1644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 38.40440923254051,
+      "learning_rate": 1.983824457420029e-05,
+      "loss": 0.9825,
+      "step": 1645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.9054024404263,
+      "learning_rate": 1.983794234080629e-05,
+      "loss": 1.0005,
+      "step": 1646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 50.22344449313249,
+      "learning_rate": 1.9837639827626973e-05,
+      "loss": 0.9829,
+      "step": 1647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 40.53580621009136,
+      "learning_rate": 1.983733703467093e-05,
+      "loss": 0.8086,
+      "step": 1648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 39.44620114445755,
+      "learning_rate": 1.983703396194678e-05,
+      "loss": 0.8687,
+      "step": 1649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.07555683473164,
+      "learning_rate": 1.9836730609463134e-05,
+      "loss": 0.8297,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 35.947607168219214,
+      "learning_rate": 1.983642697722863e-05,
+      "loss": 0.7738,
+      "step": 1651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 46.19755258972223,
+      "learning_rate": 1.9836123065251892e-05,
+      "loss": 1.0746,
+      "step": 1652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 52.617399949039246,
+      "learning_rate": 1.9835818873541572e-05,
+      "loss": 0.9922,
+      "step": 1653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 45.14678520615656,
+      "learning_rate": 1.983551440210631e-05,
+      "loss": 1.0255,
+      "step": 1654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.28876322965585,
+      "learning_rate": 1.9835209650954775e-05,
+      "loss": 0.9192,
+      "step": 1655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 40.48596057562144,
+      "learning_rate": 1.9834904620095636e-05,
+      "loss": 1.0066,
+      "step": 1656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.01946709129277,
+      "learning_rate": 1.9834599309537557e-05,
+      "loss": 0.9557,
+      "step": 1657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 50.09857045549721,
+      "learning_rate": 1.9834293719289227e-05,
+      "loss": 0.9243,
+      "step": 1658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 38.43511603801268,
+      "learning_rate": 1.9833987849359338e-05,
+      "loss": 0.8327,
+      "step": 1659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 38.71686307667344,
+      "learning_rate": 1.9833681699756592e-05,
+      "loss": 0.9341,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.695163550991765,
+      "learning_rate": 1.9833375270489682e-05,
+      "loss": 0.88,
+      "step": 1661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 40.709217522011215,
+      "learning_rate": 1.983306856156734e-05,
+      "loss": 1.0482,
+      "step": 1662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.80193931783447,
+      "learning_rate": 1.9832761572998274e-05,
+      "loss": 0.9756,
+      "step": 1663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 43.12909260555721,
+      "learning_rate": 1.9832454304791224e-05,
+      "loss": 0.8489,
+      "step": 1664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 47.981349783869796,
+      "learning_rate": 1.9832146756954925e-05,
+      "loss": 0.9644,
+      "step": 1665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 48.8622240212435,
+      "learning_rate": 1.9831838929498123e-05,
+      "loss": 0.9338,
+      "step": 1666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 35.936696725243685,
+      "learning_rate": 1.983153082242957e-05,
+      "loss": 0.8861,
+      "step": 1667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 52.961184665245554,
+      "learning_rate": 1.983122243575804e-05,
+      "loss": 0.9675,
+      "step": 1668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 67.15031512677179,
+      "learning_rate": 1.9830913769492288e-05,
+      "loss": 0.9203,
+      "step": 1669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 54.10846002010569,
+      "learning_rate": 1.9830604823641103e-05,
+      "loss": 0.9205,
+      "step": 1670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 46.37855986532713,
+      "learning_rate": 1.9830295598213263e-05,
+      "loss": 0.9443,
+      "step": 1671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 77.9188208661316,
+      "learning_rate": 1.9829986093217575e-05,
+      "loss": 0.9333,
+      "step": 1672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 43.66993311623356,
+      "learning_rate": 1.9829676308662826e-05,
+      "loss": 1.0238,
+      "step": 1673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 42.35607363430691,
+      "learning_rate": 1.9829366244557837e-05,
+      "loss": 0.9256,
+      "step": 1674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 45.40989776209613,
+      "learning_rate": 1.982905590091142e-05,
+      "loss": 0.8402,
+      "step": 1675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 43.24668266275117,
+      "learning_rate": 1.98287452777324e-05,
+      "loss": 0.9851,
+      "step": 1676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 45.01278218088471,
+      "learning_rate": 1.982843437502962e-05,
+      "loss": 0.9468,
+      "step": 1677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 48.27159063013828,
+      "learning_rate": 1.982812319281191e-05,
+      "loss": 0.9445,
+      "step": 1678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 46.60483531219292,
+      "learning_rate": 1.982781173108813e-05,
+      "loss": 0.9569,
+      "step": 1679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 45.03057622041477,
+      "learning_rate": 1.982749998986713e-05,
+      "loss": 0.934,
+      "step": 1680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 44.350235220834755,
+      "learning_rate": 1.982718796915778e-05,
+      "loss": 0.9642,
+      "step": 1681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 67.65363379655446,
+      "learning_rate": 1.9826875668968956e-05,
+      "loss": 0.9148,
+      "step": 1682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.05341058518353,
+      "learning_rate": 1.9826563089309534e-05,
+      "loss": 1.0249,
+      "step": 1683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.94118077971714,
+      "learning_rate": 1.982625023018841e-05,
+      "loss": 0.9138,
+      "step": 1684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 42.750682803049514,
+      "learning_rate": 1.982593709161447e-05,
+      "loss": 1.0796,
+      "step": 1685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 48.00140636092497,
+      "learning_rate": 1.9825623673596634e-05,
+      "loss": 0.9952,
+      "step": 1686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 55.33247112795475,
+      "learning_rate": 1.9825309976143807e-05,
+      "loss": 0.9645,
+      "step": 1687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 38.83740173293034,
+      "learning_rate": 1.982499599926491e-05,
+      "loss": 1.0268,
+      "step": 1688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 35.98687761940548,
+      "learning_rate": 1.9824681742968876e-05,
+      "loss": 0.9125,
+      "step": 1689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 38.631831993874854,
+      "learning_rate": 1.9824367207264638e-05,
+      "loss": 0.9119,
+      "step": 1690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 35.050169015713614,
+      "learning_rate": 1.9824052392161148e-05,
+      "loss": 0.9374,
+      "step": 1691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 39.94028071601414,
+      "learning_rate": 1.982373729766735e-05,
+      "loss": 0.8649,
+      "step": 1692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 42.83210551633367,
+      "learning_rate": 1.9823421923792213e-05,
+      "loss": 0.9911,
+      "step": 1693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 45.060005954923206,
+      "learning_rate": 1.98231062705447e-05,
+      "loss": 0.9157,
+      "step": 1694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 50.340139220480566,
+      "learning_rate": 1.9822790337933795e-05,
+      "loss": 0.8621,
+      "step": 1695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.856578703602544,
+      "learning_rate": 1.9822474125968478e-05,
+      "loss": 1.0349,
+      "step": 1696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 76.38076681164203,
+      "learning_rate": 1.9822157634657742e-05,
+      "loss": 0.9695,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 47.36344129020158,
+      "learning_rate": 1.9821840864010588e-05,
+      "loss": 0.8906,
+      "step": 1698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 49.935038117989784,
+      "learning_rate": 1.9821523814036026e-05,
+      "loss": 0.9935,
+      "step": 1699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 48.97929220157807,
+      "learning_rate": 1.9821206484743073e-05,
+      "loss": 0.9395,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 49.27214119958381,
+      "learning_rate": 1.9820888876140752e-05,
+      "loss": 1.0474,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 65.097802137228,
+      "learning_rate": 1.9820570988238096e-05,
+      "loss": 0.9985,
+      "step": 1702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 46.058361390318495,
+      "learning_rate": 1.982025282104415e-05,
+      "loss": 0.9233,
+      "step": 1703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 39.78388569169659,
+      "learning_rate": 1.981993437456795e-05,
+      "loss": 0.9963,
+      "step": 1704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 49.53450248942034,
+      "learning_rate": 1.9819615648818568e-05,
+      "loss": 0.8803,
+      "step": 1705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 35.25607954855406,
+      "learning_rate": 1.9819296643805058e-05,
+      "loss": 0.9016,
+      "step": 1706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.11091123593415,
+      "learning_rate": 1.98189773595365e-05,
+      "loss": 0.8865,
+      "step": 1707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 41.04798510690841,
+      "learning_rate": 1.9818657796021963e-05,
+      "loss": 0.9362,
+      "step": 1708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 38.03819865595202,
+      "learning_rate": 1.9818337953270544e-05,
+      "loss": 0.9371,
+      "step": 1709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 43.47311197385301,
+      "learning_rate": 1.981801783129134e-05,
+      "loss": 1.0196,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 49.485191360162226,
+      "learning_rate": 1.981769743009345e-05,
+      "loss": 0.9777,
+      "step": 1711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 80.80068890495086,
+      "learning_rate": 1.9817376749685994e-05,
+      "loss": 0.9387,
+      "step": 1712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 40.723585709683896,
+      "learning_rate": 1.981705579007808e-05,
+      "loss": 0.9699,
+      "step": 1713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 41.70366901887561,
+      "learning_rate": 1.9816734551278843e-05,
+      "loss": 0.8824,
+      "step": 1714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 39.62995092708024,
+      "learning_rate": 1.981641303329742e-05,
+      "loss": 0.9249,
+      "step": 1715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 39.24164537332156,
+      "learning_rate": 1.981609123614295e-05,
+      "loss": 0.9806,
+      "step": 1716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 48.62489075145854,
+      "learning_rate": 1.981576915982459e-05,
+      "loss": 0.9267,
+      "step": 1717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.81997425201524,
+      "learning_rate": 1.9815446804351495e-05,
+      "loss": 0.9208,
+      "step": 1718
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 36.480236258186146,
+      "learning_rate": 1.9815124169732834e-05,
+      "loss": 0.9038,
+      "step": 1719
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 36.17602134981626,
+      "learning_rate": 1.9814801255977786e-05,
+      "loss": 0.9605,
+      "step": 1720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 42.30023811309386,
+      "learning_rate": 1.9814478063095533e-05,
+      "loss": 0.8569,
+      "step": 1721
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 42.40189052991463,
+      "learning_rate": 1.9814154591095262e-05,
+      "loss": 0.957,
+      "step": 1722
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 44.64986980607209,
+      "learning_rate": 1.9813830839986177e-05,
+      "loss": 1.0372,
+      "step": 1723
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 40.49525864833756,
+      "learning_rate": 1.9813506809777483e-05,
+      "loss": 0.9518,
+      "step": 1724
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 45.85804558927745,
+      "learning_rate": 1.9813182500478395e-05,
+      "loss": 0.963,
+      "step": 1725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 38.632699896452586,
+      "learning_rate": 1.981285791209814e-05,
+      "loss": 0.8807,
+      "step": 1726
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 44.37494187955677,
+      "learning_rate": 1.9812533044645945e-05,
+      "loss": 0.8942,
+      "step": 1727
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 55.827356887858755,
+      "learning_rate": 1.981220789813105e-05,
+      "loss": 0.9689,
+      "step": 1728
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 44.77321393461895,
+      "learning_rate": 1.9811882472562703e-05,
+      "loss": 0.9891,
+      "step": 1729
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 45.450398774805414,
+      "learning_rate": 1.9811556767950158e-05,
+      "loss": 0.959,
+      "step": 1730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 40.78983502404438,
+      "learning_rate": 1.981123078430268e-05,
+      "loss": 0.9873,
+      "step": 1731
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.21190271561343,
+      "learning_rate": 1.9810904521629534e-05,
+      "loss": 0.9254,
+      "step": 1732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.12415384362269,
+      "learning_rate": 1.9810577979940006e-05,
+      "loss": 0.9531,
+      "step": 1733
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 62.16838628812669,
+      "learning_rate": 1.9810251159243378e-05,
+      "loss": 0.9677,
+      "step": 1734
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 58.715794511753685,
+      "learning_rate": 1.9809924059548944e-05,
+      "loss": 0.8963,
+      "step": 1735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 50.82927517977719,
+      "learning_rate": 1.980959668086601e-05,
+      "loss": 0.9707,
+      "step": 1736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 57.41035984966417,
+      "learning_rate": 1.9809269023203887e-05,
+      "loss": 0.9986,
+      "step": 1737
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 44.8300194697324,
+      "learning_rate": 1.9808941086571888e-05,
+      "loss": 0.963,
+      "step": 1738
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 41.057644166473224,
+      "learning_rate": 1.9808612870979346e-05,
+      "loss": 0.9677,
+      "step": 1739
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 43.71647358353723,
+      "learning_rate": 1.9808284376435588e-05,
+      "loss": 0.9357,
+      "step": 1740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 44.36372069879879,
+      "learning_rate": 1.9807955602949962e-05,
+      "loss": 0.9604,
+      "step": 1741
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 35.11205735462317,
+      "learning_rate": 1.9807626550531816e-05,
+      "loss": 0.941,
+      "step": 1742
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 56.699466439079636,
+      "learning_rate": 1.9807297219190508e-05,
+      "loss": 0.9582,
+      "step": 1743
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 35.8317246090743,
+      "learning_rate": 1.9806967608935403e-05,
+      "loss": 1.0735,
+      "step": 1744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 35.51939343902189,
+      "learning_rate": 1.980663771977588e-05,
+      "loss": 0.9092,
+      "step": 1745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 33.994540533547976,
+      "learning_rate": 1.9806307551721313e-05,
+      "loss": 0.8177,
+      "step": 1746
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.11538248685494,
+      "learning_rate": 1.9805977104781096e-05,
+      "loss": 0.9039,
+      "step": 1747
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 51.293110082145624,
+      "learning_rate": 1.9805646378964626e-05,
+      "loss": 0.9569,
+      "step": 1748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 41.386192853465836,
+      "learning_rate": 1.9805315374281314e-05,
+      "loss": 1.0171,
+      "step": 1749
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 36.23961616303369,
+      "learning_rate": 1.9804984090740563e-05,
+      "loss": 1.0006,
+      "step": 1750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 48.66524965311384,
+      "learning_rate": 1.9804652528351806e-05,
+      "loss": 0.9728,
+      "step": 1751
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 35.282123889355056,
+      "learning_rate": 1.980432068712446e-05,
+      "loss": 0.943,
+      "step": 1752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 52.24901195668808,
+      "learning_rate": 1.9803988567067975e-05,
+      "loss": 0.9313,
+      "step": 1753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.562093888053475,
+      "learning_rate": 1.9803656168191786e-05,
+      "loss": 0.9532,
+      "step": 1754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 54.489956534052084,
+      "learning_rate": 1.9803323490505352e-05,
+      "loss": 1.0769,
+      "step": 1755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 46.63837319821641,
+      "learning_rate": 1.9802990534018135e-05,
+      "loss": 1.0727,
+      "step": 1756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 33.726027402505515,
+      "learning_rate": 1.9802657298739598e-05,
+      "loss": 0.852,
+      "step": 1757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 38.25440268952543,
+      "learning_rate": 1.9802323784679225e-05,
+      "loss": 0.9148,
+      "step": 1758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 37.65231296329484,
+      "learning_rate": 1.9801989991846495e-05,
+      "loss": 0.9636,
+      "step": 1759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 33.1051126124488,
+      "learning_rate": 1.9801655920250903e-05,
+      "loss": 0.9216,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 38.709645756514995,
+      "learning_rate": 1.9801321569901955e-05,
+      "loss": 0.8799,
+      "step": 1761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 36.64893859377587,
+      "learning_rate": 1.980098694080915e-05,
+      "loss": 0.9209,
+      "step": 1762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 38.9292039314872,
+      "learning_rate": 1.9800652032982013e-05,
+      "loss": 0.9164,
+      "step": 1763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 43.724593845616354,
+      "learning_rate": 1.9800316846430064e-05,
+      "loss": 0.986,
+      "step": 1764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 43.158158538404486,
+      "learning_rate": 1.979998138116284e-05,
+      "loss": 0.9782,
+      "step": 1765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 41.87863624499235,
+      "learning_rate": 1.9799645637189875e-05,
+      "loss": 0.9256,
+      "step": 1766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.42995954560792,
+      "learning_rate": 1.979930961452072e-05,
+      "loss": 0.8853,
+      "step": 1767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.202641393608694,
+      "learning_rate": 1.9798973313164934e-05,
+      "loss": 0.9457,
+      "step": 1768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 46.20067635614966,
+      "learning_rate": 1.979863673313208e-05,
+      "loss": 0.9912,
+      "step": 1769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 47.9784460749356,
+      "learning_rate": 1.979829987443173e-05,
+      "loss": 0.8467,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 47.37263936827087,
+      "learning_rate": 1.9797962737073456e-05,
+      "loss": 1.0047,
+      "step": 1771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.79054195011821,
+      "learning_rate": 1.9797625321066863e-05,
+      "loss": 1.0394,
+      "step": 1772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.93985911894659,
+      "learning_rate": 1.9797287626421534e-05,
+      "loss": 0.8211,
+      "step": 1773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 49.52994740214367,
+      "learning_rate": 1.9796949653147073e-05,
+      "loss": 0.8834,
+      "step": 1774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 39.834700776747106,
+      "learning_rate": 1.97966114012531e-05,
+      "loss": 0.9924,
+      "step": 1775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 39.575956273890654,
+      "learning_rate": 1.9796272870749225e-05,
+      "loss": 0.859,
+      "step": 1776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 35.329272019174134,
+      "learning_rate": 1.9795934061645084e-05,
+      "loss": 0.9272,
+      "step": 1777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 52.42919501464259,
+      "learning_rate": 1.9795594973950305e-05,
+      "loss": 0.9026,
+      "step": 1778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 33.60652868613978,
+      "learning_rate": 1.979525560767454e-05,
+      "loss": 0.8098,
+      "step": 1779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 32.747525827846275,
+      "learning_rate": 1.9794915962827427e-05,
+      "loss": 0.974,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.494488785138834,
+      "learning_rate": 1.9794576039418643e-05,
+      "loss": 0.9298,
+      "step": 1781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 38.30607594994956,
+      "learning_rate": 1.979423583745784e-05,
+      "loss": 0.9574,
+      "step": 1782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 40.54003755106643,
+      "learning_rate": 1.9793895356954704e-05,
+      "loss": 0.8778,
+      "step": 1783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 35.043525243194544,
+      "learning_rate": 1.9793554597918908e-05,
+      "loss": 0.9835,
+      "step": 1784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 39.31991866179985,
+      "learning_rate": 1.9793213560360153e-05,
+      "loss": 0.9365,
+      "step": 1785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.786207894048744,
+      "learning_rate": 1.9792872244288132e-05,
+      "loss": 0.9637,
+      "step": 1786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 45.19028877949398,
+      "learning_rate": 1.9792530649712552e-05,
+      "loss": 0.9052,
+      "step": 1787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 38.63288758119205,
+      "learning_rate": 1.979218877664313e-05,
+      "loss": 0.9724,
+      "step": 1788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.58923980598765,
+      "learning_rate": 1.9791846625089588e-05,
+      "loss": 0.9304,
+      "step": 1789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.36415143524066,
+      "learning_rate": 1.9791504195061655e-05,
+      "loss": 0.8728,
+      "step": 1790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 51.82043204889723,
+      "learning_rate": 1.9791161486569072e-05,
+      "loss": 1.1036,
+      "step": 1791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.53040410596488,
+      "learning_rate": 1.979081849962158e-05,
+      "loss": 0.9018,
+      "step": 1792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 40.238589707142076,
+      "learning_rate": 1.979047523422894e-05,
+      "loss": 0.9751,
+      "step": 1793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.82341539777168,
+      "learning_rate": 1.9790131690400913e-05,
+      "loss": 1.0215,
+      "step": 1794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 36.0302319444882,
+      "learning_rate": 1.9789787868147265e-05,
+      "loss": 0.8062,
+      "step": 1795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 53.80192961866265,
+      "learning_rate": 1.9789443767477778e-05,
+      "loss": 0.8995,
+      "step": 1796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 47.34068837857535,
+      "learning_rate": 1.978909938840224e-05,
+      "loss": 1.0194,
+      "step": 1797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 44.59506840083555,
+      "learning_rate": 1.9788754730930436e-05,
+      "loss": 0.8828,
+      "step": 1798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 40.196033948053206,
+      "learning_rate": 1.9788409795072175e-05,
+      "loss": 0.9254,
+      "step": 1799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 34.88457456711238,
+      "learning_rate": 1.978806458083727e-05,
+      "loss": 0.9773,
+      "step": 1800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 33.60009841549743,
+      "learning_rate": 1.978771908823553e-05,
+      "loss": 0.8718,
+      "step": 1801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 50.672181355846845,
+      "learning_rate": 1.978737331727679e-05,
+      "loss": 0.9225,
+      "step": 1802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 41.39115003380261,
+      "learning_rate": 1.9787027267970873e-05,
+      "loss": 0.9292,
+      "step": 1803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 107.52987703723416,
+      "learning_rate": 1.9786680940327626e-05,
+      "loss": 0.9053,
+      "step": 1804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.26066411744663,
+      "learning_rate": 1.97863343343569e-05,
+      "loss": 1.0228,
+      "step": 1805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.89156130687658,
+      "learning_rate": 1.978598745006855e-05,
+      "loss": 0.9954,
+      "step": 1806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 41.26392981540295,
+      "learning_rate": 1.978564028747244e-05,
+      "loss": 0.9081,
+      "step": 1807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 48.96046089365259,
+      "learning_rate": 1.9785292846578446e-05,
+      "loss": 0.9573,
+      "step": 1808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.535087902789456,
+      "learning_rate": 1.9784945127396446e-05,
+      "loss": 1.0458,
+      "step": 1809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 46.94636372712701,
+      "learning_rate": 1.9784597129936333e-05,
+      "loss": 0.9753,
+      "step": 1810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 40.69566910708852,
+      "learning_rate": 1.9784248854208e-05,
+      "loss": 0.924,
+      "step": 1811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 56.31508678799105,
+      "learning_rate": 1.9783900300221354e-05,
+      "loss": 1.0262,
+      "step": 1812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 38.72887496708431,
+      "learning_rate": 1.9783551467986307e-05,
+      "loss": 0.849,
+      "step": 1813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 49.3645799047476,
+      "learning_rate": 1.978320235751278e-05,
+      "loss": 0.9026,
+      "step": 1814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 44.63192236927664,
+      "learning_rate": 1.97828529688107e-05,
+      "loss": 1.0499,
+      "step": 1815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 49.86607585984535,
+      "learning_rate": 1.9782503301890004e-05,
+      "loss": 0.9848,
+      "step": 1816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.20259019893455,
+      "learning_rate": 1.9782153356760635e-05,
+      "loss": 0.9633,
+      "step": 1817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 37.85322890574863,
+      "learning_rate": 1.978180313343255e-05,
+      "loss": 0.8405,
+      "step": 1818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 48.238733676101916,
+      "learning_rate": 1.9781452631915705e-05,
+      "loss": 0.7848,
+      "step": 1819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 35.91582421128307,
+      "learning_rate": 1.9781101852220074e-05,
+      "loss": 0.8645,
+      "step": 1820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 40.32475629095515,
+      "learning_rate": 1.978075079435562e-05,
+      "loss": 0.9407,
+      "step": 1821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 54.65122142623916,
+      "learning_rate": 1.978039945833234e-05,
+      "loss": 1.0465,
+      "step": 1822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 50.037967728871415,
+      "learning_rate": 1.978004784416022e-05,
+      "loss": 0.9396,
+      "step": 1823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 45.96185707470506,
+      "learning_rate": 1.977969595184926e-05,
+      "loss": 0.8814,
+      "step": 1824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 35.67392513952214,
+      "learning_rate": 1.9779343781409466e-05,
+      "loss": 0.9298,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 32.980975009710846,
+      "learning_rate": 1.9778991332850857e-05,
+      "loss": 0.9428,
+      "step": 1826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 44.39705990774094,
+      "learning_rate": 1.9778638606183453e-05,
+      "loss": 0.9029,
+      "step": 1827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 52.40642420649363,
+      "learning_rate": 1.977828560141729e-05,
+      "loss": 0.9454,
+      "step": 1828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 47.589880402140786,
+      "learning_rate": 1.9777932318562407e-05,
+      "loss": 0.9022,
+      "step": 1829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 44.69470546758939,
+      "learning_rate": 1.9777578757628845e-05,
+      "loss": 0.9324,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 57.29275848689243,
+      "learning_rate": 1.9777224918626662e-05,
+      "loss": 0.8513,
+      "step": 1831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 42.93817227543873,
+      "learning_rate": 1.9776870801565924e-05,
+      "loss": 1.1094,
+      "step": 1832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 50.818323015214446,
+      "learning_rate": 1.97765164064567e-05,
+      "loss": 1.0278,
+      "step": 1833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.5545416935376,
+      "learning_rate": 1.977616173330907e-05,
+      "loss": 0.9639,
+      "step": 1834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 35.16323419666216,
+      "learning_rate": 1.9775806782133114e-05,
+      "loss": 0.997,
+      "step": 1835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 43.112566098068235,
+      "learning_rate": 1.9775451552938937e-05,
+      "loss": 1.0435,
+      "step": 1836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 37.96550086605016,
+      "learning_rate": 1.9775096045736634e-05,
+      "loss": 0.878,
+      "step": 1837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 40.73131630567112,
+      "learning_rate": 1.9774740260536318e-05,
+      "loss": 0.9524,
+      "step": 1838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.942704142089895,
+      "learning_rate": 1.977438419734811e-05,
+      "loss": 0.9292,
+      "step": 1839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 39.986444870627395,
+      "learning_rate": 1.977402785618213e-05,
+      "loss": 0.8652,
+      "step": 1840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 52.00867989429778,
+      "learning_rate": 1.9773671237048514e-05,
+      "loss": 1.0505,
+      "step": 1841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 49.038048724455045,
+      "learning_rate": 1.977331433995741e-05,
+      "loss": 0.9635,
+      "step": 1842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 40.255514750394305,
+      "learning_rate": 1.977295716491896e-05,
+      "loss": 0.9964,
+      "step": 1843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 34.56806601946198,
+      "learning_rate": 1.9772599711943323e-05,
+      "loss": 0.9023,
+      "step": 1844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 45.245515643311684,
+      "learning_rate": 1.977224198104067e-05,
+      "loss": 1.0623,
+      "step": 1845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 35.02723708976195,
+      "learning_rate": 1.9771883972221174e-05,
+      "loss": 0.8413,
+      "step": 1846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 38.63355846089839,
+      "learning_rate": 1.977152568549501e-05,
+      "loss": 0.8698,
+      "step": 1847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 44.6857422596353,
+      "learning_rate": 1.9771167120872375e-05,
+      "loss": 0.9907,
+      "step": 1848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.70834985606311,
+      "learning_rate": 1.977080827836346e-05,
+      "loss": 0.8756,
+      "step": 1849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 45.19874642794346,
+      "learning_rate": 1.9770449157978475e-05,
+      "loss": 0.913,
+      "step": 1850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 34.99374981318271,
+      "learning_rate": 1.9770089759727634e-05,
+      "loss": 0.8868,
+      "step": 1851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.760455762552645,
+      "learning_rate": 1.976973008362115e-05,
+      "loss": 0.9317,
+      "step": 1852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 45.48384593471152,
+      "learning_rate": 1.976937012966926e-05,
+      "loss": 0.9695,
+      "step": 1853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 48.33446867804187,
+      "learning_rate": 1.97690098978822e-05,
+      "loss": 1.0448,
+      "step": 1854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.265571862865066,
+      "learning_rate": 1.9768649388270213e-05,
+      "loss": 0.9279,
+      "step": 1855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 42.90946286057533,
+      "learning_rate": 1.976828860084355e-05,
+      "loss": 0.99,
+      "step": 1856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 48.29161344738656,
+      "learning_rate": 1.9767927535612473e-05,
+      "loss": 0.8288,
+      "step": 1857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 49.37748925416795,
+      "learning_rate": 1.9767566192587253e-05,
+      "loss": 0.9038,
+      "step": 1858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 39.17238116916856,
+      "learning_rate": 1.9767204571778163e-05,
+      "loss": 0.9901,
+      "step": 1859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 38.253985708429376,
+      "learning_rate": 1.976684267319549e-05,
+      "loss": 0.8612,
+      "step": 1860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 39.41733369740051,
+      "learning_rate": 1.9766480496849527e-05,
+      "loss": 0.9139,
+      "step": 1861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.460497318832985,
+      "learning_rate": 1.9766118042750566e-05,
+      "loss": 0.8895,
+      "step": 1862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.877784909690014,
+      "learning_rate": 1.9765755310908925e-05,
+      "loss": 1.1487,
+      "step": 1863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.38544260682926,
+      "learning_rate": 1.9765392301334912e-05,
+      "loss": 1.0468,
+      "step": 1864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.99550727248092,
+      "learning_rate": 1.976502901403886e-05,
+      "loss": 0.9349,
+      "step": 1865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 50.08700329996261,
+      "learning_rate": 1.9764665449031093e-05,
+      "loss": 0.9467,
+      "step": 1866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.56541888413657,
+      "learning_rate": 1.976430160632195e-05,
+      "loss": 0.9694,
+      "step": 1867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.76984311719351,
+      "learning_rate": 1.9763937485921785e-05,
+      "loss": 0.7851,
+      "step": 1868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 55.81788505470933,
+      "learning_rate": 1.976357308784095e-05,
+      "loss": 1.0741,
+      "step": 1869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 45.124794069279936,
+      "learning_rate": 1.9763208412089802e-05,
+      "loss": 1.0091,
+      "step": 1870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 37.66348084477572,
+      "learning_rate": 1.9762843458678723e-05,
+      "loss": 0.9413,
+      "step": 1871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.401382622590305,
+      "learning_rate": 1.976247822761809e-05,
+      "loss": 0.9234,
+      "step": 1872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 60.57597890008608,
+      "learning_rate": 1.976211271891828e-05,
+      "loss": 0.9192,
+      "step": 1873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.028390162368034,
+      "learning_rate": 1.97617469325897e-05,
+      "loss": 1.0098,
+      "step": 1874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 37.47312894658428,
+      "learning_rate": 1.9761380868642745e-05,
+      "loss": 0.8994,
+      "step": 1875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 47.164477033248325,
+      "learning_rate": 1.9761014527087828e-05,
+      "loss": 0.9369,
+      "step": 1876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 44.52895113716544,
+      "learning_rate": 1.9760647907935372e-05,
+      "loss": 0.8989,
+      "step": 1877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 37.13303300859087,
+      "learning_rate": 1.9760281011195793e-05,
+      "loss": 0.8359,
+      "step": 1878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 36.95113298075448,
+      "learning_rate": 1.9759913836879535e-05,
+      "loss": 0.941,
+      "step": 1879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 42.46078678164014,
+      "learning_rate": 1.9759546384997042e-05,
+      "loss": 0.9186,
+      "step": 1880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 72.88052262830676,
+      "learning_rate": 1.975917865555875e-05,
+      "loss": 1.0043,
+      "step": 1881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 42.04804139945273,
+      "learning_rate": 1.9758810648575132e-05,
+      "loss": 1.038,
+      "step": 1882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 60.41519965484445,
+      "learning_rate": 1.9758442364056643e-05,
+      "loss": 0.9282,
+      "step": 1883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 43.20589250250855,
+      "learning_rate": 1.975807380201377e-05,
+      "loss": 0.945,
+      "step": 1884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 65.89660691193055,
+      "learning_rate": 1.9757704962456978e-05,
+      "loss": 0.9207,
+      "step": 1885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 37.4724543778879,
+      "learning_rate": 1.9757335845396768e-05,
+      "loss": 0.9491,
+      "step": 1886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 41.78192619740918,
+      "learning_rate": 1.9756966450843635e-05,
+      "loss": 0.929,
+      "step": 1887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 61.441506457933734,
+      "learning_rate": 1.9756596778808085e-05,
+      "loss": 1.0258,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 33.758096829262904,
+      "learning_rate": 1.9756226829300628e-05,
+      "loss": 0.9755,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 39.98274493574945,
+      "learning_rate": 1.9755856602331788e-05,
+      "loss": 0.9225,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 36.10041797031411,
+      "learning_rate": 1.9755486097912092e-05,
+      "loss": 0.8704,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 37.93881717261802,
+      "learning_rate": 1.9755115316052076e-05,
+      "loss": 0.9163,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 33.56644734731152,
+      "learning_rate": 1.9754744256762293e-05,
+      "loss": 0.8961,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 35.3499437296634,
+      "learning_rate": 1.9754372920053284e-05,
+      "loss": 0.8846,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 37.98275942016644,
+      "learning_rate": 1.975400130593562e-05,
+      "loss": 0.9502,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 31.759292423539186,
+      "learning_rate": 1.975362941441986e-05,
+      "loss": 0.9017,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 39.52984762908361,
+      "learning_rate": 1.975325724551659e-05,
+      "loss": 0.8822,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 43.90042066329103,
+      "learning_rate": 1.9752884799236387e-05,
+      "loss": 0.897,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 56.08513748229634,
+      "learning_rate": 1.9752512075589844e-05,
+      "loss": 1.0003,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 49.33050919946416,
+      "learning_rate": 1.9752139074587563e-05,
+      "loss": 0.9349,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 48.34597796858738,
+      "learning_rate": 1.9751765796240153e-05,
+      "loss": 1.0351,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 50.56696906220442,
+      "learning_rate": 1.9751392240558228e-05,
+      "loss": 0.9687,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 49.284464743967085,
+      "learning_rate": 1.9751018407552417e-05,
+      "loss": 0.8872,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 33.03146323980614,
+      "learning_rate": 1.9750644297233342e-05,
+      "loss": 0.8669,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.374799105885145,
+      "learning_rate": 1.9750269909611645e-05,
+      "loss": 0.9686,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 45.9373777426825,
+      "learning_rate": 1.9749895244697978e-05,
+      "loss": 1.0121,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 38.81906487664713,
+      "learning_rate": 1.9749520302502993e-05,
+      "loss": 1.0238,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.41272497248455,
+      "learning_rate": 1.9749145083037354e-05,
+      "loss": 1.0516,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 32.512316295775044,
+      "learning_rate": 1.9748769586311732e-05,
+      "loss": 0.8306,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 44.558009024336926,
+      "learning_rate": 1.9748393812336807e-05,
+      "loss": 0.9671,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 67.23676805184691,
+      "learning_rate": 1.9748017761123265e-05,
+      "loss": 0.9305,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 48.46424286950186,
+      "learning_rate": 1.97476414326818e-05,
+      "loss": 1.0829,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.76077248391258,
+      "learning_rate": 1.9747264827023113e-05,
+      "loss": 0.8792,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 42.9892044377281,
+      "learning_rate": 1.974688794415792e-05,
+      "loss": 0.9818,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 40.22999805531034,
+      "learning_rate": 1.9746510784096928e-05,
+      "loss": 0.8502,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 50.640006127794024,
+      "learning_rate": 1.974613334685088e-05,
+      "loss": 1.0101,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 34.245714490323586,
+      "learning_rate": 1.9745755632430496e-05,
+      "loss": 0.9321,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 48.270564944405116,
+      "learning_rate": 1.9745377640846523e-05,
+      "loss": 0.8998,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 37.44048716133916,
+      "learning_rate": 1.974499937210971e-05,
+      "loss": 0.9287,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.009388960603204,
+      "learning_rate": 1.974462082623082e-05,
+      "loss": 0.8731,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 37.00672556348166,
+      "learning_rate": 1.9744242003220607e-05,
+      "loss": 0.8791,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 31.88182871356302,
+      "learning_rate": 1.9743862903089857e-05,
+      "loss": 0.82,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 37.719594068967346,
+      "learning_rate": 1.9743483525849345e-05,
+      "loss": 0.8892,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 36.98435121897298,
+      "learning_rate": 1.974310387150986e-05,
+      "loss": 0.8798,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 41.895170728520206,
+      "learning_rate": 1.9742723940082203e-05,
+      "loss": 1.1022,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 35.69471375198463,
+      "learning_rate": 1.9742343731577173e-05,
+      "loss": 0.9275,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 39.99129369876711,
+      "learning_rate": 1.9741963246005592e-05,
+      "loss": 0.9458,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.18039106653632,
+      "learning_rate": 1.974158248337827e-05,
+      "loss": 1.0066,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 40.46904382193989,
+      "learning_rate": 1.9741201443706042e-05,
+      "loss": 0.8932,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 53.68908753895605,
+      "learning_rate": 1.9740820126999743e-05,
+      "loss": 1.0024,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 57.28493393426645,
+      "learning_rate": 1.9740438533270218e-05,
+      "loss": 0.9902,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 38.137722109632655,
+      "learning_rate": 1.9740056662528318e-05,
+      "loss": 0.835,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 42.63993974123163,
+      "learning_rate": 1.9739674514784906e-05,
+      "loss": 0.9698,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 39.37924650393584,
+      "learning_rate": 1.973929209005085e-05,
+      "loss": 0.8932,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 51.385783286661734,
+      "learning_rate": 1.9738909388337023e-05,
+      "loss": 1.0645,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 34.54050787848313,
+      "learning_rate": 1.973852640965431e-05,
+      "loss": 0.8307,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 40.19176837507004,
+      "learning_rate": 1.9738143154013603e-05,
+      "loss": 0.8069,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 50.764294678225404,
+      "learning_rate": 1.9737759621425797e-05,
+      "loss": 0.9789,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 44.3883431051888,
+      "learning_rate": 1.973737581190181e-05,
+      "loss": 0.9348,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 53.063190878064965,
+      "learning_rate": 1.973699172545255e-05,
+      "loss": 0.975,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 46.206733266005216,
+      "learning_rate": 1.973660736208894e-05,
+      "loss": 1.0274,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 51.352604107816916,
+      "learning_rate": 1.973622272182191e-05,
+      "loss": 0.8246,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 40.960373686337554,
+      "learning_rate": 1.9735837804662404e-05,
+      "loss": 0.9804,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 50.81787641640241,
+      "learning_rate": 1.9735452610621366e-05,
+      "loss": 0.9508,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 48.85708037948439,
+      "learning_rate": 1.9735067139709753e-05,
+      "loss": 0.9797,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 34.607868080829945,
+      "learning_rate": 1.973468139193852e-05,
+      "loss": 0.8169,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 50.930336992686975,
+      "learning_rate": 1.9734295367318648e-05,
+      "loss": 0.8415,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 44.303794253161215,
+      "learning_rate": 1.9733909065861107e-05,
+      "loss": 0.9385,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 70.30361073230313,
+      "learning_rate": 1.9733522487576884e-05,
+      "loss": 1.0028,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 40.85268797266097,
+      "learning_rate": 1.973313563247698e-05,
+      "loss": 0.8854,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 39.618791611290476,
+      "learning_rate": 1.9732748500572395e-05,
+      "loss": 0.9445,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 40.396157847545865,
+      "learning_rate": 1.973236109187413e-05,
+      "loss": 0.8853,
+      "step": 1952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 41.47911550511351,
+      "learning_rate": 1.973197340639321e-05,
+      "loss": 0.8913,
+      "step": 1953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 41.64481315807208,
+      "learning_rate": 1.973158544414066e-05,
+      "loss": 0.853,
+      "step": 1954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 44.617602929352294,
+      "learning_rate": 1.9731197205127514e-05,
+      "loss": 1.2224,
+      "step": 1955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.06326067663211,
+      "learning_rate": 1.973080868936481e-05,
+      "loss": 0.9848,
+      "step": 1956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 43.103353651567716,
+      "learning_rate": 1.9730419896863603e-05,
+      "loss": 1.0348,
+      "step": 1957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.36992129833771,
+      "learning_rate": 1.9730030827634942e-05,
+      "loss": 1.0377,
+      "step": 1958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 34.71831325390212,
+      "learning_rate": 1.9729641481689897e-05,
+      "loss": 0.9436,
+      "step": 1959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 33.64082780733471,
+      "learning_rate": 1.972925185903954e-05,
+      "loss": 0.9001,
+      "step": 1960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 30.720587317470184,
+      "learning_rate": 1.9728861959694957e-05,
+      "loss": 0.9045,
+      "step": 1961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 46.575177058130606,
+      "learning_rate": 1.9728471783667226e-05,
+      "loss": 0.9736,
+      "step": 1962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 35.491841026729354,
+      "learning_rate": 1.9728081330967448e-05,
+      "loss": 0.8981,
+      "step": 1963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 47.999962786819026,
+      "learning_rate": 1.972769060160673e-05,
+      "loss": 0.9982,
+      "step": 1964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 39.0769118921875,
+      "learning_rate": 1.9727299595596178e-05,
+      "loss": 0.9957,
+      "step": 1965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 35.098600247721635,
+      "learning_rate": 1.9726908312946918e-05,
+      "loss": 0.9297,
+      "step": 1966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 49.09664044045236,
+      "learning_rate": 1.9726516753670078e-05,
+      "loss": 0.9448,
+      "step": 1967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 53.715551044755635,
+      "learning_rate": 1.972612491777679e-05,
+      "loss": 0.9061,
+      "step": 1968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 46.19875480410515,
+      "learning_rate": 1.9725732805278198e-05,
+      "loss": 0.9533,
+      "step": 1969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 35.68537715804488,
+      "learning_rate": 1.9725340416185456e-05,
+      "loss": 0.8787,
+      "step": 1970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 37.877595957307875,
+      "learning_rate": 1.9724947750509718e-05,
+      "loss": 0.7946,
+      "step": 1971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 40.203205984770136,
+      "learning_rate": 1.9724554808262157e-05,
+      "loss": 0.9599,
+      "step": 1972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 39.21478588237914,
+      "learning_rate": 1.9724161589453948e-05,
+      "loss": 1.0299,
+      "step": 1973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 61.90784106455745,
+      "learning_rate": 1.9723768094096268e-05,
+      "loss": 0.8887,
+      "step": 1974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 39.42000424803097,
+      "learning_rate": 1.9723374322200315e-05,
+      "loss": 0.9657,
+      "step": 1975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.31780509146242,
+      "learning_rate": 1.9722980273777284e-05,
+      "loss": 0.8386,
+      "step": 1976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 38.57185585375677,
+      "learning_rate": 1.9722585948838378e-05,
+      "loss": 1.0077,
+      "step": 1977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 39.46848055399471,
+      "learning_rate": 1.972219134739482e-05,
+      "loss": 0.9601,
+      "step": 1978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 48.116374974390126,
+      "learning_rate": 1.9721796469457827e-05,
+      "loss": 0.9946,
+      "step": 1979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 44.17085010701091,
+      "learning_rate": 1.9721401315038628e-05,
+      "loss": 0.8805,
+      "step": 1980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.299012481245924,
+      "learning_rate": 1.972100588414846e-05,
+      "loss": 0.855,
+      "step": 1981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 43.5174317017398,
+      "learning_rate": 1.9720610176798574e-05,
+      "loss": 0.8918,
+      "step": 1982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 51.8968854950909,
+      "learning_rate": 1.972021419300022e-05,
+      "loss": 0.9368,
+      "step": 1983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.430300746609774,
+      "learning_rate": 1.9719817932764662e-05,
+      "loss": 0.8742,
+      "step": 1984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 72.67765442285433,
+      "learning_rate": 1.9719421396103167e-05,
+      "loss": 1.0295,
+      "step": 1985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 50.37256198028609,
+      "learning_rate": 1.9719024583027012e-05,
+      "loss": 0.9543,
+      "step": 1986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 45.30318504614231,
+      "learning_rate": 1.9718627493547486e-05,
+      "loss": 0.9129,
+      "step": 1987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 37.14742509523343,
+      "learning_rate": 1.9718230127675877e-05,
+      "loss": 0.9244,
+      "step": 1988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 54.687560760464294,
+      "learning_rate": 1.971783248542349e-05,
+      "loss": 0.95,
+      "step": 1989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 41.25754585028253,
+      "learning_rate": 1.9717434566801633e-05,
+      "loss": 0.9029,
+      "step": 1990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 37.653898224262974,
+      "learning_rate": 1.9717036371821615e-05,
+      "loss": 0.9133,
+      "step": 1991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 55.1738652961733,
+      "learning_rate": 1.9716637900494775e-05,
+      "loss": 0.9875,
+      "step": 1992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 34.393459749184196,
+      "learning_rate": 1.9716239152832434e-05,
+      "loss": 0.8719,
+      "step": 1993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 47.67177066397713,
+      "learning_rate": 1.9715840128845934e-05,
+      "loss": 0.9265,
+      "step": 1994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 59.22658435059801,
+      "learning_rate": 1.9715440828546626e-05,
+      "loss": 0.9704,
+      "step": 1995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 60.58443761428696,
+      "learning_rate": 1.9715041251945863e-05,
+      "loss": 0.8751,
+      "step": 1996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 48.81361109486187,
+      "learning_rate": 1.9714641399055007e-05,
+      "loss": 0.9894,
+      "step": 1997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 47.79263926378049,
+      "learning_rate": 1.9714241269885436e-05,
+      "loss": 0.9663,
+      "step": 1998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 47.22702153363782,
+      "learning_rate": 1.9713840864448528e-05,
+      "loss": 0.8274,
+      "step": 1999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.4814052069914,
+      "learning_rate": 1.9713440182755665e-05,
+      "loss": 0.9337,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.70858415979683,
+      "learning_rate": 1.9713039224818247e-05,
+      "loss": 0.9371,
+      "step": 2001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 51.164359231211584,
+      "learning_rate": 1.9712637990647674e-05,
+      "loss": 0.8898,
+      "step": 2002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.76731106077269,
+      "learning_rate": 1.971223648025536e-05,
+      "loss": 0.8404,
+      "step": 2003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 38.153967483696015,
+      "learning_rate": 1.9711834693652722e-05,
+      "loss": 0.8983,
+      "step": 2004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 43.646265519702546,
+      "learning_rate": 1.9711432630851182e-05,
+      "loss": 1.0081,
+      "step": 2005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 43.94505471115795,
+      "learning_rate": 1.9711030291862187e-05,
+      "loss": 0.9461,
+      "step": 2006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.04147980720624,
+      "learning_rate": 1.9710627676697167e-05,
+      "loss": 0.8983,
+      "step": 2007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 53.559882199070266,
+      "learning_rate": 1.9710224785367575e-05,
+      "loss": 1.1164,
+      "step": 2008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 46.8973146348978,
+      "learning_rate": 1.9709821617884873e-05,
+      "loss": 0.901,
+      "step": 2009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 38.96814707126717,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.8888,
+      "step": 2010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.3148179553728,
+      "learning_rate": 1.9709014454506e-05,
+      "loss": 0.8118,
+      "step": 2011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.38023328004317,
+      "learning_rate": 1.9708610458632783e-05,
+      "loss": 0.9609,
+      "step": 2012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 37.10323199052147,
+      "learning_rate": 1.970820618665237e-05,
+      "loss": 0.8572,
+      "step": 2013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 33.7614489315841,
+      "learning_rate": 1.9707801638576246e-05,
+      "loss": 0.8661,
+      "step": 2014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 44.883302738709524,
+      "learning_rate": 1.9707396814415927e-05,
+      "loss": 0.9252,
+      "step": 2015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 42.290801751949836,
+      "learning_rate": 1.970699171418292e-05,
+      "loss": 1.0246,
+      "step": 2016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 44.8175933635448,
+      "learning_rate": 1.9706586337888744e-05,
+      "loss": 0.9543,
+      "step": 2017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 56.43733332747713,
+      "learning_rate": 1.9706180685544933e-05,
+      "loss": 1.0058,
+      "step": 2018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 43.633113740514645,
+      "learning_rate": 1.9705774757163022e-05,
+      "loss": 0.9564,
+      "step": 2019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 40.866789678592795,
+      "learning_rate": 1.9705368552754554e-05,
+      "loss": 0.969,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.88047300002952,
+      "learning_rate": 1.970496207233108e-05,
+      "loss": 0.9596,
+      "step": 2021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 59.52697253141213,
+      "learning_rate": 1.9704555315904164e-05,
+      "loss": 0.9337,
+      "step": 2022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 33.99313525012066,
+      "learning_rate": 1.9704148283485374e-05,
+      "loss": 0.9242,
+      "step": 2023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 46.39464793969858,
+      "learning_rate": 1.9703740975086282e-05,
+      "loss": 0.8645,
+      "step": 2024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 52.01593940619217,
+      "learning_rate": 1.970333339071847e-05,
+      "loss": 0.9439,
+      "step": 2025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 34.91462731987516,
+      "learning_rate": 1.9702925530393535e-05,
+      "loss": 0.9078,
+      "step": 2026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 44.53916019254341,
+      "learning_rate": 1.970251739412307e-05,
+      "loss": 0.881,
+      "step": 2027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 41.52084460196922,
+      "learning_rate": 1.9702108981918693e-05,
+      "loss": 1.0024,
+      "step": 2028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 45.795405004746904,
+      "learning_rate": 1.970170029379201e-05,
+      "loss": 0.95,
+      "step": 2029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 39.90568744020936,
+      "learning_rate": 1.9701291329754645e-05,
+      "loss": 0.919,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 33.53367628720997,
+      "learning_rate": 1.970088208981823e-05,
+      "loss": 0.8084,
+      "step": 2031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 44.075214970108256,
+      "learning_rate": 1.9700472573994403e-05,
+      "loss": 0.8575,
+      "step": 2032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 63.160190336666595,
+      "learning_rate": 1.970006278229481e-05,
+      "loss": 1.0784,
+      "step": 2033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.18114048250084,
+      "learning_rate": 1.9699652714731106e-05,
+      "loss": 0.9827,
+      "step": 2034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 40.17951553785599,
+      "learning_rate": 1.9699242371314952e-05,
+      "loss": 0.8823,
+      "step": 2035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 40.42730777242091,
+      "learning_rate": 1.9698831752058023e-05,
+      "loss": 0.836,
+      "step": 2036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 42.444443237233656,
+      "learning_rate": 1.969842085697199e-05,
+      "loss": 0.9646,
+      "step": 2037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.57135526666538,
+      "learning_rate": 1.969800968606854e-05,
+      "loss": 0.8723,
+      "step": 2038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 49.4997331342343,
+      "learning_rate": 1.9697598239359368e-05,
+      "loss": 0.9694,
+      "step": 2039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 68.45632750145356,
+      "learning_rate": 1.9697186516856177e-05,
+      "loss": 1.2225,
+      "step": 2040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 32.864631693099646,
+      "learning_rate": 1.9696774518570673e-05,
+      "loss": 0.8448,
+      "step": 2041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 44.79382924854905,
+      "learning_rate": 1.9696362244514576e-05,
+      "loss": 0.9191,
+      "step": 2042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 42.71543478502351,
+      "learning_rate": 1.9695949694699604e-05,
+      "loss": 1.0237,
+      "step": 2043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 52.36659920356447,
+      "learning_rate": 1.96955368691375e-05,
+      "loss": 1.0277,
+      "step": 2044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 40.768236997009254,
+      "learning_rate": 1.9695123767839995e-05,
+      "loss": 0.985,
+      "step": 2045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 56.897062742550304,
+      "learning_rate": 1.9694710390818844e-05,
+      "loss": 1.0191,
+      "step": 2046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.30091413273537,
+      "learning_rate": 1.9694296738085802e-05,
+      "loss": 0.8481,
+      "step": 2047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 44.48660325799618,
+      "learning_rate": 1.9693882809652626e-05,
+      "loss": 0.9266,
+      "step": 2048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 39.73090397790765,
+      "learning_rate": 1.96934686055311e-05,
+      "loss": 0.8516,
+      "step": 2049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 37.67714192972363,
+      "learning_rate": 1.9693054125732995e-05,
+      "loss": 0.8334,
+      "step": 2050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 39.58918024080155,
+      "learning_rate": 1.96926393702701e-05,
+      "loss": 0.8357,
+      "step": 2051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.58447838430565,
+      "learning_rate": 1.9692224339154215e-05,
+      "loss": 0.8356,
+      "step": 2052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 36.393067451167795,
+      "learning_rate": 1.9691809032397135e-05,
+      "loss": 0.9464,
+      "step": 2053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 43.81485301731299,
+      "learning_rate": 1.9691393450010675e-05,
+      "loss": 0.8734,
+      "step": 2054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 51.981031928935714,
+      "learning_rate": 1.969097759200666e-05,
+      "loss": 0.9604,
+      "step": 2055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 59.31784147724984,
+      "learning_rate": 1.969056145839691e-05,
+      "loss": 0.9301,
+      "step": 2056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 113.42588169931486,
+      "learning_rate": 1.9690145049193257e-05,
+      "loss": 0.8417,
+      "step": 2057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 34.60439514086263,
+      "learning_rate": 1.968972836440755e-05,
+      "loss": 0.8871,
+      "step": 2058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 34.37947033684956,
+      "learning_rate": 1.9689311404051638e-05,
+      "loss": 0.7997,
+      "step": 2059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 34.93754165455855,
+      "learning_rate": 1.9688894168137372e-05,
+      "loss": 0.9014,
+      "step": 2060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 50.43540302630284,
+      "learning_rate": 1.9688476656676628e-05,
+      "loss": 0.8815,
+      "step": 2061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 35.70670459256949,
+      "learning_rate": 1.9688058869681277e-05,
+      "loss": 0.9594,
+      "step": 2062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 47.4020050501973,
+      "learning_rate": 1.96876408071632e-05,
+      "loss": 0.9637,
+      "step": 2063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 52.1411003660413,
+      "learning_rate": 1.968722246913428e-05,
+      "loss": 0.9052,
+      "step": 2064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 49.63302889428026,
+      "learning_rate": 1.9686803855606422e-05,
+      "loss": 0.9828,
+      "step": 2065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 54.737203992629965,
+      "learning_rate": 1.968638496659153e-05,
+      "loss": 0.9398,
+      "step": 2066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 43.323695124518736,
+      "learning_rate": 1.9685965802101517e-05,
+      "loss": 0.9237,
+      "step": 2067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 46.04107850103285,
+      "learning_rate": 1.96855463621483e-05,
+      "loss": 0.894,
+      "step": 2068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 34.11522461188951,
+      "learning_rate": 1.9685126646743814e-05,
+      "loss": 0.8216,
+      "step": 2069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 46.719575906553175,
+      "learning_rate": 1.9684706655899988e-05,
+      "loss": 0.9587,
+      "step": 2070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 40.5576433468853,
+      "learning_rate": 1.968428638962877e-05,
+      "loss": 0.8441,
+      "step": 2071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 39.86729316615193,
+      "learning_rate": 1.9683865847942116e-05,
+      "loss": 0.9371,
+      "step": 2072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.61561227186857,
+      "learning_rate": 1.968344503085198e-05,
+      "loss": 0.9345,
+      "step": 2073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 41.201406573965066,
+      "learning_rate": 1.968302393837033e-05,
+      "loss": 0.9021,
+      "step": 2074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 50.424353652995165,
+      "learning_rate": 1.9682602570509147e-05,
+      "loss": 0.9211,
+      "step": 2075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 35.078007065334745,
+      "learning_rate": 1.968218092728041e-05,
+      "loss": 0.8926,
+      "step": 2076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 66.31484771944578,
+      "learning_rate": 1.968175900869611e-05,
+      "loss": 1.0405,
+      "step": 2077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 38.891956666712964,
+      "learning_rate": 1.968133681476825e-05,
+      "loss": 0.9437,
+      "step": 2078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 43.40384659179332,
+      "learning_rate": 1.9680914345508834e-05,
+      "loss": 0.9784,
+      "step": 2079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 43.40385455119131,
+      "learning_rate": 1.9680491600929874e-05,
+      "loss": 0.9137,
+      "step": 2080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 46.2198211875933,
+      "learning_rate": 1.9680068581043396e-05,
+      "loss": 1.0134,
+      "step": 2081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.126234894581785,
+      "learning_rate": 1.9679645285861433e-05,
+      "loss": 0.8459,
+      "step": 2082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 36.620488113481606,
+      "learning_rate": 1.9679221715396018e-05,
+      "loss": 1.0121,
+      "step": 2083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 59.97455985037757,
+      "learning_rate": 1.9678797869659204e-05,
+      "loss": 0.8519,
+      "step": 2084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 55.231802040614205,
+      "learning_rate": 1.9678373748663037e-05,
+      "loss": 0.9816,
+      "step": 2085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.239689228877985,
+      "learning_rate": 1.967794935241958e-05,
+      "loss": 0.9362,
+      "step": 2086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 33.51823372458398,
+      "learning_rate": 1.967752468094091e-05,
+      "loss": 0.9038,
+      "step": 2087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 36.06846438426851,
+      "learning_rate": 1.9677099734239098e-05,
+      "loss": 0.8278,
+      "step": 2088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.24657457174211,
+      "learning_rate": 1.967667451232623e-05,
+      "loss": 0.9813,
+      "step": 2089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 35.19727564342109,
+      "learning_rate": 1.9676249015214396e-05,
+      "loss": 0.879,
+      "step": 2090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 64.18686989583614,
+      "learning_rate": 1.9675823242915705e-05,
+      "loss": 0.963,
+      "step": 2091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.67546951702241,
+      "learning_rate": 1.967539719544226e-05,
+      "loss": 0.9372,
+      "step": 2092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.61031260446119,
+      "learning_rate": 1.967497087280618e-05,
+      "loss": 0.8555,
+      "step": 2093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.89283019269291,
+      "learning_rate": 1.9674544275019583e-05,
+      "loss": 0.8636,
+      "step": 2094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.13072993953378,
+      "learning_rate": 1.9674117402094612e-05,
+      "loss": 0.8481,
+      "step": 2095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.75160628650432,
+      "learning_rate": 1.96736902540434e-05,
+      "loss": 0.8988,
+      "step": 2096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.778521669108166,
+      "learning_rate": 1.9673262830878094e-05,
+      "loss": 0.9493,
+      "step": 2097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 42.72371470045396,
+      "learning_rate": 1.9672835132610856e-05,
+      "loss": 0.9959,
+      "step": 2098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 56.85397891075003,
+      "learning_rate": 1.967240715925384e-05,
+      "loss": 1.0735,
+      "step": 2099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.718666860367144,
+      "learning_rate": 1.9671978910819225e-05,
+      "loss": 0.9531,
+      "step": 2100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.12923996890373,
+      "learning_rate": 1.9671550387319188e-05,
+      "loss": 0.9647,
+      "step": 2101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 47.10208763327146,
+      "learning_rate": 1.9671121588765913e-05,
+      "loss": 0.9242,
+      "step": 2102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 39.61570019300715,
+      "learning_rate": 1.96706925151716e-05,
+      "loss": 0.8689,
+      "step": 2103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 49.96745563903815,
+      "learning_rate": 1.967026316654845e-05,
+      "loss": 0.9629,
+      "step": 2104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 56.41071185695636,
+      "learning_rate": 1.966983354290867e-05,
+      "loss": 1.0337,
+      "step": 2105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 49.785578389284304,
+      "learning_rate": 1.9669403644264485e-05,
+      "loss": 1.014,
+      "step": 2106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.168827461450164,
+      "learning_rate": 1.966897347062811e-05,
+      "loss": 0.9236,
+      "step": 2107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.74118838371481,
+      "learning_rate": 1.9668543022011792e-05,
+      "loss": 0.9301,
+      "step": 2108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 39.90725404436684,
+      "learning_rate": 1.9668112298427764e-05,
+      "loss": 0.9378,
+      "step": 2109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 53.547672449456975,
+      "learning_rate": 1.9667681299888277e-05,
+      "loss": 0.8781,
+      "step": 2110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 49.77354040031952,
+      "learning_rate": 1.966725002640559e-05,
+      "loss": 0.8684,
+      "step": 2111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 35.54571109081911,
+      "learning_rate": 1.9666818477991967e-05,
+      "loss": 0.9012,
+      "step": 2112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 46.679321047569516,
+      "learning_rate": 1.9666386654659683e-05,
+      "loss": 1.0178,
+      "step": 2113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 40.55545016772949,
+      "learning_rate": 1.9665954556421014e-05,
+      "loss": 0.878,
+      "step": 2114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.02785677173726,
+      "learning_rate": 1.9665522183288253e-05,
+      "loss": 0.9498,
+      "step": 2115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 48.91413177987374,
+      "learning_rate": 1.9665089535273698e-05,
+      "loss": 0.9607,
+      "step": 2116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 56.90380304847961,
+      "learning_rate": 1.966465661238965e-05,
+      "loss": 1.0023,
+      "step": 2117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 40.66784747681542,
+      "learning_rate": 1.966422341464842e-05,
+      "loss": 0.9054,
+      "step": 2118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.404087281388094,
+      "learning_rate": 1.966378994206233e-05,
+      "loss": 1.0534,
+      "step": 2119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 46.55565837290104,
+      "learning_rate": 1.9663356194643704e-05,
+      "loss": 0.9762,
+      "step": 2120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.08661426625853,
+      "learning_rate": 1.9662922172404887e-05,
+      "loss": 0.9575,
+      "step": 2121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.86053556331476,
+      "learning_rate": 1.966248787535821e-05,
+      "loss": 0.9763,
+      "step": 2122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.42928261106415,
+      "learning_rate": 1.9662053303516036e-05,
+      "loss": 0.8763,
+      "step": 2123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 42.90351947895097,
+      "learning_rate": 1.9661618456890714e-05,
+      "loss": 0.9571,
+      "step": 2124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 48.697510602213455,
+      "learning_rate": 1.9661183335494617e-05,
+      "loss": 0.9152,
+      "step": 2125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 50.88632554095389,
+      "learning_rate": 1.9660747939340116e-05,
+      "loss": 0.9518,
+      "step": 2126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 40.58350538469278,
+      "learning_rate": 1.9660312268439593e-05,
+      "loss": 0.9438,
+      "step": 2127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.66369697414751,
+      "learning_rate": 1.9659876322805448e-05,
+      "loss": 0.9663,
+      "step": 2128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 36.51073742706115,
+      "learning_rate": 1.965944010245006e-05,
+      "loss": 0.9943,
+      "step": 2129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 44.17972363538808,
+      "learning_rate": 1.9659003607385857e-05,
+      "loss": 0.8038,
+      "step": 2130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 36.30316742747361,
+      "learning_rate": 1.9658566837625236e-05,
+      "loss": 0.8926,
+      "step": 2131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.732620643511225,
+      "learning_rate": 1.9658129793180625e-05,
+      "loss": 1.0092,
+      "step": 2132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.35846179255831,
+      "learning_rate": 1.965769247406445e-05,
+      "loss": 0.9464,
+      "step": 2133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 64.03720526790312,
+      "learning_rate": 1.9657254880289155e-05,
+      "loss": 1.0219,
+      "step": 2134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 38.2841248939172,
+      "learning_rate": 1.965681701186718e-05,
+      "loss": 0.9314,
+      "step": 2135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 40.650723978284184,
+      "learning_rate": 1.9656378868810978e-05,
+      "loss": 0.979,
+      "step": 2136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.35888762178088,
+      "learning_rate": 1.965594045113301e-05,
+      "loss": 0.9017,
+      "step": 2137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 44.10813749106397,
+      "learning_rate": 1.965550175884574e-05,
+      "loss": 0.9634,
+      "step": 2138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 33.84160294419143,
+      "learning_rate": 1.9655062791961654e-05,
+      "loss": 0.985,
+      "step": 2139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.672962679433354,
+      "learning_rate": 1.9654623550493227e-05,
+      "loss": 0.8215,
+      "step": 2140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 35.73102529527498,
+      "learning_rate": 1.9654184034452955e-05,
+      "loss": 0.844,
+      "step": 2141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.254625126212524,
+      "learning_rate": 1.9653744243853333e-05,
+      "loss": 0.9426,
+      "step": 2142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.748260833013504,
+      "learning_rate": 1.9653304178706877e-05,
+      "loss": 0.9268,
+      "step": 2143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 43.703627527898526,
+      "learning_rate": 1.96528638390261e-05,
+      "loss": 0.849,
+      "step": 2144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.61798186448817,
+      "learning_rate": 1.9652423224823515e-05,
+      "loss": 0.9364,
+      "step": 2145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 46.648735110602175,
+      "learning_rate": 1.965198233611166e-05,
+      "loss": 0.9517,
+      "step": 2146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.836629254546246,
+      "learning_rate": 1.9651541172903076e-05,
+      "loss": 0.8677,
+      "step": 2147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.76780118657907,
+      "learning_rate": 1.965109973521031e-05,
+      "loss": 0.7999,
+      "step": 2148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 44.122156802639964,
+      "learning_rate": 1.965065802304591e-05,
+      "loss": 0.9359,
+      "step": 2149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 46.70529988579648,
+      "learning_rate": 1.965021603642244e-05,
+      "loss": 0.9582,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 48.09909123139196,
+      "learning_rate": 1.9649773775352475e-05,
+      "loss": 0.9467,
+      "step": 2151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 33.74432205899476,
+      "learning_rate": 1.9649331239848587e-05,
+      "loss": 0.8038,
+      "step": 2152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.84876350126184,
+      "learning_rate": 1.9648888429923364e-05,
+      "loss": 1.0438,
+      "step": 2153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 52.2978683391018,
+      "learning_rate": 1.9648445345589395e-05,
+      "loss": 0.8243,
+      "step": 2154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 50.752723221462496,
+      "learning_rate": 1.9648001986859288e-05,
+      "loss": 0.9688,
+      "step": 2155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 38.55438477185719,
+      "learning_rate": 1.9647558353745645e-05,
+      "loss": 0.9278,
+      "step": 2156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 48.344816254075155,
+      "learning_rate": 1.964711444626109e-05,
+      "loss": 1.0269,
+      "step": 2157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 45.33997630026576,
+      "learning_rate": 1.9646670264418242e-05,
+      "loss": 0.9754,
+      "step": 2158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.57198990633403,
+      "learning_rate": 1.9646225808229734e-05,
+      "loss": 1.032,
+      "step": 2159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 45.893086051476054,
+      "learning_rate": 1.9645781077708208e-05,
+      "loss": 0.9555,
+      "step": 2160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 44.73401234779899,
+      "learning_rate": 1.9645336072866312e-05,
+      "loss": 0.9132,
+      "step": 2161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.50111170392456,
+      "learning_rate": 1.9644890793716697e-05,
+      "loss": 1.0355,
+      "step": 2162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 46.423819530415166,
+      "learning_rate": 1.9644445240272033e-05,
+      "loss": 0.9227,
+      "step": 2163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 45.04524303686171,
+      "learning_rate": 1.964399941254499e-05,
+      "loss": 0.8519,
+      "step": 2164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.482295360074716,
+      "learning_rate": 1.964355331054824e-05,
+      "loss": 0.9057,
+      "step": 2165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 47.427504988025575,
+      "learning_rate": 1.9643106934294476e-05,
+      "loss": 1.0283,
+      "step": 2166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 57.94359233556189,
+      "learning_rate": 1.9642660283796395e-05,
+      "loss": 1.0112,
+      "step": 2167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.52394950956603,
+      "learning_rate": 1.9642213359066694e-05,
+      "loss": 0.9205,
+      "step": 2168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 42.452621445742416,
+      "learning_rate": 1.9641766160118086e-05,
+      "loss": 0.8986,
+      "step": 2169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 46.445469846210294,
+      "learning_rate": 1.964131868696329e-05,
+      "loss": 0.9816,
+      "step": 2170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 43.78778430934628,
+      "learning_rate": 1.964087093961503e-05,
+      "loss": 0.8495,
+      "step": 2171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 35.846437029241066,
+      "learning_rate": 1.9640422918086038e-05,
+      "loss": 0.9071,
+      "step": 2172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 45.5359794710489,
+      "learning_rate": 1.963997462238906e-05,
+      "loss": 0.7745,
+      "step": 2173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.01746644978546,
+      "learning_rate": 1.963952605253684e-05,
+      "loss": 0.8943,
+      "step": 2174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 38.70636974989446,
+      "learning_rate": 1.963907720854214e-05,
+      "loss": 0.9375,
+      "step": 2175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 43.8996415618914,
+      "learning_rate": 1.9638628090417727e-05,
+      "loss": 0.868,
+      "step": 2176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.6227861623317,
+      "learning_rate": 1.9638178698176364e-05,
+      "loss": 0.9266,
+      "step": 2177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 38.1767025150112,
+      "learning_rate": 1.963772903183084e-05,
+      "loss": 0.9186,
+      "step": 2178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.69940370327326,
+      "learning_rate": 1.963727909139394e-05,
+      "loss": 0.8435,
+      "step": 2179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.86304696531993,
+      "learning_rate": 1.9636828876878456e-05,
+      "loss": 1.0119,
+      "step": 2180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 40.05621493310177,
+      "learning_rate": 1.96363783882972e-05,
+      "loss": 0.9741,
+      "step": 2181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.41368732859839,
+      "learning_rate": 1.963592762566298e-05,
+      "loss": 1.0215,
+      "step": 2182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 32.312822115268396,
+      "learning_rate": 1.9635476588988615e-05,
+      "loss": 0.8714,
+      "step": 2183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 47.95370340355815,
+      "learning_rate": 1.963502527828693e-05,
+      "loss": 1.0553,
+      "step": 2184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.81950778179926,
+      "learning_rate": 1.9634573693570767e-05,
+      "loss": 0.9789,
+      "step": 2185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 37.43971254819171,
+      "learning_rate": 1.963412183485296e-05,
+      "loss": 0.8617,
+      "step": 2186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 33.034331225823685,
+      "learning_rate": 1.9633669702146366e-05,
+      "loss": 0.8488,
+      "step": 2187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 64.21177084465916,
+      "learning_rate": 1.963321729546384e-05,
+      "loss": 1.0121,
+      "step": 2188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 34.468972233931716,
+      "learning_rate": 1.963276461481825e-05,
+      "loss": 0.9594,
+      "step": 2189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.79308823303605,
+      "learning_rate": 1.963231166022247e-05,
+      "loss": 0.896,
+      "step": 2190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 44.222665178028315,
+      "learning_rate": 1.9631858431689382e-05,
+      "loss": 0.9939,
+      "step": 2191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 33.121853459460354,
+      "learning_rate": 1.963140492923187e-05,
+      "loss": 0.8283,
+      "step": 2192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 38.89873032115857,
+      "learning_rate": 1.963095115286284e-05,
+      "loss": 0.8681,
+      "step": 2193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 40.674632887188835,
+      "learning_rate": 1.963049710259519e-05,
+      "loss": 0.9277,
+      "step": 2194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 48.25342032806202,
+      "learning_rate": 1.963004277844184e-05,
+      "loss": 0.8978,
+      "step": 2195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 42.12440270133433,
+      "learning_rate": 1.9629588180415707e-05,
+      "loss": 1.0513,
+      "step": 2196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 44.002132456213104,
+      "learning_rate": 1.9629133308529717e-05,
+      "loss": 0.8877,
+      "step": 2197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.932490206231066,
+      "learning_rate": 1.962867816279681e-05,
+      "loss": 0.9273,
+      "step": 2198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 36.36695617536652,
+      "learning_rate": 1.9628222743229927e-05,
+      "loss": 0.7187,
+      "step": 2199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 51.05854980989844,
+      "learning_rate": 1.9627767049842022e-05,
+      "loss": 0.8455,
+      "step": 2200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.40906431114824,
+      "learning_rate": 1.9627311082646053e-05,
+      "loss": 0.7823,
+      "step": 2201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 40.082389649262055,
+      "learning_rate": 1.962685484165499e-05,
+      "loss": 1.0672,
+      "step": 2202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.795605323972495,
+      "learning_rate": 1.9626398326881805e-05,
+      "loss": 0.9157,
+      "step": 2203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.848769498666876,
+      "learning_rate": 1.9625941538339484e-05,
+      "loss": 1.0016,
+      "step": 2204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 76.0112964867954,
+      "learning_rate": 1.962548447604102e-05,
+      "loss": 0.9184,
+      "step": 2205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 49.60466705330496,
+      "learning_rate": 1.9625027139999404e-05,
+      "loss": 1.1158,
+      "step": 2206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 44.708587472743794,
+      "learning_rate": 1.962456953022765e-05,
+      "loss": 0.8676,
+      "step": 2207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.54230053162045,
+      "learning_rate": 1.9624111646738767e-05,
+      "loss": 0.9571,
+      "step": 2208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 57.12704903164387,
+      "learning_rate": 1.962365348954578e-05,
+      "loss": 0.9892,
+      "step": 2209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.7242548032162,
+      "learning_rate": 1.9623195058661716e-05,
+      "loss": 0.8394,
+      "step": 2210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 37.66561474991053,
+      "learning_rate": 1.9622736354099612e-05,
+      "loss": 0.9152,
+      "step": 2211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 39.80017637496474,
+      "learning_rate": 1.9622277375872518e-05,
+      "loss": 0.9923,
+      "step": 2212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 38.29131890984082,
+      "learning_rate": 1.9621818123993482e-05,
+      "loss": 0.887,
+      "step": 2213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 41.08692360227945,
+      "learning_rate": 1.962135859847557e-05,
+      "loss": 1.0135,
+      "step": 2214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 39.662597761445454,
+      "learning_rate": 1.9620898799331848e-05,
+      "loss": 0.995,
+      "step": 2215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 40.71048605871178,
+      "learning_rate": 1.962043872657539e-05,
+      "loss": 0.9969,
+      "step": 2216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.46070804762061,
+      "learning_rate": 1.9619978380219282e-05,
+      "loss": 0.981,
+      "step": 2217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.61636858047549,
+      "learning_rate": 1.9619517760276618e-05,
+      "loss": 0.99,
+      "step": 2218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 31.814241485294563,
+      "learning_rate": 1.9619056866760497e-05,
+      "loss": 0.7719,
+      "step": 2219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 35.82324935538231,
+      "learning_rate": 1.9618595699684026e-05,
+      "loss": 0.9145,
+      "step": 2220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 50.44207434727657,
+      "learning_rate": 1.9618134259060318e-05,
+      "loss": 0.8355,
+      "step": 2221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.932532354893674,
+      "learning_rate": 1.96176725449025e-05,
+      "loss": 0.8791,
+      "step": 2222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 41.4169395899591,
+      "learning_rate": 1.9617210557223697e-05,
+      "loss": 0.9662,
+      "step": 2223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 48.61742817780684,
+      "learning_rate": 1.9616748296037052e-05,
+      "loss": 0.9076,
+      "step": 2224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 32.71086631896531,
+      "learning_rate": 1.9616285761355717e-05,
+      "loss": 0.8354,
+      "step": 2225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 31.930240011230087,
+      "learning_rate": 1.9615822953192833e-05,
+      "loss": 0.8772,
+      "step": 2226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 48.437550201389996,
+      "learning_rate": 1.9615359871561574e-05,
+      "loss": 0.9152,
+      "step": 2227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 47.0728768792195,
+      "learning_rate": 1.9614896516475107e-05,
+      "loss": 0.8333,
+      "step": 2228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 37.401577268072366,
+      "learning_rate": 1.96144328879466e-05,
+      "loss": 0.9034,
+      "step": 2229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 41.28548069542289,
+      "learning_rate": 1.9613968985989253e-05,
+      "loss": 0.8933,
+      "step": 2230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 33.125843253288735,
+      "learning_rate": 1.961350481061625e-05,
+      "loss": 0.8433,
+      "step": 2231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 44.30124620402678,
+      "learning_rate": 1.9613040361840792e-05,
+      "loss": 0.9274,
+      "step": 2232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 44.34469383047169,
+      "learning_rate": 1.961257563967609e-05,
+      "loss": 0.8903,
+      "step": 2233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 35.9314168077181,
+      "learning_rate": 1.9612110644135365e-05,
+      "loss": 0.9096,
+      "step": 2234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.15430941968391,
+      "learning_rate": 1.961164537523183e-05,
+      "loss": 0.8959,
+      "step": 2235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.614544836359656,
+      "learning_rate": 1.9611179832978725e-05,
+      "loss": 0.9837,
+      "step": 2236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 43.829467426487476,
+      "learning_rate": 1.961071401738929e-05,
+      "loss": 0.9423,
+      "step": 2237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 47.90512924724758,
+      "learning_rate": 1.961024792847677e-05,
+      "loss": 0.9021,
+      "step": 2238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 40.6057081691625,
+      "learning_rate": 1.960978156625442e-05,
+      "loss": 0.9261,
+      "step": 2239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 44.1471220329147,
+      "learning_rate": 1.96093149307355e-05,
+      "loss": 0.8491,
+      "step": 2240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 40.22761787338488,
+      "learning_rate": 1.960884802193329e-05,
+      "loss": 1.0687,
+      "step": 2241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 33.7495466131204,
+      "learning_rate": 1.960838083986106e-05,
+      "loss": 0.8281,
+      "step": 2242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 43.267005107630624,
+      "learning_rate": 1.9607913384532102e-05,
+      "loss": 0.8952,
+      "step": 2243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 37.35859938151033,
+      "learning_rate": 1.9607445655959707e-05,
+      "loss": 0.9154,
+      "step": 2244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 46.558632821905924,
+      "learning_rate": 1.9606977654157176e-05,
+      "loss": 1.1299,
+      "step": 2245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 46.468279407121905,
+      "learning_rate": 1.9606509379137823e-05,
+      "loss": 1.0714,
+      "step": 2246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 60.97889725139372,
+      "learning_rate": 1.960604083091496e-05,
+      "loss": 1.0245,
+      "step": 2247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 37.32644106937675,
+      "learning_rate": 1.9605572009501918e-05,
+      "loss": 0.892,
+      "step": 2248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 37.95650847152425,
+      "learning_rate": 1.9605102914912025e-05,
+      "loss": 0.9492,
+      "step": 2249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 31.15236510847071,
+      "learning_rate": 1.960463354715863e-05,
+      "loss": 0.8763,
+      "step": 2250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 40.26866332356682,
+      "learning_rate": 1.9604163906255068e-05,
+      "loss": 0.9463,
+      "step": 2251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 36.48547575309129,
+      "learning_rate": 1.9603693992214707e-05,
+      "loss": 0.9096,
+      "step": 2252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.21064210598088,
+      "learning_rate": 1.9603223805050903e-05,
+      "loss": 0.9171,
+      "step": 2253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 37.92755497163613,
+      "learning_rate": 1.9602753344777036e-05,
+      "loss": 0.847,
+      "step": 2254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 43.35913731415357,
+      "learning_rate": 1.960228261140648e-05,
+      "loss": 0.8437,
+      "step": 2255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 36.8523035255962,
+      "learning_rate": 1.960181160495262e-05,
+      "loss": 1.0046,
+      "step": 2256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 48.152139962356884,
+      "learning_rate": 1.960134032542886e-05,
+      "loss": 0.8815,
+      "step": 2257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 51.92626945549053,
+      "learning_rate": 1.9600868772848595e-05,
+      "loss": 0.8296,
+      "step": 2258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 50.01691489208386,
+      "learning_rate": 1.9600396947225238e-05,
+      "loss": 0.9826,
+      "step": 2259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 44.895691224276746,
+      "learning_rate": 1.959992484857221e-05,
+      "loss": 0.9578,
+      "step": 2260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 77.07206323938863,
+      "learning_rate": 1.9599452476902934e-05,
+      "loss": 1.004,
+      "step": 2261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 35.703264999949894,
+      "learning_rate": 1.9598979832230846e-05,
+      "loss": 0.8884,
+      "step": 2262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 34.97361583452066,
+      "learning_rate": 1.9598506914569383e-05,
+      "loss": 0.9957,
+      "step": 2263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 34.28291521226951,
+      "learning_rate": 1.9598033723932004e-05,
+      "loss": 0.9394,
+      "step": 2264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 39.44737836324258,
+      "learning_rate": 1.959756026033216e-05,
+      "loss": 0.8432,
+      "step": 2265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 59.20723690956892,
+      "learning_rate": 1.9597086523783312e-05,
+      "loss": 0.9087,
+      "step": 2266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.61034239736383,
+      "learning_rate": 1.959661251429894e-05,
+      "loss": 0.9016,
+      "step": 2267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 48.00683688110582,
+      "learning_rate": 1.959613823189252e-05,
+      "loss": 0.9605,
+      "step": 2268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 43.63386002993266,
+      "learning_rate": 1.9595663676577543e-05,
+      "loss": 0.8856,
+      "step": 2269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 38.21924682443983,
+      "learning_rate": 1.9595188848367507e-05,
+      "loss": 0.8538,
+      "step": 2270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 38.60412278184011,
+      "learning_rate": 1.959471374727591e-05,
+      "loss": 0.9061,
+      "step": 2271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 40.76347604207125,
+      "learning_rate": 1.959423837331627e-05,
+      "loss": 0.9253,
+      "step": 2272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 41.048186683099054,
+      "learning_rate": 1.95937627265021e-05,
+      "loss": 0.8471,
+      "step": 2273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 41.94209072268284,
+      "learning_rate": 1.9593286806846935e-05,
+      "loss": 0.9261,
+      "step": 2274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 52.39155144104144,
+      "learning_rate": 1.95928106143643e-05,
+      "loss": 1.0032,
+      "step": 2275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 38.81923125126408,
+      "learning_rate": 1.9592334149067745e-05,
+      "loss": 0.823,
+      "step": 2276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.627496266454735,
+      "learning_rate": 1.9591857410970818e-05,
+      "loss": 0.9007,
+      "step": 2277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 36.725147907382514,
+      "learning_rate": 1.9591380400087077e-05,
+      "loss": 0.953,
+      "step": 2278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.35764798280915,
+      "learning_rate": 1.9590903116430087e-05,
+      "loss": 0.8676,
+      "step": 2279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 38.34694030970036,
+      "learning_rate": 1.9590425560013427e-05,
+      "loss": 0.8718,
+      "step": 2280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 52.45665266159011,
+      "learning_rate": 1.9589947730850668e-05,
+      "loss": 0.8658,
+      "step": 2281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.29805764523074,
+      "learning_rate": 1.958946962895541e-05,
+      "loss": 0.7659,
+      "step": 2282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.71528693875412,
+      "learning_rate": 1.9588991254341243e-05,
+      "loss": 0.7741,
+      "step": 2283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 38.55191670426183,
+      "learning_rate": 1.9588512607021777e-05,
+      "loss": 1.0049,
+      "step": 2284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 42.525809977878104,
+      "learning_rate": 1.958803368701062e-05,
+      "loss": 1.0129,
+      "step": 2285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 39.300639090182806,
+      "learning_rate": 1.9587554494321394e-05,
+      "loss": 0.975,
+      "step": 2286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 35.83257977706129,
+      "learning_rate": 1.958707502896773e-05,
+      "loss": 0.849,
+      "step": 2287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 33.74842543990989,
+      "learning_rate": 1.9586595290963254e-05,
+      "loss": 0.8218,
+      "step": 2288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 40.693329113737775,
+      "learning_rate": 1.958611528032162e-05,
+      "loss": 0.9061,
+      "step": 2289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 30.252314707572637,
+      "learning_rate": 1.9585634997056472e-05,
+      "loss": 0.8472,
+      "step": 2290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 55.81104092543657,
+      "learning_rate": 1.9585154441181475e-05,
+      "loss": 0.8919,
+      "step": 2291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.309577180276015,
+      "learning_rate": 1.9584673612710293e-05,
+      "loss": 1.01,
+      "step": 2292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 30.838217852725922,
+      "learning_rate": 1.95841925116566e-05,
+      "loss": 0.9248,
+      "step": 2293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 40.453331641431554,
+      "learning_rate": 1.9583711138034075e-05,
+      "loss": 0.9044,
+      "step": 2294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 38.98437903027714,
+      "learning_rate": 1.958322949185642e-05,
+      "loss": 0.8823,
+      "step": 2295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 47.56789888876823,
+      "learning_rate": 1.9582747573137315e-05,
+      "loss": 0.9641,
+      "step": 2296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 33.77235332145695,
+      "learning_rate": 1.958226538189048e-05,
+      "loss": 0.8662,
+      "step": 2297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 48.15282243317566,
+      "learning_rate": 1.958178291812962e-05,
+      "loss": 0.9911,
+      "step": 2298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 44.77797239757984,
+      "learning_rate": 1.9581300181868462e-05,
+      "loss": 0.9121,
+      "step": 2299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 52.05640850402539,
+      "learning_rate": 1.958081717312073e-05,
+      "loss": 0.9853,
+      "step": 2300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 36.18208018088686,
+      "learning_rate": 1.9580333891900164e-05,
+      "loss": 0.8905,
+      "step": 2301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 41.40124795606118,
+      "learning_rate": 1.9579850338220502e-05,
+      "loss": 0.9544,
+      "step": 2302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.63177700219378,
+      "learning_rate": 1.9579366512095504e-05,
+      "loss": 0.967,
+      "step": 2303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 34.8866682256201,
+      "learning_rate": 1.9578882413538928e-05,
+      "loss": 0.8978,
+      "step": 2304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 58.97732339303237,
+      "learning_rate": 1.9578398042564534e-05,
+      "loss": 0.7714,
+      "step": 2305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 43.019587424697164,
+      "learning_rate": 1.9577913399186106e-05,
+      "loss": 0.9464,
+      "step": 2306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 48.959217246442975,
+      "learning_rate": 1.9577428483417424e-05,
+      "loss": 0.9627,
+      "step": 2307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 43.78097170319963,
+      "learning_rate": 1.957694329527228e-05,
+      "loss": 0.8388,
+      "step": 2308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.598507662854274,
+      "learning_rate": 1.9576457834764468e-05,
+      "loss": 0.8242,
+      "step": 2309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 44.49862269616809,
+      "learning_rate": 1.9575972101907798e-05,
+      "loss": 0.8494,
+      "step": 2310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 35.84291737528102,
+      "learning_rate": 1.9575486096716086e-05,
+      "loss": 0.8885,
+      "step": 2311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 42.015552666808624,
+      "learning_rate": 1.957499981920315e-05,
+      "loss": 0.9598,
+      "step": 2312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 42.742667662472485,
+      "learning_rate": 1.957451326938282e-05,
+      "loss": 0.8076,
+      "step": 2313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 40.05525656515943,
+      "learning_rate": 1.957402644726893e-05,
+      "loss": 0.9883,
+      "step": 2314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 46.473463192709396,
+      "learning_rate": 1.9573539352875333e-05,
+      "loss": 0.879,
+      "step": 2315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 48.05355090398908,
+      "learning_rate": 1.9573051986215873e-05,
+      "loss": 0.9606,
+      "step": 2316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 35.39551020709206,
+      "learning_rate": 1.9572564347304418e-05,
+      "loss": 0.8206,
+      "step": 2317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 35.94487181312329,
+      "learning_rate": 1.957207643615483e-05,
+      "loss": 0.848,
+      "step": 2318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 43.2563326559943,
+      "learning_rate": 1.9571588252780992e-05,
+      "loss": 1.0297,
+      "step": 2319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 36.88338240700508,
+      "learning_rate": 1.9571099797196778e-05,
+      "loss": 0.8322,
+      "step": 2320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 33.96808806285918,
+      "learning_rate": 1.957061106941609e-05,
+      "loss": 0.8406,
+      "step": 2321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 77.28784487204841,
+      "learning_rate": 1.9570122069452817e-05,
+      "loss": 0.9451,
+      "step": 2322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 47.10057108587539,
+      "learning_rate": 1.9569632797320876e-05,
+      "loss": 0.9986,
+      "step": 2323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 31.05580892205502,
+      "learning_rate": 1.9569143253034172e-05,
+      "loss": 0.8796,
+      "step": 2324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 31.292579740840367,
+      "learning_rate": 1.9568653436606632e-05,
+      "loss": 0.9447,
+      "step": 2325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 41.16621923525256,
+      "learning_rate": 1.9568163348052186e-05,
+      "loss": 1.0245,
+      "step": 2326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.398057715014055,
+      "learning_rate": 1.9567672987384774e-05,
+      "loss": 0.9454,
+      "step": 2327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 34.28085338220523,
+      "learning_rate": 1.9567182354618338e-05,
+      "loss": 0.8575,
+      "step": 2328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 42.97793572469525,
+      "learning_rate": 1.956669144976683e-05,
+      "loss": 0.8881,
+      "step": 2329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 38.480665132638926,
+      "learning_rate": 1.9566200272844216e-05,
+      "loss": 0.9067,
+      "step": 2330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 38.851142005017955,
+      "learning_rate": 1.956570882386446e-05,
+      "loss": 0.9448,
+      "step": 2331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 83.39732448004914,
+      "learning_rate": 1.9565217102841543e-05,
+      "loss": 0.9975,
+      "step": 2332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 42.80047014336974,
+      "learning_rate": 1.9564725109789447e-05,
+      "loss": 0.8749,
+      "step": 2333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 60.96888556531326,
+      "learning_rate": 1.9564232844722166e-05,
+      "loss": 0.8843,
+      "step": 2334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 44.51061733970457,
+      "learning_rate": 1.9563740307653693e-05,
+      "loss": 0.8653,
+      "step": 2335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 43.2129893294247,
+      "learning_rate": 1.9563247498598045e-05,
+      "loss": 0.9538,
+      "step": 2336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 35.11022245354559,
+      "learning_rate": 1.956275441756923e-05,
+      "loss": 0.8749,
+      "step": 2337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 40.06526658359254,
+      "learning_rate": 1.9562261064581272e-05,
+      "loss": 0.968,
+      "step": 2338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 55.55546193125571,
+      "learning_rate": 1.956176743964821e-05,
+      "loss": 1.0018,
+      "step": 2339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 37.06747378382788,
+      "learning_rate": 1.9561273542784067e-05,
+      "loss": 0.9031,
+      "step": 2340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 48.80184559419821,
+      "learning_rate": 1.9560779374002902e-05,
+      "loss": 0.9905,
+      "step": 2341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 41.610829541707076,
+      "learning_rate": 1.9560284933318763e-05,
+      "loss": 0.8024,
+      "step": 2342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 45.26194678713585,
+      "learning_rate": 1.9559790220745715e-05,
+      "loss": 0.9682,
+      "step": 2343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.029573747298244,
+      "learning_rate": 1.9559295236297822e-05,
+      "loss": 0.8839,
+      "step": 2344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 40.25862859560955,
+      "learning_rate": 1.9558799979989165e-05,
+      "loss": 0.8241,
+      "step": 2345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 40.41415827658878,
+      "learning_rate": 1.9558304451833828e-05,
+      "loss": 1.084,
+      "step": 2346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 37.182765986344585,
+      "learning_rate": 1.955780865184591e-05,
+      "loss": 0.9605,
+      "step": 2347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 43.17459512066811,
+      "learning_rate": 1.95573125800395e-05,
+      "loss": 0.9534,
+      "step": 2348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 84.69140340107333,
+      "learning_rate": 1.955681623642871e-05,
+      "loss": 0.9146,
+      "step": 2349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 56.54606859317827,
+      "learning_rate": 1.955631962102766e-05,
+      "loss": 0.9836,
+      "step": 2350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 33.6728874709284,
+      "learning_rate": 1.955582273385047e-05,
+      "loss": 0.8555,
+      "step": 2351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 38.94331674617954,
+      "learning_rate": 1.955532557491127e-05,
+      "loss": 0.9058,
+      "step": 2352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 44.13822421676511,
+      "learning_rate": 1.95548281442242e-05,
+      "loss": 0.8789,
+      "step": 2353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 44.670355691921934,
+      "learning_rate": 1.9554330441803406e-05,
+      "loss": 0.9672,
+      "step": 2354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.02964231902746,
+      "learning_rate": 1.9553832467663048e-05,
+      "loss": 0.9402,
+      "step": 2355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 53.24178759955771,
+      "learning_rate": 1.955333422181728e-05,
+      "loss": 1.0847,
+      "step": 2356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 49.592033803627075,
+      "learning_rate": 1.955283570428027e-05,
+      "loss": 0.9612,
+      "step": 2357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 41.76564873588735,
+      "learning_rate": 1.955233691506621e-05,
+      "loss": 0.9675,
+      "step": 2358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 42.90480175755367,
+      "learning_rate": 1.9551837854189274e-05,
+      "loss": 0.9855,
+      "step": 2359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 31.999373576991747,
+      "learning_rate": 1.9551338521663656e-05,
+      "loss": 0.8231,
+      "step": 2360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 39.81069079829133,
+      "learning_rate": 1.9550838917503557e-05,
+      "loss": 0.9695,
+      "step": 2361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 46.552829183614385,
+      "learning_rate": 1.9550339041723188e-05,
+      "loss": 0.863,
+      "step": 2362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.68964560374218,
+      "learning_rate": 1.954983889433676e-05,
+      "loss": 0.8712,
+      "step": 2363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 50.65451649283049,
+      "learning_rate": 1.9549338475358505e-05,
+      "loss": 0.9332,
+      "step": 2364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 31.873863211163346,
+      "learning_rate": 1.9548837784802646e-05,
+      "loss": 0.9165,
+      "step": 2365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 34.504922111427746,
+      "learning_rate": 1.9548336822683426e-05,
+      "loss": 0.843,
+      "step": 2366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 38.45009853476907,
+      "learning_rate": 1.9547835589015095e-05,
+      "loss": 1.0612,
+      "step": 2367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 48.41693325128006,
+      "learning_rate": 1.9547334083811906e-05,
+      "loss": 0.9326,
+      "step": 2368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 39.960025559924084,
+      "learning_rate": 1.954683230708812e-05,
+      "loss": 1.0433,
+      "step": 2369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 42.18658084574513,
+      "learning_rate": 1.9546330258858007e-05,
+      "loss": 0.8925,
+      "step": 2370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 35.574771527047936,
+      "learning_rate": 1.9545827939135846e-05,
+      "loss": 0.883,
+      "step": 2371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 47.10086934751283,
+      "learning_rate": 1.9545325347935923e-05,
+      "loss": 0.8598,
+      "step": 2372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 41.16594799021894,
+      "learning_rate": 1.9544822485272528e-05,
+      "loss": 0.8169,
+      "step": 2373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 35.69866430017116,
+      "learning_rate": 1.954431935115997e-05,
+      "loss": 0.8722,
+      "step": 2374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 61.16587917520833,
+      "learning_rate": 1.9543815945612548e-05,
+      "loss": 0.9859,
+      "step": 2375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.683746626792015,
+      "learning_rate": 1.9543312268644586e-05,
+      "loss": 0.8868,
+      "step": 2376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 37.180970174839686,
+      "learning_rate": 1.9542808320270405e-05,
+      "loss": 0.9579,
+      "step": 2377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 41.07220283963174,
+      "learning_rate": 1.954230410050434e-05,
+      "loss": 0.9624,
+      "step": 2378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 44.772822772392026,
+      "learning_rate": 1.9541799609360727e-05,
+      "loss": 1.1216,
+      "step": 2379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 35.36998388696637,
+      "learning_rate": 1.9541294846853914e-05,
+      "loss": 0.9116,
+      "step": 2380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 43.512232847196344,
+      "learning_rate": 1.954078981299826e-05,
+      "loss": 0.8856,
+      "step": 2381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 42.087733087214296,
+      "learning_rate": 1.954028450780812e-05,
+      "loss": 0.9426,
+      "step": 2382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 39.65345940717585,
+      "learning_rate": 1.9539778931297874e-05,
+      "loss": 0.9901,
+      "step": 2383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 34.15156076252116,
+      "learning_rate": 1.9539273083481894e-05,
+      "loss": 0.8419,
+      "step": 2384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.39311478034408,
+      "learning_rate": 1.953876696437457e-05,
+      "loss": 1.0085,
+      "step": 2385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 40.821089846861675,
+      "learning_rate": 1.953826057399029e-05,
+      "loss": 0.9878,
+      "step": 2386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 42.85337789301364,
+      "learning_rate": 1.953775391234346e-05,
+      "loss": 0.9747,
+      "step": 2387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 54.952643651863454,
+      "learning_rate": 1.953724697944849e-05,
+      "loss": 1.0897,
+      "step": 2388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 37.187491213372745,
+      "learning_rate": 1.9536739775319793e-05,
+      "loss": 0.902,
+      "step": 2389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 37.36355476817685,
+      "learning_rate": 1.9536232299971797e-05,
+      "loss": 0.8743,
+      "step": 2390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 31.822325296521043,
+      "learning_rate": 1.953572455341893e-05,
+      "loss": 0.8483,
+      "step": 2391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 47.35283252980957,
+      "learning_rate": 1.9535216535675636e-05,
+      "loss": 0.9562,
+      "step": 2392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 38.88173090446185,
+      "learning_rate": 1.9534708246756364e-05,
+      "loss": 0.8856,
+      "step": 2393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.849937783244386,
+      "learning_rate": 1.9534199686675567e-05,
+      "loss": 0.8576,
+      "step": 2394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 44.46968556141757,
+      "learning_rate": 1.9533690855447706e-05,
+      "loss": 0.9672,
+      "step": 2395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 44.42200289884626,
+      "learning_rate": 1.9533181753087255e-05,
+      "loss": 0.9391,
+      "step": 2396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 38.3100158141175,
+      "learning_rate": 1.953267237960869e-05,
+      "loss": 0.8654,
+      "step": 2397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 43.543899943680955,
+      "learning_rate": 1.9532162735026503e-05,
+      "loss": 0.8701,
+      "step": 2398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 40.64513362821259,
+      "learning_rate": 1.9531652819355184e-05,
+      "loss": 0.972,
+      "step": 2399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 44.27772787943004,
+      "learning_rate": 1.953114263260923e-05,
+      "loss": 0.9749,
+      "step": 2400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 103.29500000070905,
+      "learning_rate": 1.9530632174803158e-05,
+      "loss": 0.9348,
+      "step": 2401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 51.06488088522368,
+      "learning_rate": 1.953012144595148e-05,
+      "loss": 1.0174,
+      "step": 2402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 39.905730035750544,
+      "learning_rate": 1.9529610446068727e-05,
+      "loss": 0.8904,
+      "step": 2403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 38.16088701857184,
+      "learning_rate": 1.9529099175169425e-05,
+      "loss": 0.7966,
+      "step": 2404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 32.8430625850849,
+      "learning_rate": 1.952858763326812e-05,
+      "loss": 0.9381,
+      "step": 2405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 39.27894850174234,
+      "learning_rate": 1.9528075820379353e-05,
+      "loss": 0.8371,
+      "step": 2406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 28.046736495332166,
+      "learning_rate": 1.9527563736517683e-05,
+      "loss": 0.7721,
+      "step": 2407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 50.30161439644529,
+      "learning_rate": 1.952705138169768e-05,
+      "loss": 0.9872,
+      "step": 2408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 44.9262797921124,
+      "learning_rate": 1.95265387559339e-05,
+      "loss": 0.9383,
+      "step": 2409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 50.302592980836515,
+      "learning_rate": 1.9526025859240936e-05,
+      "loss": 0.8899,
+      "step": 2410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.8356485565209,
+      "learning_rate": 1.952551269163337e-05,
+      "loss": 0.938,
+      "step": 2411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 47.812014174642215,
+      "learning_rate": 1.9524999253125792e-05,
+      "loss": 1.0316,
+      "step": 2412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 37.416448925265506,
+      "learning_rate": 1.952448554373281e-05,
+      "loss": 0.9844,
+      "step": 2413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 33.779441213395906,
+      "learning_rate": 1.9523971563469026e-05,
+      "loss": 0.847,
+      "step": 2414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 48.75350867140566,
+      "learning_rate": 1.9523457312349067e-05,
+      "loss": 0.9335,
+      "step": 2415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 32.12436826062509,
+      "learning_rate": 1.9522942790387552e-05,
+      "loss": 0.8933,
+      "step": 2416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 47.768648665202264,
+      "learning_rate": 1.9522427997599113e-05,
+      "loss": 0.9569,
+      "step": 2417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 36.398175207708974,
+      "learning_rate": 1.9521912933998392e-05,
+      "loss": 0.9125,
+      "step": 2418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.75948071336886,
+      "learning_rate": 1.9521397599600042e-05,
+      "loss": 0.9682,
+      "step": 2419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 36.61992354250123,
+      "learning_rate": 1.952088199441871e-05,
+      "loss": 0.9564,
+      "step": 2420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 40.17110827562624,
+      "learning_rate": 1.9520366118469064e-05,
+      "loss": 0.9513,
+      "step": 2421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 34.78254059130543,
+      "learning_rate": 1.9519849971765775e-05,
+      "loss": 0.8192,
+      "step": 2422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 34.967543389414885,
+      "learning_rate": 1.951933355432352e-05,
+      "loss": 0.8375,
+      "step": 2423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 40.91887349368936,
+      "learning_rate": 1.9518816866156994e-05,
+      "loss": 0.9421,
+      "step": 2424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 34.129588503336656,
+      "learning_rate": 1.951829990728088e-05,
+      "loss": 0.9806,
+      "step": 2425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 34.09720767116474,
+      "learning_rate": 1.951778267770988e-05,
+      "loss": 0.8003,
+      "step": 2426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 49.86752360339884,
+      "learning_rate": 1.9517265177458716e-05,
+      "loss": 1.0666,
+      "step": 2427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 55.55808852440147,
+      "learning_rate": 1.9516747406542097e-05,
+      "loss": 0.9346,
+      "step": 2428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 39.984695117260806,
+      "learning_rate": 1.9516229364974748e-05,
+      "loss": 0.9331,
+      "step": 2429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.27373145376238,
+      "learning_rate": 1.95157110527714e-05,
+      "loss": 0.8625,
+      "step": 2430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 52.48037553671767,
+      "learning_rate": 1.95151924699468e-05,
+      "loss": 0.9281,
+      "step": 2431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 40.99102238605477,
+      "learning_rate": 1.9514673616515686e-05,
+      "loss": 0.9716,
+      "step": 2432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 50.38174391441594,
+      "learning_rate": 1.9514154492492827e-05,
+      "loss": 1.0248,
+      "step": 2433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 72.70216622490028,
+      "learning_rate": 1.9513635097892975e-05,
+      "loss": 0.9392,
+      "step": 2434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 35.664035976771906,
+      "learning_rate": 1.951311543273091e-05,
+      "loss": 0.9031,
+      "step": 2435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 44.904657164419106,
+      "learning_rate": 1.9512595497021406e-05,
+      "loss": 0.9133,
+      "step": 2436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 39.171174571497275,
+      "learning_rate": 1.951207529077925e-05,
+      "loss": 0.8296,
+      "step": 2437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 55.170762168877374,
+      "learning_rate": 1.9511554814019237e-05,
+      "loss": 0.9576,
+      "step": 2438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 49.58893992143856,
+      "learning_rate": 1.951103406675617e-05,
+      "loss": 0.9059,
+      "step": 2439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 48.76382576441496,
+      "learning_rate": 1.951051304900486e-05,
+      "loss": 0.9051,
+      "step": 2440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.95424692085196,
+      "learning_rate": 1.9509991760780117e-05,
+      "loss": 0.898,
+      "step": 2441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 47.714183551407295,
+      "learning_rate": 1.9509470202096774e-05,
+      "loss": 0.9791,
+      "step": 2442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 37.2310991659114,
+      "learning_rate": 1.9508948372969664e-05,
+      "loss": 0.9744,
+      "step": 2443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 45.07326673644013,
+      "learning_rate": 1.950842627341362e-05,
+      "loss": 0.9295,
+      "step": 2444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 42.51240364160248,
+      "learning_rate": 1.95079039034435e-05,
+      "loss": 0.9974,
+      "step": 2445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 31.740057837896885,
+      "learning_rate": 1.9507381263074153e-05,
+      "loss": 0.8834,
+      "step": 2446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 59.01119314313042,
+      "learning_rate": 1.9506858352320444e-05,
+      "loss": 0.897,
+      "step": 2447
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 59.65960291077185,
+      "learning_rate": 1.9506335171197243e-05,
+      "loss": 1.0394,
+      "step": 2448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 39.70546520051422,
+      "learning_rate": 1.9505811719719435e-05,
+      "loss": 0.844,
+      "step": 2449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 42.288641966471786,
+      "learning_rate": 1.95052879979019e-05,
+      "loss": 0.8209,
+      "step": 2450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 40.341893257537556,
+      "learning_rate": 1.9504764005759534e-05,
+      "loss": 0.7944,
+      "step": 2451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.6536017318004,
+      "learning_rate": 1.950423974330724e-05,
+      "loss": 0.8706,
+      "step": 2452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 47.64124956531688,
+      "learning_rate": 1.9503715210559928e-05,
+      "loss": 0.8515,
+      "step": 2453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 36.450931677665395,
+      "learning_rate": 1.9503190407532514e-05,
+      "loss": 0.8056,
+      "step": 2454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 37.663193805633185,
+      "learning_rate": 1.9502665334239926e-05,
+      "loss": 0.8756,
+      "step": 2455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 47.11883640202956,
+      "learning_rate": 1.9502139990697095e-05,
+      "loss": 0.9608,
+      "step": 2456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 32.5566576074813,
+      "learning_rate": 1.9501614376918963e-05,
+      "loss": 0.8047,
+      "step": 2457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 34.99705988592182,
+      "learning_rate": 1.9501088492920472e-05,
+      "loss": 0.889,
+      "step": 2458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 42.86081152722185,
+      "learning_rate": 1.9500562338716584e-05,
+      "loss": 0.897,
+      "step": 2459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 40.74895154739274,
+      "learning_rate": 1.9500035914322263e-05,
+      "loss": 0.9533,
+      "step": 2460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.58501707782328,
+      "learning_rate": 1.9499509219752473e-05,
+      "loss": 0.9003,
+      "step": 2461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 41.52217491433766,
+      "learning_rate": 1.94989822550222e-05,
+      "loss": 0.8525,
+      "step": 2462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 43.808957227405685,
+      "learning_rate": 1.9498455020146428e-05,
+      "loss": 0.9353,
+      "step": 2463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 57.80197743098895,
+      "learning_rate": 1.9497927515140155e-05,
+      "loss": 1.0884,
+      "step": 2464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.36118969816926,
+      "learning_rate": 1.9497399740018376e-05,
+      "loss": 0.7962,
+      "step": 2465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 38.61865526080607,
+      "learning_rate": 1.9496871694796106e-05,
+      "loss": 1.0332,
+      "step": 2466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 44.00297950550985,
+      "learning_rate": 1.949634337948836e-05,
+      "loss": 0.8829,
+      "step": 2467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.11296196441046,
+      "learning_rate": 1.9495814794110165e-05,
+      "loss": 0.8731,
+      "step": 2468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.55652447932404,
+      "learning_rate": 1.949528593867655e-05,
+      "loss": 0.899,
+      "step": 2469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 51.15675979201784,
+      "learning_rate": 1.9494756813202558e-05,
+      "loss": 0.8361,
+      "step": 2470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 53.415380801148885,
+      "learning_rate": 1.9494227417703237e-05,
+      "loss": 1.1882,
+      "step": 2471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.16422532597704,
+      "learning_rate": 1.949369775219364e-05,
+      "loss": 0.7716,
+      "step": 2472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 75.15109251729457,
+      "learning_rate": 1.9493167816688837e-05,
+      "loss": 0.8774,
+      "step": 2473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 38.209606177879486,
+      "learning_rate": 1.949263761120389e-05,
+      "loss": 0.8772,
+      "step": 2474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 62.165643855758276,
+      "learning_rate": 1.9492107135753884e-05,
+      "loss": 0.8974,
+      "step": 2475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 42.40521383460328,
+      "learning_rate": 1.9491576390353906e-05,
+      "loss": 0.94,
+      "step": 2476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 44.40855761980993,
+      "learning_rate": 1.9491045375019043e-05,
+      "loss": 0.9609,
+      "step": 2477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.20675730540131,
+      "learning_rate": 1.949051408976441e-05,
+      "loss": 0.7351,
+      "step": 2478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.43890605058369,
+      "learning_rate": 1.9489982534605098e-05,
+      "loss": 0.9674,
+      "step": 2479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 42.01314000803023,
+      "learning_rate": 1.948945070955624e-05,
+      "loss": 0.9554,
+      "step": 2480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 54.99331107646707,
+      "learning_rate": 1.9488918614632953e-05,
+      "loss": 0.7475,
+      "step": 2481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 33.837697038193085,
+      "learning_rate": 1.9488386249850372e-05,
+      "loss": 0.7735,
+      "step": 2482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.99391351719438,
+      "learning_rate": 1.9487853615223636e-05,
+      "loss": 0.8318,
+      "step": 2483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 43.70724194051685,
+      "learning_rate": 1.9487320710767898e-05,
+      "loss": 0.8901,
+      "step": 2484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 32.00298829207319,
+      "learning_rate": 1.9486787536498304e-05,
+      "loss": 0.8548,
+      "step": 2485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 50.95864854602323,
+      "learning_rate": 1.9486254092430022e-05,
+      "loss": 0.9508,
+      "step": 2486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.694920527113354,
+      "learning_rate": 1.9485720378578225e-05,
+      "loss": 0.926,
+      "step": 2487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.72274616039422,
+      "learning_rate": 1.948518639495809e-05,
+      "loss": 0.9685,
+      "step": 2488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.7607822029531,
+      "learning_rate": 1.94846521415848e-05,
+      "loss": 0.9773,
+      "step": 2489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 35.90346134136822,
+      "learning_rate": 1.9484117618473553e-05,
+      "loss": 0.9509,
+      "step": 2490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.737617843703774,
+      "learning_rate": 1.948358282563955e-05,
+      "loss": 0.9322,
+      "step": 2491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.567636192586434,
+      "learning_rate": 1.9483047763097994e-05,
+      "loss": 0.9483,
+      "step": 2492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 42.52854133293972,
+      "learning_rate": 1.9482512430864113e-05,
+      "loss": 0.8634,
+      "step": 2493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.1093780215997,
+      "learning_rate": 1.9481976828953126e-05,
+      "loss": 0.9069,
+      "step": 2494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 43.48067758009847,
+      "learning_rate": 1.9481440957380258e-05,
+      "loss": 0.8252,
+      "step": 2495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.510117556928826,
+      "learning_rate": 1.948090481616076e-05,
+      "loss": 0.9535,
+      "step": 2496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 30.759195825205495,
+      "learning_rate": 1.9480368405309876e-05,
+      "loss": 0.8768,
+      "step": 2497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.77798158414642,
+      "learning_rate": 1.9479831724842862e-05,
+      "loss": 0.7841,
+      "step": 2498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 43.49309641767311,
+      "learning_rate": 1.9479294774774974e-05,
+      "loss": 0.9425,
+      "step": 2499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 35.90603071233787,
+      "learning_rate": 1.9478757555121493e-05,
+      "loss": 0.8335,
+      "step": 2500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.12155688385849,
+      "learning_rate": 1.9478220065897687e-05,
+      "loss": 0.8868,
+      "step": 2501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.37257039581252,
+      "learning_rate": 1.947768230711885e-05,
+      "loss": 1.0077,
+      "step": 2502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.97718751801565,
+      "learning_rate": 1.9477144278800275e-05,
+      "loss": 0.9908,
+      "step": 2503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 31.918846785901124,
+      "learning_rate": 1.9476605980957258e-05,
+      "loss": 0.8595,
+      "step": 2504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 31.97164513090793,
+      "learning_rate": 1.947606741360511e-05,
+      "loss": 0.821,
+      "step": 2505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.138650867984396,
+      "learning_rate": 1.9475528576759147e-05,
+      "loss": 0.8335,
+      "step": 2506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 45.104308485670074,
+      "learning_rate": 1.9474989470434695e-05,
+      "loss": 0.9559,
+      "step": 2507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.667232299144324,
+      "learning_rate": 1.9474450094647087e-05,
+      "loss": 0.8084,
+      "step": 2508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 42.236626957712716,
+      "learning_rate": 1.947391044941166e-05,
+      "loss": 0.879,
+      "step": 2509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.64588717192694,
+      "learning_rate": 1.947337053474376e-05,
+      "loss": 0.8608,
+      "step": 2510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 48.12866732012714,
+      "learning_rate": 1.9472830350658748e-05,
+      "loss": 1.0873,
+      "step": 2511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 37.72835343297355,
+      "learning_rate": 1.9472289897171978e-05,
+      "loss": 0.8458,
+      "step": 2512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.238586294271286,
+      "learning_rate": 1.9471749174298827e-05,
+      "loss": 0.9103,
+      "step": 2513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.60619342035288,
+      "learning_rate": 1.9471208182054668e-05,
+      "loss": 0.8295,
+      "step": 2514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 51.76327314944018,
+      "learning_rate": 1.947066692045489e-05,
+      "loss": 1.0238,
+      "step": 2515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.26520969471433,
+      "learning_rate": 1.9470125389514884e-05,
+      "loss": 0.8936,
+      "step": 2516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 35.49834092387024,
+      "learning_rate": 1.9469583589250055e-05,
+      "loss": 0.8485,
+      "step": 2517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 33.02677127064508,
+      "learning_rate": 1.9469041519675806e-05,
+      "loss": 0.8669,
+      "step": 2518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 33.274644480926185,
+      "learning_rate": 1.9468499180807554e-05,
+      "loss": 0.8788,
+      "step": 2519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.06075358936108,
+      "learning_rate": 1.9467956572660727e-05,
+      "loss": 0.8946,
+      "step": 2520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.288918999636536,
+      "learning_rate": 1.9467413695250753e-05,
+      "loss": 0.906,
+      "step": 2521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.606209476351495,
+      "learning_rate": 1.946687054859307e-05,
+      "loss": 0.8586,
+      "step": 2522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 38.529908836765784,
+      "learning_rate": 1.9466327132703128e-05,
+      "loss": 0.9254,
+      "step": 2523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 46.71372963458536,
+      "learning_rate": 1.946578344759638e-05,
+      "loss": 1.0707,
+      "step": 2524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 49.280224530007324,
+      "learning_rate": 1.946523949328829e-05,
+      "loss": 0.8076,
+      "step": 2525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 39.812617069713035,
+      "learning_rate": 1.9464695269794322e-05,
+      "loss": 0.8346,
+      "step": 2526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 31.938313080054762,
+      "learning_rate": 1.9464150777129956e-05,
+      "loss": 0.9317,
+      "step": 2527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 33.17266218749073,
+      "learning_rate": 1.9463606015310684e-05,
+      "loss": 0.839,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.291324087570615,
+      "learning_rate": 1.9463060984351988e-05,
+      "loss": 0.9025,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.13898912528757,
+      "learning_rate": 1.946251568426938e-05,
+      "loss": 0.8797,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 58.60354318387693,
+      "learning_rate": 1.9461970115078356e-05,
+      "loss": 0.8743,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.68906555992743,
+      "learning_rate": 1.9461424276794435e-05,
+      "loss": 0.8677,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 34.750663613768445,
+      "learning_rate": 1.9460878169433147e-05,
+      "loss": 0.7515,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.907516749148556,
+      "learning_rate": 1.9460331793010018e-05,
+      "loss": 0.9447,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 34.82621935170014,
+      "learning_rate": 1.9459785147540585e-05,
+      "loss": 0.9232,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 38.766521844787455,
+      "learning_rate": 1.9459238233040393e-05,
+      "loss": 0.8831,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 44.286218274287734,
+      "learning_rate": 1.9458691049525003e-05,
+      "loss": 0.8313,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.17310849591034,
+      "learning_rate": 1.9458143597009974e-05,
+      "loss": 0.9278,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 48.710113254110816,
+      "learning_rate": 1.9457595875510874e-05,
+      "loss": 1.0294,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.02528236879777,
+      "learning_rate": 1.9457047885043275e-05,
+      "loss": 0.9579,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.368906896932295,
+      "learning_rate": 1.945649962562277e-05,
+      "loss": 0.8691,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.84049842732853,
+      "learning_rate": 1.9455951097264947e-05,
+      "loss": 0.8247,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.034053900300975,
+      "learning_rate": 1.9455402299985408e-05,
+      "loss": 0.9607,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.626330343212146,
+      "learning_rate": 1.9454853233799756e-05,
+      "loss": 0.9619,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.54794194497634,
+      "learning_rate": 1.9454303898723613e-05,
+      "loss": 0.9147,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.58211605944235,
+      "learning_rate": 1.9453754294772593e-05,
+      "loss": 0.9385,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.06742788027925,
+      "learning_rate": 1.9453204421962334e-05,
+      "loss": 1.0,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.225712408181536,
+      "learning_rate": 1.945265428030847e-05,
+      "loss": 0.8754,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.690091216925104,
+      "learning_rate": 1.945210386982665e-05,
+      "loss": 0.9438,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 37.67969581118276,
+      "learning_rate": 1.9451553190532523e-05,
+      "loss": 0.9555,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 37.18543453410965,
+      "learning_rate": 1.945100224244175e-05,
+      "loss": 0.9608,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.56313043427148,
+      "learning_rate": 1.9450451025570006e-05,
+      "loss": 0.9968,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.98489762756054,
+      "learning_rate": 1.9449899539932962e-05,
+      "loss": 0.9113,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 46.03377995268615,
+      "learning_rate": 1.9449347785546303e-05,
+      "loss": 0.9242,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 48.40560734853565,
+      "learning_rate": 1.9448795762425722e-05,
+      "loss": 0.9772,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 228.80534078593527,
+      "learning_rate": 1.9448243470586913e-05,
+      "loss": 0.822,
+      "step": 2556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 44.925035071637915,
+      "learning_rate": 1.944769091004559e-05,
+      "loss": 0.9118,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.77398576895923,
+      "learning_rate": 1.9447138080817466e-05,
+      "loss": 0.8609,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 46.27954130862826,
+      "learning_rate": 1.9446584982918256e-05,
+      "loss": 0.8467,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.132017693924844,
+      "learning_rate": 1.9446031616363702e-05,
+      "loss": 0.9341,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 34.88722157737975,
+      "learning_rate": 1.9445477981169528e-05,
+      "loss": 0.9363,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.513484759611536,
+      "learning_rate": 1.944492407735149e-05,
+      "loss": 0.9482,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 59.44242726148708,
+      "learning_rate": 1.9444369904925337e-05,
+      "loss": 0.8712,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 57.776845158585694,
+      "learning_rate": 1.9443815463906826e-05,
+      "loss": 1.0358,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 43.99630436766002,
+      "learning_rate": 1.9443260754311728e-05,
+      "loss": 0.9776,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.38440322265474,
+      "learning_rate": 1.944270577615582e-05,
+      "loss": 0.9533,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.29656539658545,
+      "learning_rate": 1.944215052945488e-05,
+      "loss": 0.9388,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.65399129288132,
+      "learning_rate": 1.9441595014224702e-05,
+      "loss": 0.8209,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 34.484049395816,
+      "learning_rate": 1.944103923048109e-05,
+      "loss": 0.919,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.581871665807434,
+      "learning_rate": 1.944048317823984e-05,
+      "loss": 0.936,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.66042163653235,
+      "learning_rate": 1.9439926857516777e-05,
+      "loss": 0.8893,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 41.63648541398808,
+      "learning_rate": 1.9439370268327712e-05,
+      "loss": 0.8548,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.62140932207246,
+      "learning_rate": 1.9438813410688478e-05,
+      "loss": 0.8311,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.72702626278407,
+      "learning_rate": 1.9438256284614915e-05,
+      "loss": 0.9165,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.560811601667766,
+      "learning_rate": 1.943769889012286e-05,
+      "loss": 0.9433,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.615789179166704,
+      "learning_rate": 1.9437141227228175e-05,
+      "loss": 0.8642,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.5017155511924,
+      "learning_rate": 1.943658329594671e-05,
+      "loss": 0.8636,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 38.47973795688571,
+      "learning_rate": 1.9436025096294337e-05,
+      "loss": 0.8516,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.05273683048287,
+      "learning_rate": 1.9435466628286933e-05,
+      "loss": 0.9247,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.664828258270404,
+      "learning_rate": 1.9434907891940376e-05,
+      "loss": 0.9115,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.94432528921178,
+      "learning_rate": 1.943434888727056e-05,
+      "loss": 0.9096,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.325458869316826,
+      "learning_rate": 1.9433789614293376e-05,
+      "loss": 0.9571,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 48.810219851879694,
+      "learning_rate": 1.9433230073024737e-05,
+      "loss": 0.8963,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.86409754339493,
+      "learning_rate": 1.943267026348055e-05,
+      "loss": 0.9392,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.70929899105604,
+      "learning_rate": 1.943211018567674e-05,
+      "loss": 0.8656,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 32.02207665438093,
+      "learning_rate": 1.9431549839629235e-05,
+      "loss": 0.7832,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 41.750485383141786,
+      "learning_rate": 1.943098922535397e-05,
+      "loss": 0.9866,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.95152603786499,
+      "learning_rate": 1.9430428342866888e-05,
+      "loss": 0.994,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 41.747101900169554,
+      "learning_rate": 1.942986719218394e-05,
+      "loss": 0.9503,
+      "step": 2589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 50.60625315050547,
+      "learning_rate": 1.9429305773321085e-05,
+      "loss": 1.0774,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 46.168704786081,
+      "learning_rate": 1.9428744086294293e-05,
+      "loss": 0.9625,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.001447092217894,
+      "learning_rate": 1.942818213111953e-05,
+      "loss": 0.8186,
+      "step": 2592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.00573461786083,
+      "learning_rate": 1.9427619907812788e-05,
+      "loss": 0.9845,
+      "step": 2593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 40.93375711584086,
+      "learning_rate": 1.9427057416390048e-05,
+      "loss": 0.9215,
+      "step": 2594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 46.96004540956241,
+      "learning_rate": 1.942649465686731e-05,
+      "loss": 0.9006,
+      "step": 2595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 33.23376266639948,
+      "learning_rate": 1.9425931629260578e-05,
+      "loss": 0.8426,
+      "step": 2596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.41930385453572,
+      "learning_rate": 1.9425368333585862e-05,
+      "loss": 0.8786,
+      "step": 2597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.22440537950673,
+      "learning_rate": 1.9424804769859188e-05,
+      "loss": 0.8922,
+      "step": 2598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.32082754648814,
+      "learning_rate": 1.942424093809658e-05,
+      "loss": 1.0245,
+      "step": 2599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 47.77552519608911,
+      "learning_rate": 1.9423676838314067e-05,
+      "loss": 1.0058,
+      "step": 2600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 47.86201666037235,
+      "learning_rate": 1.9423112470527703e-05,
+      "loss": 0.9654,
+      "step": 2601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.554034053440155,
+      "learning_rate": 1.9422547834753532e-05,
+      "loss": 0.9709,
+      "step": 2602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 39.6854378051851,
+      "learning_rate": 1.942198293100761e-05,
+      "loss": 0.8831,
+      "step": 2603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 34.12165529676638,
+      "learning_rate": 1.9421417759306006e-05,
+      "loss": 0.9486,
+      "step": 2604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.7110017448404,
+      "learning_rate": 1.9420852319664797e-05,
+      "loss": 0.9532,
+      "step": 2605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.648614265675135,
+      "learning_rate": 1.9420286612100054e-05,
+      "loss": 0.8445,
+      "step": 2606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 34.04109141020547,
+      "learning_rate": 1.9419720636627874e-05,
+      "loss": 0.7663,
+      "step": 2607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.36749472067259,
+      "learning_rate": 1.9419154393264346e-05,
+      "loss": 0.9498,
+      "step": 2608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 43.421251965830294,
+      "learning_rate": 1.9418587882025578e-05,
+      "loss": 0.9214,
+      "step": 2609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.88679578113446,
+      "learning_rate": 1.941802110292768e-05,
+      "loss": 0.9982,
+      "step": 2610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 48.24109550197701,
+      "learning_rate": 1.941745405598677e-05,
+      "loss": 0.8326,
+      "step": 2611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.21471939781626,
+      "learning_rate": 1.941688674121898e-05,
+      "loss": 0.9276,
+      "step": 2612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 36.94295766694226,
+      "learning_rate": 1.941631915864044e-05,
+      "loss": 0.8434,
+      "step": 2613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 48.99842123485512,
+      "learning_rate": 1.9415751308267284e-05,
+      "loss": 0.9444,
+      "step": 2614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 39.16880779027746,
+      "learning_rate": 1.9415183190115678e-05,
+      "loss": 0.996,
+      "step": 2615
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.94986823242722,
+      "learning_rate": 1.9414614804201764e-05,
+      "loss": 1.0359,
+      "step": 2616
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.54398857477098,
+      "learning_rate": 1.9414046150541712e-05,
+      "loss": 0.8539,
+      "step": 2617
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.495719322022495,
+      "learning_rate": 1.94134772291517e-05,
+      "loss": 0.8562,
+      "step": 2618
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 54.230406976214155,
+      "learning_rate": 1.9412908040047894e-05,
+      "loss": 0.8839,
+      "step": 2619
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 50.301991932714095,
+      "learning_rate": 1.94123385832465e-05,
+      "loss": 0.8699,
+      "step": 2620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 39.619689128220344,
+      "learning_rate": 1.9411768858763695e-05,
+      "loss": 0.8891,
+      "step": 2621
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.81986793183426,
+      "learning_rate": 1.941119886661569e-05,
+      "loss": 0.8262,
+      "step": 2622
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 47.631675302458504,
+      "learning_rate": 1.9410628606818696e-05,
+      "loss": 0.9751,
+      "step": 2623
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 43.68664133540556,
+      "learning_rate": 1.9410058079388933e-05,
+      "loss": 1.0252,
+      "step": 2624
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 51.534402023344434,
+      "learning_rate": 1.940948728434262e-05,
+      "loss": 0.995,
+      "step": 2625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 93.55788646222163,
+      "learning_rate": 1.940891622169599e-05,
+      "loss": 0.7995,
+      "step": 2626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 36.81660013082032,
+      "learning_rate": 1.940834489146529e-05,
+      "loss": 0.8929,
+      "step": 2627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 32.996280630347265,
+      "learning_rate": 1.9407773293666764e-05,
+      "loss": 0.9614,
+      "step": 2628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.9389345629981,
+      "learning_rate": 1.940720142831667e-05,
+      "loss": 0.9281,
+      "step": 2629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 34.28659480878678,
+      "learning_rate": 1.940662929543127e-05,
+      "loss": 0.829,
+      "step": 2630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 49.074466635768694,
+      "learning_rate": 1.9406056895026837e-05,
+      "loss": 0.8623,
+      "step": 2631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 48.30177075475011,
+      "learning_rate": 1.9405484227119646e-05,
+      "loss": 0.8749,
+      "step": 2632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.29966103793372,
+      "learning_rate": 1.9404911291725985e-05,
+      "loss": 0.8524,
+      "step": 2633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 35.03502523750974,
+      "learning_rate": 1.9404338088862152e-05,
+      "loss": 0.7641,
+      "step": 2634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 46.24314196491444,
+      "learning_rate": 1.940376461854444e-05,
+      "loss": 0.9789,
+      "step": 2635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.74438990659919,
+      "learning_rate": 1.940319088078917e-05,
+      "loss": 0.9164,
+      "step": 2636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 44.75573922760643,
+      "learning_rate": 1.9402616875612645e-05,
+      "loss": 1.0496,
+      "step": 2637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 48.49683905653694,
+      "learning_rate": 1.9402042603031202e-05,
+      "loss": 0.9029,
+      "step": 2638
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 50.70529761692459,
+      "learning_rate": 1.9401468063061165e-05,
+      "loss": 0.9921,
+      "step": 2639
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.599608572622245,
+      "learning_rate": 1.9400893255718874e-05,
+      "loss": 1.0522,
+      "step": 2640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 45.338193725209074,
+      "learning_rate": 1.9400318181020682e-05,
+      "loss": 0.9332,
+      "step": 2641
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.293326532143354,
+      "learning_rate": 1.9399742838982937e-05,
+      "loss": 0.9907,
+      "step": 2642
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 34.956787739407616,
+      "learning_rate": 1.9399167229622008e-05,
+      "loss": 0.8718,
+      "step": 2643
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 50.3050012248435,
+      "learning_rate": 1.9398591352954258e-05,
+      "loss": 0.9148,
+      "step": 2644
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 70.5750793516042,
+      "learning_rate": 1.939801520899607e-05,
+      "loss": 1.0021,
+      "step": 2645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 52.805379303995856,
+      "learning_rate": 1.9397438797763825e-05,
+      "loss": 0.9737,
+      "step": 2646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 32.222835918087,
+      "learning_rate": 1.9396862119273918e-05,
+      "loss": 0.8393,
+      "step": 2647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 35.72510135681638,
+      "learning_rate": 1.939628517354275e-05,
+      "loss": 0.9132,
+      "step": 2648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 38.99273695316897,
+      "learning_rate": 1.939570796058673e-05,
+      "loss": 0.8649,
+      "step": 2649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 39.229420782486095,
+      "learning_rate": 1.939513048042227e-05,
+      "loss": 0.9114,
+      "step": 2650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.69271846519345,
+      "learning_rate": 1.9394552733065797e-05,
+      "loss": 0.8329,
+      "step": 2651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 50.17798211663643,
+      "learning_rate": 1.939397471853374e-05,
+      "loss": 0.9949,
+      "step": 2652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 31.313636188109143,
+      "learning_rate": 1.9393396436842537e-05,
+      "loss": 0.81,
+      "step": 2653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.33022412624977,
+      "learning_rate": 1.9392817888008632e-05,
+      "loss": 0.901,
+      "step": 2654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 61.85822328409284,
+      "learning_rate": 1.939223907204848e-05,
+      "loss": 0.8576,
+      "step": 2655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.25623408517934,
+      "learning_rate": 1.9391659988978546e-05,
+      "loss": 0.9527,
+      "step": 2656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 29.95191396335605,
+      "learning_rate": 1.9391080638815295e-05,
+      "loss": 0.8413,
+      "step": 2657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 37.128826878526546,
+      "learning_rate": 1.9390501021575203e-05,
+      "loss": 0.8656,
+      "step": 2658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 53.92912805679607,
+      "learning_rate": 1.938992113727476e-05,
+      "loss": 0.9266,
+      "step": 2659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 57.052304371290745,
+      "learning_rate": 1.9389340985930447e-05,
+      "loss": 0.9535,
+      "step": 2660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.69291901436594,
+      "learning_rate": 1.938876056755877e-05,
+      "loss": 0.8689,
+      "step": 2661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 63.64034817581792,
+      "learning_rate": 1.9388179882176237e-05,
+      "loss": 0.998,
+      "step": 2662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 33.36806059427231,
+      "learning_rate": 1.938759892979936e-05,
+      "loss": 0.7825,
+      "step": 2663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.4922659400424,
+      "learning_rate": 1.9387017710444662e-05,
+      "loss": 0.9812,
+      "step": 2664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 44.2061770785472,
+      "learning_rate": 1.9386436224128668e-05,
+      "loss": 0.8519,
+      "step": 2665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 46.37618788768407,
+      "learning_rate": 1.938585447086792e-05,
+      "loss": 0.8714,
+      "step": 2666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 40.8294835614625,
+      "learning_rate": 1.9385272450678966e-05,
+      "loss": 0.7835,
+      "step": 2667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 47.30765320658186,
+      "learning_rate": 1.938469016357835e-05,
+      "loss": 0.9602,
+      "step": 2668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 38.78058772478355,
+      "learning_rate": 1.938410760958263e-05,
+      "loss": 0.9725,
+      "step": 2669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 44.59608246464812,
+      "learning_rate": 1.9383524788708387e-05,
+      "loss": 1.0388,
+      "step": 2670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.17737539419248,
+      "learning_rate": 1.9382941700972188e-05,
+      "loss": 0.9683,
+      "step": 2671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 47.80324932689949,
+      "learning_rate": 1.938235834639061e-05,
+      "loss": 0.8668,
+      "step": 2672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 34.9718799673211,
+      "learning_rate": 1.9381774724980253e-05,
+      "loss": 0.9281,
+      "step": 2673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 72.4566862092041,
+      "learning_rate": 1.9381190836757712e-05,
+      "loss": 0.9425,
+      "step": 2674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.73025488649549,
+      "learning_rate": 1.938060668173959e-05,
+      "loss": 0.9873,
+      "step": 2675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.98301875170862,
+      "learning_rate": 1.93800222599425e-05,
+      "loss": 0.9864,
+      "step": 2676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.24823431224134,
+      "learning_rate": 1.9379437571383067e-05,
+      "loss": 0.8018,
+      "step": 2677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 55.476649600305635,
+      "learning_rate": 1.9378852616077915e-05,
+      "loss": 0.9072,
+      "step": 2678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 38.25471771666584,
+      "learning_rate": 1.9378267394043678e-05,
+      "loss": 0.8931,
+      "step": 2679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 45.74381921270154,
+      "learning_rate": 1.9377681905297007e-05,
+      "loss": 0.8561,
+      "step": 2680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 35.62630788845042,
+      "learning_rate": 1.937709614985455e-05,
+      "loss": 0.8329,
+      "step": 2681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.30801128317225,
+      "learning_rate": 1.937651012773296e-05,
+      "loss": 0.764,
+      "step": 2682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 39.128982286435544,
+      "learning_rate": 1.9375923838948907e-05,
+      "loss": 0.9297,
+      "step": 2683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 37.22297094578162,
+      "learning_rate": 1.9375337283519067e-05,
+      "loss": 0.8894,
+      "step": 2684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 69.55382781968744,
+      "learning_rate": 1.9374750461460117e-05,
+      "loss": 0.9467,
+      "step": 2685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 48.21392002007394,
+      "learning_rate": 1.9374163372788748e-05,
+      "loss": 0.9189,
+      "step": 2686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.476816198711504,
+      "learning_rate": 1.9373576017521657e-05,
+      "loss": 0.8752,
+      "step": 2687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 35.780431736414485,
+      "learning_rate": 1.9372988395675547e-05,
+      "loss": 0.8021,
+      "step": 2688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.92810675404625,
+      "learning_rate": 1.9372400507267132e-05,
+      "loss": 0.8739,
+      "step": 2689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 39.762370853640064,
+      "learning_rate": 1.9371812352313125e-05,
+      "loss": 0.8198,
+      "step": 2690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.09094338806076,
+      "learning_rate": 1.9371223930830264e-05,
+      "loss": 0.9259,
+      "step": 2691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 33.39133150587644,
+      "learning_rate": 1.9370635242835272e-05,
+      "loss": 0.9216,
+      "step": 2692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 38.76881414024094,
+      "learning_rate": 1.9370046288344894e-05,
+      "loss": 0.9061,
+      "step": 2693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 39.550944121602335,
+      "learning_rate": 1.9369457067375884e-05,
+      "loss": 0.8188,
+      "step": 2694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 35.49634503636327,
+      "learning_rate": 1.9368867579944994e-05,
+      "loss": 0.8487,
+      "step": 2695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 36.26351217141352,
+      "learning_rate": 1.936827782606899e-05,
+      "loss": 0.8707,
+      "step": 2696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 38.74687915663224,
+      "learning_rate": 1.9367687805764647e-05,
+      "loss": 0.8869,
+      "step": 2697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 47.53676829009409,
+      "learning_rate": 1.936709751904874e-05,
+      "loss": 0.9417,
+      "step": 2698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.20714702222897,
+      "learning_rate": 1.936650696593806e-05,
+      "loss": 0.9233,
+      "step": 2699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 44.077794032864084,
+      "learning_rate": 1.9365916146449405e-05,
+      "loss": 0.9666,
+      "step": 2700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 38.47966790075055,
+      "learning_rate": 1.936532506059957e-05,
+      "loss": 0.8481,
+      "step": 2701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 43.70425965653874,
+      "learning_rate": 1.936473370840537e-05,
+      "loss": 0.9306,
+      "step": 2702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 32.91542565786912,
+      "learning_rate": 1.936414208988362e-05,
+      "loss": 0.7971,
+      "step": 2703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 43.13051836393502,
+      "learning_rate": 1.936355020505115e-05,
+      "loss": 0.9024,
+      "step": 2704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 40.6010416249608,
+      "learning_rate": 1.9362958053924786e-05,
+      "loss": 0.8877,
+      "step": 2705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 56.0071543483134,
+      "learning_rate": 1.9362365636521377e-05,
+      "loss": 1.028,
+      "step": 2706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.02269300406621,
+      "learning_rate": 1.9361772952857762e-05,
+      "loss": 0.9674,
+      "step": 2707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 43.85738220145097,
+      "learning_rate": 1.93611800029508e-05,
+      "loss": 0.9271,
+      "step": 2708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 60.77380571058884,
+      "learning_rate": 1.9360586786817355e-05,
+      "loss": 0.9546,
+      "step": 2709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.678006030471586,
+      "learning_rate": 1.9359993304474302e-05,
+      "loss": 0.8323,
+      "step": 2710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 66.13803681649789,
+      "learning_rate": 1.935939955593851e-05,
+      "loss": 1.0114,
+      "step": 2711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 47.11597069769519,
+      "learning_rate": 1.9358805541226872e-05,
+      "loss": 0.9668,
+      "step": 2712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 45.941012065303376,
+      "learning_rate": 1.9358211260356282e-05,
+      "loss": 0.9044,
+      "step": 2713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.069566249085035,
+      "learning_rate": 1.9357616713343633e-05,
+      "loss": 0.7842,
+      "step": 2714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 43.567326246897856,
+      "learning_rate": 1.935702190020584e-05,
+      "loss": 1.0319,
+      "step": 2715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 30.484095636179738,
+      "learning_rate": 1.9356426820959817e-05,
+      "loss": 0.879,
+      "step": 2716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 73.09634062117631,
+      "learning_rate": 1.935583147562249e-05,
+      "loss": 0.9888,
+      "step": 2717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 30.40458303745587,
+      "learning_rate": 1.9355235864210792e-05,
+      "loss": 0.8769,
+      "step": 2718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 48.32046562546788,
+      "learning_rate": 1.9354639986741653e-05,
+      "loss": 1.0724,
+      "step": 2719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 35.3222196958885,
+      "learning_rate": 1.9354043843232028e-05,
+      "loss": 0.9528,
+      "step": 2720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 41.806205395449346,
+      "learning_rate": 1.935344743369887e-05,
+      "loss": 1.0593,
+      "step": 2721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 142.51470357427434,
+      "learning_rate": 1.9352850758159136e-05,
+      "loss": 0.8699,
+      "step": 2722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.90273960095439,
+      "learning_rate": 1.9352253816629796e-05,
+      "loss": 0.798,
+      "step": 2723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 46.71277489870932,
+      "learning_rate": 1.9351656609127833e-05,
+      "loss": 0.8291,
+      "step": 2724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 65.79605778186873,
+      "learning_rate": 1.9351059135670222e-05,
+      "loss": 0.9905,
+      "step": 2725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 42.16269790055589,
+      "learning_rate": 1.9350461396273963e-05,
+      "loss": 0.988,
+      "step": 2726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.68204202295763,
+      "learning_rate": 1.934986339095605e-05,
+      "loss": 0.8996,
+      "step": 2727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 46.24156710064641,
+      "learning_rate": 1.934926511973349e-05,
+      "loss": 0.7815,
+      "step": 2728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 47.881767193075284,
+      "learning_rate": 1.9348666582623302e-05,
+      "loss": 0.8788,
+      "step": 2729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 45.63495663734132,
+      "learning_rate": 1.9348067779642506e-05,
+      "loss": 0.8576,
+      "step": 2730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 44.38014123173848,
+      "learning_rate": 1.9347468710808128e-05,
+      "loss": 1.0742,
+      "step": 2731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 67.21160461243691,
+      "learning_rate": 1.9346869376137206e-05,
+      "loss": 0.91,
+      "step": 2732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.444530639766185,
+      "learning_rate": 1.9346269775646793e-05,
+      "loss": 0.8585,
+      "step": 2733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 42.38563514978788,
+      "learning_rate": 1.9345669909353934e-05,
+      "loss": 1.0537,
+      "step": 2734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 35.781939750586545,
+      "learning_rate": 1.9345069777275685e-05,
+      "loss": 0.8946,
+      "step": 2735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.97105004773667,
+      "learning_rate": 1.934446937942912e-05,
+      "loss": 0.9429,
+      "step": 2736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.836741165913935,
+      "learning_rate": 1.9343868715831313e-05,
+      "loss": 0.8437,
+      "step": 2737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 43.55726392849164,
+      "learning_rate": 1.9343267786499346e-05,
+      "loss": 0.8559,
+      "step": 2738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 35.6929601243595,
+      "learning_rate": 1.9342666591450307e-05,
+      "loss": 0.8817,
+      "step": 2739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 105.18239023509848,
+      "learning_rate": 1.9342065130701297e-05,
+      "loss": 0.778,
+      "step": 2740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 43.36884170700649,
+      "learning_rate": 1.9341463404269422e-05,
+      "loss": 0.945,
+      "step": 2741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 39.07248895109844,
+      "learning_rate": 1.934086141217179e-05,
+      "loss": 0.953,
+      "step": 2742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.642083897409265,
+      "learning_rate": 1.9340259154425524e-05,
+      "loss": 0.8479,
+      "step": 2743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.90443488622914,
+      "learning_rate": 1.933965663104775e-05,
+      "loss": 0.7817,
+      "step": 2744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 37.22321960874793,
+      "learning_rate": 1.9339053842055606e-05,
+      "loss": 0.9729,
+      "step": 2745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 35.4897914332111,
+      "learning_rate": 1.9338450787466234e-05,
+      "loss": 0.8174,
+      "step": 2746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 41.11879071142174,
+      "learning_rate": 1.9337847467296783e-05,
+      "loss": 1.0319,
+      "step": 2747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.49264425793425,
+      "learning_rate": 1.9337243881564417e-05,
+      "loss": 0.8925,
+      "step": 2748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 53.06339664847408,
+      "learning_rate": 1.9336640030286293e-05,
+      "loss": 0.9046,
+      "step": 2749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 49.318824135305604,
+      "learning_rate": 1.933603591347959e-05,
+      "loss": 0.8286,
+      "step": 2750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.805756012175735,
+      "learning_rate": 1.9335431531161486e-05,
+      "loss": 0.9241,
+      "step": 2751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 40.9797315770309,
+      "learning_rate": 1.933482688334917e-05,
+      "loss": 1.0154,
+      "step": 2752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 39.089426339761694,
+      "learning_rate": 1.9334221970059837e-05,
+      "loss": 0.9733,
+      "step": 2753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 35.85251980083491,
+      "learning_rate": 1.9333616791310696e-05,
+      "loss": 0.8967,
+      "step": 2754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 47.77585413162795,
+      "learning_rate": 1.9333011347118953e-05,
+      "loss": 0.9687,
+      "step": 2755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 37.11772916218335,
+      "learning_rate": 1.9332405637501823e-05,
+      "loss": 0.979,
+      "step": 2756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.05573904502319,
+      "learning_rate": 1.9331799662476537e-05,
+      "loss": 0.9245,
+      "step": 2757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 42.44833616583795,
+      "learning_rate": 1.933119342206033e-05,
+      "loss": 0.9711,
+      "step": 2758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 75.02676754438906,
+      "learning_rate": 1.933058691627044e-05,
+      "loss": 0.9534,
+      "step": 2759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 53.61675200798936,
+      "learning_rate": 1.9329980145124115e-05,
+      "loss": 0.971,
+      "step": 2760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 41.83013662090238,
+      "learning_rate": 1.9329373108638614e-05,
+      "loss": 0.9817,
+      "step": 2761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 53.26839470481343,
+      "learning_rate": 1.93287658068312e-05,
+      "loss": 0.9215,
+      "step": 2762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.954808180058514,
+      "learning_rate": 1.932815823971914e-05,
+      "loss": 0.9365,
+      "step": 2763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 41.168892838187,
+      "learning_rate": 1.9327550407319717e-05,
+      "loss": 0.9096,
+      "step": 2764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 51.56440665679071,
+      "learning_rate": 1.932694230965022e-05,
+      "loss": 1.0106,
+      "step": 2765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 41.82562828663737,
+      "learning_rate": 1.9326333946727938e-05,
+      "loss": 0.9394,
+      "step": 2766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 44.695564347991024,
+      "learning_rate": 1.932572531857017e-05,
+      "loss": 0.9758,
+      "step": 2767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.22607785807831,
+      "learning_rate": 1.9325116425194235e-05,
+      "loss": 0.908,
+      "step": 2768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 32.52231585153516,
+      "learning_rate": 1.9324507266617444e-05,
+      "loss": 0.7946,
+      "step": 2769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 37.86211868653292,
+      "learning_rate": 1.932389784285712e-05,
+      "loss": 0.915,
+      "step": 2770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.15954640249255,
+      "learning_rate": 1.9323288153930595e-05,
+      "loss": 0.8509,
+      "step": 2771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 30.518000172862862,
+      "learning_rate": 1.932267819985521e-05,
+      "loss": 0.8731,
+      "step": 2772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 39.52135238078712,
+      "learning_rate": 1.932206798064831e-05,
+      "loss": 0.8256,
+      "step": 2773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 44.742465967371196,
+      "learning_rate": 1.932145749632725e-05,
+      "loss": 0.9131,
+      "step": 2774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 45.61339372348402,
+      "learning_rate": 1.932084674690939e-05,
+      "loss": 0.8891,
+      "step": 2775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.47377280099367,
+      "learning_rate": 1.9320235732412104e-05,
+      "loss": 0.8854,
+      "step": 2776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 40.86652821348321,
+      "learning_rate": 1.9319624452852765e-05,
+      "loss": 0.807,
+      "step": 2777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 44.43967107037774,
+      "learning_rate": 1.931901290824876e-05,
+      "loss": 0.8275,
+      "step": 2778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.94502056570426,
+      "learning_rate": 1.9318401098617475e-05,
+      "loss": 0.854,
+      "step": 2779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 40.769672735293476,
+      "learning_rate": 1.9317789023976314e-05,
+      "loss": 1.0916,
+      "step": 2780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 35.241737648679035,
+      "learning_rate": 1.9317176684342685e-05,
+      "loss": 0.8992,
+      "step": 2781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.320333174798115,
+      "learning_rate": 1.9316564079734005e-05,
+      "loss": 0.8133,
+      "step": 2782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.99099064348482,
+      "learning_rate": 1.931595121016769e-05,
+      "loss": 0.9266,
+      "step": 2783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 52.371876297984926,
+      "learning_rate": 1.9315338075661172e-05,
+      "loss": 0.9312,
+      "step": 2784
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 38.1402877173851,
+      "learning_rate": 1.931472467623189e-05,
+      "loss": 0.9242,
+      "step": 2785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 34.73056769944201,
+      "learning_rate": 1.9314111011897285e-05,
+      "loss": 0.9599,
+      "step": 2786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 51.76642651085661,
+      "learning_rate": 1.9313497082674813e-05,
+      "loss": 0.927,
+      "step": 2787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 42.64484157940286,
+      "learning_rate": 1.931288288858193e-05,
+      "loss": 0.9003,
+      "step": 2788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 51.21818772822542,
+      "learning_rate": 1.9312268429636108e-05,
+      "loss": 0.9053,
+      "step": 2789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 48.95262534881623,
+      "learning_rate": 1.9311653705854817e-05,
+      "loss": 0.8577,
+      "step": 2790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 41.99231481346525,
+      "learning_rate": 1.9311038717255542e-05,
+      "loss": 0.8122,
+      "step": 2791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 33.8262933142675,
+      "learning_rate": 1.9310423463855774e-05,
+      "loss": 0.7846,
+      "step": 2792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.17956050923584,
+      "learning_rate": 1.930980794567301e-05,
+      "loss": 0.8125,
+      "step": 2793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 32.432844478383565,
+      "learning_rate": 1.9309192162724756e-05,
+      "loss": 0.7904,
+      "step": 2794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.18501922194711,
+      "learning_rate": 1.9308576115028515e-05,
+      "loss": 0.8713,
+      "step": 2795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 40.855389612553246,
+      "learning_rate": 1.9307959802601824e-05,
+      "loss": 0.7728,
+      "step": 2796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 52.750631432573265,
+      "learning_rate": 1.9307343225462195e-05,
+      "loss": 0.9824,
+      "step": 2797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 49.237701623237,
+      "learning_rate": 1.930672638362717e-05,
+      "loss": 0.8726,
+      "step": 2798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.0744918705505,
+      "learning_rate": 1.9306109277114292e-05,
+      "loss": 0.7857,
+      "step": 2799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.64323017155694,
+      "learning_rate": 1.930549190594111e-05,
+      "loss": 0.9129,
+      "step": 2800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 32.55687697288922,
+      "learning_rate": 1.930487427012518e-05,
+      "loss": 0.9091,
+      "step": 2801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.50999004204346,
+      "learning_rate": 1.930425636968407e-05,
+      "loss": 0.8827,
+      "step": 2802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 40.21460553136211,
+      "learning_rate": 1.9303638204635354e-05,
+      "loss": 0.9198,
+      "step": 2803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.37602207543832,
+      "learning_rate": 1.930301977499661e-05,
+      "loss": 0.9341,
+      "step": 2804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 44.10757749640092,
+      "learning_rate": 1.930240108078542e-05,
+      "loss": 0.8495,
+      "step": 2805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 38.21562767842959,
+      "learning_rate": 1.9301782122019392e-05,
+      "loss": 0.8958,
+      "step": 2806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.008089759307126,
+      "learning_rate": 1.9301162898716116e-05,
+      "loss": 0.8261,
+      "step": 2807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.450057117797925,
+      "learning_rate": 1.9300543410893213e-05,
+      "loss": 0.8671,
+      "step": 2808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 38.34135232407019,
+      "learning_rate": 1.9299923658568294e-05,
+      "loss": 0.895,
+      "step": 2809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 47.74318506630951,
+      "learning_rate": 1.9299303641758986e-05,
+      "loss": 0.9224,
+      "step": 2810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 80.74149447781775,
+      "learning_rate": 1.9298683360482923e-05,
+      "loss": 0.9074,
+      "step": 2811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 33.97318742083961,
+      "learning_rate": 1.9298062814757746e-05,
+      "loss": 0.795,
+      "step": 2812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 33.09930968299465,
+      "learning_rate": 1.92974420046011e-05,
+      "loss": 0.8369,
+      "step": 2813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.45850892290568,
+      "learning_rate": 1.9296820930030642e-05,
+      "loss": 0.9779,
+      "step": 2814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 40.249932052128464,
+      "learning_rate": 1.9296199591064036e-05,
+      "loss": 0.8119,
+      "step": 2815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 44.479861905146024,
+      "learning_rate": 1.9295577987718953e-05,
+      "loss": 0.8945,
+      "step": 2816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 38.18594905681865,
+      "learning_rate": 1.929495612001307e-05,
+      "loss": 0.8482,
+      "step": 2817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 67.47866659360916,
+      "learning_rate": 1.9294333987964067e-05,
+      "loss": 1.059,
+      "step": 2818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 42.73283433107071,
+      "learning_rate": 1.9293711591589645e-05,
+      "loss": 0.9242,
+      "step": 2819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.24162224713948,
+      "learning_rate": 1.9293088930907505e-05,
+      "loss": 0.9388,
+      "step": 2820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.84517072570503,
+      "learning_rate": 1.9292466005935352e-05,
+      "loss": 0.7601,
+      "step": 2821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.519598468852685,
+      "learning_rate": 1.9291842816690898e-05,
+      "loss": 0.9014,
+      "step": 2822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.62872367130748,
+      "learning_rate": 1.9291219363191873e-05,
+      "loss": 0.9068,
+      "step": 2823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 50.38693725339826,
+      "learning_rate": 1.9290595645456003e-05,
+      "loss": 0.8751,
+      "step": 2824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 49.20374172450478,
+      "learning_rate": 1.9289971663501027e-05,
+      "loss": 0.9797,
+      "step": 2825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.724707301189376,
+      "learning_rate": 1.9289347417344694e-05,
+      "loss": 0.8907,
+      "step": 2826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 38.207847992876616,
+      "learning_rate": 1.928872290700475e-05,
+      "loss": 0.9558,
+      "step": 2827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 45.93093421361349,
+      "learning_rate": 1.9288098132498966e-05,
+      "loss": 0.8704,
+      "step": 2828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.16079558271555,
+      "learning_rate": 1.9287473093845102e-05,
+      "loss": 0.9142,
+      "step": 2829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 40.89157166950351,
+      "learning_rate": 1.9286847791060937e-05,
+      "loss": 0.933,
+      "step": 2830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 40.87818981404937,
+      "learning_rate": 1.928622222416425e-05,
+      "loss": 0.8909,
+      "step": 2831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 41.79034550798057,
+      "learning_rate": 1.928559639317284e-05,
+      "loss": 0.8762,
+      "step": 2832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 54.74022299855267,
+      "learning_rate": 1.9284970298104495e-05,
+      "loss": 0.9741,
+      "step": 2833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 42.2026511500795,
+      "learning_rate": 1.928434393897703e-05,
+      "loss": 0.9289,
+      "step": 2834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 41.47581229480798,
+      "learning_rate": 1.9283717315808255e-05,
+      "loss": 0.9516,
+      "step": 2835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 38.57530426429514,
+      "learning_rate": 1.928309042861599e-05,
+      "loss": 0.9904,
+      "step": 2836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 48.9834665314606,
+      "learning_rate": 1.9282463277418062e-05,
+      "loss": 0.9136,
+      "step": 2837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 34.41500118850291,
+      "learning_rate": 1.928183586223231e-05,
+      "loss": 0.9279,
+      "step": 2838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.48797152057476,
+      "learning_rate": 1.9281208183076576e-05,
+      "loss": 1.0492,
+      "step": 2839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.30105422251978,
+      "learning_rate": 1.9280580239968708e-05,
+      "loss": 0.9491,
+      "step": 2840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 33.73905232148941,
+      "learning_rate": 1.927995203292657e-05,
+      "loss": 0.8704,
+      "step": 2841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.17848843287415,
+      "learning_rate": 1.9279323561968024e-05,
+      "loss": 0.9141,
+      "step": 2842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.4448919668615,
+      "learning_rate": 1.927869482711094e-05,
+      "loss": 0.833,
+      "step": 2843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.45715518724618,
+      "learning_rate": 1.927806582837321e-05,
+      "loss": 0.8484,
+      "step": 2844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 37.33613322961608,
+      "learning_rate": 1.927743656577271e-05,
+      "loss": 0.8656,
+      "step": 2845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 32.99738717572657,
+      "learning_rate": 1.9276807039327344e-05,
+      "loss": 0.821,
+      "step": 2846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 41.65960078345298,
+      "learning_rate": 1.9276177249055012e-05,
+      "loss": 0.962,
+      "step": 2847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.395602578946544,
+      "learning_rate": 1.9275547194973626e-05,
+      "loss": 0.8316,
+      "step": 2848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 38.39167405070378,
+      "learning_rate": 1.9274916877101104e-05,
+      "loss": 0.813,
+      "step": 2849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.493353414943645,
+      "learning_rate": 1.9274286295455373e-05,
+      "loss": 0.7418,
+      "step": 2850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.3281522378687,
+      "learning_rate": 1.927365545005436e-05,
+      "loss": 0.8545,
+      "step": 2851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 37.92811928521407,
+      "learning_rate": 1.9273024340916015e-05,
+      "loss": 0.9138,
+      "step": 2852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.803946999700372,
+      "learning_rate": 1.9272392968058282e-05,
+      "loss": 0.784,
+      "step": 2853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.74735119229442,
+      "learning_rate": 1.9271761331499114e-05,
+      "loss": 0.9203,
+      "step": 2854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.727549582852475,
+      "learning_rate": 1.927112943125648e-05,
+      "loss": 0.8695,
+      "step": 2855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.91978571229579,
+      "learning_rate": 1.9270497267348348e-05,
+      "loss": 0.8725,
+      "step": 2856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 38.777395396516674,
+      "learning_rate": 1.9269864839792697e-05,
+      "loss": 0.8876,
+      "step": 2857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.68597874785245,
+      "learning_rate": 1.9269232148607515e-05,
+      "loss": 0.8815,
+      "step": 2858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.58467178857469,
+      "learning_rate": 1.926859919381079e-05,
+      "loss": 0.8422,
+      "step": 2859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 43.35626228225664,
+      "learning_rate": 1.926796597542053e-05,
+      "loss": 0.9558,
+      "step": 2860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 29.660134425884344,
+      "learning_rate": 1.9267332493454732e-05,
+      "loss": 0.8534,
+      "step": 2861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.72257746175667,
+      "learning_rate": 1.9266698747931425e-05,
+      "loss": 0.8933,
+      "step": 2862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.4153115776858,
+      "learning_rate": 1.9266064738868625e-05,
+      "loss": 0.9875,
+      "step": 2863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.04267389954405,
+      "learning_rate": 1.9265430466284362e-05,
+      "loss": 0.8854,
+      "step": 2864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.94070534941869,
+      "learning_rate": 1.9264795930196677e-05,
+      "loss": 0.7863,
+      "step": 2865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.59256609454043,
+      "learning_rate": 1.9264161130623618e-05,
+      "loss": 0.8315,
+      "step": 2866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 32.24268078363123,
+      "learning_rate": 1.9263526067583235e-05,
+      "loss": 0.818,
+      "step": 2867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 46.3236599911079,
+      "learning_rate": 1.926289074109359e-05,
+      "loss": 0.7967,
+      "step": 2868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 43.754051473216656,
+      "learning_rate": 1.9262255151172752e-05,
+      "loss": 0.9463,
+      "step": 2869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.80169044858001,
+      "learning_rate": 1.9261619297838794e-05,
+      "loss": 0.8033,
+      "step": 2870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 46.67375300906956,
+      "learning_rate": 1.92609831811098e-05,
+      "loss": 0.8548,
+      "step": 2871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 36.231203250483055,
+      "learning_rate": 1.9260346801003864e-05,
+      "loss": 0.9171,
+      "step": 2872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 44.970461655754406,
+      "learning_rate": 1.925971015753908e-05,
+      "loss": 0.9517,
+      "step": 2873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.167888166281685,
+      "learning_rate": 1.925907325073356e-05,
+      "loss": 0.9261,
+      "step": 2874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 37.57831507137075,
+      "learning_rate": 1.925843608060541e-05,
+      "loss": 0.9022,
+      "step": 2875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 60.0301677918079,
+      "learning_rate": 1.925779864717275e-05,
+      "loss": 0.9171,
+      "step": 2876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.05962808331072,
+      "learning_rate": 1.925716095045372e-05,
+      "loss": 0.8717,
+      "step": 2877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 43.594744861694906,
+      "learning_rate": 1.9256522990466445e-05,
+      "loss": 0.9466,
+      "step": 2878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 38.12190230870716,
+      "learning_rate": 1.925588476722907e-05,
+      "loss": 0.8458,
+      "step": 2879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 44.549955887028766,
+      "learning_rate": 1.9255246280759747e-05,
+      "loss": 0.8757,
+      "step": 2880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 37.3972971694335,
+      "learning_rate": 1.9254607531076633e-05,
+      "loss": 0.842,
+      "step": 2881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 50.724963697756074,
+      "learning_rate": 1.9253968518197896e-05,
+      "loss": 0.9781,
+      "step": 2882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 42.82500914324715,
+      "learning_rate": 1.925332924214171e-05,
+      "loss": 0.8648,
+      "step": 2883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 47.15569052544458,
+      "learning_rate": 1.925268970292625e-05,
+      "loss": 1.0544,
+      "step": 2884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.74535974373109,
+      "learning_rate": 1.9252049900569707e-05,
+      "loss": 0.9651,
+      "step": 2885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 37.77725474371083,
+      "learning_rate": 1.925140983509028e-05,
+      "loss": 0.8823,
+      "step": 2886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 53.52513770879058,
+      "learning_rate": 1.9250769506506164e-05,
+      "loss": 0.8333,
+      "step": 2887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 44.76281669822113,
+      "learning_rate": 1.925012891483558e-05,
+      "loss": 0.8928,
+      "step": 2888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.912869707205836,
+      "learning_rate": 1.924948806009674e-05,
+      "loss": 0.936,
+      "step": 2889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 41.68800526417135,
+      "learning_rate": 1.9248846942307867e-05,
+      "loss": 0.8888,
+      "step": 2890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 38.33288814901221,
+      "learning_rate": 1.92482055614872e-05,
+      "loss": 0.8835,
+      "step": 2891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.20316530896848,
+      "learning_rate": 1.9247563917652978e-05,
+      "loss": 0.8057,
+      "step": 2892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 52.32804279201335,
+      "learning_rate": 1.9246922010823445e-05,
+      "loss": 1.012,
+      "step": 2893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.633349968000886,
+      "learning_rate": 1.924627984101686e-05,
+      "loss": 0.8848,
+      "step": 2894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 43.559693748248776,
+      "learning_rate": 1.924563740825149e-05,
+      "loss": 0.9531,
+      "step": 2895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 46.82313696357182,
+      "learning_rate": 1.9244994712545596e-05,
+      "loss": 0.8595,
+      "step": 2896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 36.81607446657634,
+      "learning_rate": 1.924435175391746e-05,
+      "loss": 0.905,
+      "step": 2897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.71153436319752,
+      "learning_rate": 1.924370853238537e-05,
+      "loss": 0.8764,
+      "step": 2898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 41.62986745240999,
+      "learning_rate": 1.9243065047967614e-05,
+      "loss": 0.8742,
+      "step": 2899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 37.45816544349112,
+      "learning_rate": 1.92424213006825e-05,
+      "loss": 0.851,
+      "step": 2900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 54.63872072492091,
+      "learning_rate": 1.924177729054833e-05,
+      "loss": 0.9352,
+      "step": 2901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 43.454849442135696,
+      "learning_rate": 1.9241133017583416e-05,
+      "loss": 0.7502,
+      "step": 2902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.06499422011035,
+      "learning_rate": 1.9240488481806086e-05,
+      "loss": 0.8058,
+      "step": 2903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 30.732167946754664,
+      "learning_rate": 1.923984368323467e-05,
+      "loss": 0.7607,
+      "step": 2904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 42.164134210875794,
+      "learning_rate": 1.9239198621887505e-05,
+      "loss": 0.912,
+      "step": 2905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 42.765062288729325,
+      "learning_rate": 1.9238553297782937e-05,
+      "loss": 0.9869,
+      "step": 2906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 46.78930271129696,
+      "learning_rate": 1.9237907710939317e-05,
+      "loss": 0.8558,
+      "step": 2907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 42.909348236380126,
+      "learning_rate": 1.9237261861375004e-05,
+      "loss": 0.8405,
+      "step": 2908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.563165906925434,
+      "learning_rate": 1.923661574910837e-05,
+      "loss": 0.8047,
+      "step": 2909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.22936960940538,
+      "learning_rate": 1.9235969374157786e-05,
+      "loss": 0.8618,
+      "step": 2910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 46.57453500758204,
+      "learning_rate": 1.9235322736541635e-05,
+      "loss": 0.8809,
+      "step": 2911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.87740213597494,
+      "learning_rate": 1.9234675836278308e-05,
+      "loss": 0.8053,
+      "step": 2912
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 52.688389685779086,
+      "learning_rate": 1.9234028673386205e-05,
+      "loss": 0.887,
+      "step": 2913
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 43.26396008474262,
+      "learning_rate": 1.9233381247883724e-05,
+      "loss": 1.0154,
+      "step": 2914
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 40.43287477470432,
+      "learning_rate": 1.9232733559789286e-05,
+      "loss": 0.8565,
+      "step": 2915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 29.098885497810063,
+      "learning_rate": 1.9232085609121305e-05,
+      "loss": 0.8657,
+      "step": 2916
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 45.401390608525716,
+      "learning_rate": 1.9231437395898207e-05,
+      "loss": 0.8892,
+      "step": 2917
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 44.32415329328326,
+      "learning_rate": 1.9230788920138433e-05,
+      "loss": 0.9125,
+      "step": 2918
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 45.2179330015159,
+      "learning_rate": 1.923014018186042e-05,
+      "loss": 0.8768,
+      "step": 2919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 61.46548861300779,
+      "learning_rate": 1.922949118108262e-05,
+      "loss": 0.9183,
+      "step": 2920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 34.85771598270765,
+      "learning_rate": 1.9228841917823492e-05,
+      "loss": 0.9266,
+      "step": 2921
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.48853936361315,
+      "learning_rate": 1.92281923921015e-05,
+      "loss": 0.8863,
+      "step": 2922
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 37.31817400425033,
+      "learning_rate": 1.922754260393511e-05,
+      "loss": 0.8434,
+      "step": 2923
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.96641318836166,
+      "learning_rate": 1.9226892553342808e-05,
+      "loss": 0.9502,
+      "step": 2924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.48799502184359,
+      "learning_rate": 1.9226242240343082e-05,
+      "loss": 0.8722,
+      "step": 2925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 45.348073741375416,
+      "learning_rate": 1.9225591664954423e-05,
+      "loss": 0.9755,
+      "step": 2926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 39.38040546673467,
+      "learning_rate": 1.922494082719533e-05,
+      "loss": 0.9423,
+      "step": 2927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 41.834203894676946,
+      "learning_rate": 1.922428972708432e-05,
+      "loss": 0.9446,
+      "step": 2928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 34.63118196035661,
+      "learning_rate": 1.9223638364639902e-05,
+      "loss": 0.8076,
+      "step": 2929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 40.96621597041863,
+      "learning_rate": 1.9222986739880607e-05,
+      "loss": 0.905,
+      "step": 2930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 42.446982952060125,
+      "learning_rate": 1.9222334852824966e-05,
+      "loss": 0.7758,
+      "step": 2931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 47.307608791275214,
+      "learning_rate": 1.922168270349152e-05,
+      "loss": 0.855,
+      "step": 2932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 39.73808550263812,
+      "learning_rate": 1.9221030291898802e-05,
+      "loss": 0.9357,
+      "step": 2933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 30.666931931671797,
+      "learning_rate": 1.9220377618065383e-05,
+      "loss": 0.7627,
+      "step": 2934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 38.39429033145423,
+      "learning_rate": 1.921972468200982e-05,
+      "loss": 0.8665,
+      "step": 2935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 39.96036730944012,
+      "learning_rate": 1.9219071483750678e-05,
+      "loss": 0.9605,
+      "step": 2936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 33.35472290691938,
+      "learning_rate": 1.9218418023306536e-05,
+      "loss": 0.8927,
+      "step": 2937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 36.16045230259258,
+      "learning_rate": 1.9217764300695977e-05,
+      "loss": 0.8983,
+      "step": 2938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 41.03813469239394,
+      "learning_rate": 1.9217110315937597e-05,
+      "loss": 0.8329,
+      "step": 2939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 38.37941151215993,
+      "learning_rate": 1.921645606904999e-05,
+      "loss": 0.8359,
+      "step": 2940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 58.96667193787793,
+      "learning_rate": 1.9215801560051764e-05,
+      "loss": 0.9024,
+      "step": 2941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 37.45279470337073,
+      "learning_rate": 1.921514678896153e-05,
+      "loss": 0.9151,
+      "step": 2942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 39.81200556545359,
+      "learning_rate": 1.9214491755797916e-05,
+      "loss": 0.8124,
+      "step": 2943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 44.767526689383956,
+      "learning_rate": 1.9213836460579546e-05,
+      "loss": 0.9235,
+      "step": 2944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 30.914440080420608,
+      "learning_rate": 1.9213180903325056e-05,
+      "loss": 0.7661,
+      "step": 2945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 32.34035166703148,
+      "learning_rate": 1.921252508405309e-05,
+      "loss": 0.8239,
+      "step": 2946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 33.17555252008923,
+      "learning_rate": 1.92118690027823e-05,
+      "loss": 0.8883,
+      "step": 2947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 42.17786494961537,
+      "learning_rate": 1.9211212659531345e-05,
+      "loss": 0.9719,
+      "step": 2948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 32.012235511393214,
+      "learning_rate": 1.9210556054318886e-05,
+      "loss": 0.8413,
+      "step": 2949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 42.42157542648059,
+      "learning_rate": 1.9209899187163606e-05,
+      "loss": 1.0036,
+      "step": 2950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 34.566124303536306,
+      "learning_rate": 1.920924205808418e-05,
+      "loss": 0.8827,
+      "step": 2951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 42.933277641592476,
+      "learning_rate": 1.92085846670993e-05,
+      "loss": 0.8856,
+      "step": 2952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 70.54930792887457,
+      "learning_rate": 1.9207927014227653e-05,
+      "loss": 0.8645,
+      "step": 2953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 33.45313226948302,
+      "learning_rate": 1.9207269099487953e-05,
+      "loss": 0.8643,
+      "step": 2954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 38.648635552220604,
+      "learning_rate": 1.9206610922898904e-05,
+      "loss": 0.9213,
+      "step": 2955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 29.190987936070396,
+      "learning_rate": 1.9205952484479225e-05,
+      "loss": 0.8111,
+      "step": 2956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 56.589586362808355,
+      "learning_rate": 1.920529378424764e-05,
+      "loss": 0.9056,
+      "step": 2957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 64.49456328389304,
+      "learning_rate": 1.920463482222289e-05,
+      "loss": 0.8445,
+      "step": 2958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 47.81982175578024,
+      "learning_rate": 1.920397559842371e-05,
+      "loss": 0.7995,
+      "step": 2959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 33.44846381721311,
+      "learning_rate": 1.9203316112868844e-05,
+      "loss": 0.7137,
+      "step": 2960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 47.61268408918258,
+      "learning_rate": 1.9202656365577056e-05,
+      "loss": 0.8669,
+      "step": 2961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 41.42388643196546,
+      "learning_rate": 1.9201996356567104e-05,
+      "loss": 0.8453,
+      "step": 2962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 60.8896710147575,
+      "learning_rate": 1.9201336085857757e-05,
+      "loss": 0.9566,
+      "step": 2963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 44.295278724286945,
+      "learning_rate": 1.9200675553467793e-05,
+      "loss": 0.8659,
+      "step": 2964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.3742659057904,
+      "learning_rate": 1.9200014759416002e-05,
+      "loss": 0.9333,
+      "step": 2965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 41.17176600519625,
+      "learning_rate": 1.919935370372117e-05,
+      "loss": 0.9086,
+      "step": 2966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.815058741452553,
+      "learning_rate": 1.91986923864021e-05,
+      "loss": 0.8193,
+      "step": 2967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 50.84465894490366,
+      "learning_rate": 1.91980308074776e-05,
+      "loss": 0.8846,
+      "step": 2968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 53.83356461854191,
+      "learning_rate": 1.9197368966966486e-05,
+      "loss": 0.7983,
+      "step": 2969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 39.09626351501015,
+      "learning_rate": 1.919670686488758e-05,
+      "loss": 0.8669,
+      "step": 2970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 46.02837359731437,
+      "learning_rate": 1.9196044501259706e-05,
+      "loss": 0.9532,
+      "step": 2971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 33.17481926579175,
+      "learning_rate": 1.919538187610171e-05,
+      "loss": 0.8715,
+      "step": 2972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 37.31424939065766,
+      "learning_rate": 1.9194718989432434e-05,
+      "loss": 0.8692,
+      "step": 2973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 56.702195963813274,
+      "learning_rate": 1.9194055841270724e-05,
+      "loss": 0.8653,
+      "step": 2974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 36.48154450572666,
+      "learning_rate": 1.9193392431635447e-05,
+      "loss": 0.9616,
+      "step": 2975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 34.5277594135536,
+      "learning_rate": 1.9192728760545466e-05,
+      "loss": 0.9501,
+      "step": 2976
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 51.471890704016104,
+      "learning_rate": 1.9192064828019657e-05,
+      "loss": 0.7956,
+      "step": 2977
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.52205005641987,
+      "learning_rate": 1.91914006340769e-05,
+      "loss": 0.8418,
+      "step": 2978
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.93448119244672,
+      "learning_rate": 1.919073617873609e-05,
+      "loss": 0.9508,
+      "step": 2979
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.947336065252735,
+      "learning_rate": 1.9190071462016115e-05,
+      "loss": 0.899,
+      "step": 2980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 37.83085563223457,
+      "learning_rate": 1.918940648393589e-05,
+      "loss": 0.9359,
+      "step": 2981
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 69.45398351698677,
+      "learning_rate": 1.9188741244514312e-05,
+      "loss": 1.0065,
+      "step": 2982
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.917630220874486,
+      "learning_rate": 1.9188075743770312e-05,
+      "loss": 0.9084,
+      "step": 2983
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 42.07041085886105,
+      "learning_rate": 1.9187409981722814e-05,
+      "loss": 0.7592,
+      "step": 2984
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.17071234068321,
+      "learning_rate": 1.918674395839075e-05,
+      "loss": 0.8965,
+      "step": 2985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.785844819052805,
+      "learning_rate": 1.9186077673793062e-05,
+      "loss": 0.9047,
+      "step": 2986
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 44.21480878647764,
+      "learning_rate": 1.9185411127948695e-05,
+      "loss": 0.9072,
+      "step": 2987
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 38.241295646889874,
+      "learning_rate": 1.9184744320876612e-05,
+      "loss": 0.976,
+      "step": 2988
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 37.61529560659531,
+      "learning_rate": 1.9184077252595777e-05,
+      "loss": 0.886,
+      "step": 2989
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.064456737227896,
+      "learning_rate": 1.918340992312515e-05,
+      "loss": 0.8255,
+      "step": 2990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 37.18785557937522,
+      "learning_rate": 1.9182742332483722e-05,
+      "loss": 0.7957,
+      "step": 2991
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.38028406173295,
+      "learning_rate": 1.9182074480690472e-05,
+      "loss": 0.9184,
+      "step": 2992
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 46.31690770640504,
+      "learning_rate": 1.9181406367764398e-05,
+      "loss": 0.8693,
+      "step": 2993
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 34.398119292302496,
+      "learning_rate": 1.9180737993724494e-05,
+      "loss": 0.9523,
+      "step": 2994
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 50.58793367784855,
+      "learning_rate": 1.9180069358589773e-05,
+      "loss": 0.7142,
+      "step": 2995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.364264137998674,
+      "learning_rate": 1.9179400462379248e-05,
+      "loss": 0.7992,
+      "step": 2996
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 57.04671123630545,
+      "learning_rate": 1.9178731305111952e-05,
+      "loss": 1.0753,
+      "step": 2997
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.298635041896134,
+      "learning_rate": 1.91780618868069e-05,
+      "loss": 0.849,
+      "step": 2998
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 37.684140190052844,
+      "learning_rate": 1.9177392207483138e-05,
+      "loss": 0.9485,
+      "step": 2999
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 40.33241515782782,
+      "learning_rate": 1.9176722267159706e-05,
+      "loss": 0.8685,
+      "step": 3000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 47.98485144351001,
+      "learning_rate": 1.9176052065855666e-05,
+      "loss": 0.9562,
+      "step": 3001
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 39.02477908343799,
+      "learning_rate": 1.917538160359007e-05,
+      "loss": 0.9571,
+      "step": 3002
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 32.537099068743814,
+      "learning_rate": 1.9174710880381995e-05,
+      "loss": 0.8749,
+      "step": 3003
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 47.0591567423713,
+      "learning_rate": 1.9174039896250505e-05,
+      "loss": 1.0109,
+      "step": 3004
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.64615724633064,
+      "learning_rate": 1.9173368651214686e-05,
+      "loss": 0.952,
+      "step": 3005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 34.06438559116519,
+      "learning_rate": 1.917269714529363e-05,
+      "loss": 0.7441,
+      "step": 3006
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 38.71384841028974,
+      "learning_rate": 1.9172025378506434e-05,
+      "loss": 0.9963,
+      "step": 3007
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 42.2139053784029,
+      "learning_rate": 1.9171353350872203e-05,
+      "loss": 0.9353,
+      "step": 3008
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 45.855782883183345,
+      "learning_rate": 1.9170681062410042e-05,
+      "loss": 0.7944,
+      "step": 3009
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 29.72890404440658,
+      "learning_rate": 1.917000851313908e-05,
+      "loss": 0.7787,
+      "step": 3010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 38.18678281769279,
+      "learning_rate": 1.9169335703078443e-05,
+      "loss": 0.8845,
+      "step": 3011
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 55.44941473291691,
+      "learning_rate": 1.916866263224726e-05,
+      "loss": 0.7802,
+      "step": 3012
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 46.62614486556898,
+      "learning_rate": 1.9167989300664675e-05,
+      "loss": 0.9065,
+      "step": 3013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 54.330582210693876,
+      "learning_rate": 1.9167315708349835e-05,
+      "loss": 0.9944,
+      "step": 3014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 37.00367328091662,
+      "learning_rate": 1.9166641855321902e-05,
+      "loss": 0.7149,
+      "step": 3015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.290083691176456,
+      "learning_rate": 1.9165967741600038e-05,
+      "loss": 0.9567,
+      "step": 3016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.13878162793926,
+      "learning_rate": 1.916529336720341e-05,
+      "loss": 0.9739,
+      "step": 3017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 56.700281874479124,
+      "learning_rate": 1.9164618732151202e-05,
+      "loss": 0.9386,
+      "step": 3018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.82155539852996,
+      "learning_rate": 1.91639438364626e-05,
+      "loss": 0.9407,
+      "step": 3019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 32.04178570046077,
+      "learning_rate": 1.9163268680156793e-05,
+      "loss": 0.8759,
+      "step": 3020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.55718579162164,
+      "learning_rate": 1.9162593263252988e-05,
+      "loss": 0.8661,
+      "step": 3021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 42.656561012671304,
+      "learning_rate": 1.916191758577039e-05,
+      "loss": 0.9624,
+      "step": 3022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 40.01608485829379,
+      "learning_rate": 1.9161241647728213e-05,
+      "loss": 0.9195,
+      "step": 3023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 33.561249590691176,
+      "learning_rate": 1.9160565449145687e-05,
+      "loss": 0.7982,
+      "step": 3024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 31.888332779857635,
+      "learning_rate": 1.9159888990042033e-05,
+      "loss": 0.828,
+      "step": 3025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 42.86975703977446,
+      "learning_rate": 1.91592122704365e-05,
+      "loss": 0.8879,
+      "step": 3026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.037106493699167,
+      "learning_rate": 1.915853529034832e-05,
+      "loss": 0.8371,
+      "step": 3027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 47.39743221003462,
+      "learning_rate": 1.915785804979676e-05,
+      "loss": 0.8793,
+      "step": 3028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.21359654575437,
+      "learning_rate": 1.915718054880107e-05,
+      "loss": 0.8676,
+      "step": 3029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 41.3833512841328,
+      "learning_rate": 1.9156502787380527e-05,
+      "loss": 0.93,
+      "step": 3030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.826329348927224,
+      "learning_rate": 1.91558247655544e-05,
+      "loss": 0.8406,
+      "step": 3031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 50.60352281475196,
+      "learning_rate": 1.9155146483341972e-05,
+      "loss": 0.9915,
+      "step": 3032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 46.16823345353417,
+      "learning_rate": 1.9154467940762534e-05,
+      "loss": 1.1091,
+      "step": 3033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.28975880848891,
+      "learning_rate": 1.9153789137835384e-05,
+      "loss": 0.9446,
+      "step": 3034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 39.10089949161028,
+      "learning_rate": 1.9153110074579823e-05,
+      "loss": 0.9199,
+      "step": 3035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 40.93470737745477,
+      "learning_rate": 1.915243075101517e-05,
+      "loss": 0.9393,
+      "step": 3036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.07163116311857,
+      "learning_rate": 1.9151751167160734e-05,
+      "loss": 0.8872,
+      "step": 3037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 33.154750599516376,
+      "learning_rate": 1.9151071323035856e-05,
+      "loss": 0.7665,
+      "step": 3038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 34.30173825344193,
+      "learning_rate": 1.9150391218659855e-05,
+      "loss": 0.8591,
+      "step": 3039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.72059524759723,
+      "learning_rate": 1.9149710854052087e-05,
+      "loss": 0.8219,
+      "step": 3040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 40.2404751940089,
+      "learning_rate": 1.914903022923189e-05,
+      "loss": 0.8724,
+      "step": 3041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 46.344604288820044,
+      "learning_rate": 1.914834934421863e-05,
+      "loss": 1.0992,
+      "step": 3042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.60685432137719,
+      "learning_rate": 1.9147668199031664e-05,
+      "loss": 0.8819,
+      "step": 3043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 47.50290450200608,
+      "learning_rate": 1.9146986793690363e-05,
+      "loss": 0.8415,
+      "step": 3044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35.94010148956965,
+      "learning_rate": 1.914630512821411e-05,
+      "loss": 0.8949,
+      "step": 3045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 37.275385021981826,
+      "learning_rate": 1.9145623202622293e-05,
+      "loss": 0.8791,
+      "step": 3046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 30.628233859535563,
+      "learning_rate": 1.91449410169343e-05,
+      "loss": 0.8954,
+      "step": 3047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 43.815751791745356,
+      "learning_rate": 1.9144258571169533e-05,
+      "loss": 0.9353,
+      "step": 3048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 31.95320250261603,
+      "learning_rate": 1.9143575865347405e-05,
+      "loss": 0.7633,
+      "step": 3049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.88573681651624,
+      "learning_rate": 1.9142892899487323e-05,
+      "loss": 0.9822,
+      "step": 3050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 42.899742466480355,
+      "learning_rate": 1.9142209673608717e-05,
+      "loss": 0.9961,
+      "step": 3051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.3514116835606,
+      "learning_rate": 1.914152618773102e-05,
+      "loss": 0.9437,
+      "step": 3052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 43.692574685186734,
+      "learning_rate": 1.914084244187366e-05,
+      "loss": 0.8682,
+      "step": 3053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.11855284787951,
+      "learning_rate": 1.914015843605609e-05,
+      "loss": 0.8791,
+      "step": 3054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 51.32266930975463,
+      "learning_rate": 1.9139474170297764e-05,
+      "loss": 0.9138,
+      "step": 3055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.37820052246688,
+      "learning_rate": 1.9138789644618138e-05,
+      "loss": 0.8794,
+      "step": 3056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 34.90707388292338,
+      "learning_rate": 1.9138104859036678e-05,
+      "loss": 0.8783,
+      "step": 3057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.987455392355194,
+      "learning_rate": 1.913741981357286e-05,
+      "loss": 0.8103,
+      "step": 3058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 53.0636749137503,
+      "learning_rate": 1.913673450824617e-05,
+      "loss": 0.9393,
+      "step": 3059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.81355255239077,
+      "learning_rate": 1.9136048943076098e-05,
+      "loss": 0.8251,
+      "step": 3060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 45.97700420805943,
+      "learning_rate": 1.9135363118082138e-05,
+      "loss": 0.8147,
+      "step": 3061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 46.084327981342454,
+      "learning_rate": 1.913467703328379e-05,
+      "loss": 0.8427,
+      "step": 3062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 48.67992780023391,
+      "learning_rate": 1.9133990688700578e-05,
+      "loss": 0.9325,
+      "step": 3063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 30.253134655702176,
+      "learning_rate": 1.9133304084352006e-05,
+      "loss": 0.7496,
+      "step": 3064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.30012698550873,
+      "learning_rate": 1.9132617220257613e-05,
+      "loss": 0.8508,
+      "step": 3065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.365034922604014,
+      "learning_rate": 1.913193009643693e-05,
+      "loss": 0.7613,
+      "step": 3066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 48.891890531688105,
+      "learning_rate": 1.9131242712909497e-05,
+      "loss": 0.9263,
+      "step": 3067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 37.01777710944407,
+      "learning_rate": 1.913055506969486e-05,
+      "loss": 0.9058,
+      "step": 3068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.33748099387499,
+      "learning_rate": 1.9129867166812583e-05,
+      "loss": 1.0512,
+      "step": 3069
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 47.62484866341002,
+      "learning_rate": 1.9129179004282218e-05,
+      "loss": 0.9707,
+      "step": 3070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 51.54896378830825,
+      "learning_rate": 1.9128490582123346e-05,
+      "loss": 1.0148,
+      "step": 3071
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35.023173957290965,
+      "learning_rate": 1.912780190035554e-05,
+      "loss": 0.9149,
+      "step": 3072
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35.21834746082501,
+      "learning_rate": 1.9127112958998392e-05,
+      "loss": 0.8819,
+      "step": 3073
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35.241268784444905,
+      "learning_rate": 1.9126423758071486e-05,
+      "loss": 0.7196,
+      "step": 3074
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 34.258818966773184,
+      "learning_rate": 1.912573429759443e-05,
+      "loss": 0.7845,
+      "step": 3075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 42.028757439769734,
+      "learning_rate": 1.9125044577586828e-05,
+      "loss": 0.8283,
+      "step": 3076
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 34.31448988711949,
+      "learning_rate": 1.9124354598068294e-05,
+      "loss": 0.8245,
+      "step": 3077
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 36.942898228364015,
+      "learning_rate": 1.912366435905846e-05,
+      "loss": 0.936,
+      "step": 3078
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 34.930459399856744,
+      "learning_rate": 1.9122973860576942e-05,
+      "loss": 0.9639,
+      "step": 3079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 53.34273651894444,
+      "learning_rate": 1.9122283102643383e-05,
+      "loss": 0.8303,
+      "step": 3080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 43.30687382584992,
+      "learning_rate": 1.9121592085277432e-05,
+      "loss": 0.8901,
+      "step": 3081
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 51.81233521770047,
+      "learning_rate": 1.912090080849874e-05,
+      "loss": 0.8355,
+      "step": 3082
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 32.13438158310657,
+      "learning_rate": 1.9120209272326962e-05,
+      "loss": 0.8227,
+      "step": 3083
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.78890034973441,
+      "learning_rate": 1.9119517476781766e-05,
+      "loss": 0.8197,
+      "step": 3084
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 39.51495317918991,
+      "learning_rate": 1.911882542188283e-05,
+      "loss": 0.8673,
+      "step": 3085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 43.904166262449465,
+      "learning_rate": 1.9118133107649833e-05,
+      "loss": 0.8535,
+      "step": 3086
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35.54078458822127,
+      "learning_rate": 1.9117440534102466e-05,
+      "loss": 0.8415,
+      "step": 3087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 74.77340149492476,
+      "learning_rate": 1.911674770126042e-05,
+      "loss": 0.8947,
+      "step": 3088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 38.49541589148844,
+      "learning_rate": 1.9116054609143408e-05,
+      "loss": 0.9656,
+      "step": 3089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35.954459631865326,
+      "learning_rate": 1.9115361257771133e-05,
+      "loss": 0.9306,
+      "step": 3090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 30.786210062291563,
+      "learning_rate": 1.911466764716331e-05,
+      "loss": 0.871,
+      "step": 3091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 32.47701605029772,
+      "learning_rate": 1.911397377733968e-05,
+      "loss": 0.8338,
+      "step": 3092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 39.63769268096123,
+      "learning_rate": 1.9113279648319964e-05,
+      "loss": 0.8917,
+      "step": 3093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 54.770179909568554,
+      "learning_rate": 1.9112585260123906e-05,
+      "loss": 0.8962,
+      "step": 3094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 75.71655888679565,
+      "learning_rate": 1.9111890612771252e-05,
+      "loss": 0.8927,
+      "step": 3095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 37.099314557475815,
+      "learning_rate": 1.911119570628176e-05,
+      "loss": 0.8871,
+      "step": 3096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 42.41069856388116,
+      "learning_rate": 1.9110500540675194e-05,
+      "loss": 0.956,
+      "step": 3097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 31.722076297978507,
+      "learning_rate": 1.910980511597132e-05,
+      "loss": 0.8506,
+      "step": 3098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 69.72514816602171,
+      "learning_rate": 1.910910943218992e-05,
+      "loss": 0.9173,
+      "step": 3099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.16475313323675,
+      "learning_rate": 1.9108413489350775e-05,
+      "loss": 0.8759,
+      "step": 3100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 36.88742687188361,
+      "learning_rate": 1.910771728747368e-05,
+      "loss": 0.9339,
+      "step": 3101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 39.29557755074599,
+      "learning_rate": 1.910702082657843e-05,
+      "loss": 0.9463,
+      "step": 3102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 29.7334950628091,
+      "learning_rate": 1.910632410668484e-05,
+      "loss": 0.7711,
+      "step": 3103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 44.940058050444954,
+      "learning_rate": 1.9105627127812717e-05,
+      "loss": 0.9186,
+      "step": 3104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 40.52556044881115,
+      "learning_rate": 1.910492988998188e-05,
+      "loss": 0.8786,
+      "step": 3105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 31.613989372057603,
+      "learning_rate": 1.910423239321217e-05,
+      "loss": 0.9267,
+      "step": 3106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.973922728845956,
+      "learning_rate": 1.9103534637523414e-05,
+      "loss": 0.9016,
+      "step": 3107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.14990261329527,
+      "learning_rate": 1.910283662293546e-05,
+      "loss": 0.8634,
+      "step": 3108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 44.936190842087065,
+      "learning_rate": 1.9102138349468154e-05,
+      "loss": 0.9067,
+      "step": 3109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 32.15661794379678,
+      "learning_rate": 1.910143981714136e-05,
+      "loss": 0.8753,
+      "step": 3110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.02809141506681,
+      "learning_rate": 1.9100741025974943e-05,
+      "loss": 0.8552,
+      "step": 3111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.471191874794876,
+      "learning_rate": 1.9100041975988776e-05,
+      "loss": 0.9071,
+      "step": 3112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 55.1512110583226,
+      "learning_rate": 1.9099342667202733e-05,
+      "loss": 0.9042,
+      "step": 3113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 39.180013151327266,
+      "learning_rate": 1.9098643099636714e-05,
+      "loss": 0.9222,
+      "step": 3114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 42.77698401393119,
+      "learning_rate": 1.9097943273310603e-05,
+      "loss": 0.8445,
+      "step": 3115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 39.166229218549375,
+      "learning_rate": 1.909724318824431e-05,
+      "loss": 0.9375,
+      "step": 3116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 48.335754157421974,
+      "learning_rate": 1.909654284445774e-05,
+      "loss": 0.9151,
+      "step": 3117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 30.656845324196432,
+      "learning_rate": 1.9095842241970817e-05,
+      "loss": 0.7795,
+      "step": 3118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 53.670215777607886,
+      "learning_rate": 1.9095141380803457e-05,
+      "loss": 0.8942,
+      "step": 3119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 40.260962201856046,
+      "learning_rate": 1.90944402609756e-05,
+      "loss": 0.8661,
+      "step": 3120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 36.03909826369943,
+      "learning_rate": 1.909373888250718e-05,
+      "loss": 0.8662,
+      "step": 3121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.00129202225836,
+      "learning_rate": 1.9093037245418147e-05,
+      "loss": 0.8803,
+      "step": 3122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 77.63099775009997,
+      "learning_rate": 1.909233534972845e-05,
+      "loss": 1.0373,
+      "step": 3123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 45.18448392476451,
+      "learning_rate": 1.9091633195458062e-05,
+      "loss": 0.9831,
+      "step": 3124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.718563148901694,
+      "learning_rate": 1.9090930782626943e-05,
+      "loss": 0.892,
+      "step": 3125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.861172317369274,
+      "learning_rate": 1.9090228111255066e-05,
+      "loss": 0.9312,
+      "step": 3126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 48.85491365730606,
+      "learning_rate": 1.908952518136242e-05,
+      "loss": 0.864,
+      "step": 3127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 35.358625591999306,
+      "learning_rate": 1.9088821992969e-05,
+      "loss": 0.9105,
+      "step": 3128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 38.93170204898377,
+      "learning_rate": 1.9088118546094793e-05,
+      "loss": 0.8519,
+      "step": 3129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 33.23614623134373,
+      "learning_rate": 1.9087414840759812e-05,
+      "loss": 0.8628,
+      "step": 3130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 33.38827094355882,
+      "learning_rate": 1.9086710876984075e-05,
+      "loss": 0.7992,
+      "step": 3131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 34.32705021783092,
+      "learning_rate": 1.9086006654787588e-05,
+      "loss": 0.8177,
+      "step": 3132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 49.9209423043725,
+      "learning_rate": 1.908530217419039e-05,
+      "loss": 0.9766,
+      "step": 3133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 50.414592442785654,
+      "learning_rate": 1.9084597435212517e-05,
+      "loss": 0.8671,
+      "step": 3134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 45.24313610138371,
+      "learning_rate": 1.9083892437874004e-05,
+      "loss": 0.9605,
+      "step": 3135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 30.269607620645047,
+      "learning_rate": 1.90831871821949e-05,
+      "loss": 0.8868,
+      "step": 3136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 42.176135169887864,
+      "learning_rate": 1.9082481668195273e-05,
+      "loss": 0.7896,
+      "step": 3137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 38.71420812561332,
+      "learning_rate": 1.9081775895895177e-05,
+      "loss": 0.9354,
+      "step": 3138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 40.54047269458077,
+      "learning_rate": 1.9081069865314688e-05,
+      "loss": 0.9562,
+      "step": 3139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 48.299001093421325,
+      "learning_rate": 1.9080363576473883e-05,
+      "loss": 0.9622,
+      "step": 3140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.94685886408039,
+      "learning_rate": 1.907965702939285e-05,
+      "loss": 0.8581,
+      "step": 3141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.358303746501036,
+      "learning_rate": 1.9078950224091685e-05,
+      "loss": 0.8341,
+      "step": 3142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 40.53179426570455,
+      "learning_rate": 1.907824316059048e-05,
+      "loss": 0.8684,
+      "step": 3143
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 33.556555841408866,
+      "learning_rate": 1.9077535838909356e-05,
+      "loss": 0.8103,
+      "step": 3144
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 33.075613972565506,
+      "learning_rate": 1.9076828259068422e-05,
+      "loss": 0.8121,
+      "step": 3145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 35.484730341218494,
+      "learning_rate": 1.9076120421087806e-05,
+      "loss": 0.8554,
+      "step": 3146
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 31.00892253108529,
+      "learning_rate": 1.907541232498763e-05,
+      "loss": 0.813,
+      "step": 3147
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 40.34682397337635,
+      "learning_rate": 1.9074703970788032e-05,
+      "loss": 0.9262,
+      "step": 3148
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 35.913364784511856,
+      "learning_rate": 1.907399535850917e-05,
+      "loss": 0.9027,
+      "step": 3149
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 50.19466281574324,
+      "learning_rate": 1.9073286488171184e-05,
+      "loss": 0.8936,
+      "step": 3150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 40.46752008756348,
+      "learning_rate": 1.9072577359794237e-05,
+      "loss": 0.9745,
+      "step": 3151
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 30.320814762019804,
+      "learning_rate": 1.9071867973398502e-05,
+      "loss": 0.7992,
+      "step": 3152
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 35.47467863637628,
+      "learning_rate": 1.9071158329004144e-05,
+      "loss": 0.9016,
+      "step": 3153
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 35.647517024990634,
+      "learning_rate": 1.907044842663135e-05,
+      "loss": 0.9607,
+      "step": 3154
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 42.36789072090791,
+      "learning_rate": 1.906973826630031e-05,
+      "loss": 0.8814,
+      "step": 3155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 36.233997670356075,
+      "learning_rate": 1.906902784803122e-05,
+      "loss": 0.8004,
+      "step": 3156
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 48.590533576323395,
+      "learning_rate": 1.9068317171844284e-05,
+      "loss": 1.0653,
+      "step": 3157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 44.77588856793561,
+      "learning_rate": 1.9067606237759712e-05,
+      "loss": 1.0656,
+      "step": 3158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 42.095024350140356,
+      "learning_rate": 1.906689504579772e-05,
+      "loss": 0.9127,
+      "step": 3159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 35.44788478222509,
+      "learning_rate": 1.906618359597854e-05,
+      "loss": 0.7505,
+      "step": 3160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.09499353188068,
+      "learning_rate": 1.90654718883224e-05,
+      "loss": 0.8426,
+      "step": 3161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 32.29578853675105,
+      "learning_rate": 1.9064759922849544e-05,
+      "loss": 0.8694,
+      "step": 3162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 39.66900613574982,
+      "learning_rate": 1.906404769958022e-05,
+      "loss": 0.8691,
+      "step": 3163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 33.877087113529754,
+      "learning_rate": 1.9063335218534677e-05,
+      "loss": 0.9384,
+      "step": 3164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 64.66534948318417,
+      "learning_rate": 1.906262247973319e-05,
+      "loss": 0.9291,
+      "step": 3165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 43.20214546845322,
+      "learning_rate": 1.9061909483196012e-05,
+      "loss": 0.9681,
+      "step": 3166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.13569205540025,
+      "learning_rate": 1.9061196228943436e-05,
+      "loss": 0.9627,
+      "step": 3167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 39.54030854847644,
+      "learning_rate": 1.906048271699574e-05,
+      "loss": 0.8733,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 35.142412403966595,
+      "learning_rate": 1.905976894737321e-05,
+      "loss": 0.8806,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.0139999417869,
+      "learning_rate": 1.9059054920096155e-05,
+      "loss": 0.7949,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 45.035723080865964,
+      "learning_rate": 1.9058340635184877e-05,
+      "loss": 0.9196,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 51.43081631813122,
+      "learning_rate": 1.9057626092659692e-05,
+      "loss": 0.8928,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.64109933078429,
+      "learning_rate": 1.9056911292540915e-05,
+      "loss": 0.8864,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 35.15926424795143,
+      "learning_rate": 1.905619623484888e-05,
+      "loss": 0.8268,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 34.8664885392541,
+      "learning_rate": 1.9055480919603924e-05,
+      "loss": 0.7969,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 53.53580798838282,
+      "learning_rate": 1.905476534682639e-05,
+      "loss": 0.8684,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 32.8109146825109,
+      "learning_rate": 1.905404951653662e-05,
+      "loss": 0.8764,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 37.525169442192635,
+      "learning_rate": 1.9053333428754986e-05,
+      "loss": 0.8074,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 50.3668861575407,
+      "learning_rate": 1.9052617083501837e-05,
+      "loss": 0.8836,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 50.696849967227536,
+      "learning_rate": 1.905190048079756e-05,
+      "loss": 0.8887,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 33.547870825533046,
+      "learning_rate": 1.9051183620662526e-05,
+      "loss": 0.8293,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 37.93856430446156,
+      "learning_rate": 1.9050466503117127e-05,
+      "loss": 0.836,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 46.394760137695,
+      "learning_rate": 1.9049749128181752e-05,
+      "loss": 0.9227,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 42.23828565843933,
+      "learning_rate": 1.904903149587681e-05,
+      "loss": 0.9405,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.660856789208616,
+      "learning_rate": 1.90483136062227e-05,
+      "loss": 0.9209,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 62.20347489328298,
+      "learning_rate": 1.9047595459239848e-05,
+      "loss": 0.7379,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.90765852982547,
+      "learning_rate": 1.904687705494867e-05,
+      "loss": 0.8512,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.52823686537911,
+      "learning_rate": 1.9046158393369608e-05,
+      "loss": 0.9399,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 40.69227096314109,
+      "learning_rate": 1.9045439474523086e-05,
+      "loss": 0.8164,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 49.23752912940321,
+      "learning_rate": 1.904472029842956e-05,
+      "loss": 0.9787,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 51.53579417053412,
+      "learning_rate": 1.904400086510948e-05,
+      "loss": 0.7966,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.144156089989565,
+      "learning_rate": 1.9043281174583305e-05,
+      "loss": 0.9335,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 46.352368012010906,
+      "learning_rate": 1.9042561226871506e-05,
+      "loss": 0.9083,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 35.391012416175926,
+      "learning_rate": 1.9041841021994552e-05,
+      "loss": 0.8074,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 49.82957812846273,
+      "learning_rate": 1.9041120559972927e-05,
+      "loss": 0.8391,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.77362429775225,
+      "learning_rate": 1.904039984082713e-05,
+      "loss": 0.9207,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 51.574495009272034,
+      "learning_rate": 1.9039678864577642e-05,
+      "loss": 1.0544,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 37.76363665862966,
+      "learning_rate": 1.903895763124498e-05,
+      "loss": 0.8099,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.180739694236486,
+      "learning_rate": 1.9038236140849648e-05,
+      "loss": 0.906,
+      "step": 3199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 43.58731322537032,
+      "learning_rate": 1.9037514393412167e-05,
+      "loss": 0.878,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 43.89868342678134,
+      "learning_rate": 1.9036792388953066e-05,
+      "loss": 0.8097,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 36.64688291161686,
+      "learning_rate": 1.9036070127492875e-05,
+      "loss": 0.8508,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 42.514883282355534,
+      "learning_rate": 1.903534760905213e-05,
+      "loss": 0.917,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 56.528341923907504,
+      "learning_rate": 1.9034624833651393e-05,
+      "loss": 0.8508,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 42.22403034589753,
+      "learning_rate": 1.9033901801311207e-05,
+      "loss": 0.9856,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 40.712064403514475,
+      "learning_rate": 1.903317851205214e-05,
+      "loss": 0.915,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 47.91579025241054,
+      "learning_rate": 1.903245496589476e-05,
+      "loss": 0.9118,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 34.0825152116315,
+      "learning_rate": 1.9031731162859642e-05,
+      "loss": 0.8927,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.971326372566956,
+      "learning_rate": 1.9031007102967375e-05,
+      "loss": 0.9157,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 93.85317973225901,
+      "learning_rate": 1.903028278623855e-05,
+      "loss": 0.8734,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 34.17535340475933,
+      "learning_rate": 1.9029558212693764e-05,
+      "loss": 0.8196,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 36.528987367390926,
+      "learning_rate": 1.9028833382353624e-05,
+      "loss": 0.825,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 37.68152528663209,
+      "learning_rate": 1.9028108295238745e-05,
+      "loss": 0.8952,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 53.355551251736486,
+      "learning_rate": 1.902738295136975e-05,
+      "loss": 0.881,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.907189901281356,
+      "learning_rate": 1.902665735076726e-05,
+      "loss": 0.9067,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 37.2592808983416,
+      "learning_rate": 1.9025931493451917e-05,
+      "loss": 1.0035,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.06480647487497,
+      "learning_rate": 1.9025205379444362e-05,
+      "loss": 0.8915,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 38.192376782178016,
+      "learning_rate": 1.9024479008765246e-05,
+      "loss": 0.8937,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 56.6163151124931,
+      "learning_rate": 1.9023752381435227e-05,
+      "loss": 0.8561,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 37.04977550438578,
+      "learning_rate": 1.9023025497474968e-05,
+      "loss": 0.8012,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 45.867236917553896,
+      "learning_rate": 1.9022298356905146e-05,
+      "loss": 0.8321,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 48.01562009911306,
+      "learning_rate": 1.902157095974643e-05,
+      "loss": 0.843,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.78772899919821,
+      "learning_rate": 1.902084330601952e-05,
+      "loss": 0.8646,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.57561381833695,
+      "learning_rate": 1.9020115395745098e-05,
+      "loss": 0.8843,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.57765187736527,
+      "learning_rate": 1.9019387228943872e-05,
+      "loss": 0.8716,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 35.48105561131092,
+      "learning_rate": 1.9018658805636553e-05,
+      "loss": 0.8834,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 42.68630476544738,
+      "learning_rate": 1.901793012584385e-05,
+      "loss": 0.799,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 38.09577131033014,
+      "learning_rate": 1.901720118958649e-05,
+      "loss": 0.7659,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 45.535477127184215,
+      "learning_rate": 1.9016471996885202e-05,
+      "loss": 0.8251,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 31.726548390971274,
+      "learning_rate": 1.9015742547760726e-05,
+      "loss": 0.8616,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 36.50838004874559,
+      "learning_rate": 1.9015012842233807e-05,
+      "loss": 0.7665,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 42.007091525252605,
+      "learning_rate": 1.9014282880325194e-05,
+      "loss": 0.8379,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 57.02460864550313,
+      "learning_rate": 1.9013552662055652e-05,
+      "loss": 0.837,
+      "step": 3233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 48.2143481131054,
+      "learning_rate": 1.9012822187445944e-05,
+      "loss": 0.9582,
+      "step": 3234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 44.7868715153259,
+      "learning_rate": 1.9012091456516843e-05,
+      "loss": 0.8858,
+      "step": 3235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 72.04803512191849,
+      "learning_rate": 1.9011360469289138e-05,
+      "loss": 0.8053,
+      "step": 3236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 44.13880547208261,
+      "learning_rate": 1.901062922578361e-05,
+      "loss": 0.9537,
+      "step": 3237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 31.990044952675802,
+      "learning_rate": 1.9009897726021058e-05,
+      "loss": 0.7865,
+      "step": 3238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 43.14012394952277,
+      "learning_rate": 1.9009165970022282e-05,
+      "loss": 0.9031,
+      "step": 3239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 41.11909091087413,
+      "learning_rate": 1.9008433957808102e-05,
+      "loss": 0.9317,
+      "step": 3240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 45.13112691776116,
+      "learning_rate": 1.9007701689399323e-05,
+      "loss": 0.9929,
+      "step": 3241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 49.2326649932119,
+      "learning_rate": 1.9006969164816784e-05,
+      "loss": 0.9215,
+      "step": 3242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.18728570518989,
+      "learning_rate": 1.9006236384081306e-05,
+      "loss": 0.8119,
+      "step": 3243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 46.83765255394417,
+      "learning_rate": 1.9005503347213738e-05,
+      "loss": 0.8267,
+      "step": 3244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.0387183185769,
+      "learning_rate": 1.900477005423492e-05,
+      "loss": 0.9511,
+      "step": 3245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 64.78874902156043,
+      "learning_rate": 1.9004036505165708e-05,
+      "loss": 0.8026,
+      "step": 3246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.97125254459644,
+      "learning_rate": 1.9003302700026968e-05,
+      "loss": 0.939,
+      "step": 3247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 35.64930619036827,
+      "learning_rate": 1.9002568638839566e-05,
+      "loss": 0.8593,
+      "step": 3248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.51861703018147,
+      "learning_rate": 1.9001834321624378e-05,
+      "loss": 0.9814,
+      "step": 3249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 40.14346630097749,
+      "learning_rate": 1.9001099748402287e-05,
+      "loss": 0.8678,
+      "step": 3250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 34.51861269217497,
+      "learning_rate": 1.9000364919194185e-05,
+      "loss": 0.8131,
+      "step": 3251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 46.13431342361848,
+      "learning_rate": 1.899962983402097e-05,
+      "loss": 0.8471,
+      "step": 3252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 51.054300653791415,
+      "learning_rate": 1.8998894492903545e-05,
+      "loss": 0.8615,
+      "step": 3253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 36.78496289372047,
+      "learning_rate": 1.8998158895862827e-05,
+      "loss": 0.8807,
+      "step": 3254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 53.79586609341125,
+      "learning_rate": 1.8997423042919734e-05,
+      "loss": 1.0024,
+      "step": 3255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 36.751533138966806,
+      "learning_rate": 1.899668693409519e-05,
+      "loss": 0.8324,
+      "step": 3256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 43.801416208612615,
+      "learning_rate": 1.8995950569410136e-05,
+      "loss": 0.8562,
+      "step": 3257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 34.109090614269164,
+      "learning_rate": 1.8995213948885508e-05,
+      "loss": 0.8873,
+      "step": 3258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.39583946624677,
+      "learning_rate": 1.899447707254226e-05,
+      "loss": 0.8565,
+      "step": 3259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 32.683700318799325,
+      "learning_rate": 1.8993739940401342e-05,
+      "loss": 0.7496,
+      "step": 3260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 34.78382233038498,
+      "learning_rate": 1.8993002552483726e-05,
+      "loss": 0.8645,
+      "step": 3261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 40.37137361013321,
+      "learning_rate": 1.8992264908810373e-05,
+      "loss": 0.76,
+      "step": 3262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 29.901471385028614,
+      "learning_rate": 1.899152700940227e-05,
+      "loss": 0.8168,
+      "step": 3263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 34.59687112927598,
+      "learning_rate": 1.8990788854280397e-05,
+      "loss": 0.8596,
+      "step": 3264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 34.725454697677016,
+      "learning_rate": 1.899005044346575e-05,
+      "loss": 0.8142,
+      "step": 3265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 36.383223141610976,
+      "learning_rate": 1.8989311776979326e-05,
+      "loss": 0.7858,
+      "step": 3266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 34.76050103649539,
+      "learning_rate": 1.8988572854842133e-05,
+      "loss": 0.7612,
+      "step": 3267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 50.82744473864912,
+      "learning_rate": 1.898783367707519e-05,
+      "loss": 0.8878,
+      "step": 3268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 29.27273105474439,
+      "learning_rate": 1.898709424369951e-05,
+      "loss": 0.795,
+      "step": 3269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 63.02763758810447,
+      "learning_rate": 1.898635455473613e-05,
+      "loss": 0.9821,
+      "step": 3270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 35.60348744615979,
+      "learning_rate": 1.898561461020608e-05,
+      "loss": 0.8024,
+      "step": 3271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.561684521008196,
+      "learning_rate": 1.8984874410130413e-05,
+      "loss": 0.8294,
+      "step": 3272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 56.67548712868961,
+      "learning_rate": 1.898413395453017e-05,
+      "loss": 1.0093,
+      "step": 3273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 52.860529811151046,
+      "learning_rate": 1.8983393243426413e-05,
+      "loss": 0.9712,
+      "step": 3274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 41.19740486293299,
+      "learning_rate": 1.8982652276840205e-05,
+      "loss": 0.8387,
+      "step": 3275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.49447188384125,
+      "learning_rate": 1.8981911054792625e-05,
+      "loss": 0.9758,
+      "step": 3276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.43442782627515,
+      "learning_rate": 1.8981169577304747e-05,
+      "loss": 0.7945,
+      "step": 3277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 44.23153915127491,
+      "learning_rate": 1.898042784439766e-05,
+      "loss": 0.8975,
+      "step": 3278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 50.297791960014074,
+      "learning_rate": 1.897968585609246e-05,
+      "loss": 0.9212,
+      "step": 3279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.99149289085495,
+      "learning_rate": 1.8978943612410245e-05,
+      "loss": 0.9005,
+      "step": 3280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.38099450872873,
+      "learning_rate": 1.8978201113372128e-05,
+      "loss": 0.7472,
+      "step": 3281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 35.304229078294334,
+      "learning_rate": 1.8977458358999222e-05,
+      "loss": 0.8992,
+      "step": 3282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.85049912918817,
+      "learning_rate": 1.8976715349312652e-05,
+      "loss": 0.8885,
+      "step": 3283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 43.17519558904773,
+      "learning_rate": 1.897597208433355e-05,
+      "loss": 1.0083,
+      "step": 3284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.147885041084976,
+      "learning_rate": 1.8975228564083052e-05,
+      "loss": 0.8671,
+      "step": 3285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 38.48872664362402,
+      "learning_rate": 1.8974484788582303e-05,
+      "loss": 0.9185,
+      "step": 3286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 53.00946391378794,
+      "learning_rate": 1.897374075785246e-05,
+      "loss": 0.9573,
+      "step": 3287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 29.17895617200539,
+      "learning_rate": 1.8972996471914674e-05,
+      "loss": 0.8786,
+      "step": 3288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.538266697287156,
+      "learning_rate": 1.8972251930790124e-05,
+      "loss": 0.989,
+      "step": 3289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 59.968254699569954,
+      "learning_rate": 1.8971507134499974e-05,
+      "loss": 0.8914,
+      "step": 3290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 48.40076310525633,
+      "learning_rate": 1.8970762083065408e-05,
+      "loss": 0.77,
+      "step": 3291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 35.15890422041696,
+      "learning_rate": 1.897001677650762e-05,
+      "loss": 0.7924,
+      "step": 3292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 40.486026042691094,
+      "learning_rate": 1.8969271214847795e-05,
+      "loss": 0.9061,
+      "step": 3293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 45.630679735926456,
+      "learning_rate": 1.896852539810715e-05,
+      "loss": 0.9268,
+      "step": 3294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 43.90488308445158,
+      "learning_rate": 1.8967779326306884e-05,
+      "loss": 0.7986,
+      "step": 3295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 42.920726620504105,
+      "learning_rate": 1.8967032999468226e-05,
+      "loss": 0.8207,
+      "step": 3296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 60.92507701519844,
+      "learning_rate": 1.8966286417612393e-05,
+      "loss": 0.9751,
+      "step": 3297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.787425317793065,
+      "learning_rate": 1.8965539580760615e-05,
+      "loss": 0.9518,
+      "step": 3298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 44.44117238768728,
+      "learning_rate": 1.8964792488934143e-05,
+      "loss": 0.9575,
+      "step": 3299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 41.68950867853599,
+      "learning_rate": 1.8964045142154212e-05,
+      "loss": 0.8548,
+      "step": 3300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 31.873867179337754,
+      "learning_rate": 1.896329754044208e-05,
+      "loss": 0.7148,
+      "step": 3301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 33.89002378852934,
+      "learning_rate": 1.8962549683819013e-05,
+      "loss": 0.9151,
+      "step": 3302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 36.004356074371174,
+      "learning_rate": 1.8961801572306276e-05,
+      "loss": 0.8417,
+      "step": 3303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 43.946058538589305,
+      "learning_rate": 1.8961053205925143e-05,
+      "loss": 0.9261,
+      "step": 3304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.39870726450851,
+      "learning_rate": 1.89603045846969e-05,
+      "loss": 0.7951,
+      "step": 3305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 43.435364902632664,
+      "learning_rate": 1.8959555708642835e-05,
+      "loss": 0.8233,
+      "step": 3306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 42.59149970017733,
+      "learning_rate": 1.895880657778425e-05,
+      "loss": 0.9164,
+      "step": 3307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.94757518811878,
+      "learning_rate": 1.895805719214244e-05,
+      "loss": 0.8529,
+      "step": 3308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 47.76255703133735,
+      "learning_rate": 1.8957307551738727e-05,
+      "loss": 0.8911,
+      "step": 3309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 33.089381948199815,
+      "learning_rate": 1.895655765659443e-05,
+      "loss": 0.8912,
+      "step": 3310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.97875645022401,
+      "learning_rate": 1.8955807506730872e-05,
+      "loss": 0.8964,
+      "step": 3311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 48.54139991435921,
+      "learning_rate": 1.8955057102169385e-05,
+      "loss": 0.9511,
+      "step": 3312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 41.33417545442614,
+      "learning_rate": 1.8954306442931315e-05,
+      "loss": 0.8119,
+      "step": 3313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 33.61019787918265,
+      "learning_rate": 1.8953555529038006e-05,
+      "loss": 0.758,
+      "step": 3314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 40.704719380436565,
+      "learning_rate": 1.8952804360510816e-05,
+      "loss": 0.9344,
+      "step": 3315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 40.520897380644215,
+      "learning_rate": 1.8952052937371105e-05,
+      "loss": 0.9141,
+      "step": 3316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 99.75934500648266,
+      "learning_rate": 1.8951301259640252e-05,
+      "loss": 0.9608,
+      "step": 3317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 33.58627869316692,
+      "learning_rate": 1.895054932733962e-05,
+      "loss": 0.927,
+      "step": 3318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 85.23637160664624,
+      "learning_rate": 1.8949797140490607e-05,
+      "loss": 0.9472,
+      "step": 3319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 33.861230412984895,
+      "learning_rate": 1.8949044699114592e-05,
+      "loss": 0.7174,
+      "step": 3320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 34.96555105786744,
+      "learning_rate": 1.8948292003232987e-05,
+      "loss": 0.7923,
+      "step": 3321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 40.50338629117439,
+      "learning_rate": 1.894753905286719e-05,
+      "loss": 0.9557,
+      "step": 3322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.01436875380792,
+      "learning_rate": 1.8946785848038614e-05,
+      "loss": 0.9206,
+      "step": 3323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 45.834318931848564,
+      "learning_rate": 1.8946032388768687e-05,
+      "loss": 0.9506,
+      "step": 3324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.076413236860404,
+      "learning_rate": 1.8945278675078828e-05,
+      "loss": 0.9713,
+      "step": 3325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 44.749840816022875,
+      "learning_rate": 1.894452470699048e-05,
+      "loss": 0.8138,
+      "step": 3326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.1833274796902,
+      "learning_rate": 1.894377048452508e-05,
+      "loss": 0.8908,
+      "step": 3327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 90.70664472034747,
+      "learning_rate": 1.8943016007704078e-05,
+      "loss": 1.0055,
+      "step": 3328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 36.05846756020525,
+      "learning_rate": 1.8942261276548932e-05,
+      "loss": 0.8556,
+      "step": 3329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.134909176916516,
+      "learning_rate": 1.894150629108111e-05,
+      "loss": 0.9196,
+      "step": 3330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 31.646684494944374,
+      "learning_rate": 1.8940751051322075e-05,
+      "loss": 0.9159,
+      "step": 3331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.506729077930274,
+      "learning_rate": 1.8939995557293315e-05,
+      "loss": 0.9026,
+      "step": 3332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 35.78581511847956,
+      "learning_rate": 1.8939239809016306e-05,
+      "loss": 0.9163,
+      "step": 3333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 44.700513582778136,
+      "learning_rate": 1.893848380651255e-05,
+      "loss": 0.8278,
+      "step": 3334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 43.0056041652959,
+      "learning_rate": 1.893772754980354e-05,
+      "loss": 0.9128,
+      "step": 3335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.00411038378717,
+      "learning_rate": 1.8936971038910792e-05,
+      "loss": 0.831,
+      "step": 3336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.19426286895383,
+      "learning_rate": 1.8936214273855813e-05,
+      "loss": 0.9833,
+      "step": 3337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 31.757193381267665,
+      "learning_rate": 1.8935457254660128e-05,
+      "loss": 0.8412,
+      "step": 3338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 30.26270503825924,
+      "learning_rate": 1.8934699981345265e-05,
+      "loss": 0.7865,
+      "step": 3339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.919243557496145,
+      "learning_rate": 1.893394245393276e-05,
+      "loss": 0.8679,
+      "step": 3340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 35.55381904937881,
+      "learning_rate": 1.893318467244416e-05,
+      "loss": 0.7781,
+      "step": 3341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 45.48749280731907,
+      "learning_rate": 1.893242663690101e-05,
+      "loss": 0.9325,
+      "step": 3342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 35.44347484954709,
+      "learning_rate": 1.8931668347324877e-05,
+      "loss": 0.7908,
+      "step": 3343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.658447820964845,
+      "learning_rate": 1.893090980373732e-05,
+      "loss": 0.887,
+      "step": 3344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 42.337034609331326,
+      "learning_rate": 1.8930151006159908e-05,
+      "loss": 0.9158,
+      "step": 3345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 35.27951110392971,
+      "learning_rate": 1.892939195461423e-05,
+      "loss": 0.8649,
+      "step": 3346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.09430210568328,
+      "learning_rate": 1.8928632649121867e-05,
+      "loss": 0.8827,
+      "step": 3347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.448565594281426,
+      "learning_rate": 1.8927873089704416e-05,
+      "loss": 0.9621,
+      "step": 3348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 41.23440568524571,
+      "learning_rate": 1.8927113276383475e-05,
+      "loss": 0.8913,
+      "step": 3349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 34.69230451903881,
+      "learning_rate": 1.8926353209180655e-05,
+      "loss": 0.8128,
+      "step": 3350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 34.15150627777014,
+      "learning_rate": 1.892559288811757e-05,
+      "loss": 0.893,
+      "step": 3351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 31.238097399845213,
+      "learning_rate": 1.892483231321585e-05,
+      "loss": 1.005,
+      "step": 3352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.89109983112196,
+      "learning_rate": 1.8924071484497114e-05,
+      "loss": 0.884,
+      "step": 3353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 31.78220423388345,
+      "learning_rate": 1.892331040198301e-05,
+      "loss": 0.8024,
+      "step": 3354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 36.24547328308967,
+      "learning_rate": 1.8922549065695172e-05,
+      "loss": 0.8809,
+      "step": 3355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 48.75104606190805,
+      "learning_rate": 1.8921787475655265e-05,
+      "loss": 0.8756,
+      "step": 3356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 44.139975910646875,
+      "learning_rate": 1.8921025631884938e-05,
+      "loss": 0.8885,
+      "step": 3357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 39.220001206823525,
+      "learning_rate": 1.8920263534405858e-05,
+      "loss": 0.8929,
+      "step": 3358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 34.533807630454604,
+      "learning_rate": 1.8919501183239705e-05,
+      "loss": 0.8765,
+      "step": 3359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 37.286545171990454,
+      "learning_rate": 1.8918738578408157e-05,
+      "loss": 0.806,
+      "step": 3360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 29.618518659562003,
+      "learning_rate": 1.8917975719932898e-05,
+      "loss": 0.7547,
+      "step": 3361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 32.73596638382401,
+      "learning_rate": 1.891721260783563e-05,
+      "loss": 0.8547,
+      "step": 3362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.96333737322556,
+      "learning_rate": 1.8916449242138047e-05,
+      "loss": 0.9537,
+      "step": 3363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 39.98822077524588,
+      "learning_rate": 1.8915685622861864e-05,
+      "loss": 0.7995,
+      "step": 3364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 42.449984948567845,
+      "learning_rate": 1.8914921750028804e-05,
+      "loss": 0.9191,
+      "step": 3365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 46.635230775141274,
+      "learning_rate": 1.891415762366058e-05,
+      "loss": 0.9021,
+      "step": 3366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.23501184180221,
+      "learning_rate": 1.8913393243778927e-05,
+      "loss": 0.8723,
+      "step": 3367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 34.76248549783755,
+      "learning_rate": 1.8912628610405585e-05,
+      "loss": 0.8645,
+      "step": 3368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 42.68277868800257,
+      "learning_rate": 1.89118637235623e-05,
+      "loss": 0.8805,
+      "step": 3369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 43.763858954286455,
+      "learning_rate": 1.891109858327082e-05,
+      "loss": 0.9587,
+      "step": 3370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 41.257578633329466,
+      "learning_rate": 1.8910333189552913e-05,
+      "loss": 0.8642,
+      "step": 3371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 42.80209840706345,
+      "learning_rate": 1.8909567542430344e-05,
+      "loss": 0.9428,
+      "step": 3372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 33.02894513146644,
+      "learning_rate": 1.890880164192488e-05,
+      "loss": 0.806,
+      "step": 3373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 34.63157806846927,
+      "learning_rate": 1.8908035488058316e-05,
+      "loss": 0.8514,
+      "step": 3374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.01785607173977,
+      "learning_rate": 1.8907269080852432e-05,
+      "loss": 0.9508,
+      "step": 3375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 30.696386452963544,
+      "learning_rate": 1.8906502420329022e-05,
+      "loss": 0.8079,
+      "step": 3376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 45.64861325124543,
+      "learning_rate": 1.8905735506509897e-05,
+      "loss": 0.8827,
+      "step": 3377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 44.7288892702131,
+      "learning_rate": 1.890496833941686e-05,
+      "loss": 1.0891,
+      "step": 3378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.408920275292985,
+      "learning_rate": 1.890420091907174e-05,
+      "loss": 0.816,
+      "step": 3379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 42.305980704170835,
+      "learning_rate": 1.8903433245496348e-05,
+      "loss": 0.8949,
+      "step": 3380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 44.47345112059678,
+      "learning_rate": 1.8902665318712527e-05,
+      "loss": 0.9544,
+      "step": 3381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 51.762408754498175,
+      "learning_rate": 1.8901897138742107e-05,
+      "loss": 0.9281,
+      "step": 3382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 32.46987498943958,
+      "learning_rate": 1.890112870560694e-05,
+      "loss": 0.8561,
+      "step": 3383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 40.02294407781834,
+      "learning_rate": 1.8900360019328885e-05,
+      "loss": 0.8559,
+      "step": 3384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 34.534937817871175,
+      "learning_rate": 1.8899591079929796e-05,
+      "loss": 0.883,
+      "step": 3385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 39.498531537585556,
+      "learning_rate": 1.8898821887431543e-05,
+      "loss": 0.9269,
+      "step": 3386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 39.24542387630519,
+      "learning_rate": 1.8898052441855997e-05,
+      "loss": 0.8847,
+      "step": 3387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.54105652148557,
+      "learning_rate": 1.8897282743225048e-05,
+      "loss": 0.8125,
+      "step": 3388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 40.58762197661215,
+      "learning_rate": 1.8896512791560584e-05,
+      "loss": 0.9281,
+      "step": 3389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 44.29702026235724,
+      "learning_rate": 1.8895742586884502e-05,
+      "loss": 0.9812,
+      "step": 3390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.19952765559693,
+      "learning_rate": 1.88949721292187e-05,
+      "loss": 0.8207,
+      "step": 3391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 43.38983997573057,
+      "learning_rate": 1.8894201418585094e-05,
+      "loss": 0.8373,
+      "step": 3392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 46.778962430403084,
+      "learning_rate": 1.8893430455005604e-05,
+      "loss": 0.7356,
+      "step": 3393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 31.431984533259527,
+      "learning_rate": 1.8892659238502154e-05,
+      "loss": 0.7546,
+      "step": 3394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.28771744883353,
+      "learning_rate": 1.889188776909668e-05,
+      "loss": 0.9869,
+      "step": 3395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 41.35750273245121,
+      "learning_rate": 1.8891116046811116e-05,
+      "loss": 0.8845,
+      "step": 3396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 32.86534174403244,
+      "learning_rate": 1.8890344071667415e-05,
+      "loss": 0.749,
+      "step": 3397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 67.99286787413047,
+      "learning_rate": 1.888957184368753e-05,
+      "loss": 0.8928,
+      "step": 3398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 46.79788248190827,
+      "learning_rate": 1.888879936289342e-05,
+      "loss": 0.8541,
+      "step": 3399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 40.56982375448701,
+      "learning_rate": 1.8888026629307056e-05,
+      "loss": 0.8153,
+      "step": 3400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.63260423294986,
+      "learning_rate": 1.8887253642950414e-05,
+      "loss": 0.8662,
+      "step": 3401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 55.71145141775363,
+      "learning_rate": 1.8886480403845477e-05,
+      "loss": 0.8145,
+      "step": 3402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 33.6354742151081,
+      "learning_rate": 1.888570691201424e-05,
+      "loss": 0.9046,
+      "step": 3403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 46.683074046872925,
+      "learning_rate": 1.8884933167478688e-05,
+      "loss": 0.9717,
+      "step": 3404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.87132392958934,
+      "learning_rate": 1.888415917026084e-05,
+      "loss": 0.8334,
+      "step": 3405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.40131441856297,
+      "learning_rate": 1.8883384920382703e-05,
+      "loss": 0.8798,
+      "step": 3406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 40.90291381357078,
+      "learning_rate": 1.8882610417866295e-05,
+      "loss": 0.8607,
+      "step": 3407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.404743737891934,
+      "learning_rate": 1.888183566273364e-05,
+      "loss": 0.9186,
+      "step": 3408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.66698967112252,
+      "learning_rate": 1.888106065500678e-05,
+      "loss": 0.9273,
+      "step": 3409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 40.91110760760829,
+      "learning_rate": 1.8880285394707743e-05,
+      "loss": 0.9419,
+      "step": 3410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 29.280502867556017,
+      "learning_rate": 1.8879509881858587e-05,
+      "loss": 0.8061,
+      "step": 3411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.8614120164234,
+      "learning_rate": 1.8878734116481364e-05,
+      "loss": 0.9715,
+      "step": 3412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 37.69874814203913,
+      "learning_rate": 1.887795809859814e-05,
+      "loss": 0.8813,
+      "step": 3413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 27.947405405352814,
+      "learning_rate": 1.8877181828230978e-05,
+      "loss": 0.752,
+      "step": 3414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 33.846212890984475,
+      "learning_rate": 1.887640530540196e-05,
+      "loss": 0.8012,
+      "step": 3415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.6529589425247,
+      "learning_rate": 1.8875628530133164e-05,
+      "loss": 0.7932,
+      "step": 3416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 28.76469931700776,
+      "learning_rate": 1.887485150244669e-05,
+      "loss": 0.8687,
+      "step": 3417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 32.021678510509595,
+      "learning_rate": 1.8874074222364626e-05,
+      "loss": 0.8756,
+      "step": 3418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 34.220322664256834,
+      "learning_rate": 1.8873296689909084e-05,
+      "loss": 0.8779,
+      "step": 3419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 35.02402601868699,
+      "learning_rate": 1.8872518905102177e-05,
+      "loss": 0.9194,
+      "step": 3420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.86144943743689,
+      "learning_rate": 1.887174086796602e-05,
+      "loss": 0.812,
+      "step": 3421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 31.547031993579242,
+      "learning_rate": 1.8870962578522746e-05,
+      "loss": 0.8357,
+      "step": 3422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.93623953749237,
+      "learning_rate": 1.887018403679448e-05,
+      "loss": 0.9661,
+      "step": 3423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 50.55679705962011,
+      "learning_rate": 1.8869405242803373e-05,
+      "loss": 0.8039,
+      "step": 3424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 39.11717982641609,
+      "learning_rate": 1.886862619657157e-05,
+      "loss": 0.8618,
+      "step": 3425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.328522474630425,
+      "learning_rate": 1.8867846898121223e-05,
+      "loss": 0.8022,
+      "step": 3426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 43.74793363868428,
+      "learning_rate": 1.8867067347474498e-05,
+      "loss": 0.9239,
+      "step": 3427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 38.63326836636215,
+      "learning_rate": 1.8866287544653566e-05,
+      "loss": 0.9471,
+      "step": 3428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 43.89444660349363,
+      "learning_rate": 1.8865507489680602e-05,
+      "loss": 0.8819,
+      "step": 3429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 29.350893154640207,
+      "learning_rate": 1.8864727182577792e-05,
+      "loss": 0.8737,
+      "step": 3430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 45.206508342395544,
+      "learning_rate": 1.8863946623367324e-05,
+      "loss": 0.8639,
+      "step": 3431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 39.829731989243044,
+      "learning_rate": 1.88631658120714e-05,
+      "loss": 0.8177,
+      "step": 3432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 39.27116960268551,
+      "learning_rate": 1.886238474871223e-05,
+      "loss": 0.995,
+      "step": 3433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 34.95825792667469,
+      "learning_rate": 1.8861603433312017e-05,
+      "loss": 0.7334,
+      "step": 3434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 30.66643305660154,
+      "learning_rate": 1.8860821865892988e-05,
+      "loss": 0.8799,
+      "step": 3435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 31.588799614205616,
+      "learning_rate": 1.8860040046477367e-05,
+      "loss": 0.7894,
+      "step": 3436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.78850319988882,
+      "learning_rate": 1.8859257975087395e-05,
+      "loss": 0.8141,
+      "step": 3437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 47.241183906646015,
+      "learning_rate": 1.8858475651745304e-05,
+      "loss": 0.7915,
+      "step": 3438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 46.434030893571766,
+      "learning_rate": 1.8857693076473348e-05,
+      "loss": 0.9377,
+      "step": 3439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 48.594284162621314,
+      "learning_rate": 1.8856910249293783e-05,
+      "loss": 0.8399,
+      "step": 3440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 41.831399973930054,
+      "learning_rate": 1.8856127170228873e-05,
+      "loss": 0.8848,
+      "step": 3441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 36.161575413160506,
+      "learning_rate": 1.8855343839300885e-05,
+      "loss": 0.896,
+      "step": 3442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 41.868717918036346,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.8273,
+      "step": 3443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 41.11124964825311,
+      "learning_rate": 1.88537764219448e-05,
+      "loss": 0.9873,
+      "step": 3444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.00184343708885,
+      "learning_rate": 1.885299233556128e-05,
+      "loss": 0.876,
+      "step": 3445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 53.98265277417469,
+      "learning_rate": 1.8852207997403835e-05,
+      "loss": 1.0623,
+      "step": 3446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.00850631436122,
+      "learning_rate": 1.8851423407494774e-05,
+      "loss": 0.8315,
+      "step": 3447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.525491930664955,
+      "learning_rate": 1.8850638565856406e-05,
+      "loss": 0.8245,
+      "step": 3448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.467692913150024,
+      "learning_rate": 1.8849853472511057e-05,
+      "loss": 0.8192,
+      "step": 3449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.948545360714256,
+      "learning_rate": 1.8849068127481055e-05,
+      "loss": 0.7923,
+      "step": 3450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.42567454675402,
+      "learning_rate": 1.884828253078873e-05,
+      "loss": 0.899,
+      "step": 3451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 27.322479901668956,
+      "learning_rate": 1.8847496682456422e-05,
+      "loss": 0.7515,
+      "step": 3452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 54.33703520027362,
+      "learning_rate": 1.884671058250649e-05,
+      "loss": 0.9327,
+      "step": 3453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 38.4050139437018,
+      "learning_rate": 1.8845924230961278e-05,
+      "loss": 0.8041,
+      "step": 3454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.7037795993888,
+      "learning_rate": 1.884513762784316e-05,
+      "loss": 0.8699,
+      "step": 3455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 39.106032530070344,
+      "learning_rate": 1.8844350773174502e-05,
+      "loss": 0.8521,
+      "step": 3456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.79261436248193,
+      "learning_rate": 1.884356366697768e-05,
+      "loss": 0.8973,
+      "step": 3457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 49.059509254221815,
+      "learning_rate": 1.8842776309275083e-05,
+      "loss": 0.825,
+      "step": 3458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 41.48138095662038,
+      "learning_rate": 1.8841988700089096e-05,
+      "loss": 0.9494,
+      "step": 3459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 38.85017848814585,
+      "learning_rate": 1.8841200839442128e-05,
+      "loss": 0.8979,
+      "step": 3460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 46.89198359727007,
+      "learning_rate": 1.8840412727356576e-05,
+      "loss": 0.8817,
+      "step": 3461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 41.112381944402784,
+      "learning_rate": 1.883962436385486e-05,
+      "loss": 0.8321,
+      "step": 3462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 36.152050061220564,
+      "learning_rate": 1.88388357489594e-05,
+      "loss": 0.8964,
+      "step": 3463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.06574267075914,
+      "learning_rate": 1.883804688269262e-05,
+      "loss": 0.8327,
+      "step": 3464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 36.66344243778506,
+      "learning_rate": 1.8837257765076956e-05,
+      "loss": 0.8684,
+      "step": 3465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 44.634814574885375,
+      "learning_rate": 1.883646839613485e-05,
+      "loss": 0.9158,
+      "step": 3466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.57250477941282,
+      "learning_rate": 1.8835678775888752e-05,
+      "loss": 0.7699,
+      "step": 3467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.323365500445504,
+      "learning_rate": 1.883488890436112e-05,
+      "loss": 0.7862,
+      "step": 3468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 31.375492844377273,
+      "learning_rate": 1.8834098781574415e-05,
+      "loss": 0.7263,
+      "step": 3469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.79771535668224,
+      "learning_rate": 1.883330840755111e-05,
+      "loss": 0.7954,
+      "step": 3470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.76282432130738,
+      "learning_rate": 1.883251778231368e-05,
+      "loss": 0.8613,
+      "step": 3471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 66.19365353026629,
+      "learning_rate": 1.883172690588461e-05,
+      "loss": 0.7546,
+      "step": 3472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 44.43123870175495,
+      "learning_rate": 1.8830935778286393e-05,
+      "loss": 0.9374,
+      "step": 3473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.93990565333775,
+      "learning_rate": 1.8830144399541533e-05,
+      "loss": 0.8022,
+      "step": 3474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 42.04928824914398,
+      "learning_rate": 1.8829352769672525e-05,
+      "loss": 0.781,
+      "step": 3475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 38.64057273876637,
+      "learning_rate": 1.8828560888701895e-05,
+      "loss": 0.9109,
+      "step": 3476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 38.56566254493636,
+      "learning_rate": 1.8827768756652155e-05,
+      "loss": 0.9038,
+      "step": 3477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.704727255025105,
+      "learning_rate": 1.8826976373545838e-05,
+      "loss": 0.8298,
+      "step": 3478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.075708739803126,
+      "learning_rate": 1.8826183739405475e-05,
+      "loss": 0.7056,
+      "step": 3479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 53.907678455831736,
+      "learning_rate": 1.8825390854253605e-05,
+      "loss": 0.8552,
+      "step": 3480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 31.82995839894339,
+      "learning_rate": 1.882459771811279e-05,
+      "loss": 0.7898,
+      "step": 3481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 39.10996013330131,
+      "learning_rate": 1.8823804331005573e-05,
+      "loss": 0.9374,
+      "step": 3482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 34.12185975881544,
+      "learning_rate": 1.8823010692954523e-05,
+      "loss": 0.7801,
+      "step": 3483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.74996563436284,
+      "learning_rate": 1.8822216803982214e-05,
+      "loss": 0.8568,
+      "step": 3484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.02275897366054,
+      "learning_rate": 1.8821422664111218e-05,
+      "loss": 0.8237,
+      "step": 3485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 47.42317945015422,
+      "learning_rate": 1.882062827336412e-05,
+      "loss": 0.896,
+      "step": 3486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.82562682354783,
+      "learning_rate": 1.8819833631763514e-05,
+      "loss": 0.8585,
+      "step": 3487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 35.75714008014939,
+      "learning_rate": 1.8819038739332e-05,
+      "loss": 0.83,
+      "step": 3488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 42.979603859321514,
+      "learning_rate": 1.8818243596092182e-05,
+      "loss": 0.9287,
+      "step": 3489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.73853447575994,
+      "learning_rate": 1.8817448202066677e-05,
+      "loss": 0.8212,
+      "step": 3490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.21119311427957,
+      "learning_rate": 1.88166525572781e-05,
+      "loss": 0.7755,
+      "step": 3491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.79659010882023,
+      "learning_rate": 1.8815856661749085e-05,
+      "loss": 0.8209,
+      "step": 3492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 32.10825041869053,
+      "learning_rate": 1.8815060515502262e-05,
+      "loss": 0.8552,
+      "step": 3493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 36.15529579215691,
+      "learning_rate": 1.8814264118560274e-05,
+      "loss": 0.8506,
+      "step": 3494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 42.2486540280896,
+      "learning_rate": 1.8813467470945774e-05,
+      "loss": 0.8847,
+      "step": 3495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 45.67072958983024,
+      "learning_rate": 1.8812670572681413e-05,
+      "loss": 1.0456,
+      "step": 3496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.85249528240063,
+      "learning_rate": 1.8811873423789853e-05,
+      "loss": 0.8985,
+      "step": 3497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 37.728574947573776,
+      "learning_rate": 1.8811076024293774e-05,
+      "loss": 0.8935,
+      "step": 3498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 42.00618543661066,
+      "learning_rate": 1.881027837421584e-05,
+      "loss": 0.8607,
+      "step": 3499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 34.495306925914214,
+      "learning_rate": 1.880948047357875e-05,
+      "loss": 0.8443,
+      "step": 3500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 35.92331028311358,
+      "learning_rate": 1.8808682322405184e-05,
+      "loss": 0.8877,
+      "step": 3501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 45.64600820898174,
+      "learning_rate": 1.8807883920717844e-05,
+      "loss": 0.9383,
+      "step": 3502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 31.893677579874907,
+      "learning_rate": 1.8807085268539443e-05,
+      "loss": 0.7928,
+      "step": 3503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 37.448747247744016,
+      "learning_rate": 1.8806286365892685e-05,
+      "loss": 0.885,
+      "step": 3504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 37.31019451616182,
+      "learning_rate": 1.8805487212800297e-05,
+      "loss": 0.8502,
+      "step": 3505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 33.84291662808651,
+      "learning_rate": 1.8804687809285003e-05,
+      "loss": 0.8051,
+      "step": 3506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.16059364008045,
+      "learning_rate": 1.8803888155369538e-05,
+      "loss": 0.8405,
+      "step": 3507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 43.53791292001859,
+      "learning_rate": 1.8803088251076642e-05,
+      "loss": 0.8766,
+      "step": 3508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 37.68379569946093,
+      "learning_rate": 1.880228809642907e-05,
+      "loss": 0.8419,
+      "step": 3509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.7430383273818,
+      "learning_rate": 1.880148769144957e-05,
+      "loss": 0.7927,
+      "step": 3510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 32.576165793074274,
+      "learning_rate": 1.880068703616091e-05,
+      "loss": 0.8266,
+      "step": 3511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 69.62720774329786,
+      "learning_rate": 1.8799886130585858e-05,
+      "loss": 0.8624,
+      "step": 3512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 37.09932837765187,
+      "learning_rate": 1.8799084974747193e-05,
+      "loss": 0.9797,
+      "step": 3513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.43129698567381,
+      "learning_rate": 1.8798283568667698e-05,
+      "loss": 0.9903,
+      "step": 3514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 47.87945919335358,
+      "learning_rate": 1.8797481912370167e-05,
+      "loss": 0.8846,
+      "step": 3515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 46.25426307826018,
+      "learning_rate": 1.8796680005877397e-05,
+      "loss": 0.9225,
+      "step": 3516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 32.40169596883942,
+      "learning_rate": 1.879587784921219e-05,
+      "loss": 0.8157,
+      "step": 3517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 34.88603613504591,
+      "learning_rate": 1.8795075442397367e-05,
+      "loss": 0.9731,
+      "step": 3518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.76862065811143,
+      "learning_rate": 1.879427278545574e-05,
+      "loss": 0.8829,
+      "step": 3519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 31.878873413951535,
+      "learning_rate": 1.8793469878410142e-05,
+      "loss": 0.7022,
+      "step": 3520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.730292590182735,
+      "learning_rate": 1.8792666721283406e-05,
+      "loss": 0.9074,
+      "step": 3521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.35663122992348,
+      "learning_rate": 1.8791863314098368e-05,
+      "loss": 0.8209,
+      "step": 3522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 59.882669264291756,
+      "learning_rate": 1.879105965687788e-05,
+      "loss": 0.884,
+      "step": 3523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 35.34046338599543,
+      "learning_rate": 1.87902557496448e-05,
+      "loss": 0.8127,
+      "step": 3524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 46.64724757782433,
+      "learning_rate": 1.878945159242199e-05,
+      "loss": 0.8585,
+      "step": 3525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 30.681209610962757,
+      "learning_rate": 1.878864718523232e-05,
+      "loss": 0.8275,
+      "step": 3526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.165309842858136,
+      "learning_rate": 1.8787842528098657e-05,
+      "loss": 0.9033,
+      "step": 3527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.78593627378911,
+      "learning_rate": 1.87870376210439e-05,
+      "loss": 0.9972,
+      "step": 3528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 34.485200375149994,
+      "learning_rate": 1.878623246409093e-05,
+      "loss": 0.8758,
+      "step": 3529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.11075123951013,
+      "learning_rate": 1.878542705726265e-05,
+      "loss": 0.8239,
+      "step": 3530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.5261612744003,
+      "learning_rate": 1.8784621400581966e-05,
+      "loss": 0.8792,
+      "step": 3531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.370432606961366,
+      "learning_rate": 1.8783815494071786e-05,
+      "loss": 0.9322,
+      "step": 3532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.394574832182634,
+      "learning_rate": 1.8783009337755032e-05,
+      "loss": 0.9094,
+      "step": 3533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 43.483275073029354,
+      "learning_rate": 1.878220293165463e-05,
+      "loss": 0.827,
+      "step": 3534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 72.1168710335743,
+      "learning_rate": 1.8781396275793516e-05,
+      "loss": 0.8115,
+      "step": 3535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 35.172474093865894,
+      "learning_rate": 1.8780589370194626e-05,
+      "loss": 0.8569,
+      "step": 3536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.100520586733616,
+      "learning_rate": 1.8779782214880912e-05,
+      "loss": 0.8732,
+      "step": 3537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.87090819838783,
+      "learning_rate": 1.8778974809875326e-05,
+      "loss": 0.8192,
+      "step": 3538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.04819442685792,
+      "learning_rate": 1.8778167155200836e-05,
+      "loss": 0.9211,
+      "step": 3539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 34.50544177205516,
+      "learning_rate": 1.8777359250880404e-05,
+      "loss": 0.6896,
+      "step": 3540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.99805298179358,
+      "learning_rate": 1.8776551096937014e-05,
+      "loss": 0.8305,
+      "step": 3541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 37.502298405465716,
+      "learning_rate": 1.8775742693393638e-05,
+      "loss": 0.8153,
+      "step": 3542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.5840998713631,
+      "learning_rate": 1.8774934040273278e-05,
+      "loss": 0.851,
+      "step": 3543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 36.19458179325019,
+      "learning_rate": 1.8774125137598926e-05,
+      "loss": 0.7836,
+      "step": 3544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 31.630698949554827,
+      "learning_rate": 1.877331598539359e-05,
+      "loss": 0.8352,
+      "step": 3545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 35.885459584661255,
+      "learning_rate": 1.8772506583680273e-05,
+      "loss": 0.8399,
+      "step": 3546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.55157247097922,
+      "learning_rate": 1.8771696932482008e-05,
+      "loss": 0.7689,
+      "step": 3547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 73.25886449458197,
+      "learning_rate": 1.8770887031821813e-05,
+      "loss": 0.8087,
+      "step": 3548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 38.83151958524879,
+      "learning_rate": 1.8770076881722717e-05,
+      "loss": 0.8261,
+      "step": 3549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 31.90901574861052,
+      "learning_rate": 1.8769266482207766e-05,
+      "loss": 0.8471,
+      "step": 3550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 44.62552979365848,
+      "learning_rate": 1.8768455833300007e-05,
+      "loss": 0.8758,
+      "step": 3551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.25423611982839,
+      "learning_rate": 1.8767644935022496e-05,
+      "loss": 0.8637,
+      "step": 3552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 43.48426310024194,
+      "learning_rate": 1.876683378739829e-05,
+      "loss": 0.7189,
+      "step": 3553
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.267285419386994,
+      "learning_rate": 1.876602239045046e-05,
+      "loss": 0.7404,
+      "step": 3554
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 37.01741925226956,
+      "learning_rate": 1.876521074420208e-05,
+      "loss": 0.9459,
+      "step": 3555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 42.0595671267507,
+      "learning_rate": 1.8764398848676232e-05,
+      "loss": 0.9053,
+      "step": 3556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.788678523738255,
+      "learning_rate": 1.876358670389601e-05,
+      "loss": 0.8765,
+      "step": 3557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 30.256645140720256,
+      "learning_rate": 1.8762774309884508e-05,
+      "loss": 0.7965,
+      "step": 3558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 36.19459066941287,
+      "learning_rate": 1.876196166666483e-05,
+      "loss": 0.8195,
+      "step": 3559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 49.604892630340245,
+      "learning_rate": 1.876114877426009e-05,
+      "loss": 0.7408,
+      "step": 3560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 53.220365516571455,
+      "learning_rate": 1.8760335632693404e-05,
+      "loss": 0.7648,
+      "step": 3561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 36.33858472309329,
+      "learning_rate": 1.8759522241987896e-05,
+      "loss": 0.8692,
+      "step": 3562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 38.70589131897922,
+      "learning_rate": 1.87587086021667e-05,
+      "loss": 0.8715,
+      "step": 3563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.8431887936408,
+      "learning_rate": 1.8757894713252956e-05,
+      "loss": 0.8643,
+      "step": 3564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.10622017450645,
+      "learning_rate": 1.8757080575269808e-05,
+      "loss": 0.9143,
+      "step": 3565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 36.866099030112466,
+      "learning_rate": 1.8756266188240413e-05,
+      "loss": 0.8628,
+      "step": 3566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 40.34171588181302,
+      "learning_rate": 1.875545155218793e-05,
+      "loss": 0.976,
+      "step": 3567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.42540632920246,
+      "learning_rate": 1.8754636667135523e-05,
+      "loss": 0.841,
+      "step": 3568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.3023238058019,
+      "learning_rate": 1.8753821533106373e-05,
+      "loss": 0.8515,
+      "step": 3569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 40.449958839442715,
+      "learning_rate": 1.8753006150123662e-05,
+      "loss": 0.86,
+      "step": 3570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 33.48527535257398,
+      "learning_rate": 1.8752190518210572e-05,
+      "loss": 0.846,
+      "step": 3571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.0977052598155,
+      "learning_rate": 1.8751374637390304e-05,
+      "loss": 0.8851,
+      "step": 3572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 30.911690575429702,
+      "learning_rate": 1.8750558507686065e-05,
+      "loss": 0.8158,
+      "step": 3573
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 31.595843863599722,
+      "learning_rate": 1.8749742129121058e-05,
+      "loss": 0.8172,
+      "step": 3574
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 56.93827257773281,
+      "learning_rate": 1.87489255017185e-05,
+      "loss": 0.9761,
+      "step": 3575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 41.907339413312506,
+      "learning_rate": 1.8748108625501623e-05,
+      "loss": 0.8297,
+      "step": 3576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 38.02876168278475,
+      "learning_rate": 1.8747291500493653e-05,
+      "loss": 0.9019,
+      "step": 3577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 42.577113859436935,
+      "learning_rate": 1.874647412671783e-05,
+      "loss": 0.757,
+      "step": 3578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 44.79997531951923,
+      "learning_rate": 1.87456565041974e-05,
+      "loss": 0.7811,
+      "step": 3579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 29.6307629762288,
+      "learning_rate": 1.8744838632955617e-05,
+      "loss": 0.8438,
+      "step": 3580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 29.093002219582825,
+      "learning_rate": 1.8744020513015733e-05,
+      "loss": 0.8407,
+      "step": 3581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.25983749198563,
+      "learning_rate": 1.8743202144401027e-05,
+      "loss": 0.7368,
+      "step": 3582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 53.836661626639874,
+      "learning_rate": 1.874238352713476e-05,
+      "loss": 0.9904,
+      "step": 3583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 31.747616017567676,
+      "learning_rate": 1.8741564661240223e-05,
+      "loss": 0.8134,
+      "step": 3584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 59.43151537332189,
+      "learning_rate": 1.87407455467407e-05,
+      "loss": 0.941,
+      "step": 3585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 46.14964572609329,
+      "learning_rate": 1.8739926183659492e-05,
+      "loss": 1.0615,
+      "step": 3586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 42.79662849101798,
+      "learning_rate": 1.873910657201989e-05,
+      "loss": 0.8106,
+      "step": 3587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 38.92916097482055,
+      "learning_rate": 1.873828671184521e-05,
+      "loss": 0.8527,
+      "step": 3588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 37.33870121173328,
+      "learning_rate": 1.873746660315877e-05,
+      "loss": 0.8448,
+      "step": 3589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 31.905741382098096,
+      "learning_rate": 1.873664624598389e-05,
+      "loss": 0.8751,
+      "step": 3590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.440762062447284,
+      "learning_rate": 1.8735825640343904e-05,
+      "loss": 0.8885,
+      "step": 3591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 44.35108597811732,
+      "learning_rate": 1.8735004786262142e-05,
+      "loss": 0.7694,
+      "step": 3592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 28.918576073520285,
+      "learning_rate": 1.8734183683761957e-05,
+      "loss": 0.7844,
+      "step": 3593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.457769306660715,
+      "learning_rate": 1.8733362332866698e-05,
+      "loss": 0.7104,
+      "step": 3594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 41.81297428848567,
+      "learning_rate": 1.873254073359972e-05,
+      "loss": 0.807,
+      "step": 3595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.51270200719367,
+      "learning_rate": 1.8731718885984396e-05,
+      "loss": 0.8245,
+      "step": 3596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 39.749002525122215,
+      "learning_rate": 1.8730896790044092e-05,
+      "loss": 0.8639,
+      "step": 3597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 37.320877674623425,
+      "learning_rate": 1.873007444580219e-05,
+      "loss": 0.8803,
+      "step": 3598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 35.35162489743285,
+      "learning_rate": 1.872925185328208e-05,
+      "loss": 0.8252,
+      "step": 3599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 44.23413867947949,
+      "learning_rate": 1.872842901250715e-05,
+      "loss": 0.8347,
+      "step": 3600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 37.71337787894826,
+      "learning_rate": 1.872760592350081e-05,
+      "loss": 0.7414,
+      "step": 3601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 37.58846439231882,
+      "learning_rate": 1.872678258628646e-05,
+      "loss": 0.9823,
+      "step": 3602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 40.58786079615275,
+      "learning_rate": 1.872595900088752e-05,
+      "loss": 0.8235,
+      "step": 3603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 42.88383179389864,
+      "learning_rate": 1.872513516732741e-05,
+      "loss": 0.8502,
+      "step": 3604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 46.12621854707687,
+      "learning_rate": 1.872431108562956e-05,
+      "loss": 0.7759,
+      "step": 3605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 51.15102685585607,
+      "learning_rate": 1.8723486755817405e-05,
+      "loss": 0.9419,
+      "step": 3606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 32.39172801574702,
+      "learning_rate": 1.872266217791439e-05,
+      "loss": 0.8123,
+      "step": 3607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 43.07695145196687,
+      "learning_rate": 1.8721837351943964e-05,
+      "loss": 0.8133,
+      "step": 3608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 37.15973860627168,
+      "learning_rate": 1.8721012277929588e-05,
+      "loss": 0.8554,
+      "step": 3609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 35.677117602849314,
+      "learning_rate": 1.8720186955894722e-05,
+      "loss": 0.8155,
+      "step": 3610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 30.20867280681643,
+      "learning_rate": 1.8719361385862843e-05,
+      "loss": 0.7996,
+      "step": 3611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 32.62588705851974,
+      "learning_rate": 1.8718535567857426e-05,
+      "loss": 0.8815,
+      "step": 3612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 35.48089280169585,
+      "learning_rate": 1.8717709501901956e-05,
+      "loss": 0.8312,
+      "step": 3613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 29.585295386910065,
+      "learning_rate": 1.8716883188019932e-05,
+      "loss": 0.7721,
+      "step": 3614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.901932448497035,
+      "learning_rate": 1.8716056626234848e-05,
+      "loss": 0.8645,
+      "step": 3615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 35.32351138507579,
+      "learning_rate": 1.871522981657021e-05,
+      "loss": 0.8564,
+      "step": 3616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 31.40551284000831,
+      "learning_rate": 1.8714402759049535e-05,
+      "loss": 0.8286,
+      "step": 3617
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 33.967153892463614,
+      "learning_rate": 1.8713575453696344e-05,
+      "loss": 0.8722,
+      "step": 3618
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.473184061436598,
+      "learning_rate": 1.8712747900534164e-05,
+      "loss": 0.7912,
+      "step": 3619
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.952289377165535,
+      "learning_rate": 1.871192009958653e-05,
+      "loss": 0.8912,
+      "step": 3620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 39.01575201378231,
+      "learning_rate": 1.8711092050876982e-05,
+      "loss": 0.8449,
+      "step": 3621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 37.28757481991113,
+      "learning_rate": 1.8710263754429076e-05,
+      "loss": 0.9068,
+      "step": 3622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 36.787168180632875,
+      "learning_rate": 1.870943521026636e-05,
+      "loss": 0.8595,
+      "step": 3623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.67861076519463,
+      "learning_rate": 1.8708606418412408e-05,
+      "loss": 0.8,
+      "step": 3624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.934667631985334,
+      "learning_rate": 1.8707777378890778e-05,
+      "loss": 0.7929,
+      "step": 3625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.40569822697669,
+      "learning_rate": 1.870694809172506e-05,
+      "loss": 0.7304,
+      "step": 3626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 34.528861616571746,
+      "learning_rate": 1.8706118556938826e-05,
+      "loss": 0.786,
+      "step": 3627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 38.15128057237973,
+      "learning_rate": 1.870528877455567e-05,
+      "loss": 0.8869,
+      "step": 3628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 41.231125507337374,
+      "learning_rate": 1.8704458744599202e-05,
+      "loss": 0.8895,
+      "step": 3629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.48863260570629,
+      "learning_rate": 1.8703628467093017e-05,
+      "loss": 0.8197,
+      "step": 3630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 31.73937412105792,
+      "learning_rate": 1.8702797942060725e-05,
+      "loss": 0.8037,
+      "step": 3631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 39.41973078245322,
+      "learning_rate": 1.8701967169525958e-05,
+      "loss": 0.907,
+      "step": 3632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 38.215983314840116,
+      "learning_rate": 1.8701136149512328e-05,
+      "loss": 0.7782,
+      "step": 3633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 40.29930686934394,
+      "learning_rate": 1.870030488204348e-05,
+      "loss": 0.9614,
+      "step": 3634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.06448170744942,
+      "learning_rate": 1.869947336714305e-05,
+      "loss": 0.8048,
+      "step": 3635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.85715476748551,
+      "learning_rate": 1.8698641604834687e-05,
+      "loss": 0.8328,
+      "step": 3636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 33.69407092185535,
+      "learning_rate": 1.8697809595142042e-05,
+      "loss": 0.8443,
+      "step": 3637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 34.01684135676873,
+      "learning_rate": 1.869697733808878e-05,
+      "loss": 0.8048,
+      "step": 3638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 42.90419177559416,
+      "learning_rate": 1.8696144833698574e-05,
+      "loss": 0.8785,
+      "step": 3639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.984629136285136,
+      "learning_rate": 1.8695312081995096e-05,
+      "loss": 0.8203,
+      "step": 3640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 46.78749631921956,
+      "learning_rate": 1.8694479083002027e-05,
+      "loss": 0.9381,
+      "step": 3641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 33.652491665170245,
+      "learning_rate": 1.8693645836743062e-05,
+      "loss": 0.8695,
+      "step": 3642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.735048770524628,
+      "learning_rate": 1.8692812343241892e-05,
+      "loss": 0.8004,
+      "step": 3643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 44.64779880497386,
+      "learning_rate": 1.8691978602522226e-05,
+      "loss": 0.8924,
+      "step": 3644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 37.65628448738007,
+      "learning_rate": 1.869114461460777e-05,
+      "loss": 0.859,
+      "step": 3645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.383588391845834,
+      "learning_rate": 1.8690310379522247e-05,
+      "loss": 0.8015,
+      "step": 3646
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 45.30126221837233,
+      "learning_rate": 1.8689475897289383e-05,
+      "loss": 0.8581,
+      "step": 3647
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.567122891833645,
+      "learning_rate": 1.8688641167932906e-05,
+      "loss": 0.7383,
+      "step": 3648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 40.95098802098885,
+      "learning_rate": 1.8687806191476555e-05,
+      "loss": 0.9702,
+      "step": 3649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 40.34912075989763,
+      "learning_rate": 1.8686970967944082e-05,
+      "loss": 0.9071,
+      "step": 3650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 40.806165767377664,
+      "learning_rate": 1.8686135497359233e-05,
+      "loss": 0.7808,
+      "step": 3651
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 39.09635433606196,
+      "learning_rate": 1.868529977974577e-05,
+      "loss": 0.7977,
+      "step": 3652
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 34.50240014890937,
+      "learning_rate": 1.868446381512747e-05,
+      "loss": 0.8771,
+      "step": 3653
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 34.15659098616396,
+      "learning_rate": 1.868362760352809e-05,
+      "loss": 0.7837,
+      "step": 3654
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.84200905893017,
+      "learning_rate": 1.8682791144971426e-05,
+      "loss": 0.8172,
+      "step": 3655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 36.848398010882356,
+      "learning_rate": 1.8681954439481262e-05,
+      "loss": 0.8943,
+      "step": 3656
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 33.129491177677835,
+      "learning_rate": 1.868111748708139e-05,
+      "loss": 0.7675,
+      "step": 3657
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.60109644591686,
+      "learning_rate": 1.8680280287795612e-05,
+      "loss": 0.8419,
+      "step": 3658
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 38.201355719600585,
+      "learning_rate": 1.8679442841647747e-05,
+      "loss": 0.9071,
+      "step": 3659
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 34.295361246057595,
+      "learning_rate": 1.8678605148661602e-05,
+      "loss": 0.847,
+      "step": 3660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.60449005592201,
+      "learning_rate": 1.8677767208861e-05,
+      "loss": 0.795,
+      "step": 3661
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 37.24057670424315,
+      "learning_rate": 1.867692902226978e-05,
+      "loss": 0.8317,
+      "step": 3662
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 38.72081202757483,
+      "learning_rate": 1.8676090588911774e-05,
+      "loss": 0.8206,
+      "step": 3663
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 29.82266326025614,
+      "learning_rate": 1.8675251908810824e-05,
+      "loss": 0.8367,
+      "step": 3664
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 30.49022942134237,
+      "learning_rate": 1.867441298199079e-05,
+      "loss": 0.8378,
+      "step": 3665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 50.302910922819486,
+      "learning_rate": 1.867357380847552e-05,
+      "loss": 0.7651,
+      "step": 3666
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 36.62171737424742,
+      "learning_rate": 1.8672734388288883e-05,
+      "loss": 0.8321,
+      "step": 3667
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 39.00830458902333,
+      "learning_rate": 1.867189472145476e-05,
+      "loss": 0.8105,
+      "step": 3668
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 36.14643339525662,
+      "learning_rate": 1.867105480799702e-05,
+      "loss": 0.7674,
+      "step": 3669
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 36.20516670275071,
+      "learning_rate": 1.867021464793955e-05,
+      "loss": 0.8511,
+      "step": 3670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.995063258774863,
+      "learning_rate": 1.8669374241306256e-05,
+      "loss": 0.8213,
+      "step": 3671
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 37.16352796059882,
+      "learning_rate": 1.8668533588121026e-05,
+      "loss": 0.8235,
+      "step": 3672
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 39.4371528187023,
+      "learning_rate": 1.866769268840777e-05,
+      "loss": 0.8698,
+      "step": 3673
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 41.67512121955254,
+      "learning_rate": 1.8666851542190406e-05,
+      "loss": 0.8233,
+      "step": 3674
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 39.11450033708174,
+      "learning_rate": 1.8666010149492853e-05,
+      "loss": 0.7933,
+      "step": 3675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 33.28286133926409,
+      "learning_rate": 1.8665168510339043e-05,
+      "loss": 0.8227,
+      "step": 3676
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 31.282866665796256,
+      "learning_rate": 1.866432662475291e-05,
+      "loss": 0.7975,
+      "step": 3677
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 40.24108404128938,
+      "learning_rate": 1.8663484492758393e-05,
+      "loss": 0.8852,
+      "step": 3678
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 40.69883921605807,
+      "learning_rate": 1.8662642114379444e-05,
+      "loss": 0.9712,
+      "step": 3679
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 35.7002145628532,
+      "learning_rate": 1.8661799489640025e-05,
+      "loss": 0.7485,
+      "step": 3680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 41.027926924191995,
+      "learning_rate": 1.8660956618564095e-05,
+      "loss": 0.8138,
+      "step": 3681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 42.768655412461484,
+      "learning_rate": 1.8660113501175622e-05,
+      "loss": 0.9246,
+      "step": 3682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 40.59913425696989,
+      "learning_rate": 1.8659270137498588e-05,
+      "loss": 0.9074,
+      "step": 3683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.11663033825596,
+      "learning_rate": 1.865842652755698e-05,
+      "loss": 0.8354,
+      "step": 3684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.272103222784494,
+      "learning_rate": 1.8657582671374787e-05,
+      "loss": 0.9301,
+      "step": 3685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.22453630823408,
+      "learning_rate": 1.8656738568976003e-05,
+      "loss": 0.8547,
+      "step": 3686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.92172175165434,
+      "learning_rate": 1.865589422038464e-05,
+      "loss": 0.8896,
+      "step": 3687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 55.37778350516409,
+      "learning_rate": 1.865504962562471e-05,
+      "loss": 0.9688,
+      "step": 3688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 40.75499561908368,
+      "learning_rate": 1.865420478472023e-05,
+      "loss": 0.8171,
+      "step": 3689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 42.010308828419625,
+      "learning_rate": 1.865335969769523e-05,
+      "loss": 0.8551,
+      "step": 3690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 42.80354910511817,
+      "learning_rate": 1.865251436457374e-05,
+      "loss": 0.8138,
+      "step": 3691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 38.33439189997277,
+      "learning_rate": 1.8651668785379806e-05,
+      "loss": 0.9124,
+      "step": 3692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 50.06267269452392,
+      "learning_rate": 1.865082296013747e-05,
+      "loss": 0.88,
+      "step": 3693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 44.387823578394475,
+      "learning_rate": 1.864997688887079e-05,
+      "loss": 0.9636,
+      "step": 3694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 45.388922528431216,
+      "learning_rate": 1.864913057160383e-05,
+      "loss": 0.8779,
+      "step": 3695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 48.99753157567883,
+      "learning_rate": 1.8648284008360654e-05,
+      "loss": 0.9369,
+      "step": 3696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 39.41151712476317,
+      "learning_rate": 1.864743719916534e-05,
+      "loss": 0.8935,
+      "step": 3697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 41.96299617861477,
+      "learning_rate": 1.8646590144041972e-05,
+      "loss": 0.8672,
+      "step": 3698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 26.034231855461172,
+      "learning_rate": 1.8645742843014635e-05,
+      "loss": 0.7259,
+      "step": 3699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 33.5055210488921,
+      "learning_rate": 1.8644895296107432e-05,
+      "loss": 0.8367,
+      "step": 3700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 38.07965724712678,
+      "learning_rate": 1.8644047503344465e-05,
+      "loss": 0.967,
+      "step": 3701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 44.92738600918675,
+      "learning_rate": 1.8643199464749843e-05,
+      "loss": 0.9514,
+      "step": 3702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 69.0576363624536,
+      "learning_rate": 1.8642351180347682e-05,
+      "loss": 0.8368,
+      "step": 3703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 31.459933450353216,
+      "learning_rate": 1.864150265016211e-05,
+      "loss": 0.7503,
+      "step": 3704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 39.87647795255116,
+      "learning_rate": 1.8640653874217257e-05,
+      "loss": 0.9083,
+      "step": 3705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 43.82768283559755,
+      "learning_rate": 1.8639804852537264e-05,
+      "loss": 0.8675,
+      "step": 3706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.4962296397766,
+      "learning_rate": 1.8638955585146277e-05,
+      "loss": 0.7914,
+      "step": 3707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.69182758848715,
+      "learning_rate": 1.8638106072068443e-05,
+      "loss": 0.8044,
+      "step": 3708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 42.560991873889,
+      "learning_rate": 1.863725631332793e-05,
+      "loss": 0.8297,
+      "step": 3709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.74222335396725,
+      "learning_rate": 1.8636406308948894e-05,
+      "loss": 0.8595,
+      "step": 3710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.298562132194604,
+      "learning_rate": 1.863555605895552e-05,
+      "loss": 0.8529,
+      "step": 3711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 47.07071786286489,
+      "learning_rate": 1.863470556337198e-05,
+      "loss": 0.8113,
+      "step": 3712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 33.101750783630145,
+      "learning_rate": 1.8633854822222467e-05,
+      "loss": 0.8128,
+      "step": 3713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.1266698301196,
+      "learning_rate": 1.8633003835531172e-05,
+      "loss": 0.8295,
+      "step": 3714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 45.90721642481248,
+      "learning_rate": 1.86321526033223e-05,
+      "loss": 0.8512,
+      "step": 3715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.888305077833245,
+      "learning_rate": 1.8631301125620056e-05,
+      "loss": 0.7238,
+      "step": 3716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 46.24394448305561,
+      "learning_rate": 1.8630449402448655e-05,
+      "loss": 0.9651,
+      "step": 3717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.75656559816042,
+      "learning_rate": 1.8629597433832326e-05,
+      "loss": 0.6916,
+      "step": 3718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 52.626154907311715,
+      "learning_rate": 1.8628745219795292e-05,
+      "loss": 0.9645,
+      "step": 3719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 40.66766456092318,
+      "learning_rate": 1.862789276036179e-05,
+      "loss": 0.9276,
+      "step": 3720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.72976795650737,
+      "learning_rate": 1.862704005555607e-05,
+      "loss": 0.8607,
+      "step": 3721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 43.65498252454108,
+      "learning_rate": 1.8626187105402373e-05,
+      "loss": 0.9038,
+      "step": 3722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 38.18725198143865,
+      "learning_rate": 1.8625333909924964e-05,
+      "loss": 0.8428,
+      "step": 3723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.357752420144344,
+      "learning_rate": 1.8624480469148103e-05,
+      "loss": 0.9059,
+      "step": 3724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 46.01208432472787,
+      "learning_rate": 1.862362678309606e-05,
+      "loss": 0.8621,
+      "step": 3725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 53.147455164040224,
+      "learning_rate": 1.862277285179312e-05,
+      "loss": 0.8537,
+      "step": 3726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 29.88031373989767,
+      "learning_rate": 1.8621918675263564e-05,
+      "loss": 0.8329,
+      "step": 3727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 39.70924554853113,
+      "learning_rate": 1.862106425353168e-05,
+      "loss": 1.0238,
+      "step": 3728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.068838393175255,
+      "learning_rate": 1.8620209586621778e-05,
+      "loss": 0.733,
+      "step": 3729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.20984838703015,
+      "learning_rate": 1.8619354674558155e-05,
+      "loss": 0.8602,
+      "step": 3730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 38.95058664252281,
+      "learning_rate": 1.8618499517365127e-05,
+      "loss": 0.9604,
+      "step": 3731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.80272368125809,
+      "learning_rate": 1.8617644115067013e-05,
+      "loss": 0.8423,
+      "step": 3732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 54.94677042508312,
+      "learning_rate": 1.8616788467688146e-05,
+      "loss": 0.893,
+      "step": 3733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 50.68240654232968,
+      "learning_rate": 1.8615932575252855e-05,
+      "loss": 0.8333,
+      "step": 3734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 49.641388711646904,
+      "learning_rate": 1.8615076437785474e-05,
+      "loss": 0.8023,
+      "step": 3735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 46.50654317083899,
+      "learning_rate": 1.8614220055310367e-05,
+      "loss": 0.8931,
+      "step": 3736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.882076578039,
+      "learning_rate": 1.8613363427851875e-05,
+      "loss": 0.8653,
+      "step": 3737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 27.95776192687122,
+      "learning_rate": 1.8612506555434372e-05,
+      "loss": 0.7792,
+      "step": 3738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 44.35672348713611,
+      "learning_rate": 1.861164943808222e-05,
+      "loss": 0.9949,
+      "step": 3739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 41.29781422753971,
+      "learning_rate": 1.861079207581979e-05,
+      "loss": 0.8266,
+      "step": 3740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 38.366624980664135,
+      "learning_rate": 1.8609934468671474e-05,
+      "loss": 0.7959,
+      "step": 3741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 47.23506330618922,
+      "learning_rate": 1.8609076616661658e-05,
+      "loss": 0.8323,
+      "step": 3742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.45749435432909,
+      "learning_rate": 1.860821851981474e-05,
+      "loss": 0.8067,
+      "step": 3743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.743174781436316,
+      "learning_rate": 1.8607360178155123e-05,
+      "loss": 0.7867,
+      "step": 3744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 33.46607756074027,
+      "learning_rate": 1.8606501591707217e-05,
+      "loss": 0.7838,
+      "step": 3745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.11344877444539,
+      "learning_rate": 1.860564276049544e-05,
+      "loss": 0.7818,
+      "step": 3746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.45929287486633,
+      "learning_rate": 1.860478368454422e-05,
+      "loss": 0.7809,
+      "step": 3747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.868552449912702,
+      "learning_rate": 1.860392436387798e-05,
+      "loss": 0.8637,
+      "step": 3748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 44.147755450352086,
+      "learning_rate": 1.8603064798521168e-05,
+      "loss": 0.9093,
+      "step": 3749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 35.43780710198279,
+      "learning_rate": 1.8602204988498224e-05,
+      "loss": 0.8711,
+      "step": 3750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.76516526855235,
+      "learning_rate": 1.8601344933833605e-05,
+      "loss": 0.791,
+      "step": 3751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.26828559925446,
+      "learning_rate": 1.8600484634551767e-05,
+      "loss": 0.8889,
+      "step": 3752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 45.74892581137222,
+      "learning_rate": 1.8599624090677174e-05,
+      "loss": 0.9201,
+      "step": 3753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.725762522781785,
+      "learning_rate": 1.8598763302234306e-05,
+      "loss": 0.8307,
+      "step": 3754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.929721140976156,
+      "learning_rate": 1.859790226924764e-05,
+      "loss": 0.8627,
+      "step": 3755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 35.762496341437135,
+      "learning_rate": 1.8597040991741663e-05,
+      "loss": 0.9115,
+      "step": 3756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 34.34607656371855,
+      "learning_rate": 1.8596179469740863e-05,
+      "loss": 0.825,
+      "step": 3757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.502660628627126,
+      "learning_rate": 1.8595317703269756e-05,
+      "loss": 0.8755,
+      "step": 3758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.64071953279383,
+      "learning_rate": 1.859445569235284e-05,
+      "loss": 0.8666,
+      "step": 3759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.69869328959497,
+      "learning_rate": 1.8593593437014627e-05,
+      "loss": 0.7927,
+      "step": 3760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.44296533450193,
+      "learning_rate": 1.8592730937279647e-05,
+      "loss": 0.8474,
+      "step": 3761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 35.369409941727994,
+      "learning_rate": 1.8591868193172423e-05,
+      "loss": 0.8279,
+      "step": 3762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.51210428093346,
+      "learning_rate": 1.8591005204717498e-05,
+      "loss": 0.8617,
+      "step": 3763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 41.86472175627781,
+      "learning_rate": 1.859014197193941e-05,
+      "loss": 0.8927,
+      "step": 3764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 35.214522486282846,
+      "learning_rate": 1.8589278494862707e-05,
+      "loss": 0.8403,
+      "step": 3765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 42.164911668364454,
+      "learning_rate": 1.858841477351195e-05,
+      "loss": 0.8794,
+      "step": 3766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 28.920549610667777,
+      "learning_rate": 1.85875508079117e-05,
+      "loss": 0.6778,
+      "step": 3767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 41.94885045159954,
+      "learning_rate": 1.858668659808653e-05,
+      "loss": 0.8825,
+      "step": 3768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 46.24982480196375,
+      "learning_rate": 1.8585822144061013e-05,
+      "loss": 0.8502,
+      "step": 3769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 41.35392555275426,
+      "learning_rate": 1.858495744585974e-05,
+      "loss": 0.8722,
+      "step": 3770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 66.03091832089832,
+      "learning_rate": 1.8584092503507296e-05,
+      "loss": 0.7666,
+      "step": 3771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 39.356094787010505,
+      "learning_rate": 1.8583227317028287e-05,
+      "loss": 0.9033,
+      "step": 3772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.81416327286668,
+      "learning_rate": 1.8582361886447315e-05,
+      "loss": 0.8492,
+      "step": 3773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.30032797586033,
+      "learning_rate": 1.8581496211788985e-05,
+      "loss": 0.8067,
+      "step": 3774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.313522813610334,
+      "learning_rate": 1.8580630293077925e-05,
+      "loss": 0.8577,
+      "step": 3775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.12287444381929,
+      "learning_rate": 1.857976413033876e-05,
+      "loss": 0.791,
+      "step": 3776
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.203096677572425,
+      "learning_rate": 1.8578897723596124e-05,
+      "loss": 0.8249,
+      "step": 3777
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 42.67898843057624,
+      "learning_rate": 1.8578031072874655e-05,
+      "loss": 0.8466,
+      "step": 3778
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 31.882305321635606,
+      "learning_rate": 1.8577164178198995e-05,
+      "loss": 0.7664,
+      "step": 3779
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.7510071846465,
+      "learning_rate": 1.857629703959381e-05,
+      "loss": 0.9192,
+      "step": 3780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 37.426001163207246,
+      "learning_rate": 1.8575429657083753e-05,
+      "loss": 0.9176,
+      "step": 3781
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 60.21654296219648,
+      "learning_rate": 1.8574562030693492e-05,
+      "loss": 0.8759,
+      "step": 3782
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.01697330419029,
+      "learning_rate": 1.85736941604477e-05,
+      "loss": 0.9414,
+      "step": 3783
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 39.93424159922574,
+      "learning_rate": 1.8572826046371063e-05,
+      "loss": 0.9246,
+      "step": 3784
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 41.445847701806564,
+      "learning_rate": 1.8571957688488268e-05,
+      "loss": 0.9046,
+      "step": 3785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 43.55398444596177,
+      "learning_rate": 1.8571089086824014e-05,
+      "loss": 0.7793,
+      "step": 3786
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.96351361023642,
+      "learning_rate": 1.8570220241402996e-05,
+      "loss": 0.795,
+      "step": 3787
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 48.52713764106678,
+      "learning_rate": 1.8569351152249932e-05,
+      "loss": 0.7943,
+      "step": 3788
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.709239581926624,
+      "learning_rate": 1.856848181938953e-05,
+      "loss": 0.8408,
+      "step": 3789
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 30.21918970476512,
+      "learning_rate": 1.856761224284652e-05,
+      "loss": 0.8278,
+      "step": 3790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.2204231026834,
+      "learning_rate": 1.856674242264563e-05,
+      "loss": 0.868,
+      "step": 3791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.046607865500626,
+      "learning_rate": 1.8565872358811594e-05,
+      "loss": 0.8461,
+      "step": 3792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.16510144228093,
+      "learning_rate": 1.856500205136916e-05,
+      "loss": 0.7753,
+      "step": 3793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 45.78134728118664,
+      "learning_rate": 1.8564131500343082e-05,
+      "loss": 0.8832,
+      "step": 3794
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.92003642188907,
+      "learning_rate": 1.8563260705758114e-05,
+      "loss": 0.8269,
+      "step": 3795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 33.47836802720298,
+      "learning_rate": 1.8562389667639016e-05,
+      "loss": 0.771,
+      "step": 3796
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.822276357452076,
+      "learning_rate": 1.8561518386010567e-05,
+      "loss": 0.8036,
+      "step": 3797
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 37.19955369838001,
+      "learning_rate": 1.856064686089754e-05,
+      "loss": 0.9037,
+      "step": 3798
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 46.730119287602676,
+      "learning_rate": 1.855977509232473e-05,
+      "loss": 0.9689,
+      "step": 3799
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 39.5725265561354,
+      "learning_rate": 1.8558903080316922e-05,
+      "loss": 0.8776,
+      "step": 3800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.79472888383258,
+      "learning_rate": 1.8558030824898915e-05,
+      "loss": 0.8973,
+      "step": 3801
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.961917583898774,
+      "learning_rate": 1.8557158326095515e-05,
+      "loss": 0.852,
+      "step": 3802
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.59277407999158,
+      "learning_rate": 1.8556285583931547e-05,
+      "loss": 0.9851,
+      "step": 3803
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 67.25716892306285,
+      "learning_rate": 1.8555412598431814e-05,
+      "loss": 0.95,
+      "step": 3804
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.97626459615256,
+      "learning_rate": 1.8554539369621156e-05,
+      "loss": 0.8015,
+      "step": 3805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 37.80747140282093,
+      "learning_rate": 1.85536658975244e-05,
+      "loss": 0.896,
+      "step": 3806
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 34.930586927655604,
+      "learning_rate": 1.8552792182166393e-05,
+      "loss": 0.799,
+      "step": 3807
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 41.4719835930872,
+      "learning_rate": 1.8551918223571975e-05,
+      "loss": 0.9199,
+      "step": 3808
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 31.729741564472757,
+      "learning_rate": 1.855104402176601e-05,
+      "loss": 0.8882,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 42.366991180859834,
+      "learning_rate": 1.8550169576773353e-05,
+      "loss": 0.9282,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.3046488369247,
+      "learning_rate": 1.8549294888618872e-05,
+      "loss": 0.9739,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.42095061177867,
+      "learning_rate": 1.8548419957327454e-05,
+      "loss": 0.7905,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 42.234043172251795,
+      "learning_rate": 1.8547544782923963e-05,
+      "loss": 0.9064,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 33.994639720940505,
+      "learning_rate": 1.8546669365433305e-05,
+      "loss": 0.8713,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 38.625935095668176,
+      "learning_rate": 1.8545793704880367e-05,
+      "loss": 0.8873,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 36.008752474162215,
+      "learning_rate": 1.854491780129006e-05,
+      "loss": 0.8459,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 41.35849950402045,
+      "learning_rate": 1.8544041654687282e-05,
+      "loss": 0.8932,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 40.80949371505238,
+      "learning_rate": 1.8543165265096962e-05,
+      "loss": 0.8783,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.784155252180824,
+      "learning_rate": 1.854228863254402e-05,
+      "loss": 0.8727,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 36.638299148366166,
+      "learning_rate": 1.854141175705338e-05,
+      "loss": 0.9574,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.54110077385603,
+      "learning_rate": 1.8540534638649994e-05,
+      "loss": 0.8129,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 44.42757992871256,
+      "learning_rate": 1.8539657277358798e-05,
+      "loss": 0.7708,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 44.09896273928518,
+      "learning_rate": 1.8538779673204743e-05,
+      "loss": 0.7676,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 50.45584008881877,
+      "learning_rate": 1.8537901826212786e-05,
+      "loss": 0.9708,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.79744242210299,
+      "learning_rate": 1.8537023736407898e-05,
+      "loss": 0.7351,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.571456552387062,
+      "learning_rate": 1.853614540381505e-05,
+      "loss": 0.7495,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.40006947593549,
+      "learning_rate": 1.853526682845922e-05,
+      "loss": 0.8872,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.77710973378541,
+      "learning_rate": 1.8534388010365397e-05,
+      "loss": 0.7702,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 50.665484058021974,
+      "learning_rate": 1.8533508949558564e-05,
+      "loss": 1.0239,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 56.558257669930605,
+      "learning_rate": 1.8532629646063737e-05,
+      "loss": 0.9649,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 45.23960995334817,
+      "learning_rate": 1.853175009990591e-05,
+      "loss": 0.9094,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.31046207062651,
+      "learning_rate": 1.85308703111101e-05,
+      "loss": 0.812,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.12410553608808,
+      "learning_rate": 1.852999027970133e-05,
+      "loss": 0.8809,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.259116163004933,
+      "learning_rate": 1.8529110005704627e-05,
+      "loss": 0.8102,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.38925783801321,
+      "learning_rate": 1.8528229489145023e-05,
+      "loss": 0.8396,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 40.92476508970891,
+      "learning_rate": 1.8527348730047563e-05,
+      "loss": 0.9174,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 30.632414632566366,
+      "learning_rate": 1.8526467728437292e-05,
+      "loss": 0.8308,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.880351678946795,
+      "learning_rate": 1.852558648433927e-05,
+      "loss": 0.7906,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 39.047157014864105,
+      "learning_rate": 1.852470499777855e-05,
+      "loss": 0.8602,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 48.79842791252925,
+      "learning_rate": 1.8523823268780214e-05,
+      "loss": 0.8598,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.342397822292398,
+      "learning_rate": 1.8522941297369324e-05,
+      "loss": 0.7983,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.705114555779865,
+      "learning_rate": 1.8522059083570972e-05,
+      "loss": 0.9586,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 45.38916305840242,
+      "learning_rate": 1.8521176627410246e-05,
+      "loss": 0.8446,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 44.375014044531085,
+      "learning_rate": 1.852029392891224e-05,
+      "loss": 0.9226,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 51.62718561362804,
+      "learning_rate": 1.8519410988102058e-05,
+      "loss": 0.9104,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.87320238013035,
+      "learning_rate": 1.8518527805004812e-05,
+      "loss": 0.8822,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 39.14063845575219,
+      "learning_rate": 1.851764437964562e-05,
+      "loss": 0.9039,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 35.14225778140603,
+      "learning_rate": 1.8516760712049603e-05,
+      "loss": 0.8888,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 39.45251215694833,
+      "learning_rate": 1.8515876802241894e-05,
+      "loss": 0.8468,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.69424516891094,
+      "learning_rate": 1.8514992650247627e-05,
+      "loss": 0.9611,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.01648407466969,
+      "learning_rate": 1.8514108256091953e-05,
+      "loss": 0.8782,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.947279916826126,
+      "learning_rate": 1.8513223619800022e-05,
+      "loss": 0.9245,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 29.494760541019836,
+      "learning_rate": 1.851233874139699e-05,
+      "loss": 0.7802,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 36.767170424107036,
+      "learning_rate": 1.8511453620908027e-05,
+      "loss": 0.901,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.21985243438542,
+      "learning_rate": 1.8510568258358298e-05,
+      "loss": 0.8142,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 44.47355796894949,
+      "learning_rate": 1.850968265377299e-05,
+      "loss": 0.8238,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 45.97447122308144,
+      "learning_rate": 1.850879680717728e-05,
+      "loss": 0.9571,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 36.73326911056695,
+      "learning_rate": 1.850791071859637e-05,
+      "loss": 0.9071,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 30.96306431271091,
+      "learning_rate": 1.8507024388055456e-05,
+      "loss": 0.8293,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 28.3827629141423,
+      "learning_rate": 1.8506137815579744e-05,
+      "loss": 0.8775,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 57.33308066083248,
+      "learning_rate": 1.8505251001194445e-05,
+      "loss": 0.9665,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 65.02580742926848,
+      "learning_rate": 1.850436394492479e-05,
+      "loss": 0.9286,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.35051901132318,
+      "learning_rate": 1.8503476646795996e-05,
+      "loss": 0.8119,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 35.75379754284062,
+      "learning_rate": 1.85025891068333e-05,
+      "loss": 0.8467,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 29.66037593716485,
+      "learning_rate": 1.8501701325061943e-05,
+      "loss": 0.817,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 35.24729485311183,
+      "learning_rate": 1.8500813301507174e-05,
+      "loss": 0.7748,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 43.28327584603749,
+      "learning_rate": 1.8499925036194247e-05,
+      "loss": 0.8821,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.035660997803102,
+      "learning_rate": 1.849903652914843e-05,
+      "loss": 0.8414,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 51.135408978731355,
+      "learning_rate": 1.849814778039498e-05,
+      "loss": 0.9086,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 42.76151761842015,
+      "learning_rate": 1.849725878995918e-05,
+      "loss": 0.7525,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 34.96436121367604,
+      "learning_rate": 1.849636955786631e-05,
+      "loss": 0.7784,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 37.38157041712444,
+      "learning_rate": 1.8495480084141662e-05,
+      "loss": 0.9633,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 39.64447901315731,
+      "learning_rate": 1.849459036881053e-05,
+      "loss": 0.8589,
+      "step": 3873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.03262763189548,
+      "learning_rate": 1.8493700411898213e-05,
+      "loss": 0.835,
+      "step": 3874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 31.224322614965452,
+      "learning_rate": 1.8492810213430027e-05,
+      "loss": 0.7995,
+      "step": 3875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.281008220987424,
+      "learning_rate": 1.849191977343129e-05,
+      "loss": 0.815,
+      "step": 3876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 59.10060016985713,
+      "learning_rate": 1.849102909192732e-05,
+      "loss": 0.8191,
+      "step": 3877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 51.2642688798574,
+      "learning_rate": 1.849013816894345e-05,
+      "loss": 0.962,
+      "step": 3878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 38.39144421806887,
+      "learning_rate": 1.8489247004505017e-05,
+      "loss": 0.8644,
+      "step": 3879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 35.40099305454539,
+      "learning_rate": 1.8488355598637362e-05,
+      "loss": 0.8623,
+      "step": 3880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 84.5129281712341,
+      "learning_rate": 1.8487463951365842e-05,
+      "loss": 0.9036,
+      "step": 3881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 64.58832329043908,
+      "learning_rate": 1.8486572062715814e-05,
+      "loss": 0.9839,
+      "step": 3882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 42.54660999609224,
+      "learning_rate": 1.848567993271264e-05,
+      "loss": 0.931,
+      "step": 3883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 35.919841727065254,
+      "learning_rate": 1.848478756138169e-05,
+      "loss": 0.909,
+      "step": 3884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.317431162670914,
+      "learning_rate": 1.8483894948748348e-05,
+      "loss": 0.8593,
+      "step": 3885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 33.86734601504003,
+      "learning_rate": 1.8483002094837998e-05,
+      "loss": 0.8322,
+      "step": 3886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 34.47244084536765,
+      "learning_rate": 1.848210899967603e-05,
+      "loss": 0.9243,
+      "step": 3887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 38.731756293382716,
+      "learning_rate": 1.848121566328784e-05,
+      "loss": 0.8268,
+      "step": 3888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 45.30027935581933,
+      "learning_rate": 1.848032208569884e-05,
+      "loss": 0.8678,
+      "step": 3889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.74066406829359,
+      "learning_rate": 1.8479428266934442e-05,
+      "loss": 0.8371,
+      "step": 3890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.302863966042246,
+      "learning_rate": 1.8478534207020066e-05,
+      "loss": 0.804,
+      "step": 3891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 45.857049028460196,
+      "learning_rate": 1.8477639905981135e-05,
+      "loss": 0.9279,
+      "step": 3892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 40.83411061980877,
+      "learning_rate": 1.8476745363843082e-05,
+      "loss": 0.9017,
+      "step": 3893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.139738372080366,
+      "learning_rate": 1.847585058063135e-05,
+      "loss": 0.789,
+      "step": 3894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 45.15319063059243,
+      "learning_rate": 1.847495555637139e-05,
+      "loss": 0.9077,
+      "step": 3895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 34.25122261649637,
+      "learning_rate": 1.8474060291088646e-05,
+      "loss": 0.8423,
+      "step": 3896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 35.04964288129644,
+      "learning_rate": 1.8473164784808592e-05,
+      "loss": 0.8065,
+      "step": 3897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 47.09753239256187,
+      "learning_rate": 1.8472269037556683e-05,
+      "loss": 0.8257,
+      "step": 3898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.767369804045465,
+      "learning_rate": 1.84713730493584e-05,
+      "loss": 0.8064,
+      "step": 3899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.08410981482975,
+      "learning_rate": 1.847047682023922e-05,
+      "loss": 0.8252,
+      "step": 3900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 40.58308957253214,
+      "learning_rate": 1.8469580350224636e-05,
+      "loss": 0.9086,
+      "step": 3901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 34.18337335385814,
+      "learning_rate": 1.8468683639340145e-05,
+      "loss": 0.8453,
+      "step": 3902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 31.81150643571471,
+      "learning_rate": 1.846778668761124e-05,
+      "loss": 0.9309,
+      "step": 3903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.62748729408393,
+      "learning_rate": 1.846688949506344e-05,
+      "loss": 0.847,
+      "step": 3904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.720903909695416,
+      "learning_rate": 1.8465992061722255e-05,
+      "loss": 0.9065,
+      "step": 3905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 37.30448380819104,
+      "learning_rate": 1.8465094387613204e-05,
+      "loss": 0.89,
+      "step": 3906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 35.27624860780106,
+      "learning_rate": 1.8464196472761825e-05,
+      "loss": 0.8567,
+      "step": 3907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 29.107209195187636,
+      "learning_rate": 1.8463298317193647e-05,
+      "loss": 0.7,
+      "step": 3908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 37.633206746144985,
+      "learning_rate": 1.846239992093422e-05,
+      "loss": 0.8558,
+      "step": 3909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.03979873802727,
+      "learning_rate": 1.8461501284009085e-05,
+      "loss": 0.9851,
+      "step": 3910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.51423594061296,
+      "learning_rate": 1.8460602406443802e-05,
+      "loss": 0.89,
+      "step": 3911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.6684035578228,
+      "learning_rate": 1.845970328826394e-05,
+      "loss": 0.7862,
+      "step": 3912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 52.16859407896344,
+      "learning_rate": 1.8458803929495062e-05,
+      "loss": 0.866,
+      "step": 3913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.01632456948365,
+      "learning_rate": 1.8457904330162752e-05,
+      "loss": 0.8501,
+      "step": 3914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 38.40603163356796,
+      "learning_rate": 1.8457004490292588e-05,
+      "loss": 0.9043,
+      "step": 3915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 33.02539213382339,
+      "learning_rate": 1.8456104409910165e-05,
+      "loss": 0.8828,
+      "step": 3916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 41.9775929795749,
+      "learning_rate": 1.845520408904108e-05,
+      "loss": 0.8626,
+      "step": 3917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.69639023767191,
+      "learning_rate": 1.8454303527710934e-05,
+      "loss": 0.7546,
+      "step": 3918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 34.138682630091914,
+      "learning_rate": 1.8453402725945343e-05,
+      "loss": 0.8634,
+      "step": 3919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.29587360105856,
+      "learning_rate": 1.8452501683769924e-05,
+      "loss": 0.8379,
+      "step": 3920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 33.37522721302542,
+      "learning_rate": 1.8451600401210303e-05,
+      "loss": 0.7773,
+      "step": 3921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 46.07071089835106,
+      "learning_rate": 1.8450698878292106e-05,
+      "loss": 0.843,
+      "step": 3922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 33.109525488475,
+      "learning_rate": 1.844979711504098e-05,
+      "loss": 0.8008,
+      "step": 3923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.949682067231436,
+      "learning_rate": 1.8448895111482565e-05,
+      "loss": 0.7987,
+      "step": 3924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.64891936211383,
+      "learning_rate": 1.8447992867642518e-05,
+      "loss": 0.7941,
+      "step": 3925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.95384093981482,
+      "learning_rate": 1.8447090383546494e-05,
+      "loss": 0.7512,
+      "step": 3926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 43.654965020658246,
+      "learning_rate": 1.8446187659220163e-05,
+      "loss": 0.955,
+      "step": 3927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 42.60864211299532,
+      "learning_rate": 1.8445284694689193e-05,
+      "loss": 0.8686,
+      "step": 3928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 28.022009969254167,
+      "learning_rate": 1.8444381489979267e-05,
+      "loss": 0.7299,
+      "step": 3929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 31.204620527971382,
+      "learning_rate": 1.8443478045116072e-05,
+      "loss": 0.7717,
+      "step": 3930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 40.418897937193066,
+      "learning_rate": 1.84425743601253e-05,
+      "loss": 0.9207,
+      "step": 3931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.85958981953926,
+      "learning_rate": 1.8441670435032654e-05,
+      "loss": 0.8207,
+      "step": 3932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 50.554560713475645,
+      "learning_rate": 1.8440766269863838e-05,
+      "loss": 0.9,
+      "step": 3933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.679136019437045,
+      "learning_rate": 1.8439861864644567e-05,
+      "loss": 0.8033,
+      "step": 3934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 41.8460296539778,
+      "learning_rate": 1.843895721940056e-05,
+      "loss": 0.8806,
+      "step": 3935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.321974775473784,
+      "learning_rate": 1.843805233415755e-05,
+      "loss": 0.8079,
+      "step": 3936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 32.04250585269808,
+      "learning_rate": 1.8437147208941265e-05,
+      "loss": 0.8105,
+      "step": 3937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.02720745766553,
+      "learning_rate": 1.843624184377745e-05,
+      "loss": 0.8428,
+      "step": 3938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.480018120325134,
+      "learning_rate": 1.8435336238691853e-05,
+      "loss": 0.8098,
+      "step": 3939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.585298658614626,
+      "learning_rate": 1.8434430393710224e-05,
+      "loss": 0.8374,
+      "step": 3940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 28.967852015810376,
+      "learning_rate": 1.8433524308858334e-05,
+      "loss": 0.6953,
+      "step": 3941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.937595134496426,
+      "learning_rate": 1.843261798416194e-05,
+      "loss": 0.7834,
+      "step": 3942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 38.048253332468335,
+      "learning_rate": 1.8431711419646826e-05,
+      "loss": 0.8733,
+      "step": 3943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.15568692387909,
+      "learning_rate": 1.8430804615338774e-05,
+      "loss": 0.9014,
+      "step": 3944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.46540445629664,
+      "learning_rate": 1.842989757126357e-05,
+      "loss": 1.002,
+      "step": 3945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 38.947901192262684,
+      "learning_rate": 1.8428990287447006e-05,
+      "loss": 0.8633,
+      "step": 3946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.92738727647155,
+      "learning_rate": 1.8428082763914893e-05,
+      "loss": 0.8181,
+      "step": 3947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 33.31034459052933,
+      "learning_rate": 1.842717500069304e-05,
+      "loss": 0.9007,
+      "step": 3948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 69.32804971363724,
+      "learning_rate": 1.8426266997807252e-05,
+      "loss": 0.8032,
+      "step": 3949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 37.050064332548395,
+      "learning_rate": 1.8425358755283365e-05,
+      "loss": 0.7354,
+      "step": 3950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 33.537016800205095,
+      "learning_rate": 1.84244502731472e-05,
+      "loss": 0.7975,
+      "step": 3951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 49.14493109874919,
+      "learning_rate": 1.84235415514246e-05,
+      "loss": 0.8159,
+      "step": 3952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 30.28659597394548,
+      "learning_rate": 1.8422632590141407e-05,
+      "loss": 0.7378,
+      "step": 3953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 44.74889001854937,
+      "learning_rate": 1.8421723389323468e-05,
+      "loss": 0.8753,
+      "step": 3954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 43.61208393590551,
+      "learning_rate": 1.8420813948996643e-05,
+      "loss": 0.8276,
+      "step": 3955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 31.380654995642633,
+      "learning_rate": 1.8419904269186797e-05,
+      "loss": 0.7997,
+      "step": 3956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.40373473947023,
+      "learning_rate": 1.8418994349919798e-05,
+      "loss": 0.8512,
+      "step": 3957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 49.255723431883396,
+      "learning_rate": 1.8418084191221522e-05,
+      "loss": 0.8166,
+      "step": 3958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 42.28001794479047,
+      "learning_rate": 1.8417173793117857e-05,
+      "loss": 0.8001,
+      "step": 3959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 51.93024877798989,
+      "learning_rate": 1.8416263155634695e-05,
+      "loss": 0.9826,
+      "step": 3960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 40.489182302014875,
+      "learning_rate": 1.8415352278797932e-05,
+      "loss": 0.9021,
+      "step": 3961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 33.882253623330655,
+      "learning_rate": 1.8414441162633473e-05,
+      "loss": 0.8898,
+      "step": 3962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 44.33756039510205,
+      "learning_rate": 1.8413529807167228e-05,
+      "loss": 0.8197,
+      "step": 3963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 42.47776219938426,
+      "learning_rate": 1.8412618212425116e-05,
+      "loss": 0.8104,
+      "step": 3964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.29105606762553,
+      "learning_rate": 1.841170637843307e-05,
+      "loss": 0.7312,
+      "step": 3965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.69138924270864,
+      "learning_rate": 1.8410794305217006e-05,
+      "loss": 0.7913,
+      "step": 3966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.818433654778744,
+      "learning_rate": 1.8409881992802882e-05,
+      "loss": 0.9093,
+      "step": 3967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 41.09730265530251,
+      "learning_rate": 1.8408969441216628e-05,
+      "loss": 0.8475,
+      "step": 3968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 34.90453757556937,
+      "learning_rate": 1.8408056650484204e-05,
+      "loss": 0.8728,
+      "step": 3969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 43.83929264633483,
+      "learning_rate": 1.8407143620631564e-05,
+      "loss": 0.8935,
+      "step": 3970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 31.80935160602969,
+      "learning_rate": 1.840623035168468e-05,
+      "loss": 0.9255,
+      "step": 3971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 45.644689387454385,
+      "learning_rate": 1.8405316843669524e-05,
+      "loss": 0.9088,
+      "step": 3972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 35.15071321756858,
+      "learning_rate": 1.840440309661207e-05,
+      "loss": 0.8779,
+      "step": 3973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 37.664612141562706,
+      "learning_rate": 1.8403489110538312e-05,
+      "loss": 0.8874,
+      "step": 3974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.852383488796185,
+      "learning_rate": 1.8402574885474238e-05,
+      "loss": 0.8488,
+      "step": 3975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 39.48752671084175,
+      "learning_rate": 1.840166042144585e-05,
+      "loss": 0.8867,
+      "step": 3976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.38062694784759,
+      "learning_rate": 1.8400745718479155e-05,
+      "loss": 0.7921,
+      "step": 3977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 34.00150129916549,
+      "learning_rate": 1.8399830776600164e-05,
+      "loss": 0.8428,
+      "step": 3978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 33.48444689708686,
+      "learning_rate": 1.83989155958349e-05,
+      "loss": 0.9227,
+      "step": 3979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 35.08410510636115,
+      "learning_rate": 1.839800017620939e-05,
+      "loss": 0.8762,
+      "step": 3980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 41.58664674815204,
+      "learning_rate": 1.8397084517749668e-05,
+      "loss": 0.8782,
+      "step": 3981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 35.227239958001164,
+      "learning_rate": 1.8396168620481772e-05,
+      "loss": 0.823,
+      "step": 3982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 42.373604619280286,
+      "learning_rate": 1.8395252484431758e-05,
+      "loss": 0.8438,
+      "step": 3983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 38.106502560777614,
+      "learning_rate": 1.839433610962567e-05,
+      "loss": 0.7681,
+      "step": 3984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.56247589893036,
+      "learning_rate": 1.8393419496089577e-05,
+      "loss": 0.8838,
+      "step": 3985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 41.148429903920906,
+      "learning_rate": 1.839250264384954e-05,
+      "loss": 0.8702,
+      "step": 3986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 37.37632633011982,
+      "learning_rate": 1.839158555293164e-05,
+      "loss": 0.8401,
+      "step": 3987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 36.07922334845357,
+      "learning_rate": 1.839066822336195e-05,
+      "loss": 0.8358,
+      "step": 3988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 42.771715716541095,
+      "learning_rate": 1.8389750655166572e-05,
+      "loss": 0.9097,
+      "step": 3989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 35.949323215268585,
+      "learning_rate": 1.8388832848371588e-05,
+      "loss": 0.7038,
+      "step": 3990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.66010793502806,
+      "learning_rate": 1.8387914803003106e-05,
+      "loss": 0.7992,
+      "step": 3991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 40.71357820680106,
+      "learning_rate": 1.8386996519087236e-05,
+      "loss": 0.865,
+      "step": 3992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 45.188904675802014,
+      "learning_rate": 1.8386077996650086e-05,
+      "loss": 0.9796,
+      "step": 3993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 41.79415118602653,
+      "learning_rate": 1.838515923571779e-05,
+      "loss": 0.8042,
+      "step": 3994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 37.163356258665715,
+      "learning_rate": 1.838424023631647e-05,
+      "loss": 0.8576,
+      "step": 3995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 38.32512347306619,
+      "learning_rate": 1.838332099847226e-05,
+      "loss": 0.8604,
+      "step": 3996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 56.95730735796127,
+      "learning_rate": 1.8382401522211304e-05,
+      "loss": 0.8386,
+      "step": 3997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 30.300917797855483,
+      "learning_rate": 1.838148180755975e-05,
+      "loss": 0.7576,
+      "step": 3998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 44.58480218030409,
+      "learning_rate": 1.8380561854543763e-05,
+      "loss": 0.781,
+      "step": 3999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 33.52320916856622,
+      "learning_rate": 1.8379641663189493e-05,
+      "loss": 0.7646,
+      "step": 4000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 43.135609977391574,
+      "learning_rate": 1.837872123352312e-05,
+      "loss": 0.9031,
+      "step": 4001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.821867786753515,
+      "learning_rate": 1.837780056557082e-05,
+      "loss": 0.9506,
+      "step": 4002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 43.85256243268588,
+      "learning_rate": 1.8376879659358767e-05,
+      "loss": 0.8326,
+      "step": 4003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 50.65309945627797,
+      "learning_rate": 1.8375958514913158e-05,
+      "loss": 0.8626,
+      "step": 4004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 40.68364433618735,
+      "learning_rate": 1.8375037132260187e-05,
+      "loss": 0.8955,
+      "step": 4005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 37.07553852896583,
+      "learning_rate": 1.837411551142606e-05,
+      "loss": 0.7803,
+      "step": 4006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 44.15668368548394,
+      "learning_rate": 1.837319365243699e-05,
+      "loss": 0.7946,
+      "step": 4007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 32.09342569709219,
+      "learning_rate": 1.8372271555319185e-05,
+      "loss": 0.7693,
+      "step": 4008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 39.08888886780606,
+      "learning_rate": 1.837134922009888e-05,
+      "loss": 0.8961,
+      "step": 4009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 37.264624355485026,
+      "learning_rate": 1.8370426646802298e-05,
+      "loss": 0.947,
+      "step": 4010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.66602052133725,
+      "learning_rate": 1.8369503835455676e-05,
+      "loss": 0.8079,
+      "step": 4011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 42.61742870287029,
+      "learning_rate": 1.8368580786085265e-05,
+      "loss": 0.8372,
+      "step": 4012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 42.21826343270113,
+      "learning_rate": 1.836765749871731e-05,
+      "loss": 0.9563,
+      "step": 4013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.70197404137336,
+      "learning_rate": 1.8366733973378074e-05,
+      "loss": 0.8721,
+      "step": 4014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.48190020513931,
+      "learning_rate": 1.8365810210093814e-05,
+      "loss": 0.8128,
+      "step": 4015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.85903905906476,
+      "learning_rate": 1.836488620889081e-05,
+      "loss": 0.862,
+      "step": 4016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 37.0064175395011,
+      "learning_rate": 1.836396196979533e-05,
+      "loss": 0.7945,
+      "step": 4017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.44780283052182,
+      "learning_rate": 1.8363037492833667e-05,
+      "loss": 0.8569,
+      "step": 4018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.80925253091538,
+      "learning_rate": 1.836211277803211e-05,
+      "loss": 0.7711,
+      "step": 4019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 29.5037748540182,
+      "learning_rate": 1.836118782541696e-05,
+      "loss": 0.8905,
+      "step": 4020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 29.156728871410618,
+      "learning_rate": 1.8360262635014513e-05,
+      "loss": 0.7759,
+      "step": 4021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 47.535885275721256,
+      "learning_rate": 1.8359337206851094e-05,
+      "loss": 0.8794,
+      "step": 4022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 57.221685679038536,
+      "learning_rate": 1.8358411540953008e-05,
+      "loss": 0.8222,
+      "step": 4023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 50.03474638502479,
+      "learning_rate": 1.835748563734659e-05,
+      "loss": 0.8736,
+      "step": 4024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 41.15543307027282,
+      "learning_rate": 1.8356559496058172e-05,
+      "loss": 0.8288,
+      "step": 4025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 33.49162855275866,
+      "learning_rate": 1.835563311711409e-05,
+      "loss": 0.842,
+      "step": 4026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.73177642912373,
+      "learning_rate": 1.835470650054069e-05,
+      "loss": 0.8584,
+      "step": 4027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 42.78592357088483,
+      "learning_rate": 1.8353779646364323e-05,
+      "loss": 0.8799,
+      "step": 4028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 32.91353380207359,
+      "learning_rate": 1.835285255461135e-05,
+      "loss": 0.7892,
+      "step": 4029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 43.0110887435076,
+      "learning_rate": 1.8351925225308136e-05,
+      "loss": 0.8115,
+      "step": 4030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.22489168065693,
+      "learning_rate": 1.8350997658481054e-05,
+      "loss": 0.8342,
+      "step": 4031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.81012831152815,
+      "learning_rate": 1.835006985415649e-05,
+      "loss": 0.7909,
+      "step": 4032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.112010070897334,
+      "learning_rate": 1.8349141812360815e-05,
+      "loss": 0.8502,
+      "step": 4033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 42.31276152852542,
+      "learning_rate": 1.8348213533120437e-05,
+      "loss": 0.9221,
+      "step": 4034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 33.302816690566154,
+      "learning_rate": 1.8347285016461745e-05,
+      "loss": 0.8251,
+      "step": 4035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.03269320893796,
+      "learning_rate": 1.8346356262411157e-05,
+      "loss": 0.752,
+      "step": 4036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 28.841657606051996,
+      "learning_rate": 1.8345427270995072e-05,
+      "loss": 0.7861,
+      "step": 4037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 32.76970017121682,
+      "learning_rate": 1.8344498042239922e-05,
+      "loss": 0.7424,
+      "step": 4038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 37.62106175897106,
+      "learning_rate": 1.8343568576172127e-05,
+      "loss": 0.7575,
+      "step": 4039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.68181973762707,
+      "learning_rate": 1.834263887281812e-05,
+      "loss": 0.7857,
+      "step": 4040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 32.70519355395893,
+      "learning_rate": 1.834170893220435e-05,
+      "loss": 0.7532,
+      "step": 4041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.699759637962302,
+      "learning_rate": 1.8340778754357252e-05,
+      "loss": 0.8426,
+      "step": 4042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.67350540350453,
+      "learning_rate": 1.833984833930329e-05,
+      "loss": 0.8137,
+      "step": 4043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 47.01220514721797,
+      "learning_rate": 1.8338917687068916e-05,
+      "loss": 0.873,
+      "step": 4044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.842124378945115,
+      "learning_rate": 1.8337986797680604e-05,
+      "loss": 0.8843,
+      "step": 4045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 41.501649703132436,
+      "learning_rate": 1.8337055671164826e-05,
+      "loss": 0.8582,
+      "step": 4046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.39074499979376,
+      "learning_rate": 1.833612430754806e-05,
+      "loss": 0.8567,
+      "step": 4047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 35.04304986930952,
+      "learning_rate": 1.8335192706856792e-05,
+      "loss": 0.8262,
+      "step": 4048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.97734443092778,
+      "learning_rate": 1.8334260869117524e-05,
+      "loss": 0.92,
+      "step": 4049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 33.755637079588574,
+      "learning_rate": 1.833332879435675e-05,
+      "loss": 0.8218,
+      "step": 4050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 36.30430873068264,
+      "learning_rate": 1.833239648260098e-05,
+      "loss": 0.8106,
+      "step": 4051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 37.94054427022461,
+      "learning_rate": 1.833146393387673e-05,
+      "loss": 0.8228,
+      "step": 4052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.26662986435987,
+      "learning_rate": 1.8330531148210516e-05,
+      "loss": 0.9697,
+      "step": 4053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 29.08388961890298,
+      "learning_rate": 1.8329598125628873e-05,
+      "loss": 0.7822,
+      "step": 4054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 32.23101300760293,
+      "learning_rate": 1.8328664866158332e-05,
+      "loss": 0.8261,
+      "step": 4055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 56.11378238612207,
+      "learning_rate": 1.8327731369825432e-05,
+      "loss": 0.9005,
+      "step": 4056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 45.25889301376595,
+      "learning_rate": 1.8326797636656723e-05,
+      "loss": 0.8681,
+      "step": 4057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 33.29864045452212,
+      "learning_rate": 1.8325863666678764e-05,
+      "loss": 0.8755,
+      "step": 4058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 44.462530573763885,
+      "learning_rate": 1.832492945991811e-05,
+      "loss": 0.7037,
+      "step": 4059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 30.247910127880566,
+      "learning_rate": 1.8323995016401332e-05,
+      "loss": 0.8662,
+      "step": 4060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 43.71715421129662,
+      "learning_rate": 1.8323060336155008e-05,
+      "loss": 0.867,
+      "step": 4061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.80680225462874,
+      "learning_rate": 1.8322125419205713e-05,
+      "loss": 0.8923,
+      "step": 4062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.118217640843664,
+      "learning_rate": 1.832119026558004e-05,
+      "loss": 0.7454,
+      "step": 4063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.115304082933314,
+      "learning_rate": 1.8320254875304585e-05,
+      "loss": 0.8042,
+      "step": 4064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 39.053335793307774,
+      "learning_rate": 1.831931924840595e-05,
+      "loss": 0.838,
+      "step": 4065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 44.95306822899504,
+      "learning_rate": 1.831838338491074e-05,
+      "loss": 0.8983,
+      "step": 4066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 41.461029777186965,
+      "learning_rate": 1.8317447284845573e-05,
+      "loss": 0.9216,
+      "step": 4067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 45.552649255960986,
+      "learning_rate": 1.831651094823707e-05,
+      "loss": 0.9655,
+      "step": 4068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.42295983224519,
+      "learning_rate": 1.831557437511186e-05,
+      "loss": 0.8743,
+      "step": 4069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 73.72985442495519,
+      "learning_rate": 1.8314637565496584e-05,
+      "loss": 0.9784,
+      "step": 4070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 39.0686673476096,
+      "learning_rate": 1.8313700519417876e-05,
+      "loss": 0.9001,
+      "step": 4071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 37.74620181674877,
+      "learning_rate": 1.8312763236902387e-05,
+      "loss": 0.8241,
+      "step": 4072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 36.41409411098547,
+      "learning_rate": 1.831182571797678e-05,
+      "loss": 0.8535,
+      "step": 4073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.22553819357881,
+      "learning_rate": 1.8310887962667706e-05,
+      "loss": 0.8582,
+      "step": 4074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 30.349060583894357,
+      "learning_rate": 1.830994997100184e-05,
+      "loss": 0.7536,
+      "step": 4075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 33.17321353463204,
+      "learning_rate": 1.8309011743005863e-05,
+      "loss": 0.7489,
+      "step": 4076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 38.99034168396991,
+      "learning_rate": 1.8308073278706448e-05,
+      "loss": 0.8356,
+      "step": 4077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 29.50429127079701,
+      "learning_rate": 1.8307134578130293e-05,
+      "loss": 0.8826,
+      "step": 4078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 39.814595531127544,
+      "learning_rate": 1.8306195641304088e-05,
+      "loss": 0.751,
+      "step": 4079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 42.264538690000606,
+      "learning_rate": 1.8305256468254537e-05,
+      "loss": 0.8227,
+      "step": 4080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.56483475730463,
+      "learning_rate": 1.830431705900835e-05,
+      "loss": 0.8417,
+      "step": 4081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 45.24689466664522,
+      "learning_rate": 1.8303377413592248e-05,
+      "loss": 0.7676,
+      "step": 4082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 35.45656724327605,
+      "learning_rate": 1.8302437532032945e-05,
+      "loss": 0.8356,
+      "step": 4083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 31.46370763829105,
+      "learning_rate": 1.830149741435718e-05,
+      "loss": 0.8406,
+      "step": 4084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 35.5556951961325,
+      "learning_rate": 1.830055706059168e-05,
+      "loss": 0.8751,
+      "step": 4085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 33.81141801537874,
+      "learning_rate": 1.8299616470763196e-05,
+      "loss": 0.7935,
+      "step": 4086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 36.14311257858524,
+      "learning_rate": 1.829867564489847e-05,
+      "loss": 0.7641,
+      "step": 4087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 30.971392267641512,
+      "learning_rate": 1.8297734583024273e-05,
+      "loss": 0.8067,
+      "step": 4088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.70642996107409,
+      "learning_rate": 1.8296793285167348e-05,
+      "loss": 0.8392,
+      "step": 4089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 50.88671707735464,
+      "learning_rate": 1.8295851751354482e-05,
+      "loss": 0.9396,
+      "step": 4090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 40.63799360380897,
+      "learning_rate": 1.8294909981612443e-05,
+      "loss": 0.8109,
+      "step": 4091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 35.098278018494156,
+      "learning_rate": 1.8293967975968016e-05,
+      "loss": 0.9222,
+      "step": 4092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 43.662352269946865,
+      "learning_rate": 1.8293025734447992e-05,
+      "loss": 0.9155,
+      "step": 4093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 42.38445658083527,
+      "learning_rate": 1.8292083257079166e-05,
+      "loss": 0.8619,
+      "step": 4094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.02571784697994,
+      "learning_rate": 1.8291140543888347e-05,
+      "loss": 0.861,
+      "step": 4095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 49.816611226815574,
+      "learning_rate": 1.8290197594902337e-05,
+      "loss": 0.9143,
+      "step": 4096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 33.28669410199872,
+      "learning_rate": 1.8289254410147958e-05,
+      "loss": 0.7771,
+      "step": 4097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 42.252351780055825,
+      "learning_rate": 1.8288310989652035e-05,
+      "loss": 0.9049,
+      "step": 4098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 37.44051312021615,
+      "learning_rate": 1.828736733344139e-05,
+      "loss": 0.7633,
+      "step": 4099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 56.10912262390538,
+      "learning_rate": 1.8286423441542874e-05,
+      "loss": 0.8392,
+      "step": 4100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 45.93999013055338,
+      "learning_rate": 1.828547931398332e-05,
+      "loss": 0.886,
+      "step": 4101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 39.99470005686441,
+      "learning_rate": 1.8284534950789578e-05,
+      "loss": 0.9379,
+      "step": 4102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 47.07386962619349,
+      "learning_rate": 1.828359035198851e-05,
+      "loss": 0.9452,
+      "step": 4103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 44.669485620014036,
+      "learning_rate": 1.8282645517606977e-05,
+      "loss": 0.7954,
+      "step": 4104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 39.55394105698566,
+      "learning_rate": 1.8281700447671852e-05,
+      "loss": 0.9655,
+      "step": 4105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 44.79206946376696,
+      "learning_rate": 1.828075514221001e-05,
+      "loss": 0.7659,
+      "step": 4106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 39.12587722428657,
+      "learning_rate": 1.8279809601248342e-05,
+      "loss": 0.8793,
+      "step": 4107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 36.75871461902451,
+      "learning_rate": 1.8278863824813726e-05,
+      "loss": 0.9763,
+      "step": 4108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 41.49090657800549,
+      "learning_rate": 1.8277917812933068e-05,
+      "loss": 0.7825,
+      "step": 4109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 42.574592732191235,
+      "learning_rate": 1.827697156563327e-05,
+      "loss": 0.7591,
+      "step": 4110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 41.299034232515815,
+      "learning_rate": 1.827602508294124e-05,
+      "loss": 0.767,
+      "step": 4111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.174159870650065,
+      "learning_rate": 1.82750783648839e-05,
+      "loss": 0.7724,
+      "step": 4112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 47.889689609671144,
+      "learning_rate": 1.8274131411488172e-05,
+      "loss": 0.8578,
+      "step": 4113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 43.0863154819264,
+      "learning_rate": 1.8273184222780984e-05,
+      "loss": 0.7822,
+      "step": 4114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 36.51688879370035,
+      "learning_rate": 1.827223679878928e-05,
+      "loss": 0.9558,
+      "step": 4115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 31.389641911106885,
+      "learning_rate": 1.8271289139539997e-05,
+      "loss": 0.8221,
+      "step": 4116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 41.070889171944096,
+      "learning_rate": 1.8270341245060095e-05,
+      "loss": 0.8552,
+      "step": 4117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 40.96994789636049,
+      "learning_rate": 1.8269393115376522e-05,
+      "loss": 0.7877,
+      "step": 4118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 39.20494117650184,
+      "learning_rate": 1.8268444750516246e-05,
+      "loss": 0.7924,
+      "step": 4119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.53973265150692,
+      "learning_rate": 1.8267496150506242e-05,
+      "loss": 0.8427,
+      "step": 4120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 31.345811656265102,
+      "learning_rate": 1.826654731537348e-05,
+      "loss": 0.8037,
+      "step": 4121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.49639798524182,
+      "learning_rate": 1.826559824514495e-05,
+      "loss": 0.889,
+      "step": 4122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 42.777391397635846,
+      "learning_rate": 1.8264648939847642e-05,
+      "loss": 0.8382,
+      "step": 4123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 44.35579532316618,
+      "learning_rate": 1.826369939950855e-05,
+      "loss": 0.9222,
+      "step": 4124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 30.497497889805338,
+      "learning_rate": 1.8262749624154684e-05,
+      "loss": 0.8366,
+      "step": 4125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 30.81500245778713,
+      "learning_rate": 1.8261799613813053e-05,
+      "loss": 0.6687,
+      "step": 4126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 36.741488573977016,
+      "learning_rate": 1.826084936851067e-05,
+      "loss": 0.7739,
+      "step": 4127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 35.16424294039611,
+      "learning_rate": 1.8259898888274564e-05,
+      "loss": 0.8732,
+      "step": 4128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 49.751919637478665,
+      "learning_rate": 1.8258948173131766e-05,
+      "loss": 0.8711,
+      "step": 4129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 35.427734031784496,
+      "learning_rate": 1.8257997223109317e-05,
+      "loss": 1.008,
+      "step": 4130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 32.84915640022142,
+      "learning_rate": 1.825704603823425e-05,
+      "loss": 0.7806,
+      "step": 4131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 40.41181413025426,
+      "learning_rate": 1.8256094618533633e-05,
+      "loss": 0.8919,
+      "step": 4132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 30.439290959425264,
+      "learning_rate": 1.825514296403451e-05,
+      "loss": 0.7867,
+      "step": 4133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 67.80278195831815,
+      "learning_rate": 1.825419107476395e-05,
+      "loss": 0.7547,
+      "step": 4134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 29.046473442367688,
+      "learning_rate": 1.8253238950749025e-05,
+      "loss": 0.7991,
+      "step": 4135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 28.16098755704673,
+      "learning_rate": 1.8252286592016812e-05,
+      "loss": 0.8042,
+      "step": 4136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 39.51881130370007,
+      "learning_rate": 1.8251333998594395e-05,
+      "loss": 0.9077,
+      "step": 4137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 40.72267510210145,
+      "learning_rate": 1.825038117050887e-05,
+      "loss": 0.778,
+      "step": 4138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 93.88506583180668,
+      "learning_rate": 1.824942810778733e-05,
+      "loss": 0.8217,
+      "step": 4139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 42.21946376872015,
+      "learning_rate": 1.8248474810456874e-05,
+      "loss": 0.9316,
+      "step": 4140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 35.69889851935845,
+      "learning_rate": 1.8247521278544625e-05,
+      "loss": 0.7887,
+      "step": 4141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 45.55053819742499,
+      "learning_rate": 1.8246567512077694e-05,
+      "loss": 0.9168,
+      "step": 4142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 31.724668117948607,
+      "learning_rate": 1.824561351108321e-05,
+      "loss": 0.9122,
+      "step": 4143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 44.44202926345238,
+      "learning_rate": 1.8244659275588298e-05,
+      "loss": 0.8764,
+      "step": 4144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 46.37514312033149,
+      "learning_rate": 1.82437048056201e-05,
+      "loss": 0.7632,
+      "step": 4145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 39.32211947140385,
+      "learning_rate": 1.8242750101205757e-05,
+      "loss": 0.7927,
+      "step": 4146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 42.70226096302596,
+      "learning_rate": 1.8241795162372426e-05,
+      "loss": 0.9264,
+      "step": 4147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 46.899435538227934,
+      "learning_rate": 1.8240839989147264e-05,
+      "loss": 0.8214,
+      "step": 4148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 33.634813340912295,
+      "learning_rate": 1.823988458155743e-05,
+      "loss": 0.8582,
+      "step": 4149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 34.12263146589142,
+      "learning_rate": 1.8238928939630097e-05,
+      "loss": 0.8371,
+      "step": 4150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 46.539496926480865,
+      "learning_rate": 1.823797306339245e-05,
+      "loss": 0.8622,
+      "step": 4151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 40.693625098968184,
+      "learning_rate": 1.8237016952871664e-05,
+      "loss": 0.7343,
+      "step": 4152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 33.66213907628399,
+      "learning_rate": 1.8236060608094938e-05,
+      "loss": 0.749,
+      "step": 4153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 43.395396535173475,
+      "learning_rate": 1.8235104029089465e-05,
+      "loss": 0.7345,
+      "step": 4154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.14045626217383,
+      "learning_rate": 1.823414721588245e-05,
+      "loss": 0.8286,
+      "step": 4155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 34.366799864216404,
+      "learning_rate": 1.8233190168501104e-05,
+      "loss": 0.8534,
+      "step": 4156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.659730702096304,
+      "learning_rate": 1.8232232886972643e-05,
+      "loss": 0.8707,
+      "step": 4157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 33.83143155542379,
+      "learning_rate": 1.8231275371324303e-05,
+      "loss": 0.7288,
+      "step": 4158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.5683849641979,
+      "learning_rate": 1.82303176215833e-05,
+      "loss": 0.7846,
+      "step": 4159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 35.61253137396317,
+      "learning_rate": 1.8229359637776883e-05,
+      "loss": 0.9112,
+      "step": 4160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 34.403700939412786,
+      "learning_rate": 1.8228401419932287e-05,
+      "loss": 0.8273,
+      "step": 4161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 41.4022324768696,
+      "learning_rate": 1.822744296807677e-05,
+      "loss": 0.8272,
+      "step": 4162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.58478303901938,
+      "learning_rate": 1.8226484282237587e-05,
+      "loss": 0.8452,
+      "step": 4163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 70.54997525758803,
+      "learning_rate": 1.822552536244201e-05,
+      "loss": 0.8312,
+      "step": 4164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 31.772307929958416,
+      "learning_rate": 1.8224566208717296e-05,
+      "loss": 0.7848,
+      "step": 4165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.06875517096025,
+      "learning_rate": 1.8223606821090733e-05,
+      "loss": 0.8842,
+      "step": 4166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 34.340062655906635,
+      "learning_rate": 1.82226471995896e-05,
+      "loss": 0.8107,
+      "step": 4167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 40.15036614798577,
+      "learning_rate": 1.8221687344241193e-05,
+      "loss": 0.9347,
+      "step": 4168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 34.307401684822615,
+      "learning_rate": 1.822072725507281e-05,
+      "loss": 0.8183,
+      "step": 4169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 33.32328052964039,
+      "learning_rate": 1.8219766932111747e-05,
+      "loss": 0.7893,
+      "step": 4170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 44.59696911939625,
+      "learning_rate": 1.8218806375385326e-05,
+      "loss": 0.7642,
+      "step": 4171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.87125560231133,
+      "learning_rate": 1.8217845584920858e-05,
+      "loss": 0.8333,
+      "step": 4172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 98.53296049879441,
+      "learning_rate": 1.821688456074567e-05,
+      "loss": 0.7626,
+      "step": 4173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 37.98302456374721,
+      "learning_rate": 1.8215923302887085e-05,
+      "loss": 0.8989,
+      "step": 4174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.98710850105251,
+      "learning_rate": 1.8214961811372455e-05,
+      "loss": 0.9469,
+      "step": 4175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 32.29408773677898,
+      "learning_rate": 1.8214000086229115e-05,
+      "loss": 0.7271,
+      "step": 4176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 37.174125824522264,
+      "learning_rate": 1.8213038127484415e-05,
+      "loss": 0.9249,
+      "step": 4177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.35061644932631,
+      "learning_rate": 1.821207593516572e-05,
+      "loss": 0.8344,
+      "step": 4178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.224506827317214,
+      "learning_rate": 1.8211113509300384e-05,
+      "loss": 0.824,
+      "step": 4179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 71.34010036255044,
+      "learning_rate": 1.8210150849915787e-05,
+      "loss": 0.7917,
+      "step": 4180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.87318415555972,
+      "learning_rate": 1.82091879570393e-05,
+      "loss": 0.8707,
+      "step": 4181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 39.437549929579674,
+      "learning_rate": 1.8208224830698312e-05,
+      "loss": 0.9568,
+      "step": 4182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 43.43831743527794,
+      "learning_rate": 1.820726147092021e-05,
+      "loss": 0.835,
+      "step": 4183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.25355925680115,
+      "learning_rate": 1.8206297877732393e-05,
+      "loss": 0.7701,
+      "step": 4184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.349386091439676,
+      "learning_rate": 1.8205334051162264e-05,
+      "loss": 0.794,
+      "step": 4185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 50.393469945750844,
+      "learning_rate": 1.8204369991237237e-05,
+      "loss": 0.7594,
+      "step": 4186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 46.60301232077166,
+      "learning_rate": 1.8203405697984726e-05,
+      "loss": 0.8104,
+      "step": 4187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 35.29198971495035,
+      "learning_rate": 1.8202441171432156e-05,
+      "loss": 0.8509,
+      "step": 4188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 59.669338550983426,
+      "learning_rate": 1.8201476411606956e-05,
+      "loss": 0.8508,
+      "step": 4189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 57.71803465824864,
+      "learning_rate": 1.8200511418536563e-05,
+      "loss": 0.9183,
+      "step": 4190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 32.16533567379444,
+      "learning_rate": 1.819954619224843e-05,
+      "loss": 0.7585,
+      "step": 4191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 44.733237824690995,
+      "learning_rate": 1.8198580732769992e-05,
+      "loss": 0.8179,
+      "step": 4192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 54.9008576595387,
+      "learning_rate": 1.819761504012872e-05,
+      "loss": 0.8377,
+      "step": 4193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.5796635237141,
+      "learning_rate": 1.8196649114352065e-05,
+      "loss": 0.8333,
+      "step": 4194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.38039177152057,
+      "learning_rate": 1.819568295546751e-05,
+      "loss": 0.7953,
+      "step": 4195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.26465618235788,
+      "learning_rate": 1.819471656350252e-05,
+      "loss": 0.8661,
+      "step": 4196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 39.28585702529589,
+      "learning_rate": 1.819374993848459e-05,
+      "loss": 0.8818,
+      "step": 4197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 39.207287947082435,
+      "learning_rate": 1.8192783080441204e-05,
+      "loss": 0.822,
+      "step": 4198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 33.91561759190139,
+      "learning_rate": 1.819181598939986e-05,
+      "loss": 0.8833,
+      "step": 4199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 44.757612521257,
+      "learning_rate": 1.819084866538806e-05,
+      "loss": 0.9004,
+      "step": 4200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.43946582450326,
+      "learning_rate": 1.8189881108433317e-05,
+      "loss": 0.8595,
+      "step": 4201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 41.50431050469485,
+      "learning_rate": 1.8188913318563145e-05,
+      "loss": 0.816,
+      "step": 4202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 37.738319055009015,
+      "learning_rate": 1.8187945295805066e-05,
+      "loss": 0.7402,
+      "step": 4203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 66.5108856330017,
+      "learning_rate": 1.8186977040186615e-05,
+      "loss": 0.9577,
+      "step": 4204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 45.053771354789404,
+      "learning_rate": 1.8186008551735327e-05,
+      "loss": 0.8182,
+      "step": 4205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 39.19225997245897,
+      "learning_rate": 1.8185039830478745e-05,
+      "loss": 0.8848,
+      "step": 4206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 40.39612914605484,
+      "learning_rate": 1.8184070876444417e-05,
+      "loss": 0.8056,
+      "step": 4207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 41.59883352579111,
+      "learning_rate": 1.8183101689659902e-05,
+      "loss": 0.8843,
+      "step": 4208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 29.891917721701095,
+      "learning_rate": 1.818213227015276e-05,
+      "loss": 0.8687,
+      "step": 4209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.79548633341359,
+      "learning_rate": 1.8181162617950563e-05,
+      "loss": 0.7837,
+      "step": 4210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 59.08450401591883,
+      "learning_rate": 1.8180192733080887e-05,
+      "loss": 0.8161,
+      "step": 4211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.94712306721468,
+      "learning_rate": 1.817922261557132e-05,
+      "loss": 0.796,
+      "step": 4212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.086764244949244,
+      "learning_rate": 1.817825226544944e-05,
+      "loss": 0.7548,
+      "step": 4213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 42.018952538318345,
+      "learning_rate": 1.817728168274285e-05,
+      "loss": 0.8406,
+      "step": 4214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 36.31373355431388,
+      "learning_rate": 1.8176310867479155e-05,
+      "loss": 0.8614,
+      "step": 4215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 30.34644836386719,
+      "learning_rate": 1.8175339819685965e-05,
+      "loss": 0.8125,
+      "step": 4216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.16443194207598,
+      "learning_rate": 1.8174368539390888e-05,
+      "loss": 0.8648,
+      "step": 4217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.59823573372111,
+      "learning_rate": 1.8173397026621554e-05,
+      "loss": 0.8225,
+      "step": 4218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 67.52157586323159,
+      "learning_rate": 1.817242528140559e-05,
+      "loss": 0.9049,
+      "step": 4219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.13540421678036,
+      "learning_rate": 1.8171453303770628e-05,
+      "loss": 0.9172,
+      "step": 4220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.943279790082826,
+      "learning_rate": 1.8170481093744317e-05,
+      "loss": 0.7693,
+      "step": 4221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.07087480085554,
+      "learning_rate": 1.81695086513543e-05,
+      "loss": 0.862,
+      "step": 4222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 37.9772022807887,
+      "learning_rate": 1.8168535976628242e-05,
+      "loss": 0.8647,
+      "step": 4223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 31.492662044385938,
+      "learning_rate": 1.8167563069593796e-05,
+      "loss": 0.8397,
+      "step": 4224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 47.62057822017819,
+      "learning_rate": 1.8166589930278637e-05,
+      "loss": 0.9154,
+      "step": 4225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 36.07782475944214,
+      "learning_rate": 1.816561655871043e-05,
+      "loss": 0.7755,
+      "step": 4226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 30.90154249405993,
+      "learning_rate": 1.816464295491687e-05,
+      "loss": 0.7879,
+      "step": 4227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.05557216222447,
+      "learning_rate": 1.816366911892564e-05,
+      "loss": 0.8173,
+      "step": 4228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 41.67823741217232,
+      "learning_rate": 1.8162695050764437e-05,
+      "loss": 0.9366,
+      "step": 4229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.046716207999296,
+      "learning_rate": 1.8161720750460963e-05,
+      "loss": 0.8983,
+      "step": 4230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.0457174271552,
+      "learning_rate": 1.8160746218042923e-05,
+      "loss": 0.7752,
+      "step": 4231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.533589914711456,
+      "learning_rate": 1.8159771453538034e-05,
+      "loss": 0.7551,
+      "step": 4232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 56.57832228935123,
+      "learning_rate": 1.8158796456974023e-05,
+      "loss": 0.9709,
+      "step": 4233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.36647992261564,
+      "learning_rate": 1.8157821228378607e-05,
+      "loss": 0.8027,
+      "step": 4234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.486394874266004,
+      "learning_rate": 1.8156845767779532e-05,
+      "loss": 0.8423,
+      "step": 4235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 33.20924542912321,
+      "learning_rate": 1.815587007520453e-05,
+      "loss": 0.689,
+      "step": 4236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 37.911656889579156,
+      "learning_rate": 1.815489415068136e-05,
+      "loss": 0.8438,
+      "step": 4237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 40.31042347372579,
+      "learning_rate": 1.8153917994237765e-05,
+      "loss": 0.8688,
+      "step": 4238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 30.950080845130994,
+      "learning_rate": 1.8152941605901518e-05,
+      "loss": 0.8043,
+      "step": 4239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 34.982806444463606,
+      "learning_rate": 1.8151964985700375e-05,
+      "loss": 0.8545,
+      "step": 4240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 39.63342630883699,
+      "learning_rate": 1.815098813366212e-05,
+      "loss": 0.9212,
+      "step": 4241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 37.21755081089754,
+      "learning_rate": 1.8150011049814534e-05,
+      "loss": 0.8998,
+      "step": 4242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 37.052951760026616,
+      "learning_rate": 1.8149033734185396e-05,
+      "loss": 0.9241,
+      "step": 4243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 32.378875947786916,
+      "learning_rate": 1.814805618680251e-05,
+      "loss": 0.8949,
+      "step": 4244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 45.363235861222506,
+      "learning_rate": 1.814707840769367e-05,
+      "loss": 0.8377,
+      "step": 4245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 36.393826273032936,
+      "learning_rate": 1.8146100396886685e-05,
+      "loss": 0.8157,
+      "step": 4246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 33.66311397707912,
+      "learning_rate": 1.814512215440937e-05,
+      "loss": 0.7336,
+      "step": 4247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 31.711607936060197,
+      "learning_rate": 1.8144143680289547e-05,
+      "loss": 0.8127,
+      "step": 4248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 32.83868930348852,
+      "learning_rate": 1.8143164974555042e-05,
+      "loss": 0.8078,
+      "step": 4249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 37.64875409875197,
+      "learning_rate": 1.8142186037233687e-05,
+      "loss": 0.8327,
+      "step": 4250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 31.999588293168898,
+      "learning_rate": 1.8141206868353327e-05,
+      "loss": 0.7545,
+      "step": 4251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 32.48605765076054,
+      "learning_rate": 1.8140227467941803e-05,
+      "loss": 0.7457,
+      "step": 4252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 39.281286920911,
+      "learning_rate": 1.8139247836026977e-05,
+      "loss": 0.7979,
+      "step": 4253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 33.12449465672076,
+      "learning_rate": 1.8138267972636697e-05,
+      "loss": 0.8423,
+      "step": 4254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 45.86620161724993,
+      "learning_rate": 1.8137287877798842e-05,
+      "loss": 0.816,
+      "step": 4255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 31.380544180227023,
+      "learning_rate": 1.8136307551541276e-05,
+      "loss": 0.7565,
+      "step": 4256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 43.222013619218345,
+      "learning_rate": 1.8135326993891886e-05,
+      "loss": 0.7538,
+      "step": 4257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.589072234274774,
+      "learning_rate": 1.8134346204878553e-05,
+      "loss": 0.7812,
+      "step": 4258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 36.18311571499153,
+      "learning_rate": 1.8133365184529176e-05,
+      "loss": 0.9111,
+      "step": 4259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 40.16671219436682,
+      "learning_rate": 1.8132383932871645e-05,
+      "loss": 0.9219,
+      "step": 4260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 38.897475001434195,
+      "learning_rate": 1.8131402449933875e-05,
+      "loss": 0.7659,
+      "step": 4261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 30.96043692450535,
+      "learning_rate": 1.813042073574378e-05,
+      "loss": 0.7404,
+      "step": 4262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 47.26446024318667,
+      "learning_rate": 1.8129438790329272e-05,
+      "loss": 0.9737,
+      "step": 4263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.25888827471488,
+      "learning_rate": 1.8128456613718278e-05,
+      "loss": 0.816,
+      "step": 4264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 40.187878715190124,
+      "learning_rate": 1.812747420593874e-05,
+      "loss": 0.9527,
+      "step": 4265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.5444172092803,
+      "learning_rate": 1.8126491567018582e-05,
+      "loss": 0.7738,
+      "step": 4266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 38.25443897170479,
+      "learning_rate": 1.8125508696985763e-05,
+      "loss": 0.8477,
+      "step": 4267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 40.07164513031833,
+      "learning_rate": 1.8124525595868228e-05,
+      "loss": 0.7872,
+      "step": 4268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 29.448110093807383,
+      "learning_rate": 1.8123542263693935e-05,
+      "loss": 0.7084,
+      "step": 4269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.46278136864134,
+      "learning_rate": 1.8122558700490857e-05,
+      "loss": 0.8883,
+      "step": 4270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 45.76229186730985,
+      "learning_rate": 1.812157490628696e-05,
+      "loss": 0.8127,
+      "step": 4271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.30917640390204,
+      "learning_rate": 1.812059088111022e-05,
+      "loss": 0.8212,
+      "step": 4272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 48.59095852061694,
+      "learning_rate": 1.8119606624988632e-05,
+      "loss": 0.8857,
+      "step": 4273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 43.12192564247464,
+      "learning_rate": 1.8118622137950174e-05,
+      "loss": 0.7084,
+      "step": 4274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 61.840660681143895,
+      "learning_rate": 1.8117637420022858e-05,
+      "loss": 0.8428,
+      "step": 4275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 55.012030122462605,
+      "learning_rate": 1.8116652471234682e-05,
+      "loss": 0.8199,
+      "step": 4276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.565547869492846,
+      "learning_rate": 1.8115667291613653e-05,
+      "loss": 0.7733,
+      "step": 4277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 42.162987172817886,
+      "learning_rate": 1.8114681881187795e-05,
+      "loss": 0.9752,
+      "step": 4278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 47.62345106254286,
+      "learning_rate": 1.8113696239985134e-05,
+      "loss": 0.8894,
+      "step": 4279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 40.87146975936677,
+      "learning_rate": 1.81127103680337e-05,
+      "loss": 0.8948,
+      "step": 4280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 50.519883042145004,
+      "learning_rate": 1.8111724265361525e-05,
+      "loss": 0.9611,
+      "step": 4281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 44.04993347992134,
+      "learning_rate": 1.8110737931996658e-05,
+      "loss": 0.7925,
+      "step": 4282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 42.03388572285031,
+      "learning_rate": 1.8109751367967148e-05,
+      "loss": 0.9212,
+      "step": 4283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 39.6681996761871,
+      "learning_rate": 1.8108764573301054e-05,
+      "loss": 0.8899,
+      "step": 4284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 45.043518280072675,
+      "learning_rate": 1.810777754802644e-05,
+      "loss": 0.7737,
+      "step": 4285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 36.60259511299144,
+      "learning_rate": 1.8106790292171378e-05,
+      "loss": 0.8436,
+      "step": 4286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.71014355269347,
+      "learning_rate": 1.8105802805763938e-05,
+      "loss": 0.817,
+      "step": 4287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.6262249161854,
+      "learning_rate": 1.810481508883221e-05,
+      "loss": 0.726,
+      "step": 4288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.211855046066574,
+      "learning_rate": 1.810382714140428e-05,
+      "loss": 0.7594,
+      "step": 4289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 36.45160747431933,
+      "learning_rate": 1.8102838963508247e-05,
+      "loss": 0.8533,
+      "step": 4290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.523105770841184,
+      "learning_rate": 1.8101850555172214e-05,
+      "loss": 0.8625,
+      "step": 4291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 55.836491758297186,
+      "learning_rate": 1.8100861916424293e-05,
+      "loss": 0.9328,
+      "step": 4292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.10917545244632,
+      "learning_rate": 1.8099873047292596e-05,
+      "loss": 0.7255,
+      "step": 4293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.31939793344635,
+      "learning_rate": 1.8098883947805247e-05,
+      "loss": 0.8146,
+      "step": 4294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.800347561454416,
+      "learning_rate": 1.8097894617990378e-05,
+      "loss": 1.0073,
+      "step": 4295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.65502756415406,
+      "learning_rate": 1.809690505787612e-05,
+      "loss": 0.8047,
+      "step": 4296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 37.96298176469484,
+      "learning_rate": 1.8095915267490625e-05,
+      "loss": 0.8328,
+      "step": 4297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 30.58673447780054,
+      "learning_rate": 1.809492524686203e-05,
+      "loss": 0.762,
+      "step": 4298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.38794474922535,
+      "learning_rate": 1.80939349960185e-05,
+      "loss": 0.8601,
+      "step": 4299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 39.4931907941417,
+      "learning_rate": 1.8092944514988193e-05,
+      "loss": 0.8495,
+      "step": 4300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.16686216785152,
+      "learning_rate": 1.8091953803799275e-05,
+      "loss": 0.9107,
+      "step": 4301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 41.28325200045267,
+      "learning_rate": 1.8090962862479927e-05,
+      "loss": 0.8782,
+      "step": 4302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 46.349552730142946,
+      "learning_rate": 1.8089971691058327e-05,
+      "loss": 0.8504,
+      "step": 4303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 42.83967387302332,
+      "learning_rate": 1.8088980289562666e-05,
+      "loss": 0.7901,
+      "step": 4304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 51.117060989253645,
+      "learning_rate": 1.8087988658021137e-05,
+      "loss": 0.8309,
+      "step": 4305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 76.169626081084,
+      "learning_rate": 1.8086996796461943e-05,
+      "loss": 0.9376,
+      "step": 4306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.60265470061314,
+      "learning_rate": 1.808600470491329e-05,
+      "loss": 0.8264,
+      "step": 4307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.282565142715654,
+      "learning_rate": 1.8085012383403392e-05,
+      "loss": 0.8216,
+      "step": 4308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 37.9664147118025,
+      "learning_rate": 1.8084019831960474e-05,
+      "loss": 0.8777,
+      "step": 4309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.35571933728974,
+      "learning_rate": 1.808302705061276e-05,
+      "loss": 0.7819,
+      "step": 4310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 50.47102142985892,
+      "learning_rate": 1.8082034039388486e-05,
+      "loss": 0.9198,
+      "step": 4311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 38.215658775574816,
+      "learning_rate": 1.808104079831589e-05,
+      "loss": 0.8867,
+      "step": 4312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 38.175644896480485,
+      "learning_rate": 1.808004732742322e-05,
+      "loss": 0.8322,
+      "step": 4313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 31.47812332672899,
+      "learning_rate": 1.8079053626738732e-05,
+      "loss": 0.8695,
+      "step": 4314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 74.5962835040662,
+      "learning_rate": 1.8078059696290685e-05,
+      "loss": 0.9325,
+      "step": 4315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.95369493521391,
+      "learning_rate": 1.8077065536107347e-05,
+      "loss": 0.8176,
+      "step": 4316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.90950611846044,
+      "learning_rate": 1.8076071146216988e-05,
+      "loss": 0.816,
+      "step": 4317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.15882940027539,
+      "learning_rate": 1.8075076526647892e-05,
+      "loss": 0.8113,
+      "step": 4318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.59621884405428,
+      "learning_rate": 1.807408167742834e-05,
+      "loss": 0.7883,
+      "step": 4319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 68.51381512712324,
+      "learning_rate": 1.8073086598586634e-05,
+      "loss": 0.8794,
+      "step": 4320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 46.01562012444099,
+      "learning_rate": 1.8072091290151063e-05,
+      "loss": 0.8571,
+      "step": 4321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.20874394128949,
+      "learning_rate": 1.807109575214994e-05,
+      "loss": 0.8499,
+      "step": 4322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 47.93128646508965,
+      "learning_rate": 1.8070099984611575e-05,
+      "loss": 0.7889,
+      "step": 4323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 38.79616335501529,
+      "learning_rate": 1.8069103987564286e-05,
+      "loss": 0.9286,
+      "step": 4324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 90.93662515855233,
+      "learning_rate": 1.8068107761036402e-05,
+      "loss": 0.7976,
+      "step": 4325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 57.538547950199764,
+      "learning_rate": 1.806711130505625e-05,
+      "loss": 0.9165,
+      "step": 4326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 40.11735583959991,
+      "learning_rate": 1.806611461965217e-05,
+      "loss": 0.8386,
+      "step": 4327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 42.53754065167247,
+      "learning_rate": 1.8065117704852518e-05,
+      "loss": 0.82,
+      "step": 4328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 40.7635276316301,
+      "learning_rate": 1.8064120560685626e-05,
+      "loss": 0.8291,
+      "step": 4329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 36.23874870180551,
+      "learning_rate": 1.806312318717987e-05,
+      "loss": 0.8226,
+      "step": 4330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 30.17171900198705,
+      "learning_rate": 1.8062125584363605e-05,
+      "loss": 0.7619,
+      "step": 4331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.9299669343848,
+      "learning_rate": 1.8061127752265203e-05,
+      "loss": 0.809,
+      "step": 4332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.36962700226758,
+      "learning_rate": 1.8060129690913045e-05,
+      "loss": 0.8152,
+      "step": 4333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 51.26299783520275,
+      "learning_rate": 1.8059131400335512e-05,
+      "loss": 0.9602,
+      "step": 4334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 38.95569701258995,
+      "learning_rate": 1.8058132880560998e-05,
+      "loss": 0.9229,
+      "step": 4335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.139951540145525,
+      "learning_rate": 1.80571341316179e-05,
+      "loss": 0.8277,
+      "step": 4336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 43.41439680266299,
+      "learning_rate": 1.8056135153534616e-05,
+      "loss": 0.9002,
+      "step": 4337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 46.66360281809349,
+      "learning_rate": 1.8055135946339563e-05,
+      "loss": 0.8946,
+      "step": 4338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.50600933491211,
+      "learning_rate": 1.8054136510061156e-05,
+      "loss": 0.7587,
+      "step": 4339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 38.824267699685464,
+      "learning_rate": 1.8053136844727822e-05,
+      "loss": 0.9187,
+      "step": 4340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 35.43330469477212,
+      "learning_rate": 1.8052136950367983e-05,
+      "loss": 0.8464,
+      "step": 4341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.47703051254821,
+      "learning_rate": 1.8051136827010077e-05,
+      "loss": 0.8129,
+      "step": 4342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 46.28393517019228,
+      "learning_rate": 1.8050136474682553e-05,
+      "loss": 0.8274,
+      "step": 4343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 36.62120851544517,
+      "learning_rate": 1.8049135893413858e-05,
+      "loss": 0.8535,
+      "step": 4344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 26.937258093117247,
+      "learning_rate": 1.8048135083232446e-05,
+      "loss": 0.7862,
+      "step": 4345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 35.50608110728466,
+      "learning_rate": 1.804713404416678e-05,
+      "loss": 0.7974,
+      "step": 4346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 30.728089085882548,
+      "learning_rate": 1.8046132776245327e-05,
+      "loss": 0.8203,
+      "step": 4347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 40.84807386911693,
+      "learning_rate": 1.804513127949657e-05,
+      "loss": 0.8947,
+      "step": 4348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 35.51675537431852,
+      "learning_rate": 1.8044129553948978e-05,
+      "loss": 0.8672,
+      "step": 4349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.01019013092321,
+      "learning_rate": 1.8043127599631053e-05,
+      "loss": 0.7141,
+      "step": 4350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 46.2474761789777,
+      "learning_rate": 1.8042125416571284e-05,
+      "loss": 0.7759,
+      "step": 4351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 39.61444971204329,
+      "learning_rate": 1.8041123004798173e-05,
+      "loss": 0.7869,
+      "step": 4352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 48.75899893037386,
+      "learning_rate": 1.8040120364340223e-05,
+      "loss": 0.7879,
+      "step": 4353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.54148512816905,
+      "learning_rate": 1.8039117495225956e-05,
+      "loss": 0.7592,
+      "step": 4354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.45336875768884,
+      "learning_rate": 1.803811439748389e-05,
+      "loss": 0.8791,
+      "step": 4355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 45.05377233907671,
+      "learning_rate": 1.8037111071142552e-05,
+      "loss": 0.8386,
+      "step": 4356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 38.80126912570589,
+      "learning_rate": 1.8036107516230478e-05,
+      "loss": 0.8134,
+      "step": 4357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 41.24995157065583,
+      "learning_rate": 1.8035103732776206e-05,
+      "loss": 1.0147,
+      "step": 4358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 48.97629126325947,
+      "learning_rate": 1.803409972080829e-05,
+      "loss": 0.8928,
+      "step": 4359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 33.70645560200643,
+      "learning_rate": 1.803309548035527e-05,
+      "loss": 0.9031,
+      "step": 4360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 36.25832712950346,
+      "learning_rate": 1.8032091011445714e-05,
+      "loss": 0.7687,
+      "step": 4361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.900275884503515,
+      "learning_rate": 1.803108631410819e-05,
+      "loss": 0.8645,
+      "step": 4362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 43.22502436022185,
+      "learning_rate": 1.803008138837127e-05,
+      "loss": 0.8479,
+      "step": 4363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 33.88742630013754,
+      "learning_rate": 1.8029076234263534e-05,
+      "loss": 0.8301,
+      "step": 4364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 30.727806549144624,
+      "learning_rate": 1.8028070851813566e-05,
+      "loss": 0.8816,
+      "step": 4365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 41.92027765928289,
+      "learning_rate": 1.802706524104996e-05,
+      "loss": 0.8396,
+      "step": 4366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.877369218003416,
+      "learning_rate": 1.8026059402001313e-05,
+      "loss": 0.8748,
+      "step": 4367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.364513003638884,
+      "learning_rate": 1.8025053334696232e-05,
+      "loss": 0.7096,
+      "step": 4368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 40.02539816020377,
+      "learning_rate": 1.8024047039163325e-05,
+      "loss": 0.8942,
+      "step": 4369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.09102431082761,
+      "learning_rate": 1.802304051543122e-05,
+      "loss": 0.8634,
+      "step": 4370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.266368156396524,
+      "learning_rate": 1.8022033763528534e-05,
+      "loss": 0.8001,
+      "step": 4371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.29219097572491,
+      "learning_rate": 1.80210267834839e-05,
+      "loss": 0.9265,
+      "step": 4372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.097343079585784,
+      "learning_rate": 1.802001957532596e-05,
+      "loss": 0.7901,
+      "step": 4373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 35.20227664377362,
+      "learning_rate": 1.801901213908335e-05,
+      "loss": 0.8231,
+      "step": 4374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 33.97298752745037,
+      "learning_rate": 1.801800447478473e-05,
+      "loss": 0.8422,
+      "step": 4375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.13153539596514,
+      "learning_rate": 1.8016996582458746e-05,
+      "loss": 0.7825,
+      "step": 4376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 31.57007932399853,
+      "learning_rate": 1.8015988462134073e-05,
+      "loss": 0.7969,
+      "step": 4377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 36.59856937509929,
+      "learning_rate": 1.801498011383938e-05,
+      "loss": 0.8713,
+      "step": 4378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 36.725435085069485,
+      "learning_rate": 1.8013971537603336e-05,
+      "loss": 0.7678,
+      "step": 4379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 34.4809988846934,
+      "learning_rate": 1.8012962733454636e-05,
+      "loss": 0.8348,
+      "step": 4380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 31.388355482589848,
+      "learning_rate": 1.8011953701421957e-05,
+      "loss": 0.7846,
+      "step": 4381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 33.99949965739946,
+      "learning_rate": 1.8010944441534004e-05,
+      "loss": 0.9066,
+      "step": 4382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 39.0089740912566,
+      "learning_rate": 1.800993495381948e-05,
+      "loss": 0.9292,
+      "step": 4383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 39.02781050640346,
+      "learning_rate": 1.800892523830709e-05,
+      "loss": 0.7848,
+      "step": 4384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 39.71812161744646,
+      "learning_rate": 1.8007915295025552e-05,
+      "loss": 0.9642,
+      "step": 4385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 38.581725606454064,
+      "learning_rate": 1.8006905124003583e-05,
+      "loss": 0.8245,
+      "step": 4386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 41.74923443520643,
+      "learning_rate": 1.8005894725269918e-05,
+      "loss": 0.8753,
+      "step": 4387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 37.386671785869765,
+      "learning_rate": 1.8004884098853296e-05,
+      "loss": 0.7934,
+      "step": 4388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 44.735867831483795,
+      "learning_rate": 1.8003873244782447e-05,
+      "loss": 0.8776,
+      "step": 4389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 44.31112678993445,
+      "learning_rate": 1.8002862163086127e-05,
+      "loss": 0.8237,
+      "step": 4390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.37750664510711,
+      "learning_rate": 1.8001850853793084e-05,
+      "loss": 0.6928,
+      "step": 4391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 40.03267513481018,
+      "learning_rate": 1.8000839316932094e-05,
+      "loss": 0.896,
+      "step": 4392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 36.60496590830405,
+      "learning_rate": 1.7999827552531905e-05,
+      "loss": 0.813,
+      "step": 4393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 48.26923337855144,
+      "learning_rate": 1.7998815560621305e-05,
+      "loss": 0.9092,
+      "step": 4394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 31.123369917256852,
+      "learning_rate": 1.7997803341229067e-05,
+      "loss": 0.8061,
+      "step": 4395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 41.621905037085554,
+      "learning_rate": 1.7996790894383982e-05,
+      "loss": 0.7803,
+      "step": 4396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 39.83137601601733,
+      "learning_rate": 1.799577822011484e-05,
+      "loss": 0.8425,
+      "step": 4397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 34.74624853330191,
+      "learning_rate": 1.7994765318450446e-05,
+      "loss": 0.6932,
+      "step": 4398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 65.52226268096706,
+      "learning_rate": 1.7993752189419602e-05,
+      "loss": 0.8463,
+      "step": 4399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 33.154871035563644,
+      "learning_rate": 1.7992738833051122e-05,
+      "loss": 0.7025,
+      "step": 4400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 35.17605478200514,
+      "learning_rate": 1.7991725249373822e-05,
+      "loss": 0.6951,
+      "step": 4401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.88025365848802,
+      "learning_rate": 1.7990711438416536e-05,
+      "loss": 0.8182,
+      "step": 4402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 71.49156712773639,
+      "learning_rate": 1.798969740020809e-05,
+      "loss": 0.8159,
+      "step": 4403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 35.16208829086121,
+      "learning_rate": 1.798868313477732e-05,
+      "loss": 0.782,
+      "step": 4404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 61.744951227100735,
+      "learning_rate": 1.7987668642153075e-05,
+      "loss": 0.9461,
+      "step": 4405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.08118605838648,
+      "learning_rate": 1.798665392236421e-05,
+      "loss": 0.7291,
+      "step": 4406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 44.16955987822865,
+      "learning_rate": 1.7985638975439578e-05,
+      "loss": 0.858,
+      "step": 4407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 31.84133818517788,
+      "learning_rate": 1.7984623801408046e-05,
+      "loss": 0.8274,
+      "step": 4408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 33.10047531233242,
+      "learning_rate": 1.798360840029848e-05,
+      "loss": 0.8325,
+      "step": 4409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 34.822371115801,
+      "learning_rate": 1.7982592772139766e-05,
+      "loss": 0.8097,
+      "step": 4410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 41.170519082255694,
+      "learning_rate": 1.7981576916960784e-05,
+      "loss": 0.7597,
+      "step": 4411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 26.33104953087427,
+      "learning_rate": 1.798056083479042e-05,
+      "loss": 0.7058,
+      "step": 4412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 37.03553458427842,
+      "learning_rate": 1.7979544525657578e-05,
+      "loss": 0.7071,
+      "step": 4413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 37.01683878488614,
+      "learning_rate": 1.797852798959115e-05,
+      "loss": 0.9029,
+      "step": 4414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 38.203803644997926,
+      "learning_rate": 1.7977511226620062e-05,
+      "loss": 0.8235,
+      "step": 4415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.84336455361253,
+      "learning_rate": 1.7976494236773217e-05,
+      "loss": 0.7565,
+      "step": 4416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 42.39283680866565,
+      "learning_rate": 1.7975477020079544e-05,
+      "loss": 0.8341,
+      "step": 4417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 39.049180385985245,
+      "learning_rate": 1.7974459576567968e-05,
+      "loss": 0.7608,
+      "step": 4418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 35.00900114323432,
+      "learning_rate": 1.797344190626743e-05,
+      "loss": 0.832,
+      "step": 4419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 38.034035222572555,
+      "learning_rate": 1.7972424009206865e-05,
+      "loss": 0.7471,
+      "step": 4420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 41.41526238589198,
+      "learning_rate": 1.7971405885415228e-05,
+      "loss": 0.7818,
+      "step": 4421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.58319439862445,
+      "learning_rate": 1.7970387534921467e-05,
+      "loss": 0.7834,
+      "step": 4422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 37.900815063081524,
+      "learning_rate": 1.796936895775455e-05,
+      "loss": 0.7618,
+      "step": 4423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 35.861020983316706,
+      "learning_rate": 1.7968350153943443e-05,
+      "loss": 0.8662,
+      "step": 4424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 42.393345437248854,
+      "learning_rate": 1.7967331123517114e-05,
+      "loss": 0.7513,
+      "step": 4425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 37.18427830052051,
+      "learning_rate": 1.7966311866504552e-05,
+      "loss": 0.866,
+      "step": 4426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 39.46020518008376,
+      "learning_rate": 1.796529238293474e-05,
+      "loss": 0.8492,
+      "step": 4427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 36.46394752908989,
+      "learning_rate": 1.7964272672836675e-05,
+      "loss": 0.8622,
+      "step": 4428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 31.49891466587051,
+      "learning_rate": 1.7963252736239347e-05,
+      "loss": 0.8159,
+      "step": 4429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 38.08437925310394,
+      "learning_rate": 1.7962232573171772e-05,
+      "loss": 0.8661,
+      "step": 4430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 38.03793194764054,
+      "learning_rate": 1.7961212183662967e-05,
+      "loss": 0.8754,
+      "step": 4431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 57.862468079487904,
+      "learning_rate": 1.796019156774194e-05,
+      "loss": 0.7845,
+      "step": 4432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 36.35224457098228,
+      "learning_rate": 1.795917072543772e-05,
+      "loss": 0.8641,
+      "step": 4433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 52.12673824257008,
+      "learning_rate": 1.7958149656779343e-05,
+      "loss": 0.7766,
+      "step": 4434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 37.59430323801727,
+      "learning_rate": 1.7957128361795844e-05,
+      "loss": 0.8347,
+      "step": 4435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 30.691640819242107,
+      "learning_rate": 1.7956106840516268e-05,
+      "loss": 0.7583,
+      "step": 4436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.83465191005586,
+      "learning_rate": 1.7955085092969668e-05,
+      "loss": 0.8403,
+      "step": 4437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 36.17175123551948,
+      "learning_rate": 1.7954063119185105e-05,
+      "loss": 0.8922,
+      "step": 4438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 48.15552792017054,
+      "learning_rate": 1.7953040919191635e-05,
+      "loss": 0.823,
+      "step": 4439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 34.11287736233173,
+      "learning_rate": 1.7952018493018335e-05,
+      "loss": 0.8424,
+      "step": 4440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 46.29486144492863,
+      "learning_rate": 1.7950995840694282e-05,
+      "loss": 0.9958,
+      "step": 4441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 33.62073350094872,
+      "learning_rate": 1.794997296224856e-05,
+      "loss": 0.8683,
+      "step": 4442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 53.79255229458273,
+      "learning_rate": 1.7948949857710256e-05,
+      "loss": 0.8488,
+      "step": 4443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 52.61625029614949,
+      "learning_rate": 1.7947926527108467e-05,
+      "loss": 0.8713,
+      "step": 4444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 34.99679463558949,
+      "learning_rate": 1.7946902970472297e-05,
+      "loss": 0.8083,
+      "step": 4445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 39.60778633257366,
+      "learning_rate": 1.7945879187830857e-05,
+      "loss": 0.945,
+      "step": 4446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 46.589752184382164,
+      "learning_rate": 1.794485517921326e-05,
+      "loss": 0.8961,
+      "step": 4447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 32.388571358042654,
+      "learning_rate": 1.794383094464863e-05,
+      "loss": 0.7125,
+      "step": 4448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 30.041809539790272,
+      "learning_rate": 1.7942806484166096e-05,
+      "loss": 0.8621,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 54.55900686646639,
+      "learning_rate": 1.794178179779479e-05,
+      "loss": 0.8112,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.307472536710925,
+      "learning_rate": 1.7940756885563855e-05,
+      "loss": 0.8212,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.81422113384207,
+      "learning_rate": 1.793973174750244e-05,
+      "loss": 0.8279,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 37.71375939476207,
+      "learning_rate": 1.79387063836397e-05,
+      "loss": 0.8721,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 34.67907758306626,
+      "learning_rate": 1.7937680794004793e-05,
+      "loss": 0.7828,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.93059123831946,
+      "learning_rate": 1.793665497862689e-05,
+      "loss": 0.8146,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 43.43934714142867,
+      "learning_rate": 1.793562893753516e-05,
+      "loss": 0.9712,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.537111786410584,
+      "learning_rate": 1.7934602670758786e-05,
+      "loss": 0.8467,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 41.65053036879858,
+      "learning_rate": 1.7933576178326952e-05,
+      "loss": 0.7824,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 31.998457390886003,
+      "learning_rate": 1.7932549460268856e-05,
+      "loss": 0.7458,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.89957803859208,
+      "learning_rate": 1.7931522516613688e-05,
+      "loss": 0.7593,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 29.42953314967296,
+      "learning_rate": 1.7930495347390665e-05,
+      "loss": 0.78,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 29.631208799309803,
+      "learning_rate": 1.792946795262899e-05,
+      "loss": 0.7346,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.92974627211905,
+      "learning_rate": 1.7928440332357885e-05,
+      "loss": 0.9288,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 30.317508169742293,
+      "learning_rate": 1.7927412486606578e-05,
+      "loss": 0.8189,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 37.79991687079276,
+      "learning_rate": 1.7926384415404298e-05,
+      "loss": 0.8514,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.036141804677534,
+      "learning_rate": 1.792535611878028e-05,
+      "loss": 0.8878,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 47.252358544234156,
+      "learning_rate": 1.792432759676377e-05,
+      "loss": 0.9395,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 50.524432716223885,
+      "learning_rate": 1.792329884938402e-05,
+      "loss": 0.9638,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.14700510795713,
+      "learning_rate": 1.7922269876670282e-05,
+      "loss": 0.8762,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.86760874555552,
+      "learning_rate": 1.7921240678651823e-05,
+      "loss": 0.8632,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 31.307921800167023,
+      "learning_rate": 1.7920211255357918e-05,
+      "loss": 0.8283,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 42.5254135107246,
+      "learning_rate": 1.7919181606817836e-05,
+      "loss": 0.8756,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 34.24400801696322,
+      "learning_rate": 1.791815173306086e-05,
+      "loss": 0.7652,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.004398479623745,
+      "learning_rate": 1.7917121634116284e-05,
+      "loss": 0.7947,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.269743746238824,
+      "learning_rate": 1.7916091310013396e-05,
+      "loss": 0.7745,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.13296750327934,
+      "learning_rate": 1.7915060760781505e-05,
+      "loss": 0.7976,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.40453411981753,
+      "learning_rate": 1.7914029986449916e-05,
+      "loss": 0.8288,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 48.23468180102567,
+      "learning_rate": 1.791299898704794e-05,
+      "loss": 0.817,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.93206436928158,
+      "learning_rate": 1.7911967762604905e-05,
+      "loss": 0.7884,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 30.318653186115863,
+      "learning_rate": 1.7910936313150137e-05,
+      "loss": 0.8857,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 28.44449879384271,
+      "learning_rate": 1.7909904638712963e-05,
+      "loss": 0.804,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 37.33018073468687,
+      "learning_rate": 1.790887273932273e-05,
+      "loss": 0.8688,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 34.34736729143187,
+      "learning_rate": 1.7907840615008787e-05,
+      "loss": 0.6752,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 37.134783075535495,
+      "learning_rate": 1.790680826580048e-05,
+      "loss": 0.8133,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.480754220413324,
+      "learning_rate": 1.7905775691727168e-05,
+      "loss": 0.742,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 45.30382934746926,
+      "learning_rate": 1.7904742892818225e-05,
+      "loss": 0.7743,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 31.473603363956027,
+      "learning_rate": 1.7903709869103018e-05,
+      "loss": 0.7722,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.40309404231392,
+      "learning_rate": 1.7902676620610925e-05,
+      "loss": 0.8136,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 47.369395737324176,
+      "learning_rate": 1.7901643147371328e-05,
+      "loss": 0.8781,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 38.26692966341454,
+      "learning_rate": 1.7900609449413626e-05,
+      "loss": 0.7815,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 54.523563733045634,
+      "learning_rate": 1.7899575526767214e-05,
+      "loss": 1.042,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 48.970313548119734,
+      "learning_rate": 1.7898541379461494e-05,
+      "loss": 0.7316,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 36.55284956379626,
+      "learning_rate": 1.7897507007525873e-05,
+      "loss": 0.8872,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.82285433756532,
+      "learning_rate": 1.789647241098978e-05,
+      "loss": 0.7503,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 30.34767122006842,
+      "learning_rate": 1.7895437589882627e-05,
+      "loss": 0.8116,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 38.33455413380948,
+      "learning_rate": 1.7894402544233846e-05,
+      "loss": 0.8199,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 37.18703433474529,
+      "learning_rate": 1.789336727407288e-05,
+      "loss": 0.8164,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 42.252109325393356,
+      "learning_rate": 1.7892331779429164e-05,
+      "loss": 0.8488,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 72.23228207641637,
+      "learning_rate": 1.789129606033215e-05,
+      "loss": 0.7194,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 44.39266289582022,
+      "learning_rate": 1.789026011681129e-05,
+      "loss": 0.8787,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 41.98051057215192,
+      "learning_rate": 1.7889223948896046e-05,
+      "loss": 0.8763,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 38.59914860515292,
+      "learning_rate": 1.788818755661589e-05,
+      "loss": 0.8389,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 31.979436404595166,
+      "learning_rate": 1.7887150940000296e-05,
+      "loss": 0.818,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 31.76558228586968,
+      "learning_rate": 1.788611409907874e-05,
+      "loss": 0.8788,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 41.13160547005193,
+      "learning_rate": 1.7885077033880714e-05,
+      "loss": 0.9171,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 31.501921882734525,
+      "learning_rate": 1.788403974443571e-05,
+      "loss": 0.8279,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 36.24133310957579,
+      "learning_rate": 1.788300223077323e-05,
+      "loss": 0.8052,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 40.07942894195668,
+      "learning_rate": 1.7881964492922774e-05,
+      "loss": 0.8905,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 35.74850948934569,
+      "learning_rate": 1.7880926530913863e-05,
+      "loss": 0.7479,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 41.66584888450459,
+      "learning_rate": 1.7879888344776008e-05,
+      "loss": 0.8923,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 33.198972475570464,
+      "learning_rate": 1.787884993453874e-05,
+      "loss": 0.839,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.73244527735356,
+      "learning_rate": 1.7877811300231587e-05,
+      "loss": 0.8745,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 44.309430049660534,
+      "learning_rate": 1.7876772441884093e-05,
+      "loss": 0.8019,
+      "step": 4513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 38.45801873283341,
+      "learning_rate": 1.7875733359525793e-05,
+      "loss": 0.7494,
+      "step": 4514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 32.28626270648248,
+      "learning_rate": 1.7874694053186246e-05,
+      "loss": 0.8473,
+      "step": 4515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 32.26481265322391,
+      "learning_rate": 1.7873654522895006e-05,
+      "loss": 0.8394,
+      "step": 4516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 38.887041152851495,
+      "learning_rate": 1.7872614768681638e-05,
+      "loss": 0.8155,
+      "step": 4517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 47.106813599650394,
+      "learning_rate": 1.7871574790575713e-05,
+      "loss": 0.7886,
+      "step": 4518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 33.29150036486617,
+      "learning_rate": 1.7870534588606804e-05,
+      "loss": 0.8225,
+      "step": 4519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 46.399173014563466,
+      "learning_rate": 1.7869494162804492e-05,
+      "loss": 0.8746,
+      "step": 4520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 30.698514938781056,
+      "learning_rate": 1.7868453513198375e-05,
+      "loss": 0.8431,
+      "step": 4521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.90649098580204,
+      "learning_rate": 1.786741263981804e-05,
+      "loss": 0.902,
+      "step": 4522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 48.6027902884355,
+      "learning_rate": 1.7866371542693094e-05,
+      "loss": 0.8981,
+      "step": 4523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 31.810966495958073,
+      "learning_rate": 1.786533022185314e-05,
+      "loss": 0.8259,
+      "step": 4524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 32.89640528612709,
+      "learning_rate": 1.78642886773278e-05,
+      "loss": 0.8197,
+      "step": 4525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 31.96314241780406,
+      "learning_rate": 1.7863246909146688e-05,
+      "loss": 0.8594,
+      "step": 4526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 38.494730520666515,
+      "learning_rate": 1.786220491733943e-05,
+      "loss": 0.7587,
+      "step": 4527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 36.61975341297174,
+      "learning_rate": 1.7861162701935672e-05,
+      "loss": 0.8286,
+      "step": 4528
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 46.784462858579666,
+      "learning_rate": 1.7860120262965038e-05,
+      "loss": 0.8838,
+      "step": 4529
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.069798316848804,
+      "learning_rate": 1.7859077600457184e-05,
+      "loss": 0.7248,
+      "step": 4530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 31.952385656481876,
+      "learning_rate": 1.7858034714441764e-05,
+      "loss": 0.7404,
+      "step": 4531
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 43.22407474663792,
+      "learning_rate": 1.7856991604948428e-05,
+      "loss": 0.846,
+      "step": 4532
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 44.577874868542246,
+      "learning_rate": 1.7855948272006848e-05,
+      "loss": 0.8039,
+      "step": 4533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 31.985337254557688,
+      "learning_rate": 1.7854904715646697e-05,
+      "loss": 0.7222,
+      "step": 4534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.98421749555908,
+      "learning_rate": 1.785386093589765e-05,
+      "loss": 0.8298,
+      "step": 4535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.53323509690004,
+      "learning_rate": 1.7852816932789392e-05,
+      "loss": 0.8512,
+      "step": 4536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 36.775113387098365,
+      "learning_rate": 1.7851772706351614e-05,
+      "loss": 0.7102,
+      "step": 4537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 37.33081344561271,
+      "learning_rate": 1.785072825661401e-05,
+      "loss": 0.7863,
+      "step": 4538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.340285852390146,
+      "learning_rate": 1.7849683583606293e-05,
+      "loss": 0.7616,
+      "step": 4539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 45.80047068541428,
+      "learning_rate": 1.7848638687358164e-05,
+      "loss": 0.8332,
+      "step": 4540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 40.694428678725785,
+      "learning_rate": 1.7847593567899344e-05,
+      "loss": 0.8181,
+      "step": 4541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 56.102896766536205,
+      "learning_rate": 1.7846548225259553e-05,
+      "loss": 0.8988,
+      "step": 4542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 28.777940491503124,
+      "learning_rate": 1.7845502659468522e-05,
+      "loss": 0.7274,
+      "step": 4543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 36.379558415252085,
+      "learning_rate": 1.784445687055598e-05,
+      "loss": 0.7966,
+      "step": 4544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 30.080694242456097,
+      "learning_rate": 1.7843410858551674e-05,
+      "loss": 0.7556,
+      "step": 4545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 40.84725945827103,
+      "learning_rate": 1.7842364623485356e-05,
+      "loss": 0.7942,
+      "step": 4546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 43.20295127571593,
+      "learning_rate": 1.784131816538677e-05,
+      "loss": 0.7821,
+      "step": 4547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 43.337280274640946,
+      "learning_rate": 1.7840271484285687e-05,
+      "loss": 0.8767,
+      "step": 4548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 34.67191227708474,
+      "learning_rate": 1.783922458021187e-05,
+      "loss": 0.8836,
+      "step": 4549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 40.24810406265272,
+      "learning_rate": 1.783817745319509e-05,
+      "loss": 0.9156,
+      "step": 4550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 28.533603178420204,
+      "learning_rate": 1.783713010326513e-05,
+      "loss": 0.7626,
+      "step": 4551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 37.84449536119519,
+      "learning_rate": 1.7836082530451772e-05,
+      "loss": 0.8264,
+      "step": 4552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 39.714263317451255,
+      "learning_rate": 1.783503473478481e-05,
+      "loss": 0.8246,
+      "step": 4553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 47.72754209471317,
+      "learning_rate": 1.7833986716294046e-05,
+      "loss": 0.9071,
+      "step": 4554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 50.413142216131924,
+      "learning_rate": 1.783293847500928e-05,
+      "loss": 0.9078,
+      "step": 4555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 38.132807935805495,
+      "learning_rate": 1.783189001096033e-05,
+      "loss": 0.8205,
+      "step": 4556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 42.6883461087823,
+      "learning_rate": 1.7830841324177006e-05,
+      "loss": 0.8781,
+      "step": 4557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 37.35519081976875,
+      "learning_rate": 1.7829792414689135e-05,
+      "loss": 0.8545,
+      "step": 4558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 37.12266193438579,
+      "learning_rate": 1.782874328252655e-05,
+      "loss": 0.896,
+      "step": 4559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 54.494955682481866,
+      "learning_rate": 1.7827693927719086e-05,
+      "loss": 0.7548,
+      "step": 4560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 29.76055998473373,
+      "learning_rate": 1.7826644350296588e-05,
+      "loss": 0.7749,
+      "step": 4561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 48.51615116245637,
+      "learning_rate": 1.7825594550288898e-05,
+      "loss": 0.7932,
+      "step": 4562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 31.76835080979871,
+      "learning_rate": 1.782454452772588e-05,
+      "loss": 0.7666,
+      "step": 4563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 40.18717559220769,
+      "learning_rate": 1.782349428263739e-05,
+      "loss": 0.809,
+      "step": 4564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 41.189983049950904,
+      "learning_rate": 1.7822443815053305e-05,
+      "loss": 0.7893,
+      "step": 4565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 30.632033137997322,
+      "learning_rate": 1.782139312500349e-05,
+      "loss": 0.6796,
+      "step": 4566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 40.89786290855438,
+      "learning_rate": 1.7820342212517826e-05,
+      "loss": 0.7711,
+      "step": 4567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 44.61076457035631,
+      "learning_rate": 1.7819291077626212e-05,
+      "loss": 0.8119,
+      "step": 4568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 36.930173544270666,
+      "learning_rate": 1.781823972035853e-05,
+      "loss": 0.768,
+      "step": 4569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 41.38516661328961,
+      "learning_rate": 1.7817188140744682e-05,
+      "loss": 0.8262,
+      "step": 4570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 50.24107284389005,
+      "learning_rate": 1.7816136338814577e-05,
+      "loss": 0.9111,
+      "step": 4571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 40.977002823521204,
+      "learning_rate": 1.781508431459813e-05,
+      "loss": 0.8088,
+      "step": 4572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 37.502587343603366,
+      "learning_rate": 1.7814032068125254e-05,
+      "loss": 0.8518,
+      "step": 4573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 35.45537727770656,
+      "learning_rate": 1.781297959942588e-05,
+      "loss": 0.7603,
+      "step": 4574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 34.01891899368275,
+      "learning_rate": 1.7811926908529934e-05,
+      "loss": 0.9095,
+      "step": 4575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 48.54802991993028,
+      "learning_rate": 1.781087399546736e-05,
+      "loss": 0.8515,
+      "step": 4576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 42.68886626529993,
+      "learning_rate": 1.7809820860268096e-05,
+      "loss": 0.9588,
+      "step": 4577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 33.86765984686512,
+      "learning_rate": 1.7808767502962094e-05,
+      "loss": 0.7788,
+      "step": 4578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 37.295465091925514,
+      "learning_rate": 1.7807713923579313e-05,
+      "loss": 0.9595,
+      "step": 4579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.47123166001738,
+      "learning_rate": 1.7806660122149714e-05,
+      "loss": 0.7735,
+      "step": 4580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.92395688495697,
+      "learning_rate": 1.7805606098703273e-05,
+      "loss": 0.8038,
+      "step": 4581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 44.98799021253211,
+      "learning_rate": 1.7804551853269957e-05,
+      "loss": 0.9557,
+      "step": 4582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 47.390147234743715,
+      "learning_rate": 1.7803497385879752e-05,
+      "loss": 0.6888,
+      "step": 4583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 38.65466945257189,
+      "learning_rate": 1.7802442696562646e-05,
+      "loss": 0.7259,
+      "step": 4584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.89486025315833,
+      "learning_rate": 1.7801387785348637e-05,
+      "loss": 0.7669,
+      "step": 4585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.764076004052484,
+      "learning_rate": 1.780033265226772e-05,
+      "loss": 0.78,
+      "step": 4586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 44.27797331494932,
+      "learning_rate": 1.7799277297349908e-05,
+      "loss": 0.9415,
+      "step": 4587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 29.684504203222044,
+      "learning_rate": 1.7798221720625205e-05,
+      "loss": 0.7376,
+      "step": 4588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.72075733632768,
+      "learning_rate": 1.779716592212365e-05,
+      "loss": 0.7781,
+      "step": 4589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.45041337787776,
+      "learning_rate": 1.7796109901875246e-05,
+      "loss": 0.8653,
+      "step": 4590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 31.57293040846766,
+      "learning_rate": 1.7795053659910044e-05,
+      "loss": 0.8545,
+      "step": 4591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.39634979631655,
+      "learning_rate": 1.7793997196258074e-05,
+      "loss": 0.7533,
+      "step": 4592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 38.584893662984754,
+      "learning_rate": 1.7792940510949383e-05,
+      "loss": 0.8146,
+      "step": 4593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 45.96450057833248,
+      "learning_rate": 1.7791883604014023e-05,
+      "loss": 0.8946,
+      "step": 4594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 29.703182824413155,
+      "learning_rate": 1.7790826475482047e-05,
+      "loss": 0.7865,
+      "step": 4595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.22436547366476,
+      "learning_rate": 1.7789769125383528e-05,
+      "loss": 0.8919,
+      "step": 4596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 33.321121953722574,
+      "learning_rate": 1.7788711553748533e-05,
+      "loss": 0.8656,
+      "step": 4597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 42.791749166070495,
+      "learning_rate": 1.7787653760607134e-05,
+      "loss": 0.7512,
+      "step": 4598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 27.76891794679808,
+      "learning_rate": 1.778659574598942e-05,
+      "loss": 0.7162,
+      "step": 4599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.2067654517167,
+      "learning_rate": 1.7785537509925478e-05,
+      "loss": 0.8042,
+      "step": 4600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 36.643649967479334,
+      "learning_rate": 1.77844790524454e-05,
+      "loss": 0.8993,
+      "step": 4601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.918942763106564,
+      "learning_rate": 1.77834203735793e-05,
+      "loss": 0.8329,
+      "step": 4602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 49.97448784363947,
+      "learning_rate": 1.778236147335727e-05,
+      "loss": 0.8558,
+      "step": 4603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 46.4721728246279,
+      "learning_rate": 1.7781302351809437e-05,
+      "loss": 0.8531,
+      "step": 4604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 44.09373912094706,
+      "learning_rate": 1.7780243008965915e-05,
+      "loss": 0.8115,
+      "step": 4605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.69885422741556,
+      "learning_rate": 1.7779183444856833e-05,
+      "loss": 0.8246,
+      "step": 4606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 46.871316856643624,
+      "learning_rate": 1.7778123659512326e-05,
+      "loss": 0.8726,
+      "step": 4607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 29.24534346528984,
+      "learning_rate": 1.777706365296253e-05,
+      "loss": 0.7914,
+      "step": 4608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.00668358697844,
+      "learning_rate": 1.7776003425237592e-05,
+      "loss": 0.7798,
+      "step": 4609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.119997143387536,
+      "learning_rate": 1.7774942976367668e-05,
+      "loss": 0.8016,
+      "step": 4610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.66169049734333,
+      "learning_rate": 1.7773882306382913e-05,
+      "loss": 0.7563,
+      "step": 4611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 46.35623986340654,
+      "learning_rate": 1.7772821415313493e-05,
+      "loss": 0.9658,
+      "step": 4612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.014571656426316,
+      "learning_rate": 1.777176030318958e-05,
+      "loss": 0.7202,
+      "step": 4613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 33.51149084029472,
+      "learning_rate": 1.777069897004135e-05,
+      "loss": 0.8768,
+      "step": 4614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.90394797054705,
+      "learning_rate": 1.7769637415898982e-05,
+      "loss": 0.7635,
+      "step": 4615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 40.513408654700406,
+      "learning_rate": 1.776857564079268e-05,
+      "loss": 0.8102,
+      "step": 4616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.20500570373888,
+      "learning_rate": 1.7767513644752624e-05,
+      "loss": 0.8236,
+      "step": 4617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 42.1720075177672,
+      "learning_rate": 1.7766451427809026e-05,
+      "loss": 0.8035,
+      "step": 4618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 36.28571094356303,
+      "learning_rate": 1.7765388989992093e-05,
+      "loss": 0.9117,
+      "step": 4619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.25731704434649,
+      "learning_rate": 1.776432633133204e-05,
+      "loss": 0.9251,
+      "step": 4620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 38.17502358966091,
+      "learning_rate": 1.776326345185909e-05,
+      "loss": 0.8585,
+      "step": 4621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 31.49195594106272,
+      "learning_rate": 1.7762200351603465e-05,
+      "loss": 0.8891,
+      "step": 4622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.74855260621493,
+      "learning_rate": 1.7761137030595406e-05,
+      "loss": 0.8207,
+      "step": 4623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.28233829192871,
+      "learning_rate": 1.776007348886515e-05,
+      "loss": 0.8638,
+      "step": 4624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.01625138960224,
+      "learning_rate": 1.775900972644294e-05,
+      "loss": 0.8252,
+      "step": 4625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.61551675226949,
+      "learning_rate": 1.7757945743359033e-05,
+      "loss": 0.8366,
+      "step": 4626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.585419369408264,
+      "learning_rate": 1.775688153964369e-05,
+      "loss": 0.8342,
+      "step": 4627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 36.37235726263671,
+      "learning_rate": 1.775581711532717e-05,
+      "loss": 0.8884,
+      "step": 4628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.163724988788516,
+      "learning_rate": 1.7754752470439755e-05,
+      "loss": 0.82,
+      "step": 4629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.53467521261489,
+      "learning_rate": 1.7753687605011707e-05,
+      "loss": 0.7316,
+      "step": 4630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 30.886326692928968,
+      "learning_rate": 1.7752622519073327e-05,
+      "loss": 0.7926,
+      "step": 4631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 36.80125664218796,
+      "learning_rate": 1.7751557212654893e-05,
+      "loss": 0.878,
+      "step": 4632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 54.578227089197505,
+      "learning_rate": 1.7750491685786708e-05,
+      "loss": 0.8557,
+      "step": 4633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 39.88945660970042,
+      "learning_rate": 1.7749425938499073e-05,
+      "loss": 0.7267,
+      "step": 4634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 35.50282207877208,
+      "learning_rate": 1.77483599708223e-05,
+      "loss": 0.8641,
+      "step": 4635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 36.6953494433387,
+      "learning_rate": 1.7747293782786697e-05,
+      "loss": 0.8037,
+      "step": 4636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.95412880518087,
+      "learning_rate": 1.7746227374422597e-05,
+      "loss": 0.7838,
+      "step": 4637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 44.13597625829035,
+      "learning_rate": 1.7745160745760317e-05,
+      "loss": 0.9124,
+      "step": 4638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 39.781152927373135,
+      "learning_rate": 1.7744093896830196e-05,
+      "loss": 0.8535,
+      "step": 4639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 38.984799626241426,
+      "learning_rate": 1.7743026827662578e-05,
+      "loss": 0.8814,
+      "step": 4640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 38.76545245413952,
+      "learning_rate": 1.7741959538287807e-05,
+      "loss": 0.878,
+      "step": 4641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 35.94961307553512,
+      "learning_rate": 1.7740892028736233e-05,
+      "loss": 0.7788,
+      "step": 4642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 38.674729767824985,
+      "learning_rate": 1.7739824299038217e-05,
+      "loss": 0.8418,
+      "step": 4643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 37.177048774091084,
+      "learning_rate": 1.773875634922413e-05,
+      "loss": 0.8228,
+      "step": 4644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 35.03897678530617,
+      "learning_rate": 1.7737688179324334e-05,
+      "loss": 0.8406,
+      "step": 4645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 41.88665037781915,
+      "learning_rate": 1.773661978936922e-05,
+      "loss": 0.8071,
+      "step": 4646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.10344779452163,
+      "learning_rate": 1.7735551179389157e-05,
+      "loss": 0.7856,
+      "step": 4647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.505037465404875,
+      "learning_rate": 1.7734482349414547e-05,
+      "loss": 0.8875,
+      "step": 4648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.1888056232033,
+      "learning_rate": 1.7733413299475787e-05,
+      "loss": 0.8782,
+      "step": 4649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 38.80570159563931,
+      "learning_rate": 1.7732344029603273e-05,
+      "loss": 0.8268,
+      "step": 4650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 37.97610226312851,
+      "learning_rate": 1.773127453982742e-05,
+      "loss": 0.8921,
+      "step": 4651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 30.308088744575684,
+      "learning_rate": 1.7730204830178638e-05,
+      "loss": 0.6744,
+      "step": 4652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 42.39891437454231,
+      "learning_rate": 1.7729134900687354e-05,
+      "loss": 0.785,
+      "step": 4653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 35.95358554589878,
+      "learning_rate": 1.7728064751383997e-05,
+      "loss": 0.7974,
+      "step": 4654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.67069730273437,
+      "learning_rate": 1.7726994382299e-05,
+      "loss": 0.7792,
+      "step": 4655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 33.21895836148448,
+      "learning_rate": 1.7725923793462798e-05,
+      "loss": 0.7682,
+      "step": 4656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 33.297796371794156,
+      "learning_rate": 1.7724852984905844e-05,
+      "loss": 0.8098,
+      "step": 4657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 55.39982916327071,
+      "learning_rate": 1.7723781956658593e-05,
+      "loss": 0.7694,
+      "step": 4658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 45.261043843473544,
+      "learning_rate": 1.77227107087515e-05,
+      "loss": 0.8063,
+      "step": 4659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 39.04628940370605,
+      "learning_rate": 1.7721639241215034e-05,
+      "loss": 0.7218,
+      "step": 4660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 37.36917961691583,
+      "learning_rate": 1.772056755407966e-05,
+      "loss": 0.8178,
+      "step": 4661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.07709556224428,
+      "learning_rate": 1.771949564737587e-05,
+      "loss": 0.8056,
+      "step": 4662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.21959195299535,
+      "learning_rate": 1.771842352113413e-05,
+      "loss": 0.8941,
+      "step": 4663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.78002518024013,
+      "learning_rate": 1.7717351175384948e-05,
+      "loss": 0.8465,
+      "step": 4664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.42703574096037,
+      "learning_rate": 1.7716278610158806e-05,
+      "loss": 0.8171,
+      "step": 4665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 56.30613442831577,
+      "learning_rate": 1.7715205825486218e-05,
+      "loss": 0.8765,
+      "step": 4666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.18574273929956,
+      "learning_rate": 1.771413282139769e-05,
+      "loss": 0.7579,
+      "step": 4667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 48.59568861255925,
+      "learning_rate": 1.7713059597923735e-05,
+      "loss": 0.8942,
+      "step": 4668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 39.82883439867067,
+      "learning_rate": 1.771198615509488e-05,
+      "loss": 0.7521,
+      "step": 4669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 28.74034480632029,
+      "learning_rate": 1.771091249294165e-05,
+      "loss": 0.8329,
+      "step": 4670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 43.204138207583355,
+      "learning_rate": 1.7709838611494577e-05,
+      "loss": 0.8802,
+      "step": 4671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 31.090141102288737,
+      "learning_rate": 1.7708764510784205e-05,
+      "loss": 0.7796,
+      "step": 4672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 31.769809321592852,
+      "learning_rate": 1.770769019084108e-05,
+      "loss": 0.8245,
+      "step": 4673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 33.9862086430612,
+      "learning_rate": 1.7706615651695754e-05,
+      "loss": 0.8613,
+      "step": 4674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 28.16933840621024,
+      "learning_rate": 1.7705540893378787e-05,
+      "loss": 0.7902,
+      "step": 4675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 35.128804445410836,
+      "learning_rate": 1.7704465915920745e-05,
+      "loss": 0.8727,
+      "step": 4676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.34693318757914,
+      "learning_rate": 1.77033907193522e-05,
+      "loss": 0.8266,
+      "step": 4677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 37.98131086637444,
+      "learning_rate": 1.770231530370373e-05,
+      "loss": 0.9564,
+      "step": 4678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.37092544335839,
+      "learning_rate": 1.7701239669005916e-05,
+      "loss": 0.9198,
+      "step": 4679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 26.361817288048535,
+      "learning_rate": 1.7700163815289353e-05,
+      "loss": 0.7786,
+      "step": 4680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 38.44160042577947,
+      "learning_rate": 1.7699087742584636e-05,
+      "loss": 0.8251,
+      "step": 4681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.11793322144014,
+      "learning_rate": 1.769801145092237e-05,
+      "loss": 0.883,
+      "step": 4682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.90151633866146,
+      "learning_rate": 1.7696934940333155e-05,
+      "loss": 0.8126,
+      "step": 4683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 106.6205876590131,
+      "learning_rate": 1.7695858210847615e-05,
+      "loss": 0.9946,
+      "step": 4684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 38.212725603550126,
+      "learning_rate": 1.769478126249637e-05,
+      "loss": 0.7581,
+      "step": 4685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.248786779481605,
+      "learning_rate": 1.769370409531005e-05,
+      "loss": 0.8289,
+      "step": 4686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.27281294828197,
+      "learning_rate": 1.7692626709319283e-05,
+      "loss": 0.7786,
+      "step": 4687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.61949211340322,
+      "learning_rate": 1.769154910455471e-05,
+      "loss": 0.7701,
+      "step": 4688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.98002915500381,
+      "learning_rate": 1.7690471281046982e-05,
+      "loss": 0.7994,
+      "step": 4689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.41305680556562,
+      "learning_rate": 1.768939323882675e-05,
+      "loss": 0.8136,
+      "step": 4690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 47.15857803786584,
+      "learning_rate": 1.7688314977924674e-05,
+      "loss": 0.7763,
+      "step": 4691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.39005865800432,
+      "learning_rate": 1.768723649837142e-05,
+      "loss": 0.7874,
+      "step": 4692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.65175045353817,
+      "learning_rate": 1.7686157800197652e-05,
+      "loss": 0.7694,
+      "step": 4693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 44.41141601300414,
+      "learning_rate": 1.7685078883434054e-05,
+      "loss": 1.0579,
+      "step": 4694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 46.19899765636501,
+      "learning_rate": 1.7683999748111306e-05,
+      "loss": 0.7866,
+      "step": 4695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.59491889448439,
+      "learning_rate": 1.7682920394260102e-05,
+      "loss": 0.7948,
+      "step": 4696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.17775814518374,
+      "learning_rate": 1.768184082191114e-05,
+      "loss": 0.8491,
+      "step": 4697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 38.1945668875533,
+      "learning_rate": 1.7680761031095113e-05,
+      "loss": 0.9136,
+      "step": 4698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.843823807174665,
+      "learning_rate": 1.7679681021842738e-05,
+      "loss": 0.8476,
+      "step": 4699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.06750964593941,
+      "learning_rate": 1.767860079418473e-05,
+      "loss": 0.7884,
+      "step": 4700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 31.1762843634519,
+      "learning_rate": 1.7677520348151805e-05,
+      "loss": 0.7977,
+      "step": 4701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.78188504093971,
+      "learning_rate": 1.767643968377469e-05,
+      "loss": 0.8756,
+      "step": 4702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 40.52061267521392,
+      "learning_rate": 1.7675358801084122e-05,
+      "loss": 0.9173,
+      "step": 4703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 38.17087462742625,
+      "learning_rate": 1.767427770011084e-05,
+      "loss": 0.9042,
+      "step": 4704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 39.07964904326922,
+      "learning_rate": 1.767319638088559e-05,
+      "loss": 0.8539,
+      "step": 4705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 36.37452119007445,
+      "learning_rate": 1.7672114843439127e-05,
+      "loss": 0.9836,
+      "step": 4706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 40.430395289320785,
+      "learning_rate": 1.76710330878022e-05,
+      "loss": 0.9036,
+      "step": 4707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 50.380285047350924,
+      "learning_rate": 1.7669951114005584e-05,
+      "loss": 0.7786,
+      "step": 4708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 29.28834063926821,
+      "learning_rate": 1.766886892208004e-05,
+      "loss": 0.7443,
+      "step": 4709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 41.104424341279966,
+      "learning_rate": 1.7667786512056352e-05,
+      "loss": 0.8526,
+      "step": 4710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.51404121326082,
+      "learning_rate": 1.7666703883965305e-05,
+      "loss": 0.8747,
+      "step": 4711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.21337321292179,
+      "learning_rate": 1.766562103783768e-05,
+      "loss": 0.832,
+      "step": 4712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 47.25829103176803,
+      "learning_rate": 1.766453797370428e-05,
+      "loss": 0.8656,
+      "step": 4713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 26.524057644545262,
+      "learning_rate": 1.76634546915959e-05,
+      "loss": 0.817,
+      "step": 4714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 36.54613226975361,
+      "learning_rate": 1.7662371191543355e-05,
+      "loss": 0.7871,
+      "step": 4715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 35.355136386353095,
+      "learning_rate": 1.7661287473577457e-05,
+      "loss": 0.7285,
+      "step": 4716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 62.52472563130101,
+      "learning_rate": 1.766020353772902e-05,
+      "loss": 0.7729,
+      "step": 4717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 38.18279580143498,
+      "learning_rate": 1.765911938402888e-05,
+      "loss": 0.7867,
+      "step": 4718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 58.367980485976666,
+      "learning_rate": 1.7658035012507863e-05,
+      "loss": 0.8862,
+      "step": 4719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 45.13483430633609,
+      "learning_rate": 1.7656950423196807e-05,
+      "loss": 0.8995,
+      "step": 4720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 39.749240184215765,
+      "learning_rate": 1.7655865616126564e-05,
+      "loss": 0.883,
+      "step": 4721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 40.09551271412619,
+      "learning_rate": 1.7654780591327983e-05,
+      "loss": 0.7159,
+      "step": 4722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 44.840355423633994,
+      "learning_rate": 1.7653695348831915e-05,
+      "loss": 0.8607,
+      "step": 4723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 47.632162128780415,
+      "learning_rate": 1.7652609888669234e-05,
+      "loss": 0.8998,
+      "step": 4724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 44.756862529754976,
+      "learning_rate": 1.76515242108708e-05,
+      "loss": 0.8563,
+      "step": 4725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 33.05557153863458,
+      "learning_rate": 1.7650438315467494e-05,
+      "loss": 0.8338,
+      "step": 4726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.07890021252679,
+      "learning_rate": 1.7649352202490198e-05,
+      "loss": 0.8163,
+      "step": 4727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 36.94095495044535,
+      "learning_rate": 1.7648265871969803e-05,
+      "loss": 0.8128,
+      "step": 4728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 32.844899038582696,
+      "learning_rate": 1.76471793239372e-05,
+      "loss": 0.8142,
+      "step": 4729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 30.776594284311567,
+      "learning_rate": 1.7646092558423288e-05,
+      "loss": 0.812,
+      "step": 4730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 30.90739308047544,
+      "learning_rate": 1.7645005575458977e-05,
+      "loss": 0.8113,
+      "step": 4731
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 33.34450518531395,
+      "learning_rate": 1.764391837507518e-05,
+      "loss": 0.7146,
+      "step": 4732
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 33.946769442530695,
+      "learning_rate": 1.7642830957302815e-05,
+      "loss": 0.7724,
+      "step": 4733
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 46.181314552335316,
+      "learning_rate": 1.7641743322172812e-05,
+      "loss": 0.7583,
+      "step": 4734
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.87250544629357,
+      "learning_rate": 1.7640655469716096e-05,
+      "loss": 0.6967,
+      "step": 4735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 35.601844282544015,
+      "learning_rate": 1.7639567399963607e-05,
+      "loss": 0.8219,
+      "step": 4736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 39.53391428778435,
+      "learning_rate": 1.7638479112946294e-05,
+      "loss": 0.8757,
+      "step": 4737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 26.160395436178188,
+      "learning_rate": 1.7637390608695102e-05,
+      "loss": 0.7897,
+      "step": 4738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 42.84805739749949,
+      "learning_rate": 1.7636301887240987e-05,
+      "loss": 0.8444,
+      "step": 4739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 40.76174848086594,
+      "learning_rate": 1.7635212948614915e-05,
+      "loss": 0.7817,
+      "step": 4740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.971120056888296,
+      "learning_rate": 1.7634123792847854e-05,
+      "loss": 0.9167,
+      "step": 4741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 36.52625821405363,
+      "learning_rate": 1.7633034419970775e-05,
+      "loss": 0.7883,
+      "step": 4742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.48123181526948,
+      "learning_rate": 1.7631944830014663e-05,
+      "loss": 0.7952,
+      "step": 4743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.83993836792046,
+      "learning_rate": 1.763085502301051e-05,
+      "loss": 0.766,
+      "step": 4744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 35.51845715462624,
+      "learning_rate": 1.76297649989893e-05,
+      "loss": 0.7808,
+      "step": 4745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 29.236826302022944,
+      "learning_rate": 1.7628674757982037e-05,
+      "loss": 0.7147,
+      "step": 4746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 49.158776542692685,
+      "learning_rate": 1.7627584300019727e-05,
+      "loss": 0.8562,
+      "step": 4747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.008640263137984,
+      "learning_rate": 1.7626493625133377e-05,
+      "loss": 0.7748,
+      "step": 4748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.65726129279186,
+      "learning_rate": 1.7625402733354015e-05,
+      "loss": 0.8333,
+      "step": 4749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 38.4520610800099,
+      "learning_rate": 1.7624311624712657e-05,
+      "loss": 0.901,
+      "step": 4750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 39.211741017957394,
+      "learning_rate": 1.762322029924034e-05,
+      "loss": 0.8175,
+      "step": 4751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.52524918418639,
+      "learning_rate": 1.7622128756968095e-05,
+      "loss": 0.8208,
+      "step": 4752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 29.98775648776199,
+      "learning_rate": 1.762103699792697e-05,
+      "loss": 0.8159,
+      "step": 4753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.05248310246926,
+      "learning_rate": 1.7619945022148008e-05,
+      "loss": 0.8023,
+      "step": 4754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 47.12589750775963,
+      "learning_rate": 1.7618852829662264e-05,
+      "loss": 0.8848,
+      "step": 4755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 40.036886763841245,
+      "learning_rate": 1.7617760420500806e-05,
+      "loss": 0.7955,
+      "step": 4756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 41.453037317947505,
+      "learning_rate": 1.7616667794694697e-05,
+      "loss": 0.8794,
+      "step": 4757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.23914709220439,
+      "learning_rate": 1.7615574952275012e-05,
+      "loss": 0.8838,
+      "step": 4758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 42.11801812922503,
+      "learning_rate": 1.761448189327283e-05,
+      "loss": 0.7962,
+      "step": 4759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 41.87525598533162,
+      "learning_rate": 1.761338861771924e-05,
+      "loss": 0.8595,
+      "step": 4760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 38.6852620082984,
+      "learning_rate": 1.761229512564533e-05,
+      "loss": 0.8921,
+      "step": 4761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.02002354031211,
+      "learning_rate": 1.7611201417082196e-05,
+      "loss": 0.812,
+      "step": 4762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 49.81626961108361,
+      "learning_rate": 1.761010749206095e-05,
+      "loss": 0.8148,
+      "step": 4763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 39.34853821721003,
+      "learning_rate": 1.7609013350612696e-05,
+      "loss": 0.8482,
+      "step": 4764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 35.27826464332492,
+      "learning_rate": 1.7607918992768554e-05,
+      "loss": 0.7856,
+      "step": 4765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 46.35139840278788,
+      "learning_rate": 1.7606824418559648e-05,
+      "loss": 0.7731,
+      "step": 4766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 47.91553999723417,
+      "learning_rate": 1.7605729628017108e-05,
+      "loss": 1.0022,
+      "step": 4767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.52829168255526,
+      "learning_rate": 1.760463462117206e-05,
+      "loss": 0.7763,
+      "step": 4768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 38.936315944052076,
+      "learning_rate": 1.7603539398055658e-05,
+      "loss": 0.964,
+      "step": 4769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.08256757502181,
+      "learning_rate": 1.760244395869904e-05,
+      "loss": 0.8036,
+      "step": 4770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.38275892873419,
+      "learning_rate": 1.7601348303133364e-05,
+      "loss": 0.7153,
+      "step": 4771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 46.61933124024932,
+      "learning_rate": 1.760025243138979e-05,
+      "loss": 0.7796,
+      "step": 4772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.417312832505196,
+      "learning_rate": 1.7599156343499482e-05,
+      "loss": 0.7494,
+      "step": 4773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 42.564623366618335,
+      "learning_rate": 1.7598060039493613e-05,
+      "loss": 0.8248,
+      "step": 4774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 44.851721793395036,
+      "learning_rate": 1.759696351940336e-05,
+      "loss": 0.7183,
+      "step": 4775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.18057012021348,
+      "learning_rate": 1.759586678325991e-05,
+      "loss": 0.8012,
+      "step": 4776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 47.48153140763093,
+      "learning_rate": 1.7594769831094452e-05,
+      "loss": 0.8111,
+      "step": 4777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 32.262013434133394,
+      "learning_rate": 1.759367266293818e-05,
+      "loss": 0.8104,
+      "step": 4778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 38.03751325165176,
+      "learning_rate": 1.7592575278822304e-05,
+      "loss": 0.9996,
+      "step": 4779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 47.76299993197966,
+      "learning_rate": 1.7591477678778027e-05,
+      "loss": 0.8153,
+      "step": 4780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.47676416775597,
+      "learning_rate": 1.7590379862836565e-05,
+      "loss": 0.7888,
+      "step": 4781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.614168065343335,
+      "learning_rate": 1.758928183102914e-05,
+      "loss": 0.7699,
+      "step": 4782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.278664926389652,
+      "learning_rate": 1.7588183583386982e-05,
+      "loss": 0.7294,
+      "step": 4783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 59.325059145943364,
+      "learning_rate": 1.7587085119941318e-05,
+      "loss": 0.7447,
+      "step": 4784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.408788433208,
+      "learning_rate": 1.758598644072339e-05,
+      "loss": 0.8258,
+      "step": 4785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 43.18671420937898,
+      "learning_rate": 1.7584887545764452e-05,
+      "loss": 0.8411,
+      "step": 4786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.88548285302328,
+      "learning_rate": 1.7583788435095746e-05,
+      "loss": 0.9163,
+      "step": 4787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 42.180348367984344,
+      "learning_rate": 1.758268910874853e-05,
+      "loss": 0.7618,
+      "step": 4788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 43.81132485613591,
+      "learning_rate": 1.7581589566754076e-05,
+      "loss": 0.8661,
+      "step": 4789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.35362252244443,
+      "learning_rate": 1.7580489809143648e-05,
+      "loss": 0.9055,
+      "step": 4790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.326747669648995,
+      "learning_rate": 1.7579389835948525e-05,
+      "loss": 0.8065,
+      "step": 4791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.97973979880907,
+      "learning_rate": 1.757828964719999e-05,
+      "loss": 0.8068,
+      "step": 4792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 31.42210386962093,
+      "learning_rate": 1.7577189242929325e-05,
+      "loss": 0.8443,
+      "step": 4793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.983030223663388,
+      "learning_rate": 1.7576088623167838e-05,
+      "loss": 0.7344,
+      "step": 4794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.45301182774694,
+      "learning_rate": 1.7574987787946817e-05,
+      "loss": 0.9917,
+      "step": 4795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 40.12760709748163,
+      "learning_rate": 1.7573886737297575e-05,
+      "loss": 0.839,
+      "step": 4796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.98347456385385,
+      "learning_rate": 1.757278547125143e-05,
+      "loss": 0.7684,
+      "step": 4797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.18169533441535,
+      "learning_rate": 1.7571683989839693e-05,
+      "loss": 0.7411,
+      "step": 4798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 29.798461371742246,
+      "learning_rate": 1.757058229309369e-05,
+      "loss": 0.7694,
+      "step": 4799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 46.730760348464976,
+      "learning_rate": 1.7569480381044758e-05,
+      "loss": 0.756,
+      "step": 4800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.67268428937909,
+      "learning_rate": 1.756837825372423e-05,
+      "loss": 0.8194,
+      "step": 4801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.223307448342368,
+      "learning_rate": 1.7567275911163454e-05,
+      "loss": 0.7761,
+      "step": 4802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 44.10636888368814,
+      "learning_rate": 1.7566173353393778e-05,
+      "loss": 0.8367,
+      "step": 4803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 28.3690673028783,
+      "learning_rate": 1.7565070580446556e-05,
+      "loss": 0.7049,
+      "step": 4804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 35.69748552052936,
+      "learning_rate": 1.7563967592353152e-05,
+      "loss": 0.8208,
+      "step": 4805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 39.98910117680531,
+      "learning_rate": 1.7562864389144936e-05,
+      "loss": 0.8318,
+      "step": 4806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.612982575805717,
+      "learning_rate": 1.756176097085328e-05,
+      "loss": 0.7783,
+      "step": 4807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 42.611370206421235,
+      "learning_rate": 1.7560657337509567e-05,
+      "loss": 0.9096,
+      "step": 4808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 38.24853623774561,
+      "learning_rate": 1.755955348914518e-05,
+      "loss": 0.7967,
+      "step": 4809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 31.215385645216912,
+      "learning_rate": 1.7558449425791515e-05,
+      "loss": 0.6803,
+      "step": 4810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.75677857845137,
+      "learning_rate": 1.7557345147479968e-05,
+      "loss": 0.8485,
+      "step": 4811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 32.84176910698676,
+      "learning_rate": 1.755624065424195e-05,
+      "loss": 0.8138,
+      "step": 4812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.60059080090927,
+      "learning_rate": 1.7555135946108866e-05,
+      "loss": 0.8436,
+      "step": 4813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.731933786471814,
+      "learning_rate": 1.7554031023112134e-05,
+      "loss": 0.9565,
+      "step": 4814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.523992583688226,
+      "learning_rate": 1.755292588528318e-05,
+      "loss": 0.8149,
+      "step": 4815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 35.36210239155855,
+      "learning_rate": 1.755182053265343e-05,
+      "loss": 0.8636,
+      "step": 4816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 31.732224782713306,
+      "learning_rate": 1.7550714965254325e-05,
+      "loss": 0.7145,
+      "step": 4817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.374934655709694,
+      "learning_rate": 1.7549609183117305e-05,
+      "loss": 0.7748,
+      "step": 4818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 31.842656005395757,
+      "learning_rate": 1.7548503186273812e-05,
+      "loss": 0.7911,
+      "step": 4819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.366619052508646,
+      "learning_rate": 1.7547396974755307e-05,
+      "loss": 0.7925,
+      "step": 4820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 33.412212100288905,
+      "learning_rate": 1.7546290548593242e-05,
+      "loss": 0.7753,
+      "step": 4821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 33.29369643751118,
+      "learning_rate": 1.7545183907819094e-05,
+      "loss": 0.8214,
+      "step": 4822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 33.398559902033796,
+      "learning_rate": 1.7544077052464327e-05,
+      "loss": 0.7548,
+      "step": 4823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 27.91615583512297,
+      "learning_rate": 1.7542969982560424e-05,
+      "loss": 0.7323,
+      "step": 4824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.02845408336298,
+      "learning_rate": 1.754186269813886e-05,
+      "loss": 0.7293,
+      "step": 4825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.196384564732625,
+      "learning_rate": 1.7540755199231145e-05,
+      "loss": 0.9429,
+      "step": 4826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 46.24420618934783,
+      "learning_rate": 1.7539647485868753e-05,
+      "loss": 0.8172,
+      "step": 4827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 41.98043450964389,
+      "learning_rate": 1.75385395580832e-05,
+      "loss": 0.827,
+      "step": 4828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 45.31993475067579,
+      "learning_rate": 1.7537431415905995e-05,
+      "loss": 0.8761,
+      "step": 4829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.570403006249535,
+      "learning_rate": 1.7536323059368644e-05,
+      "loss": 0.8901,
+      "step": 4830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 35.26300519779449,
+      "learning_rate": 1.7535214488502677e-05,
+      "loss": 0.7563,
+      "step": 4831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 31.422223063115133,
+      "learning_rate": 1.753410570333962e-05,
+      "loss": 0.6196,
+      "step": 4832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 49.37625993014435,
+      "learning_rate": 1.7532996703911002e-05,
+      "loss": 0.7603,
+      "step": 4833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 28.915835289358988,
+      "learning_rate": 1.7531887490248364e-05,
+      "loss": 0.7383,
+      "step": 4834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.12493393310937,
+      "learning_rate": 1.7530778062383253e-05,
+      "loss": 0.8373,
+      "step": 4835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.18317691892754,
+      "learning_rate": 1.752966842034722e-05,
+      "loss": 0.7089,
+      "step": 4836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 33.82862633031915,
+      "learning_rate": 1.752855856417182e-05,
+      "loss": 0.8551,
+      "step": 4837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 62.605551408463334,
+      "learning_rate": 1.752744849388862e-05,
+      "loss": 0.8139,
+      "step": 4838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.17296623367938,
+      "learning_rate": 1.7526338209529184e-05,
+      "loss": 0.7525,
+      "step": 4839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.588740671426738,
+      "learning_rate": 1.7525227711125098e-05,
+      "loss": 0.7636,
+      "step": 4840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 44.87681372969888,
+      "learning_rate": 1.7524116998707933e-05,
+      "loss": 0.8372,
+      "step": 4841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 35.62217585428832,
+      "learning_rate": 1.7523006072309286e-05,
+      "loss": 0.8879,
+      "step": 4842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.86051827553697,
+      "learning_rate": 1.7521894931960742e-05,
+      "loss": 0.8641,
+      "step": 4843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 46.344415542854996,
+      "learning_rate": 1.7520783577693912e-05,
+      "loss": 0.8274,
+      "step": 4844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.745205101014825,
+      "learning_rate": 1.7519672009540394e-05,
+      "loss": 0.8568,
+      "step": 4845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 39.481083879578776,
+      "learning_rate": 1.7518560227531806e-05,
+      "loss": 0.7791,
+      "step": 4846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 42.23976891692848,
+      "learning_rate": 1.7517448231699758e-05,
+      "loss": 0.7767,
+      "step": 4847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 48.945593046585365,
+      "learning_rate": 1.7516336022075883e-05,
+      "loss": 0.8919,
+      "step": 4848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 60.55664799528257,
+      "learning_rate": 1.751522359869181e-05,
+      "loss": 0.7836,
+      "step": 4849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 40.53112228244021,
+      "learning_rate": 1.751411096157917e-05,
+      "loss": 0.8972,
+      "step": 4850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 29.93172089856882,
+      "learning_rate": 1.7512998110769613e-05,
+      "loss": 0.793,
+      "step": 4851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.60251108607501,
+      "learning_rate": 1.7511885046294782e-05,
+      "loss": 0.7639,
+      "step": 4852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 54.51503702394932,
+      "learning_rate": 1.751077176818634e-05,
+      "loss": 0.8471,
+      "step": 4853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.75204270248763,
+      "learning_rate": 1.7509658276475936e-05,
+      "loss": 0.7966,
+      "step": 4854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.61097097032781,
+      "learning_rate": 1.750854457119525e-05,
+      "loss": 0.8081,
+      "step": 4855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 40.14597395690984,
+      "learning_rate": 1.7507430652375943e-05,
+      "loss": 0.8179,
+      "step": 4856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.22583880686744,
+      "learning_rate": 1.7506316520049704e-05,
+      "loss": 0.7719,
+      "step": 4857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.64595652526385,
+      "learning_rate": 1.750520217424821e-05,
+      "loss": 0.7886,
+      "step": 4858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 38.08351181678489,
+      "learning_rate": 1.750408761500316e-05,
+      "loss": 0.7598,
+      "step": 4859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 35.1892386323691,
+      "learning_rate": 1.7502972842346248e-05,
+      "loss": 0.8609,
+      "step": 4860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 39.078597391194975,
+      "learning_rate": 1.7501857856309176e-05,
+      "loss": 0.8838,
+      "step": 4861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 27.083343242985798,
+      "learning_rate": 1.7500742656923658e-05,
+      "loss": 0.71,
+      "step": 4862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 39.42075147021865,
+      "learning_rate": 1.7499627244221403e-05,
+      "loss": 0.8323,
+      "step": 4863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.86985622388786,
+      "learning_rate": 1.7498511618234134e-05,
+      "loss": 0.786,
+      "step": 4864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 33.518765574033225,
+      "learning_rate": 1.7497395778993585e-05,
+      "loss": 0.7353,
+      "step": 4865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.38403740078959,
+      "learning_rate": 1.749627972653149e-05,
+      "loss": 0.7808,
+      "step": 4866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.42225460097725,
+      "learning_rate": 1.749516346087958e-05,
+      "loss": 0.7582,
+      "step": 4867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 28.761204588663276,
+      "learning_rate": 1.7494046982069604e-05,
+      "loss": 0.8025,
+      "step": 4868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 35.363923438108955,
+      "learning_rate": 1.749293029013332e-05,
+      "loss": 0.756,
+      "step": 4869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 39.060258956710975,
+      "learning_rate": 1.749181338510248e-05,
+      "loss": 0.7586,
+      "step": 4870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.07659501906523,
+      "learning_rate": 1.7490696267008845e-05,
+      "loss": 0.7451,
+      "step": 4871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 38.410003307995886,
+      "learning_rate": 1.74895789358842e-05,
+      "loss": 0.8505,
+      "step": 4872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.878812191767366,
+      "learning_rate": 1.7488461391760304e-05,
+      "loss": 0.8061,
+      "step": 4873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 38.25147771162514,
+      "learning_rate": 1.748734363466895e-05,
+      "loss": 0.7371,
+      "step": 4874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.38034782425675,
+      "learning_rate": 1.748622566464192e-05,
+      "loss": 0.8319,
+      "step": 4875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.02972353391641,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.7847,
+      "step": 4876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 49.691949014193675,
+      "learning_rate": 1.7483989085908027e-05,
+      "loss": 0.8049,
+      "step": 4877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 63.608003245934505,
+      "learning_rate": 1.748287047726477e-05,
+      "loss": 0.7569,
+      "step": 4878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 39.53314363972589,
+      "learning_rate": 1.7481751655813056e-05,
+      "loss": 0.7853,
+      "step": 4879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.66467344050915,
+      "learning_rate": 1.74806326215847e-05,
+      "loss": 0.8329,
+      "step": 4880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 38.30135134088367,
+      "learning_rate": 1.747951337461153e-05,
+      "loss": 0.7245,
+      "step": 4881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.43474035339374,
+      "learning_rate": 1.747839391492537e-05,
+      "loss": 0.7823,
+      "step": 4882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 61.14176071441906,
+      "learning_rate": 1.7477274242558064e-05,
+      "loss": 0.9224,
+      "step": 4883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 45.23860923965065,
+      "learning_rate": 1.7476154357541455e-05,
+      "loss": 0.9008,
+      "step": 4884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.292606940705785,
+      "learning_rate": 1.747503425990739e-05,
+      "loss": 0.8055,
+      "step": 4885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 33.70555782247546,
+      "learning_rate": 1.747391394968772e-05,
+      "loss": 0.8142,
+      "step": 4886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.98703877375695,
+      "learning_rate": 1.7472793426914313e-05,
+      "loss": 0.713,
+      "step": 4887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.759951751302346,
+      "learning_rate": 1.7471672691619028e-05,
+      "loss": 0.8421,
+      "step": 4888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.66316430383517,
+      "learning_rate": 1.7470551743833747e-05,
+      "loss": 0.7622,
+      "step": 4889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.47955119491132,
+      "learning_rate": 1.746943058359034e-05,
+      "loss": 0.7338,
+      "step": 4890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 32.97420790995664,
+      "learning_rate": 1.74683092109207e-05,
+      "loss": 0.862,
+      "step": 4891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 44.577195942393544,
+      "learning_rate": 1.7467187625856713e-05,
+      "loss": 0.9122,
+      "step": 4892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.744099658227764,
+      "learning_rate": 1.7466065828430277e-05,
+      "loss": 0.7822,
+      "step": 4893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.419319216991916,
+      "learning_rate": 1.7464943818673302e-05,
+      "loss": 0.8593,
+      "step": 4894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 34.792104588491746,
+      "learning_rate": 1.7463821596617686e-05,
+      "loss": 0.8281,
+      "step": 4895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 55.52313367600281,
+      "learning_rate": 1.7462699162295353e-05,
+      "loss": 0.8277,
+      "step": 4896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.501291833513086,
+      "learning_rate": 1.746157651573822e-05,
+      "loss": 0.7568,
+      "step": 4897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 39.6790781291655,
+      "learning_rate": 1.7460453656978217e-05,
+      "loss": 0.8818,
+      "step": 4898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 30.91737593602406,
+      "learning_rate": 1.7459330586047273e-05,
+      "loss": 0.8396,
+      "step": 4899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 57.382450060547015,
+      "learning_rate": 1.7458207302977333e-05,
+      "loss": 0.7898,
+      "step": 4900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 48.687324817425555,
+      "learning_rate": 1.7457083807800342e-05,
+      "loss": 0.7517,
+      "step": 4901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 36.69040146618191,
+      "learning_rate": 1.745596010054825e-05,
+      "loss": 0.8803,
+      "step": 4902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 45.05617340161811,
+      "learning_rate": 1.745483618125301e-05,
+      "loss": 0.8558,
+      "step": 4903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 68.3813911951851,
+      "learning_rate": 1.745371204994659e-05,
+      "loss": 0.8859,
+      "step": 4904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 31.670199665612174,
+      "learning_rate": 1.7452587706660967e-05,
+      "loss": 0.7665,
+      "step": 4905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 40.184124325757196,
+      "learning_rate": 1.7451463151428104e-05,
+      "loss": 0.7977,
+      "step": 4906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 32.61323231364325,
+      "learning_rate": 1.7450338384279985e-05,
+      "loss": 0.7898,
+      "step": 4907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.3100016954939,
+      "learning_rate": 1.7449213405248607e-05,
+      "loss": 0.7112,
+      "step": 4908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 56.57249484615816,
+      "learning_rate": 1.7448088214365953e-05,
+      "loss": 0.8171,
+      "step": 4909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 31.659313558274686,
+      "learning_rate": 1.744696281166403e-05,
+      "loss": 0.7549,
+      "step": 4910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 30.91840779516302,
+      "learning_rate": 1.7445837197174836e-05,
+      "loss": 0.7022,
+      "step": 4911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 30.208393906280474,
+      "learning_rate": 1.744471137093039e-05,
+      "loss": 0.7595,
+      "step": 4912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 39.04159952942984,
+      "learning_rate": 1.7443585332962715e-05,
+      "loss": 0.8025,
+      "step": 4913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 33.60260836533053,
+      "learning_rate": 1.744245908330382e-05,
+      "loss": 0.8603,
+      "step": 4914
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 43.068558625219694,
+      "learning_rate": 1.7441332621985742e-05,
+      "loss": 0.9236,
+      "step": 4915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 33.63320956289304,
+      "learning_rate": 1.744020594904052e-05,
+      "loss": 0.7696,
+      "step": 4916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 33.7802220310765,
+      "learning_rate": 1.7439079064500193e-05,
+      "loss": 0.7691,
+      "step": 4917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 46.66338753192148,
+      "learning_rate": 1.7437951968396808e-05,
+      "loss": 0.8435,
+      "step": 4918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 33.44376640426356,
+      "learning_rate": 1.743682466076242e-05,
+      "loss": 0.7086,
+      "step": 4919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 37.611880635378526,
+      "learning_rate": 1.7435697141629087e-05,
+      "loss": 0.7365,
+      "step": 4920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 37.24266457177092,
+      "learning_rate": 1.7434569411028883e-05,
+      "loss": 0.7638,
+      "step": 4921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 39.589664672988555,
+      "learning_rate": 1.743344146899387e-05,
+      "loss": 0.7876,
+      "step": 4922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 36.10562302472779,
+      "learning_rate": 1.7432313315556132e-05,
+      "loss": 0.7607,
+      "step": 4923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 34.075700760238675,
+      "learning_rate": 1.743118495074775e-05,
+      "loss": 0.7564,
+      "step": 4924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 46.554810397233496,
+      "learning_rate": 1.7430056374600813e-05,
+      "loss": 0.8172,
+      "step": 4925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 36.97858630875624,
+      "learning_rate": 1.7428927587147422e-05,
+      "loss": 0.8025,
+      "step": 4926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 40.45034586589995,
+      "learning_rate": 1.7427798588419674e-05,
+      "loss": 0.7656,
+      "step": 4927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 28.887613901914882,
+      "learning_rate": 1.7426669378449685e-05,
+      "loss": 0.7589,
+      "step": 4928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 43.471450151379294,
+      "learning_rate": 1.7425539957269557e-05,
+      "loss": 0.8135,
+      "step": 4929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 42.38922294831844,
+      "learning_rate": 1.742441032491142e-05,
+      "loss": 0.8488,
+      "step": 4930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 38.040460453916964,
+      "learning_rate": 1.7423280481407393e-05,
+      "loss": 0.8937,
+      "step": 4931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 48.09952057938557,
+      "learning_rate": 1.7422150426789613e-05,
+      "loss": 0.8927,
+      "step": 4932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 39.29859292918463,
+      "learning_rate": 1.7421020161090216e-05,
+      "loss": 0.7984,
+      "step": 4933
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 34.92413564102554,
+      "learning_rate": 1.741988968434135e-05,
+      "loss": 0.8007,
+      "step": 4934
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 38.96428003500927,
+      "learning_rate": 1.741875899657516e-05,
+      "loss": 0.7557,
+      "step": 4935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 40.968130547840104,
+      "learning_rate": 1.7417628097823802e-05,
+      "loss": 0.7416,
+      "step": 4936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 42.51356850178339,
+      "learning_rate": 1.741649698811944e-05,
+      "loss": 0.7433,
+      "step": 4937
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 104.84527104691408,
+      "learning_rate": 1.7415365667494248e-05,
+      "loss": 0.7819,
+      "step": 4938
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.597462006183303,
+      "learning_rate": 1.741423413598039e-05,
+      "loss": 0.6858,
+      "step": 4939
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 30.655627253095492,
+      "learning_rate": 1.741310239361005e-05,
+      "loss": 0.794,
+      "step": 4940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 34.23207876425586,
+      "learning_rate": 1.7411970440415418e-05,
+      "loss": 0.8808,
+      "step": 4941
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 38.975983160243324,
+      "learning_rate": 1.7410838276428677e-05,
+      "loss": 0.8312,
+      "step": 4942
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 36.052501826303796,
+      "learning_rate": 1.7409705901682033e-05,
+      "loss": 0.6793,
+      "step": 4943
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 52.67522328116262,
+      "learning_rate": 1.740857331620769e-05,
+      "loss": 0.8434,
+      "step": 4944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 50.042409754224515,
+      "learning_rate": 1.740744052003785e-05,
+      "loss": 0.9015,
+      "step": 4945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 32.391735812331675,
+      "learning_rate": 1.7406307513204742e-05,
+      "loss": 0.759,
+      "step": 4946
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 43.84245825867821,
+      "learning_rate": 1.7405174295740577e-05,
+      "loss": 0.9237,
+      "step": 4947
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 35.48385611675041,
+      "learning_rate": 1.7404040867677587e-05,
+      "loss": 0.7397,
+      "step": 4948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 44.407381820482804,
+      "learning_rate": 1.740290722904801e-05,
+      "loss": 0.8133,
+      "step": 4949
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 39.631277375331365,
+      "learning_rate": 1.7401773379884078e-05,
+      "loss": 0.7895,
+      "step": 4950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 41.500604458508796,
+      "learning_rate": 1.7400639320218042e-05,
+      "loss": 0.7443,
+      "step": 4951
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 45.06580643818052,
+      "learning_rate": 1.739950505008215e-05,
+      "loss": 0.8657,
+      "step": 4952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 37.144442062439346,
+      "learning_rate": 1.7398370569508667e-05,
+      "loss": 0.9296,
+      "step": 4953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 35.966977600256776,
+      "learning_rate": 1.739723587852985e-05,
+      "loss": 0.7404,
+      "step": 4954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 32.058464402358155,
+      "learning_rate": 1.7396100977177975e-05,
+      "loss": 0.8787,
+      "step": 4955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 50.22378965773624,
+      "learning_rate": 1.7394965865485312e-05,
+      "loss": 0.7409,
+      "step": 4956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 31.95015490468998,
+      "learning_rate": 1.7393830543484147e-05,
+      "loss": 0.7796,
+      "step": 4957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 39.05178371072671,
+      "learning_rate": 1.7392695011206768e-05,
+      "loss": 0.8192,
+      "step": 4958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 35.090100500311316,
+      "learning_rate": 1.7391559268685464e-05,
+      "loss": 0.7572,
+      "step": 4959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 37.35386264774466,
+      "learning_rate": 1.739042331595254e-05,
+      "loss": 0.7945,
+      "step": 4960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 32.7489288715472,
+      "learning_rate": 1.73892871530403e-05,
+      "loss": 0.7696,
+      "step": 4961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.19288951703399,
+      "learning_rate": 1.7388150779981057e-05,
+      "loss": 0.8762,
+      "step": 4962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 40.06149603765547,
+      "learning_rate": 1.738701419680713e-05,
+      "loss": 0.8113,
+      "step": 4963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.73541489268041,
+      "learning_rate": 1.7385877403550836e-05,
+      "loss": 0.7793,
+      "step": 4964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 38.83664885816973,
+      "learning_rate": 1.738474040024451e-05,
+      "loss": 0.9325,
+      "step": 4965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 45.89264821783732,
+      "learning_rate": 1.738360318692049e-05,
+      "loss": 0.7629,
+      "step": 4966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.64422443894643,
+      "learning_rate": 1.738246576361111e-05,
+      "loss": 0.8877,
+      "step": 4967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 27.061930335482604,
+      "learning_rate": 1.7381328130348727e-05,
+      "loss": 0.6861,
+      "step": 4968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.65712454062364,
+      "learning_rate": 1.7380190287165686e-05,
+      "loss": 0.8245,
+      "step": 4969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 27.99836596123177,
+      "learning_rate": 1.7379052234094353e-05,
+      "loss": 0.7346,
+      "step": 4970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 31.611585595393727,
+      "learning_rate": 1.737791397116709e-05,
+      "loss": 0.7925,
+      "step": 4971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 35.40128725797409,
+      "learning_rate": 1.737677549841627e-05,
+      "loss": 0.7509,
+      "step": 4972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 46.03601204453365,
+      "learning_rate": 1.7375636815874273e-05,
+      "loss": 0.8129,
+      "step": 4973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.10431379382538,
+      "learning_rate": 1.7374497923573473e-05,
+      "loss": 0.7552,
+      "step": 4974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.11794383688522,
+      "learning_rate": 1.7373358821546272e-05,
+      "loss": 0.8103,
+      "step": 4975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 42.061449953210634,
+      "learning_rate": 1.7372219509825056e-05,
+      "loss": 0.8714,
+      "step": 4976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 46.6139284033204,
+      "learning_rate": 1.737107998844223e-05,
+      "loss": 0.6798,
+      "step": 4977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.25122372436203,
+      "learning_rate": 1.7369940257430203e-05,
+      "loss": 0.7573,
+      "step": 4978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 31.496700433837187,
+      "learning_rate": 1.7368800316821387e-05,
+      "loss": 0.709,
+      "step": 4979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 35.87586865054663,
+      "learning_rate": 1.7367660166648197e-05,
+      "loss": 0.7418,
+      "step": 4980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 43.88230682456576,
+      "learning_rate": 1.7366519806943067e-05,
+      "loss": 0.854,
+      "step": 4981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.13652958076635,
+      "learning_rate": 1.736537923773842e-05,
+      "loss": 0.7568,
+      "step": 4982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 39.22750538519203,
+      "learning_rate": 1.736423845906669e-05,
+      "loss": 0.8622,
+      "step": 4983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 41.66997421329305,
+      "learning_rate": 1.7363097470960336e-05,
+      "loss": 0.8736,
+      "step": 4984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 39.43640599494142,
+      "learning_rate": 1.736195627345179e-05,
+      "loss": 0.7992,
+      "step": 4985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 41.77724065781295,
+      "learning_rate": 1.736081486657352e-05,
+      "loss": 0.7101,
+      "step": 4986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.92263921801814,
+      "learning_rate": 1.7359673250357977e-05,
+      "loss": 0.841,
+      "step": 4987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.54515903418986,
+      "learning_rate": 1.7358531424837633e-05,
+      "loss": 0.7748,
+      "step": 4988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 32.87207614022273,
+      "learning_rate": 1.7357389390044965e-05,
+      "loss": 0.7853,
+      "step": 4989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 33.364489986460995,
+      "learning_rate": 1.735624714601244e-05,
+      "loss": 0.7363,
+      "step": 4990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 52.812884878979496,
+      "learning_rate": 1.735510469277255e-05,
+      "loss": 0.8387,
+      "step": 4991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.02665406536362,
+      "learning_rate": 1.7353962030357786e-05,
+      "loss": 0.8179,
+      "step": 4992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 48.13381852936145,
+      "learning_rate": 1.7352819158800646e-05,
+      "loss": 0.8696,
+      "step": 4993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 28.27547349895926,
+      "learning_rate": 1.735167607813363e-05,
+      "loss": 0.812,
+      "step": 4994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 26.837133735747955,
+      "learning_rate": 1.7350532788389248e-05,
+      "loss": 0.7602,
+      "step": 4995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.645330493624876,
+      "learning_rate": 1.734938928960001e-05,
+      "loss": 0.762,
+      "step": 4996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 44.32534176489784,
+      "learning_rate": 1.7348245581798438e-05,
+      "loss": 0.7904,
+      "step": 4997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 26.65762586497551,
+      "learning_rate": 1.7347101665017064e-05,
+      "loss": 0.7369,
+      "step": 4998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 38.78899230028698,
+      "learning_rate": 1.7345957539288417e-05,
+      "loss": 0.8097,
+      "step": 4999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.3553924295597,
+      "learning_rate": 1.7344813204645033e-05,
+      "loss": 0.7015,
+      "step": 5000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 32.63814405231698,
+      "learning_rate": 1.734366866111946e-05,
+      "loss": 0.7557,
+      "step": 5001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 39.99140397332864,
+      "learning_rate": 1.7342523908744246e-05,
+      "loss": 0.7616,
+      "step": 5002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.666869602885534,
+      "learning_rate": 1.7341378947551946e-05,
+      "loss": 0.8019,
+      "step": 5003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 33.55969433982883,
+      "learning_rate": 1.7340233777575125e-05,
+      "loss": 0.9274,
+      "step": 5004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.27107630627118,
+      "learning_rate": 1.733908839884635e-05,
+      "loss": 0.7681,
+      "step": 5005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 33.72863003435633,
+      "learning_rate": 1.733794281139819e-05,
+      "loss": 0.7643,
+      "step": 5006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.90784561057036,
+      "learning_rate": 1.733679701526323e-05,
+      "loss": 0.7969,
+      "step": 5007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 28.554916558344996,
+      "learning_rate": 1.7335651010474057e-05,
+      "loss": 0.7396,
+      "step": 5008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 44.34828919827853,
+      "learning_rate": 1.7334504797063257e-05,
+      "loss": 0.7734,
+      "step": 5009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.887950969952605,
+      "learning_rate": 1.733335837506344e-05,
+      "loss": 0.8536,
+      "step": 5010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.7867235673454,
+      "learning_rate": 1.733221174450719e-05,
+      "loss": 0.8874,
+      "step": 5011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 35.775934462535474,
+      "learning_rate": 1.7331064905427133e-05,
+      "loss": 0.7146,
+      "step": 5012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.78607540811798,
+      "learning_rate": 1.7329917857855875e-05,
+      "loss": 0.8105,
+      "step": 5013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 30.662580297328407,
+      "learning_rate": 1.7328770601826047e-05,
+      "loss": 0.7843,
+      "step": 5014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 33.581526767251184,
+      "learning_rate": 1.7327623137370265e-05,
+      "loss": 0.7888,
+      "step": 5015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 38.23822064189841,
+      "learning_rate": 1.732647546452117e-05,
+      "loss": 0.8326,
+      "step": 5016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.72447794773593,
+      "learning_rate": 1.7325327583311396e-05,
+      "loss": 0.8427,
+      "step": 5017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 27.586187659171607,
+      "learning_rate": 1.7324179493773593e-05,
+      "loss": 0.7343,
+      "step": 5018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.35723620027698,
+      "learning_rate": 1.732303119594041e-05,
+      "loss": 0.7967,
+      "step": 5019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 28.608246758640487,
+      "learning_rate": 1.73218826898445e-05,
+      "loss": 0.7737,
+      "step": 5020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 27.00978148292552,
+      "learning_rate": 1.7320733975518533e-05,
+      "loss": 0.6995,
+      "step": 5021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 28.806944214228977,
+      "learning_rate": 1.7319585052995177e-05,
+      "loss": 0.7315,
+      "step": 5022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 37.671638200063896,
+      "learning_rate": 1.7318435922307097e-05,
+      "loss": 0.8234,
+      "step": 5023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 32.31521466655634,
+      "learning_rate": 1.7317286583486983e-05,
+      "loss": 0.7976,
+      "step": 5024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.88829069960274,
+      "learning_rate": 1.7316137036567523e-05,
+      "loss": 0.8156,
+      "step": 5025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 43.169937996032246,
+      "learning_rate": 1.7314987281581403e-05,
+      "loss": 0.8514,
+      "step": 5026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.3418817352573,
+      "learning_rate": 1.731383731856132e-05,
+      "loss": 0.8509,
+      "step": 5027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.671129226399955,
+      "learning_rate": 1.731268714753999e-05,
+      "loss": 0.7509,
+      "step": 5028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.496025996588912,
+      "learning_rate": 1.7311536768550107e-05,
+      "loss": 0.763,
+      "step": 5029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 30.539692112757802,
+      "learning_rate": 1.73103861816244e-05,
+      "loss": 0.823,
+      "step": 5030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.66230507001225,
+      "learning_rate": 1.7309235386795585e-05,
+      "loss": 0.807,
+      "step": 5031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 35.88859790861851,
+      "learning_rate": 1.7308084384096395e-05,
+      "loss": 0.8778,
+      "step": 5032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.29258526038848,
+      "learning_rate": 1.7306933173559554e-05,
+      "loss": 0.8197,
+      "step": 5033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 46.758076163992996,
+      "learning_rate": 1.7305781755217812e-05,
+      "loss": 0.9424,
+      "step": 5034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.704231705391866,
+      "learning_rate": 1.730463012910391e-05,
+      "loss": 0.7853,
+      "step": 5035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 39.67288878604023,
+      "learning_rate": 1.73034782952506e-05,
+      "loss": 0.7981,
+      "step": 5036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 35.53602946402501,
+      "learning_rate": 1.7302326253690643e-05,
+      "loss": 0.7528,
+      "step": 5037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 52.737212027635806,
+      "learning_rate": 1.7301174004456794e-05,
+      "loss": 0.7582,
+      "step": 5038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 32.87025463100786,
+      "learning_rate": 1.730002154758183e-05,
+      "loss": 0.7599,
+      "step": 5039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 37.790844168469505,
+      "learning_rate": 1.729886888309852e-05,
+      "loss": 0.8042,
+      "step": 5040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.42756752031988,
+      "learning_rate": 1.7297716011039654e-05,
+      "loss": 0.712,
+      "step": 5041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 44.794358256295205,
+      "learning_rate": 1.729656293143801e-05,
+      "loss": 0.7772,
+      "step": 5042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 37.129206781911826,
+      "learning_rate": 1.7295409644326387e-05,
+      "loss": 0.7938,
+      "step": 5043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 53.57971815586895,
+      "learning_rate": 1.7294256149737577e-05,
+      "loss": 0.9336,
+      "step": 5044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 50.621346487854844,
+      "learning_rate": 1.7293102447704395e-05,
+      "loss": 0.7733,
+      "step": 5045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 30.511734932790304,
+      "learning_rate": 1.7291948538259644e-05,
+      "loss": 0.8305,
+      "step": 5046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 40.19425522550944,
+      "learning_rate": 1.7290794421436138e-05,
+      "loss": 0.8599,
+      "step": 5047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 47.47019849303787,
+      "learning_rate": 1.7289640097266712e-05,
+      "loss": 0.9037,
+      "step": 5048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 42.06598268009268,
+      "learning_rate": 1.728848556578418e-05,
+      "loss": 0.8299,
+      "step": 5049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.48595276237911,
+      "learning_rate": 1.7287330827021382e-05,
+      "loss": 0.7681,
+      "step": 5050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 32.70381554951204,
+      "learning_rate": 1.7286175881011158e-05,
+      "loss": 0.7023,
+      "step": 5051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 46.48836144543862,
+      "learning_rate": 1.7285020727786354e-05,
+      "loss": 0.8739,
+      "step": 5052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 29.70738774373515,
+      "learning_rate": 1.7283865367379826e-05,
+      "loss": 0.8191,
+      "step": 5053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 71.64197966148684,
+      "learning_rate": 1.7282709799824428e-05,
+      "loss": 0.7576,
+      "step": 5054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.34408011030499,
+      "learning_rate": 1.7281554025153018e-05,
+      "loss": 0.781,
+      "step": 5055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 39.46481173114824,
+      "learning_rate": 1.7280398043398478e-05,
+      "loss": 0.829,
+      "step": 5056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 29.004706409861033,
+      "learning_rate": 1.727924185459367e-05,
+      "loss": 0.7539,
+      "step": 5057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.16895426359312,
+      "learning_rate": 1.7278085458771485e-05,
+      "loss": 0.7864,
+      "step": 5058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 39.34267946442697,
+      "learning_rate": 1.7276928855964805e-05,
+      "loss": 0.8299,
+      "step": 5059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.03315542894981,
+      "learning_rate": 1.727577204620653e-05,
+      "loss": 0.7741,
+      "step": 5060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 47.04774358752717,
+      "learning_rate": 1.727461502952955e-05,
+      "loss": 0.797,
+      "step": 5061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.31606094050113,
+      "learning_rate": 1.7273457805966773e-05,
+      "loss": 0.9267,
+      "step": 5062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.110256685594415,
+      "learning_rate": 1.7272300375551116e-05,
+      "loss": 0.6833,
+      "step": 5063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 35.88550060061239,
+      "learning_rate": 1.7271142738315485e-05,
+      "loss": 0.8335,
+      "step": 5064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 45.95819956165442,
+      "learning_rate": 1.7269984894292814e-05,
+      "loss": 0.8934,
+      "step": 5065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 39.72529281042563,
+      "learning_rate": 1.726882684351602e-05,
+      "loss": 0.8191,
+      "step": 5066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.881050783035136,
+      "learning_rate": 1.7267668586018044e-05,
+      "loss": 0.753,
+      "step": 5067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.215183634763957,
+      "learning_rate": 1.7266510121831824e-05,
+      "loss": 0.6797,
+      "step": 5068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.80976990487334,
+      "learning_rate": 1.726535145099031e-05,
+      "loss": 0.6919,
+      "step": 5069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.90542018265878,
+      "learning_rate": 1.726419257352645e-05,
+      "loss": 0.801,
+      "step": 5070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 40.561611747714444,
+      "learning_rate": 1.7263033489473202e-05,
+      "loss": 0.9085,
+      "step": 5071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 41.82711544535658,
+      "learning_rate": 1.7261874198863533e-05,
+      "loss": 0.8411,
+      "step": 5072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 59.968768216717265,
+      "learning_rate": 1.7260714701730405e-05,
+      "loss": 0.8908,
+      "step": 5073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 39.28918668340847,
+      "learning_rate": 1.72595549981068e-05,
+      "loss": 0.8053,
+      "step": 5074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.63321471283312,
+      "learning_rate": 1.72583950880257e-05,
+      "loss": 0.8145,
+      "step": 5075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.74210329347961,
+      "learning_rate": 1.7257234971520086e-05,
+      "loss": 0.7312,
+      "step": 5076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 40.105248738673076,
+      "learning_rate": 1.7256074648622958e-05,
+      "loss": 0.9223,
+      "step": 5077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.92712141079162,
+      "learning_rate": 1.725491411936731e-05,
+      "loss": 0.7552,
+      "step": 5078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 46.74721081210999,
+      "learning_rate": 1.7253753383786148e-05,
+      "loss": 0.7523,
+      "step": 5079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.67314853751872,
+      "learning_rate": 1.7252592441912487e-05,
+      "loss": 0.7966,
+      "step": 5080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.805488078867285,
+      "learning_rate": 1.7251431293779334e-05,
+      "loss": 0.7913,
+      "step": 5081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.3231830410297,
+      "learning_rate": 1.725026993941972e-05,
+      "loss": 0.7401,
+      "step": 5082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 48.41357569367443,
+      "learning_rate": 1.724910837886667e-05,
+      "loss": 0.7317,
+      "step": 5083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.84642456697402,
+      "learning_rate": 1.7247946612153215e-05,
+      "loss": 0.7462,
+      "step": 5084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 36.07852797120047,
+      "learning_rate": 1.7246784639312403e-05,
+      "loss": 0.7888,
+      "step": 5085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 37.50161706616816,
+      "learning_rate": 1.7245622460377274e-05,
+      "loss": 0.9922,
+      "step": 5086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.8771303476472,
+      "learning_rate": 1.724446007538088e-05,
+      "loss": 0.8459,
+      "step": 5087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 32.9706536272264,
+      "learning_rate": 1.724329748435628e-05,
+      "loss": 0.8465,
+      "step": 5088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 30.091528773900812,
+      "learning_rate": 1.7242134687336535e-05,
+      "loss": 0.8033,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.89972894284594,
+      "learning_rate": 1.7240971684354717e-05,
+      "loss": 0.9029,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.35270343072886,
+      "learning_rate": 1.72398084754439e-05,
+      "loss": 0.8098,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.443565431204426,
+      "learning_rate": 1.7238645060637165e-05,
+      "loss": 0.8313,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 32.586244245579365,
+      "learning_rate": 1.7237481439967602e-05,
+      "loss": 0.7365,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 38.16940280249246,
+      "learning_rate": 1.7236317613468294e-05,
+      "loss": 0.7964,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.389979174874696,
+      "learning_rate": 1.7235153581172353e-05,
+      "loss": 0.7502,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 52.11096388926628,
+      "learning_rate": 1.7233989343112872e-05,
+      "loss": 0.8075,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 30.173664717534574,
+      "learning_rate": 1.723282489932297e-05,
+      "loss": 0.7811,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 50.94343559927222,
+      "learning_rate": 1.7231660249835756e-05,
+      "loss": 0.8638,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 41.97311747897976,
+      "learning_rate": 1.7230495394684353e-05,
+      "loss": 0.8456,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.38365541327769,
+      "learning_rate": 1.7229330333901895e-05,
+      "loss": 0.8134,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 32.20456135973908,
+      "learning_rate": 1.722816506752151e-05,
+      "loss": 0.7918,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.943220308794675,
+      "learning_rate": 1.722699959557634e-05,
+      "loss": 0.7273,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.52099267018878,
+      "learning_rate": 1.7225833918099527e-05,
+      "loss": 0.7535,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.15709067886925,
+      "learning_rate": 1.7224668035124224e-05,
+      "loss": 0.7191,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.67909173180867,
+      "learning_rate": 1.722350194668359e-05,
+      "loss": 0.7922,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 36.83308611581154,
+      "learning_rate": 1.7222335652810788e-05,
+      "loss": 0.8742,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.493262044770955,
+      "learning_rate": 1.7221169153538978e-05,
+      "loss": 0.8303,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 36.715043264740444,
+      "learning_rate": 1.7220002448901346e-05,
+      "loss": 0.7449,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 42.42953559737319,
+      "learning_rate": 1.721883553893107e-05,
+      "loss": 0.8874,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 37.3068992334574,
+      "learning_rate": 1.7217668423661332e-05,
+      "loss": 0.8511,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.97050287530967,
+      "learning_rate": 1.7216501103125326e-05,
+      "loss": 0.8492,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.936873773862185,
+      "learning_rate": 1.721533357735625e-05,
+      "loss": 0.7793,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 52.38473675156963,
+      "learning_rate": 1.721416584638731e-05,
+      "loss": 0.8287,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.45893511925122,
+      "learning_rate": 1.721299791025171e-05,
+      "loss": 0.7972,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.80643302896004,
+      "learning_rate": 1.7211829768982672e-05,
+      "loss": 0.831,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 58.90090150709918,
+      "learning_rate": 1.7210661422613412e-05,
+      "loss": 0.8074,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.47771773619314,
+      "learning_rate": 1.720949287117716e-05,
+      "loss": 0.8218,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.20350155700851,
+      "learning_rate": 1.720832411470715e-05,
+      "loss": 0.9219,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 37.71783667393334,
+      "learning_rate": 1.720715515323662e-05,
+      "loss": 0.9536,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.268357887231275,
+      "learning_rate": 1.7205985986798808e-05,
+      "loss": 0.9059,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 28.3548610110804,
+      "learning_rate": 1.7204816615426972e-05,
+      "loss": 0.7794,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 62.07816411287663,
+      "learning_rate": 1.7203647039154367e-05,
+      "loss": 0.8454,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 46.077904034198255,
+      "learning_rate": 1.7202477258014256e-05,
+      "loss": 0.8156,
+      "step": 5123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 30.14654450059931,
+      "learning_rate": 1.72013072720399e-05,
+      "loss": 0.7657,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.2677337207469,
+      "learning_rate": 1.7200137081264584e-05,
+      "loss": 0.7522,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 36.038399886037915,
+      "learning_rate": 1.719896668572158e-05,
+      "loss": 0.8094,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 41.933703192402376,
+      "learning_rate": 1.7197796085444173e-05,
+      "loss": 0.8859,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.276640695524,
+      "learning_rate": 1.7196625280465656e-05,
+      "loss": 0.7366,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.82632707292792,
+      "learning_rate": 1.7195454270819326e-05,
+      "loss": 0.7404,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.880529670256806,
+      "learning_rate": 1.7194283056538486e-05,
+      "loss": 0.8113,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 46.12239961608156,
+      "learning_rate": 1.7193111637656445e-05,
+      "loss": 0.7646,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 37.126329179853194,
+      "learning_rate": 1.7191940014206518e-05,
+      "loss": 0.7662,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 49.13875395058653,
+      "learning_rate": 1.719076818622202e-05,
+      "loss": 0.9188,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 36.697453852794126,
+      "learning_rate": 1.7189596153736285e-05,
+      "loss": 0.7878,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 89.8635180570736,
+      "learning_rate": 1.7188423916782637e-05,
+      "loss": 0.7245,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 28.79423837084046,
+      "learning_rate": 1.7187251475394423e-05,
+      "loss": 0.8005,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 47.5723453186999,
+      "learning_rate": 1.7186078829604978e-05,
+      "loss": 0.8363,
+      "step": 5137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.862127914079736,
+      "learning_rate": 1.7184905979447655e-05,
+      "loss": 0.7754,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.039170081037085,
+      "learning_rate": 1.7183732924955808e-05,
+      "loss": 0.6347,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 40.080681422983425,
+      "learning_rate": 1.7182559666162802e-05,
+      "loss": 0.7972,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.023820460155484,
+      "learning_rate": 1.7181386203102e-05,
+      "loss": 0.7211,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.7662518402055,
+      "learning_rate": 1.7180212535806776e-05,
+      "loss": 0.7743,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.36796318162202,
+      "learning_rate": 1.717903866431051e-05,
+      "loss": 0.7896,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.41142828140268,
+      "learning_rate": 1.717786458864658e-05,
+      "loss": 0.7145,
+      "step": 5144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.00754158443171,
+      "learning_rate": 1.717669030884838e-05,
+      "loss": 0.7975,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.196421144685246,
+      "learning_rate": 1.7175515824949304e-05,
+      "loss": 0.8306,
+      "step": 5146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.55178166020633,
+      "learning_rate": 1.7174341136982763e-05,
+      "loss": 0.811,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 50.43569136145615,
+      "learning_rate": 1.717316624498215e-05,
+      "loss": 0.8649,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 35.633546192299455,
+      "learning_rate": 1.7171991148980888e-05,
+      "loss": 0.7819,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 63.27894248993861,
+      "learning_rate": 1.717081584901239e-05,
+      "loss": 0.9472,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.28176802293354,
+      "learning_rate": 1.7169640345110088e-05,
+      "loss": 0.8558,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.24081840101921,
+      "learning_rate": 1.7168464637307408e-05,
+      "loss": 0.8536,
+      "step": 5152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 34.74655398771107,
+      "learning_rate": 1.7167288725637786e-05,
+      "loss": 0.7806,
+      "step": 5153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 38.3977358035046,
+      "learning_rate": 1.7166112610134665e-05,
+      "loss": 0.8345,
+      "step": 5154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 45.618585302826745,
+      "learning_rate": 1.7164936290831494e-05,
+      "loss": 0.9153,
+      "step": 5155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 50.38187074779283,
+      "learning_rate": 1.7163759767761727e-05,
+      "loss": 0.7384,
+      "step": 5156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 37.86819148164116,
+      "learning_rate": 1.716258304095882e-05,
+      "loss": 0.8259,
+      "step": 5157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 37.569862244038454,
+      "learning_rate": 1.716140611045625e-05,
+      "loss": 0.8339,
+      "step": 5158
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 32.268966472223994,
+      "learning_rate": 1.716022897628747e-05,
+      "loss": 0.7374,
+      "step": 5159
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.47156648922627,
+      "learning_rate": 1.7159051638485967e-05,
+      "loss": 0.7966,
+      "step": 5160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 39.52926665747446,
+      "learning_rate": 1.715787409708523e-05,
+      "loss": 0.7662,
+      "step": 5161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 40.436361234924,
+      "learning_rate": 1.7156696352118735e-05,
+      "loss": 0.7266,
+      "step": 5162
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 40.87521995695235,
+      "learning_rate": 1.7155518403619983e-05,
+      "loss": 0.8403,
+      "step": 5163
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 38.53378449679669,
+      "learning_rate": 1.7154340251622476e-05,
+      "loss": 0.7864,
+      "step": 5164
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 34.69940461711581,
+      "learning_rate": 1.7153161896159717e-05,
+      "loss": 0.6637,
+      "step": 5165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 31.744368519308534,
+      "learning_rate": 1.7151983337265217e-05,
+      "loss": 0.8038,
+      "step": 5166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 43.328276085409605,
+      "learning_rate": 1.7150804574972496e-05,
+      "loss": 0.7838,
+      "step": 5167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.345211369854745,
+      "learning_rate": 1.7149625609315076e-05,
+      "loss": 0.8679,
+      "step": 5168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 38.62677476256058,
+      "learning_rate": 1.7148446440326485e-05,
+      "loss": 0.8056,
+      "step": 5169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 37.828810555234675,
+      "learning_rate": 1.714726706804026e-05,
+      "loss": 0.7802,
+      "step": 5170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 46.83647653048411,
+      "learning_rate": 1.714608749248994e-05,
+      "loss": 0.9744,
+      "step": 5171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 32.81836869572777,
+      "learning_rate": 1.714490771370907e-05,
+      "loss": 0.8288,
+      "step": 5172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 51.41053924290353,
+      "learning_rate": 1.7143727731731204e-05,
+      "loss": 0.8813,
+      "step": 5173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.19807268104505,
+      "learning_rate": 1.7142547546589903e-05,
+      "loss": 0.6279,
+      "step": 5174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 29.97908496956013,
+      "learning_rate": 1.714136715831873e-05,
+      "loss": 0.7703,
+      "step": 5175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 27.25331796779669,
+      "learning_rate": 1.7140186566951248e-05,
+      "loss": 0.7277,
+      "step": 5176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 44.51045047012689,
+      "learning_rate": 1.7139005772521038e-05,
+      "loss": 0.8625,
+      "step": 5177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 31.339230710789842,
+      "learning_rate": 1.7137824775061682e-05,
+      "loss": 0.7518,
+      "step": 5178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 30.718461366153125,
+      "learning_rate": 1.7136643574606764e-05,
+      "loss": 0.8075,
+      "step": 5179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.87028468744026,
+      "learning_rate": 1.7135462171189877e-05,
+      "loss": 0.8731,
+      "step": 5180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.176575429865714,
+      "learning_rate": 1.713428056484462e-05,
+      "loss": 0.7614,
+      "step": 5181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 51.36210221679422,
+      "learning_rate": 1.7133098755604595e-05,
+      "loss": 0.8681,
+      "step": 5182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.06087723622869,
+      "learning_rate": 1.713191674350342e-05,
+      "loss": 0.7767,
+      "step": 5183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 49.84696389482679,
+      "learning_rate": 1.71307345285747e-05,
+      "loss": 0.8023,
+      "step": 5184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 31.552621526820356,
+      "learning_rate": 1.7129552110852064e-05,
+      "loss": 0.7633,
+      "step": 5185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.21390147171926,
+      "learning_rate": 1.7128369490369134e-05,
+      "loss": 0.8261,
+      "step": 5186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 42.937829570255055,
+      "learning_rate": 1.7127186667159547e-05,
+      "loss": 0.8537,
+      "step": 5187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 30.341546910744682,
+      "learning_rate": 1.712600364125694e-05,
+      "loss": 0.7564,
+      "step": 5188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 42.27926218303943,
+      "learning_rate": 1.712482041269496e-05,
+      "loss": 0.8537,
+      "step": 5189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 31.40799875657263,
+      "learning_rate": 1.7123636981507256e-05,
+      "loss": 0.7932,
+      "step": 5190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 37.93360516384276,
+      "learning_rate": 1.712245334772748e-05,
+      "loss": 0.7638,
+      "step": 5191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 31.413116421858025,
+      "learning_rate": 1.7121269511389298e-05,
+      "loss": 0.6713,
+      "step": 5192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.6034482113734,
+      "learning_rate": 1.7120085472526377e-05,
+      "loss": 0.83,
+      "step": 5193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 61.29994170651302,
+      "learning_rate": 1.711890123117239e-05,
+      "loss": 0.7888,
+      "step": 5194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 32.41749243473104,
+      "learning_rate": 1.711771678736102e-05,
+      "loss": 0.8099,
+      "step": 5195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 30.813912632035507,
+      "learning_rate": 1.7116532141125947e-05,
+      "loss": 0.6798,
+      "step": 5196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.76421139091256,
+      "learning_rate": 1.711534729250086e-05,
+      "loss": 0.8388,
+      "step": 5197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 46.9031822426377,
+      "learning_rate": 1.711416224151946e-05,
+      "loss": 0.8138,
+      "step": 5198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 39.02349338775695,
+      "learning_rate": 1.7112976988215445e-05,
+      "loss": 0.8958,
+      "step": 5199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.95634525453887,
+      "learning_rate": 1.711179153262253e-05,
+      "loss": 0.8032,
+      "step": 5200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 38.914004981582984,
+      "learning_rate": 1.711060587477442e-05,
+      "loss": 0.7946,
+      "step": 5201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 28.22078836457076,
+      "learning_rate": 1.710942001470484e-05,
+      "loss": 0.7343,
+      "step": 5202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 45.00955601000089,
+      "learning_rate": 1.7108233952447516e-05,
+      "loss": 0.8703,
+      "step": 5203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 29.103519958614555,
+      "learning_rate": 1.7107047688036175e-05,
+      "loss": 0.6684,
+      "step": 5204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 39.3261222890196,
+      "learning_rate": 1.7105861221504555e-05,
+      "loss": 0.8151,
+      "step": 5205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 33.41759762493801,
+      "learning_rate": 1.71046745528864e-05,
+      "loss": 0.762,
+      "step": 5206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 31.590600272721893,
+      "learning_rate": 1.7103487682215454e-05,
+      "loss": 0.8158,
+      "step": 5207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 37.42398550178562,
+      "learning_rate": 1.710230060952548e-05,
+      "loss": 0.7453,
+      "step": 5208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 33.55163535468737,
+      "learning_rate": 1.7101113334850224e-05,
+      "loss": 0.833,
+      "step": 5209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 34.806806848736045,
+      "learning_rate": 1.709992585822346e-05,
+      "loss": 0.7913,
+      "step": 5210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 52.15814345427277,
+      "learning_rate": 1.709873817967896e-05,
+      "loss": 0.79,
+      "step": 5211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.188992612630535,
+      "learning_rate": 1.70975502992505e-05,
+      "loss": 0.7609,
+      "step": 5212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 30.351028716549823,
+      "learning_rate": 1.7096362216971857e-05,
+      "loss": 0.6969,
+      "step": 5213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 45.033530382051865,
+      "learning_rate": 1.7095173932876827e-05,
+      "loss": 0.7665,
+      "step": 5214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 42.669988634507156,
+      "learning_rate": 1.7093985446999203e-05,
+      "loss": 0.8002,
+      "step": 5215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 53.136878056768666,
+      "learning_rate": 1.7092796759372778e-05,
+      "loss": 0.7912,
+      "step": 5216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 32.86375489115221,
+      "learning_rate": 1.7091607870031362e-05,
+      "loss": 0.7784,
+      "step": 5217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 37.29169121249508,
+      "learning_rate": 1.7090418779008772e-05,
+      "loss": 0.833,
+      "step": 5218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 37.49650112359606,
+      "learning_rate": 1.7089229486338815e-05,
+      "loss": 0.7932,
+      "step": 5219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 36.55912938353801,
+      "learning_rate": 1.708803999205532e-05,
+      "loss": 0.8069,
+      "step": 5220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 39.451785222334784,
+      "learning_rate": 1.7086850296192118e-05,
+      "loss": 0.8037,
+      "step": 5221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 37.24671253956563,
+      "learning_rate": 1.7085660398783032e-05,
+      "loss": 0.7321,
+      "step": 5222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 55.367867847911604,
+      "learning_rate": 1.7084470299861915e-05,
+      "loss": 0.85,
+      "step": 5223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 35.375317245247054,
+      "learning_rate": 1.7083279999462603e-05,
+      "loss": 0.7703,
+      "step": 5224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 49.78232418608751,
+      "learning_rate": 1.7082089497618955e-05,
+      "loss": 0.7482,
+      "step": 5225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 34.8149833238631,
+      "learning_rate": 1.708089879436482e-05,
+      "loss": 0.8486,
+      "step": 5226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 28.966615028265814,
+      "learning_rate": 1.707970788973407e-05,
+      "loss": 0.759,
+      "step": 5227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 29.369644569659545,
+      "learning_rate": 1.7078516783760568e-05,
+      "loss": 0.7164,
+      "step": 5228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 32.03933218986899,
+      "learning_rate": 1.707732547647819e-05,
+      "loss": 0.8329,
+      "step": 5229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 62.482838769425605,
+      "learning_rate": 1.7076133967920813e-05,
+      "loss": 0.7755,
+      "step": 5230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 37.32115271095438,
+      "learning_rate": 1.7074942258122326e-05,
+      "loss": 0.7933,
+      "step": 5231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 45.77703643260786,
+      "learning_rate": 1.707375034711662e-05,
+      "loss": 0.8616,
+      "step": 5232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 49.44075969843824,
+      "learning_rate": 1.7072558234937597e-05,
+      "loss": 0.7077,
+      "step": 5233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 48.253222847070944,
+      "learning_rate": 1.707136592161915e-05,
+      "loss": 0.8954,
+      "step": 5234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 39.030517540710335,
+      "learning_rate": 1.7070173407195193e-05,
+      "loss": 0.8276,
+      "step": 5235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 35.22192568199553,
+      "learning_rate": 1.7068980691699638e-05,
+      "loss": 0.9077,
+      "step": 5236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 34.37538858367459,
+      "learning_rate": 1.7067787775166414e-05,
+      "loss": 0.7523,
+      "step": 5237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 53.99934190772972,
+      "learning_rate": 1.7066594657629435e-05,
+      "loss": 0.9374,
+      "step": 5238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 34.47292521515005,
+      "learning_rate": 1.706540133912264e-05,
+      "loss": 0.8517,
+      "step": 5239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 43.53765894258167,
+      "learning_rate": 1.7064207819679964e-05,
+      "loss": 0.8128,
+      "step": 5240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 36.613340898093014,
+      "learning_rate": 1.7063014099335353e-05,
+      "loss": 0.7507,
+      "step": 5241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 32.23269346421295,
+      "learning_rate": 1.7061820178122746e-05,
+      "loss": 0.7759,
+      "step": 5242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 32.99519338787257,
+      "learning_rate": 1.7060626056076107e-05,
+      "loss": 0.7905,
+      "step": 5243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 35.48947633653072,
+      "learning_rate": 1.705943173322939e-05,
+      "loss": 0.751,
+      "step": 5244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 43.96125618165092,
+      "learning_rate": 1.705823720961657e-05,
+      "loss": 0.8356,
+      "step": 5245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 38.396425609329235,
+      "learning_rate": 1.705704248527161e-05,
+      "loss": 0.847,
+      "step": 5246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 42.906745821650325,
+      "learning_rate": 1.7055847560228495e-05,
+      "loss": 0.8895,
+      "step": 5247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 41.56280268508717,
+      "learning_rate": 1.7054652434521197e-05,
+      "loss": 0.7695,
+      "step": 5248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 35.22647910278492,
+      "learning_rate": 1.7053457108183713e-05,
+      "loss": 0.7263,
+      "step": 5249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 31.764753525803144,
+      "learning_rate": 1.7052261581250034e-05,
+      "loss": 0.6884,
+      "step": 5250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 34.62692928275734,
+      "learning_rate": 1.705106585375416e-05,
+      "loss": 0.7463,
+      "step": 5251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 80.47370354081235,
+      "learning_rate": 1.70498699257301e-05,
+      "loss": 0.7815,
+      "step": 5252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.48064635036863,
+      "learning_rate": 1.7048673797211863e-05,
+      "loss": 0.735,
+      "step": 5253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 59.12570240821319,
+      "learning_rate": 1.7047477468233467e-05,
+      "loss": 0.7559,
+      "step": 5254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 36.797034452380394,
+      "learning_rate": 1.7046280938828932e-05,
+      "loss": 0.8432,
+      "step": 5255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 43.10801920844877,
+      "learning_rate": 1.704508420903229e-05,
+      "loss": 0.8349,
+      "step": 5256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 25.986117154898917,
+      "learning_rate": 1.7043887278877577e-05,
+      "loss": 0.7099,
+      "step": 5257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 38.471747668163026,
+      "learning_rate": 1.7042690148398825e-05,
+      "loss": 0.8477,
+      "step": 5258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 42.41236034497362,
+      "learning_rate": 1.704149281763009e-05,
+      "loss": 0.812,
+      "step": 5259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 38.28367333040455,
+      "learning_rate": 1.7040295286605415e-05,
+      "loss": 0.8064,
+      "step": 5260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 34.491078679718946,
+      "learning_rate": 1.7039097555358862e-05,
+      "loss": 0.7304,
+      "step": 5261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.36557750350174,
+      "learning_rate": 1.7037899623924495e-05,
+      "loss": 0.7727,
+      "step": 5262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 29.810570272533845,
+      "learning_rate": 1.7036701492336378e-05,
+      "loss": 0.7028,
+      "step": 5263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 39.93528777250336,
+      "learning_rate": 1.7035503160628585e-05,
+      "loss": 0.8923,
+      "step": 5264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.79513350614886,
+      "learning_rate": 1.7034304628835198e-05,
+      "loss": 0.7924,
+      "step": 5265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 41.67048663875688,
+      "learning_rate": 1.70331058969903e-05,
+      "loss": 0.7773,
+      "step": 5266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 32.27195198151114,
+      "learning_rate": 1.703190696512799e-05,
+      "loss": 0.7451,
+      "step": 5267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.63686241373506,
+      "learning_rate": 1.703070783328236e-05,
+      "loss": 0.795,
+      "step": 5268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 45.12274926191526,
+      "learning_rate": 1.702950850148751e-05,
+      "loss": 0.8828,
+      "step": 5269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 40.58101325613033,
+      "learning_rate": 1.7028308969777548e-05,
+      "loss": 0.8033,
+      "step": 5270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 36.27718796115884,
+      "learning_rate": 1.7027109238186593e-05,
+      "loss": 0.8674,
+      "step": 5271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 45.6320194070318,
+      "learning_rate": 1.7025909306748757e-05,
+      "loss": 0.8689,
+      "step": 5272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 39.80940846008241,
+      "learning_rate": 1.7024709175498174e-05,
+      "loss": 0.7366,
+      "step": 5273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 39.188400678003404,
+      "learning_rate": 1.7023508844468974e-05,
+      "loss": 0.8305,
+      "step": 5274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 43.62156929364274,
+      "learning_rate": 1.702230831369529e-05,
+      "loss": 0.7581,
+      "step": 5275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 37.8301047744984,
+      "learning_rate": 1.7021107583211266e-05,
+      "loss": 0.8204,
+      "step": 5276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 29.192559324502998,
+      "learning_rate": 1.7019906653051045e-05,
+      "loss": 0.6642,
+      "step": 5277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 36.25905884974198,
+      "learning_rate": 1.701870552324879e-05,
+      "loss": 0.8098,
+      "step": 5278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 40.98559275892947,
+      "learning_rate": 1.7017504193838654e-05,
+      "loss": 0.7627,
+      "step": 5279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 29.28559481325602,
+      "learning_rate": 1.7016302664854802e-05,
+      "loss": 0.7186,
+      "step": 5280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 38.75514431308044,
+      "learning_rate": 1.701510093633141e-05,
+      "loss": 0.7734,
+      "step": 5281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 34.1294270462935,
+      "learning_rate": 1.701389900830265e-05,
+      "loss": 0.8058,
+      "step": 5282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 33.38356647671118,
+      "learning_rate": 1.70126968808027e-05,
+      "loss": 0.7097,
+      "step": 5283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 39.81801073630219,
+      "learning_rate": 1.7011494553865762e-05,
+      "loss": 0.7969,
+      "step": 5284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.13741819474422,
+      "learning_rate": 1.7010292027526016e-05,
+      "loss": 0.6976,
+      "step": 5285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 43.75903646606092,
+      "learning_rate": 1.7009089301817666e-05,
+      "loss": 0.9417,
+      "step": 5286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.72631336678354,
+      "learning_rate": 1.7007886376774917e-05,
+      "loss": 0.7855,
+      "step": 5287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 40.12647038733625,
+      "learning_rate": 1.7006683252431977e-05,
+      "loss": 0.8791,
+      "step": 5288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 36.99222569690251,
+      "learning_rate": 1.7005479928823066e-05,
+      "loss": 0.87,
+      "step": 5289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 38.97186625363324,
+      "learning_rate": 1.7004276405982404e-05,
+      "loss": 0.7702,
+      "step": 5290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.43537704056181,
+      "learning_rate": 1.7003072683944214e-05,
+      "loss": 0.8178,
+      "step": 5291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 39.32524233208045,
+      "learning_rate": 1.700186876274274e-05,
+      "loss": 0.9265,
+      "step": 5292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 33.883905285397006,
+      "learning_rate": 1.700066464241221e-05,
+      "loss": 0.7144,
+      "step": 5293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 29.663528883652713,
+      "learning_rate": 1.6999460322986876e-05,
+      "loss": 0.7464,
+      "step": 5294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.731026485948114,
+      "learning_rate": 1.6998255804500983e-05,
+      "loss": 0.6296,
+      "step": 5295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 41.10157834175818,
+      "learning_rate": 1.699705108698879e-05,
+      "loss": 0.9089,
+      "step": 5296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.58316007091022,
+      "learning_rate": 1.6995846170484554e-05,
+      "loss": 0.8098,
+      "step": 5297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.39902640825038,
+      "learning_rate": 1.699464105502255e-05,
+      "loss": 0.8216,
+      "step": 5298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.650818251657135,
+      "learning_rate": 1.6993435740637042e-05,
+      "loss": 0.7667,
+      "step": 5299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.3924312509563,
+      "learning_rate": 1.6992230227362317e-05,
+      "loss": 0.7349,
+      "step": 5300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 33.33332924644127,
+      "learning_rate": 1.6991024515232652e-05,
+      "loss": 0.7731,
+      "step": 5301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 29.23681328298516,
+      "learning_rate": 1.698981860428234e-05,
+      "loss": 0.7736,
+      "step": 5302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 34.38810043046496,
+      "learning_rate": 1.6988612494545673e-05,
+      "loss": 0.841,
+      "step": 5303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 43.634853310529145,
+      "learning_rate": 1.698740618605696e-05,
+      "loss": 0.9113,
+      "step": 5304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 39.41484324624311,
+      "learning_rate": 1.69861996788505e-05,
+      "loss": 0.8088,
+      "step": 5305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.61871517465187,
+      "learning_rate": 1.6984992972960606e-05,
+      "loss": 0.7427,
+      "step": 5306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 29.104672625974047,
+      "learning_rate": 1.69837860684216e-05,
+      "loss": 0.6272,
+      "step": 5307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.618553390074148,
+      "learning_rate": 1.6982578965267805e-05,
+      "loss": 0.6919,
+      "step": 5308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.0169802790262,
+      "learning_rate": 1.6981371663533543e-05,
+      "loss": 0.8639,
+      "step": 5309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.12387141980176,
+      "learning_rate": 1.698016416325316e-05,
+      "loss": 0.8164,
+      "step": 5310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 37.39327779908822,
+      "learning_rate": 1.6978956464460993e-05,
+      "loss": 0.7621,
+      "step": 5311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 40.84844124200145,
+      "learning_rate": 1.6977748567191384e-05,
+      "loss": 0.789,
+      "step": 5312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 33.87427039961903,
+      "learning_rate": 1.697654047147869e-05,
+      "loss": 0.7096,
+      "step": 5313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 38.59662424728334,
+      "learning_rate": 1.6975332177357258e-05,
+      "loss": 0.8119,
+      "step": 5314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 37.4175503646951,
+      "learning_rate": 1.697412368486147e-05,
+      "loss": 0.7714,
+      "step": 5315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.623780374180644,
+      "learning_rate": 1.697291499402568e-05,
+      "loss": 0.7345,
+      "step": 5316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 39.955245855713784,
+      "learning_rate": 1.6971706104884263e-05,
+      "loss": 0.9608,
+      "step": 5317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.76248702852539,
+      "learning_rate": 1.6970497017471608e-05,
+      "loss": 0.8766,
+      "step": 5318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 43.37530421004678,
+      "learning_rate": 1.6969287731822092e-05,
+      "loss": 0.8403,
+      "step": 5319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.977600351095187,
+      "learning_rate": 1.6968078247970116e-05,
+      "loss": 0.7474,
+      "step": 5320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 36.711469978843226,
+      "learning_rate": 1.6966868565950062e-05,
+      "loss": 0.7442,
+      "step": 5321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 39.44940415163268,
+      "learning_rate": 1.6965658685796347e-05,
+      "loss": 0.8814,
+      "step": 5322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.182461928438705,
+      "learning_rate": 1.6964448607543375e-05,
+      "loss": 0.8549,
+      "step": 5323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 55.32966908517096,
+      "learning_rate": 1.696323833122555e-05,
+      "loss": 0.7637,
+      "step": 5324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 41.98645629096656,
+      "learning_rate": 1.696202785687731e-05,
+      "loss": 0.8651,
+      "step": 5325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 41.7142194595296,
+      "learning_rate": 1.6960817184533068e-05,
+      "loss": 0.8926,
+      "step": 5326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 45.98738916274346,
+      "learning_rate": 1.6959606314227254e-05,
+      "loss": 0.8471,
+      "step": 5327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 39.18336994973152,
+      "learning_rate": 1.695839524599431e-05,
+      "loss": 0.9136,
+      "step": 5328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.74264477943489,
+      "learning_rate": 1.6957183979868673e-05,
+      "loss": 0.7551,
+      "step": 5329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.275834846350264,
+      "learning_rate": 1.6955972515884798e-05,
+      "loss": 0.8129,
+      "step": 5330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 47.25363402041072,
+      "learning_rate": 1.695476085407713e-05,
+      "loss": 0.8215,
+      "step": 5331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.417543033153006,
+      "learning_rate": 1.695354899448013e-05,
+      "loss": 0.6471,
+      "step": 5332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 37.866937540880905,
+      "learning_rate": 1.695233693712827e-05,
+      "loss": 0.7623,
+      "step": 5333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.603475332190243,
+      "learning_rate": 1.695112468205601e-05,
+      "loss": 0.8139,
+      "step": 5334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 29.012768713196895,
+      "learning_rate": 1.6949912229297834e-05,
+      "loss": 0.7391,
+      "step": 5335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.79841744651831,
+      "learning_rate": 1.6948699578888218e-05,
+      "loss": 0.8241,
+      "step": 5336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.53745649852301,
+      "learning_rate": 1.6947486730861652e-05,
+      "loss": 0.6983,
+      "step": 5337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 34.84722181395112,
+      "learning_rate": 1.694627368525263e-05,
+      "loss": 0.792,
+      "step": 5338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 34.81021932188735,
+      "learning_rate": 1.6945060442095642e-05,
+      "loss": 0.7681,
+      "step": 5339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.544710002823233,
+      "learning_rate": 1.69438470014252e-05,
+      "loss": 0.7556,
+      "step": 5340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 34.23254682847062,
+      "learning_rate": 1.6942633363275812e-05,
+      "loss": 0.6796,
+      "step": 5341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 36.5634586078146,
+      "learning_rate": 1.694141952768199e-05,
+      "loss": 0.8106,
+      "step": 5342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 38.34684350142461,
+      "learning_rate": 1.694020549467826e-05,
+      "loss": 0.7848,
+      "step": 5343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.25594004326694,
+      "learning_rate": 1.6938991264299145e-05,
+      "loss": 0.7741,
+      "step": 5344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 34.008113892172936,
+      "learning_rate": 1.693777683657918e-05,
+      "loss": 0.8162,
+      "step": 5345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.418250733736567,
+      "learning_rate": 1.6936562211552895e-05,
+      "loss": 0.6992,
+      "step": 5346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 44.89383338432954,
+      "learning_rate": 1.693534738925484e-05,
+      "loss": 0.9204,
+      "step": 5347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.97744234298571,
+      "learning_rate": 1.6934132369719563e-05,
+      "loss": 0.7225,
+      "step": 5348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.49599486934203,
+      "learning_rate": 1.693291715298162e-05,
+      "loss": 0.7465,
+      "step": 5349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.534946420219594,
+      "learning_rate": 1.6931701739075568e-05,
+      "loss": 0.6916,
+      "step": 5350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 36.68917757712148,
+      "learning_rate": 1.6930486128035974e-05,
+      "loss": 0.7781,
+      "step": 5351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 39.12042026406659,
+      "learning_rate": 1.6929270319897407e-05,
+      "loss": 0.695,
+      "step": 5352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.2130147421744,
+      "learning_rate": 1.6928054314694443e-05,
+      "loss": 0.7025,
+      "step": 5353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.22130112348361,
+      "learning_rate": 1.6926838112461668e-05,
+      "loss": 0.8801,
+      "step": 5354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.22412882098711,
+      "learning_rate": 1.692562171323367e-05,
+      "loss": 0.8201,
+      "step": 5355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 45.27376743357707,
+      "learning_rate": 1.6924405117045046e-05,
+      "loss": 0.8974,
+      "step": 5356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.19840774050565,
+      "learning_rate": 1.692318832393039e-05,
+      "loss": 0.8082,
+      "step": 5357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.84488766634932,
+      "learning_rate": 1.6921971333924304e-05,
+      "loss": 0.8581,
+      "step": 5358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.27528691513623,
+      "learning_rate": 1.6920754147061406e-05,
+      "loss": 0.6787,
+      "step": 5359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.953400847123383,
+      "learning_rate": 1.691953676337631e-05,
+      "loss": 0.7487,
+      "step": 5360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 36.81917615756225,
+      "learning_rate": 1.691831918290363e-05,
+      "loss": 0.8407,
+      "step": 5361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.190812280931763,
+      "learning_rate": 1.6917101405678004e-05,
+      "loss": 0.7507,
+      "step": 5362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 43.70750198567611,
+      "learning_rate": 1.691588343173406e-05,
+      "loss": 0.7727,
+      "step": 5363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 40.042691664555285,
+      "learning_rate": 1.6914665261106434e-05,
+      "loss": 0.8728,
+      "step": 5364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 38.07631753256235,
+      "learning_rate": 1.691344689382978e-05,
+      "loss": 0.7254,
+      "step": 5365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 39.4665103134933,
+      "learning_rate": 1.6912228329938734e-05,
+      "loss": 0.8307,
+      "step": 5366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 40.02809840890962,
+      "learning_rate": 1.691100956946796e-05,
+      "loss": 0.7992,
+      "step": 5367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.366018565537175,
+      "learning_rate": 1.6909790612452114e-05,
+      "loss": 0.761,
+      "step": 5368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 37.467929796113395,
+      "learning_rate": 1.690857145892587e-05,
+      "loss": 0.7798,
+      "step": 5369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.282123168617225,
+      "learning_rate": 1.6907352108923892e-05,
+      "loss": 0.6965,
+      "step": 5370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 32.53540197987939,
+      "learning_rate": 1.690613256248086e-05,
+      "loss": 0.7451,
+      "step": 5371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 37.38164410175297,
+      "learning_rate": 1.690491281963146e-05,
+      "loss": 0.8494,
+      "step": 5372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.12547186796387,
+      "learning_rate": 1.690369288041038e-05,
+      "loss": 0.7525,
+      "step": 5373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.68303343818953,
+      "learning_rate": 1.690247274485231e-05,
+      "loss": 0.7053,
+      "step": 5374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 34.47678751540344,
+      "learning_rate": 1.690125241299195e-05,
+      "loss": 0.7203,
+      "step": 5375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 58.537522028304494,
+      "learning_rate": 1.6900031884864013e-05,
+      "loss": 0.8031,
+      "step": 5376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 43.018665784060325,
+      "learning_rate": 1.6898811160503204e-05,
+      "loss": 0.7735,
+      "step": 5377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.94845048910579,
+      "learning_rate": 1.6897590239944242e-05,
+      "loss": 0.6295,
+      "step": 5378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 34.95670274031741,
+      "learning_rate": 1.6896369123221852e-05,
+      "loss": 0.8347,
+      "step": 5379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 55.179475436342564,
+      "learning_rate": 1.6895147810370755e-05,
+      "loss": 0.8099,
+      "step": 5380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 32.98724684507912,
+      "learning_rate": 1.6893926301425685e-05,
+      "loss": 0.819,
+      "step": 5381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 46.06853489052516,
+      "learning_rate": 1.6892704596421386e-05,
+      "loss": 0.7988,
+      "step": 5382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 49.057975818002426,
+      "learning_rate": 1.6891482695392603e-05,
+      "loss": 0.8518,
+      "step": 5383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 34.22977702517022,
+      "learning_rate": 1.6890260598374083e-05,
+      "loss": 0.8008,
+      "step": 5384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 38.75115387336856,
+      "learning_rate": 1.688903830540058e-05,
+      "loss": 0.8891,
+      "step": 5385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 69.95643884965102,
+      "learning_rate": 1.6887815816506858e-05,
+      "loss": 0.7951,
+      "step": 5386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.84579544015169,
+      "learning_rate": 1.6886593131727687e-05,
+      "loss": 0.7375,
+      "step": 5387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 51.921660954059746,
+      "learning_rate": 1.688537025109783e-05,
+      "loss": 0.7466,
+      "step": 5388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 41.71893081911598,
+      "learning_rate": 1.6884147174652077e-05,
+      "loss": 0.8046,
+      "step": 5389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 30.754478401783782,
+      "learning_rate": 1.68829239024252e-05,
+      "loss": 0.7543,
+      "step": 5390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 37.88287955805468,
+      "learning_rate": 1.6881700434451996e-05,
+      "loss": 0.8421,
+      "step": 5391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 43.40401837219635,
+      "learning_rate": 1.6880476770767256e-05,
+      "loss": 0.8723,
+      "step": 5392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.484047350969284,
+      "learning_rate": 1.6879252911405782e-05,
+      "loss": 0.7354,
+      "step": 5393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 32.33323175354532,
+      "learning_rate": 1.6878028856402382e-05,
+      "loss": 0.8456,
+      "step": 5394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 34.48182977877295,
+      "learning_rate": 1.6876804605791864e-05,
+      "loss": 0.7497,
+      "step": 5395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 26.621067920876595,
+      "learning_rate": 1.6875580159609044e-05,
+      "loss": 0.7739,
+      "step": 5396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.68423987160165,
+      "learning_rate": 1.6874355517888747e-05,
+      "loss": 0.7413,
+      "step": 5397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 56.20320713523991,
+      "learning_rate": 1.6873130680665798e-05,
+      "loss": 0.7332,
+      "step": 5398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 33.93167126728356,
+      "learning_rate": 1.6871905647975038e-05,
+      "loss": 0.6839,
+      "step": 5399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 32.35863319405068,
+      "learning_rate": 1.6870680419851297e-05,
+      "loss": 0.7623,
+      "step": 5400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.27414337567245,
+      "learning_rate": 1.686945499632943e-05,
+      "loss": 0.7128,
+      "step": 5401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 38.865894806034184,
+      "learning_rate": 1.6868229377444276e-05,
+      "loss": 0.9048,
+      "step": 5402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 31.968487705858813,
+      "learning_rate": 1.6867003563230697e-05,
+      "loss": 0.7114,
+      "step": 5403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.52907663211665,
+      "learning_rate": 1.6865777553723558e-05,
+      "loss": 0.8599,
+      "step": 5404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.21536197968653,
+      "learning_rate": 1.6864551348957717e-05,
+      "loss": 0.7664,
+      "step": 5405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.780210234673003,
+      "learning_rate": 1.686332494896805e-05,
+      "loss": 0.7597,
+      "step": 5406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.99297395403117,
+      "learning_rate": 1.686209835378944e-05,
+      "loss": 0.8041,
+      "step": 5407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 39.306196580736206,
+      "learning_rate": 1.6860871563456766e-05,
+      "loss": 0.7178,
+      "step": 5408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 37.90109723598902,
+      "learning_rate": 1.685964457800492e-05,
+      "loss": 0.7447,
+      "step": 5409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 42.68849408289402,
+      "learning_rate": 1.6858417397468792e-05,
+      "loss": 0.7398,
+      "step": 5410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 35.851099938579715,
+      "learning_rate": 1.6857190021883287e-05,
+      "loss": 0.7375,
+      "step": 5411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 43.83227481881455,
+      "learning_rate": 1.685596245128331e-05,
+      "loss": 0.8272,
+      "step": 5412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.021408396471074,
+      "learning_rate": 1.685473468570377e-05,
+      "loss": 0.791,
+      "step": 5413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 33.237730135771905,
+      "learning_rate": 1.6853506725179584e-05,
+      "loss": 0.8194,
+      "step": 5414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.92364394090205,
+      "learning_rate": 1.6852278569745678e-05,
+      "loss": 0.8121,
+      "step": 5415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.33534277226855,
+      "learning_rate": 1.685105021943698e-05,
+      "loss": 0.7629,
+      "step": 5416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 41.88358546253362,
+      "learning_rate": 1.6849821674288418e-05,
+      "loss": 0.7765,
+      "step": 5417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 36.73556771039734,
+      "learning_rate": 1.6848592934334934e-05,
+      "loss": 0.8525,
+      "step": 5418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.160586419287846,
+      "learning_rate": 1.6847363999611475e-05,
+      "loss": 0.8016,
+      "step": 5419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 34.55043741920916,
+      "learning_rate": 1.6846134870152987e-05,
+      "loss": 0.7396,
+      "step": 5420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 52.29502157363862,
+      "learning_rate": 1.684490554599443e-05,
+      "loss": 0.887,
+      "step": 5421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 34.94435513006042,
+      "learning_rate": 1.6843676027170764e-05,
+      "loss": 0.811,
+      "step": 5422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.59822723034376,
+      "learning_rate": 1.6842446313716957e-05,
+      "loss": 0.7808,
+      "step": 5423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.18019373758243,
+      "learning_rate": 1.6841216405667976e-05,
+      "loss": 0.7762,
+      "step": 5424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.859356876802554,
+      "learning_rate": 1.6839986303058803e-05,
+      "loss": 0.6843,
+      "step": 5425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 44.555653296231576,
+      "learning_rate": 1.6838756005924425e-05,
+      "loss": 0.8405,
+      "step": 5426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 47.12167518685176,
+      "learning_rate": 1.6837525514299823e-05,
+      "loss": 0.8201,
+      "step": 5427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.852569053007784,
+      "learning_rate": 1.6836294828219997e-05,
+      "loss": 0.747,
+      "step": 5428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.83947884155379,
+      "learning_rate": 1.6835063947719943e-05,
+      "loss": 0.7375,
+      "step": 5429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.91299999431539,
+      "learning_rate": 1.683383287283467e-05,
+      "loss": 0.8885,
+      "step": 5430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 42.367222398521626,
+      "learning_rate": 1.683260160359919e-05,
+      "loss": 0.7285,
+      "step": 5431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 35.134170092038026,
+      "learning_rate": 1.6831370140048513e-05,
+      "loss": 0.7644,
+      "step": 5432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.600119439780904,
+      "learning_rate": 1.6830138482217667e-05,
+      "loss": 0.7258,
+      "step": 5433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 39.82920317539256,
+      "learning_rate": 1.6828906630141678e-05,
+      "loss": 0.8123,
+      "step": 5434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 43.44248963932753,
+      "learning_rate": 1.682767458385558e-05,
+      "loss": 0.8552,
+      "step": 5435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.974290347184002,
+      "learning_rate": 1.682644234339441e-05,
+      "loss": 0.8357,
+      "step": 5436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 35.366863477607524,
+      "learning_rate": 1.6825209908793217e-05,
+      "loss": 0.78,
+      "step": 5437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 38.15309426194505,
+      "learning_rate": 1.682397728008704e-05,
+      "loss": 0.7867,
+      "step": 5438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.78893537580816,
+      "learning_rate": 1.6822744457310948e-05,
+      "loss": 0.8161,
+      "step": 5439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 29.88552169187451,
+      "learning_rate": 1.682151144049999e-05,
+      "loss": 0.6269,
+      "step": 5440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.291620051589913,
+      "learning_rate": 1.682027822968924e-05,
+      "loss": 0.8028,
+      "step": 5441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 36.097825669880855,
+      "learning_rate": 1.6819044824913762e-05,
+      "loss": 0.7613,
+      "step": 5442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 33.724446724947754,
+      "learning_rate": 1.6817811226208643e-05,
+      "loss": 0.6991,
+      "step": 5443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.07711634863188,
+      "learning_rate": 1.681657743360896e-05,
+      "loss": 0.7184,
+      "step": 5444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 39.49819238063704,
+      "learning_rate": 1.6815343447149803e-05,
+      "loss": 0.7668,
+      "step": 5445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 41.55313242846027,
+      "learning_rate": 1.681410926686626e-05,
+      "loss": 0.8217,
+      "step": 5446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 34.62552870336449,
+      "learning_rate": 1.6812874892793443e-05,
+      "loss": 0.7342,
+      "step": 5447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 37.02587609552094,
+      "learning_rate": 1.6811640324966446e-05,
+      "loss": 0.7055,
+      "step": 5448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 41.151974863681154,
+      "learning_rate": 1.681040556342038e-05,
+      "loss": 0.7317,
+      "step": 5449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 44.29018847611668,
+      "learning_rate": 1.680917060819037e-05,
+      "loss": 0.8103,
+      "step": 5450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 36.54570028921404,
+      "learning_rate": 1.6807935459311528e-05,
+      "loss": 0.762,
+      "step": 5451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 33.166529436322755,
+      "learning_rate": 1.6806700116818982e-05,
+      "loss": 0.7938,
+      "step": 5452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.595008793028672,
+      "learning_rate": 1.680546458074787e-05,
+      "loss": 0.7838,
+      "step": 5453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 29.053828381368366,
+      "learning_rate": 1.6804228851133326e-05,
+      "loss": 0.7058,
+      "step": 5454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 48.22107248405464,
+      "learning_rate": 1.6802992928010496e-05,
+      "loss": 0.7963,
+      "step": 5455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.466448141877507,
+      "learning_rate": 1.680175681141452e-05,
+      "loss": 0.8258,
+      "step": 5456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 36.60253420978055,
+      "learning_rate": 1.6800520501380564e-05,
+      "loss": 0.7839,
+      "step": 5457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.320430294959465,
+      "learning_rate": 1.679928399794378e-05,
+      "loss": 0.8466,
+      "step": 5458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.899605159181263,
+      "learning_rate": 1.6798047301139338e-05,
+      "loss": 0.6675,
+      "step": 5459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.811566452833492,
+      "learning_rate": 1.679681041100241e-05,
+      "loss": 0.6973,
+      "step": 5460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 33.15980240022687,
+      "learning_rate": 1.6795573327568168e-05,
+      "loss": 0.8349,
+      "step": 5461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 43.09817856128369,
+      "learning_rate": 1.6794336050871797e-05,
+      "loss": 0.9267,
+      "step": 5462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 39.11192564318563,
+      "learning_rate": 1.6793098580948482e-05,
+      "loss": 0.9164,
+      "step": 5463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 40.75033970556662,
+      "learning_rate": 1.6791860917833417e-05,
+      "loss": 0.7594,
+      "step": 5464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 36.888679230112075,
+      "learning_rate": 1.67906230615618e-05,
+      "loss": 0.8055,
+      "step": 5465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 38.72089062639324,
+      "learning_rate": 1.6789385012168836e-05,
+      "loss": 0.8104,
+      "step": 5466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 49.2203536822924,
+      "learning_rate": 1.6788146769689734e-05,
+      "loss": 0.8168,
+      "step": 5467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 38.81347572095349,
+      "learning_rate": 1.678690833415971e-05,
+      "loss": 0.6945,
+      "step": 5468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 48.67380964864246,
+      "learning_rate": 1.678566970561398e-05,
+      "loss": 0.815,
+      "step": 5469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 38.34495363468198,
+      "learning_rate": 1.678443088408778e-05,
+      "loss": 0.809,
+      "step": 5470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.08560194792557,
+      "learning_rate": 1.6783191869616327e-05,
+      "loss": 0.7828,
+      "step": 5471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 38.37894528141901,
+      "learning_rate": 1.678195266223487e-05,
+      "loss": 0.8568,
+      "step": 5472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.190140790117063,
+      "learning_rate": 1.6780713261978646e-05,
+      "loss": 0.7277,
+      "step": 5473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.07525026586066,
+      "learning_rate": 1.67794736688829e-05,
+      "loss": 0.8198,
+      "step": 5474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.407922678492966,
+      "learning_rate": 1.6778233882982894e-05,
+      "loss": 0.8674,
+      "step": 5475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.70909897972485,
+      "learning_rate": 1.6776993904313875e-05,
+      "loss": 0.7782,
+      "step": 5476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 46.821580609647036,
+      "learning_rate": 1.677575373291112e-05,
+      "loss": 0.8064,
+      "step": 5477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.065685914922504,
+      "learning_rate": 1.6774513368809887e-05,
+      "loss": 0.8624,
+      "step": 5478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.451848436468374,
+      "learning_rate": 1.677327281204546e-05,
+      "loss": 0.831,
+      "step": 5479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.14871378838998,
+      "learning_rate": 1.6772032062653115e-05,
+      "loss": 0.7864,
+      "step": 5480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 37.956698548210404,
+      "learning_rate": 1.677079112066814e-05,
+      "loss": 0.8305,
+      "step": 5481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 25.21835808053543,
+      "learning_rate": 1.6769549986125827e-05,
+      "loss": 0.6,
+      "step": 5482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.27839380647795,
+      "learning_rate": 1.6768308659061474e-05,
+      "loss": 0.6937,
+      "step": 5483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 31.65231116099106,
+      "learning_rate": 1.6767067139510383e-05,
+      "loss": 0.7532,
+      "step": 5484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.933747565271467,
+      "learning_rate": 1.6765825427507855e-05,
+      "loss": 0.6432,
+      "step": 5485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.254257846570454,
+      "learning_rate": 1.6764583523089214e-05,
+      "loss": 0.7954,
+      "step": 5486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.59711199763788,
+      "learning_rate": 1.6763341426289773e-05,
+      "loss": 0.8399,
+      "step": 5487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 61.117698639013895,
+      "learning_rate": 1.676209913714486e-05,
+      "loss": 0.8365,
+      "step": 5488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.042157633624235,
+      "learning_rate": 1.6760856655689804e-05,
+      "loss": 0.7089,
+      "step": 5489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 46.13837199744862,
+      "learning_rate": 1.675961398195994e-05,
+      "loss": 0.746,
+      "step": 5490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 42.23276652760159,
+      "learning_rate": 1.6758371115990607e-05,
+      "loss": 0.8004,
+      "step": 5491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 37.243271540042514,
+      "learning_rate": 1.6757128057817154e-05,
+      "loss": 0.8293,
+      "step": 5492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.298473373047045,
+      "learning_rate": 1.675588480747493e-05,
+      "loss": 0.8503,
+      "step": 5493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 39.284036860696034,
+      "learning_rate": 1.6754641364999297e-05,
+      "loss": 0.798,
+      "step": 5494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 42.82245322313615,
+      "learning_rate": 1.675339773042561e-05,
+      "loss": 0.7754,
+      "step": 5495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.60961711769292,
+      "learning_rate": 1.6752153903789247e-05,
+      "loss": 0.8353,
+      "step": 5496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 41.908794032854345,
+      "learning_rate": 1.6750909885125575e-05,
+      "loss": 0.8527,
+      "step": 5497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 35.23964694689539,
+      "learning_rate": 1.6749665674469975e-05,
+      "loss": 0.7557,
+      "step": 5498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.39780020472204,
+      "learning_rate": 1.6748421271857835e-05,
+      "loss": 0.7806,
+      "step": 5499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 34.313462478396396,
+      "learning_rate": 1.6747176677324535e-05,
+      "loss": 0.8144,
+      "step": 5500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 44.75392614222508,
+      "learning_rate": 1.674593189090548e-05,
+      "loss": 0.8198,
+      "step": 5501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.18888846387216,
+      "learning_rate": 1.674468691263607e-05,
+      "loss": 0.7034,
+      "step": 5502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.83827851396546,
+      "learning_rate": 1.674344174255171e-05,
+      "loss": 0.7856,
+      "step": 5503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 35.35663278880605,
+      "learning_rate": 1.674219638068781e-05,
+      "loss": 0.8366,
+      "step": 5504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.267722680200386,
+      "learning_rate": 1.6740950827079786e-05,
+      "loss": 0.7404,
+      "step": 5505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 43.880141769343595,
+      "learning_rate": 1.673970508176307e-05,
+      "loss": 0.843,
+      "step": 5506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 34.81686796868508,
+      "learning_rate": 1.673845914477308e-05,
+      "loss": 0.7826,
+      "step": 5507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.821757204001685,
+      "learning_rate": 1.6737213016145256e-05,
+      "loss": 0.7557,
+      "step": 5508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.914270724781225,
+      "learning_rate": 1.6735966695915036e-05,
+      "loss": 0.7507,
+      "step": 5509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.175792584614634,
+      "learning_rate": 1.673472018411786e-05,
+      "loss": 0.6683,
+      "step": 5510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 39.98280987316738,
+      "learning_rate": 1.6733473480789183e-05,
+      "loss": 0.8426,
+      "step": 5511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 35.29912479250429,
+      "learning_rate": 1.673222658596446e-05,
+      "loss": 0.6702,
+      "step": 5512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 50.158114622638195,
+      "learning_rate": 1.673097949967915e-05,
+      "loss": 0.8446,
+      "step": 5513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 47.20141327225191,
+      "learning_rate": 1.6729732221968722e-05,
+      "loss": 0.8095,
+      "step": 5514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.19505989114455,
+      "learning_rate": 1.6728484752868644e-05,
+      "loss": 0.8298,
+      "step": 5515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 39.10100274690467,
+      "learning_rate": 1.6727237092414397e-05,
+      "loss": 0.8163,
+      "step": 5516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 41.37141740622103,
+      "learning_rate": 1.6725989240641466e-05,
+      "loss": 0.8438,
+      "step": 5517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 30.829608496785664,
+      "learning_rate": 1.672474119758533e-05,
+      "loss": 0.7053,
+      "step": 5518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.01546153990779,
+      "learning_rate": 1.6723492963281492e-05,
+      "loss": 0.7737,
+      "step": 5519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.303883871485155,
+      "learning_rate": 1.6722244537765444e-05,
+      "loss": 0.6871,
+      "step": 5520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.39444658596558,
+      "learning_rate": 1.6720995921072698e-05,
+      "loss": 0.8417,
+      "step": 5521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 31.556661048746083,
+      "learning_rate": 1.6719747113238754e-05,
+      "loss": 0.678,
+      "step": 5522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 46.71237470831029,
+      "learning_rate": 1.6718498114299138e-05,
+      "loss": 0.799,
+      "step": 5523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 29.873430949071285,
+      "learning_rate": 1.6717248924289363e-05,
+      "loss": 0.6799,
+      "step": 5524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 35.994725059448015,
+      "learning_rate": 1.671599954324496e-05,
+      "loss": 0.7559,
+      "step": 5525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 63.52483306616625,
+      "learning_rate": 1.6714749971201457e-05,
+      "loss": 0.7798,
+      "step": 5526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 40.44194730626839,
+      "learning_rate": 1.6713500208194395e-05,
+      "loss": 0.7625,
+      "step": 5527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.4078691977721,
+      "learning_rate": 1.6712250254259313e-05,
+      "loss": 0.827,
+      "step": 5528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.77529602085356,
+      "learning_rate": 1.6711000109431757e-05,
+      "loss": 0.89,
+      "step": 5529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.366399323095145,
+      "learning_rate": 1.6709749773747292e-05,
+      "loss": 0.684,
+      "step": 5530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 38.75943889715687,
+      "learning_rate": 1.670849924724146e-05,
+      "loss": 0.8935,
+      "step": 5531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.27347856034948,
+      "learning_rate": 1.670724852994984e-05,
+      "loss": 0.9149,
+      "step": 5532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 31.666324725730913,
+      "learning_rate": 1.6705997621907993e-05,
+      "loss": 0.7788,
+      "step": 5533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 40.11955117279806,
+      "learning_rate": 1.6704746523151498e-05,
+      "loss": 0.7905,
+      "step": 5534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 34.77302768374081,
+      "learning_rate": 1.670349523371593e-05,
+      "loss": 0.6735,
+      "step": 5535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.71168149832154,
+      "learning_rate": 1.670224375363688e-05,
+      "loss": 0.7277,
+      "step": 5536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 37.3549253652994,
+      "learning_rate": 1.670099208294994e-05,
+      "loss": 0.8628,
+      "step": 5537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 57.74940779824081,
+      "learning_rate": 1.6699740221690706e-05,
+      "loss": 0.7262,
+      "step": 5538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 31.466520948362234,
+      "learning_rate": 1.669848816989478e-05,
+      "loss": 0.8024,
+      "step": 5539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 28.12201662134974,
+      "learning_rate": 1.6697235927597763e-05,
+      "loss": 0.7421,
+      "step": 5540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 31.76959367720379,
+      "learning_rate": 1.6695983494835283e-05,
+      "loss": 0.8552,
+      "step": 5541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 28.62172593212158,
+      "learning_rate": 1.669473087164294e-05,
+      "loss": 0.7372,
+      "step": 5542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 38.421853830673754,
+      "learning_rate": 1.6693478058056375e-05,
+      "loss": 0.8027,
+      "step": 5543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 41.73566529604843,
+      "learning_rate": 1.6692225054111207e-05,
+      "loss": 0.7263,
+      "step": 5544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 32.9772762392151,
+      "learning_rate": 1.669097185984307e-05,
+      "loss": 0.7679,
+      "step": 5545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.844002198161014,
+      "learning_rate": 1.668971847528761e-05,
+      "loss": 0.9172,
+      "step": 5546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 34.536597908822195,
+      "learning_rate": 1.668846490048047e-05,
+      "loss": 0.7548,
+      "step": 5547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 41.84503658444864,
+      "learning_rate": 1.6687211135457304e-05,
+      "loss": 0.839,
+      "step": 5548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 36.29222441039219,
+      "learning_rate": 1.668595718025376e-05,
+      "loss": 0.7487,
+      "step": 5549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.98077762857578,
+      "learning_rate": 1.6684703034905507e-05,
+      "loss": 0.8229,
+      "step": 5550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 39.38336803284348,
+      "learning_rate": 1.668344869944821e-05,
+      "loss": 0.8605,
+      "step": 5551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 32.37901281283361,
+      "learning_rate": 1.6682194173917543e-05,
+      "loss": 0.7467,
+      "step": 5552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 30.96645048748342,
+      "learning_rate": 1.6680939458349184e-05,
+      "loss": 0.7653,
+      "step": 5553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.45761453862763,
+      "learning_rate": 1.667968455277881e-05,
+      "loss": 0.7559,
+      "step": 5554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 32.09854119928493,
+      "learning_rate": 1.667842945724212e-05,
+      "loss": 0.7531,
+      "step": 5555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 34.9587225492602,
+      "learning_rate": 1.6677174171774798e-05,
+      "loss": 0.728,
+      "step": 5556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 26.867076466456556,
+      "learning_rate": 1.6675918696412552e-05,
+      "loss": 0.6782,
+      "step": 5557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 30.42233976255009,
+      "learning_rate": 1.6674663031191084e-05,
+      "loss": 0.7976,
+      "step": 5558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 40.100545505014374,
+      "learning_rate": 1.66734071761461e-05,
+      "loss": 0.7189,
+      "step": 5559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 44.34426008042682,
+      "learning_rate": 1.6672151131313324e-05,
+      "loss": 0.7713,
+      "step": 5560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 39.547947314799,
+      "learning_rate": 1.6670894896728472e-05,
+      "loss": 0.903,
+      "step": 5561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.898390199383115,
+      "learning_rate": 1.666963847242727e-05,
+      "loss": 0.7582,
+      "step": 5562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 34.513335522019815,
+      "learning_rate": 1.666838185844545e-05,
+      "loss": 0.6825,
+      "step": 5563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 32.702607480185925,
+      "learning_rate": 1.6667125054818756e-05,
+      "loss": 0.7641,
+      "step": 5564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 31.200888861624804,
+      "learning_rate": 1.6665868061582922e-05,
+      "loss": 0.7514,
+      "step": 5565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.48949438096133,
+      "learning_rate": 1.66646108787737e-05,
+      "loss": 0.7691,
+      "step": 5566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.095465739592306,
+      "learning_rate": 1.6663353506426845e-05,
+      "loss": 0.8169,
+      "step": 5567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 40.245157939787234,
+      "learning_rate": 1.6662095944578112e-05,
+      "loss": 0.7641,
+      "step": 5568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 44.52428136006793,
+      "learning_rate": 1.6660838193263268e-05,
+      "loss": 0.8345,
+      "step": 5569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 37.44349572724622,
+      "learning_rate": 1.6659580252518083e-05,
+      "loss": 0.7756,
+      "step": 5570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 44.93078704799243,
+      "learning_rate": 1.665832212237833e-05,
+      "loss": 0.8225,
+      "step": 5571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 40.45490501144662,
+      "learning_rate": 1.6657063802879795e-05,
+      "loss": 0.731,
+      "step": 5572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 36.44607205596041,
+      "learning_rate": 1.6655805294058256e-05,
+      "loss": 0.8114,
+      "step": 5573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 51.609406381447954,
+      "learning_rate": 1.665454659594951e-05,
+      "loss": 0.698,
+      "step": 5574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 51.785837429539136,
+      "learning_rate": 1.665328770858935e-05,
+      "loss": 0.771,
+      "step": 5575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.22463090288307,
+      "learning_rate": 1.665202863201358e-05,
+      "loss": 0.7587,
+      "step": 5576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 38.32861006842307,
+      "learning_rate": 1.6650769366258007e-05,
+      "loss": 0.7469,
+      "step": 5577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.28203083250641,
+      "learning_rate": 1.6649509911358444e-05,
+      "loss": 0.8405,
+      "step": 5578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 38.30598449067303,
+      "learning_rate": 1.6648250267350708e-05,
+      "loss": 0.8488,
+      "step": 5579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 44.573577140781666,
+      "learning_rate": 1.6646990434270625e-05,
+      "loss": 0.7671,
+      "step": 5580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 37.781845861110924,
+      "learning_rate": 1.664573041215402e-05,
+      "loss": 0.6999,
+      "step": 5581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 42.946458451136316,
+      "learning_rate": 1.6644470201036732e-05,
+      "loss": 0.722,
+      "step": 5582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.141789733798475,
+      "learning_rate": 1.6643209800954597e-05,
+      "loss": 0.7686,
+      "step": 5583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 49.77545365510497,
+      "learning_rate": 1.664194921194346e-05,
+      "loss": 0.7818,
+      "step": 5584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 97.61728123617996,
+      "learning_rate": 1.664068843403918e-05,
+      "loss": 0.8913,
+      "step": 5585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 29.893150801159482,
+      "learning_rate": 1.6639427467277597e-05,
+      "loss": 0.6874,
+      "step": 5586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 32.64128539429433,
+      "learning_rate": 1.663816631169459e-05,
+      "loss": 0.7725,
+      "step": 5587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 42.132045869733304,
+      "learning_rate": 1.663690496732601e-05,
+      "loss": 0.8387,
+      "step": 5588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 35.86717494664422,
+      "learning_rate": 1.663564343420773e-05,
+      "loss": 0.7645,
+      "step": 5589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 29.01950571529155,
+      "learning_rate": 1.6634381712375643e-05,
+      "loss": 0.6583,
+      "step": 5590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.71536529565451,
+      "learning_rate": 1.6633119801865617e-05,
+      "loss": 0.8313,
+      "step": 5591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 38.60702417334493,
+      "learning_rate": 1.6631857702713544e-05,
+      "loss": 0.8165,
+      "step": 5592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.610228356984784,
+      "learning_rate": 1.6630595414955314e-05,
+      "loss": 0.8144,
+      "step": 5593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 36.84354117142607,
+      "learning_rate": 1.6629332938626836e-05,
+      "loss": 0.7484,
+      "step": 5594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 45.222816690405985,
+      "learning_rate": 1.6628070273764002e-05,
+      "loss": 0.7382,
+      "step": 5595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 27.16785541129587,
+      "learning_rate": 1.6626807420402727e-05,
+      "loss": 0.7184,
+      "step": 5596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.33180754785543,
+      "learning_rate": 1.662554437857893e-05,
+      "loss": 0.8334,
+      "step": 5597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 36.26100522076626,
+      "learning_rate": 1.662428114832852e-05,
+      "loss": 0.7146,
+      "step": 5598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 45.40497679215052,
+      "learning_rate": 1.6623017729687432e-05,
+      "loss": 0.7505,
+      "step": 5599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 39.59774873889614,
+      "learning_rate": 1.6621754122691596e-05,
+      "loss": 0.8147,
+      "step": 5600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 36.31075660210405,
+      "learning_rate": 1.6620490327376947e-05,
+      "loss": 0.8481,
+      "step": 5601
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 39.70888193211187,
+      "learning_rate": 1.6619226343779424e-05,
+      "loss": 0.8138,
+      "step": 5602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 36.372233400300566,
+      "learning_rate": 1.6617962171934975e-05,
+      "loss": 0.7303,
+      "step": 5603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 38.52832293936467,
+      "learning_rate": 1.6616697811879553e-05,
+      "loss": 0.8548,
+      "step": 5604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 43.789130085085624,
+      "learning_rate": 1.661543326364912e-05,
+      "loss": 0.7266,
+      "step": 5605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 29.882315310661685,
+      "learning_rate": 1.661416852727963e-05,
+      "loss": 0.7876,
+      "step": 5606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 26.207253178128614,
+      "learning_rate": 1.661290360280706e-05,
+      "loss": 0.7688,
+      "step": 5607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.90703234909998,
+      "learning_rate": 1.6611638490267375e-05,
+      "loss": 0.9015,
+      "step": 5608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 28.17823414218221,
+      "learning_rate": 1.6610373189696565e-05,
+      "loss": 0.6845,
+      "step": 5609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 37.66352183933712,
+      "learning_rate": 1.6609107701130603e-05,
+      "loss": 0.7732,
+      "step": 5610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 54.13742747366924,
+      "learning_rate": 1.660784202460549e-05,
+      "loss": 0.8701,
+      "step": 5611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.836394178621504,
+      "learning_rate": 1.6606576160157212e-05,
+      "loss": 0.7616,
+      "step": 5612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 27.462838983158264,
+      "learning_rate": 1.6605310107821774e-05,
+      "loss": 0.6223,
+      "step": 5613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 30.467242906906336,
+      "learning_rate": 1.660404386763518e-05,
+      "loss": 0.7288,
+      "step": 5614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 31.776817318354688,
+      "learning_rate": 1.660277743963344e-05,
+      "loss": 0.7415,
+      "step": 5615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 44.06121112480549,
+      "learning_rate": 1.6601510823852574e-05,
+      "loss": 0.8893,
+      "step": 5616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 30.919000022591337,
+      "learning_rate": 1.66002440203286e-05,
+      "loss": 0.8251,
+      "step": 5617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 33.11464474602247,
+      "learning_rate": 1.659897702909755e-05,
+      "loss": 0.845,
+      "step": 5618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 41.925693003978246,
+      "learning_rate": 1.659770985019545e-05,
+      "loss": 0.8743,
+      "step": 5619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 40.876613736742996,
+      "learning_rate": 1.6596442483658345e-05,
+      "loss": 0.9045,
+      "step": 5620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.858962233740264,
+      "learning_rate": 1.6595174929522273e-05,
+      "loss": 0.7189,
+      "step": 5621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 28.387585336043752,
+      "learning_rate": 1.6593907187823284e-05,
+      "loss": 0.7873,
+      "step": 5622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.13072659926449,
+      "learning_rate": 1.6592639258597432e-05,
+      "loss": 0.7993,
+      "step": 5623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 44.62085139868248,
+      "learning_rate": 1.6591371141880776e-05,
+      "loss": 0.7218,
+      "step": 5624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.54678767075494,
+      "learning_rate": 1.6590102837709382e-05,
+      "loss": 0.8701,
+      "step": 5625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 37.18347596908175,
+      "learning_rate": 1.6588834346119315e-05,
+      "loss": 0.8308,
+      "step": 5626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 29.712151445680206,
+      "learning_rate": 1.6587565667146657e-05,
+      "loss": 0.8346,
+      "step": 5627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 37.244375277854324,
+      "learning_rate": 1.6586296800827483e-05,
+      "loss": 0.7916,
+      "step": 5628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 28.723230612220938,
+      "learning_rate": 1.6585027747197882e-05,
+      "loss": 0.7939,
+      "step": 5629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 35.71490016892329,
+      "learning_rate": 1.6583758506293945e-05,
+      "loss": 0.7917,
+      "step": 5630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 35.00350584795794,
+      "learning_rate": 1.6582489078151765e-05,
+      "loss": 0.7642,
+      "step": 5631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 30.155923594551428,
+      "learning_rate": 1.6581219462807444e-05,
+      "loss": 0.8217,
+      "step": 5632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.83618426101868,
+      "learning_rate": 1.6579949660297098e-05,
+      "loss": 0.7827,
+      "step": 5633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 36.74051559768097,
+      "learning_rate": 1.657867967065683e-05,
+      "loss": 0.7769,
+      "step": 5634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.45230054496029,
+      "learning_rate": 1.657740949392276e-05,
+      "loss": 0.7241,
+      "step": 5635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 31.176944762097325,
+      "learning_rate": 1.657613913013101e-05,
+      "loss": 0.7136,
+      "step": 5636
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.13805087443695,
+      "learning_rate": 1.657486857931771e-05,
+      "loss": 0.7228,
+      "step": 5637
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 28.813082004386867,
+      "learning_rate": 1.6573597841518995e-05,
+      "loss": 0.6801,
+      "step": 5638
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 33.74770908351136,
+      "learning_rate": 1.6572326916771008e-05,
+      "loss": 0.7926,
+      "step": 5639
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 65.09592491615037,
+      "learning_rate": 1.657105580510988e-05,
+      "loss": 0.8685,
+      "step": 5640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 45.79060323687749,
+      "learning_rate": 1.6569784506571772e-05,
+      "loss": 0.6388,
+      "step": 5641
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.734725556966055,
+      "learning_rate": 1.6568513021192837e-05,
+      "loss": 0.772,
+      "step": 5642
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 41.04427210030042,
+      "learning_rate": 1.6567241349009235e-05,
+      "loss": 0.9014,
+      "step": 5643
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 44.901629364848496,
+      "learning_rate": 1.6565969490057128e-05,
+      "loss": 0.9032,
+      "step": 5644
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 45.88761617676471,
+      "learning_rate": 1.656469744437269e-05,
+      "loss": 0.8356,
+      "step": 5645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 26.035046116337707,
+      "learning_rate": 1.65634252119921e-05,
+      "loss": 0.7565,
+      "step": 5646
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 35.505922045181045,
+      "learning_rate": 1.6562152792951534e-05,
+      "loss": 0.8634,
+      "step": 5647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.56495995116557,
+      "learning_rate": 1.656088018728718e-05,
+      "loss": 0.7339,
+      "step": 5648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 35.15689649538494,
+      "learning_rate": 1.6559607395035234e-05,
+      "loss": 0.7909,
+      "step": 5649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 35.71066735398505,
+      "learning_rate": 1.655833441623189e-05,
+      "loss": 0.8109,
+      "step": 5650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 33.67736180822884,
+      "learning_rate": 1.6557061250913352e-05,
+      "loss": 0.8614,
+      "step": 5651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.33178041274874,
+      "learning_rate": 1.6555787899115827e-05,
+      "loss": 0.6862,
+      "step": 5652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 40.41754648620873,
+      "learning_rate": 1.6554514360875528e-05,
+      "loss": 0.8337,
+      "step": 5653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 44.09576199467601,
+      "learning_rate": 1.655324063622868e-05,
+      "loss": 0.8279,
+      "step": 5654
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 29.947739365204892,
+      "learning_rate": 1.6551966725211497e-05,
+      "loss": 0.6711,
+      "step": 5655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 35.902941346344186,
+      "learning_rate": 1.6550692627860213e-05,
+      "loss": 0.8458,
+      "step": 5656
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.86584510757679,
+      "learning_rate": 1.6549418344211066e-05,
+      "loss": 0.7594,
+      "step": 5657
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 38.340953768471685,
+      "learning_rate": 1.6548143874300292e-05,
+      "loss": 0.8399,
+      "step": 5658
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.43166318282136,
+      "learning_rate": 1.654686921816413e-05,
+      "loss": 0.7824,
+      "step": 5659
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 31.412413854818627,
+      "learning_rate": 1.6545594375838846e-05,
+      "loss": 0.7995,
+      "step": 5660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.93909281817236,
+      "learning_rate": 1.6544319347360685e-05,
+      "loss": 0.8314,
+      "step": 5661
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 31.907659609720525,
+      "learning_rate": 1.6543044132765907e-05,
+      "loss": 0.744,
+      "step": 5662
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 41.53730800066555,
+      "learning_rate": 1.6541768732090784e-05,
+      "loss": 0.8646,
+      "step": 5663
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.477427603454686,
+      "learning_rate": 1.6540493145371582e-05,
+      "loss": 0.8736,
+      "step": 5664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 37.18409080585917,
+      "learning_rate": 1.6539217372644585e-05,
+      "loss": 0.7901,
+      "step": 5665
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 30.979239938098186,
+      "learning_rate": 1.653794141394607e-05,
+      "loss": 0.7237,
+      "step": 5666
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 32.868241351483,
+      "learning_rate": 1.6536665269312324e-05,
+      "loss": 0.7462,
+      "step": 5667
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 31.729412042007556,
+      "learning_rate": 1.6535388938779644e-05,
+      "loss": 0.6954,
+      "step": 5668
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 30.93352678390212,
+      "learning_rate": 1.6534112422384324e-05,
+      "loss": 0.7576,
+      "step": 5669
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 264.18563390884515,
+      "learning_rate": 1.653283572016267e-05,
+      "loss": 0.6366,
+      "step": 5670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 41.70880390146947,
+      "learning_rate": 1.653155883215099e-05,
+      "loss": 0.7955,
+      "step": 5671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 40.180886280587096,
+      "learning_rate": 1.6530281758385597e-05,
+      "loss": 0.7264,
+      "step": 5672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 30.356269508749822,
+      "learning_rate": 1.652900449890281e-05,
+      "loss": 0.7259,
+      "step": 5673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.49891418644811,
+      "learning_rate": 1.6527727053738957e-05,
+      "loss": 0.6441,
+      "step": 5674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 46.526996418131546,
+      "learning_rate": 1.6526449422930363e-05,
+      "loss": 0.8912,
+      "step": 5675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.085558218489595,
+      "learning_rate": 1.6525171606513368e-05,
+      "loss": 0.7951,
+      "step": 5676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 48.46695086093278,
+      "learning_rate": 1.6523893604524304e-05,
+      "loss": 0.7318,
+      "step": 5677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.43911903289044,
+      "learning_rate": 1.652261541699953e-05,
+      "loss": 0.7902,
+      "step": 5678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 43.92872560775319,
+      "learning_rate": 1.6521337043975384e-05,
+      "loss": 0.8021,
+      "step": 5679
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 42.8940484881559,
+      "learning_rate": 1.652005848548823e-05,
+      "loss": 0.7589,
+      "step": 5680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.33495944984067,
+      "learning_rate": 1.6518779741574422e-05,
+      "loss": 0.7753,
+      "step": 5681
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 54.06968859028868,
+      "learning_rate": 1.6517500812270335e-05,
+      "loss": 0.7749,
+      "step": 5682
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 38.372204187918015,
+      "learning_rate": 1.6516221697612338e-05,
+      "loss": 0.7054,
+      "step": 5683
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.37951302142681,
+      "learning_rate": 1.6514942397636807e-05,
+      "loss": 0.7548,
+      "step": 5684
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 40.62735973840738,
+      "learning_rate": 1.6513662912380123e-05,
+      "loss": 0.7877,
+      "step": 5685
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 48.16833062361216,
+      "learning_rate": 1.651238324187868e-05,
+      "loss": 0.8083,
+      "step": 5686
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 49.43003293539683,
+      "learning_rate": 1.6511103386168867e-05,
+      "loss": 0.955,
+      "step": 5687
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 45.941040675931625,
+      "learning_rate": 1.650982334528708e-05,
+      "loss": 0.8191,
+      "step": 5688
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 41.62659845920154,
+      "learning_rate": 1.6508543119269727e-05,
+      "loss": 0.8207,
+      "step": 5689
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 41.61484654629396,
+      "learning_rate": 1.6507262708153215e-05,
+      "loss": 0.7209,
+      "step": 5690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 28.77402281754922,
+      "learning_rate": 1.6505982111973956e-05,
+      "loss": 0.6521,
+      "step": 5691
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 40.41876242929282,
+      "learning_rate": 1.6504701330768372e-05,
+      "loss": 0.737,
+      "step": 5692
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.96108980375239,
+      "learning_rate": 1.650342036457289e-05,
+      "loss": 0.7688,
+      "step": 5693
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.70807169028669,
+      "learning_rate": 1.6502139213423936e-05,
+      "loss": 0.7455,
+      "step": 5694
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 42.004575383651456,
+      "learning_rate": 1.650085787735795e-05,
+      "loss": 0.8861,
+      "step": 5695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 39.45144550034641,
+      "learning_rate": 1.649957635641136e-05,
+      "loss": 0.7994,
+      "step": 5696
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.095704908398275,
+      "learning_rate": 1.6498294650620628e-05,
+      "loss": 0.748,
+      "step": 5697
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 89.74945977242332,
+      "learning_rate": 1.6497012760022197e-05,
+      "loss": 0.8188,
+      "step": 5698
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 42.82750585109187,
+      "learning_rate": 1.649573068465252e-05,
+      "loss": 0.7401,
+      "step": 5699
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.298642168448765,
+      "learning_rate": 1.6494448424548065e-05,
+      "loss": 0.7613,
+      "step": 5700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 39.43637908575863,
+      "learning_rate": 1.6493165979745294e-05,
+      "loss": 0.7773,
+      "step": 5701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.99684928031548,
+      "learning_rate": 1.649188335028068e-05,
+      "loss": 0.905,
+      "step": 5702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 34.792584357274855,
+      "learning_rate": 1.6490600536190705e-05,
+      "loss": 0.7521,
+      "step": 5703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 37.04929874492694,
+      "learning_rate": 1.6489317537511846e-05,
+      "loss": 0.7775,
+      "step": 5704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 40.5847921112543,
+      "learning_rate": 1.6488034354280592e-05,
+      "loss": 0.7173,
+      "step": 5705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 45.72290940571102,
+      "learning_rate": 1.648675098653344e-05,
+      "loss": 0.7828,
+      "step": 5706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 44.0913646223102,
+      "learning_rate": 1.648546743430688e-05,
+      "loss": 0.7706,
+      "step": 5707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.35162433267516,
+      "learning_rate": 1.648418369763742e-05,
+      "loss": 0.6799,
+      "step": 5708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.690183670639335,
+      "learning_rate": 1.648289977656157e-05,
+      "loss": 0.764,
+      "step": 5709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 49.477532831890144,
+      "learning_rate": 1.6481615671115845e-05,
+      "loss": 0.7908,
+      "step": 5710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.646324791089434,
+      "learning_rate": 1.6480331381336757e-05,
+      "loss": 0.6975,
+      "step": 5711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 40.57900559355229,
+      "learning_rate": 1.647904690726084e-05,
+      "loss": 0.8987,
+      "step": 5712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 24.147692011985942,
+      "learning_rate": 1.6477762248924616e-05,
+      "loss": 0.6034,
+      "step": 5713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 36.552556223005865,
+      "learning_rate": 1.6476477406364623e-05,
+      "loss": 0.685,
+      "step": 5714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 34.84197444411836,
+      "learning_rate": 1.6475192379617405e-05,
+      "loss": 0.7571,
+      "step": 5715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 41.85229836194952,
+      "learning_rate": 1.64739071687195e-05,
+      "loss": 0.7419,
+      "step": 5716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 42.95944803433861,
+      "learning_rate": 1.6472621773707463e-05,
+      "loss": 0.7781,
+      "step": 5717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 47.59036065479776,
+      "learning_rate": 1.647133619461785e-05,
+      "loss": 0.8603,
+      "step": 5718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 34.887362770540534,
+      "learning_rate": 1.647005043148722e-05,
+      "loss": 0.8574,
+      "step": 5719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.94142570762854,
+      "learning_rate": 1.646876448435214e-05,
+      "loss": 0.8143,
+      "step": 5720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.040008733427264,
+      "learning_rate": 1.6467478353249186e-05,
+      "loss": 0.7849,
+      "step": 5721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 41.1740258482049,
+      "learning_rate": 1.6466192038214928e-05,
+      "loss": 0.8721,
+      "step": 5722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 37.662761634907966,
+      "learning_rate": 1.646490553928595e-05,
+      "loss": 0.8217,
+      "step": 5723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.011222722630826,
+      "learning_rate": 1.6463618856498844e-05,
+      "loss": 0.6815,
+      "step": 5724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.65990665799672,
+      "learning_rate": 1.64623319898902e-05,
+      "loss": 0.7354,
+      "step": 5725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 39.45170255910591,
+      "learning_rate": 1.6461044939496606e-05,
+      "loss": 0.8519,
+      "step": 5726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.00195605397898,
+      "learning_rate": 1.645975770535468e-05,
+      "loss": 0.7817,
+      "step": 5727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.40135189292939,
+      "learning_rate": 1.6458470287501024e-05,
+      "loss": 0.7781,
+      "step": 5728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 50.0583223369026,
+      "learning_rate": 1.6457182685972248e-05,
+      "loss": 0.7754,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 40.70888337095927,
+      "learning_rate": 1.6455894900804974e-05,
+      "loss": 0.76,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.16852624686304,
+      "learning_rate": 1.6454606932035827e-05,
+      "loss": 0.8962,
+      "step": 5731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 36.52422148327198,
+      "learning_rate": 1.6453318779701434e-05,
+      "loss": 0.8457,
+      "step": 5732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.71282941955379,
+      "learning_rate": 1.645203044383843e-05,
+      "loss": 0.8015,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 29.39993475741329,
+      "learning_rate": 1.645074192448346e-05,
+      "loss": 0.753,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 37.96211629302468,
+      "learning_rate": 1.6449453221673154e-05,
+      "loss": 0.7719,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.43949784260018,
+      "learning_rate": 1.6448164335444178e-05,
+      "loss": 0.6988,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 32.2315766621817,
+      "learning_rate": 1.6446875265833178e-05,
+      "loss": 0.8184,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.798139312542546,
+      "learning_rate": 1.6445586012876814e-05,
+      "loss": 0.8339,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 30.243601197114945,
+      "learning_rate": 1.644429657661176e-05,
+      "loss": 0.7589,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.7097002969543,
+      "learning_rate": 1.6443006957074675e-05,
+      "loss": 0.7595,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 32.13111244117306,
+      "learning_rate": 1.6441717154302246e-05,
+      "loss": 0.8108,
+      "step": 5741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.26037808416415,
+      "learning_rate": 1.644042716833115e-05,
+      "loss": 0.7773,
+      "step": 5742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.68546511205847,
+      "learning_rate": 1.6439136999198068e-05,
+      "loss": 0.7384,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 37.56107962361785,
+      "learning_rate": 1.64378466469397e-05,
+      "loss": 0.7103,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.570672486048085,
+      "learning_rate": 1.643655611159274e-05,
+      "loss": 0.6432,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 44.39447715343128,
+      "learning_rate": 1.6435265393193886e-05,
+      "loss": 0.7908,
+      "step": 5746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.04789290223371,
+      "learning_rate": 1.6433974491779853e-05,
+      "loss": 0.7013,
+      "step": 5747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.6197875812808,
+      "learning_rate": 1.6432683407387346e-05,
+      "loss": 0.7332,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 37.03192838860277,
+      "learning_rate": 1.6431392140053088e-05,
+      "loss": 0.7902,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 39.424852809861044,
+      "learning_rate": 1.6430100689813797e-05,
+      "loss": 0.6827,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 36.05526224754879,
+      "learning_rate": 1.642880905670621e-05,
+      "loss": 0.8242,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.627067579929836,
+      "learning_rate": 1.6427517240767046e-05,
+      "loss": 0.8142,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.306810341799267,
+      "learning_rate": 1.6426225242033055e-05,
+      "loss": 0.7552,
+      "step": 5753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 32.03214364456708,
+      "learning_rate": 1.6424933060540978e-05,
+      "loss": 0.7199,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 48.39305176653056,
+      "learning_rate": 1.6423640696327564e-05,
+      "loss": 0.7896,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.99034431562039,
+      "learning_rate": 1.6422348149429566e-05,
+      "loss": 0.7976,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.48577779829967,
+      "learning_rate": 1.6421055419883744e-05,
+      "loss": 0.7138,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 40.27377676168785,
+      "learning_rate": 1.641976250772686e-05,
+      "loss": 0.8399,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.092633563510596,
+      "learning_rate": 1.6418469412995685e-05,
+      "loss": 0.847,
+      "step": 5759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 48.1040246211626,
+      "learning_rate": 1.6417176135726998e-05,
+      "loss": 0.793,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.0813468309122,
+      "learning_rate": 1.641588267595757e-05,
+      "loss": 0.8627,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 45.12797156189562,
+      "learning_rate": 1.6414589033724197e-05,
+      "loss": 0.7549,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 42.87023543417555,
+      "learning_rate": 1.6413295209063663e-05,
+      "loss": 0.7406,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 33.35313074781507,
+      "learning_rate": 1.6412001202012768e-05,
+      "loss": 0.7739,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 41.26510694901531,
+      "learning_rate": 1.6410707012608303e-05,
+      "loss": 0.8114,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.68335363969928,
+      "learning_rate": 1.640941264088709e-05,
+      "loss": 0.7233,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 29.026315737875294,
+      "learning_rate": 1.6408118086885925e-05,
+      "loss": 0.7118,
+      "step": 5767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 41.95106368302013,
+      "learning_rate": 1.6406823350641628e-05,
+      "loss": 0.865,
+      "step": 5768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 41.43998419103173,
+      "learning_rate": 1.6405528432191027e-05,
+      "loss": 0.7834,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.83614776883227,
+      "learning_rate": 1.6404233331570944e-05,
+      "loss": 0.8169,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 30.419946555530217,
+      "learning_rate": 1.6402938048818208e-05,
+      "loss": 0.6901,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.6441071497583,
+      "learning_rate": 1.640164258396966e-05,
+      "loss": 0.7299,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.92661683161436,
+      "learning_rate": 1.6400346937062147e-05,
+      "loss": 0.7916,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 36.37651051010087,
+      "learning_rate": 1.6399051108132507e-05,
+      "loss": 0.8647,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 36.605406268860996,
+      "learning_rate": 1.6397755097217598e-05,
+      "loss": 0.7709,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.76734582870041,
+      "learning_rate": 1.6396458904354275e-05,
+      "loss": 0.7369,
+      "step": 5776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 37.113978226909325,
+      "learning_rate": 1.6395162529579406e-05,
+      "loss": 0.7451,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.03028038688392,
+      "learning_rate": 1.6393865972929853e-05,
+      "loss": 0.8854,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 33.61498645666446,
+      "learning_rate": 1.6392569234442492e-05,
+      "loss": 0.7967,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 30.94036139925058,
+      "learning_rate": 1.63912723141542e-05,
+      "loss": 0.7251,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 36.245847431295154,
+      "learning_rate": 1.6389975212101867e-05,
+      "loss": 0.8765,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 40.65629750629438,
+      "learning_rate": 1.6388677928322376e-05,
+      "loss": 0.8619,
+      "step": 5782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 32.01503080279281,
+      "learning_rate": 1.638738046285262e-05,
+      "loss": 0.7202,
+      "step": 5783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.530706325339093,
+      "learning_rate": 1.63860828157295e-05,
+      "loss": 0.8038,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 32.61559769644914,
+      "learning_rate": 1.6384784986989918e-05,
+      "loss": 0.7396,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 45.66331556768221,
+      "learning_rate": 1.638348697667079e-05,
+      "loss": 0.8474,
+      "step": 5786
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.11892243068202,
+      "learning_rate": 1.6382188784809027e-05,
+      "loss": 0.6435,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 39.38933465789394,
+      "learning_rate": 1.638089041144155e-05,
+      "loss": 0.8058,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.16593333786399,
+      "learning_rate": 1.6379591856605274e-05,
+      "loss": 0.773,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 45.38881562347808,
+      "learning_rate": 1.6378293120337144e-05,
+      "loss": 0.8401,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 45.56366830800953,
+      "learning_rate": 1.637699420267409e-05,
+      "loss": 0.8627,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.10740556273525,
+      "learning_rate": 1.6375695103653047e-05,
+      "loss": 0.843,
+      "step": 5792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 46.064796261118964,
+      "learning_rate": 1.637439582331097e-05,
+      "loss": 0.8207,
+      "step": 5793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 49.47301755367353,
+      "learning_rate": 1.63730963616848e-05,
+      "loss": 0.8389,
+      "step": 5794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 29.207995482616287,
+      "learning_rate": 1.6371796718811496e-05,
+      "loss": 0.705,
+      "step": 5795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 42.512251512967204,
+      "learning_rate": 1.637049689472803e-05,
+      "loss": 0.8448,
+      "step": 5796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 35.1612741466172,
+      "learning_rate": 1.636919688947135e-05,
+      "loss": 0.734,
+      "step": 5797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 36.26655877853456,
+      "learning_rate": 1.636789670307844e-05,
+      "loss": 0.7862,
+      "step": 5798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 42.56029848183211,
+      "learning_rate": 1.636659633558627e-05,
+      "loss": 0.8272,
+      "step": 5799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 34.06938580549726,
+      "learning_rate": 1.6365295787031827e-05,
+      "loss": 0.7604,
+      "step": 5800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 28.191315124203754,
+      "learning_rate": 1.6363995057452098e-05,
+      "loss": 0.6906,
+      "step": 5801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 36.33783759769629,
+      "learning_rate": 1.6362694146884067e-05,
+      "loss": 0.8204,
+      "step": 5802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 45.07949771476886,
+      "learning_rate": 1.636139305536474e-05,
+      "loss": 0.7937,
+      "step": 5803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.54406762800532,
+      "learning_rate": 1.6360091782931115e-05,
+      "loss": 0.655,
+      "step": 5804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 34.397915505445475,
+      "learning_rate": 1.63587903296202e-05,
+      "loss": 0.8173,
+      "step": 5805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 35.02142477963862,
+      "learning_rate": 1.635748869546901e-05,
+      "loss": 0.7226,
+      "step": 5806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 31.48465596492529,
+      "learning_rate": 1.6356186880514556e-05,
+      "loss": 0.6571,
+      "step": 5807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 43.19114674630548,
+      "learning_rate": 1.635488488479387e-05,
+      "loss": 0.8587,
+      "step": 5808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 31.167276881783952,
+      "learning_rate": 1.635358270834397e-05,
+      "loss": 0.7467,
+      "step": 5809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 36.23761096238632,
+      "learning_rate": 1.6352280351201898e-05,
+      "loss": 0.8048,
+      "step": 5810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.07381373075287,
+      "learning_rate": 1.635097781340469e-05,
+      "loss": 0.861,
+      "step": 5811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.09094807455748,
+      "learning_rate": 1.6349675094989387e-05,
+      "loss": 0.8153,
+      "step": 5812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 41.43971124563174,
+      "learning_rate": 1.6348372195993037e-05,
+      "loss": 0.7667,
+      "step": 5813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 46.5169747758696,
+      "learning_rate": 1.63470691164527e-05,
+      "loss": 0.6652,
+      "step": 5814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 29.86357742671659,
+      "learning_rate": 1.6345765856405424e-05,
+      "loss": 0.6805,
+      "step": 5815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 41.33156689335333,
+      "learning_rate": 1.6344462415888283e-05,
+      "loss": 0.7893,
+      "step": 5816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 46.39205434466015,
+      "learning_rate": 1.6343158794938342e-05,
+      "loss": 0.7808,
+      "step": 5817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 36.960527371405135,
+      "learning_rate": 1.6341854993592674e-05,
+      "loss": 0.8298,
+      "step": 5818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 31.974930781959852,
+      "learning_rate": 1.634055101188836e-05,
+      "loss": 0.7291,
+      "step": 5819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 30.020541458158185,
+      "learning_rate": 1.6339246849862488e-05,
+      "loss": 0.6871,
+      "step": 5820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.168125673845815,
+      "learning_rate": 1.633794250755214e-05,
+      "loss": 0.6841,
+      "step": 5821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 27.610027184633623,
+      "learning_rate": 1.6336637984994418e-05,
+      "loss": 0.7138,
+      "step": 5822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.88242544274227,
+      "learning_rate": 1.6335333282226414e-05,
+      "loss": 0.8843,
+      "step": 5823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 45.090596411942684,
+      "learning_rate": 1.6334028399285243e-05,
+      "loss": 0.7098,
+      "step": 5824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 63.10209571846402,
+      "learning_rate": 1.6332723336208007e-05,
+      "loss": 0.7966,
+      "step": 5825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 51.14178365717119,
+      "learning_rate": 1.633141809303182e-05,
+      "loss": 0.8033,
+      "step": 5826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.42925493378733,
+      "learning_rate": 1.633011266979381e-05,
+      "loss": 0.782,
+      "step": 5827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 24.544547479960915,
+      "learning_rate": 1.6328807066531102e-05,
+      "loss": 0.6258,
+      "step": 5828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 33.80562984668576,
+      "learning_rate": 1.6327501283280816e-05,
+      "loss": 0.715,
+      "step": 5829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.06916941416905,
+      "learning_rate": 1.6326195320080103e-05,
+      "loss": 0.8334,
+      "step": 5830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.70380431668692,
+      "learning_rate": 1.632488917696609e-05,
+      "loss": 0.7239,
+      "step": 5831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.387629397342764,
+      "learning_rate": 1.632358285397593e-05,
+      "loss": 0.6939,
+      "step": 5832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 40.89084010346407,
+      "learning_rate": 1.6322276351146774e-05,
+      "loss": 0.7683,
+      "step": 5833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 48.35330125570362,
+      "learning_rate": 1.6320969668515773e-05,
+      "loss": 0.9275,
+      "step": 5834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.71590608496326,
+      "learning_rate": 1.63196628061201e-05,
+      "loss": 0.7368,
+      "step": 5835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 31.681266615489424,
+      "learning_rate": 1.6318355763996908e-05,
+      "loss": 0.7641,
+      "step": 5836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 45.196084258431355,
+      "learning_rate": 1.6317048542183375e-05,
+      "loss": 0.8875,
+      "step": 5837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 28.931111355571108,
+      "learning_rate": 1.6315741140716675e-05,
+      "loss": 0.6785,
+      "step": 5838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 31.41516828547467,
+      "learning_rate": 1.6314433559633992e-05,
+      "loss": 0.7401,
+      "step": 5839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.73776041644203,
+      "learning_rate": 1.6313125798972514e-05,
+      "loss": 0.7571,
+      "step": 5840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 42.446062805632955,
+      "learning_rate": 1.6311817858769434e-05,
+      "loss": 0.8534,
+      "step": 5841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 41.67570302378294,
+      "learning_rate": 1.6310509739061946e-05,
+      "loss": 1.0233,
+      "step": 5842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 34.156103177466164,
+      "learning_rate": 1.6309201439887248e-05,
+      "loss": 0.7899,
+      "step": 5843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 35.46351068594047,
+      "learning_rate": 1.6307892961282555e-05,
+      "loss": 0.7712,
+      "step": 5844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 35.14311416718346,
+      "learning_rate": 1.6306584303285077e-05,
+      "loss": 0.7163,
+      "step": 5845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 43.209484213152145,
+      "learning_rate": 1.6305275465932027e-05,
+      "loss": 0.9377,
+      "step": 5846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 39.904041109660625,
+      "learning_rate": 1.6303966449260636e-05,
+      "loss": 0.7436,
+      "step": 5847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 37.966728392105125,
+      "learning_rate": 1.630265725330812e-05,
+      "loss": 0.7192,
+      "step": 5848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 35.69077360835198,
+      "learning_rate": 1.6301347878111726e-05,
+      "loss": 0.7461,
+      "step": 5849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 46.57349916129555,
+      "learning_rate": 1.630003832370868e-05,
+      "loss": 0.8995,
+      "step": 5850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 39.03333286075825,
+      "learning_rate": 1.629872859013623e-05,
+      "loss": 0.7199,
+      "step": 5851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.5791608028274,
+      "learning_rate": 1.6297418677431625e-05,
+      "loss": 0.7464,
+      "step": 5852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.30051118177526,
+      "learning_rate": 1.629610858563212e-05,
+      "loss": 0.79,
+      "step": 5853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 30.777636669327407,
+      "learning_rate": 1.6294798314774963e-05,
+      "loss": 0.6861,
+      "step": 5854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 48.46903532179101,
+      "learning_rate": 1.6293487864897425e-05,
+      "loss": 0.8407,
+      "step": 5855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.71252252798795,
+      "learning_rate": 1.6292177236036776e-05,
+      "loss": 0.6836,
+      "step": 5856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 29.475761443131496,
+      "learning_rate": 1.629086642823029e-05,
+      "loss": 0.6814,
+      "step": 5857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 44.981071835846535,
+      "learning_rate": 1.628955544151524e-05,
+      "loss": 0.8406,
+      "step": 5858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 36.83811277598921,
+      "learning_rate": 1.6288244275928912e-05,
+      "loss": 0.7281,
+      "step": 5859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 40.76412433568547,
+      "learning_rate": 1.6286932931508596e-05,
+      "loss": 0.7718,
+      "step": 5860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.39182855971103,
+      "learning_rate": 1.6285621408291584e-05,
+      "loss": 0.7589,
+      "step": 5861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.1220877582596,
+      "learning_rate": 1.6284309706315178e-05,
+      "loss": 0.9099,
+      "step": 5862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 42.14467205095155,
+      "learning_rate": 1.6282997825616676e-05,
+      "loss": 0.7512,
+      "step": 5863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 28.965064200078775,
+      "learning_rate": 1.62816857662334e-05,
+      "loss": 0.8316,
+      "step": 5864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.16614973607111,
+      "learning_rate": 1.6280373528202648e-05,
+      "loss": 0.7108,
+      "step": 5865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.618553058112326,
+      "learning_rate": 1.627906111156175e-05,
+      "loss": 0.7316,
+      "step": 5866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 29.278274531516484,
+      "learning_rate": 1.6277748516348025e-05,
+      "loss": 0.7182,
+      "step": 5867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.89115075612823,
+      "learning_rate": 1.6276435742598807e-05,
+      "loss": 0.7685,
+      "step": 5868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 35.52592494117536,
+      "learning_rate": 1.6275122790351426e-05,
+      "loss": 0.8542,
+      "step": 5869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.74343576264143,
+      "learning_rate": 1.6273809659643226e-05,
+      "loss": 0.9111,
+      "step": 5870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 42.884203064914445,
+      "learning_rate": 1.6272496350511547e-05,
+      "loss": 0.7839,
+      "step": 5871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 39.646915067560926,
+      "learning_rate": 1.6271182862993743e-05,
+      "loss": 0.8952,
+      "step": 5872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 41.36571115483506,
+      "learning_rate": 1.6269869197127162e-05,
+      "loss": 0.8229,
+      "step": 5873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.574873937095838,
+      "learning_rate": 1.6268555352949175e-05,
+      "loss": 0.7516,
+      "step": 5874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.717646469991614,
+      "learning_rate": 1.626724133049714e-05,
+      "loss": 0.7752,
+      "step": 5875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 42.20649418239832,
+      "learning_rate": 1.6265927129808426e-05,
+      "loss": 0.9162,
+      "step": 5876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.776173288181486,
+      "learning_rate": 1.6264612750920406e-05,
+      "loss": 0.7044,
+      "step": 5877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.58634189381308,
+      "learning_rate": 1.6263298193870468e-05,
+      "loss": 0.801,
+      "step": 5878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.15297097382667,
+      "learning_rate": 1.6261983458695996e-05,
+      "loss": 0.8019,
+      "step": 5879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.325012253651106,
+      "learning_rate": 1.6260668545434375e-05,
+      "loss": 0.696,
+      "step": 5880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.435625715728285,
+      "learning_rate": 1.6259353454123003e-05,
+      "loss": 0.7669,
+      "step": 5881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.774285422209307,
+      "learning_rate": 1.625803818479928e-05,
+      "loss": 0.8454,
+      "step": 5882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 41.49286077172578,
+      "learning_rate": 1.6256722737500613e-05,
+      "loss": 0.7945,
+      "step": 5883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 33.12207810452746,
+      "learning_rate": 1.625540711226441e-05,
+      "loss": 0.8023,
+      "step": 5884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 33.53431236702752,
+      "learning_rate": 1.625409130912809e-05,
+      "loss": 0.7912,
+      "step": 5885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 31.979129899756618,
+      "learning_rate": 1.6252775328129067e-05,
+      "loss": 0.7249,
+      "step": 5886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.62480309266942,
+      "learning_rate": 1.6251459169304776e-05,
+      "loss": 0.7454,
+      "step": 5887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 31.666454131684745,
+      "learning_rate": 1.6250142832692643e-05,
+      "loss": 0.7177,
+      "step": 5888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.729513691584636,
+      "learning_rate": 1.6248826318330103e-05,
+      "loss": 0.7783,
+      "step": 5889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 35.551811053290294,
+      "learning_rate": 1.6247509626254597e-05,
+      "loss": 0.6866,
+      "step": 5890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 41.42550878079964,
+      "learning_rate": 1.6246192756503572e-05,
+      "loss": 0.8436,
+      "step": 5891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 35.02470441607743,
+      "learning_rate": 1.624487570911448e-05,
+      "loss": 0.6558,
+      "step": 5892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 50.259633957956446,
+      "learning_rate": 1.6243558484124778e-05,
+      "loss": 0.8475,
+      "step": 5893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 33.5989826190281,
+      "learning_rate": 1.6242241081571923e-05,
+      "loss": 0.7026,
+      "step": 5894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.37070268930533,
+      "learning_rate": 1.6240923501493383e-05,
+      "loss": 0.7143,
+      "step": 5895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.774352962600005,
+      "learning_rate": 1.6239605743926632e-05,
+      "loss": 0.7838,
+      "step": 5896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 36.47714311044093,
+      "learning_rate": 1.6238287808909137e-05,
+      "loss": 0.773,
+      "step": 5897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 41.029075309259404,
+      "learning_rate": 1.6236969696478393e-05,
+      "loss": 0.7227,
+      "step": 5898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 44.229842832011364,
+      "learning_rate": 1.6235651406671875e-05,
+      "loss": 0.7507,
+      "step": 5899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.65543230158702,
+      "learning_rate": 1.623433293952708e-05,
+      "loss": 0.8022,
+      "step": 5900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.16723327271666,
+      "learning_rate": 1.6233014295081505e-05,
+      "loss": 0.7988,
+      "step": 5901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.563386459600498,
+      "learning_rate": 1.623169547337265e-05,
+      "loss": 0.7085,
+      "step": 5902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 35.23774910818807,
+      "learning_rate": 1.6230376474438018e-05,
+      "loss": 0.7514,
+      "step": 5903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 28.185968268617852,
+      "learning_rate": 1.6229057298315123e-05,
+      "loss": 0.7295,
+      "step": 5904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.450152753656646,
+      "learning_rate": 1.6227737945041485e-05,
+      "loss": 0.7868,
+      "step": 5905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 45.48629694828706,
+      "learning_rate": 1.6226418414654625e-05,
+      "loss": 0.9134,
+      "step": 5906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 39.98937140757488,
+      "learning_rate": 1.6225098707192063e-05,
+      "loss": 0.8719,
+      "step": 5907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 29.863731242170747,
+      "learning_rate": 1.622377882269134e-05,
+      "loss": 0.7188,
+      "step": 5908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.457063458415753,
+      "learning_rate": 1.6222458761189984e-05,
+      "loss": 0.7271,
+      "step": 5909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 40.27006223874129,
+      "learning_rate": 1.622113852272554e-05,
+      "loss": 0.7347,
+      "step": 5910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.30028898653128,
+      "learning_rate": 1.621981810733556e-05,
+      "loss": 0.7116,
+      "step": 5911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.98703163134623,
+      "learning_rate": 1.6218497515057593e-05,
+      "loss": 0.7912,
+      "step": 5912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 44.25297538538618,
+      "learning_rate": 1.621717674592919e-05,
+      "loss": 0.8519,
+      "step": 5913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.39330206922661,
+      "learning_rate": 1.621585579998792e-05,
+      "loss": 0.6879,
+      "step": 5914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.59105550174588,
+      "learning_rate": 1.6214534677271344e-05,
+      "loss": 0.7783,
+      "step": 5915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 56.27052262507723,
+      "learning_rate": 1.6213213377817035e-05,
+      "loss": 0.8197,
+      "step": 5916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 31.616121176573834,
+      "learning_rate": 1.6211891901662576e-05,
+      "loss": 0.778,
+      "step": 5917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 33.64601043245142,
+      "learning_rate": 1.621057024884555e-05,
+      "loss": 0.8122,
+      "step": 5918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 37.73663218442713,
+      "learning_rate": 1.620924841940353e-05,
+      "loss": 0.8379,
+      "step": 5919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 32.95715365285955,
+      "learning_rate": 1.6207926413374124e-05,
+      "loss": 0.7405,
+      "step": 5920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.86324601780971,
+      "learning_rate": 1.620660423079492e-05,
+      "loss": 0.8375,
+      "step": 5921
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 36.68596200139966,
+      "learning_rate": 1.620528187170352e-05,
+      "loss": 0.7344,
+      "step": 5922
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 31.34705906949797,
+      "learning_rate": 1.6203959336137538e-05,
+      "loss": 0.7976,
+      "step": 5923
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 32.80405265700755,
+      "learning_rate": 1.620263662413458e-05,
+      "loss": 0.8014,
+      "step": 5924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 31.495946789899804,
+      "learning_rate": 1.6201313735732265e-05,
+      "loss": 0.75,
+      "step": 5925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.184346337023012,
+      "learning_rate": 1.619999067096821e-05,
+      "loss": 0.7651,
+      "step": 5926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 28.224551480026477,
+      "learning_rate": 1.6198667429880054e-05,
+      "loss": 0.6459,
+      "step": 5927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 34.57093072198363,
+      "learning_rate": 1.619734401250542e-05,
+      "loss": 0.8054,
+      "step": 5928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 35.28935758759948,
+      "learning_rate": 1.6196020418881947e-05,
+      "loss": 0.7933,
+      "step": 5929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 63.83293698870899,
+      "learning_rate": 1.6194696649047278e-05,
+      "loss": 0.7935,
+      "step": 5930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 40.87573568749469,
+      "learning_rate": 1.619337270303906e-05,
+      "loss": 0.9134,
+      "step": 5931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 43.190980668863645,
+      "learning_rate": 1.6192048580894946e-05,
+      "loss": 0.8043,
+      "step": 5932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 42.39558620646416,
+      "learning_rate": 1.6190724282652594e-05,
+      "loss": 0.9359,
+      "step": 5933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 34.85922332826428,
+      "learning_rate": 1.6189399808349663e-05,
+      "loss": 0.8561,
+      "step": 5934
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 39.17076810742023,
+      "learning_rate": 1.6188075158023825e-05,
+      "loss": 0.846,
+      "step": 5935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 43.01817714382345,
+      "learning_rate": 1.6186750331712744e-05,
+      "loss": 0.7333,
+      "step": 5936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.382760909749518,
+      "learning_rate": 1.6185425329454106e-05,
+      "loss": 0.6379,
+      "step": 5937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 41.247981894862015,
+      "learning_rate": 1.618410015128559e-05,
+      "loss": 0.7614,
+      "step": 5938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 31.873352741237408,
+      "learning_rate": 1.6182774797244882e-05,
+      "loss": 0.7651,
+      "step": 5939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 28.43802350998133,
+      "learning_rate": 1.6181449267369678e-05,
+      "loss": 0.8032,
+      "step": 5940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 30.138141970796305,
+      "learning_rate": 1.6180123561697672e-05,
+      "loss": 0.726,
+      "step": 5941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 42.84489733761701,
+      "learning_rate": 1.6178797680266566e-05,
+      "loss": 0.7979,
+      "step": 5942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 36.08179124828941,
+      "learning_rate": 1.6177471623114068e-05,
+      "loss": 0.8058,
+      "step": 5943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 37.530235835848124,
+      "learning_rate": 1.6176145390277893e-05,
+      "loss": 0.8058,
+      "step": 5944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 31.939842578652875,
+      "learning_rate": 1.6174818981795756e-05,
+      "loss": 0.7506,
+      "step": 5945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.97993763207787,
+      "learning_rate": 1.6173492397705376e-05,
+      "loss": 0.7787,
+      "step": 5946
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 34.602977180285805,
+      "learning_rate": 1.6172165638044486e-05,
+      "loss": 0.8186,
+      "step": 5947
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.37622004861404,
+      "learning_rate": 1.6170838702850814e-05,
+      "loss": 0.7853,
+      "step": 5948
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 30.799499144445118,
+      "learning_rate": 1.61695115921621e-05,
+      "loss": 0.7965,
+      "step": 5949
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 40.92575075571305,
+      "learning_rate": 1.6168184306016092e-05,
+      "loss": 0.8539,
+      "step": 5950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 42.24935489365191,
+      "learning_rate": 1.6166856844450522e-05,
+      "loss": 0.7628,
+      "step": 5951
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.22522085114983,
+      "learning_rate": 1.6165529207503156e-05,
+      "loss": 0.6823,
+      "step": 5952
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 47.72934566793367,
+      "learning_rate": 1.616420139521174e-05,
+      "loss": 0.7347,
+      "step": 5953
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 40.327561050486125,
+      "learning_rate": 1.6162873407614047e-05,
+      "loss": 0.7678,
+      "step": 5954
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 41.07276518204678,
+      "learning_rate": 1.616154524474784e-05,
+      "loss": 0.7784,
+      "step": 5955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 30.971882828153472,
+      "learning_rate": 1.616021690665089e-05,
+      "loss": 0.7062,
+      "step": 5956
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 35.88784435164143,
+      "learning_rate": 1.615888839336097e-05,
+      "loss": 0.7197,
+      "step": 5957
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 27.282602401194463,
+      "learning_rate": 1.6157559704915873e-05,
+      "loss": 0.6793,
+      "step": 5958
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 41.90624800873367,
+      "learning_rate": 1.6156230841353376e-05,
+      "loss": 0.7081,
+      "step": 5959
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 32.778799698186845,
+      "learning_rate": 1.6154901802711277e-05,
+      "loss": 0.6583,
+      "step": 5960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 41.37184568119285,
+      "learning_rate": 1.615357258902737e-05,
+      "loss": 0.7512,
+      "step": 5961
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 34.50190100408138,
+      "learning_rate": 1.6152243200339462e-05,
+      "loss": 0.823,
+      "step": 5962
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 32.6366568746842,
+      "learning_rate": 1.6150913636685356e-05,
+      "loss": 0.7408,
+      "step": 5963
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 50.411419460900284,
+      "learning_rate": 1.614958389810286e-05,
+      "loss": 0.8976,
+      "step": 5964
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 39.34318934761514,
+      "learning_rate": 1.6148253984629798e-05,
+      "loss": 0.7891,
+      "step": 5965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 37.068198034601785,
+      "learning_rate": 1.614692389630399e-05,
+      "loss": 0.8014,
+      "step": 5966
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 36.194508303809336,
+      "learning_rate": 1.6145593633163258e-05,
+      "loss": 0.8392,
+      "step": 5967
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 31.981254585687736,
+      "learning_rate": 1.614426319524544e-05,
+      "loss": 0.6922,
+      "step": 5968
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 31.699039476114528,
+      "learning_rate": 1.6142932582588375e-05,
+      "loss": 0.6938,
+      "step": 5969
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 41.05715047633059,
+      "learning_rate": 1.61416017952299e-05,
+      "loss": 0.7297,
+      "step": 5970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 41.966679951623014,
+      "learning_rate": 1.614027083320786e-05,
+      "loss": 0.904,
+      "step": 5971
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.726589656180025,
+      "learning_rate": 1.613893969656011e-05,
+      "loss": 0.8163,
+      "step": 5972
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 36.39511699313658,
+      "learning_rate": 1.6137608385324505e-05,
+      "loss": 0.7617,
+      "step": 5973
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 36.38379522875893,
+      "learning_rate": 1.613627689953891e-05,
+      "loss": 0.7877,
+      "step": 5974
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.783901472833197,
+      "learning_rate": 1.613494523924119e-05,
+      "loss": 0.7667,
+      "step": 5975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.194651951977,
+      "learning_rate": 1.6133613404469216e-05,
+      "loss": 0.7664,
+      "step": 5976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 30.86805782380088,
+      "learning_rate": 1.6132281395260863e-05,
+      "loss": 0.7186,
+      "step": 5977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 45.57236041210932,
+      "learning_rate": 1.6130949211654017e-05,
+      "loss": 0.8066,
+      "step": 5978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 52.01511411911586,
+      "learning_rate": 1.612961685368656e-05,
+      "loss": 0.7897,
+      "step": 5979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 39.15172139106732,
+      "learning_rate": 1.6128284321396385e-05,
+      "loss": 0.7727,
+      "step": 5980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.42192988237367,
+      "learning_rate": 1.6126951614821388e-05,
+      "loss": 0.8015,
+      "step": 5981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 38.97181936747496,
+      "learning_rate": 1.6125618733999476e-05,
+      "loss": 0.8011,
+      "step": 5982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 35.58198259069516,
+      "learning_rate": 1.6124285678968546e-05,
+      "loss": 0.7874,
+      "step": 5983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.162056576741893,
+      "learning_rate": 1.6122952449766516e-05,
+      "loss": 0.7268,
+      "step": 5984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 33.49987764478516,
+      "learning_rate": 1.6121619046431297e-05,
+      "loss": 0.8282,
+      "step": 5985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 35.4244508181143,
+      "learning_rate": 1.612028546900082e-05,
+      "loss": 0.753,
+      "step": 5986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 54.57380843108972,
+      "learning_rate": 1.6118951717513e-05,
+      "loss": 0.8578,
+      "step": 5987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 34.57192925244885,
+      "learning_rate": 1.611761779200577e-05,
+      "loss": 0.7024,
+      "step": 5988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 31.296011609544095,
+      "learning_rate": 1.6116283692517075e-05,
+      "loss": 0.7558,
+      "step": 5989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 42.823091029495224,
+      "learning_rate": 1.6114949419084846e-05,
+      "loss": 0.8428,
+      "step": 5990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.103248473934144,
+      "learning_rate": 1.6113614971747032e-05,
+      "loss": 0.7982,
+      "step": 5991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 40.6694535579885,
+      "learning_rate": 1.6112280350541584e-05,
+      "loss": 0.7739,
+      "step": 5992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 41.56695277519131,
+      "learning_rate": 1.6110945555506462e-05,
+      "loss": 0.7777,
+      "step": 5993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.606180323232646,
+      "learning_rate": 1.6109610586679624e-05,
+      "loss": 0.7114,
+      "step": 5994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 38.31953743900342,
+      "learning_rate": 1.610827544409903e-05,
+      "loss": 0.7238,
+      "step": 5995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.366075883836864,
+      "learning_rate": 1.6106940127802658e-05,
+      "loss": 0.6882,
+      "step": 5996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.49883634201629,
+      "learning_rate": 1.610560463782848e-05,
+      "loss": 0.7839,
+      "step": 5997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 38.260257872596334,
+      "learning_rate": 1.610426897421448e-05,
+      "loss": 0.8223,
+      "step": 5998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 31.82307593664259,
+      "learning_rate": 1.610293313699864e-05,
+      "loss": 0.7416,
+      "step": 5999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 38.63808847162498,
+      "learning_rate": 1.610159712621895e-05,
+      "loss": 0.795,
+      "step": 6000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.382884904635,
+      "learning_rate": 1.610026094191341e-05,
+      "loss": 0.7744,
+      "step": 6001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 45.667274399992294,
+      "learning_rate": 1.6098924584120017e-05,
+      "loss": 0.9027,
+      "step": 6002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 64.56907931960087,
+      "learning_rate": 1.6097588052876774e-05,
+      "loss": 0.7887,
+      "step": 6003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 46.07909013805226,
+      "learning_rate": 1.60962513482217e-05,
+      "loss": 0.6545,
+      "step": 6004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 38.39930372687248,
+      "learning_rate": 1.6094914470192796e-05,
+      "loss": 0.8252,
+      "step": 6005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 48.165376386333094,
+      "learning_rate": 1.6093577418828094e-05,
+      "loss": 0.7925,
+      "step": 6006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 33.297247875171884,
+      "learning_rate": 1.6092240194165618e-05,
+      "loss": 0.762,
+      "step": 6007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 40.029713013778334,
+      "learning_rate": 1.6090902796243386e-05,
+      "loss": 0.8149,
+      "step": 6008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 31.580014790911406,
+      "learning_rate": 1.608956522509945e-05,
+      "loss": 0.8029,
+      "step": 6009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 45.691300438908996,
+      "learning_rate": 1.6088227480771835e-05,
+      "loss": 0.8702,
+      "step": 6010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.68706662782433,
+      "learning_rate": 1.6086889563298594e-05,
+      "loss": 0.7286,
+      "step": 6011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 35.50360673729188,
+      "learning_rate": 1.6085551472717775e-05,
+      "loss": 0.7393,
+      "step": 6012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 49.13463721251158,
+      "learning_rate": 1.6084213209067433e-05,
+      "loss": 0.8008,
+      "step": 6013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.208258834396226,
+      "learning_rate": 1.6082874772385628e-05,
+      "loss": 0.7093,
+      "step": 6014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 45.06685163461848,
+      "learning_rate": 1.6081536162710418e-05,
+      "loss": 0.7722,
+      "step": 6015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 41.314235962422465,
+      "learning_rate": 1.6080197380079883e-05,
+      "loss": 0.7743,
+      "step": 6016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 48.11138555990754,
+      "learning_rate": 1.6078858424532088e-05,
+      "loss": 0.7304,
+      "step": 6017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 39.78629200006678,
+      "learning_rate": 1.6077519296105112e-05,
+      "loss": 0.8293,
+      "step": 6018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.78465610442436,
+      "learning_rate": 1.6076179994837045e-05,
+      "loss": 0.7306,
+      "step": 6019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 29.70971773679563,
+      "learning_rate": 1.6074840520765976e-05,
+      "loss": 0.6711,
+      "step": 6020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 43.75275148731302,
+      "learning_rate": 1.607350087393e-05,
+      "loss": 0.7828,
+      "step": 6021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.64749920094595,
+      "learning_rate": 1.60721610543672e-05,
+      "loss": 0.8075,
+      "step": 6022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 55.09076066759696,
+      "learning_rate": 1.6070821062115704e-05,
+      "loss": 0.6614,
+      "step": 6023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 41.95572494605629,
+      "learning_rate": 1.60694808972136e-05,
+      "loss": 0.8987,
+      "step": 6024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 43.063549505104035,
+      "learning_rate": 1.606814055969901e-05,
+      "loss": 0.7276,
+      "step": 6025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 37.41308007789513,
+      "learning_rate": 1.606680004961006e-05,
+      "loss": 0.7342,
+      "step": 6026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.89397425690169,
+      "learning_rate": 1.6065459366984856e-05,
+      "loss": 0.7617,
+      "step": 6027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 43.866702213473445,
+      "learning_rate": 1.6064118511861543e-05,
+      "loss": 0.8072,
+      "step": 6028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 40.77441655050883,
+      "learning_rate": 1.6062777484278245e-05,
+      "loss": 0.8654,
+      "step": 6029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.54238731003136,
+      "learning_rate": 1.60614362842731e-05,
+      "loss": 0.9236,
+      "step": 6030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 29.417055767491973,
+      "learning_rate": 1.6060094911884255e-05,
+      "loss": 0.7248,
+      "step": 6031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 39.55297426525206,
+      "learning_rate": 1.6058753367149854e-05,
+      "loss": 0.8103,
+      "step": 6032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 38.33707437953753,
+      "learning_rate": 1.6057411650108053e-05,
+      "loss": 0.77,
+      "step": 6033
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 40.29783949302345,
+      "learning_rate": 1.6056069760797006e-05,
+      "loss": 0.7275,
+      "step": 6034
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.995393344819675,
+      "learning_rate": 1.605472769925488e-05,
+      "loss": 0.8051,
+      "step": 6035
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.613350816406445,
+      "learning_rate": 1.6053385465519838e-05,
+      "loss": 0.7281,
+      "step": 6036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.90857254700324,
+      "learning_rate": 1.6052043059630058e-05,
+      "loss": 0.7755,
+      "step": 6037
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 37.44912209761444,
+      "learning_rate": 1.6050700481623712e-05,
+      "loss": 0.75,
+      "step": 6038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 27.925873684086483,
+      "learning_rate": 1.604935773153898e-05,
+      "loss": 0.688,
+      "step": 6039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 53.635037447572344,
+      "learning_rate": 1.6048014809414057e-05,
+      "loss": 0.8438,
+      "step": 6040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.98608861191888,
+      "learning_rate": 1.604667171528713e-05,
+      "loss": 0.8785,
+      "step": 6041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 34.27351507035071,
+      "learning_rate": 1.60453284491964e-05,
+      "loss": 0.8457,
+      "step": 6042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 31.47493355655399,
+      "learning_rate": 1.6043985011180062e-05,
+      "loss": 0.7319,
+      "step": 6043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 34.192866913581966,
+      "learning_rate": 1.6042641401276326e-05,
+      "loss": 0.8889,
+      "step": 6044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 35.763930544329995,
+      "learning_rate": 1.6041297619523405e-05,
+      "loss": 0.7361,
+      "step": 6045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 33.03581959379931,
+      "learning_rate": 1.603995366595951e-05,
+      "loss": 0.7711,
+      "step": 6046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.74819460151574,
+      "learning_rate": 1.603860954062287e-05,
+      "loss": 0.8179,
+      "step": 6047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 42.144459092795024,
+      "learning_rate": 1.6037265243551704e-05,
+      "loss": 0.7442,
+      "step": 6048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 32.313753686634435,
+      "learning_rate": 1.6035920774784247e-05,
+      "loss": 0.7058,
+      "step": 6049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 42.42466758121239,
+      "learning_rate": 1.6034576134358736e-05,
+      "loss": 0.9029,
+      "step": 6050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 44.466272431207805,
+      "learning_rate": 1.6033231322313406e-05,
+      "loss": 0.7822,
+      "step": 6051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 41.38809664108115,
+      "learning_rate": 1.603188633868651e-05,
+      "loss": 0.8205,
+      "step": 6052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 35.92894649052539,
+      "learning_rate": 1.6030541183516292e-05,
+      "loss": 0.6781,
+      "step": 6053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 41.18396160341724,
+      "learning_rate": 1.6029195856841015e-05,
+      "loss": 0.8313,
+      "step": 6054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 36.839017335975925,
+      "learning_rate": 1.602785035869893e-05,
+      "loss": 0.7794,
+      "step": 6055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 32.22025285947191,
+      "learning_rate": 1.602650468912831e-05,
+      "loss": 0.7376,
+      "step": 6056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 50.31667012057441,
+      "learning_rate": 1.602515884816742e-05,
+      "loss": 0.8099,
+      "step": 6057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 39.849742519547185,
+      "learning_rate": 1.602381283585454e-05,
+      "loss": 0.7333,
+      "step": 6058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 48.018336176401746,
+      "learning_rate": 1.6022466652227944e-05,
+      "loss": 0.7617,
+      "step": 6059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.465216596443774,
+      "learning_rate": 1.602112029732592e-05,
+      "loss": 0.6405,
+      "step": 6060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 36.69104104767477,
+      "learning_rate": 1.6019773771186756e-05,
+      "loss": 0.7546,
+      "step": 6061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 44.6608705018917,
+      "learning_rate": 1.601842707384875e-05,
+      "loss": 0.7891,
+      "step": 6062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.81239100057584,
+      "learning_rate": 1.60170802053502e-05,
+      "loss": 0.8701,
+      "step": 6063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.1660187915962,
+      "learning_rate": 1.6015733165729406e-05,
+      "loss": 0.6713,
+      "step": 6064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 34.108858521378636,
+      "learning_rate": 1.601438595502468e-05,
+      "loss": 0.7477,
+      "step": 6065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 32.08601668436085,
+      "learning_rate": 1.6013038573274336e-05,
+      "loss": 0.67,
+      "step": 6066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.651242514412388,
+      "learning_rate": 1.6011691020516696e-05,
+      "loss": 0.7097,
+      "step": 6067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 41.276736247318695,
+      "learning_rate": 1.6010343296790074e-05,
+      "loss": 0.7672,
+      "step": 6068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 89.08608373161451,
+      "learning_rate": 1.6008995402132807e-05,
+      "loss": 0.7929,
+      "step": 6069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 39.76942270427294,
+      "learning_rate": 1.6007647336583227e-05,
+      "loss": 0.8202,
+      "step": 6070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 49.935479563683636,
+      "learning_rate": 1.600629910017967e-05,
+      "loss": 0.8381,
+      "step": 6071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.845521269287513,
+      "learning_rate": 1.600495069296048e-05,
+      "loss": 0.73,
+      "step": 6072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 50.782664732959994,
+      "learning_rate": 1.6003602114964007e-05,
+      "loss": 0.705,
+      "step": 6073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.668949150543085,
+      "learning_rate": 1.60022533662286e-05,
+      "loss": 0.7878,
+      "step": 6074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.892711010651915,
+      "learning_rate": 1.600090444679262e-05,
+      "loss": 0.8121,
+      "step": 6075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 44.871485630973226,
+      "learning_rate": 1.5999555356694426e-05,
+      "loss": 0.8653,
+      "step": 6076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 35.43398873360602,
+      "learning_rate": 1.599820609597239e-05,
+      "loss": 0.7044,
+      "step": 6077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 42.4730793877396,
+      "learning_rate": 1.5996856664664876e-05,
+      "loss": 0.8578,
+      "step": 6078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 50.00432840685671,
+      "learning_rate": 1.599550706281027e-05,
+      "loss": 0.8478,
+      "step": 6079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.306924098870525,
+      "learning_rate": 1.5994157290446953e-05,
+      "loss": 0.7751,
+      "step": 6080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.437679254998315,
+      "learning_rate": 1.5992807347613305e-05,
+      "loss": 0.7785,
+      "step": 6081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.195071767121277,
+      "learning_rate": 1.5991457234347725e-05,
+      "loss": 0.7859,
+      "step": 6082
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 29.936378812401685,
+      "learning_rate": 1.5990106950688605e-05,
+      "loss": 0.702,
+      "step": 6083
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 29.791377544111644,
+      "learning_rate": 1.5988756496674345e-05,
+      "loss": 0.8108,
+      "step": 6084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.956309764731536,
+      "learning_rate": 1.5987405872343358e-05,
+      "loss": 0.7194,
+      "step": 6085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 35.23553619839645,
+      "learning_rate": 1.5986055077734046e-05,
+      "loss": 0.7292,
+      "step": 6086
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 37.10928202085392,
+      "learning_rate": 1.5984704112884834e-05,
+      "loss": 0.9207,
+      "step": 6087
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.60749998940977,
+      "learning_rate": 1.5983352977834132e-05,
+      "loss": 0.8256,
+      "step": 6088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.278535477684024,
+      "learning_rate": 1.5982001672620376e-05,
+      "loss": 0.7663,
+      "step": 6089
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 36.94206416788704,
+      "learning_rate": 1.5980650197281994e-05,
+      "loss": 0.7718,
+      "step": 6090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.46925769620003,
+      "learning_rate": 1.597929855185741e-05,
+      "loss": 0.7713,
+      "step": 6091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 35.506049953649885,
+      "learning_rate": 1.597794673638508e-05,
+      "loss": 0.8679,
+      "step": 6092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 42.4645866166591,
+      "learning_rate": 1.597659475090344e-05,
+      "loss": 0.7397,
+      "step": 6093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.737768407815764,
+      "learning_rate": 1.5975242595450946e-05,
+      "loss": 0.7158,
+      "step": 6094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 30.169395123608375,
+      "learning_rate": 1.5973890270066042e-05,
+      "loss": 0.713,
+      "step": 6095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 38.00815955196688,
+      "learning_rate": 1.5972537774787194e-05,
+      "loss": 0.8449,
+      "step": 6096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.71498651451123,
+      "learning_rate": 1.597118510965287e-05,
+      "loss": 0.7964,
+      "step": 6097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 38.28697455130009,
+      "learning_rate": 1.5969832274701528e-05,
+      "loss": 0.7669,
+      "step": 6098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 46.70755757763254,
+      "learning_rate": 1.5968479269971654e-05,
+      "loss": 0.816,
+      "step": 6099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 37.61304838516709,
+      "learning_rate": 1.596712609550172e-05,
+      "loss": 0.8507,
+      "step": 6100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.921338663623008,
+      "learning_rate": 1.5965772751330207e-05,
+      "loss": 0.7094,
+      "step": 6101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 30.33934037670246,
+      "learning_rate": 1.596441923749561e-05,
+      "loss": 0.7313,
+      "step": 6102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 29.138078413646834,
+      "learning_rate": 1.5963065554036418e-05,
+      "loss": 0.7055,
+      "step": 6103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.85607183755162,
+      "learning_rate": 1.596171170099113e-05,
+      "loss": 0.7454,
+      "step": 6104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 39.60014457218676,
+      "learning_rate": 1.596035767839825e-05,
+      "loss": 0.8359,
+      "step": 6105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 33.592051518995106,
+      "learning_rate": 1.595900348629628e-05,
+      "loss": 0.7737,
+      "step": 6106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 29.212330224095133,
+      "learning_rate": 1.5957649124723743e-05,
+      "loss": 0.7143,
+      "step": 6107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 35.15728821980441,
+      "learning_rate": 1.5956294593719147e-05,
+      "loss": 0.6855,
+      "step": 6108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 40.95077307626961,
+      "learning_rate": 1.5954939893321016e-05,
+      "loss": 0.7748,
+      "step": 6109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 31.968932959516867,
+      "learning_rate": 1.595358502356788e-05,
+      "loss": 0.7886,
+      "step": 6110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 46.49387775644092,
+      "learning_rate": 1.5952229984498265e-05,
+      "loss": 0.8587,
+      "step": 6111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 34.526377145093846,
+      "learning_rate": 1.5950874776150715e-05,
+      "loss": 0.7482,
+      "step": 6112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 32.23623257645687,
+      "learning_rate": 1.5949519398563766e-05,
+      "loss": 0.7306,
+      "step": 6113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.58144721812578,
+      "learning_rate": 1.594816385177597e-05,
+      "loss": 0.7634,
+      "step": 6114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.48159819824958,
+      "learning_rate": 1.5946808135825866e-05,
+      "loss": 0.7592,
+      "step": 6115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 44.91648320489099,
+      "learning_rate": 1.5945452250752022e-05,
+      "loss": 0.8389,
+      "step": 6116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 34.66934957499868,
+      "learning_rate": 1.5944096196592994e-05,
+      "loss": 0.8228,
+      "step": 6117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.01848286442453,
+      "learning_rate": 1.5942739973387345e-05,
+      "loss": 0.6817,
+      "step": 6118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 44.707438981116375,
+      "learning_rate": 1.594138358117365e-05,
+      "loss": 0.8551,
+      "step": 6119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.25009285832038,
+      "learning_rate": 1.594002701999048e-05,
+      "loss": 0.6411,
+      "step": 6120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.6693070177499,
+      "learning_rate": 1.5938670289876418e-05,
+      "loss": 0.8214,
+      "step": 6121
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 31.910970523609624,
+      "learning_rate": 1.5937313390870045e-05,
+      "loss": 0.7618,
+      "step": 6122
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 43.52007919882833,
+      "learning_rate": 1.5935956323009954e-05,
+      "loss": 0.7045,
+      "step": 6123
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 41.00768921438047,
+      "learning_rate": 1.5934599086334733e-05,
+      "loss": 0.7896,
+      "step": 6124
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 47.25589259809259,
+      "learning_rate": 1.593324168088299e-05,
+      "loss": 0.7839,
+      "step": 6125
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 53.88571684826427,
+      "learning_rate": 1.5931884106693326e-05,
+      "loss": 0.8895,
+      "step": 6126
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 31.08678419595151,
+      "learning_rate": 1.5930526363804344e-05,
+      "loss": 0.7227,
+      "step": 6127
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.94809012498954,
+      "learning_rate": 1.5929168452254664e-05,
+      "loss": 0.7566,
+      "step": 6128
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 34.020785548746375,
+      "learning_rate": 1.5927810372082903e-05,
+      "loss": 0.7371,
+      "step": 6129
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.005163650267924,
+      "learning_rate": 1.592645212332768e-05,
+      "loss": 0.7514,
+      "step": 6130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.85202149652624,
+      "learning_rate": 1.5925093706027625e-05,
+      "loss": 0.8334,
+      "step": 6131
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.674923462061077,
+      "learning_rate": 1.5923735120221372e-05,
+      "loss": 0.6416,
+      "step": 6132
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.14565050521252,
+      "learning_rate": 1.5922376365947557e-05,
+      "loss": 0.7729,
+      "step": 6133
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 40.20234758297774,
+      "learning_rate": 1.5921017443244825e-05,
+      "loss": 0.7986,
+      "step": 6134
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.565893450935747,
+      "learning_rate": 1.5919658352151818e-05,
+      "loss": 0.6825,
+      "step": 6135
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.06028843251805,
+      "learning_rate": 1.5918299092707194e-05,
+      "loss": 0.7613,
+      "step": 6136
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.80841048328571,
+      "learning_rate": 1.5916939664949602e-05,
+      "loss": 0.7348,
+      "step": 6137
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 34.01689127368655,
+      "learning_rate": 1.591558006891771e-05,
+      "loss": 0.7033,
+      "step": 6138
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 39.804129102269826,
+      "learning_rate": 1.5914220304650177e-05,
+      "loss": 0.8391,
+      "step": 6139
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 46.2545785029543,
+      "learning_rate": 1.5912860372185683e-05,
+      "loss": 0.7491,
+      "step": 6140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 41.455024048822764,
+      "learning_rate": 1.5911500271562895e-05,
+      "loss": 0.6915,
+      "step": 6141
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 33.39358490966108,
+      "learning_rate": 1.5910140002820503e-05,
+      "loss": 0.7429,
+      "step": 6142
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.4524948428097,
+      "learning_rate": 1.5908779565997182e-05,
+      "loss": 0.8786,
+      "step": 6143
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 55.38640402998999,
+      "learning_rate": 1.5907418961131628e-05,
+      "loss": 0.7575,
+      "step": 6144
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 53.23081090767748,
+      "learning_rate": 1.5906058188262534e-05,
+      "loss": 0.7204,
+      "step": 6145
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 31.044887841977896,
+      "learning_rate": 1.5904697247428602e-05,
+      "loss": 0.7507,
+      "step": 6146
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.02478508396346,
+      "learning_rate": 1.5903336138668532e-05,
+      "loss": 0.8132,
+      "step": 6147
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.548795506040634,
+      "learning_rate": 1.5901974862021036e-05,
+      "loss": 0.6485,
+      "step": 6148
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 41.2708491839334,
+      "learning_rate": 1.5900613417524827e-05,
+      "loss": 0.7851,
+      "step": 6149
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 33.29595372443946,
+      "learning_rate": 1.5899251805218625e-05,
+      "loss": 0.788,
+      "step": 6150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 34.54861543809788,
+      "learning_rate": 1.5897890025141153e-05,
+      "loss": 0.8019,
+      "step": 6151
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 33.68714154748114,
+      "learning_rate": 1.589652807733114e-05,
+      "loss": 0.8429,
+      "step": 6152
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 39.430982997375466,
+      "learning_rate": 1.5895165961827315e-05,
+      "loss": 0.8307,
+      "step": 6153
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 43.23282829250464,
+      "learning_rate": 1.5893803678668424e-05,
+      "loss": 0.7451,
+      "step": 6154
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 96.27308405710907,
+      "learning_rate": 1.5892441227893197e-05,
+      "loss": 0.676,
+      "step": 6155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.520683123291622,
+      "learning_rate": 1.589107860954039e-05,
+      "loss": 0.7071,
+      "step": 6156
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.48221230797411,
+      "learning_rate": 1.588971582364876e-05,
+      "loss": 0.7563,
+      "step": 6157
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.16787111928027,
+      "learning_rate": 1.588835287025705e-05,
+      "loss": 0.8468,
+      "step": 6158
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.70111359207891,
+      "learning_rate": 1.588698974940403e-05,
+      "loss": 0.7663,
+      "step": 6159
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.87058343338205,
+      "learning_rate": 1.5885626461128467e-05,
+      "loss": 0.7157,
+      "step": 6160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 49.937179110836446,
+      "learning_rate": 1.588426300546913e-05,
+      "loss": 0.7478,
+      "step": 6161
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 35.9731439612693,
+      "learning_rate": 1.5882899382464795e-05,
+      "loss": 0.7708,
+      "step": 6162
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.88819986202249,
+      "learning_rate": 1.5881535592154245e-05,
+      "loss": 0.7453,
+      "step": 6163
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.443819384638864,
+      "learning_rate": 1.5880171634576262e-05,
+      "loss": 0.8249,
+      "step": 6164
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.99078103068131,
+      "learning_rate": 1.587880750976964e-05,
+      "loss": 0.7784,
+      "step": 6165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 44.60848306542507,
+      "learning_rate": 1.5877443217773166e-05,
+      "loss": 0.8683,
+      "step": 6166
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 36.611340380870345,
+      "learning_rate": 1.5876078758625647e-05,
+      "loss": 0.7458,
+      "step": 6167
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.349130799879404,
+      "learning_rate": 1.5874714132365887e-05,
+      "loss": 0.8232,
+      "step": 6168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 35.47720234858209,
+      "learning_rate": 1.5873349339032693e-05,
+      "loss": 0.8082,
+      "step": 6169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 31.325981359796167,
+      "learning_rate": 1.587198437866488e-05,
+      "loss": 0.6776,
+      "step": 6170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 29.1275965797893,
+      "learning_rate": 1.5870619251301267e-05,
+      "loss": 0.676,
+      "step": 6171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.28827567265984,
+      "learning_rate": 1.5869253956980672e-05,
+      "loss": 0.7801,
+      "step": 6172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 34.35793305928032,
+      "learning_rate": 1.5867888495741934e-05,
+      "loss": 0.685,
+      "step": 6173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 38.63467287959134,
+      "learning_rate": 1.5866522867623876e-05,
+      "loss": 0.7154,
+      "step": 6174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 33.076478393684184,
+      "learning_rate": 1.586515707266534e-05,
+      "loss": 0.8293,
+      "step": 6175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 41.66179622349868,
+      "learning_rate": 1.5863791110905172e-05,
+      "loss": 0.6856,
+      "step": 6176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 31.559294807928257,
+      "learning_rate": 1.586242498238221e-05,
+      "loss": 0.7411,
+      "step": 6177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 43.32309895971086,
+      "learning_rate": 1.5861058687135314e-05,
+      "loss": 0.7335,
+      "step": 6178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 37.12932363372855,
+      "learning_rate": 1.5859692225203335e-05,
+      "loss": 0.7779,
+      "step": 6179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 41.0722767587119,
+      "learning_rate": 1.585832559662514e-05,
+      "loss": 0.6994,
+      "step": 6180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.71958729056959,
+      "learning_rate": 1.585695880143959e-05,
+      "loss": 0.7695,
+      "step": 6181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.48615581926855,
+      "learning_rate": 1.5855591839685556e-05,
+      "loss": 0.6695,
+      "step": 6182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 38.861105490370456,
+      "learning_rate": 1.585422471140192e-05,
+      "loss": 0.7694,
+      "step": 6183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.23868621366819,
+      "learning_rate": 1.5852857416627552e-05,
+      "loss": 0.7106,
+      "step": 6184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 36.25325088064803,
+      "learning_rate": 1.585148995540135e-05,
+      "loss": 0.7886,
+      "step": 6185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 39.21140941298651,
+      "learning_rate": 1.585012232776219e-05,
+      "loss": 0.7893,
+      "step": 6186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 43.23099940570527,
+      "learning_rate": 1.584875453374898e-05,
+      "loss": 0.8169,
+      "step": 6187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.1332785444828,
+      "learning_rate": 1.5847386573400605e-05,
+      "loss": 0.7367,
+      "step": 6188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 39.44598945978174,
+      "learning_rate": 1.584601844675598e-05,
+      "loss": 0.7893,
+      "step": 6189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 49.7797513018994,
+      "learning_rate": 1.584465015385401e-05,
+      "loss": 0.8847,
+      "step": 6190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 36.51291751903637,
+      "learning_rate": 1.584328169473361e-05,
+      "loss": 0.8798,
+      "step": 6191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 40.746630947699565,
+      "learning_rate": 1.58419130694337e-05,
+      "loss": 0.8104,
+      "step": 6192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 43.085998975030336,
+      "learning_rate": 1.5840544277993193e-05,
+      "loss": 0.7732,
+      "step": 6193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.73023492310968,
+      "learning_rate": 1.5839175320451027e-05,
+      "loss": 0.6922,
+      "step": 6194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 27.37000589453244,
+      "learning_rate": 1.583780619684613e-05,
+      "loss": 0.7463,
+      "step": 6195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 43.09047805226593,
+      "learning_rate": 1.5836436907217438e-05,
+      "loss": 0.7422,
+      "step": 6196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.14377812692994,
+      "learning_rate": 1.58350674516039e-05,
+      "loss": 0.6956,
+      "step": 6197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.50833191612388,
+      "learning_rate": 1.583369783004445e-05,
+      "loss": 0.7759,
+      "step": 6198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 38.50369277187479,
+      "learning_rate": 1.5832328042578047e-05,
+      "loss": 0.6946,
+      "step": 6199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 34.455828059377495,
+      "learning_rate": 1.5830958089243652e-05,
+      "loss": 0.6944,
+      "step": 6200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.34081784336651,
+      "learning_rate": 1.5829587970080217e-05,
+      "loss": 0.7073,
+      "step": 6201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.86342973493056,
+      "learning_rate": 1.582821768512671e-05,
+      "loss": 0.8783,
+      "step": 6202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.612283830216594,
+      "learning_rate": 1.5826847234422102e-05,
+      "loss": 0.7883,
+      "step": 6203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 37.77466693443534,
+      "learning_rate": 1.5825476618005363e-05,
+      "loss": 0.7816,
+      "step": 6204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.21028003214984,
+      "learning_rate": 1.582410583591548e-05,
+      "loss": 0.8587,
+      "step": 6205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 33.862535161098336,
+      "learning_rate": 1.5822734888191437e-05,
+      "loss": 0.7342,
+      "step": 6206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 31.748747913572746,
+      "learning_rate": 1.5821363774872218e-05,
+      "loss": 0.7707,
+      "step": 6207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 42.00900940256781,
+      "learning_rate": 1.5819992495996818e-05,
+      "loss": 0.7039,
+      "step": 6208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 49.58292611646794,
+      "learning_rate": 1.5818621051604234e-05,
+      "loss": 0.857,
+      "step": 6209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 37.67444538854922,
+      "learning_rate": 1.5817249441733475e-05,
+      "loss": 0.8162,
+      "step": 6210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 37.341889396507774,
+      "learning_rate": 1.5815877666423543e-05,
+      "loss": 0.8306,
+      "step": 6211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 41.08862773460567,
+      "learning_rate": 1.5814505725713454e-05,
+      "loss": 0.7725,
+      "step": 6212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.792196626629156,
+      "learning_rate": 1.581313361964222e-05,
+      "loss": 0.8159,
+      "step": 6213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 37.102703138799974,
+      "learning_rate": 1.5811761348248872e-05,
+      "loss": 0.7817,
+      "step": 6214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 47.27926027456744,
+      "learning_rate": 1.5810388911572424e-05,
+      "loss": 0.8123,
+      "step": 6215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 33.85309959177553,
+      "learning_rate": 1.5809016309651916e-05,
+      "loss": 0.7106,
+      "step": 6216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 41.98869476980661,
+      "learning_rate": 1.5807643542526387e-05,
+      "loss": 0.784,
+      "step": 6217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 110.85820625877328,
+      "learning_rate": 1.580627061023487e-05,
+      "loss": 0.7547,
+      "step": 6218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 34.529019538339156,
+      "learning_rate": 1.5804897512816416e-05,
+      "loss": 0.7075,
+      "step": 6219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.70585293608087,
+      "learning_rate": 1.580352425031007e-05,
+      "loss": 0.7552,
+      "step": 6220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 38.28622775353695,
+      "learning_rate": 1.5802150822754888e-05,
+      "loss": 0.6913,
+      "step": 6221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 33.21510975397075,
+      "learning_rate": 1.580077723018993e-05,
+      "loss": 0.7563,
+      "step": 6222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 42.83105982258546,
+      "learning_rate": 1.5799403472654266e-05,
+      "loss": 0.8138,
+      "step": 6223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 41.91400529585576,
+      "learning_rate": 1.5798029550186957e-05,
+      "loss": 0.7325,
+      "step": 6224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 36.95094295132178,
+      "learning_rate": 1.5796655462827077e-05,
+      "loss": 0.7856,
+      "step": 6225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.32376692969233,
+      "learning_rate": 1.5795281210613708e-05,
+      "loss": 0.7045,
+      "step": 6226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 45.527760180152555,
+      "learning_rate": 1.5793906793585935e-05,
+      "loss": 0.8927,
+      "step": 6227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 38.38242263331953,
+      "learning_rate": 1.5792532211782837e-05,
+      "loss": 0.7306,
+      "step": 6228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 38.23535637634698,
+      "learning_rate": 1.5791157465243517e-05,
+      "loss": 0.8265,
+      "step": 6229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 41.26333786311069,
+      "learning_rate": 1.5789782554007063e-05,
+      "loss": 0.7835,
+      "step": 6230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 39.80997543774508,
+      "learning_rate": 1.578840747811258e-05,
+      "loss": 0.6851,
+      "step": 6231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 38.8372208746391,
+      "learning_rate": 1.5787032237599173e-05,
+      "loss": 0.8172,
+      "step": 6232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 46.87665212390033,
+      "learning_rate": 1.5785656832505956e-05,
+      "loss": 0.6824,
+      "step": 6233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.153662336272944,
+      "learning_rate": 1.5784281262872046e-05,
+      "loss": 0.6937,
+      "step": 6234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.09630194958301,
+      "learning_rate": 1.5782905528736558e-05,
+      "loss": 0.7398,
+      "step": 6235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 40.45670772760885,
+      "learning_rate": 1.578152963013862e-05,
+      "loss": 0.8381,
+      "step": 6236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 36.630382816038896,
+      "learning_rate": 1.578015356711736e-05,
+      "loss": 0.6851,
+      "step": 6237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 40.17069412092314,
+      "learning_rate": 1.5778777339711914e-05,
+      "loss": 0.7441,
+      "step": 6238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 29.055334938920545,
+      "learning_rate": 1.577740094796142e-05,
+      "loss": 0.6577,
+      "step": 6239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.77290914678107,
+      "learning_rate": 1.5776024391905026e-05,
+      "loss": 0.8386,
+      "step": 6240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 37.87800787322315,
+      "learning_rate": 1.5774647671581878e-05,
+      "loss": 0.814,
+      "step": 6241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 29.09628138608043,
+      "learning_rate": 1.5773270787031124e-05,
+      "loss": 0.6603,
+      "step": 6242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 39.01014551222123,
+      "learning_rate": 1.577189373829193e-05,
+      "loss": 0.7958,
+      "step": 6243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 45.92394134347886,
+      "learning_rate": 1.5770516525403453e-05,
+      "loss": 0.7708,
+      "step": 6244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 28.617228685787303,
+      "learning_rate": 1.576913914840486e-05,
+      "loss": 0.6817,
+      "step": 6245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 38.17649669361294,
+      "learning_rate": 1.5767761607335327e-05,
+      "loss": 0.8354,
+      "step": 6246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 59.9066698739188,
+      "learning_rate": 1.5766383902234026e-05,
+      "loss": 0.7086,
+      "step": 6247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 43.87700213241302,
+      "learning_rate": 1.5765006033140142e-05,
+      "loss": 0.8438,
+      "step": 6248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 34.01246762286422,
+      "learning_rate": 1.5763628000092858e-05,
+      "loss": 0.7463,
+      "step": 6249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 48.598824788629166,
+      "learning_rate": 1.5762249803131365e-05,
+      "loss": 0.837,
+      "step": 6250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 34.39857662050635,
+      "learning_rate": 1.5760871442294856e-05,
+      "loss": 0.6996,
+      "step": 6251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 45.49185579373235,
+      "learning_rate": 1.5759492917622537e-05,
+      "loss": 0.9143,
+      "step": 6252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 38.740630624531775,
+      "learning_rate": 1.5758114229153606e-05,
+      "loss": 0.7933,
+      "step": 6253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.340926578495978,
+      "learning_rate": 1.575673537692728e-05,
+      "loss": 0.7354,
+      "step": 6254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.30558545768053,
+      "learning_rate": 1.575535636098276e-05,
+      "loss": 0.7554,
+      "step": 6255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 40.41983454110295,
+      "learning_rate": 1.5753977181359277e-05,
+      "loss": 0.7336,
+      "step": 6256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 43.30996238683962,
+      "learning_rate": 1.5752597838096046e-05,
+      "loss": 0.7494,
+      "step": 6257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 45.653002900905506,
+      "learning_rate": 1.57512183312323e-05,
+      "loss": 0.793,
+      "step": 6258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 34.96975524521072,
+      "learning_rate": 1.574983866080727e-05,
+      "loss": 0.7312,
+      "step": 6259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 49.83976888991086,
+      "learning_rate": 1.5748458826860185e-05,
+      "loss": 0.8668,
+      "step": 6260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 50.16372487112437,
+      "learning_rate": 1.5747078829430302e-05,
+      "loss": 0.7499,
+      "step": 6261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 46.727160173327135,
+      "learning_rate": 1.5745698668556856e-05,
+      "loss": 0.8469,
+      "step": 6262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 36.27138049906921,
+      "learning_rate": 1.5744318344279103e-05,
+      "loss": 0.7047,
+      "step": 6263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 34.23471013805682,
+      "learning_rate": 1.5742937856636294e-05,
+      "loss": 0.6749,
+      "step": 6264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 36.35409998773765,
+      "learning_rate": 1.5741557205667688e-05,
+      "loss": 0.7542,
+      "step": 6265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.55009405481795,
+      "learning_rate": 1.574017639141256e-05,
+      "loss": 0.6954,
+      "step": 6266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 43.06516710224371,
+      "learning_rate": 1.5738795413910174e-05,
+      "loss": 0.7327,
+      "step": 6267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 39.54644725120593,
+      "learning_rate": 1.57374142731998e-05,
+      "loss": 0.836,
+      "step": 6268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 47.02754324059468,
+      "learning_rate": 1.573603296932072e-05,
+      "loss": 0.6796,
+      "step": 6269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 39.073280809062126,
+      "learning_rate": 1.5734651502312218e-05,
+      "loss": 0.7049,
+      "step": 6270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 34.903323714252195,
+      "learning_rate": 1.5733269872213583e-05,
+      "loss": 0.7083,
+      "step": 6271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 30.436931992543673,
+      "learning_rate": 1.5731888079064107e-05,
+      "loss": 0.8132,
+      "step": 6272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 40.48381859512211,
+      "learning_rate": 1.5730506122903086e-05,
+      "loss": 0.8034,
+      "step": 6273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 37.45882088864048,
+      "learning_rate": 1.5729124003769826e-05,
+      "loss": 0.7926,
+      "step": 6274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.67475555141688,
+      "learning_rate": 1.572774172170363e-05,
+      "loss": 0.7937,
+      "step": 6275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 28.784252442970896,
+      "learning_rate": 1.5726359276743808e-05,
+      "loss": 0.6938,
+      "step": 6276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 38.070672759776905,
+      "learning_rate": 1.5724976668929678e-05,
+      "loss": 0.7039,
+      "step": 6277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 38.33809127650166,
+      "learning_rate": 1.5723593898300562e-05,
+      "loss": 0.8177,
+      "step": 6278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 38.65326776002434,
+      "learning_rate": 1.572221096489578e-05,
+      "loss": 0.7799,
+      "step": 6279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.586323137780994,
+      "learning_rate": 1.572082786875467e-05,
+      "loss": 0.7184,
+      "step": 6280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.658907452098884,
+      "learning_rate": 1.5719444609916564e-05,
+      "loss": 0.7277,
+      "step": 6281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 63.597251492017065,
+      "learning_rate": 1.5718061188420793e-05,
+      "loss": 0.7764,
+      "step": 6282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 36.98603514735428,
+      "learning_rate": 1.571667760430671e-05,
+      "loss": 0.6896,
+      "step": 6283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 37.64761677148423,
+      "learning_rate": 1.5715293857613662e-05,
+      "loss": 0.7189,
+      "step": 6284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 56.95917828995185,
+      "learning_rate": 1.5713909948380995e-05,
+      "loss": 0.7385,
+      "step": 6285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.489139601509876,
+      "learning_rate": 1.5712525876648076e-05,
+      "loss": 0.7554,
+      "step": 6286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 36.55213764426167,
+      "learning_rate": 1.5711141642454258e-05,
+      "loss": 0.8128,
+      "step": 6287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 40.44060454913779,
+      "learning_rate": 1.5709757245838918e-05,
+      "loss": 0.822,
+      "step": 6288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 37.16269537460238,
+      "learning_rate": 1.570837268684142e-05,
+      "loss": 0.7057,
+      "step": 6289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 43.551048683784494,
+      "learning_rate": 1.5706987965501142e-05,
+      "loss": 0.9086,
+      "step": 6290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 33.61863003631538,
+      "learning_rate": 1.5705603081857463e-05,
+      "loss": 0.6787,
+      "step": 6291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.26308894471869,
+      "learning_rate": 1.5704218035949773e-05,
+      "loss": 0.7217,
+      "step": 6292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 42.55996716170909,
+      "learning_rate": 1.570283282781746e-05,
+      "loss": 0.8522,
+      "step": 6293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 42.39204131592571,
+      "learning_rate": 1.5701447457499914e-05,
+      "loss": 0.9423,
+      "step": 6294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 29.182185414139163,
+      "learning_rate": 1.570006192503654e-05,
+      "loss": 0.7155,
+      "step": 6295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 34.45132016668472,
+      "learning_rate": 1.5698676230466735e-05,
+      "loss": 0.7742,
+      "step": 6296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.31756583817327,
+      "learning_rate": 1.5697290373829913e-05,
+      "loss": 0.7492,
+      "step": 6297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 37.93941415045858,
+      "learning_rate": 1.5695904355165486e-05,
+      "loss": 0.7488,
+      "step": 6298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.43396751926294,
+      "learning_rate": 1.5694518174512873e-05,
+      "loss": 0.7881,
+      "step": 6299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 36.82072984639811,
+      "learning_rate": 1.5693131831911494e-05,
+      "loss": 0.9006,
+      "step": 6300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 37.33938970927259,
+      "learning_rate": 1.5691745327400776e-05,
+      "loss": 0.7222,
+      "step": 6301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.95465256183283,
+      "learning_rate": 1.569035866102015e-05,
+      "loss": 0.6752,
+      "step": 6302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 33.11481283555725,
+      "learning_rate": 1.568897183280905e-05,
+      "loss": 0.737,
+      "step": 6303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 28.34702756283657,
+      "learning_rate": 1.5687584842806925e-05,
+      "loss": 0.7423,
+      "step": 6304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.575742906163214,
+      "learning_rate": 1.568619769105321e-05,
+      "loss": 0.7248,
+      "step": 6305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 40.93101824049575,
+      "learning_rate": 1.568481037758736e-05,
+      "loss": 0.7969,
+      "step": 6306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 43.302413470554704,
+      "learning_rate": 1.5683422902448827e-05,
+      "loss": 0.8613,
+      "step": 6307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 36.49385691576631,
+      "learning_rate": 1.5682035265677074e-05,
+      "loss": 0.808,
+      "step": 6308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.52826241912484,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.7954,
+      "step": 6309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 40.75050682794123,
+      "learning_rate": 1.5679259507391755e-05,
+      "loss": 0.7848,
+      "step": 6310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.77322577551529,
+      "learning_rate": 1.5677871385957134e-05,
+      "loss": 0.8062,
+      "step": 6311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.345381169549654,
+      "learning_rate": 1.567648310304717e-05,
+      "loss": 0.757,
+      "step": 6312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.749873442660434,
+      "learning_rate": 1.567509465870135e-05,
+      "loss": 0.8059,
+      "step": 6313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 46.683627880850985,
+      "learning_rate": 1.567370605295915e-05,
+      "loss": 0.8429,
+      "step": 6314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 55.13808592054982,
+      "learning_rate": 1.5672317285860076e-05,
+      "loss": 0.6558,
+      "step": 6315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.51643380920909,
+      "learning_rate": 1.5670928357443617e-05,
+      "loss": 0.7367,
+      "step": 6316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.843888980398695,
+      "learning_rate": 1.566953926774927e-05,
+      "loss": 0.7835,
+      "step": 6317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 36.46829233805824,
+      "learning_rate": 1.5668150016816545e-05,
+      "loss": 0.8174,
+      "step": 6318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 43.186752737953135,
+      "learning_rate": 1.5666760604684947e-05,
+      "loss": 0.8211,
+      "step": 6319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 30.150663338666057,
+      "learning_rate": 1.5665371031393994e-05,
+      "loss": 0.6556,
+      "step": 6320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.454037920679205,
+      "learning_rate": 1.56639812969832e-05,
+      "loss": 0.7248,
+      "step": 6321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.4587960279429,
+      "learning_rate": 1.5662591401492096e-05,
+      "loss": 0.6987,
+      "step": 6322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.686154573994685,
+      "learning_rate": 1.5661201344960203e-05,
+      "loss": 0.6241,
+      "step": 6323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 32.52409523769148,
+      "learning_rate": 1.5659811127427053e-05,
+      "loss": 0.6426,
+      "step": 6324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 45.88682391303285,
+      "learning_rate": 1.5658420748932187e-05,
+      "loss": 0.6901,
+      "step": 6325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 35.768673044882135,
+      "learning_rate": 1.5657030209515146e-05,
+      "loss": 0.794,
+      "step": 6326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.28803491680358,
+      "learning_rate": 1.5655639509215476e-05,
+      "loss": 0.9265,
+      "step": 6327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 40.70099914036718,
+      "learning_rate": 1.565424864807273e-05,
+      "loss": 0.7726,
+      "step": 6328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.36941806443384,
+      "learning_rate": 1.565285762612645e-05,
+      "loss": 0.8246,
+      "step": 6329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.45682551966342,
+      "learning_rate": 1.5651466443416213e-05,
+      "loss": 0.833,
+      "step": 6330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.73378203688559,
+      "learning_rate": 1.5650075099981573e-05,
+      "loss": 0.7179,
+      "step": 6331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 28.364734296692333,
+      "learning_rate": 1.564868359586211e-05,
+      "loss": 0.7039,
+      "step": 6332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 40.937437811651385,
+      "learning_rate": 1.564729193109738e-05,
+      "loss": 0.7791,
+      "step": 6333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.05711630380712,
+      "learning_rate": 1.5645900105726976e-05,
+      "loss": 0.7333,
+      "step": 6334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.98670678981215,
+      "learning_rate": 1.5644508119790477e-05,
+      "loss": 0.787,
+      "step": 6335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.887956831922228,
+      "learning_rate": 1.5643115973327464e-05,
+      "loss": 0.6708,
+      "step": 6336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.64874611087064,
+      "learning_rate": 1.5641723666377536e-05,
+      "loss": 0.7903,
+      "step": 6337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.339466040507,
+      "learning_rate": 1.564033119898029e-05,
+      "loss": 0.8174,
+      "step": 6338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 35.512921157804826,
+      "learning_rate": 1.5638938571175324e-05,
+      "loss": 0.8734,
+      "step": 6339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.36391324980581,
+      "learning_rate": 1.563754578300224e-05,
+      "loss": 0.7878,
+      "step": 6340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.6296139639421,
+      "learning_rate": 1.5636152834500654e-05,
+      "loss": 0.7425,
+      "step": 6341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.64003454652172,
+      "learning_rate": 1.5634759725710178e-05,
+      "loss": 0.7626,
+      "step": 6342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 30.461805653632176,
+      "learning_rate": 1.5633366456670433e-05,
+      "loss": 0.7503,
+      "step": 6343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.262680197176984,
+      "learning_rate": 1.563197302742104e-05,
+      "loss": 0.7152,
+      "step": 6344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.676188641429984,
+      "learning_rate": 1.5630579438001626e-05,
+      "loss": 0.7127,
+      "step": 6345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 42.42811372802747,
+      "learning_rate": 1.562918568845183e-05,
+      "loss": 0.7676,
+      "step": 6346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.11649481240458,
+      "learning_rate": 1.562779177881129e-05,
+      "loss": 0.7655,
+      "step": 6347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 34.769095388641325,
+      "learning_rate": 1.5626397709119638e-05,
+      "loss": 0.7565,
+      "step": 6348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.91623762731489,
+      "learning_rate": 1.562500347941653e-05,
+      "loss": 0.7673,
+      "step": 6349
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.07916978169029,
+      "learning_rate": 1.5623609089741608e-05,
+      "loss": 0.6613,
+      "step": 6350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.01036669469204,
+      "learning_rate": 1.5622214540134536e-05,
+      "loss": 0.8151,
+      "step": 6351
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 32.90133559290096,
+      "learning_rate": 1.5620819830634975e-05,
+      "loss": 0.6974,
+      "step": 6352
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 42.55285844536578,
+      "learning_rate": 1.5619424961282585e-05,
+      "loss": 0.7783,
+      "step": 6353
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 47.063994273746104,
+      "learning_rate": 1.5618029932117035e-05,
+      "loss": 0.7619,
+      "step": 6354
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 37.34061401160153,
+      "learning_rate": 1.5616634743177996e-05,
+      "loss": 0.7027,
+      "step": 6355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 30.080740242749815,
+      "learning_rate": 1.5615239394505154e-05,
+      "loss": 0.6612,
+      "step": 6356
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 32.41212887167164,
+      "learning_rate": 1.5613843886138192e-05,
+      "loss": 0.6485,
+      "step": 6357
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 35.0530151723233,
+      "learning_rate": 1.561244821811679e-05,
+      "loss": 0.7802,
+      "step": 6358
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.54506823745433,
+      "learning_rate": 1.5611052390480646e-05,
+      "loss": 0.7935,
+      "step": 6359
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.225015711439795,
+      "learning_rate": 1.5609656403269452e-05,
+      "loss": 0.7392,
+      "step": 6360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 36.24461596066712,
+      "learning_rate": 1.560826025652291e-05,
+      "loss": 0.7375,
+      "step": 6361
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 53.88807798512339,
+      "learning_rate": 1.560686395028073e-05,
+      "loss": 0.7521,
+      "step": 6362
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 33.21090222278087,
+      "learning_rate": 1.560546748458262e-05,
+      "loss": 0.6798,
+      "step": 6363
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 37.80608105677706,
+      "learning_rate": 1.5604070859468292e-05,
+      "loss": 0.7741,
+      "step": 6364
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.617613749743725,
+      "learning_rate": 1.5602674074977467e-05,
+      "loss": 0.7938,
+      "step": 6365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 32.90950235594475,
+      "learning_rate": 1.560127713114987e-05,
+      "loss": 0.7913,
+      "step": 6366
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.381962648450916,
+      "learning_rate": 1.559988002802523e-05,
+      "loss": 0.8259,
+      "step": 6367
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.49432358799507,
+      "learning_rate": 1.5598482765643273e-05,
+      "loss": 0.7462,
+      "step": 6368
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 40.51955696913461,
+      "learning_rate": 1.5597085344043742e-05,
+      "loss": 0.7441,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 38.01704495024518,
+      "learning_rate": 1.5595687763266378e-05,
+      "loss": 0.7379,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 41.69714261106131,
+      "learning_rate": 1.5594290023350933e-05,
+      "loss": 0.7664,
+      "step": 6371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 56.244437921825025,
+      "learning_rate": 1.5592892124337145e-05,
+      "loss": 0.7175,
+      "step": 6372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.79167130902349,
+      "learning_rate": 1.559149406626478e-05,
+      "loss": 0.7063,
+      "step": 6373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.76778480250201,
+      "learning_rate": 1.5590095849173597e-05,
+      "loss": 0.7881,
+      "step": 6374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 36.00000076823763,
+      "learning_rate": 1.5588697473103355e-05,
+      "loss": 0.731,
+      "step": 6375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.945560949201536,
+      "learning_rate": 1.5587298938093828e-05,
+      "loss": 0.7631,
+      "step": 6376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 54.931848037802865,
+      "learning_rate": 1.5585900244184785e-05,
+      "loss": 0.8192,
+      "step": 6377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 36.647001490029446,
+      "learning_rate": 1.5584501391416008e-05,
+      "loss": 0.7704,
+      "step": 6378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.42806056718561,
+      "learning_rate": 1.558310237982728e-05,
+      "loss": 0.681,
+      "step": 6379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.55949000850301,
+      "learning_rate": 1.558170320945838e-05,
+      "loss": 0.5974,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 42.35783333894488,
+      "learning_rate": 1.5580303880349113e-05,
+      "loss": 0.7896,
+      "step": 6381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.07916960468503,
+      "learning_rate": 1.5578904392539268e-05,
+      "loss": 0.8067,
+      "step": 6382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.452696214979,
+      "learning_rate": 1.557750474606864e-05,
+      "loss": 0.7784,
+      "step": 6383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.18429622524046,
+      "learning_rate": 1.5576104940977045e-05,
+      "loss": 0.7994,
+      "step": 6384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.354271282057354,
+      "learning_rate": 1.5574704977304286e-05,
+      "loss": 0.7571,
+      "step": 6385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 40.62386965426229,
+      "learning_rate": 1.557330485509018e-05,
+      "loss": 0.7433,
+      "step": 6386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.063829346104626,
+      "learning_rate": 1.5571904574374543e-05,
+      "loss": 0.7194,
+      "step": 6387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.574604585198955,
+      "learning_rate": 1.55705041351972e-05,
+      "loss": 0.7802,
+      "step": 6388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.119413673163876,
+      "learning_rate": 1.5569103537597972e-05,
+      "loss": 0.6967,
+      "step": 6389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.50995656698537,
+      "learning_rate": 1.55677027816167e-05,
+      "loss": 0.8075,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 30.423969980371883,
+      "learning_rate": 1.5566301867293223e-05,
+      "loss": 0.7645,
+      "step": 6391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 41.7839244537222,
+      "learning_rate": 1.5564900794667372e-05,
+      "loss": 0.8469,
+      "step": 6392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 30.251475250904868,
+      "learning_rate": 1.5563499563779002e-05,
+      "loss": 0.6507,
+      "step": 6393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 38.03874517916803,
+      "learning_rate": 1.5562098174667957e-05,
+      "loss": 0.9462,
+      "step": 6394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 48.97902063434249,
+      "learning_rate": 1.556069662737409e-05,
+      "loss": 0.7322,
+      "step": 6395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.210384704397065,
+      "learning_rate": 1.5559294921937272e-05,
+      "loss": 0.725,
+      "step": 6396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.559472522275485,
+      "learning_rate": 1.555789305839735e-05,
+      "loss": 0.7591,
+      "step": 6397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.76410638471326,
+      "learning_rate": 1.5556491036794204e-05,
+      "loss": 0.7501,
+      "step": 6398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 29.926672267558153,
+      "learning_rate": 1.5555088857167703e-05,
+      "loss": 0.6808,
+      "step": 6399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 40.46152052944486,
+      "learning_rate": 1.5553686519557726e-05,
+      "loss": 0.7601,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 45.40779277073482,
+      "learning_rate": 1.5552284024004154e-05,
+      "loss": 0.7556,
+      "step": 6401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.158177535819846,
+      "learning_rate": 1.555088137054687e-05,
+      "loss": 0.7009,
+      "step": 6402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 36.33555605451731,
+      "learning_rate": 1.554947855922577e-05,
+      "loss": 0.826,
+      "step": 6403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 36.44787127802791,
+      "learning_rate": 1.5548075590080745e-05,
+      "loss": 0.8605,
+      "step": 6404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 28.873561321682327,
+      "learning_rate": 1.5546672463151695e-05,
+      "loss": 0.648,
+      "step": 6405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.306904936636336,
+      "learning_rate": 1.5545269178478523e-05,
+      "loss": 0.7376,
+      "step": 6406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.298822367314074,
+      "learning_rate": 1.554386573610114e-05,
+      "loss": 0.6489,
+      "step": 6407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.12596585177785,
+      "learning_rate": 1.554246213605946e-05,
+      "loss": 0.7939,
+      "step": 6408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.58634791314885,
+      "learning_rate": 1.55410583783934e-05,
+      "loss": 0.82,
+      "step": 6409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.63382579592899,
+      "learning_rate": 1.5539654463142878e-05,
+      "loss": 0.807,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 31.209225329056167,
+      "learning_rate": 1.5538250390347825e-05,
+      "loss": 0.7408,
+      "step": 6411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.07514807595159,
+      "learning_rate": 1.5536846160048172e-05,
+      "loss": 0.709,
+      "step": 6412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 41.50264226016454,
+      "learning_rate": 1.553544177228385e-05,
+      "loss": 0.793,
+      "step": 6413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.37645841706572,
+      "learning_rate": 1.5534037227094807e-05,
+      "loss": 0.8248,
+      "step": 6414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.24587287697316,
+      "learning_rate": 1.553263252452098e-05,
+      "loss": 0.7439,
+      "step": 6415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 30.878191381452027,
+      "learning_rate": 1.553122766460232e-05,
+      "loss": 0.6702,
+      "step": 6416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 34.715609723488754,
+      "learning_rate": 1.552982264737878e-05,
+      "loss": 0.8254,
+      "step": 6417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 45.74634891803995,
+      "learning_rate": 1.5528417472890324e-05,
+      "loss": 0.8544,
+      "step": 6418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 42.47003961475933,
+      "learning_rate": 1.5527012141176904e-05,
+      "loss": 0.7788,
+      "step": 6419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.82988629658568,
+      "learning_rate": 1.5525606652278493e-05,
+      "loss": 0.7457,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.5959804562831,
+      "learning_rate": 1.5524201006235063e-05,
+      "loss": 0.8194,
+      "step": 6421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 38.33443459331794,
+      "learning_rate": 1.5522795203086584e-05,
+      "loss": 0.7163,
+      "step": 6422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.82107720314123,
+      "learning_rate": 1.5521389242873044e-05,
+      "loss": 0.7402,
+      "step": 6423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 39.098764934936355,
+      "learning_rate": 1.551998312563442e-05,
+      "loss": 0.7698,
+      "step": 6424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 38.19932725748382,
+      "learning_rate": 1.551857685141071e-05,
+      "loss": 0.7541,
+      "step": 6425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.81700649656317,
+      "learning_rate": 1.5517170420241897e-05,
+      "loss": 0.8039,
+      "step": 6426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.372181254393254,
+      "learning_rate": 1.551576383216799e-05,
+      "loss": 0.7934,
+      "step": 6427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 49.15053807446248,
+      "learning_rate": 1.5514357087228985e-05,
+      "loss": 0.7462,
+      "step": 6428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.02398244892471,
+      "learning_rate": 1.551295018546489e-05,
+      "loss": 0.7432,
+      "step": 6429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 37.566531328387946,
+      "learning_rate": 1.5511543126915713e-05,
+      "loss": 0.7687,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.44014265577977,
+      "learning_rate": 1.5510135911621474e-05,
+      "loss": 0.7577,
+      "step": 6431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 36.8173631069187,
+      "learning_rate": 1.5508728539622196e-05,
+      "loss": 0.7876,
+      "step": 6432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.647522861716034,
+      "learning_rate": 1.5507321010957903e-05,
+      "loss": 0.7428,
+      "step": 6433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 28.73759699602255,
+      "learning_rate": 1.550591332566862e-05,
+      "loss": 0.6516,
+      "step": 6434
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 33.60086823034228,
+      "learning_rate": 1.550450548379438e-05,
+      "loss": 0.6345,
+      "step": 6435
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 44.62881219699557,
+      "learning_rate": 1.5503097485375228e-05,
+      "loss": 0.6894,
+      "step": 6436
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 36.63758242404898,
+      "learning_rate": 1.5501689330451203e-05,
+      "loss": 0.7091,
+      "step": 6437
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 40.8155955474072,
+      "learning_rate": 1.5500281019062347e-05,
+      "loss": 0.7804,
+      "step": 6438
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.21422261672208,
+      "learning_rate": 1.5498872551248722e-05,
+      "loss": 0.8241,
+      "step": 6439
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 37.71468397725981,
+      "learning_rate": 1.5497463927050375e-05,
+      "loss": 0.8223,
+      "step": 6440
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 44.68329449051607,
+      "learning_rate": 1.5496055146507368e-05,
+      "loss": 0.8753,
+      "step": 6441
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 31.409320684979424,
+      "learning_rate": 1.5494646209659775e-05,
+      "loss": 0.7194,
+      "step": 6442
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 35.09806490125812,
+      "learning_rate": 1.549323711654765e-05,
+      "loss": 0.697,
+      "step": 6443
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 39.12922390937105,
+      "learning_rate": 1.549182786721108e-05,
+      "loss": 0.8265,
+      "step": 6444
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 31.53474084454596,
+      "learning_rate": 1.5490418461690137e-05,
+      "loss": 0.739,
+      "step": 6445
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 29.670996959806335,
+      "learning_rate": 1.5489008900024903e-05,
+      "loss": 0.6712,
+      "step": 6446
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 49.7942609768191,
+      "learning_rate": 1.5487599182255467e-05,
+      "loss": 0.7111,
+      "step": 6447
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 40.585155913507116,
+      "learning_rate": 1.5486189308421922e-05,
+      "loss": 0.6969,
+      "step": 6448
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 28.772518419827794,
+      "learning_rate": 1.5484779278564363e-05,
+      "loss": 0.7387,
+      "step": 6449
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.50407030448146,
+      "learning_rate": 1.5483369092722888e-05,
+      "loss": 0.7308,
+      "step": 6450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 37.82121857312479,
+      "learning_rate": 1.5481958750937605e-05,
+      "loss": 0.8193,
+      "step": 6451
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 28.891938760945255,
+      "learning_rate": 1.548054825324862e-05,
+      "loss": 0.6515,
+      "step": 6452
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 29.094149133011573,
+      "learning_rate": 1.547913759969605e-05,
+      "loss": 0.636,
+      "step": 6453
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 35.01191996523285,
+      "learning_rate": 1.5477726790320012e-05,
+      "loss": 0.7042,
+      "step": 6454
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.25587538980833,
+      "learning_rate": 1.547631582516063e-05,
+      "loss": 0.8202,
+      "step": 6455
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 30.63015613357947,
+      "learning_rate": 1.5474904704258027e-05,
+      "loss": 0.6663,
+      "step": 6456
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.255044621316166,
+      "learning_rate": 1.5473493427652337e-05,
+      "loss": 0.757,
+      "step": 6457
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 33.23349197642214,
+      "learning_rate": 1.54720819953837e-05,
+      "loss": 0.7079,
+      "step": 6458
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 29.1993507081179,
+      "learning_rate": 1.547067040749225e-05,
+      "loss": 0.6721,
+      "step": 6459
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.21835169046609,
+      "learning_rate": 1.546925866401813e-05,
+      "loss": 0.7782,
+      "step": 6460
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 32.565049626262656,
+      "learning_rate": 1.5467846765001496e-05,
+      "loss": 0.7037,
+      "step": 6461
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 45.16321290923997,
+      "learning_rate": 1.54664347104825e-05,
+      "loss": 0.7508,
+      "step": 6462
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.41211688816999,
+      "learning_rate": 1.5465022500501294e-05,
+      "loss": 0.7709,
+      "step": 6463
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.05044865961256,
+      "learning_rate": 1.5463610135098048e-05,
+      "loss": 0.8591,
+      "step": 6464
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 32.36055110265848,
+      "learning_rate": 1.5462197614312926e-05,
+      "loss": 0.7057,
+      "step": 6465
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 39.7649018107607,
+      "learning_rate": 1.54607849381861e-05,
+      "loss": 0.623,
+      "step": 6466
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 54.27064553763419,
+      "learning_rate": 1.5459372106757742e-05,
+      "loss": 0.7842,
+      "step": 6467
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 50.71610571830707,
+      "learning_rate": 1.5457959120068036e-05,
+      "loss": 0.7633,
+      "step": 6468
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 31.43375664146506,
+      "learning_rate": 1.5456545978157168e-05,
+      "loss": 0.717,
+      "step": 6469
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 36.45767023010314,
+      "learning_rate": 1.545513268106532e-05,
+      "loss": 0.7055,
+      "step": 6470
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 56.04736877087775,
+      "learning_rate": 1.5453719228832692e-05,
+      "loss": 0.7246,
+      "step": 6471
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 42.77865275042755,
+      "learning_rate": 1.5452305621499483e-05,
+      "loss": 0.8298,
+      "step": 6472
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 32.088114242707945,
+      "learning_rate": 1.5450891859105884e-05,
+      "loss": 0.7562,
+      "step": 6473
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 29.492153542606435,
+      "learning_rate": 1.5449477941692114e-05,
+      "loss": 0.7142,
+      "step": 6474
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 32.902677151855464,
+      "learning_rate": 1.544806386929838e-05,
+      "loss": 0.7566,
+      "step": 6475
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 30.860931463464542,
+      "learning_rate": 1.5446649641964895e-05,
+      "loss": 0.683,
+      "step": 6476
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 31.793268457526374,
+      "learning_rate": 1.544523525973188e-05,
+      "loss": 0.749,
+      "step": 6477
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.419071138586716,
+      "learning_rate": 1.544382072263956e-05,
+      "loss": 0.7727,
+      "step": 6478
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 35.85018966151123,
+      "learning_rate": 1.5442406030728166e-05,
+      "loss": 0.7681,
+      "step": 6479
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 30.030391307705056,
+      "learning_rate": 1.5440991184037924e-05,
+      "loss": 0.6939,
+      "step": 6480
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 38.02074476292715,
+      "learning_rate": 1.5439576182609077e-05,
+      "loss": 0.7471,
+      "step": 6481
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.48175556655875,
+      "learning_rate": 1.5438161026481866e-05,
+      "loss": 0.77,
+      "step": 6482
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 38.25798909273341,
+      "learning_rate": 1.5436745715696535e-05,
+      "loss": 0.7045,
+      "step": 6483
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.46274309913146,
+      "learning_rate": 1.543533025029334e-05,
+      "loss": 0.6876,
+      "step": 6484
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 41.09503977096829,
+      "learning_rate": 1.5433914630312527e-05,
+      "loss": 0.6185,
+      "step": 6485
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.53524577984932,
+      "learning_rate": 1.5432498855794364e-05,
+      "loss": 0.8262,
+      "step": 6486
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 33.55368725459333,
+      "learning_rate": 1.543108292677911e-05,
+      "loss": 0.6495,
+      "step": 6487
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 36.34302325846431,
+      "learning_rate": 1.5429666843307035e-05,
+      "loss": 0.7907,
+      "step": 6488
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 33.76563025690309,
+      "learning_rate": 1.5428250605418414e-05,
+      "loss": 0.7128,
+      "step": 6489
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 50.451220549987895,
+      "learning_rate": 1.542683421315352e-05,
+      "loss": 0.7618,
+      "step": 6490
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 30.60811789220703,
+      "learning_rate": 1.5425417666552635e-05,
+      "loss": 0.7299,
+      "step": 6491
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 36.40267642577642,
+      "learning_rate": 1.5424000965656042e-05,
+      "loss": 0.7295,
+      "step": 6492
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.84496908407363,
+      "learning_rate": 1.542258411050404e-05,
+      "loss": 0.7054,
+      "step": 6493
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 31.950617169969686,
+      "learning_rate": 1.5421167101136917e-05,
+      "loss": 0.6779,
+      "step": 6494
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 35.56736781052767,
+      "learning_rate": 1.5419749937594967e-05,
+      "loss": 0.735,
+      "step": 6495
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 38.4475503017127,
+      "learning_rate": 1.5418332619918507e-05,
+      "loss": 0.6568,
+      "step": 6496
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 39.89646806668498,
+      "learning_rate": 1.541691514814783e-05,
+      "loss": 0.704,
+      "step": 6497
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 36.33404256882354,
+      "learning_rate": 1.541549752232326e-05,
+      "loss": 0.6937,
+      "step": 6498
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.745047301990766,
+      "learning_rate": 1.541407974248511e-05,
+      "loss": 0.6956,
+      "step": 6499
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.22385250501582,
+      "learning_rate": 1.5412661808673694e-05,
+      "loss": 0.7342,
+      "step": 6500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 41.35817676418218,
+      "learning_rate": 1.5411243720929342e-05,
+      "loss": 0.8195,
+      "step": 6501
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 44.35982700237202,
+      "learning_rate": 1.5409825479292388e-05,
+      "loss": 0.7737,
+      "step": 6502
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.99931886245773,
+      "learning_rate": 1.5408407083803162e-05,
+      "loss": 0.7152,
+      "step": 6503
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 45.240693800359345,
+      "learning_rate": 1.5406988534502002e-05,
+      "loss": 0.8647,
+      "step": 6504
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 33.66759015499039,
+      "learning_rate": 1.5405569831429247e-05,
+      "loss": 0.689,
+      "step": 6505
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.92832801274304,
+      "learning_rate": 1.5404150974625254e-05,
+      "loss": 0.7444,
+      "step": 6506
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.76756475501133,
+      "learning_rate": 1.5402731964130365e-05,
+      "loss": 0.725,
+      "step": 6507
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 31.135135631924406,
+      "learning_rate": 1.5401312799984943e-05,
+      "loss": 0.7072,
+      "step": 6508
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 42.193860028134324,
+      "learning_rate": 1.539989348222934e-05,
+      "loss": 0.7594,
+      "step": 6509
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.068559847954674,
+      "learning_rate": 1.5398474010903927e-05,
+      "loss": 0.7837,
+      "step": 6510
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.415286770998584,
+      "learning_rate": 1.5397054386049072e-05,
+      "loss": 0.6572,
+      "step": 6511
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.649259006652464,
+      "learning_rate": 1.5395634607705145e-05,
+      "loss": 0.8937,
+      "step": 6512
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.017587376466416,
+      "learning_rate": 1.539421467591253e-05,
+      "loss": 0.6757,
+      "step": 6513
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 43.945069297287105,
+      "learning_rate": 1.5392794590711605e-05,
+      "loss": 0.8104,
+      "step": 6514
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 46.60194833472662,
+      "learning_rate": 1.5391374352142752e-05,
+      "loss": 0.7307,
+      "step": 6515
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.11390988255453,
+      "learning_rate": 1.538995396024637e-05,
+      "loss": 0.7467,
+      "step": 6516
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.73552277920571,
+      "learning_rate": 1.5388533415062848e-05,
+      "loss": 0.5773,
+      "step": 6517
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 31.236182992045002,
+      "learning_rate": 1.5387112716632594e-05,
+      "loss": 0.6768,
+      "step": 6518
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 32.38790900201465,
+      "learning_rate": 1.5385691864995998e-05,
+      "loss": 0.727,
+      "step": 6519
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 28.89050199277201,
+      "learning_rate": 1.5384270860193477e-05,
+      "loss": 0.7004,
+      "step": 6520
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 50.226483478920095,
+      "learning_rate": 1.5382849702265447e-05,
+      "loss": 0.7442,
+      "step": 6521
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 35.80758553716117,
+      "learning_rate": 1.538142839125232e-05,
+      "loss": 0.6726,
+      "step": 6522
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 44.18543381021634,
+      "learning_rate": 1.538000692719451e-05,
+      "loss": 0.7928,
+      "step": 6523
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.31193050405256,
+      "learning_rate": 1.5378585310132458e-05,
+      "loss": 0.7386,
+      "step": 6524
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 46.815370570994354,
+      "learning_rate": 1.5377163540106582e-05,
+      "loss": 0.829,
+      "step": 6525
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.106663760862595,
+      "learning_rate": 1.5375741617157324e-05,
+      "loss": 0.7179,
+      "step": 6526
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 46.47145352070168,
+      "learning_rate": 1.5374319541325114e-05,
+      "loss": 0.7521,
+      "step": 6527
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.80882466658841,
+      "learning_rate": 1.5372897312650406e-05,
+      "loss": 0.7887,
+      "step": 6528
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 37.216013108716325,
+      "learning_rate": 1.5371474931173638e-05,
+      "loss": 0.7598,
+      "step": 6529
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.17807209804017,
+      "learning_rate": 1.5370052396935268e-05,
+      "loss": 0.839,
+      "step": 6530
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 35.89100542477183,
+      "learning_rate": 1.5368629709975745e-05,
+      "loss": 0.6285,
+      "step": 6531
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.93150267347535,
+      "learning_rate": 1.5367206870335536e-05,
+      "loss": 0.8456,
+      "step": 6532
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 43.327162697881036,
+      "learning_rate": 1.5365783878055103e-05,
+      "loss": 0.8118,
+      "step": 6533
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.09224143346064,
+      "learning_rate": 1.5364360733174916e-05,
+      "loss": 0.6961,
+      "step": 6534
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 44.39828215703952,
+      "learning_rate": 1.536293743573545e-05,
+      "loss": 0.8356,
+      "step": 6535
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 40.10014848800526,
+      "learning_rate": 1.5361513985777175e-05,
+      "loss": 0.6727,
+      "step": 6536
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 37.16703455077199,
+      "learning_rate": 1.536009038334058e-05,
+      "loss": 0.7517,
+      "step": 6537
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.77400604515236,
+      "learning_rate": 1.5358666628466154e-05,
+      "loss": 0.7302,
+      "step": 6538
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 31.989079836120215,
+      "learning_rate": 1.535724272119438e-05,
+      "loss": 0.7308,
+      "step": 6539
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 31.10132598403732,
+      "learning_rate": 1.535581866156576e-05,
+      "loss": 0.6397,
+      "step": 6540
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.355378850623104,
+      "learning_rate": 1.535439444962079e-05,
+      "loss": 0.7701,
+      "step": 6541
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 54.558488626270254,
+      "learning_rate": 1.535297008539997e-05,
+      "loss": 0.83,
+      "step": 6542
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 32.347158491978085,
+      "learning_rate": 1.5351545568943818e-05,
+      "loss": 0.6995,
+      "step": 6543
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.08947990226638,
+      "learning_rate": 1.5350120900292833e-05,
+      "loss": 0.8036,
+      "step": 6544
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 36.93163275230647,
+      "learning_rate": 1.5348696079487547e-05,
+      "loss": 0.7469,
+      "step": 6545
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.02747837250658,
+      "learning_rate": 1.534727110656847e-05,
+      "loss": 0.7102,
+      "step": 6546
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.549917365179354,
+      "learning_rate": 1.534584598157613e-05,
+      "loss": 0.6881,
+      "step": 6547
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.21225294657848,
+      "learning_rate": 1.5344420704551058e-05,
+      "loss": 0.7626,
+      "step": 6548
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 37.65957382538381,
+      "learning_rate": 1.534299527553379e-05,
+      "loss": 0.8705,
+      "step": 6549
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.00247182696392,
+      "learning_rate": 1.534156969456486e-05,
+      "loss": 0.6645,
+      "step": 6550
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 50.38596231815618,
+      "learning_rate": 1.5340143961684813e-05,
+      "loss": 0.8575,
+      "step": 6551
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 37.032678555646754,
+      "learning_rate": 1.5338718076934195e-05,
+      "loss": 0.6441,
+      "step": 6552
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 45.509897025504614,
+      "learning_rate": 1.5337292040353555e-05,
+      "loss": 0.6996,
+      "step": 6553
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 31.219206572774787,
+      "learning_rate": 1.5335865851983456e-05,
+      "loss": 0.6134,
+      "step": 6554
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.28525902100368,
+      "learning_rate": 1.5334439511864453e-05,
+      "loss": 0.7378,
+      "step": 6555
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 33.12917713962665,
+      "learning_rate": 1.533301302003711e-05,
+      "loss": 0.6817,
+      "step": 6556
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 37.12839915066236,
+      "learning_rate": 1.5331586376541997e-05,
+      "loss": 0.7475,
+      "step": 6557
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 34.11397060552613,
+      "learning_rate": 1.5330159581419687e-05,
+      "loss": 0.8011,
+      "step": 6558
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 46.05265217846575,
+      "learning_rate": 1.532873263471075e-05,
+      "loss": 0.7813,
+      "step": 6559
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 48.916486367191766,
+      "learning_rate": 1.5327305536455786e-05,
+      "loss": 0.7429,
+      "step": 6560
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 39.79112622990855,
+      "learning_rate": 1.5325878286695362e-05,
+      "loss": 0.7387,
+      "step": 6561
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.4618758963308,
+      "learning_rate": 1.5324450885470078e-05,
+      "loss": 0.7096,
+      "step": 6562
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 32.22949036400924,
+      "learning_rate": 1.5323023332820517e-05,
+      "loss": 0.7203,
+      "step": 6563
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 52.47083886528526,
+      "learning_rate": 1.5321595628787297e-05,
+      "loss": 0.7308,
+      "step": 6564
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 32.55741457849073,
+      "learning_rate": 1.5320167773411004e-05,
+      "loss": 0.6813,
+      "step": 6565
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 39.05143620478962,
+      "learning_rate": 1.5318739766732255e-05,
+      "loss": 0.7819,
+      "step": 6566
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.432352248567696,
+      "learning_rate": 1.5317311608791656e-05,
+      "loss": 0.7737,
+      "step": 6567
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 44.43348284294623,
+      "learning_rate": 1.5315883299629825e-05,
+      "loss": 0.8234,
+      "step": 6568
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 43.000688777441646,
+      "learning_rate": 1.531445483928738e-05,
+      "loss": 0.8474,
+      "step": 6569
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 33.64634326079926,
+      "learning_rate": 1.531302622780496e-05,
+      "loss": 0.6788,
+      "step": 6570
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 33.29667583318295,
+      "learning_rate": 1.5311597465223173e-05,
+      "loss": 0.689,
+      "step": 6571
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 37.22493970442091,
+      "learning_rate": 1.531016855158266e-05,
+      "loss": 0.7211,
+      "step": 6572
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.01038578532493,
+      "learning_rate": 1.5308739486924064e-05,
+      "loss": 0.744,
+      "step": 6573
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 36.35692856069979,
+      "learning_rate": 1.530731027128802e-05,
+      "loss": 0.7548,
+      "step": 6574
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.35917961026885,
+      "learning_rate": 1.5305880904715177e-05,
+      "loss": 0.7579,
+      "step": 6575
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 35.44786719065504,
+      "learning_rate": 1.530445138724619e-05,
+      "loss": 0.7457,
+      "step": 6576
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 35.78288159377702,
+      "learning_rate": 1.53030217189217e-05,
+      "loss": 0.7597,
+      "step": 6577
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 33.495027806529166,
+      "learning_rate": 1.5301591899782376e-05,
+      "loss": 0.7052,
+      "step": 6578
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 37.04745593635137,
+      "learning_rate": 1.5300161929868886e-05,
+      "loss": 0.7543,
+      "step": 6579
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.43239750593648,
+      "learning_rate": 1.5298731809221886e-05,
+      "loss": 0.7431,
+      "step": 6580
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.7958464519732,
+      "learning_rate": 1.5297301537882055e-05,
+      "loss": 0.7884,
+      "step": 6581
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 46.14249643211601,
+      "learning_rate": 1.5295871115890066e-05,
+      "loss": 0.7528,
+      "step": 6582
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 75.33874181164184,
+      "learning_rate": 1.52944405432866e-05,
+      "loss": 1.0509,
+      "step": 6583
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 50.77934663932202,
+      "learning_rate": 1.5293009820112345e-05,
+      "loss": 0.6352,
+      "step": 6584
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.80694797766112,
+      "learning_rate": 1.5291578946407985e-05,
+      "loss": 0.7228,
+      "step": 6585
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.72427789914974,
+      "learning_rate": 1.5290147922214212e-05,
+      "loss": 0.7893,
+      "step": 6586
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.709947769853294,
+      "learning_rate": 1.5288716747571735e-05,
+      "loss": 0.6774,
+      "step": 6587
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.80570507338805,
+      "learning_rate": 1.5287285422521238e-05,
+      "loss": 0.802,
+      "step": 6588
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.11013208914867,
+      "learning_rate": 1.5285853947103446e-05,
+      "loss": 0.7533,
+      "step": 6589
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.85474293639716,
+      "learning_rate": 1.5284422321359054e-05,
+      "loss": 0.838,
+      "step": 6590
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 43.78030891686756,
+      "learning_rate": 1.5282990545328782e-05,
+      "loss": 0.8005,
+      "step": 6591
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 35.491436388824646,
+      "learning_rate": 1.5281558619053353e-05,
+      "loss": 0.6391,
+      "step": 6592
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 32.926416816148375,
+      "learning_rate": 1.5280126542573484e-05,
+      "loss": 0.7321,
+      "step": 6593
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 30.27098445675209,
+      "learning_rate": 1.5278694315929906e-05,
+      "loss": 0.671,
+      "step": 6594
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 42.42619447021321,
+      "learning_rate": 1.527726193916335e-05,
+      "loss": 0.7264,
+      "step": 6595
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 30.647588768184878,
+      "learning_rate": 1.5275829412314547e-05,
+      "loss": 0.6999,
+      "step": 6596
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 36.44704150057439,
+      "learning_rate": 1.5274396735424244e-05,
+      "loss": 0.7102,
+      "step": 6597
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.924531963109516,
+      "learning_rate": 1.5272963908533184e-05,
+      "loss": 0.7572,
+      "step": 6598
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 33.998188321147026,
+      "learning_rate": 1.5271530931682116e-05,
+      "loss": 0.64,
+      "step": 6599
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.17720198831111,
+      "learning_rate": 1.5270097804911794e-05,
+      "loss": 0.6956,
+      "step": 6600
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 32.11718752969358,
+      "learning_rate": 1.526866452826297e-05,
+      "loss": 0.7338,
+      "step": 6601
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.653906599998855,
+      "learning_rate": 1.526723110177641e-05,
+      "loss": 0.6859,
+      "step": 6602
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 28.5527014296073,
+      "learning_rate": 1.5265797525492878e-05,
+      "loss": 0.6545,
+      "step": 6603
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 37.0360740642044,
+      "learning_rate": 1.526436379945315e-05,
+      "loss": 0.7673,
+      "step": 6604
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 31.981042417173494,
+      "learning_rate": 1.5262929923697986e-05,
+      "loss": 0.653,
+      "step": 6605
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.32882087220988,
+      "learning_rate": 1.526149589826818e-05,
+      "loss": 0.6778,
+      "step": 6606
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.74151676024518,
+      "learning_rate": 1.5260061723204506e-05,
+      "loss": 0.8138,
+      "step": 6607
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.01896295474441,
+      "learning_rate": 1.5258627398547754e-05,
+      "loss": 0.7185,
+      "step": 6608
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 37.31716651745008,
+      "learning_rate": 1.5257192924338715e-05,
+      "loss": 0.6693,
+      "step": 6609
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.730893562207086,
+      "learning_rate": 1.5255758300618184e-05,
+      "loss": 0.8248,
+      "step": 6610
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.75675344067897,
+      "learning_rate": 1.5254323527426964e-05,
+      "loss": 0.8154,
+      "step": 6611
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.251852906918515,
+      "learning_rate": 1.5252888604805853e-05,
+      "loss": 0.77,
+      "step": 6612
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.38936865381853,
+      "learning_rate": 1.5251453532795663e-05,
+      "loss": 0.7543,
+      "step": 6613
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.53557538866698,
+      "learning_rate": 1.5250018311437212e-05,
+      "loss": 0.7297,
+      "step": 6614
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 39.34992965754968,
+      "learning_rate": 1.5248582940771306e-05,
+      "loss": 0.798,
+      "step": 6615
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 36.170229346655866,
+      "learning_rate": 1.5247147420838774e-05,
+      "loss": 0.6618,
+      "step": 6616
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 44.80788129682913,
+      "learning_rate": 1.5245711751680438e-05,
+      "loss": 0.6682,
+      "step": 6617
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 39.47246950493249,
+      "learning_rate": 1.524427593333713e-05,
+      "loss": 0.7004,
+      "step": 6618
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 40.952644947881375,
+      "learning_rate": 1.524283996584968e-05,
+      "loss": 0.7694,
+      "step": 6619
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 50.96002903838655,
+      "learning_rate": 1.524140384925893e-05,
+      "loss": 0.7096,
+      "step": 6620
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 32.01766613308412,
+      "learning_rate": 1.5239967583605719e-05,
+      "loss": 0.7453,
+      "step": 6621
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 38.69766367148885,
+      "learning_rate": 1.5238531168930894e-05,
+      "loss": 0.7624,
+      "step": 6622
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 62.576662348918795,
+      "learning_rate": 1.5237094605275311e-05,
+      "loss": 0.6608,
+      "step": 6623
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 33.81625057015855,
+      "learning_rate": 1.5235657892679818e-05,
+      "loss": 0.7392,
+      "step": 6624
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 34.14442480315298,
+      "learning_rate": 1.523422103118528e-05,
+      "loss": 0.8005,
+      "step": 6625
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 31.2493939302801,
+      "learning_rate": 1.5232784020832556e-05,
+      "loss": 0.6875,
+      "step": 6626
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.26535712600475,
+      "learning_rate": 1.5231346861662518e-05,
+      "loss": 0.6875,
+      "step": 6627
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 33.12695227574011,
+      "learning_rate": 1.5229909553716032e-05,
+      "loss": 0.7407,
+      "step": 6628
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 32.85185185689419,
+      "learning_rate": 1.5228472097033979e-05,
+      "loss": 0.6774,
+      "step": 6629
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 32.97563625933651,
+      "learning_rate": 1.5227034491657235e-05,
+      "loss": 0.6482,
+      "step": 6630
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 87.32033858068917,
+      "learning_rate": 1.5225596737626695e-05,
+      "loss": 0.658,
+      "step": 6631
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.91226527122947,
+      "learning_rate": 1.5224158834983234e-05,
+      "loss": 0.7085,
+      "step": 6632
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 34.482154764347065,
+      "learning_rate": 1.5222720783767755e-05,
+      "loss": 0.7395,
+      "step": 6633
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.698520314098005,
+      "learning_rate": 1.522128258402115e-05,
+      "loss": 0.7833,
+      "step": 6634
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 65.57822900836105,
+      "learning_rate": 1.5219844235784326e-05,
+      "loss": 0.8833,
+      "step": 6635
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 41.905297299108675,
+      "learning_rate": 1.5218405739098183e-05,
+      "loss": 0.8482,
+      "step": 6636
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.55540146673638,
+      "learning_rate": 1.5216967094003633e-05,
+      "loss": 0.7229,
+      "step": 6637
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.994400145108585,
+      "learning_rate": 1.5215528300541593e-05,
+      "loss": 0.7938,
+      "step": 6638
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 43.61140725550899,
+      "learning_rate": 1.5214089358752979e-05,
+      "loss": 0.6688,
+      "step": 6639
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 45.97983229166435,
+      "learning_rate": 1.5212650268678711e-05,
+      "loss": 0.7423,
+      "step": 6640
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 42.78144595628643,
+      "learning_rate": 1.5211211030359719e-05,
+      "loss": 0.811,
+      "step": 6641
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.403361504418065,
+      "learning_rate": 1.5209771643836937e-05,
+      "loss": 0.7252,
+      "step": 6642
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 41.63949055513042,
+      "learning_rate": 1.5208332109151295e-05,
+      "loss": 0.8382,
+      "step": 6643
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.885361277123984,
+      "learning_rate": 1.5206892426343736e-05,
+      "loss": 0.6998,
+      "step": 6644
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 34.74435846242326,
+      "learning_rate": 1.5205452595455198e-05,
+      "loss": 0.7181,
+      "step": 6645
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 33.74180611820819,
+      "learning_rate": 1.5204012616526637e-05,
+      "loss": 0.7166,
+      "step": 6646
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 34.28402916728017,
+      "learning_rate": 1.5202572489598998e-05,
+      "loss": 0.7324,
+      "step": 6647
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 25.88239594506322,
+      "learning_rate": 1.5201132214713249e-05,
+      "loss": 0.6749,
+      "step": 6648
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.30328793636488,
+      "learning_rate": 1.5199691791910333e-05,
+      "loss": 0.733,
+      "step": 6649
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.62101093544564,
+      "learning_rate": 1.5198251221231232e-05,
+      "loss": 0.7471,
+      "step": 6650
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 52.472647044518176,
+      "learning_rate": 1.51968105027169e-05,
+      "loss": 0.6993,
+      "step": 6651
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.27936236774046,
+      "learning_rate": 1.5195369636408322e-05,
+      "loss": 0.7882,
+      "step": 6652
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.03461122343587,
+      "learning_rate": 1.519392862234647e-05,
+      "loss": 0.8171,
+      "step": 6653
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 39.69650838544951,
+      "learning_rate": 1.5192487460572327e-05,
+      "loss": 0.6955,
+      "step": 6654
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 33.4663659287583,
+      "learning_rate": 1.5191046151126876e-05,
+      "loss": 0.7369,
+      "step": 6655
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.994475590857576,
+      "learning_rate": 1.5189604694051111e-05,
+      "loss": 0.7292,
+      "step": 6656
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.78751779465759,
+      "learning_rate": 1.5188163089386022e-05,
+      "loss": 0.7268,
+      "step": 6657
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 39.390626731064486,
+      "learning_rate": 1.5186721337172614e-05,
+      "loss": 0.7753,
+      "step": 6658
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 33.22105149974513,
+      "learning_rate": 1.5185279437451881e-05,
+      "loss": 0.6425,
+      "step": 6659
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 30.980276010750917,
+      "learning_rate": 1.5183837390264833e-05,
+      "loss": 0.7358,
+      "step": 6660
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 43.770122813117204,
+      "learning_rate": 1.5182395195652487e-05,
+      "loss": 0.843,
+      "step": 6661
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 45.295864926906376,
+      "learning_rate": 1.518095285365585e-05,
+      "loss": 0.9047,
+      "step": 6662
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 42.251637678177175,
+      "learning_rate": 1.5179510364315945e-05,
+      "loss": 0.7911,
+      "step": 6663
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.879904929850426,
+      "learning_rate": 1.5178067727673797e-05,
+      "loss": 0.6994,
+      "step": 6664
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 41.28374893514192,
+      "learning_rate": 1.517662494377043e-05,
+      "loss": 0.7443,
+      "step": 6665
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.6371010277499,
+      "learning_rate": 1.5175182012646877e-05,
+      "loss": 0.7868,
+      "step": 6666
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 30.667584659910162,
+      "learning_rate": 1.5173738934344178e-05,
+      "loss": 0.7068,
+      "step": 6667
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 45.20129203046916,
+      "learning_rate": 1.5172295708903368e-05,
+      "loss": 0.7421,
+      "step": 6668
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 35.81102587210756,
+      "learning_rate": 1.5170852336365494e-05,
+      "loss": 0.7604,
+      "step": 6669
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.16071403463051,
+      "learning_rate": 1.5169408816771605e-05,
+      "loss": 0.6739,
+      "step": 6670
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.90018638024386,
+      "learning_rate": 1.5167965150162753e-05,
+      "loss": 0.6994,
+      "step": 6671
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 36.47261686180973,
+      "learning_rate": 1.5166521336579996e-05,
+      "loss": 0.6776,
+      "step": 6672
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 35.31592495268274,
+      "learning_rate": 1.5165077376064394e-05,
+      "loss": 0.7398,
+      "step": 6673
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.888147514596966,
+      "learning_rate": 1.516363326865701e-05,
+      "loss": 0.8223,
+      "step": 6674
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 46.7535806396426,
+      "learning_rate": 1.5162189014398923e-05,
+      "loss": 0.6748,
+      "step": 6675
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.142647479128414,
+      "learning_rate": 1.5160744613331196e-05,
+      "loss": 0.6751,
+      "step": 6676
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 43.14764772680412,
+      "learning_rate": 1.5159300065494912e-05,
+      "loss": 0.7287,
+      "step": 6677
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.40744149231345,
+      "learning_rate": 1.515785537093115e-05,
+      "loss": 0.7348,
+      "step": 6678
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 34.047774612425144,
+      "learning_rate": 1.5156410529681e-05,
+      "loss": 0.6995,
+      "step": 6679
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 29.844131283546346,
+      "learning_rate": 1.5154965541785554e-05,
+      "loss": 0.757,
+      "step": 6680
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 32.92334003203233,
+      "learning_rate": 1.51535204072859e-05,
+      "loss": 0.6755,
+      "step": 6681
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 37.20270232081073,
+      "learning_rate": 1.515207512622314e-05,
+      "loss": 0.7455,
+      "step": 6682
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 35.348240544319694,
+      "learning_rate": 1.515062969863838e-05,
+      "loss": 0.8003,
+      "step": 6683
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 31.252706708036055,
+      "learning_rate": 1.514918412457272e-05,
+      "loss": 0.6587,
+      "step": 6684
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 40.36669871966609,
+      "learning_rate": 1.5147738404067279e-05,
+      "loss": 0.7829,
+      "step": 6685
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 35.09531129402507,
+      "learning_rate": 1.5146292537163168e-05,
+      "loss": 0.7389,
+      "step": 6686
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 34.61013224098404,
+      "learning_rate": 1.5144846523901508e-05,
+      "loss": 0.8322,
+      "step": 6687
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 44.346403261667106,
+      "learning_rate": 1.5143400364323424e-05,
+      "loss": 0.782,
+      "step": 6688
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 29.826837610805196,
+      "learning_rate": 1.5141954058470041e-05,
+      "loss": 0.7343,
+      "step": 6689
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 38.00937365063275,
+      "learning_rate": 1.5140507606382496e-05,
+      "loss": 0.7467,
+      "step": 6690
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.10079630791744,
+      "learning_rate": 1.5139061008101915e-05,
+      "loss": 0.6803,
+      "step": 6691
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.54847138413754,
+      "learning_rate": 1.5137614263669451e-05,
+      "loss": 0.726,
+      "step": 6692
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 39.63959137046747,
+      "learning_rate": 1.5136167373126239e-05,
+      "loss": 0.7306,
+      "step": 6693
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.12139705281089,
+      "learning_rate": 1.5134720336513433e-05,
+      "loss": 0.7963,
+      "step": 6694
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 37.75797185660954,
+      "learning_rate": 1.5133273153872186e-05,
+      "loss": 0.7817,
+      "step": 6695
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 37.3301587291289,
+      "learning_rate": 1.5131825825243651e-05,
+      "loss": 0.672,
+      "step": 6696
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 32.15964180618175,
+      "learning_rate": 1.5130378350668996e-05,
+      "loss": 0.6931,
+      "step": 6697
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 31.896738134632248,
+      "learning_rate": 1.5128930730189377e-05,
+      "loss": 0.6907,
+      "step": 6698
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 39.29314032643356,
+      "learning_rate": 1.5127482963845968e-05,
+      "loss": 0.7081,
+      "step": 6699
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 39.361933332394045,
+      "learning_rate": 1.5126035051679946e-05,
+      "loss": 0.7432,
+      "step": 6700
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.60139388429282,
+      "learning_rate": 1.5124586993732483e-05,
+      "loss": 0.7736,
+      "step": 6701
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.697199291935654,
+      "learning_rate": 1.5123138790044767e-05,
+      "loss": 0.6929,
+      "step": 6702
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 35.72010116157539,
+      "learning_rate": 1.5121690440657977e-05,
+      "loss": 0.6731,
+      "step": 6703
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.06064055730866,
+      "learning_rate": 1.512024194561331e-05,
+      "loss": 0.726,
+      "step": 6704
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 35.10592641221199,
+      "learning_rate": 1.5118793304951955e-05,
+      "loss": 0.7611,
+      "step": 6705
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.00063639316732,
+      "learning_rate": 1.5117344518715115e-05,
+      "loss": 0.7664,
+      "step": 6706
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 47.793408091357115,
+      "learning_rate": 1.511589558694399e-05,
+      "loss": 0.824,
+      "step": 6707
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.33643602306591,
+      "learning_rate": 1.5114446509679783e-05,
+      "loss": 0.754,
+      "step": 6708
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 41.359763512193545,
+      "learning_rate": 1.5112997286963715e-05,
+      "loss": 0.8384,
+      "step": 6709
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 35.85847424807392,
+      "learning_rate": 1.5111547918836994e-05,
+      "loss": 0.6925,
+      "step": 6710
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 30.5984974062137,
+      "learning_rate": 1.5110098405340838e-05,
+      "loss": 0.6387,
+      "step": 6711
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 29.679356588480484,
+      "learning_rate": 1.5108648746516475e-05,
+      "loss": 0.6752,
+      "step": 6712
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.94130678134244,
+      "learning_rate": 1.5107198942405128e-05,
+      "loss": 0.7366,
+      "step": 6713
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 38.99480697402596,
+      "learning_rate": 1.5105748993048032e-05,
+      "loss": 0.7095,
+      "step": 6714
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.970026244112155,
+      "learning_rate": 1.5104298898486423e-05,
+      "loss": 0.7312,
+      "step": 6715
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.65089004396635,
+      "learning_rate": 1.5102848658761535e-05,
+      "loss": 0.7646,
+      "step": 6716
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 44.15241287416381,
+      "learning_rate": 1.5101398273914623e-05,
+      "loss": 0.7408,
+      "step": 6717
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 28.40588391679775,
+      "learning_rate": 1.509994774398692e-05,
+      "loss": 0.6821,
+      "step": 6718
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 40.23176659582315,
+      "learning_rate": 1.5098497069019693e-05,
+      "loss": 0.7991,
+      "step": 6719
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.40265880001684,
+      "learning_rate": 1.5097046249054187e-05,
+      "loss": 0.809,
+      "step": 6720
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.21614890513324,
+      "learning_rate": 1.5095595284131674e-05,
+      "loss": 0.6883,
+      "step": 6721
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.59549974719612,
+      "learning_rate": 1.5094144174293406e-05,
+      "loss": 0.6344,
+      "step": 6722
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.348625847067936,
+      "learning_rate": 1.509269291958066e-05,
+      "loss": 0.7335,
+      "step": 6723
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 30.504083559388267,
+      "learning_rate": 1.5091241520034705e-05,
+      "loss": 0.6068,
+      "step": 6724
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 43.84100587703736,
+      "learning_rate": 1.508978997569682e-05,
+      "loss": 0.7301,
+      "step": 6725
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.79977612848182,
+      "learning_rate": 1.5088338286608287e-05,
+      "loss": 0.7182,
+      "step": 6726
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.61421225198847,
+      "learning_rate": 1.508688645281039e-05,
+      "loss": 0.6919,
+      "step": 6727
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 58.438089688023666,
+      "learning_rate": 1.5085434474344416e-05,
+      "loss": 0.864,
+      "step": 6728
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 39.23959634569865,
+      "learning_rate": 1.508398235125166e-05,
+      "loss": 0.8164,
+      "step": 6729
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.877177833566236,
+      "learning_rate": 1.5082530083573421e-05,
+      "loss": 0.7382,
+      "step": 6730
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 42.606472922334135,
+      "learning_rate": 1.5081077671350998e-05,
+      "loss": 0.7048,
+      "step": 6731
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 44.50185637941741,
+      "learning_rate": 1.5079625114625701e-05,
+      "loss": 0.6744,
+      "step": 6732
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 37.398989591872414,
+      "learning_rate": 1.5078172413438836e-05,
+      "loss": 0.6737,
+      "step": 6733
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 32.9622345149517,
+      "learning_rate": 1.507671956783172e-05,
+      "loss": 0.7235,
+      "step": 6734
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 46.496219619776745,
+      "learning_rate": 1.5075266577845668e-05,
+      "loss": 0.7813,
+      "step": 6735
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 40.681050839847224,
+      "learning_rate": 1.5073813443522005e-05,
+      "loss": 0.7693,
+      "step": 6736
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 32.51332952342409,
+      "learning_rate": 1.5072360164902052e-05,
+      "loss": 0.5999,
+      "step": 6737
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 39.5795531318308,
+      "learning_rate": 1.5070906742027145e-05,
+      "loss": 0.7813,
+      "step": 6738
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 41.49927829206808,
+      "learning_rate": 1.5069453174938618e-05,
+      "loss": 0.7275,
+      "step": 6739
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 37.4199975580075,
+      "learning_rate": 1.5067999463677807e-05,
+      "loss": 0.7146,
+      "step": 6740
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 35.555015654653666,
+      "learning_rate": 1.5066545608286056e-05,
+      "loss": 0.7182,
+      "step": 6741
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.7885602510202,
+      "learning_rate": 1.5065091608804712e-05,
+      "loss": 0.6071,
+      "step": 6742
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.954580652708735,
+      "learning_rate": 1.5063637465275127e-05,
+      "loss": 0.6774,
+      "step": 6743
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 41.243231241149225,
+      "learning_rate": 1.5062183177738652e-05,
+      "loss": 0.8135,
+      "step": 6744
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 38.406716873894524,
+      "learning_rate": 1.5060728746236648e-05,
+      "loss": 0.79,
+      "step": 6745
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 36.08619595369231,
+      "learning_rate": 1.5059274170810483e-05,
+      "loss": 0.6721,
+      "step": 6746
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 45.47622980423052,
+      "learning_rate": 1.505781945150152e-05,
+      "loss": 0.7941,
+      "step": 6747
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 32.68138191486874,
+      "learning_rate": 1.5056364588351127e-05,
+      "loss": 0.7185,
+      "step": 6748
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 35.27217973428929,
+      "learning_rate": 1.5054909581400683e-05,
+      "loss": 0.7469,
+      "step": 6749
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.717223160249894,
+      "learning_rate": 1.505345443069157e-05,
+      "loss": 0.7161,
+      "step": 6750
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 45.05106951886689,
+      "learning_rate": 1.505199913626517e-05,
+      "loss": 0.7415,
+      "step": 6751
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 49.90426937561006,
+      "learning_rate": 1.5050543698162865e-05,
+      "loss": 0.8875,
+      "step": 6752
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 81.89293366212878,
+      "learning_rate": 1.5049088116426056e-05,
+      "loss": 0.6935,
+      "step": 6753
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 34.58522729206938,
+      "learning_rate": 1.5047632391096132e-05,
+      "loss": 0.7089,
+      "step": 6754
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 45.669134195459826,
+      "learning_rate": 1.5046176522214497e-05,
+      "loss": 0.7475,
+      "step": 6755
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 44.34353129015075,
+      "learning_rate": 1.5044720509822553e-05,
+      "loss": 0.8171,
+      "step": 6756
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 35.9236314235469,
+      "learning_rate": 1.5043264353961711e-05,
+      "loss": 0.7318,
+      "step": 6757
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 41.66169022151703,
+      "learning_rate": 1.5041808054673379e-05,
+      "loss": 0.7873,
+      "step": 6758
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 46.17729903375198,
+      "learning_rate": 1.5040351611998975e-05,
+      "loss": 0.7279,
+      "step": 6759
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 35.87630579894676,
+      "learning_rate": 1.5038895025979919e-05,
+      "loss": 0.6738,
+      "step": 6760
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 40.272002952918854,
+      "learning_rate": 1.503743829665764e-05,
+      "loss": 0.7333,
+      "step": 6761
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 42.277500716588314,
+      "learning_rate": 1.503598142407356e-05,
+      "loss": 0.7938,
+      "step": 6762
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.17794001486774,
+      "learning_rate": 1.5034524408269115e-05,
+      "loss": 0.666,
+      "step": 6763
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 39.015066776093825,
+      "learning_rate": 1.503306724928574e-05,
+      "loss": 0.8171,
+      "step": 6764
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 33.90232607729719,
+      "learning_rate": 1.5031609947164876e-05,
+      "loss": 0.703,
+      "step": 6765
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 41.23347066832245,
+      "learning_rate": 1.503015250194797e-05,
+      "loss": 0.7815,
+      "step": 6766
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 40.77767309528523,
+      "learning_rate": 1.5028694913676469e-05,
+      "loss": 0.7617,
+      "step": 6767
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 30.8435594732551,
+      "learning_rate": 1.5027237182391825e-05,
+      "loss": 0.6894,
+      "step": 6768
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 64.09907164444449,
+      "learning_rate": 1.5025779308135499e-05,
+      "loss": 0.7634,
+      "step": 6769
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 29.377968341923822,
+      "learning_rate": 1.5024321290948948e-05,
+      "loss": 0.6287,
+      "step": 6770
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 39.6553203242323,
+      "learning_rate": 1.502286313087364e-05,
+      "loss": 0.7492,
+      "step": 6771
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 43.407680015544415,
+      "learning_rate": 1.5021404827951039e-05,
+      "loss": 0.6804,
+      "step": 6772
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 44.28861886241521,
+      "learning_rate": 1.5019946382222626e-05,
+      "loss": 0.6904,
+      "step": 6773
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 31.69698733069452,
+      "learning_rate": 1.5018487793729875e-05,
+      "loss": 0.6515,
+      "step": 6774
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 40.62281746503568,
+      "learning_rate": 1.5017029062514263e-05,
+      "loss": 0.6589,
+      "step": 6775
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 40.16841158631588,
+      "learning_rate": 1.5015570188617283e-05,
+      "loss": 0.7802,
+      "step": 6776
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.559654421435035,
+      "learning_rate": 1.5014111172080416e-05,
+      "loss": 0.6931,
+      "step": 6777
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 46.274355872651896,
+      "learning_rate": 1.5012652012945169e-05,
+      "loss": 0.7631,
+      "step": 6778
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.85387623817869,
+      "learning_rate": 1.5011192711253021e-05,
+      "loss": 0.6721,
+      "step": 6779
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 41.45822971976414,
+      "learning_rate": 1.5009733267045492e-05,
+      "loss": 0.7521,
+      "step": 6780
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 47.16587154046225,
+      "learning_rate": 1.5008273680364074e-05,
+      "loss": 0.7776,
+      "step": 6781
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 31.24580746183449,
+      "learning_rate": 1.5006813951250286e-05,
+      "loss": 0.777,
+      "step": 6782
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 33.047635643684494,
+      "learning_rate": 1.5005354079745636e-05,
+      "loss": 0.629,
+      "step": 6783
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.99338040226048,
+      "learning_rate": 1.5003894065891644e-05,
+      "loss": 0.6778,
+      "step": 6784
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 29.99989143590357,
+      "learning_rate": 1.5002433909729832e-05,
+      "loss": 0.6651,
+      "step": 6785
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 32.76314669556266,
+      "learning_rate": 1.5000973611301729e-05,
+      "loss": 0.7118,
+      "step": 6786
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.35593745264672,
+      "learning_rate": 1.4999513170648856e-05,
+      "loss": 0.7752,
+      "step": 6787
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 42.251366068407854,
+      "learning_rate": 1.4998052587812759e-05,
+      "loss": 0.8317,
+      "step": 6788
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.40466176497801,
+      "learning_rate": 1.4996591862834965e-05,
+      "loss": 0.767,
+      "step": 6789
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.416616867320016,
+      "learning_rate": 1.4995130995757029e-05,
+      "loss": 0.7257,
+      "step": 6790
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 35.44649773193235,
+      "learning_rate": 1.499366998662048e-05,
+      "loss": 0.8359,
+      "step": 6791
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 32.34720233238984,
+      "learning_rate": 1.4992208835466884e-05,
+      "loss": 0.6921,
+      "step": 6792
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 37.4238614504799,
+      "learning_rate": 1.4990747542337787e-05,
+      "loss": 0.6394,
+      "step": 6793
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 52.01686628699182,
+      "learning_rate": 1.498928610727475e-05,
+      "loss": 0.7743,
+      "step": 6794
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 47.260644299090174,
+      "learning_rate": 1.4987824530319335e-05,
+      "loss": 0.7633,
+      "step": 6795
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.205922320405506,
+      "learning_rate": 1.4986362811513107e-05,
+      "loss": 0.7607,
+      "step": 6796
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 37.9109102609481,
+      "learning_rate": 1.4984900950897638e-05,
+      "loss": 0.7864,
+      "step": 6797
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.83471586907093,
+      "learning_rate": 1.4983438948514501e-05,
+      "loss": 0.7565,
+      "step": 6798
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 37.71078671997602,
+      "learning_rate": 1.4981976804405278e-05,
+      "loss": 0.7921,
+      "step": 6799
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 34.37085300052996,
+      "learning_rate": 1.4980514518611546e-05,
+      "loss": 0.8009,
+      "step": 6800
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 32.6835050749234,
+      "learning_rate": 1.49790520911749e-05,
+      "loss": 0.7049,
+      "step": 6801
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 34.3809172375138,
+      "learning_rate": 1.497758952213692e-05,
+      "loss": 0.6774,
+      "step": 6802
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.64462705033676,
+      "learning_rate": 1.4976126811539207e-05,
+      "loss": 0.7938,
+      "step": 6803
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 30.448379957233104,
+      "learning_rate": 1.4974663959423358e-05,
+      "loss": 0.6519,
+      "step": 6804
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.64487875381338,
+      "learning_rate": 1.497320096583098e-05,
+      "loss": 0.7244,
+      "step": 6805
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 38.69204580441811,
+      "learning_rate": 1.497173783080367e-05,
+      "loss": 0.8097,
+      "step": 6806
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 29.21435476690333,
+      "learning_rate": 1.497027455438305e-05,
+      "loss": 0.6806,
+      "step": 6807
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 35.38672408948092,
+      "learning_rate": 1.4968811136610724e-05,
+      "loss": 0.6808,
+      "step": 6808
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 29.850358768524362,
+      "learning_rate": 1.4967347577528319e-05,
+      "loss": 0.6647,
+      "step": 6809
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 32.69182364057988,
+      "learning_rate": 1.4965883877177454e-05,
+      "loss": 0.7229,
+      "step": 6810
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 31.38671738269093,
+      "learning_rate": 1.4964420035599756e-05,
+      "loss": 0.6658,
+      "step": 6811
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 44.871950779127,
+      "learning_rate": 1.4962956052836858e-05,
+      "loss": 0.8241,
+      "step": 6812
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 34.39758177729729,
+      "learning_rate": 1.4961491928930394e-05,
+      "loss": 0.6444,
+      "step": 6813
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 34.57515121476543,
+      "learning_rate": 1.4960027663921996e-05,
+      "loss": 0.7536,
+      "step": 6814
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 34.826477593693205,
+      "learning_rate": 1.495856325785332e-05,
+      "loss": 0.6963,
+      "step": 6815
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 39.694896975863,
+      "learning_rate": 1.4957098710766e-05,
+      "loss": 0.8122,
+      "step": 6816
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 44.36614519074801,
+      "learning_rate": 1.4955634022701695e-05,
+      "loss": 0.7045,
+      "step": 6817
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 54.313603577634304,
+      "learning_rate": 1.4954169193702058e-05,
+      "loss": 0.7639,
+      "step": 6818
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 31.53379171087366,
+      "learning_rate": 1.4952704223808745e-05,
+      "loss": 0.7115,
+      "step": 6819
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 30.998085351237116,
+      "learning_rate": 1.4951239113063425e-05,
+      "loss": 0.6119,
+      "step": 6820
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 38.98882208402129,
+      "learning_rate": 1.4949773861507754e-05,
+      "loss": 0.8186,
+      "step": 6821
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.83118359862254,
+      "learning_rate": 1.4948308469183419e-05,
+      "loss": 0.8429,
+      "step": 6822
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 40.691394285537356,
+      "learning_rate": 1.494684293613208e-05,
+      "loss": 0.7841,
+      "step": 6823
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.23209645946992,
+      "learning_rate": 1.4945377262395425e-05,
+      "loss": 0.7147,
+      "step": 6824
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 42.34123779389496,
+      "learning_rate": 1.4943911448015129e-05,
+      "loss": 0.751,
+      "step": 6825
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.895798870670085,
+      "learning_rate": 1.4942445493032887e-05,
+      "loss": 0.6822,
+      "step": 6826
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 31.709224789069015,
+      "learning_rate": 1.4940979397490386e-05,
+      "loss": 0.6431,
+      "step": 6827
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 32.407538326776056,
+      "learning_rate": 1.493951316142932e-05,
+      "loss": 0.7319,
+      "step": 6828
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.41556245946919,
+      "learning_rate": 1.493804678489139e-05,
+      "loss": 0.8479,
+      "step": 6829
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.2560887832878,
+      "learning_rate": 1.4936580267918299e-05,
+      "loss": 0.7969,
+      "step": 6830
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 32.51663485805247,
+      "learning_rate": 1.4935113610551751e-05,
+      "loss": 0.7395,
+      "step": 6831
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 34.83572010520829,
+      "learning_rate": 1.4933646812833461e-05,
+      "loss": 0.7826,
+      "step": 6832
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 29.417488865173105,
+      "learning_rate": 1.493217987480514e-05,
+      "loss": 0.6498,
+      "step": 6833
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.17986826688208,
+      "learning_rate": 1.4930712796508511e-05,
+      "loss": 0.7751,
+      "step": 6834
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.06513317100816,
+      "learning_rate": 1.4929245577985292e-05,
+      "loss": 0.8438,
+      "step": 6835
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 40.0000320434442,
+      "learning_rate": 1.4927778219277214e-05,
+      "loss": 0.8402,
+      "step": 6836
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 39.477043026631065,
+      "learning_rate": 1.4926310720426005e-05,
+      "loss": 0.8616,
+      "step": 6837
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 40.457805535160404,
+      "learning_rate": 1.4924843081473401e-05,
+      "loss": 0.7267,
+      "step": 6838
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 32.03748857733872,
+      "learning_rate": 1.492337530246114e-05,
+      "loss": 0.7059,
+      "step": 6839
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 41.0213512586862,
+      "learning_rate": 1.4921907383430964e-05,
+      "loss": 0.6695,
+      "step": 6840
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 38.87450862770233,
+      "learning_rate": 1.4920439324424627e-05,
+      "loss": 0.6583,
+      "step": 6841
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.91862464445656,
+      "learning_rate": 1.491897112548387e-05,
+      "loss": 0.6773,
+      "step": 6842
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 43.913470212847486,
+      "learning_rate": 1.4917502786650451e-05,
+      "loss": 0.6807,
+      "step": 6843
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 42.0937256901479,
+      "learning_rate": 1.4916034307966132e-05,
+      "loss": 0.77,
+      "step": 6844
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.37392379754713,
+      "learning_rate": 1.4914565689472671e-05,
+      "loss": 0.7356,
+      "step": 6845
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.08953604025653,
+      "learning_rate": 1.4913096931211839e-05,
+      "loss": 0.6786,
+      "step": 6846
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 29.852976969413113,
+      "learning_rate": 1.4911628033225402e-05,
+      "loss": 0.6671,
+      "step": 6847
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.8847761869731,
+      "learning_rate": 1.4910158995555138e-05,
+      "loss": 0.6723,
+      "step": 6848
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 50.75441123319525,
+      "learning_rate": 1.4908689818242826e-05,
+      "loss": 0.748,
+      "step": 6849
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.98798276965403,
+      "learning_rate": 1.4907220501330245e-05,
+      "loss": 0.7003,
+      "step": 6850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.8734395804177,
+      "learning_rate": 1.4905751044859185e-05,
+      "loss": 0.7158,
+      "step": 6851
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 30.748403228611952,
+      "learning_rate": 1.4904281448871431e-05,
+      "loss": 0.7598,
+      "step": 6852
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 29.926794650664633,
+      "learning_rate": 1.4902811713408784e-05,
+      "loss": 0.6042,
+      "step": 6853
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 51.20032740363479,
+      "learning_rate": 1.4901341838513044e-05,
+      "loss": 0.7039,
+      "step": 6854
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 33.270434525085584,
+      "learning_rate": 1.4899871824226004e-05,
+      "loss": 0.7481,
+      "step": 6855
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.980210575370734,
+      "learning_rate": 1.4898401670589478e-05,
+      "loss": 0.7676,
+      "step": 6856
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.67833763126691,
+      "learning_rate": 1.4896931377645273e-05,
+      "loss": 0.7286,
+      "step": 6857
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 40.37707496484075,
+      "learning_rate": 1.4895460945435205e-05,
+      "loss": 0.7036,
+      "step": 6858
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 27.538023904465202,
+      "learning_rate": 1.489399037400109e-05,
+      "loss": 0.6456,
+      "step": 6859
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.502919403665004,
+      "learning_rate": 1.4892519663384755e-05,
+      "loss": 0.7825,
+      "step": 6860
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 30.6000058769395,
+      "learning_rate": 1.489104881362802e-05,
+      "loss": 0.7038,
+      "step": 6861
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 38.17258478005233,
+      "learning_rate": 1.4889577824772719e-05,
+      "loss": 0.7864,
+      "step": 6862
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 26.702442555554605,
+      "learning_rate": 1.4888106696860684e-05,
+      "loss": 0.6208,
+      "step": 6863
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 37.345889738063626,
+      "learning_rate": 1.4886635429933756e-05,
+      "loss": 0.7974,
+      "step": 6864
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 33.799193039549635,
+      "learning_rate": 1.4885164024033771e-05,
+      "loss": 0.7972,
+      "step": 6865
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 40.42921811628379,
+      "learning_rate": 1.4883692479202582e-05,
+      "loss": 0.8163,
+      "step": 6866
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 41.68916327777184,
+      "learning_rate": 1.4882220795482037e-05,
+      "loss": 0.8218,
+      "step": 6867
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 34.547531018181445,
+      "learning_rate": 1.4880748972913987e-05,
+      "loss": 0.6825,
+      "step": 6868
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.092138394371055,
+      "learning_rate": 1.4879277011540289e-05,
+      "loss": 0.6888,
+      "step": 6869
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 31.359807317972646,
+      "learning_rate": 1.4877804911402809e-05,
+      "loss": 0.6797,
+      "step": 6870
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 39.204664334681105,
+      "learning_rate": 1.4876332672543412e-05,
+      "loss": 0.7694,
+      "step": 6871
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 42.28121437922723,
+      "learning_rate": 1.4874860295003965e-05,
+      "loss": 0.7091,
+      "step": 6872
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 40.04032313897666,
+      "learning_rate": 1.4873387778826344e-05,
+      "loss": 0.7054,
+      "step": 6873
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 41.35726232144363,
+      "learning_rate": 1.4871915124052426e-05,
+      "loss": 0.7493,
+      "step": 6874
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.251485623138855,
+      "learning_rate": 1.4870442330724088e-05,
+      "loss": 0.7343,
+      "step": 6875
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.95634489797863,
+      "learning_rate": 1.4868969398883224e-05,
+      "loss": 0.6991,
+      "step": 6876
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 34.891627036370465,
+      "learning_rate": 1.4867496328571716e-05,
+      "loss": 0.7244,
+      "step": 6877
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 38.058581646957755,
+      "learning_rate": 1.4866023119831461e-05,
+      "loss": 0.8002,
+      "step": 6878
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 35.401916023169875,
+      "learning_rate": 1.4864549772704353e-05,
+      "loss": 0.7413,
+      "step": 6879
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 44.00730628344307,
+      "learning_rate": 1.4863076287232297e-05,
+      "loss": 0.7354,
+      "step": 6880
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.57268674396649,
+      "learning_rate": 1.4861602663457195e-05,
+      "loss": 0.6791,
+      "step": 6881
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.130716831778365,
+      "learning_rate": 1.4860128901420958e-05,
+      "loss": 0.8511,
+      "step": 6882
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 52.93654647022428,
+      "learning_rate": 1.4858655001165498e-05,
+      "loss": 0.7131,
+      "step": 6883
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 31.287642554592008,
+      "learning_rate": 1.4857180962732731e-05,
+      "loss": 0.7609,
+      "step": 6884
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 51.98389960422197,
+      "learning_rate": 1.4855706786164579e-05,
+      "loss": 0.7465,
+      "step": 6885
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 44.84649847559841,
+      "learning_rate": 1.4854232471502967e-05,
+      "loss": 0.7823,
+      "step": 6886
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.12145300452661,
+      "learning_rate": 1.4852758018789826e-05,
+      "loss": 0.6662,
+      "step": 6887
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.6259489196909,
+      "learning_rate": 1.4851283428067083e-05,
+      "loss": 0.7602,
+      "step": 6888
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 40.041353307179946,
+      "learning_rate": 1.4849808699376675e-05,
+      "loss": 0.8259,
+      "step": 6889
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 33.96117658862226,
+      "learning_rate": 1.4848333832760549e-05,
+      "loss": 0.7638,
+      "step": 6890
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.218162846507127,
+      "learning_rate": 1.4846858828260642e-05,
+      "loss": 0.6913,
+      "step": 6891
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 43.712006311244075,
+      "learning_rate": 1.4845383685918906e-05,
+      "loss": 0.801,
+      "step": 6892
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 32.09607751238317,
+      "learning_rate": 1.4843908405777294e-05,
+      "loss": 0.711,
+      "step": 6893
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 51.90917130987727,
+      "learning_rate": 1.4842432987877758e-05,
+      "loss": 0.6856,
+      "step": 6894
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.028058152426475,
+      "learning_rate": 1.4840957432262265e-05,
+      "loss": 0.7293,
+      "step": 6895
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 28.710300738481195,
+      "learning_rate": 1.483948173897277e-05,
+      "loss": 0.641,
+      "step": 6896
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 29.47350757701571,
+      "learning_rate": 1.4838005908051245e-05,
+      "loss": 0.6009,
+      "step": 6897
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.29101860126152,
+      "learning_rate": 1.4836529939539665e-05,
+      "loss": 0.7001,
+      "step": 6898
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 40.20854795955372,
+      "learning_rate": 1.4835053833480003e-05,
+      "loss": 0.7575,
+      "step": 6899
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 41.32810058881475,
+      "learning_rate": 1.4833577589914234e-05,
+      "loss": 0.7464,
+      "step": 6900
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 39.71760557433301,
+      "learning_rate": 1.4832101208884352e-05,
+      "loss": 0.8691,
+      "step": 6901
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 28.866684100106887,
+      "learning_rate": 1.4830624690432331e-05,
+      "loss": 0.6483,
+      "step": 6902
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 43.98947156307333,
+      "learning_rate": 1.4829148034600174e-05,
+      "loss": 0.7612,
+      "step": 6903
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 37.350075394325806,
+      "learning_rate": 1.4827671241429873e-05,
+      "loss": 0.7675,
+      "step": 6904
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.067672455759094,
+      "learning_rate": 1.4826194310963424e-05,
+      "loss": 0.7687,
+      "step": 6905
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.68505098899955,
+      "learning_rate": 1.4824717243242835e-05,
+      "loss": 0.6847,
+      "step": 6906
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 37.93025246385048,
+      "learning_rate": 1.4823240038310108e-05,
+      "loss": 0.7624,
+      "step": 6907
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 37.27194401170206,
+      "learning_rate": 1.4821762696207254e-05,
+      "loss": 0.7262,
+      "step": 6908
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.44269193924497,
+      "learning_rate": 1.4820285216976288e-05,
+      "loss": 0.7451,
+      "step": 6909
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 44.34963113659351,
+      "learning_rate": 1.4818807600659237e-05,
+      "loss": 0.746,
+      "step": 6910
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.87781935750085,
+      "learning_rate": 1.4817329847298112e-05,
+      "loss": 0.7824,
+      "step": 6911
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 37.590729957590696,
+      "learning_rate": 1.4815851956934948e-05,
+      "loss": 0.7547,
+      "step": 6912
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 31.327215072088634,
+      "learning_rate": 1.4814373929611768e-05,
+      "loss": 0.8039,
+      "step": 6913
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 40.61275456079582,
+      "learning_rate": 1.4812895765370609e-05,
+      "loss": 0.6869,
+      "step": 6914
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 40.801045859479956,
+      "learning_rate": 1.4811417464253512e-05,
+      "loss": 0.771,
+      "step": 6915
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 39.00423440503875,
+      "learning_rate": 1.4809939026302518e-05,
+      "loss": 0.7233,
+      "step": 6916
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 35.59567432041337,
+      "learning_rate": 1.4808460451559669e-05,
+      "loss": 0.6967,
+      "step": 6917
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 35.39964997204723,
+      "learning_rate": 1.4806981740067018e-05,
+      "loss": 0.7744,
+      "step": 6918
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.5980364582748,
+      "learning_rate": 1.480550289186662e-05,
+      "loss": 0.6787,
+      "step": 6919
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.25105541306532,
+      "learning_rate": 1.480402390700053e-05,
+      "loss": 0.8011,
+      "step": 6920
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 35.02285223222329,
+      "learning_rate": 1.4802544785510811e-05,
+      "loss": 0.6426,
+      "step": 6921
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.609621636028447,
+      "learning_rate": 1.4801065527439526e-05,
+      "loss": 0.6867,
+      "step": 6922
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.20137771249105,
+      "learning_rate": 1.4799586132828747e-05,
+      "loss": 0.6878,
+      "step": 6923
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.7860731908971,
+      "learning_rate": 1.4798106601720545e-05,
+      "loss": 0.7915,
+      "step": 6924
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 31.457427021048975,
+      "learning_rate": 1.4796626934157e-05,
+      "loss": 0.7061,
+      "step": 6925
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 51.556283370289414,
+      "learning_rate": 1.4795147130180192e-05,
+      "loss": 0.7007,
+      "step": 6926
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.03843306366588,
+      "learning_rate": 1.4793667189832201e-05,
+      "loss": 0.7227,
+      "step": 6927
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.008997448464186,
+      "learning_rate": 1.4792187113155121e-05,
+      "loss": 0.7249,
+      "step": 6928
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 32.757509462613875,
+      "learning_rate": 1.4790706900191043e-05,
+      "loss": 0.6354,
+      "step": 6929
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 41.7106966510799,
+      "learning_rate": 1.4789226550982062e-05,
+      "loss": 0.7963,
+      "step": 6930
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 51.01076152805164,
+      "learning_rate": 1.4787746065570282e-05,
+      "loss": 0.8097,
+      "step": 6931
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 39.288084954407815,
+      "learning_rate": 1.4786265443997805e-05,
+      "loss": 0.7053,
+      "step": 6932
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 40.23720694343667,
+      "learning_rate": 1.4784784686306737e-05,
+      "loss": 0.7006,
+      "step": 6933
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.30409094441645,
+      "learning_rate": 1.478330379253919e-05,
+      "loss": 0.7385,
+      "step": 6934
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 43.66623447413023,
+      "learning_rate": 1.4781822762737284e-05,
+      "loss": 0.7702,
+      "step": 6935
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.11045968802226,
+      "learning_rate": 1.4780341596943133e-05,
+      "loss": 0.7014,
+      "step": 6936
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.460921865509633,
+      "learning_rate": 1.4778860295198869e-05,
+      "loss": 0.6248,
+      "step": 6937
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 37.387690874159006,
+      "learning_rate": 1.4777378857546606e-05,
+      "loss": 0.8395,
+      "step": 6938
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.124547645350205,
+      "learning_rate": 1.4775897284028491e-05,
+      "loss": 0.6125,
+      "step": 6939
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.824700851416644,
+      "learning_rate": 1.4774415574686645e-05,
+      "loss": 0.7466,
+      "step": 6940
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 32.1347964782882,
+      "learning_rate": 1.4772933729563215e-05,
+      "loss": 0.6727,
+      "step": 6941
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 36.289211953700246,
+      "learning_rate": 1.4771451748700344e-05,
+      "loss": 0.774,
+      "step": 6942
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 37.60106669229264,
+      "learning_rate": 1.4769969632140173e-05,
+      "loss": 0.7467,
+      "step": 6943
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.46835105830445,
+      "learning_rate": 1.4768487379924857e-05,
+      "loss": 0.6862,
+      "step": 6944
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.12160686419359,
+      "learning_rate": 1.476700499209655e-05,
+      "loss": 0.6678,
+      "step": 6945
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.74307438273175,
+      "learning_rate": 1.4765522468697406e-05,
+      "loss": 0.8059,
+      "step": 6946
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 41.86414610416184,
+      "learning_rate": 1.4764039809769598e-05,
+      "loss": 0.7524,
+      "step": 6947
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 34.17447177699195,
+      "learning_rate": 1.4762557015355278e-05,
+      "loss": 0.6966,
+      "step": 6948
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 40.035193711359234,
+      "learning_rate": 1.4761074085496625e-05,
+      "loss": 0.7824,
+      "step": 6949
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.55838242290202,
+      "learning_rate": 1.4759591020235812e-05,
+      "loss": 0.7072,
+      "step": 6950
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 44.18134339970678,
+      "learning_rate": 1.4758107819615014e-05,
+      "loss": 0.7013,
+      "step": 6951
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 37.54884560533769,
+      "learning_rate": 1.4756624483676412e-05,
+      "loss": 0.6982,
+      "step": 6952
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 35.73765137618015,
+      "learning_rate": 1.475514101246219e-05,
+      "loss": 0.6821,
+      "step": 6953
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.378476333345674,
+      "learning_rate": 1.4753657406014545e-05,
+      "loss": 0.7293,
+      "step": 6954
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 33.87879774757339,
+      "learning_rate": 1.475217366437566e-05,
+      "loss": 0.7502,
+      "step": 6955
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 41.4955787745239,
+      "learning_rate": 1.4750689787587735e-05,
+      "loss": 0.7157,
+      "step": 6956
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 34.38081786774007,
+      "learning_rate": 1.4749205775692974e-05,
+      "loss": 0.6757,
+      "step": 6957
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 34.70482853912374,
+      "learning_rate": 1.4747721628733578e-05,
+      "loss": 0.6749,
+      "step": 6958
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.79640915654777,
+      "learning_rate": 1.4746237346751757e-05,
+      "loss": 0.6443,
+      "step": 6959
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.99490094812081,
+      "learning_rate": 1.4744752929789723e-05,
+      "loss": 0.6927,
+      "step": 6960
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.541044829473805,
+      "learning_rate": 1.474326837788969e-05,
+      "loss": 0.8436,
+      "step": 6961
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.65352481928555,
+      "learning_rate": 1.4741783691093879e-05,
+      "loss": 0.7254,
+      "step": 6962
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 44.80697415277909,
+      "learning_rate": 1.4740298869444512e-05,
+      "loss": 0.7839,
+      "step": 6963
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 39.76635053231329,
+      "learning_rate": 1.4738813912983823e-05,
+      "loss": 0.8954,
+      "step": 6964
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.237315670165366,
+      "learning_rate": 1.4737328821754034e-05,
+      "loss": 0.774,
+      "step": 6965
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.023089201102447,
+      "learning_rate": 1.4735843595797388e-05,
+      "loss": 0.7083,
+      "step": 6966
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 29.334518115859925,
+      "learning_rate": 1.4734358235156117e-05,
+      "loss": 0.6417,
+      "step": 6967
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 40.77811836700421,
+      "learning_rate": 1.4732872739872469e-05,
+      "loss": 0.8144,
+      "step": 6968
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.515300790100774,
+      "learning_rate": 1.4731387109988689e-05,
+      "loss": 0.7359,
+      "step": 6969
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 40.92742449989693,
+      "learning_rate": 1.4729901345547027e-05,
+      "loss": 0.8006,
+      "step": 6970
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.82283607811186,
+      "learning_rate": 1.4728415446589736e-05,
+      "loss": 0.6834,
+      "step": 6971
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.438049941059116,
+      "learning_rate": 1.4726929413159077e-05,
+      "loss": 0.6163,
+      "step": 6972
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 34.49617610693055,
+      "learning_rate": 1.472544324529731e-05,
+      "loss": 0.7881,
+      "step": 6973
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.851905693379265,
+      "learning_rate": 1.4723956943046705e-05,
+      "loss": 0.8295,
+      "step": 6974
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 35.94918079782506,
+      "learning_rate": 1.4722470506449525e-05,
+      "loss": 0.7495,
+      "step": 6975
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 30.30661407821583,
+      "learning_rate": 1.4720983935548048e-05,
+      "loss": 0.6921,
+      "step": 6976
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.52290841154895,
+      "learning_rate": 1.4719497230384548e-05,
+      "loss": 0.789,
+      "step": 6977
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.36473729432884,
+      "learning_rate": 1.4718010391001306e-05,
+      "loss": 0.6629,
+      "step": 6978
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 35.70833680825382,
+      "learning_rate": 1.4716523417440613e-05,
+      "loss": 0.8211,
+      "step": 6979
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 37.44633618425326,
+      "learning_rate": 1.4715036309744749e-05,
+      "loss": 0.723,
+      "step": 6980
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.420519296268886,
+      "learning_rate": 1.4713549067956011e-05,
+      "loss": 0.8051,
+      "step": 6981
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 37.41238580540373,
+      "learning_rate": 1.4712061692116697e-05,
+      "loss": 0.7034,
+      "step": 6982
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 42.69842347086814,
+      "learning_rate": 1.4710574182269103e-05,
+      "loss": 0.734,
+      "step": 6983
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 43.95212593122959,
+      "learning_rate": 1.4709086538455537e-05,
+      "loss": 0.8096,
+      "step": 6984
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 30.26624577505509,
+      "learning_rate": 1.4707598760718305e-05,
+      "loss": 0.5911,
+      "step": 6985
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 47.779772728786234,
+      "learning_rate": 1.4706110849099715e-05,
+      "loss": 0.7272,
+      "step": 6986
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 29.432974113113403,
+      "learning_rate": 1.4704622803642089e-05,
+      "loss": 0.7694,
+      "step": 6987
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.92558305214258,
+      "learning_rate": 1.470313462438774e-05,
+      "loss": 0.8555,
+      "step": 6988
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 35.225332279682185,
+      "learning_rate": 1.4701646311378994e-05,
+      "loss": 0.7047,
+      "step": 6989
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 49.98682583342122,
+      "learning_rate": 1.4700157864658176e-05,
+      "loss": 0.8214,
+      "step": 6990
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.895966171867,
+      "learning_rate": 1.4698669284267619e-05,
+      "loss": 0.7039,
+      "step": 6991
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 33.0006006070592,
+      "learning_rate": 1.4697180570249655e-05,
+      "loss": 0.7714,
+      "step": 6992
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.33779295121273,
+      "learning_rate": 1.4695691722646626e-05,
+      "loss": 0.6827,
+      "step": 6993
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.57740884020325,
+      "learning_rate": 1.4694202741500871e-05,
+      "loss": 0.7444,
+      "step": 6994
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 33.20117502029057,
+      "learning_rate": 1.4692713626854735e-05,
+      "loss": 0.7363,
+      "step": 6995
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 33.24036732227967,
+      "learning_rate": 1.4691224378750569e-05,
+      "loss": 0.668,
+      "step": 6996
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 40.58496531991768,
+      "learning_rate": 1.4689734997230724e-05,
+      "loss": 0.6276,
+      "step": 6997
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.839959178196317,
+      "learning_rate": 1.4688245482337565e-05,
+      "loss": 0.687,
+      "step": 6998
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 34.83603098262846,
+      "learning_rate": 1.468675583411344e-05,
+      "loss": 0.7258,
+      "step": 6999
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.50928250312206,
+      "learning_rate": 1.4685266052600724e-05,
+      "loss": 0.7222,
+      "step": 7000
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 30.378158605262854,
+      "learning_rate": 1.468377613784178e-05,
+      "loss": 0.7035,
+      "step": 7001
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 36.47736508582735,
+      "learning_rate": 1.4682286089878985e-05,
+      "loss": 0.7135,
+      "step": 7002
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 40.479803683375984,
+      "learning_rate": 1.4680795908754713e-05,
+      "loss": 0.7484,
+      "step": 7003
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 42.26021652168017,
+      "learning_rate": 1.4679305594511344e-05,
+      "loss": 0.7704,
+      "step": 7004
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 37.57392993756122,
+      "learning_rate": 1.4677815147191256e-05,
+      "loss": 0.8069,
+      "step": 7005
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.7351002612648,
+      "learning_rate": 1.4676324566836848e-05,
+      "loss": 0.7213,
+      "step": 7006
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 33.79480032652886,
+      "learning_rate": 1.4674833853490502e-05,
+      "loss": 0.7372,
+      "step": 7007
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 32.1321323321861,
+      "learning_rate": 1.4673343007194618e-05,
+      "loss": 0.6994,
+      "step": 7008
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 48.23749624322984,
+      "learning_rate": 1.4671852027991589e-05,
+      "loss": 0.7008,
+      "step": 7009
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 28.515467813136972,
+      "learning_rate": 1.4670360915923824e-05,
+      "loss": 0.6781,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 39.31795223738355,
+      "learning_rate": 1.4668869671033726e-05,
+      "loss": 0.7028,
+      "step": 7011
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.56447522337749,
+      "learning_rate": 1.4667378293363706e-05,
+      "loss": 0.6359,
+      "step": 7012
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 31.257003373264226,
+      "learning_rate": 1.4665886782956178e-05,
+      "loss": 0.6376,
+      "step": 7013
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 43.139566930101275,
+      "learning_rate": 1.4664395139853558e-05,
+      "loss": 0.8133,
+      "step": 7014
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 30.223834877259428,
+      "learning_rate": 1.466290336409827e-05,
+      "loss": 0.6546,
+      "step": 7015
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.78267618674685,
+      "learning_rate": 1.4661411455732735e-05,
+      "loss": 0.741,
+      "step": 7016
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 43.09305608742579,
+      "learning_rate": 1.4659919414799389e-05,
+      "loss": 0.6782,
+      "step": 7017
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 43.04525834663489,
+      "learning_rate": 1.465842724134066e-05,
+      "loss": 0.814,
+      "step": 7018
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.34000319030026,
+      "learning_rate": 1.4656934935398983e-05,
+      "loss": 0.6583,
+      "step": 7019
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 37.2117465434041,
+      "learning_rate": 1.4655442497016802e-05,
+      "loss": 0.765,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 32.34454147549891,
+      "learning_rate": 1.4653949926236562e-05,
+      "loss": 0.676,
+      "step": 7021
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 32.302303409821086,
+      "learning_rate": 1.4652457223100704e-05,
+      "loss": 0.6198,
+      "step": 7022
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 27.408225628795964,
+      "learning_rate": 1.4650964387651692e-05,
+      "loss": 0.5476,
+      "step": 7023
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 42.701442044674586,
+      "learning_rate": 1.4649471419931964e-05,
+      "loss": 0.7814,
+      "step": 7024
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 34.685873811423534,
+      "learning_rate": 1.4647978319983998e-05,
+      "loss": 0.6708,
+      "step": 7025
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.89804001011928,
+      "learning_rate": 1.4646485087850239e-05,
+      "loss": 0.8045,
+      "step": 7026
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.3561468166037,
+      "learning_rate": 1.4644991723573167e-05,
+      "loss": 0.6488,
+      "step": 7027
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 27.789584828707582,
+      "learning_rate": 1.4643498227195247e-05,
+      "loss": 0.68,
+      "step": 7028
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.90943622297069,
+      "learning_rate": 1.4642004598758955e-05,
+      "loss": 0.7079,
+      "step": 7029
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 37.31158288950004,
+      "learning_rate": 1.4640510838306765e-05,
+      "loss": 0.6878,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 47.37410959197133,
+      "learning_rate": 1.4639016945881164e-05,
+      "loss": 0.8378,
+      "step": 7031
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 42.08494794159142,
+      "learning_rate": 1.4637522921524632e-05,
+      "loss": 0.6713,
+      "step": 7032
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 29.405073330857018,
+      "learning_rate": 1.4636028765279662e-05,
+      "loss": 0.6325,
+      "step": 7033
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 35.18424573319424,
+      "learning_rate": 1.4634534477188745e-05,
+      "loss": 0.5776,
+      "step": 7034
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 38.47586654730779,
+      "learning_rate": 1.4633040057294379e-05,
+      "loss": 0.6996,
+      "step": 7035
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 39.87558219149824,
+      "learning_rate": 1.4631545505639066e-05,
+      "loss": 0.7071,
+      "step": 7036
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 27.73843900750911,
+      "learning_rate": 1.4630050822265306e-05,
+      "loss": 0.6503,
+      "step": 7037
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.33773686210783,
+      "learning_rate": 1.4628556007215612e-05,
+      "loss": 0.6684,
+      "step": 7038
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 29.48460975153249,
+      "learning_rate": 1.462706106053249e-05,
+      "loss": 0.7017,
+      "step": 7039
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 46.41774285311339,
+      "learning_rate": 1.4625565982258461e-05,
+      "loss": 0.7608,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 31.133075436351685,
+      "learning_rate": 1.4624070772436037e-05,
+      "loss": 0.635,
+      "step": 7041
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 45.790978990874926,
+      "learning_rate": 1.4622575431107748e-05,
+      "loss": 0.6374,
+      "step": 7042
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 39.22168362044765,
+      "learning_rate": 1.4621079958316118e-05,
+      "loss": 0.686,
+      "step": 7043
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 35.37867093668724,
+      "learning_rate": 1.4619584354103675e-05,
+      "loss": 0.7521,
+      "step": 7044
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 51.98589290401316,
+      "learning_rate": 1.4618088618512957e-05,
+      "loss": 0.7315,
+      "step": 7045
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 39.84483246772731,
+      "learning_rate": 1.4616592751586503e-05,
+      "loss": 0.8392,
+      "step": 7046
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 35.77443940425191,
+      "learning_rate": 1.4615096753366849e-05,
+      "loss": 0.7658,
+      "step": 7047
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.1215340366516,
+      "learning_rate": 1.4613600623896542e-05,
+      "loss": 0.6977,
+      "step": 7048
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.39060125497376,
+      "learning_rate": 1.4612104363218132e-05,
+      "loss": 0.7647,
+      "step": 7049
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 32.37241993200664,
+      "learning_rate": 1.4610607971374175e-05,
+      "loss": 0.7378,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 29.846043349821294,
+      "learning_rate": 1.460911144840722e-05,
+      "loss": 0.6685,
+      "step": 7051
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 32.10484961481167,
+      "learning_rate": 1.4607614794359836e-05,
+      "loss": 0.6564,
+      "step": 7052
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.7022672303687,
+      "learning_rate": 1.4606118009274582e-05,
+      "loss": 0.7422,
+      "step": 7053
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.50117831745048,
+      "learning_rate": 1.4604621093194024e-05,
+      "loss": 0.7062,
+      "step": 7054
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 34.797796530954784,
+      "learning_rate": 1.4603124046160737e-05,
+      "loss": 0.749,
+      "step": 7055
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.03791792379473,
+      "learning_rate": 1.4601626868217296e-05,
+      "loss": 0.7028,
+      "step": 7056
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 37.35392530029633,
+      "learning_rate": 1.4600129559406278e-05,
+      "loss": 0.7013,
+      "step": 7057
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 43.423180186122956,
+      "learning_rate": 1.4598632119770266e-05,
+      "loss": 0.8206,
+      "step": 7058
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 31.46810633617612,
+      "learning_rate": 1.4597134549351845e-05,
+      "loss": 0.6796,
+      "step": 7059
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 34.3169058488184,
+      "learning_rate": 1.459563684819361e-05,
+      "loss": 0.696,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 29.664670687316303,
+      "learning_rate": 1.4594139016338149e-05,
+      "loss": 0.6593,
+      "step": 7061
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 33.65110168467423,
+      "learning_rate": 1.4592641053828063e-05,
+      "loss": 0.6425,
+      "step": 7062
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 40.17486110110149,
+      "learning_rate": 1.4591142960705951e-05,
+      "loss": 0.8083,
+      "step": 7063
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 40.5481355621843,
+      "learning_rate": 1.4589644737014418e-05,
+      "loss": 0.7411,
+      "step": 7064
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 47.675265356364555,
+      "learning_rate": 1.4588146382796075e-05,
+      "loss": 0.7698,
+      "step": 7065
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 43.325134570335955,
+      "learning_rate": 1.4586647898093534e-05,
+      "loss": 0.6938,
+      "step": 7066
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 26.202906569569645,
+      "learning_rate": 1.4585149282949409e-05,
+      "loss": 0.6445,
+      "step": 7067
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 36.55237908115507,
+      "learning_rate": 1.458365053740632e-05,
+      "loss": 0.7584,
+      "step": 7068
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 34.8562298992688,
+      "learning_rate": 1.4582151661506894e-05,
+      "loss": 0.622,
+      "step": 7069
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 30.591984402731388,
+      "learning_rate": 1.458065265529375e-05,
+      "loss": 0.6814,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 41.54041103708832,
+      "learning_rate": 1.4579153518809527e-05,
+      "loss": 0.8368,
+      "step": 7071
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 42.89282901569469,
+      "learning_rate": 1.4577654252096856e-05,
+      "loss": 0.6899,
+      "step": 7072
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 37.72563615468607,
+      "learning_rate": 1.4576154855198377e-05,
+      "loss": 0.7545,
+      "step": 7073
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 30.228573694510402,
+      "learning_rate": 1.457465532815673e-05,
+      "loss": 0.6422,
+      "step": 7074
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.764859868667834,
+      "learning_rate": 1.4573155671014563e-05,
+      "loss": 0.676,
+      "step": 7075
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 39.45103577171519,
+      "learning_rate": 1.457165588381452e-05,
+      "loss": 0.7243,
+      "step": 7076
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 37.614175532132386,
+      "learning_rate": 1.4570155966599264e-05,
+      "loss": 0.715,
+      "step": 7077
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.38463243147605,
+      "learning_rate": 1.456865591941144e-05,
+      "loss": 0.6631,
+      "step": 7078
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 39.75837006405634,
+      "learning_rate": 1.4567155742293717e-05,
+      "loss": 0.6409,
+      "step": 7079
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.71408247467522,
+      "learning_rate": 1.4565655435288756e-05,
+      "loss": 0.6598,
+      "step": 7080
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 42.638905849446516,
+      "learning_rate": 1.4564154998439225e-05,
+      "loss": 0.9319,
+      "step": 7081
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.394034578404415,
+      "learning_rate": 1.4562654431787797e-05,
+      "loss": 0.5991,
+      "step": 7082
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.726112313928713,
+      "learning_rate": 1.4561153735377147e-05,
+      "loss": 0.7078,
+      "step": 7083
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.88204574860842,
+      "learning_rate": 1.455965290924995e-05,
+      "loss": 0.6946,
+      "step": 7084
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 36.631302157893565,
+      "learning_rate": 1.4558151953448891e-05,
+      "loss": 0.7629,
+      "step": 7085
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 40.341001838084146,
+      "learning_rate": 1.4556650868016661e-05,
+      "loss": 0.7438,
+      "step": 7086
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.04561026642824,
+      "learning_rate": 1.4555149652995942e-05,
+      "loss": 0.7745,
+      "step": 7087
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 39.99162532040896,
+      "learning_rate": 1.4553648308429433e-05,
+      "loss": 0.7071,
+      "step": 7088
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 41.30551981281295,
+      "learning_rate": 1.455214683435983e-05,
+      "loss": 0.767,
+      "step": 7089
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 37.299620298391034,
+      "learning_rate": 1.4550645230829837e-05,
+      "loss": 0.6978,
+      "step": 7090
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 59.412773757390546,
+      "learning_rate": 1.4549143497882152e-05,
+      "loss": 0.7492,
+      "step": 7091
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.515225051920524,
+      "learning_rate": 1.4547641635559488e-05,
+      "loss": 0.7132,
+      "step": 7092
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 29.48872364443546,
+      "learning_rate": 1.4546139643904553e-05,
+      "loss": 0.6667,
+      "step": 7093
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.396744354710858,
+      "learning_rate": 1.4544637522960072e-05,
+      "loss": 0.7103,
+      "step": 7094
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 31.87098767889087,
+      "learning_rate": 1.4543135272768753e-05,
+      "loss": 0.663,
+      "step": 7095
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.93113659375826,
+      "learning_rate": 1.454163289337333e-05,
+      "loss": 0.6681,
+      "step": 7096
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 39.61896782457271,
+      "learning_rate": 1.454013038481652e-05,
+      "loss": 0.8107,
+      "step": 7097
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 48.03057386310835,
+      "learning_rate": 1.4538627747141058e-05,
+      "loss": 0.6563,
+      "step": 7098
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 40.20426435223811,
+      "learning_rate": 1.4537124980389678e-05,
+      "loss": 0.8006,
+      "step": 7099
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 36.391142006555526,
+      "learning_rate": 1.4535622084605119e-05,
+      "loss": 0.7312,
+      "step": 7100
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 29.329421446104988,
+      "learning_rate": 1.453411905983012e-05,
+      "loss": 0.6742,
+      "step": 7101
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 31.678445779004736,
+      "learning_rate": 1.453261590610743e-05,
+      "loss": 0.6343,
+      "step": 7102
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.060582486697356,
+      "learning_rate": 1.453111262347979e-05,
+      "loss": 0.7173,
+      "step": 7103
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.562315441970185,
+      "learning_rate": 1.4529609211989962e-05,
+      "loss": 0.7397,
+      "step": 7104
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.96553958610906,
+      "learning_rate": 1.4528105671680697e-05,
+      "loss": 0.756,
+      "step": 7105
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 53.94298856178161,
+      "learning_rate": 1.4526602002594755e-05,
+      "loss": 0.6907,
+      "step": 7106
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.34280821124752,
+      "learning_rate": 1.45250982047749e-05,
+      "loss": 0.7475,
+      "step": 7107
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 40.74918216048455,
+      "learning_rate": 1.4523594278263899e-05,
+      "loss": 0.7796,
+      "step": 7108
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 56.76744779331977,
+      "learning_rate": 1.4522090223104523e-05,
+      "loss": 0.7392,
+      "step": 7109
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.60447542998998,
+      "learning_rate": 1.4520586039339545e-05,
+      "loss": 0.8105,
+      "step": 7110
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.0195915166961,
+      "learning_rate": 1.4519081727011747e-05,
+      "loss": 0.5994,
+      "step": 7111
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 38.135918035739415,
+      "learning_rate": 1.4517577286163906e-05,
+      "loss": 0.7984,
+      "step": 7112
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.256502989100824,
+      "learning_rate": 1.4516072716838815e-05,
+      "loss": 0.7136,
+      "step": 7113
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.40188298528257,
+      "learning_rate": 1.4514568019079252e-05,
+      "loss": 0.7104,
+      "step": 7114
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.29017964488197,
+      "learning_rate": 1.4513063192928016e-05,
+      "loss": 0.721,
+      "step": 7115
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 28.63194868338216,
+      "learning_rate": 1.4511558238427905e-05,
+      "loss": 0.6511,
+      "step": 7116
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 45.84152702033353,
+      "learning_rate": 1.4510053155621714e-05,
+      "loss": 0.6782,
+      "step": 7117
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.35273805905443,
+      "learning_rate": 1.4508547944552254e-05,
+      "loss": 0.7222,
+      "step": 7118
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.0301844263486,
+      "learning_rate": 1.4507042605262324e-05,
+      "loss": 0.6921,
+      "step": 7119
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.32240822794643,
+      "learning_rate": 1.450553713779474e-05,
+      "loss": 0.648,
+      "step": 7120
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.94791651550994,
+      "learning_rate": 1.4504031542192319e-05,
+      "loss": 0.7583,
+      "step": 7121
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 45.3224764976207,
+      "learning_rate": 1.450252581849787e-05,
+      "loss": 0.7601,
+      "step": 7122
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 31.371639178092988,
+      "learning_rate": 1.4501019966754225e-05,
+      "loss": 0.6147,
+      "step": 7123
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.292858876681464,
+      "learning_rate": 1.4499513987004201e-05,
+      "loss": 0.6227,
+      "step": 7124
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.56246898964869,
+      "learning_rate": 1.4498007879290635e-05,
+      "loss": 0.7076,
+      "step": 7125
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 41.327446886928115,
+      "learning_rate": 1.4496501643656356e-05,
+      "loss": 0.6779,
+      "step": 7126
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.42359350590761,
+      "learning_rate": 1.4494995280144198e-05,
+      "loss": 0.6508,
+      "step": 7127
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 31.777762395144414,
+      "learning_rate": 1.4493488788797006e-05,
+      "loss": 0.6423,
+      "step": 7128
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 39.855282252255876,
+      "learning_rate": 1.449198216965762e-05,
+      "loss": 0.8243,
+      "step": 7129
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 33.102078348929645,
+      "learning_rate": 1.449047542276889e-05,
+      "loss": 0.6454,
+      "step": 7130
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 32.71528554680699,
+      "learning_rate": 1.4488968548173662e-05,
+      "loss": 0.6956,
+      "step": 7131
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 38.89889582486353,
+      "learning_rate": 1.4487461545914801e-05,
+      "loss": 0.7464,
+      "step": 7132
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 46.46745970627139,
+      "learning_rate": 1.4485954416035152e-05,
+      "loss": 0.7455,
+      "step": 7133
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 45.18281185792459,
+      "learning_rate": 1.4484447158577588e-05,
+      "loss": 0.693,
+      "step": 7134
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.30514116833563,
+      "learning_rate": 1.448293977358497e-05,
+      "loss": 0.6941,
+      "step": 7135
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 37.66446379356982,
+      "learning_rate": 1.4481432261100168e-05,
+      "loss": 0.7647,
+      "step": 7136
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 45.48911955975333,
+      "learning_rate": 1.4479924621166051e-05,
+      "loss": 0.7172,
+      "step": 7137
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 57.4299993090104,
+      "learning_rate": 1.4478416853825502e-05,
+      "loss": 0.8564,
+      "step": 7138
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 43.51229772260465,
+      "learning_rate": 1.4476908959121394e-05,
+      "loss": 0.8384,
+      "step": 7139
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.25150693746227,
+      "learning_rate": 1.4475400937096617e-05,
+      "loss": 0.6945,
+      "step": 7140
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 45.12125636393376,
+      "learning_rate": 1.4473892787794053e-05,
+      "loss": 0.7883,
+      "step": 7141
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.21661765241537,
+      "learning_rate": 1.4472384511256597e-05,
+      "loss": 0.6595,
+      "step": 7142
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 40.37303084776891,
+      "learning_rate": 1.4470876107527141e-05,
+      "loss": 0.75,
+      "step": 7143
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.38590464289263,
+      "learning_rate": 1.4469367576648582e-05,
+      "loss": 0.735,
+      "step": 7144
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.44124189928879,
+      "learning_rate": 1.4467858918663826e-05,
+      "loss": 0.7624,
+      "step": 7145
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.008462373319695,
+      "learning_rate": 1.4466350133615776e-05,
+      "loss": 0.656,
+      "step": 7146
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.05893438316406,
+      "learning_rate": 1.4464841221547339e-05,
+      "loss": 0.7978,
+      "step": 7147
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.17907540815813,
+      "learning_rate": 1.4463332182501431e-05,
+      "loss": 0.6486,
+      "step": 7148
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.635564135938374,
+      "learning_rate": 1.4461823016520967e-05,
+      "loss": 0.7762,
+      "step": 7149
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 34.53311993813461,
+      "learning_rate": 1.4460313723648866e-05,
+      "loss": 0.6599,
+      "step": 7150
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.76673436493848,
+      "learning_rate": 1.4458804303928053e-05,
+      "loss": 0.6542,
+      "step": 7151
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 31.894070074294483,
+      "learning_rate": 1.4457294757401452e-05,
+      "loss": 0.6871,
+      "step": 7152
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 52.21448836865962,
+      "learning_rate": 1.4455785084111997e-05,
+      "loss": 0.7698,
+      "step": 7153
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 38.680321287765615,
+      "learning_rate": 1.445427528410262e-05,
+      "loss": 0.6823,
+      "step": 7154
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.71950866430863,
+      "learning_rate": 1.4452765357416261e-05,
+      "loss": 0.791,
+      "step": 7155
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 34.8489536796038,
+      "learning_rate": 1.4451255304095857e-05,
+      "loss": 0.642,
+      "step": 7156
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 36.11268244525935,
+      "learning_rate": 1.444974512418436e-05,
+      "loss": 0.682,
+      "step": 7157
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 31.894739415458307,
+      "learning_rate": 1.444823481772471e-05,
+      "loss": 0.6937,
+      "step": 7158
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.45387145255182,
+      "learning_rate": 1.4446724384759869e-05,
+      "loss": 0.6733,
+      "step": 7159
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 29.673232736002053,
+      "learning_rate": 1.4445213825332784e-05,
+      "loss": 0.707,
+      "step": 7160
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.37186730653694,
+      "learning_rate": 1.444370313948642e-05,
+      "loss": 0.8533,
+      "step": 7161
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.23797302579458,
+      "learning_rate": 1.4442192327263737e-05,
+      "loss": 0.6844,
+      "step": 7162
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.012763777245596,
+      "learning_rate": 1.4440681388707704e-05,
+      "loss": 0.6797,
+      "step": 7163
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 41.16970685718756,
+      "learning_rate": 1.4439170323861289e-05,
+      "loss": 0.7721,
+      "step": 7164
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 43.30168409563434,
+      "learning_rate": 1.443765913276747e-05,
+      "loss": 0.7976,
+      "step": 7165
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 34.47619016002398,
+      "learning_rate": 1.4436147815469217e-05,
+      "loss": 0.5966,
+      "step": 7166
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 40.669406662091056,
+      "learning_rate": 1.4434636372009516e-05,
+      "loss": 0.7683,
+      "step": 7167
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 29.563717524677724,
+      "learning_rate": 1.4433124802431353e-05,
+      "loss": 0.6499,
+      "step": 7168
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 37.29655678545016,
+      "learning_rate": 1.4431613106777712e-05,
+      "loss": 0.7052,
+      "step": 7169
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 31.710835725706033,
+      "learning_rate": 1.4430101285091587e-05,
+      "loss": 0.6298,
+      "step": 7170
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 45.20486841884518,
+      "learning_rate": 1.4428589337415975e-05,
+      "loss": 0.7286,
+      "step": 7171
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.862147148662466,
+      "learning_rate": 1.442707726379387e-05,
+      "loss": 0.7379,
+      "step": 7172
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.253244068129646,
+      "learning_rate": 1.4425565064268276e-05,
+      "loss": 0.6972,
+      "step": 7173
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 28.249318855226647,
+      "learning_rate": 1.4424052738882203e-05,
+      "loss": 0.5969,
+      "step": 7174
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.298527165080465,
+      "learning_rate": 1.442254028767866e-05,
+      "loss": 0.5731,
+      "step": 7175
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.444156308629765,
+      "learning_rate": 1.4421027710700655e-05,
+      "loss": 0.7209,
+      "step": 7176
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 38.818156489004,
+      "learning_rate": 1.441951500799121e-05,
+      "loss": 0.7396,
+      "step": 7177
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.83022358193019,
+      "learning_rate": 1.4418002179593345e-05,
+      "loss": 0.7893,
+      "step": 7178
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 31.576526206878757,
+      "learning_rate": 1.441648922555008e-05,
+      "loss": 0.6892,
+      "step": 7179
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 31.464760820124408,
+      "learning_rate": 1.4414976145904445e-05,
+      "loss": 0.6578,
+      "step": 7180
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 37.78460091622804,
+      "learning_rate": 1.441346294069947e-05,
+      "loss": 0.8096,
+      "step": 7181
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 36.317894608759225,
+      "learning_rate": 1.4411949609978196e-05,
+      "loss": 0.7394,
+      "step": 7182
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 52.676268854596586,
+      "learning_rate": 1.441043615378365e-05,
+      "loss": 0.6804,
+      "step": 7183
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 40.6813326994348,
+      "learning_rate": 1.4408922572158883e-05,
+      "loss": 0.7209,
+      "step": 7184
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.97039832904502,
+      "learning_rate": 1.4407408865146935e-05,
+      "loss": 0.6945,
+      "step": 7185
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 34.1173030570912,
+      "learning_rate": 1.4405895032790858e-05,
+      "loss": 0.6969,
+      "step": 7186
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 29.624401730319104,
+      "learning_rate": 1.4404381075133706e-05,
+      "loss": 0.6159,
+      "step": 7187
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.41735087302295,
+      "learning_rate": 1.4402866992218529e-05,
+      "loss": 0.6868,
+      "step": 7188
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.83376508193306,
+      "learning_rate": 1.4401352784088394e-05,
+      "loss": 0.7058,
+      "step": 7189
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 28.478691193146904,
+      "learning_rate": 1.4399838450786358e-05,
+      "loss": 0.6432,
+      "step": 7190
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 48.43327907752621,
+      "learning_rate": 1.439832399235549e-05,
+      "loss": 0.7856,
+      "step": 7191
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.73546299114185,
+      "learning_rate": 1.4396809408838862e-05,
+      "loss": 0.7105,
+      "step": 7192
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 32.805264305500074,
+      "learning_rate": 1.4395294700279545e-05,
+      "loss": 0.6863,
+      "step": 7193
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 31.330454454836563,
+      "learning_rate": 1.439377986672062e-05,
+      "loss": 0.675,
+      "step": 7194
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 35.5372315591555,
+      "learning_rate": 1.4392264908205165e-05,
+      "loss": 0.7257,
+      "step": 7195
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 39.08047692421752,
+      "learning_rate": 1.4390749824776264e-05,
+      "loss": 0.7645,
+      "step": 7196
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.76840363048983,
+      "learning_rate": 1.438923461647701e-05,
+      "loss": 0.7113,
+      "step": 7197
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 45.509746723245605,
+      "learning_rate": 1.4387719283350488e-05,
+      "loss": 0.6303,
+      "step": 7198
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 29.49297071212718,
+      "learning_rate": 1.43862038254398e-05,
+      "loss": 0.6625,
+      "step": 7199
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.89861764747875,
+      "learning_rate": 1.4384688242788034e-05,
+      "loss": 0.6748,
+      "step": 7200
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 34.214195568479056,
+      "learning_rate": 1.4383172535438306e-05,
+      "loss": 0.6569,
+      "step": 7201
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 38.107051445390034,
+      "learning_rate": 1.4381656703433713e-05,
+      "loss": 0.7594,
+      "step": 7202
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 42.08870025568382,
+      "learning_rate": 1.4380140746817366e-05,
+      "loss": 0.7431,
+      "step": 7203
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.102897555978863,
+      "learning_rate": 1.4378624665632379e-05,
+      "loss": 0.645,
+      "step": 7204
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 32.39542826596562,
+      "learning_rate": 1.4377108459921867e-05,
+      "loss": 0.653,
+      "step": 7205
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.90819697002641,
+      "learning_rate": 1.4375592129728952e-05,
+      "loss": 0.6255,
+      "step": 7206
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 34.99983024556001,
+      "learning_rate": 1.4374075675096757e-05,
+      "loss": 0.7243,
+      "step": 7207
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 38.7160946148669,
+      "learning_rate": 1.4372559096068406e-05,
+      "loss": 0.6673,
+      "step": 7208
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 56.14886330501085,
+      "learning_rate": 1.4371042392687036e-05,
+      "loss": 0.7877,
+      "step": 7209
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.79427803491974,
+      "learning_rate": 1.4369525564995774e-05,
+      "loss": 0.6663,
+      "step": 7210
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 38.38699433264774,
+      "learning_rate": 1.4368008613037763e-05,
+      "loss": 0.7844,
+      "step": 7211
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 39.68991099866392,
+      "learning_rate": 1.436649153685614e-05,
+      "loss": 0.7137,
+      "step": 7212
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 34.15912325013116,
+      "learning_rate": 1.4364974336494055e-05,
+      "loss": 0.7014,
+      "step": 7213
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.66053720581611,
+      "learning_rate": 1.436345701199465e-05,
+      "loss": 0.6916,
+      "step": 7214
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 44.31107677224086,
+      "learning_rate": 1.4361939563401082e-05,
+      "loss": 0.8427,
+      "step": 7215
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 46.01714449124028,
+      "learning_rate": 1.4360421990756506e-05,
+      "loss": 0.6755,
+      "step": 7216
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 46.236641552730156,
+      "learning_rate": 1.4358904294104076e-05,
+      "loss": 0.7477,
+      "step": 7217
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 38.1009055122643,
+      "learning_rate": 1.4357386473486961e-05,
+      "loss": 0.7824,
+      "step": 7218
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.12391877119181,
+      "learning_rate": 1.435586852894832e-05,
+      "loss": 0.7137,
+      "step": 7219
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.8103833197153,
+      "learning_rate": 1.435435046053133e-05,
+      "loss": 0.7501,
+      "step": 7220
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.10968270058756,
+      "learning_rate": 1.4352832268279158e-05,
+      "loss": 0.757,
+      "step": 7221
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 29.053383221729703,
+      "learning_rate": 1.4351313952234985e-05,
+      "loss": 0.6435,
+      "step": 7222
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 39.1938822810198,
+      "learning_rate": 1.4349795512441984e-05,
+      "loss": 0.6503,
+      "step": 7223
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 38.030565993866595,
+      "learning_rate": 1.434827694894335e-05,
+      "loss": 0.6432,
+      "step": 7224
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.143205173969974,
+      "learning_rate": 1.4346758261782256e-05,
+      "loss": 0.6847,
+      "step": 7225
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.47595761372968,
+      "learning_rate": 1.4345239451001905e-05,
+      "loss": 0.7692,
+      "step": 7226
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 36.61358799709739,
+      "learning_rate": 1.4343720516645482e-05,
+      "loss": 0.6643,
+      "step": 7227
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 26.237971456361997,
+      "learning_rate": 1.4342201458756193e-05,
+      "loss": 0.6564,
+      "step": 7228
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.04244499183781,
+      "learning_rate": 1.434068227737723e-05,
+      "loss": 0.6823,
+      "step": 7229
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.87715788741333,
+      "learning_rate": 1.4339162972551806e-05,
+      "loss": 0.6193,
+      "step": 7230
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 27.760582832616844,
+      "learning_rate": 1.4337643544323124e-05,
+      "loss": 0.6425,
+      "step": 7231
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 38.8954392944275,
+      "learning_rate": 1.4336123992734396e-05,
+      "loss": 0.6979,
+      "step": 7232
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 41.7846778370798,
+      "learning_rate": 1.433460431782884e-05,
+      "loss": 0.7706,
+      "step": 7233
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.74772165520271,
+      "learning_rate": 1.4333084519649671e-05,
+      "loss": 0.6584,
+      "step": 7234
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 41.025739135569474,
+      "learning_rate": 1.4331564598240112e-05,
+      "loss": 0.7304,
+      "step": 7235
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 49.86607964653809,
+      "learning_rate": 1.4330044553643391e-05,
+      "loss": 0.8544,
+      "step": 7236
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.224160128341666,
+      "learning_rate": 1.4328524385902738e-05,
+      "loss": 0.6971,
+      "step": 7237
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.3604619941236,
+      "learning_rate": 1.432700409506138e-05,
+      "loss": 0.6355,
+      "step": 7238
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.871343769486575,
+      "learning_rate": 1.4325483681162562e-05,
+      "loss": 0.794,
+      "step": 7239
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 29.441612187257864,
+      "learning_rate": 1.4323963144249514e-05,
+      "loss": 0.6887,
+      "step": 7240
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 29.290871483675218,
+      "learning_rate": 1.4322442484365486e-05,
+      "loss": 0.6946,
+      "step": 7241
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 34.390666398366214,
+      "learning_rate": 1.432092170155372e-05,
+      "loss": 0.7478,
+      "step": 7242
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 32.29827285236702,
+      "learning_rate": 1.4319400795857472e-05,
+      "loss": 0.6732,
+      "step": 7243
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 34.81802954667069,
+      "learning_rate": 1.431787976731999e-05,
+      "loss": 0.7232,
+      "step": 7244
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.829693227045258,
+      "learning_rate": 1.4316358615984532e-05,
+      "loss": 0.6304,
+      "step": 7245
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.2702608285501,
+      "learning_rate": 1.4314837341894361e-05,
+      "loss": 0.7596,
+      "step": 7246
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 30.459601956851927,
+      "learning_rate": 1.431331594509274e-05,
+      "loss": 0.678,
+      "step": 7247
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 36.57753266934898,
+      "learning_rate": 1.4311794425622937e-05,
+      "loss": 0.6574,
+      "step": 7248
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.39541561318803,
+      "learning_rate": 1.4310272783528223e-05,
+      "loss": 0.593,
+      "step": 7249
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 40.30260630005622,
+      "learning_rate": 1.4308751018851872e-05,
+      "loss": 0.7716,
+      "step": 7250
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 33.78870343497123,
+      "learning_rate": 1.4307229131637163e-05,
+      "loss": 0.6282,
+      "step": 7251
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 43.928899311211055,
+      "learning_rate": 1.4305707121927377e-05,
+      "loss": 0.7295,
+      "step": 7252
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 51.828912700823395,
+      "learning_rate": 1.43041849897658e-05,
+      "loss": 0.7682,
+      "step": 7253
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 49.086021218053155,
+      "learning_rate": 1.4302662735195717e-05,
+      "loss": 0.9048,
+      "step": 7254
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.46704390871641,
+      "learning_rate": 1.4301140358260426e-05,
+      "loss": 0.7276,
+      "step": 7255
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 39.82662030172847,
+      "learning_rate": 1.4299617859003218e-05,
+      "loss": 0.6928,
+      "step": 7256
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 34.63408790200226,
+      "learning_rate": 1.4298095237467394e-05,
+      "loss": 0.6872,
+      "step": 7257
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 73.26896546132109,
+      "learning_rate": 1.4296572493696255e-05,
+      "loss": 0.6765,
+      "step": 7258
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 30.989516335209245,
+      "learning_rate": 1.4295049627733107e-05,
+      "loss": 0.5973,
+      "step": 7259
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 29.42063079893571,
+      "learning_rate": 1.4293526639621262e-05,
+      "loss": 0.6718,
+      "step": 7260
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 36.798603880614934,
+      "learning_rate": 1.4292003529404028e-05,
+      "loss": 0.6659,
+      "step": 7261
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.52629533408623,
+      "learning_rate": 1.4290480297124726e-05,
+      "loss": 0.6254,
+      "step": 7262
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 34.986905792276744,
+      "learning_rate": 1.4288956942826674e-05,
+      "loss": 0.677,
+      "step": 7263
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 40.6206641949886,
+      "learning_rate": 1.4287433466553198e-05,
+      "loss": 0.6932,
+      "step": 7264
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 41.496559000410045,
+      "learning_rate": 1.4285909868347621e-05,
+      "loss": 0.6485,
+      "step": 7265
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 35.09888360239657,
+      "learning_rate": 1.4284386148253273e-05,
+      "loss": 0.6653,
+      "step": 7266
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 35.75729621112763,
+      "learning_rate": 1.4282862306313488e-05,
+      "loss": 0.7166,
+      "step": 7267
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 33.91803346140302,
+      "learning_rate": 1.4281338342571609e-05,
+      "loss": 0.7289,
+      "step": 7268
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.25780392803075,
+      "learning_rate": 1.4279814257070967e-05,
+      "loss": 0.7058,
+      "step": 7269
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.34394838576358,
+      "learning_rate": 1.4278290049854917e-05,
+      "loss": 0.7072,
+      "step": 7270
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 31.429906439913502,
+      "learning_rate": 1.4276765720966797e-05,
+      "loss": 0.6978,
+      "step": 7271
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.92028591261228,
+      "learning_rate": 1.4275241270449962e-05,
+      "loss": 0.76,
+      "step": 7272
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 36.61246330595408,
+      "learning_rate": 1.4273716698347766e-05,
+      "loss": 0.7149,
+      "step": 7273
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 37.14168653307178,
+      "learning_rate": 1.4272192004703569e-05,
+      "loss": 0.8262,
+      "step": 7274
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 49.255162709151044,
+      "learning_rate": 1.4270667189560727e-05,
+      "loss": 0.7742,
+      "step": 7275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.050004533524316,
+      "learning_rate": 1.426914225296261e-05,
+      "loss": 0.6685,
+      "step": 7276
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 35.36115045679312,
+      "learning_rate": 1.4267617194952588e-05,
+      "loss": 0.6827,
+      "step": 7277
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.092200746073374,
+      "learning_rate": 1.4266092015574027e-05,
+      "loss": 0.685,
+      "step": 7278
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.468481482801224,
+      "learning_rate": 1.4264566714870301e-05,
+      "loss": 0.7346,
+      "step": 7279
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 26.58300515548285,
+      "learning_rate": 1.4263041292884795e-05,
+      "loss": 0.6185,
+      "step": 7280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 41.849999881958595,
+      "learning_rate": 1.426151574966089e-05,
+      "loss": 0.7222,
+      "step": 7281
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.478063353104424,
+      "learning_rate": 1.425999008524197e-05,
+      "loss": 0.7916,
+      "step": 7282
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.4637280192455,
+      "learning_rate": 1.4258464299671425e-05,
+      "loss": 0.7142,
+      "step": 7283
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 35.393593117583315,
+      "learning_rate": 1.4256938392992641e-05,
+      "loss": 0.717,
+      "step": 7284
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.50902059701758,
+      "learning_rate": 1.4255412365249027e-05,
+      "loss": 0.7031,
+      "step": 7285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.67209730615426,
+      "learning_rate": 1.4253886216483968e-05,
+      "loss": 0.6303,
+      "step": 7286
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 31.45291954765549,
+      "learning_rate": 1.4252359946740877e-05,
+      "loss": 0.612,
+      "step": 7287
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 39.61485633318021,
+      "learning_rate": 1.4250833556063155e-05,
+      "loss": 0.6597,
+      "step": 7288
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 33.029524625469286,
+      "learning_rate": 1.4249307044494214e-05,
+      "loss": 0.7589,
+      "step": 7289
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 39.364241283687434,
+      "learning_rate": 1.4247780412077466e-05,
+      "loss": 0.7159,
+      "step": 7290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 35.778950666485876,
+      "learning_rate": 1.4246253658856329e-05,
+      "loss": 0.7122,
+      "step": 7291
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.934516078488635,
+      "learning_rate": 1.4244726784874221e-05,
+      "loss": 0.6923,
+      "step": 7292
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 28.951471494591043,
+      "learning_rate": 1.4243199790174566e-05,
+      "loss": 0.6613,
+      "step": 7293
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.72096156222865,
+      "learning_rate": 1.4241672674800791e-05,
+      "loss": 0.7478,
+      "step": 7294
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 44.68930690020065,
+      "learning_rate": 1.4240145438796329e-05,
+      "loss": 0.7447,
+      "step": 7295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.790300612652395,
+      "learning_rate": 1.4238618082204609e-05,
+      "loss": 0.7003,
+      "step": 7296
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 31.689803282415163,
+      "learning_rate": 1.4237090605069072e-05,
+      "loss": 0.6792,
+      "step": 7297
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 31.640502669486462,
+      "learning_rate": 1.4235563007433153e-05,
+      "loss": 0.6633,
+      "step": 7298
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 37.63813458696057,
+      "learning_rate": 1.4234035289340304e-05,
+      "loss": 0.7123,
+      "step": 7299
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.4772038576249,
+      "learning_rate": 1.4232507450833966e-05,
+      "loss": 0.6874,
+      "step": 7300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 50.61184867709249,
+      "learning_rate": 1.4230979491957592e-05,
+      "loss": 0.8465,
+      "step": 7301
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 47.53367196257156,
+      "learning_rate": 1.4229451412754638e-05,
+      "loss": 0.6693,
+      "step": 7302
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.98902961419632,
+      "learning_rate": 1.4227923213268557e-05,
+      "loss": 0.6818,
+      "step": 7303
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 33.74977425393795,
+      "learning_rate": 1.4226394893542817e-05,
+      "loss": 0.7808,
+      "step": 7304
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 41.17632635038221,
+      "learning_rate": 1.4224866453620875e-05,
+      "loss": 0.7434,
+      "step": 7305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 45.57960130708813,
+      "learning_rate": 1.4223337893546206e-05,
+      "loss": 0.6771,
+      "step": 7306
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 46.508566836329315,
+      "learning_rate": 1.4221809213362275e-05,
+      "loss": 0.7491,
+      "step": 7307
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 41.71071596260063,
+      "learning_rate": 1.422028041311256e-05,
+      "loss": 0.7939,
+      "step": 7308
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 39.81150700455146,
+      "learning_rate": 1.421875149284054e-05,
+      "loss": 0.753,
+      "step": 7309
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 41.07640843162096,
+      "learning_rate": 1.4217222452589694e-05,
+      "loss": 0.6724,
+      "step": 7310
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.27430807468383,
+      "learning_rate": 1.4215693292403508e-05,
+      "loss": 0.6841,
+      "step": 7311
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 57.47311588068041,
+      "learning_rate": 1.4214164012325475e-05,
+      "loss": 0.698,
+      "step": 7312
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 36.88020324204188,
+      "learning_rate": 1.421263461239908e-05,
+      "loss": 0.7505,
+      "step": 7313
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 51.24149455692315,
+      "learning_rate": 1.421110509266782e-05,
+      "loss": 0.7679,
+      "step": 7314
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.47814093326679,
+      "learning_rate": 1.4209575453175195e-05,
+      "loss": 0.5968,
+      "step": 7315
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.93220768938293,
+      "learning_rate": 1.4208045693964707e-05,
+      "loss": 0.6457,
+      "step": 7316
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 42.754496945778804,
+      "learning_rate": 1.4206515815079862e-05,
+      "loss": 0.6778,
+      "step": 7317
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 31.000652604386413,
+      "learning_rate": 1.4204985816564167e-05,
+      "loss": 0.5926,
+      "step": 7318
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 29.45914379112906,
+      "learning_rate": 1.4203455698461135e-05,
+      "loss": 0.5812,
+      "step": 7319
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.832165450765615,
+      "learning_rate": 1.4201925460814282e-05,
+      "loss": 0.7377,
+      "step": 7320
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 37.903573173611434,
+      "learning_rate": 1.4200395103667126e-05,
+      "loss": 0.7553,
+      "step": 7321
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.15276781214283,
+      "learning_rate": 1.4198864627063194e-05,
+      "loss": 0.6737,
+      "step": 7322
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 35.083705813287345,
+      "learning_rate": 1.4197334031046004e-05,
+      "loss": 0.7487,
+      "step": 7323
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.29083790227623,
+      "learning_rate": 1.4195803315659092e-05,
+      "loss": 0.6526,
+      "step": 7324
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 32.28670475194934,
+      "learning_rate": 1.4194272480945987e-05,
+      "loss": 0.7102,
+      "step": 7325
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 38.08733071933407,
+      "learning_rate": 1.4192741526950226e-05,
+      "loss": 0.7394,
+      "step": 7326
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 36.94728525484575,
+      "learning_rate": 1.4191210453715351e-05,
+      "loss": 0.7501,
+      "step": 7327
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 36.259374563921874,
+      "learning_rate": 1.4189679261284899e-05,
+      "loss": 0.8544,
+      "step": 7328
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 37.22792052987186,
+      "learning_rate": 1.4188147949702425e-05,
+      "loss": 0.6459,
+      "step": 7329
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 28.375918411314505,
+      "learning_rate": 1.4186616519011467e-05,
+      "loss": 0.6402,
+      "step": 7330
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 39.242140487226564,
+      "learning_rate": 1.4185084969255589e-05,
+      "loss": 0.8012,
+      "step": 7331
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.08745581886665,
+      "learning_rate": 1.4183553300478339e-05,
+      "loss": 0.6958,
+      "step": 7332
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 36.97501517225242,
+      "learning_rate": 1.4182021512723283e-05,
+      "loss": 0.6862,
+      "step": 7333
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.13265311884847,
+      "learning_rate": 1.418048960603398e-05,
+      "loss": 0.6687,
+      "step": 7334
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.224221021264775,
+      "learning_rate": 1.4178957580454e-05,
+      "loss": 0.696,
+      "step": 7335
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 38.307878504895136,
+      "learning_rate": 1.417742543602691e-05,
+      "loss": 0.781,
+      "step": 7336
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 37.25207388308332,
+      "learning_rate": 1.4175893172796285e-05,
+      "loss": 0.7026,
+      "step": 7337
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.543538309630364,
+      "learning_rate": 1.4174360790805699e-05,
+      "loss": 0.7071,
+      "step": 7338
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.64080578037417,
+      "learning_rate": 1.417282829009874e-05,
+      "loss": 0.7279,
+      "step": 7339
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.35554655945481,
+      "learning_rate": 1.4171295670718976e-05,
+      "loss": 0.6693,
+      "step": 7340
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.83734442376009,
+      "learning_rate": 1.4169762932710012e-05,
+      "loss": 0.727,
+      "step": 7341
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 56.63757450702509,
+      "learning_rate": 1.4168230076115424e-05,
+      "loss": 0.7174,
+      "step": 7342
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 28.928848645902043,
+      "learning_rate": 1.4166697100978812e-05,
+      "loss": 0.6046,
+      "step": 7343
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.094179574831166,
+      "learning_rate": 1.4165164007343772e-05,
+      "loss": 0.6791,
+      "step": 7344
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 37.08491681536422,
+      "learning_rate": 1.4163630795253904e-05,
+      "loss": 0.7474,
+      "step": 7345
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 51.264991302898764,
+      "learning_rate": 1.4162097464752815e-05,
+      "loss": 0.6689,
+      "step": 7346
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.41505682858156,
+      "learning_rate": 1.4160564015884103e-05,
+      "loss": 0.646,
+      "step": 7347
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.649251680066996,
+      "learning_rate": 1.415903044869139e-05,
+      "loss": 0.6203,
+      "step": 7348
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.25614506151546,
+      "learning_rate": 1.4157496763218278e-05,
+      "loss": 0.7348,
+      "step": 7349
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.331182738108254,
+      "learning_rate": 1.4155962959508396e-05,
+      "loss": 0.6235,
+      "step": 7350
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.8218209206743,
+      "learning_rate": 1.4154429037605358e-05,
+      "loss": 0.6305,
+      "step": 7351
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.920114057653215,
+      "learning_rate": 1.4152894997552786e-05,
+      "loss": 0.7766,
+      "step": 7352
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 36.2281018214537,
+      "learning_rate": 1.415136083939431e-05,
+      "loss": 0.6746,
+      "step": 7353
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.06886334398385,
+      "learning_rate": 1.414982656317356e-05,
+      "loss": 0.7331,
+      "step": 7354
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 30.525715637722296,
+      "learning_rate": 1.414829216893417e-05,
+      "loss": 0.5496,
+      "step": 7355
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 41.987631889742566,
+      "learning_rate": 1.4146757656719781e-05,
+      "loss": 0.6197,
+      "step": 7356
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 44.508280080194304,
+      "learning_rate": 1.4145223026574027e-05,
+      "loss": 0.6769,
+      "step": 7357
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.1929455231122,
+      "learning_rate": 1.4143688278540554e-05,
+      "loss": 0.7054,
+      "step": 7358
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 43.608903511256756,
+      "learning_rate": 1.4142153412663012e-05,
+      "loss": 0.7035,
+      "step": 7359
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 56.358799021130565,
+      "learning_rate": 1.414061842898505e-05,
+      "loss": 0.7505,
+      "step": 7360
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 28.909655049791954,
+      "learning_rate": 1.413908332755032e-05,
+      "loss": 0.6817,
+      "step": 7361
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.97855108761065,
+      "learning_rate": 1.4137548108402483e-05,
+      "loss": 0.7704,
+      "step": 7362
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.969263995860715,
+      "learning_rate": 1.41360127715852e-05,
+      "loss": 0.749,
+      "step": 7363
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.898118119478845,
+      "learning_rate": 1.4134477317142133e-05,
+      "loss": 0.661,
+      "step": 7364
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.89656732764665,
+      "learning_rate": 1.4132941745116946e-05,
+      "loss": 0.696,
+      "step": 7365
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.167747035429656,
+      "learning_rate": 1.4131406055553316e-05,
+      "loss": 0.7592,
+      "step": 7366
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.40695652793906,
+      "learning_rate": 1.4129870248494913e-05,
+      "loss": 0.6497,
+      "step": 7367
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.32020737757932,
+      "learning_rate": 1.4128334323985418e-05,
+      "loss": 0.6854,
+      "step": 7368
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 31.394082098815677,
+      "learning_rate": 1.412679828206851e-05,
+      "loss": 0.6273,
+      "step": 7369
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.37180979056875,
+      "learning_rate": 1.4125262122787872e-05,
+      "loss": 0.6248,
+      "step": 7370
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 44.82338638762848,
+      "learning_rate": 1.4123725846187193e-05,
+      "loss": 0.8117,
+      "step": 7371
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 36.76809430209654,
+      "learning_rate": 1.412218945231016e-05,
+      "loss": 0.6459,
+      "step": 7372
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.63500798875825,
+      "learning_rate": 1.4120652941200477e-05,
+      "loss": 0.6683,
+      "step": 7373
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.640062606187904,
+      "learning_rate": 1.4119116312901828e-05,
+      "loss": 0.732,
+      "step": 7374
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 65.20091662676774,
+      "learning_rate": 1.4117579567457927e-05,
+      "loss": 0.7993,
+      "step": 7375
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 42.03102139265436,
+      "learning_rate": 1.4116042704912465e-05,
+      "loss": 0.6543,
+      "step": 7376
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.20676785587124,
+      "learning_rate": 1.411450572530916e-05,
+      "loss": 0.6029,
+      "step": 7377
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.150468574868015,
+      "learning_rate": 1.411296862869172e-05,
+      "loss": 0.859,
+      "step": 7378
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.83260284053353,
+      "learning_rate": 1.4111431415103858e-05,
+      "loss": 0.7517,
+      "step": 7379
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 45.73549840216616,
+      "learning_rate": 1.4109894084589291e-05,
+      "loss": 0.7531,
+      "step": 7380
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.58428932166897,
+      "learning_rate": 1.410835663719174e-05,
+      "loss": 0.7459,
+      "step": 7381
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 37.94457076543319,
+      "learning_rate": 1.410681907295493e-05,
+      "loss": 0.7099,
+      "step": 7382
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 52.786843919843456,
+      "learning_rate": 1.410528139192259e-05,
+      "loss": 0.7445,
+      "step": 7383
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 50.6756729928162,
+      "learning_rate": 1.4103743594138443e-05,
+      "loss": 0.8064,
+      "step": 7384
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 42.64720749920044,
+      "learning_rate": 1.4102205679646236e-05,
+      "loss": 0.7813,
+      "step": 7385
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 38.98508726703935,
+      "learning_rate": 1.4100667648489692e-05,
+      "loss": 0.6728,
+      "step": 7386
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 45.11887668370424,
+      "learning_rate": 1.4099129500712562e-05,
+      "loss": 0.6932,
+      "step": 7387
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 45.47877819992192,
+      "learning_rate": 1.4097591236358588e-05,
+      "loss": 0.7152,
+      "step": 7388
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 29.73095347874287,
+      "learning_rate": 1.4096052855471519e-05,
+      "loss": 0.6906,
+      "step": 7389
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 37.39443489938164,
+      "learning_rate": 1.4094514358095096e-05,
+      "loss": 0.7179,
+      "step": 7390
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 31.748597035236468,
+      "learning_rate": 1.4092975744273082e-05,
+      "loss": 0.6792,
+      "step": 7391
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 41.48337795977544,
+      "learning_rate": 1.4091437014049234e-05,
+      "loss": 0.7554,
+      "step": 7392
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.118477881026884,
+      "learning_rate": 1.408989816746731e-05,
+      "loss": 0.6918,
+      "step": 7393
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 36.2024925351206,
+      "learning_rate": 1.4088359204571076e-05,
+      "loss": 0.7202,
+      "step": 7394
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 29.058141049236703,
+      "learning_rate": 1.4086820125404299e-05,
+      "loss": 0.6666,
+      "step": 7395
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 52.33606522814022,
+      "learning_rate": 1.4085280930010745e-05,
+      "loss": 0.7474,
+      "step": 7396
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.289627947882906,
+      "learning_rate": 1.4083741618434192e-05,
+      "loss": 0.6506,
+      "step": 7397
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 34.124917187433056,
+      "learning_rate": 1.4082202190718417e-05,
+      "loss": 0.654,
+      "step": 7398
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.86117346845404,
+      "learning_rate": 1.40806626469072e-05,
+      "loss": 0.7208,
+      "step": 7399
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 43.77384764713893,
+      "learning_rate": 1.4079122987044324e-05,
+      "loss": 0.6652,
+      "step": 7400
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.96044202419583,
+      "learning_rate": 1.4077583211173575e-05,
+      "loss": 0.743,
+      "step": 7401
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 31.127893014684282,
+      "learning_rate": 1.4076043319338748e-05,
+      "loss": 0.6516,
+      "step": 7402
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.80185311712118,
+      "learning_rate": 1.407450331158363e-05,
+      "loss": 0.7536,
+      "step": 7403
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 39.687052836865185,
+      "learning_rate": 1.4072963187952023e-05,
+      "loss": 0.7175,
+      "step": 7404
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.091487458795385,
+      "learning_rate": 1.4071422948487725e-05,
+      "loss": 0.727,
+      "step": 7405
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.590946642416974,
+      "learning_rate": 1.406988259323454e-05,
+      "loss": 0.6949,
+      "step": 7406
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 34.36882156334097,
+      "learning_rate": 1.4068342122236275e-05,
+      "loss": 0.6554,
+      "step": 7407
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 44.235931199328576,
+      "learning_rate": 1.406680153553674e-05,
+      "loss": 0.7107,
+      "step": 7408
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 39.49923666083673,
+      "learning_rate": 1.4065260833179748e-05,
+      "loss": 0.7406,
+      "step": 7409
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.40377183065335,
+      "learning_rate": 1.4063720015209117e-05,
+      "loss": 0.6832,
+      "step": 7410
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.70682497476775,
+      "learning_rate": 1.4062179081668665e-05,
+      "loss": 0.7551,
+      "step": 7411
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 41.11983580471164,
+      "learning_rate": 1.4060638032602218e-05,
+      "loss": 0.6627,
+      "step": 7412
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.434999766096084,
+      "learning_rate": 1.4059096868053601e-05,
+      "loss": 0.743,
+      "step": 7413
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 34.20358961965203,
+      "learning_rate": 1.405755558806664e-05,
+      "loss": 0.636,
+      "step": 7414
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 42.337458998348,
+      "learning_rate": 1.4056014192685175e-05,
+      "loss": 0.7269,
+      "step": 7415
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 41.816380079850084,
+      "learning_rate": 1.4054472681953035e-05,
+      "loss": 0.7223,
+      "step": 7416
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 35.76356330418863,
+      "learning_rate": 1.405293105591407e-05,
+      "loss": 0.6656,
+      "step": 7417
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.30081887698914,
+      "learning_rate": 1.4051389314612112e-05,
+      "loss": 0.6541,
+      "step": 7418
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.54263868088439,
+      "learning_rate": 1.4049847458091014e-05,
+      "loss": 0.7495,
+      "step": 7419
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 43.284779913555255,
+      "learning_rate": 1.404830548639462e-05,
+      "loss": 0.657,
+      "step": 7420
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 29.574154538462913,
+      "learning_rate": 1.404676339956679e-05,
+      "loss": 0.6328,
+      "step": 7421
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.717236413073834,
+      "learning_rate": 1.4045221197651375e-05,
+      "loss": 0.7181,
+      "step": 7422
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.666474950915784,
+      "learning_rate": 1.4043678880692235e-05,
+      "loss": 0.681,
+      "step": 7423
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 40.0729361047277,
+      "learning_rate": 1.4042136448733235e-05,
+      "loss": 0.7502,
+      "step": 7424
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 45.85144622210814,
+      "learning_rate": 1.4040593901818239e-05,
+      "loss": 0.6212,
+      "step": 7425
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.526713573383105,
+      "learning_rate": 1.403905123999111e-05,
+      "loss": 0.6701,
+      "step": 7426
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 42.02982967770592,
+      "learning_rate": 1.4037508463295735e-05,
+      "loss": 0.7074,
+      "step": 7427
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 34.53883409715817,
+      "learning_rate": 1.4035965571775976e-05,
+      "loss": 0.7053,
+      "step": 7428
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.70742063775976,
+      "learning_rate": 1.403442256547572e-05,
+      "loss": 0.7257,
+      "step": 7429
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.26184730175689,
+      "learning_rate": 1.4032879444438843e-05,
+      "loss": 0.7148,
+      "step": 7430
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 40.62158868698897,
+      "learning_rate": 1.4031336208709236e-05,
+      "loss": 0.6872,
+      "step": 7431
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 39.13161136598648,
+      "learning_rate": 1.4029792858330783e-05,
+      "loss": 0.7013,
+      "step": 7432
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.46748116450701,
+      "learning_rate": 1.4028249393347382e-05,
+      "loss": 0.69,
+      "step": 7433
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.57828258112957,
+      "learning_rate": 1.4026705813802923e-05,
+      "loss": 0.7612,
+      "step": 7434
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 41.46208564256845,
+      "learning_rate": 1.4025162119741304e-05,
+      "loss": 0.6598,
+      "step": 7435
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.92302449236634,
+      "learning_rate": 1.4023618311206432e-05,
+      "loss": 0.7064,
+      "step": 7436
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 52.97648059167509,
+      "learning_rate": 1.4022074388242208e-05,
+      "loss": 0.7435,
+      "step": 7437
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 34.34603369897846,
+      "learning_rate": 1.4020530350892542e-05,
+      "loss": 0.6461,
+      "step": 7438
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 39.182068580696516,
+      "learning_rate": 1.4018986199201345e-05,
+      "loss": 0.819,
+      "step": 7439
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.23921760175371,
+      "learning_rate": 1.4017441933212532e-05,
+      "loss": 0.6018,
+      "step": 7440
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 40.191050756756276,
+      "learning_rate": 1.4015897552970017e-05,
+      "loss": 0.8217,
+      "step": 7441
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.25763758504119,
+      "learning_rate": 1.401435305851773e-05,
+      "loss": 0.715,
+      "step": 7442
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 41.60795727791671,
+      "learning_rate": 1.4012808449899584e-05,
+      "loss": 0.7378,
+      "step": 7443
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 31.770759475288635,
+      "learning_rate": 1.4011263727159521e-05,
+      "loss": 0.705,
+      "step": 7444
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 43.61107758738747,
+      "learning_rate": 1.4009718890341457e-05,
+      "loss": 0.7616,
+      "step": 7445
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 31.267548918926366,
+      "learning_rate": 1.4008173939489338e-05,
+      "loss": 0.6806,
+      "step": 7446
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.71878787293633,
+      "learning_rate": 1.4006628874647094e-05,
+      "loss": 0.7449,
+      "step": 7447
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 35.437014480574945,
+      "learning_rate": 1.400508369585867e-05,
+      "loss": 0.7842,
+      "step": 7448
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.074289397172876,
+      "learning_rate": 1.400353840316801e-05,
+      "loss": 0.7373,
+      "step": 7449
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 35.305269591497876,
+      "learning_rate": 1.4001992996619056e-05,
+      "loss": 0.6979,
+      "step": 7450
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.47687829576025,
+      "learning_rate": 1.4000447476255765e-05,
+      "loss": 0.6527,
+      "step": 7451
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 39.72889041887942,
+      "learning_rate": 1.3998901842122088e-05,
+      "loss": 0.8695,
+      "step": 7452
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.00036263606557,
+      "learning_rate": 1.3997356094261977e-05,
+      "loss": 0.7826,
+      "step": 7453
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.489939619381836,
+      "learning_rate": 1.3995810232719405e-05,
+      "loss": 0.6348,
+      "step": 7454
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 32.445834120064625,
+      "learning_rate": 1.3994264257538324e-05,
+      "loss": 0.6564,
+      "step": 7455
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 42.9142783892382,
+      "learning_rate": 1.3992718168762702e-05,
+      "loss": 0.7322,
+      "step": 7456
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 40.32990859535164,
+      "learning_rate": 1.3991171966436513e-05,
+      "loss": 0.7533,
+      "step": 7457
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 42.331369269358035,
+      "learning_rate": 1.3989625650603729e-05,
+      "loss": 0.7698,
+      "step": 7458
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.30772564934131,
+      "learning_rate": 1.3988079221308323e-05,
+      "loss": 0.7039,
+      "step": 7459
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.927169585544,
+      "learning_rate": 1.3986532678594277e-05,
+      "loss": 0.7238,
+      "step": 7460
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.15355126967723,
+      "learning_rate": 1.3984986022505578e-05,
+      "loss": 0.7433,
+      "step": 7461
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.475467572848494,
+      "learning_rate": 1.3983439253086201e-05,
+      "loss": 0.7681,
+      "step": 7462
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 39.67337554341475,
+      "learning_rate": 1.3981892370380146e-05,
+      "loss": 0.6951,
+      "step": 7463
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 41.878384979991885,
+      "learning_rate": 1.3980345374431401e-05,
+      "loss": 0.6677,
+      "step": 7464
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 40.53827992152756,
+      "learning_rate": 1.3978798265283962e-05,
+      "loss": 0.7441,
+      "step": 7465
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 41.3127660930334,
+      "learning_rate": 1.3977251042981828e-05,
+      "loss": 0.7395,
+      "step": 7466
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 31.321697940000558,
+      "learning_rate": 1.3975703707569e-05,
+      "loss": 0.6748,
+      "step": 7467
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 44.866514814310214,
+      "learning_rate": 1.3974156259089486e-05,
+      "loss": 0.6229,
+      "step": 7468
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.05237887592361,
+      "learning_rate": 1.3972608697587292e-05,
+      "loss": 0.6214,
+      "step": 7469
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 31.754886536513375,
+      "learning_rate": 1.3971061023106428e-05,
+      "loss": 0.6532,
+      "step": 7470
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.55119320073777,
+      "learning_rate": 1.3969513235690915e-05,
+      "loss": 0.7947,
+      "step": 7471
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.28814963966702,
+      "learning_rate": 1.3967965335384766e-05,
+      "loss": 0.676,
+      "step": 7472
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.228856925266456,
+      "learning_rate": 1.3966417322232006e-05,
+      "loss": 0.6685,
+      "step": 7473
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.30101979082109,
+      "learning_rate": 1.3964869196276654e-05,
+      "loss": 0.6665,
+      "step": 7474
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.687613460186526,
+      "learning_rate": 1.3963320957562743e-05,
+      "loss": 0.6204,
+      "step": 7475
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 33.58883658029388,
+      "learning_rate": 1.3961772606134303e-05,
+      "loss": 0.7628,
+      "step": 7476
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.61934079788923,
+      "learning_rate": 1.3960224142035368e-05,
+      "loss": 0.7339,
+      "step": 7477
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 43.056440641302586,
+      "learning_rate": 1.3958675565309974e-05,
+      "loss": 0.8247,
+      "step": 7478
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.587667071020974,
+      "learning_rate": 1.3957126876002164e-05,
+      "loss": 0.7107,
+      "step": 7479
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 34.5894342709498,
+      "learning_rate": 1.3955578074155978e-05,
+      "loss": 0.6942,
+      "step": 7480
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.1807951156582,
+      "learning_rate": 1.3954029159815468e-05,
+      "loss": 0.6839,
+      "step": 7481
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.29329269458678,
+      "learning_rate": 1.3952480133024682e-05,
+      "loss": 0.6836,
+      "step": 7482
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 45.05325451631379,
+      "learning_rate": 1.3950930993827671e-05,
+      "loss": 0.7679,
+      "step": 7483
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.66263146762302,
+      "learning_rate": 1.3949381742268496e-05,
+      "loss": 0.7771,
+      "step": 7484
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.14187299975265,
+      "learning_rate": 1.394783237839121e-05,
+      "loss": 0.7138,
+      "step": 7485
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 50.3262484892582,
+      "learning_rate": 1.3946282902239886e-05,
+      "loss": 0.7854,
+      "step": 7486
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 31.11523650898913,
+      "learning_rate": 1.3944733313858583e-05,
+      "loss": 0.6885,
+      "step": 7487
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 34.074037536747966,
+      "learning_rate": 1.3943183613291374e-05,
+      "loss": 0.6458,
+      "step": 7488
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 33.47289157891104,
+      "learning_rate": 1.3941633800582325e-05,
+      "loss": 0.758,
+      "step": 7489
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 55.34458935627225,
+      "learning_rate": 1.394008387577552e-05,
+      "loss": 0.6763,
+      "step": 7490
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.97445625156978,
+      "learning_rate": 1.3938533838915033e-05,
+      "loss": 0.6282,
+      "step": 7491
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.21149810273706,
+      "learning_rate": 1.3936983690044949e-05,
+      "loss": 0.733,
+      "step": 7492
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 29.23553207203607,
+      "learning_rate": 1.3935433429209352e-05,
+      "loss": 0.6703,
+      "step": 7493
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 54.89047295049537,
+      "learning_rate": 1.3933883056452331e-05,
+      "loss": 0.7477,
+      "step": 7494
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.52488707934941,
+      "learning_rate": 1.3932332571817975e-05,
+      "loss": 0.7149,
+      "step": 7495
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.35356946106225,
+      "learning_rate": 1.3930781975350382e-05,
+      "loss": 0.7649,
+      "step": 7496
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.99863443867728,
+      "learning_rate": 1.3929231267093646e-05,
+      "loss": 0.7923,
+      "step": 7497
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.7424030723469,
+      "learning_rate": 1.3927680447091876e-05,
+      "loss": 0.663,
+      "step": 7498
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.81367092731976,
+      "learning_rate": 1.3926129515389168e-05,
+      "loss": 0.617,
+      "step": 7499
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.21470378772707,
+      "learning_rate": 1.3924578472029637e-05,
+      "loss": 0.6964,
+      "step": 7500
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.308867431537735,
+      "learning_rate": 1.3923027317057388e-05,
+      "loss": 0.6222,
+      "step": 7501
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.88167187019374,
+      "learning_rate": 1.3921476050516538e-05,
+      "loss": 0.7745,
+      "step": 7502
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.32811124655986,
+      "learning_rate": 1.3919924672451201e-05,
+      "loss": 0.6859,
+      "step": 7503
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.897132273039155,
+      "learning_rate": 1.3918373182905501e-05,
+      "loss": 0.6787,
+      "step": 7504
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 48.12754589629298,
+      "learning_rate": 1.391682158192356e-05,
+      "loss": 0.681,
+      "step": 7505
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 34.781890568485444,
+      "learning_rate": 1.3915269869549504e-05,
+      "loss": 0.6608,
+      "step": 7506
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 41.439269446658884,
+      "learning_rate": 1.3913718045827462e-05,
+      "loss": 0.7273,
+      "step": 7507
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 34.78015242778737,
+      "learning_rate": 1.391216611080157e-05,
+      "loss": 0.7165,
+      "step": 7508
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 57.69363835617072,
+      "learning_rate": 1.3910614064515964e-05,
+      "loss": 0.7566,
+      "step": 7509
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 41.69297247315923,
+      "learning_rate": 1.3909061907014781e-05,
+      "loss": 0.7094,
+      "step": 7510
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.884923699040904,
+      "learning_rate": 1.390750963834216e-05,
+      "loss": 0.6445,
+      "step": 7511
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.08967910486576,
+      "learning_rate": 1.3905957258542253e-05,
+      "loss": 0.6927,
+      "step": 7512
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 40.15051730607717,
+      "learning_rate": 1.390440476765921e-05,
+      "loss": 0.6961,
+      "step": 7513
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.98961798536389,
+      "learning_rate": 1.3902852165737172e-05,
+      "loss": 0.6938,
+      "step": 7514
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.62847301661816,
+      "learning_rate": 1.390129945282031e-05,
+      "loss": 0.6966,
+      "step": 7515
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 39.39615081499145,
+      "learning_rate": 1.3899746628952766e-05,
+      "loss": 0.7706,
+      "step": 7516
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 48.37118465449237,
+      "learning_rate": 1.3898193694178714e-05,
+      "loss": 0.7032,
+      "step": 7517
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.500179011512145,
+      "learning_rate": 1.3896640648542312e-05,
+      "loss": 0.6263,
+      "step": 7518
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 31.580797119184954,
+      "learning_rate": 1.389508749208773e-05,
+      "loss": 0.671,
+      "step": 7519
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.51568691654857,
+      "learning_rate": 1.389353422485914e-05,
+      "loss": 0.6246,
+      "step": 7520
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.345738705601256,
+      "learning_rate": 1.3891980846900712e-05,
+      "loss": 0.7444,
+      "step": 7521
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 57.25147672689302,
+      "learning_rate": 1.3890427358256626e-05,
+      "loss": 0.7835,
+      "step": 7522
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.778186916831366,
+      "learning_rate": 1.388887375897106e-05,
+      "loss": 0.7182,
+      "step": 7523
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 34.946631052140845,
+      "learning_rate": 1.3887320049088202e-05,
+      "loss": 0.6548,
+      "step": 7524
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.41979573949973,
+      "learning_rate": 1.3885766228652235e-05,
+      "loss": 0.7314,
+      "step": 7525
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.61351658637805,
+      "learning_rate": 1.3884212297707349e-05,
+      "loss": 0.6993,
+      "step": 7526
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 44.30554611135533,
+      "learning_rate": 1.3882658256297738e-05,
+      "loss": 0.8039,
+      "step": 7527
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.04108115916555,
+      "learning_rate": 1.3881104104467598e-05,
+      "loss": 0.6866,
+      "step": 7528
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 41.47662740369279,
+      "learning_rate": 1.3879549842261123e-05,
+      "loss": 0.7075,
+      "step": 7529
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.18514824507306,
+      "learning_rate": 1.3877995469722527e-05,
+      "loss": 0.6896,
+      "step": 7530
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 34.8442520037014,
+      "learning_rate": 1.3876440986896004e-05,
+      "loss": 0.7483,
+      "step": 7531
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 42.71615602756385,
+      "learning_rate": 1.387488639382577e-05,
+      "loss": 0.6929,
+      "step": 7532
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 40.663220634137495,
+      "learning_rate": 1.387333169055603e-05,
+      "loss": 0.7146,
+      "step": 7533
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 39.69093951241907,
+      "learning_rate": 1.3871776877131007e-05,
+      "loss": 0.6344,
+      "step": 7534
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.90918623480406,
+      "learning_rate": 1.3870221953594912e-05,
+      "loss": 0.6949,
+      "step": 7535
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.79066857550607,
+      "learning_rate": 1.3868666919991966e-05,
+      "loss": 0.7337,
+      "step": 7536
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.81551420965018,
+      "learning_rate": 1.38671117763664e-05,
+      "loss": 0.6923,
+      "step": 7537
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 40.251610794769924,
+      "learning_rate": 1.3865556522762438e-05,
+      "loss": 0.7085,
+      "step": 7538
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 42.19109924786753,
+      "learning_rate": 1.3864001159224306e-05,
+      "loss": 0.8205,
+      "step": 7539
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.56213931381803,
+      "learning_rate": 1.3862445685796244e-05,
+      "loss": 0.6359,
+      "step": 7540
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.76571852070399,
+      "learning_rate": 1.3860890102522482e-05,
+      "loss": 0.8125,
+      "step": 7541
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 32.908482699531405,
+      "learning_rate": 1.385933440944727e-05,
+      "loss": 0.6532,
+      "step": 7542
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.24045908762394,
+      "learning_rate": 1.385777860661484e-05,
+      "loss": 0.7604,
+      "step": 7543
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 39.538242478564506,
+      "learning_rate": 1.3856222694069446e-05,
+      "loss": 0.7719,
+      "step": 7544
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.14381446074045,
+      "learning_rate": 1.3854666671855334e-05,
+      "loss": 0.7115,
+      "step": 7545
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 41.41922731114435,
+      "learning_rate": 1.3853110540016759e-05,
+      "loss": 0.7824,
+      "step": 7546
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 59.69075032915982,
+      "learning_rate": 1.3851554298597971e-05,
+      "loss": 0.7226,
+      "step": 7547
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 39.90207296569931,
+      "learning_rate": 1.384999794764323e-05,
+      "loss": 0.7033,
+      "step": 7548
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 45.04100718208049,
+      "learning_rate": 1.3848441487196804e-05,
+      "loss": 0.7704,
+      "step": 7549
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.96004037278766,
+      "learning_rate": 1.3846884917302952e-05,
+      "loss": 0.6536,
+      "step": 7550
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.03582461575856,
+      "learning_rate": 1.3845328238005942e-05,
+      "loss": 0.6902,
+      "step": 7551
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.671851301089646,
+      "learning_rate": 1.3843771449350047e-05,
+      "loss": 0.7432,
+      "step": 7552
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.798694552910778,
+      "learning_rate": 1.3842214551379542e-05,
+      "loss": 0.6871,
+      "step": 7553
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.521510968905304,
+      "learning_rate": 1.3840657544138702e-05,
+      "loss": 0.6579,
+      "step": 7554
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 29.537574312889067,
+      "learning_rate": 1.3839100427671808e-05,
+      "loss": 0.6591,
+      "step": 7555
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.302301153580146,
+      "learning_rate": 1.3837543202023143e-05,
+      "loss": 0.7171,
+      "step": 7556
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 44.30403056268957,
+      "learning_rate": 1.3835985867236995e-05,
+      "loss": 0.7855,
+      "step": 7557
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 42.92240099728698,
+      "learning_rate": 1.383442842335765e-05,
+      "loss": 0.7016,
+      "step": 7558
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.346041908321546,
+      "learning_rate": 1.383287087042941e-05,
+      "loss": 0.7399,
+      "step": 7559
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 50.39713820836202,
+      "learning_rate": 1.383131320849656e-05,
+      "loss": 0.6584,
+      "step": 7560
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.04135919115685,
+      "learning_rate": 1.3829755437603405e-05,
+      "loss": 0.6695,
+      "step": 7561
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 44.35629576731343,
+      "learning_rate": 1.3828197557794245e-05,
+      "loss": 0.6801,
+      "step": 7562
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.20275933840193,
+      "learning_rate": 1.3826639569113384e-05,
+      "loss": 0.7559,
+      "step": 7563
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 30.69521872956868,
+      "learning_rate": 1.3825081471605137e-05,
+      "loss": 0.5942,
+      "step": 7564
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.06977453943053,
+      "learning_rate": 1.3823523265313808e-05,
+      "loss": 0.6847,
+      "step": 7565
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.727470568710594,
+      "learning_rate": 1.3821964950283713e-05,
+      "loss": 0.7381,
+      "step": 7566
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 32.274752399638366,
+      "learning_rate": 1.3820406526559169e-05,
+      "loss": 0.7503,
+      "step": 7567
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 40.19925988687952,
+      "learning_rate": 1.38188479941845e-05,
+      "loss": 0.7822,
+      "step": 7568
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 37.06828041415972,
+      "learning_rate": 1.381728935320403e-05,
+      "loss": 0.7307,
+      "step": 7569
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 32.90291174957996,
+      "learning_rate": 1.3815730603662083e-05,
+      "loss": 0.737,
+      "step": 7570
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.5186359925378,
+      "learning_rate": 1.3814171745602991e-05,
+      "loss": 0.7488,
+      "step": 7571
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 44.134078161803544,
+      "learning_rate": 1.3812612779071084e-05,
+      "loss": 0.6979,
+      "step": 7572
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.866523947107474,
+      "learning_rate": 1.3811053704110697e-05,
+      "loss": 0.6264,
+      "step": 7573
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 45.11496855592525,
+      "learning_rate": 1.3809494520766179e-05,
+      "loss": 0.6747,
+      "step": 7574
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 32.93824918137412,
+      "learning_rate": 1.3807935229081859e-05,
+      "loss": 0.6097,
+      "step": 7575
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.017128312515915,
+      "learning_rate": 1.3806375829102092e-05,
+      "loss": 0.7551,
+      "step": 7576
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 30.64880253411306,
+      "learning_rate": 1.3804816320871221e-05,
+      "loss": 0.6918,
+      "step": 7577
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.335317015432885,
+      "learning_rate": 1.38032567044336e-05,
+      "loss": 0.7277,
+      "step": 7578
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.0949975660973,
+      "learning_rate": 1.3801696979833586e-05,
+      "loss": 0.7234,
+      "step": 7579
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 36.29122872020836,
+      "learning_rate": 1.3800137147115531e-05,
+      "loss": 0.7001,
+      "step": 7580
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 40.511282632599155,
+      "learning_rate": 1.37985772063238e-05,
+      "loss": 0.6716,
+      "step": 7581
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 42.07108948396527,
+      "learning_rate": 1.3797017157502754e-05,
+      "loss": 0.6948,
+      "step": 7582
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.530183613391564,
+      "learning_rate": 1.3795457000696762e-05,
+      "loss": 0.6976,
+      "step": 7583
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.093957044838106,
+      "learning_rate": 1.3793896735950195e-05,
+      "loss": 0.6889,
+      "step": 7584
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 38.78722631866253,
+      "learning_rate": 1.3792336363307419e-05,
+      "loss": 0.6654,
+      "step": 7585
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.689000167548382,
+      "learning_rate": 1.3790775882812818e-05,
+      "loss": 0.6213,
+      "step": 7586
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.70674526263149,
+      "learning_rate": 1.378921529451077e-05,
+      "loss": 0.674,
+      "step": 7587
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 39.585678608666804,
+      "learning_rate": 1.3787654598445656e-05,
+      "loss": 0.8273,
+      "step": 7588
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.33545239335055,
+      "learning_rate": 1.3786093794661858e-05,
+      "loss": 0.6907,
+      "step": 7589
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 42.02956737567939,
+      "learning_rate": 1.3784532883203769e-05,
+      "loss": 0.7362,
+      "step": 7590
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 41.3703686867916,
+      "learning_rate": 1.378297186411578e-05,
+      "loss": 0.7207,
+      "step": 7591
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 38.51723611491323,
+      "learning_rate": 1.378141073744228e-05,
+      "loss": 0.8053,
+      "step": 7592
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 44.46139993462306,
+      "learning_rate": 1.3779849503227675e-05,
+      "loss": 0.6418,
+      "step": 7593
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 44.22738739414207,
+      "learning_rate": 1.377828816151636e-05,
+      "loss": 0.7061,
+      "step": 7594
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.232284406591827,
+      "learning_rate": 1.377672671235274e-05,
+      "loss": 0.6759,
+      "step": 7595
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.722874267313294,
+      "learning_rate": 1.377516515578122e-05,
+      "loss": 0.707,
+      "step": 7596
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.893974500793902,
+      "learning_rate": 1.3773603491846213e-05,
+      "loss": 0.662,
+      "step": 7597
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.38169421944358,
+      "learning_rate": 1.3772041720592131e-05,
+      "loss": 0.7194,
+      "step": 7598
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 36.5276508597831,
+      "learning_rate": 1.3770479842063387e-05,
+      "loss": 0.712,
+      "step": 7599
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 27.772413652865968,
+      "learning_rate": 1.37689178563044e-05,
+      "loss": 0.5858,
+      "step": 7600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 37.7704625375386,
+      "learning_rate": 1.37673557633596e-05,
+      "loss": 0.6603,
+      "step": 7601
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.7350534423356,
+      "learning_rate": 1.3765793563273402e-05,
+      "loss": 0.7437,
+      "step": 7602
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.72342557167342,
+      "learning_rate": 1.376423125609024e-05,
+      "loss": 0.7153,
+      "step": 7603
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.29633637727094,
+      "learning_rate": 1.376266884185454e-05,
+      "loss": 0.688,
+      "step": 7604
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 37.005660568076145,
+      "learning_rate": 1.376110632061074e-05,
+      "loss": 0.6941,
+      "step": 7605
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 28.06160427099785,
+      "learning_rate": 1.3759543692403278e-05,
+      "loss": 0.5859,
+      "step": 7606
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 34.24859371221903,
+      "learning_rate": 1.3757980957276594e-05,
+      "loss": 0.7804,
+      "step": 7607
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 61.15406455460976,
+      "learning_rate": 1.3756418115275128e-05,
+      "loss": 0.6704,
+      "step": 7608
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 41.676378771257845,
+      "learning_rate": 1.3754855166443326e-05,
+      "loss": 0.7008,
+      "step": 7609
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 99.52960617648274,
+      "learning_rate": 1.375329211082564e-05,
+      "loss": 0.7171,
+      "step": 7610
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.15228045623503,
+      "learning_rate": 1.3751728948466526e-05,
+      "loss": 0.628,
+      "step": 7611
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.2573527485593,
+      "learning_rate": 1.3750165679410431e-05,
+      "loss": 0.6686,
+      "step": 7612
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 42.15564186104338,
+      "learning_rate": 1.3748602303701821e-05,
+      "loss": 0.6647,
+      "step": 7613
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 36.47882062957371,
+      "learning_rate": 1.3747038821385155e-05,
+      "loss": 0.6464,
+      "step": 7614
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 44.809481414044974,
+      "learning_rate": 1.3745475232504895e-05,
+      "loss": 0.7532,
+      "step": 7615
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.975165261504706,
+      "learning_rate": 1.374391153710551e-05,
+      "loss": 0.6963,
+      "step": 7616
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 36.555548791263135,
+      "learning_rate": 1.374234773523147e-05,
+      "loss": 0.6254,
+      "step": 7617
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 49.30928875193785,
+      "learning_rate": 1.3740783826927252e-05,
+      "loss": 0.7911,
+      "step": 7618
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 29.417552413699227,
+      "learning_rate": 1.373921981223733e-05,
+      "loss": 0.6342,
+      "step": 7619
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 34.77642669548382,
+      "learning_rate": 1.3737655691206184e-05,
+      "loss": 0.7111,
+      "step": 7620
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 27.927598393343494,
+      "learning_rate": 1.3736091463878296e-05,
+      "loss": 0.6152,
+      "step": 7621
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.959174413514624,
+      "learning_rate": 1.3734527130298153e-05,
+      "loss": 0.641,
+      "step": 7622
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 31.641385330672133,
+      "learning_rate": 1.373296269051024e-05,
+      "loss": 0.6316,
+      "step": 7623
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 34.86039874643792,
+      "learning_rate": 1.3731398144559056e-05,
+      "loss": 0.6799,
+      "step": 7624
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 33.22175879861365,
+      "learning_rate": 1.3729833492489091e-05,
+      "loss": 0.6397,
+      "step": 7625
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.4118010853375,
+      "learning_rate": 1.3728268734344842e-05,
+      "loss": 0.7118,
+      "step": 7626
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 41.763867804727724,
+      "learning_rate": 1.372670387017081e-05,
+      "loss": 0.7136,
+      "step": 7627
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.51977669033608,
+      "learning_rate": 1.3725138900011503e-05,
+      "loss": 0.6206,
+      "step": 7628
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 34.95241784994901,
+      "learning_rate": 1.372357382391142e-05,
+      "loss": 0.7323,
+      "step": 7629
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 37.79736676000695,
+      "learning_rate": 1.3722008641915081e-05,
+      "loss": 0.6519,
+      "step": 7630
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 43.25292719598783,
+      "learning_rate": 1.372044335406699e-05,
+      "loss": 0.7108,
+      "step": 7631
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 42.18087775919224,
+      "learning_rate": 1.3718877960411669e-05,
+      "loss": 0.6762,
+      "step": 7632
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.43016992680033,
+      "learning_rate": 1.3717312460993632e-05,
+      "loss": 0.6482,
+      "step": 7633
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 55.299558438060956,
+      "learning_rate": 1.3715746855857407e-05,
+      "loss": 0.9309,
+      "step": 7634
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 43.56080996328687,
+      "learning_rate": 1.3714181145047512e-05,
+      "loss": 0.6648,
+      "step": 7635
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 38.281695765799896,
+      "learning_rate": 1.3712615328608476e-05,
+      "loss": 0.8097,
+      "step": 7636
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 44.8569382603976,
+      "learning_rate": 1.3711049406584836e-05,
+      "loss": 0.7429,
+      "step": 7637
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.64849551679384,
+      "learning_rate": 1.3709483379021118e-05,
+      "loss": 0.6653,
+      "step": 7638
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.37827824004349,
+      "learning_rate": 1.3707917245961864e-05,
+      "loss": 0.6021,
+      "step": 7639
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.01720505367157,
+      "learning_rate": 1.3706351007451616e-05,
+      "loss": 0.7194,
+      "step": 7640
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 36.4774006649022,
+      "learning_rate": 1.3704784663534909e-05,
+      "loss": 0.6214,
+      "step": 7641
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.326977330671333,
+      "learning_rate": 1.3703218214256295e-05,
+      "loss": 0.6392,
+      "step": 7642
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 57.827549789029284,
+      "learning_rate": 1.3701651659660322e-05,
+      "loss": 0.6661,
+      "step": 7643
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 36.35883181059247,
+      "learning_rate": 1.3700084999791539e-05,
+      "loss": 0.73,
+      "step": 7644
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 44.481258849968135,
+      "learning_rate": 1.3698518234694507e-05,
+      "loss": 0.7378,
+      "step": 7645
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.50857345764872,
+      "learning_rate": 1.3696951364413776e-05,
+      "loss": 0.6339,
+      "step": 7646
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.85775521137625,
+      "learning_rate": 1.3695384388993915e-05,
+      "loss": 0.8167,
+      "step": 7647
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 41.70206469506047,
+      "learning_rate": 1.369381730847948e-05,
+      "loss": 0.6772,
+      "step": 7648
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 37.36534871803763,
+      "learning_rate": 1.3692250122915042e-05,
+      "loss": 0.6946,
+      "step": 7649
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 28.466139990947,
+      "learning_rate": 1.3690682832345172e-05,
+      "loss": 0.6066,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.149038323061056,
+      "learning_rate": 1.3689115436814442e-05,
+      "loss": 0.6711,
+      "step": 7651
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.439260690476296,
+      "learning_rate": 1.3687547936367424e-05,
+      "loss": 0.6374,
+      "step": 7652
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 37.6873375207489,
+      "learning_rate": 1.36859803310487e-05,
+      "loss": 0.7368,
+      "step": 7653
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 41.528378538778256,
+      "learning_rate": 1.3684412620902854e-05,
+      "loss": 0.749,
+      "step": 7654
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 47.933596634882214,
+      "learning_rate": 1.3682844805974466e-05,
+      "loss": 0.6683,
+      "step": 7655
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.31412548394326,
+      "learning_rate": 1.3681276886308129e-05,
+      "loss": 0.7597,
+      "step": 7656
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 33.475894361347294,
+      "learning_rate": 1.3679708861948429e-05,
+      "loss": 0.6189,
+      "step": 7657
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 39.1087773117744,
+      "learning_rate": 1.3678140732939962e-05,
+      "loss": 0.6699,
+      "step": 7658
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.09137430979773,
+      "learning_rate": 1.3676572499327322e-05,
+      "loss": 0.7027,
+      "step": 7659
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.36550548866043,
+      "learning_rate": 1.3675004161155116e-05,
+      "loss": 0.6421,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.59001469007115,
+      "learning_rate": 1.3673435718467936e-05,
+      "loss": 0.6329,
+      "step": 7661
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.764738255276576,
+      "learning_rate": 1.3671867171310398e-05,
+      "loss": 0.7462,
+      "step": 7662
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 42.34935743076374,
+      "learning_rate": 1.36702985197271e-05,
+      "loss": 0.7211,
+      "step": 7663
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.31244406585362,
+      "learning_rate": 1.3668729763762664e-05,
+      "loss": 0.6677,
+      "step": 7664
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.95587883009984,
+      "learning_rate": 1.3667160903461699e-05,
+      "loss": 0.7362,
+      "step": 7665
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.63902848917561,
+      "learning_rate": 1.366559193886882e-05,
+      "loss": 0.6521,
+      "step": 7666
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 36.801728711262875,
+      "learning_rate": 1.3664022870028655e-05,
+      "loss": 0.6955,
+      "step": 7667
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 36.5407545845111,
+      "learning_rate": 1.366245369698582e-05,
+      "loss": 0.6144,
+      "step": 7668
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.88290975024413,
+      "learning_rate": 1.3660884419784944e-05,
+      "loss": 0.7062,
+      "step": 7669
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.35644334513496,
+      "learning_rate": 1.3659315038470657e-05,
+      "loss": 0.6964,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 30.31947215802199,
+      "learning_rate": 1.3657745553087591e-05,
+      "loss": 0.7068,
+      "step": 7671
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 28.071046300990595,
+      "learning_rate": 1.3656175963680385e-05,
+      "loss": 0.6551,
+      "step": 7672
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 43.263540593142224,
+      "learning_rate": 1.3654606270293666e-05,
+      "loss": 0.8779,
+      "step": 7673
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 46.734311378305605,
+      "learning_rate": 1.3653036472972088e-05,
+      "loss": 0.8571,
+      "step": 7674
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 44.15635082614022,
+      "learning_rate": 1.365146657176029e-05,
+      "loss": 0.785,
+      "step": 7675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 30.99724595849466,
+      "learning_rate": 1.3649896566702916e-05,
+      "loss": 0.6186,
+      "step": 7676
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 31.185189408079857,
+      "learning_rate": 1.364832645784462e-05,
+      "loss": 0.6016,
+      "step": 7677
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.11664620900685,
+      "learning_rate": 1.3646756245230056e-05,
+      "loss": 0.6855,
+      "step": 7678
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.66161912859265,
+      "learning_rate": 1.3645185928903874e-05,
+      "loss": 0.6213,
+      "step": 7679
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 33.47159289014206,
+      "learning_rate": 1.3643615508910734e-05,
+      "loss": 0.7268,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.23254873749207,
+      "learning_rate": 1.3642044985295307e-05,
+      "loss": 0.7489,
+      "step": 7681
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 37.166707093740015,
+      "learning_rate": 1.3640474358102247e-05,
+      "loss": 0.7091,
+      "step": 7682
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 37.33318832949737,
+      "learning_rate": 1.3638903627376227e-05,
+      "loss": 0.6801,
+      "step": 7683
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 43.97703720869354,
+      "learning_rate": 1.3637332793161918e-05,
+      "loss": 0.7532,
+      "step": 7684
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.60137829157704,
+      "learning_rate": 1.363576185550399e-05,
+      "loss": 0.7114,
+      "step": 7685
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 43.09479316184675,
+      "learning_rate": 1.3634190814447124e-05,
+      "loss": 0.8332,
+      "step": 7686
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 35.8596893629203,
+      "learning_rate": 1.3632619670035996e-05,
+      "loss": 0.6788,
+      "step": 7687
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.313759542670034,
+      "learning_rate": 1.3631048422315288e-05,
+      "loss": 0.6769,
+      "step": 7688
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.36959760206904,
+      "learning_rate": 1.3629477071329689e-05,
+      "loss": 0.8306,
+      "step": 7689
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 33.31477850946147,
+      "learning_rate": 1.3627905617123884e-05,
+      "loss": 0.7388,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 30.724056224124926,
+      "learning_rate": 1.3626334059742567e-05,
+      "loss": 0.6635,
+      "step": 7691
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 37.69836660395778,
+      "learning_rate": 1.362476239923043e-05,
+      "loss": 0.6355,
+      "step": 7692
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 35.010609445870635,
+      "learning_rate": 1.3623190635632171e-05,
+      "loss": 0.7232,
+      "step": 7693
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 35.08387103352053,
+      "learning_rate": 1.3621618768992488e-05,
+      "loss": 0.6644,
+      "step": 7694
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 36.26522899906191,
+      "learning_rate": 1.3620046799356088e-05,
+      "loss": 0.7587,
+      "step": 7695
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 51.403081732516114,
+      "learning_rate": 1.3618474726767675e-05,
+      "loss": 0.6548,
+      "step": 7696
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 29.93817764263901,
+      "learning_rate": 1.3616902551271955e-05,
+      "loss": 0.6342,
+      "step": 7697
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 29.2925566010833,
+      "learning_rate": 1.3615330272913643e-05,
+      "loss": 0.612,
+      "step": 7698
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 31.336662913684954,
+      "learning_rate": 1.3613757891737453e-05,
+      "loss": 0.6463,
+      "step": 7699
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.48499019129547,
+      "learning_rate": 1.3612185407788104e-05,
+      "loss": 0.7404,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 33.61322954562886,
+      "learning_rate": 1.3610612821110315e-05,
+      "loss": 0.6411,
+      "step": 7701
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 39.40007300672689,
+      "learning_rate": 1.360904013174881e-05,
+      "loss": 0.6881,
+      "step": 7702
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.798679668475284,
+      "learning_rate": 1.360746733974831e-05,
+      "loss": 0.56,
+      "step": 7703
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.20005361105063,
+      "learning_rate": 1.3605894445153556e-05,
+      "loss": 0.7092,
+      "step": 7704
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 33.75402894520365,
+      "learning_rate": 1.3604321448009268e-05,
+      "loss": 0.6902,
+      "step": 7705
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 36.000337221632506,
+      "learning_rate": 1.3602748348360191e-05,
+      "loss": 0.7491,
+      "step": 7706
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.822983823158125,
+      "learning_rate": 1.3601175146251056e-05,
+      "loss": 0.7446,
+      "step": 7707
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.00031973183584,
+      "learning_rate": 1.3599601841726608e-05,
+      "loss": 0.6339,
+      "step": 7708
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 33.29025555764376,
+      "learning_rate": 1.3598028434831584e-05,
+      "loss": 0.6681,
+      "step": 7709
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 32.18036870751715,
+      "learning_rate": 1.3596454925610742e-05,
+      "loss": 0.6406,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.738828036415654,
+      "learning_rate": 1.3594881314108823e-05,
+      "loss": 0.5741,
+      "step": 7711
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.74191235465556,
+      "learning_rate": 1.3593307600370585e-05,
+      "loss": 0.7597,
+      "step": 7712
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 46.03605771774279,
+      "learning_rate": 1.359173378444078e-05,
+      "loss": 0.818,
+      "step": 7713
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 35.63497771359172,
+      "learning_rate": 1.3590159866364166e-05,
+      "loss": 0.6897,
+      "step": 7714
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 39.148824375839176,
+      "learning_rate": 1.3588585846185502e-05,
+      "loss": 0.652,
+      "step": 7715
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 41.85632420750158,
+      "learning_rate": 1.3587011723949564e-05,
+      "loss": 0.7365,
+      "step": 7716
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.277163362281755,
+      "learning_rate": 1.3585437499701104e-05,
+      "loss": 0.6778,
+      "step": 7717
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 53.86473075459473,
+      "learning_rate": 1.3583863173484902e-05,
+      "loss": 0.7287,
+      "step": 7718
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 36.938317949820565,
+      "learning_rate": 1.3582288745345728e-05,
+      "loss": 0.6319,
+      "step": 7719
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.12121904430903,
+      "learning_rate": 1.3580714215328357e-05,
+      "loss": 0.7424,
+      "step": 7720
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.47462543939115,
+      "learning_rate": 1.357913958347757e-05,
+      "loss": 0.7647,
+      "step": 7721
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 53.04057531240423,
+      "learning_rate": 1.3577564849838146e-05,
+      "loss": 0.7736,
+      "step": 7722
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 47.80970196484033,
+      "learning_rate": 1.3575990014454869e-05,
+      "loss": 0.7316,
+      "step": 7723
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 31.715118865581665,
+      "learning_rate": 1.3574415077372527e-05,
+      "loss": 0.6398,
+      "step": 7724
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 36.35958045209633,
+      "learning_rate": 1.3572840038635913e-05,
+      "loss": 0.7301,
+      "step": 7725
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.79603388077628,
+      "learning_rate": 1.357126489828982e-05,
+      "loss": 0.6491,
+      "step": 7726
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 41.845679753585216,
+      "learning_rate": 1.356968965637904e-05,
+      "loss": 0.679,
+      "step": 7727
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 28.526639028863283,
+      "learning_rate": 1.3568114312948375e-05,
+      "loss": 0.6831,
+      "step": 7728
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.90518472217152,
+      "learning_rate": 1.3566538868042627e-05,
+      "loss": 0.7067,
+      "step": 7729
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.10840402142087,
+      "learning_rate": 1.3564963321706597e-05,
+      "loss": 0.7027,
+      "step": 7730
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 28.006787239579587,
+      "learning_rate": 1.35633876739851e-05,
+      "loss": 0.5559,
+      "step": 7731
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.236142153060484,
+      "learning_rate": 1.3561811924922939e-05,
+      "loss": 0.7024,
+      "step": 7732
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.20006522847252,
+      "learning_rate": 1.3560236074564935e-05,
+      "loss": 0.6925,
+      "step": 7733
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 31.559265677287954,
+      "learning_rate": 1.3558660122955894e-05,
+      "loss": 0.6477,
+      "step": 7734
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 36.993472080826976,
+      "learning_rate": 1.3557084070140644e-05,
+      "loss": 0.6363,
+      "step": 7735
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 38.54488924042988,
+      "learning_rate": 1.3555507916164002e-05,
+      "loss": 0.7493,
+      "step": 7736
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 29.310879782530346,
+      "learning_rate": 1.3553931661070796e-05,
+      "loss": 0.6027,
+      "step": 7737
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 40.718166976584094,
+      "learning_rate": 1.3552355304905851e-05,
+      "loss": 0.6275,
+      "step": 7738
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 29.94552026642246,
+      "learning_rate": 1.3550778847714001e-05,
+      "loss": 0.7029,
+      "step": 7739
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 29.95350834576813,
+      "learning_rate": 1.3549202289540077e-05,
+      "loss": 0.6204,
+      "step": 7740
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 35.1818071811541,
+      "learning_rate": 1.3547625630428919e-05,
+      "loss": 0.6642,
+      "step": 7741
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 42.05994214220371,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.8008,
+      "step": 7742
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 50.45330715057958,
+      "learning_rate": 1.3544472009574243e-05,
+      "loss": 0.578,
+      "step": 7743
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.004373244618,
+      "learning_rate": 1.354289504792042e-05,
+      "loss": 0.7025,
+      "step": 7744
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 36.450722524888874,
+      "learning_rate": 1.3541317985508733e-05,
+      "loss": 0.7165,
+      "step": 7745
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 43.15438623844243,
+      "learning_rate": 1.3539740822384034e-05,
+      "loss": 0.6706,
+      "step": 7746
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 40.098360877295875,
+      "learning_rate": 1.3538163558591175e-05,
+      "loss": 0.6531,
+      "step": 7747
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.96524760865711,
+      "learning_rate": 1.3536586194175019e-05,
+      "loss": 0.7114,
+      "step": 7748
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.94485462675096,
+      "learning_rate": 1.3535008729180414e-05,
+      "loss": 0.6776,
+      "step": 7749
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 38.19366226797963,
+      "learning_rate": 1.3533431163652235e-05,
+      "loss": 0.6921,
+      "step": 7750
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 48.057573648878645,
+      "learning_rate": 1.3531853497635336e-05,
+      "loss": 0.7433,
+      "step": 7751
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 38.359286918461315,
+      "learning_rate": 1.3530275731174591e-05,
+      "loss": 0.5953,
+      "step": 7752
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 42.14907791457235,
+      "learning_rate": 1.352869786431487e-05,
+      "loss": 0.7349,
+      "step": 7753
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 53.049310709747466,
+      "learning_rate": 1.3527119897101046e-05,
+      "loss": 0.7028,
+      "step": 7754
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.91003762154914,
+      "learning_rate": 1.3525541829577997e-05,
+      "loss": 0.7076,
+      "step": 7755
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 38.547535436420866,
+      "learning_rate": 1.3523963661790597e-05,
+      "loss": 0.688,
+      "step": 7756
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.05735246102873,
+      "learning_rate": 1.3522385393783735e-05,
+      "loss": 0.7209,
+      "step": 7757
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.47642546425934,
+      "learning_rate": 1.3520807025602291e-05,
+      "loss": 0.6483,
+      "step": 7758
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 51.74281540353892,
+      "learning_rate": 1.3519228557291156e-05,
+      "loss": 0.7108,
+      "step": 7759
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.6128964330662,
+      "learning_rate": 1.3517649988895221e-05,
+      "loss": 0.6588,
+      "step": 7760
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.141979242779215,
+      "learning_rate": 1.3516071320459374e-05,
+      "loss": 0.6672,
+      "step": 7761
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.17653762418294,
+      "learning_rate": 1.3514492552028518e-05,
+      "loss": 0.7187,
+      "step": 7762
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 35.658365298889294,
+      "learning_rate": 1.351291368364755e-05,
+      "loss": 0.7131,
+      "step": 7763
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 41.429157947865605,
+      "learning_rate": 1.3511334715361371e-05,
+      "loss": 0.6699,
+      "step": 7764
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.180511938996695,
+      "learning_rate": 1.3509755647214888e-05,
+      "loss": 0.5853,
+      "step": 7765
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 31.38970655404988,
+      "learning_rate": 1.3508176479253003e-05,
+      "loss": 0.6709,
+      "step": 7766
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.49235433708788,
+      "learning_rate": 1.3506597211520638e-05,
+      "loss": 0.6956,
+      "step": 7767
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 37.93010575508804,
+      "learning_rate": 1.3505017844062693e-05,
+      "loss": 0.7492,
+      "step": 7768
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 39.742818921492834,
+      "learning_rate": 1.3503438376924096e-05,
+      "loss": 0.6568,
+      "step": 7769
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 41.07576145353246,
+      "learning_rate": 1.350185881014976e-05,
+      "loss": 0.6518,
+      "step": 7770
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 35.50423648566067,
+      "learning_rate": 1.3500279143784606e-05,
+      "loss": 0.6985,
+      "step": 7771
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.184798759848995,
+      "learning_rate": 1.3498699377873563e-05,
+      "loss": 0.687,
+      "step": 7772
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 38.54203171256457,
+      "learning_rate": 1.3497119512461555e-05,
+      "loss": 0.7655,
+      "step": 7773
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 36.14629560975657,
+      "learning_rate": 1.3495539547593512e-05,
+      "loss": 0.6818,
+      "step": 7774
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 38.04096176285448,
+      "learning_rate": 1.3493959483314373e-05,
+      "loss": 0.734,
+      "step": 7775
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 34.541733830658536,
+      "learning_rate": 1.349237931966907e-05,
+      "loss": 0.697,
+      "step": 7776
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 31.081022197533695,
+      "learning_rate": 1.3490799056702542e-05,
+      "loss": 0.7111,
+      "step": 7777
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 41.04398725245603,
+      "learning_rate": 1.3489218694459727e-05,
+      "loss": 0.72,
+      "step": 7778
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.55066167125525,
+      "learning_rate": 1.3487638232985575e-05,
+      "loss": 0.6481,
+      "step": 7779
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 40.79874481626565,
+      "learning_rate": 1.3486057672325035e-05,
+      "loss": 0.7168,
+      "step": 7780
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 30.74364373952017,
+      "learning_rate": 1.348447701252305e-05,
+      "loss": 0.6739,
+      "step": 7781
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.134308354937254,
+      "learning_rate": 1.348289625362458e-05,
+      "loss": 0.6476,
+      "step": 7782
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.44592370995722,
+      "learning_rate": 1.3481315395674577e-05,
+      "loss": 0.698,
+      "step": 7783
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 52.16274098574425,
+      "learning_rate": 1.3479734438718002e-05,
+      "loss": 0.8484,
+      "step": 7784
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 28.61477541410627,
+      "learning_rate": 1.3478153382799812e-05,
+      "loss": 0.6619,
+      "step": 7785
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.4452229919716,
+      "learning_rate": 1.3476572227964976e-05,
+      "loss": 0.624,
+      "step": 7786
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 32.35677604977152,
+      "learning_rate": 1.347499097425846e-05,
+      "loss": 0.6273,
+      "step": 7787
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 33.090742495060994,
+      "learning_rate": 1.3473409621725233e-05,
+      "loss": 0.6587,
+      "step": 7788
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.960055450724546,
+      "learning_rate": 1.3471828170410268e-05,
+      "loss": 0.716,
+      "step": 7789
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.830353168222516,
+      "learning_rate": 1.3470246620358543e-05,
+      "loss": 0.7406,
+      "step": 7790
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.18343390052967,
+      "learning_rate": 1.3468664971615031e-05,
+      "loss": 0.6836,
+      "step": 7791
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 34.78428061923749,
+      "learning_rate": 1.3467083224224719e-05,
+      "loss": 0.6672,
+      "step": 7792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.85427292241531,
+      "learning_rate": 1.3465501378232586e-05,
+      "loss": 0.7507,
+      "step": 7793
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.817290572415686,
+      "learning_rate": 1.3463919433683624e-05,
+      "loss": 0.692,
+      "step": 7794
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.88187503399879,
+      "learning_rate": 1.3462337390622814e-05,
+      "loss": 0.6941,
+      "step": 7795
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 29.85799172399691,
+      "learning_rate": 1.3460755249095161e-05,
+      "loss": 0.6295,
+      "step": 7796
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.76708913986543,
+      "learning_rate": 1.345917300914565e-05,
+      "loss": 0.6989,
+      "step": 7797
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.81527088091071,
+      "learning_rate": 1.3457590670819283e-05,
+      "loss": 0.7155,
+      "step": 7798
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 43.51834365598044,
+      "learning_rate": 1.3456008234161057e-05,
+      "loss": 0.6313,
+      "step": 7799
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.38652853996589,
+      "learning_rate": 1.3454425699215982e-05,
+      "loss": 0.7805,
+      "step": 7800
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.911562983746386,
+      "learning_rate": 1.3452843066029058e-05,
+      "loss": 0.6958,
+      "step": 7801
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.939347068754934,
+      "learning_rate": 1.3451260334645299e-05,
+      "loss": 0.6764,
+      "step": 7802
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 41.846824869190385,
+      "learning_rate": 1.3449677505109716e-05,
+      "loss": 0.7516,
+      "step": 7803
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 45.52694772935131,
+      "learning_rate": 1.3448094577467323e-05,
+      "loss": 0.6999,
+      "step": 7804
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 63.48578829439722,
+      "learning_rate": 1.3446511551763133e-05,
+      "loss": 0.6995,
+      "step": 7805
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 43.90480615271343,
+      "learning_rate": 1.3444928428042175e-05,
+      "loss": 0.6636,
+      "step": 7806
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 31.757814932397277,
+      "learning_rate": 1.3443345206349466e-05,
+      "loss": 0.6451,
+      "step": 7807
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.78385781248187,
+      "learning_rate": 1.3441761886730033e-05,
+      "loss": 0.7973,
+      "step": 7808
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 44.424971122066836,
+      "learning_rate": 1.3440178469228906e-05,
+      "loss": 0.7902,
+      "step": 7809
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 39.63655426692021,
+      "learning_rate": 1.3438594953891113e-05,
+      "loss": 0.7582,
+      "step": 7810
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 32.998970279378156,
+      "learning_rate": 1.3437011340761697e-05,
+      "loss": 0.6776,
+      "step": 7811
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.14958441374791,
+      "learning_rate": 1.3435427629885684e-05,
+      "loss": 0.6468,
+      "step": 7812
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 31.997518100454734,
+      "learning_rate": 1.343384382130812e-05,
+      "loss": 0.6474,
+      "step": 7813
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.246425873226315,
+      "learning_rate": 1.3432259915074048e-05,
+      "loss": 0.7298,
+      "step": 7814
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 38.02293817346297,
+      "learning_rate": 1.3430675911228512e-05,
+      "loss": 0.673,
+      "step": 7815
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 34.52846102606701,
+      "learning_rate": 1.342909180981656e-05,
+      "loss": 0.6792,
+      "step": 7816
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 46.471045495744384,
+      "learning_rate": 1.3427507610883242e-05,
+      "loss": 0.6852,
+      "step": 7817
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 30.824531964494685,
+      "learning_rate": 1.342592331447361e-05,
+      "loss": 0.6206,
+      "step": 7818
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 39.669156775065055,
+      "learning_rate": 1.342433892063273e-05,
+      "loss": 0.7119,
+      "step": 7819
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 34.168674456909024,
+      "learning_rate": 1.3422754429405647e-05,
+      "loss": 0.6263,
+      "step": 7820
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 32.31428709566793,
+      "learning_rate": 1.3421169840837433e-05,
+      "loss": 0.7081,
+      "step": 7821
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 32.01307001153817,
+      "learning_rate": 1.341958515497315e-05,
+      "loss": 0.6604,
+      "step": 7822
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 31.187360147839186,
+      "learning_rate": 1.3418000371857866e-05,
+      "loss": 0.6761,
+      "step": 7823
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 33.11780565696866,
+      "learning_rate": 1.341641549153665e-05,
+      "loss": 0.6267,
+      "step": 7824
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 52.38892744700493,
+      "learning_rate": 1.3414830514054578e-05,
+      "loss": 0.7963,
+      "step": 7825
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.05280860973006,
+      "learning_rate": 1.3413245439456721e-05,
+      "loss": 0.7077,
+      "step": 7826
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 38.60300165708062,
+      "learning_rate": 1.3411660267788163e-05,
+      "loss": 0.8147,
+      "step": 7827
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.6079965044814,
+      "learning_rate": 1.341007499909398e-05,
+      "loss": 0.721,
+      "step": 7828
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 46.82250338984794,
+      "learning_rate": 1.3408489633419261e-05,
+      "loss": 0.7406,
+      "step": 7829
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.25772639745253,
+      "learning_rate": 1.340690417080909e-05,
+      "loss": 0.634,
+      "step": 7830
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.79457886412913,
+      "learning_rate": 1.3405318611308558e-05,
+      "loss": 0.7056,
+      "step": 7831
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 28.404546659323877,
+      "learning_rate": 1.3403732954962755e-05,
+      "loss": 0.6425,
+      "step": 7832
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 32.996679745882304,
+      "learning_rate": 1.3402147201816779e-05,
+      "loss": 0.6397,
+      "step": 7833
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 39.188095946092574,
+      "learning_rate": 1.3400561351915728e-05,
+      "loss": 0.7604,
+      "step": 7834
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 33.85624629318727,
+      "learning_rate": 1.33989754053047e-05,
+      "loss": 0.6127,
+      "step": 7835
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 38.11991007863314,
+      "learning_rate": 1.3397389362028802e-05,
+      "loss": 0.7279,
+      "step": 7836
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 34.0993172272374,
+      "learning_rate": 1.3395803222133133e-05,
+      "loss": 0.7256,
+      "step": 7837
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 40.80200239648583,
+      "learning_rate": 1.3394216985662813e-05,
+      "loss": 0.7831,
+      "step": 7838
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 30.448256722473122,
+      "learning_rate": 1.3392630652662943e-05,
+      "loss": 0.539,
+      "step": 7839
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 37.496466813535044,
+      "learning_rate": 1.3391044223178641e-05,
+      "loss": 0.734,
+      "step": 7840
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.67553877259936,
+      "learning_rate": 1.338945769725503e-05,
+      "loss": 0.6435,
+      "step": 7841
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.798894472853526,
+      "learning_rate": 1.3387871074937222e-05,
+      "loss": 0.7354,
+      "step": 7842
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 37.21542372681557,
+      "learning_rate": 1.3386284356270342e-05,
+      "loss": 0.6495,
+      "step": 7843
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 25.95665856776273,
+      "learning_rate": 1.3384697541299518e-05,
+      "loss": 0.6353,
+      "step": 7844
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 29.35465710318018,
+      "learning_rate": 1.3383110630069874e-05,
+      "loss": 0.6162,
+      "step": 7845
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 34.910873564502104,
+      "learning_rate": 1.3381523622626544e-05,
+      "loss": 0.6931,
+      "step": 7846
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.812257905051563,
+      "learning_rate": 1.3379936519014658e-05,
+      "loss": 0.6664,
+      "step": 7847
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.29813785972231,
+      "learning_rate": 1.3378349319279358e-05,
+      "loss": 0.7112,
+      "step": 7848
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 39.573802767709715,
+      "learning_rate": 1.3376762023465775e-05,
+      "loss": 0.6682,
+      "step": 7849
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.05551545964744,
+      "learning_rate": 1.337517463161906e-05,
+      "loss": 0.5959,
+      "step": 7850
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.63036432048821,
+      "learning_rate": 1.337358714378435e-05,
+      "loss": 0.6093,
+      "step": 7851
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.792434794214536,
+      "learning_rate": 1.3371999560006796e-05,
+      "loss": 0.6502,
+      "step": 7852
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.496658787561017,
+      "learning_rate": 1.3370411880331549e-05,
+      "loss": 0.5224,
+      "step": 7853
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.19959889516641,
+      "learning_rate": 1.3368824104803755e-05,
+      "loss": 0.678,
+      "step": 7854
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 29.159594345044543,
+      "learning_rate": 1.3367236233468579e-05,
+      "loss": 0.6647,
+      "step": 7855
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 37.18083996157921,
+      "learning_rate": 1.3365648266371171e-05,
+      "loss": 0.7719,
+      "step": 7856
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 39.62155696880562,
+      "learning_rate": 1.3364060203556697e-05,
+      "loss": 0.7082,
+      "step": 7857
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.480859775716514,
+      "learning_rate": 1.3362472045070318e-05,
+      "loss": 0.5914,
+      "step": 7858
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 43.60858396320476,
+      "learning_rate": 1.3360883790957202e-05,
+      "loss": 0.7774,
+      "step": 7859
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 34.33849796473872,
+      "learning_rate": 1.3359295441262516e-05,
+      "loss": 0.6365,
+      "step": 7860
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 42.53156474919843,
+      "learning_rate": 1.3357706996031434e-05,
+      "loss": 0.7318,
+      "step": 7861
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.60163197982535,
+      "learning_rate": 1.3356118455309127e-05,
+      "loss": 0.5974,
+      "step": 7862
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.909223291998426,
+      "learning_rate": 1.3354529819140777e-05,
+      "loss": 0.5798,
+      "step": 7863
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 40.19474934738593,
+      "learning_rate": 1.3352941087571558e-05,
+      "loss": 0.6697,
+      "step": 7864
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 36.55038040004927,
+      "learning_rate": 1.3351352260646662e-05,
+      "loss": 0.6781,
+      "step": 7865
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.442692424539928,
+      "learning_rate": 1.3349763338411261e-05,
+      "loss": 0.6756,
+      "step": 7866
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.000320810351074,
+      "learning_rate": 1.3348174320910554e-05,
+      "loss": 0.6793,
+      "step": 7867
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.067671130816088,
+      "learning_rate": 1.3346585208189727e-05,
+      "loss": 0.723,
+      "step": 7868
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.91512158059484,
+      "learning_rate": 1.3344996000293976e-05,
+      "loss": 0.644,
+      "step": 7869
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 43.65287363571359,
+      "learning_rate": 1.3343406697268493e-05,
+      "loss": 0.8061,
+      "step": 7870
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 37.49831435547611,
+      "learning_rate": 1.334181729915848e-05,
+      "loss": 0.7786,
+      "step": 7871
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.74518662173866,
+      "learning_rate": 1.3340227806009138e-05,
+      "loss": 0.7363,
+      "step": 7872
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 42.24724161861051,
+      "learning_rate": 1.3338638217865673e-05,
+      "loss": 0.8734,
+      "step": 7873
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 45.0082986067438,
+      "learning_rate": 1.333704853477329e-05,
+      "loss": 0.6897,
+      "step": 7874
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.81375907999441,
+      "learning_rate": 1.3335458756777195e-05,
+      "loss": 0.6084,
+      "step": 7875
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 44.26679634566922,
+      "learning_rate": 1.3333868883922606e-05,
+      "loss": 0.9087,
+      "step": 7876
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.13485848218767,
+      "learning_rate": 1.3332278916254739e-05,
+      "loss": 0.7337,
+      "step": 7877
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 36.04596631013051,
+      "learning_rate": 1.3330688853818808e-05,
+      "loss": 0.6664,
+      "step": 7878
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.68917555206945,
+      "learning_rate": 1.332909869666003e-05,
+      "loss": 0.6436,
+      "step": 7879
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.29939797950553,
+      "learning_rate": 1.3327508444823637e-05,
+      "loss": 0.5788,
+      "step": 7880
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 33.91530834792871,
+      "learning_rate": 1.3325918098354848e-05,
+      "loss": 0.6086,
+      "step": 7881
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.72239010702846,
+      "learning_rate": 1.33243276572989e-05,
+      "loss": 0.7254,
+      "step": 7882
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 43.40750994134205,
+      "learning_rate": 1.3322737121701011e-05,
+      "loss": 0.806,
+      "step": 7883
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.24552499489158,
+      "learning_rate": 1.3321146491606423e-05,
+      "loss": 0.6962,
+      "step": 7884
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.763078918091345,
+      "learning_rate": 1.3319555767060373e-05,
+      "loss": 0.6484,
+      "step": 7885
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 44.39670483780834,
+      "learning_rate": 1.33179649481081e-05,
+      "loss": 0.6843,
+      "step": 7886
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.46472901393912,
+      "learning_rate": 1.3316374034794844e-05,
+      "loss": 0.6777,
+      "step": 7887
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 45.14334476420571,
+      "learning_rate": 1.331478302716585e-05,
+      "loss": 0.6822,
+      "step": 7888
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 34.15558459932981,
+      "learning_rate": 1.3313191925266362e-05,
+      "loss": 0.7123,
+      "step": 7889
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 36.755886329832755,
+      "learning_rate": 1.3311600729141641e-05,
+      "loss": 0.7049,
+      "step": 7890
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.89111080679558,
+      "learning_rate": 1.3310009438836925e-05,
+      "loss": 0.7091,
+      "step": 7891
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 34.98987856020463,
+      "learning_rate": 1.3308418054397485e-05,
+      "loss": 0.6857,
+      "step": 7892
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.026504570744414,
+      "learning_rate": 1.3306826575868565e-05,
+      "loss": 0.6415,
+      "step": 7893
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 32.07612007144827,
+      "learning_rate": 1.3305235003295434e-05,
+      "loss": 0.686,
+      "step": 7894
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 33.38596962913702,
+      "learning_rate": 1.330364333672335e-05,
+      "loss": 0.6348,
+      "step": 7895
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.09717995517327,
+      "learning_rate": 1.3302051576197586e-05,
+      "loss": 0.6725,
+      "step": 7896
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 37.418006814916794,
+      "learning_rate": 1.3300459721763405e-05,
+      "loss": 0.6158,
+      "step": 7897
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 34.35739153956673,
+      "learning_rate": 1.329886777346608e-05,
+      "loss": 0.6076,
+      "step": 7898
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 30.644997852242586,
+      "learning_rate": 1.3297275731350885e-05,
+      "loss": 0.6439,
+      "step": 7899
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 40.65064493197924,
+      "learning_rate": 1.3295683595463099e-05,
+      "loss": 0.8065,
+      "step": 7900
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.05116048493218,
+      "learning_rate": 1.3294091365847998e-05,
+      "loss": 0.617,
+      "step": 7901
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 37.223959787974636,
+      "learning_rate": 1.3292499042550869e-05,
+      "loss": 0.6509,
+      "step": 7902
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.681626552406755,
+      "learning_rate": 1.3290906625616988e-05,
+      "loss": 0.6672,
+      "step": 7903
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 52.83941783777088,
+      "learning_rate": 1.3289314115091649e-05,
+      "loss": 0.764,
+      "step": 7904
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 41.04794591255246,
+      "learning_rate": 1.3287721511020145e-05,
+      "loss": 0.705,
+      "step": 7905
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 46.603298434390936,
+      "learning_rate": 1.3286128813447759e-05,
+      "loss": 0.7149,
+      "step": 7906
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 31.915638378367976,
+      "learning_rate": 1.3284536022419796e-05,
+      "loss": 0.5946,
+      "step": 7907
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.56195910069437,
+      "learning_rate": 1.3282943137981546e-05,
+      "loss": 0.6263,
+      "step": 7908
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 39.08780531963757,
+      "learning_rate": 1.3281350160178317e-05,
+      "loss": 0.895,
+      "step": 7909
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 37.647685280730315,
+      "learning_rate": 1.3279757089055404e-05,
+      "loss": 0.6896,
+      "step": 7910
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 42.24150108314481,
+      "learning_rate": 1.3278163924658123e-05,
+      "loss": 0.6753,
+      "step": 7911
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 42.43064036429352,
+      "learning_rate": 1.3276570667031777e-05,
+      "loss": 0.6145,
+      "step": 7912
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.6066163399115,
+      "learning_rate": 1.3274977316221678e-05,
+      "loss": 0.6473,
+      "step": 7913
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.05692694333025,
+      "learning_rate": 1.3273383872273137e-05,
+      "loss": 0.7346,
+      "step": 7914
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.28125929229001,
+      "learning_rate": 1.3271790335231476e-05,
+      "loss": 0.8096,
+      "step": 7915
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.932387135895794,
+      "learning_rate": 1.327019670514201e-05,
+      "loss": 0.6493,
+      "step": 7916
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.79495419638067,
+      "learning_rate": 1.3268602982050064e-05,
+      "loss": 0.6621,
+      "step": 7917
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.048009445964404,
+      "learning_rate": 1.3267009166000961e-05,
+      "loss": 0.6871,
+      "step": 7918
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.4712235852771,
+      "learning_rate": 1.3265415257040031e-05,
+      "loss": 0.7129,
+      "step": 7919
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 41.893858210352626,
+      "learning_rate": 1.3263821255212598e-05,
+      "loss": 0.7341,
+      "step": 7920
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 30.634911179006757,
+      "learning_rate": 1.3262227160564001e-05,
+      "loss": 0.7003,
+      "step": 7921
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.67748635394358,
+      "learning_rate": 1.3260632973139569e-05,
+      "loss": 0.6684,
+      "step": 7922
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.38680335788756,
+      "learning_rate": 1.3259038692984643e-05,
+      "loss": 0.6829,
+      "step": 7923
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 39.01244634925896,
+      "learning_rate": 1.3257444320144566e-05,
+      "loss": 0.7298,
+      "step": 7924
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 36.17130769486602,
+      "learning_rate": 1.3255849854664674e-05,
+      "loss": 0.7095,
+      "step": 7925
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 36.41030852653583,
+      "learning_rate": 1.325425529659032e-05,
+      "loss": 0.6544,
+      "step": 7926
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 39.95841969960372,
+      "learning_rate": 1.3252660645966846e-05,
+      "loss": 0.7655,
+      "step": 7927
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 40.05662614369306,
+      "learning_rate": 1.3251065902839608e-05,
+      "loss": 0.7157,
+      "step": 7928
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.25116083465533,
+      "learning_rate": 1.3249471067253957e-05,
+      "loss": 0.6789,
+      "step": 7929
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.9709801734021,
+      "learning_rate": 1.3247876139255249e-05,
+      "loss": 0.6764,
+      "step": 7930
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.18824859650358,
+      "learning_rate": 1.3246281118888843e-05,
+      "loss": 0.7588,
+      "step": 7931
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.58584899817159,
+      "learning_rate": 1.3244686006200105e-05,
+      "loss": 0.6903,
+      "step": 7932
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 39.82279983245264,
+      "learning_rate": 1.324309080123439e-05,
+      "loss": 0.6802,
+      "step": 7933
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.906718932240445,
+      "learning_rate": 1.3241495504037074e-05,
+      "loss": 0.8049,
+      "step": 7934
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 45.52175082657691,
+      "learning_rate": 1.323990011465352e-05,
+      "loss": 0.6839,
+      "step": 7935
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 36.701883333200456,
+      "learning_rate": 1.3238304633129102e-05,
+      "loss": 0.7582,
+      "step": 7936
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.662673096436635,
+      "learning_rate": 1.3236709059509194e-05,
+      "loss": 0.6266,
+      "step": 7937
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.62749613789866,
+      "learning_rate": 1.3235113393839174e-05,
+      "loss": 0.6772,
+      "step": 7938
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.472389501459695,
+      "learning_rate": 1.3233517636164423e-05,
+      "loss": 0.7211,
+      "step": 7939
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 39.682822626182215,
+      "learning_rate": 1.3231921786530323e-05,
+      "loss": 0.6564,
+      "step": 7940
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.56302132370335,
+      "learning_rate": 1.3230325844982254e-05,
+      "loss": 0.683,
+      "step": 7941
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.093987981272996,
+      "learning_rate": 1.3228729811565608e-05,
+      "loss": 0.5904,
+      "step": 7942
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 42.78525749590905,
+      "learning_rate": 1.3227133686325777e-05,
+      "loss": 0.7038,
+      "step": 7943
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 43.48937285716888,
+      "learning_rate": 1.3225537469308152e-05,
+      "loss": 0.747,
+      "step": 7944
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.49067189102186,
+      "learning_rate": 1.3223941160558127e-05,
+      "loss": 0.6727,
+      "step": 7945
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 44.60208338588512,
+      "learning_rate": 1.3222344760121105e-05,
+      "loss": 0.6918,
+      "step": 7946
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 36.98699735910377,
+      "learning_rate": 1.3220748268042478e-05,
+      "loss": 0.6739,
+      "step": 7947
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 30.166273504842465,
+      "learning_rate": 1.3219151684367656e-05,
+      "loss": 0.6722,
+      "step": 7948
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.18146451276927,
+      "learning_rate": 1.3217555009142044e-05,
+      "loss": 0.65,
+      "step": 7949
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 33.607614723832114,
+      "learning_rate": 1.3215958242411048e-05,
+      "loss": 0.6329,
+      "step": 7950
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.602834180328706,
+      "learning_rate": 1.3214361384220084e-05,
+      "loss": 0.7285,
+      "step": 7951
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.46189069927054,
+      "learning_rate": 1.321276443461456e-05,
+      "loss": 0.7347,
+      "step": 7952
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 32.93958781825828,
+      "learning_rate": 1.3211167393639898e-05,
+      "loss": 0.6911,
+      "step": 7953
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 37.05088482622581,
+      "learning_rate": 1.320957026134151e-05,
+      "loss": 0.6351,
+      "step": 7954
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 46.21981853619504,
+      "learning_rate": 1.3207973037764822e-05,
+      "loss": 0.6427,
+      "step": 7955
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.9261420670342,
+      "learning_rate": 1.3206375722955261e-05,
+      "loss": 0.6799,
+      "step": 7956
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.9784796354121,
+      "learning_rate": 1.3204778316958247e-05,
+      "loss": 0.6708,
+      "step": 7957
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 41.203385307059705,
+      "learning_rate": 1.3203180819819213e-05,
+      "loss": 0.6469,
+      "step": 7958
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 30.00495426278108,
+      "learning_rate": 1.320158323158359e-05,
+      "loss": 0.61,
+      "step": 7959
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 43.75162296554601,
+      "learning_rate": 1.319998555229681e-05,
+      "loss": 0.7189,
+      "step": 7960
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 41.17181723082107,
+      "learning_rate": 1.3198387782004315e-05,
+      "loss": 0.7403,
+      "step": 7961
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 63.656521387884105,
+      "learning_rate": 1.3196789920751546e-05,
+      "loss": 0.6527,
+      "step": 7962
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.45520786439653,
+      "learning_rate": 1.3195191968583938e-05,
+      "loss": 0.7044,
+      "step": 7963
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 28.6809811176956,
+      "learning_rate": 1.3193593925546941e-05,
+      "loss": 0.6242,
+      "step": 7964
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 31.200055937533484,
+      "learning_rate": 1.3191995791686002e-05,
+      "loss": 0.6696,
+      "step": 7965
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.00419441715367,
+      "learning_rate": 1.3190397567046569e-05,
+      "loss": 0.6761,
+      "step": 7966
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 36.14221307322271,
+      "learning_rate": 1.3188799251674092e-05,
+      "loss": 0.7045,
+      "step": 7967
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 41.869879993178365,
+      "learning_rate": 1.3187200845614035e-05,
+      "loss": 0.7095,
+      "step": 7968
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 35.25685631917552,
+      "learning_rate": 1.3185602348911846e-05,
+      "loss": 0.6656,
+      "step": 7969
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 36.46905186976844,
+      "learning_rate": 1.318400376161299e-05,
+      "loss": 0.6698,
+      "step": 7970
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 33.41171406236495,
+      "learning_rate": 1.3182405083762935e-05,
+      "loss": 0.7133,
+      "step": 7971
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 49.07187119905231,
+      "learning_rate": 1.3180806315407137e-05,
+      "loss": 0.7982,
+      "step": 7972
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 41.33474551212391,
+      "learning_rate": 1.3179207456591067e-05,
+      "loss": 0.7451,
+      "step": 7973
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 37.14111276449931,
+      "learning_rate": 1.31776085073602e-05,
+      "loss": 0.7436,
+      "step": 7974
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 39.5471847624537,
+      "learning_rate": 1.3176009467760005e-05,
+      "loss": 0.8314,
+      "step": 7975
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 42.96206732924139,
+      "learning_rate": 1.3174410337835956e-05,
+      "loss": 0.7348,
+      "step": 7976
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 31.913740184935275,
+      "learning_rate": 1.3172811117633536e-05,
+      "loss": 0.5792,
+      "step": 7977
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 64.21552595598445,
+      "learning_rate": 1.3171211807198225e-05,
+      "loss": 0.618,
+      "step": 7978
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.55886990453522,
+      "learning_rate": 1.3169612406575501e-05,
+      "loss": 0.6346,
+      "step": 7979
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 40.39180683937776,
+      "learning_rate": 1.316801291581086e-05,
+      "loss": 0.6535,
+      "step": 7980
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.871580829060214,
+      "learning_rate": 1.316641333494978e-05,
+      "loss": 0.7059,
+      "step": 7981
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 33.03384044237444,
+      "learning_rate": 1.316481366403776e-05,
+      "loss": 0.6113,
+      "step": 7982
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.305397736979586,
+      "learning_rate": 1.316321390312029e-05,
+      "loss": 0.7245,
+      "step": 7983
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 52.219247862418904,
+      "learning_rate": 1.3161614052242869e-05,
+      "loss": 0.6341,
+      "step": 7984
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 39.09518127809681,
+      "learning_rate": 1.3160014111450989e-05,
+      "loss": 0.7329,
+      "step": 7985
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 37.51715668169459,
+      "learning_rate": 1.3158414080790158e-05,
+      "loss": 0.746,
+      "step": 7986
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 41.263516968424476,
+      "learning_rate": 1.315681396030588e-05,
+      "loss": 0.6512,
+      "step": 7987
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 42.08770592732322,
+      "learning_rate": 1.3155213750043658e-05,
+      "loss": 0.6828,
+      "step": 7988
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 36.1474547902872,
+      "learning_rate": 1.3153613450049006e-05,
+      "loss": 0.6702,
+      "step": 7989
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 36.073188130062924,
+      "learning_rate": 1.3152013060367428e-05,
+      "loss": 0.6724,
+      "step": 7990
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 35.11859146383809,
+      "learning_rate": 1.3150412581044446e-05,
+      "loss": 0.7315,
+      "step": 7991
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.808756282316146,
+      "learning_rate": 1.314881201212557e-05,
+      "loss": 0.6863,
+      "step": 7992
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.19311015737223,
+      "learning_rate": 1.3147211353656327e-05,
+      "loss": 0.6419,
+      "step": 7993
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 30.081265063176048,
+      "learning_rate": 1.314561060568223e-05,
+      "loss": 0.6381,
+      "step": 7994
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.489448969164606,
+      "learning_rate": 1.3144009768248808e-05,
+      "loss": 0.7452,
+      "step": 7995
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 42.024504325162894,
+      "learning_rate": 1.3142408841401586e-05,
+      "loss": 0.6279,
+      "step": 7996
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 39.96190720739866,
+      "learning_rate": 1.3140807825186098e-05,
+      "loss": 0.688,
+      "step": 7997
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 39.93444243593385,
+      "learning_rate": 1.313920671964787e-05,
+      "loss": 0.724,
+      "step": 7998
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.423100025979124,
+      "learning_rate": 1.3137605524832442e-05,
+      "loss": 0.688,
+      "step": 7999
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 40.324553516657176,
+      "learning_rate": 1.3136004240785348e-05,
+      "loss": 0.6755,
+      "step": 8000
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.678664535937756,
+      "learning_rate": 1.3134402867552125e-05,
+      "loss": 0.7316,
+      "step": 8001
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 40.40464851822727,
+      "learning_rate": 1.3132801405178322e-05,
+      "loss": 0.7554,
+      "step": 8002
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 29.859889187220734,
+      "learning_rate": 1.3131199853709477e-05,
+      "loss": 0.6221,
+      "step": 8003
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.46208703751911,
+      "learning_rate": 1.3129598213191139e-05,
+      "loss": 0.5966,
+      "step": 8004
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.91559449674953,
+      "learning_rate": 1.3127996483668862e-05,
+      "loss": 0.7009,
+      "step": 8005
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 32.38952539162791,
+      "learning_rate": 1.3126394665188191e-05,
+      "loss": 0.6912,
+      "step": 8006
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 37.459036878118106,
+      "learning_rate": 1.3124792757794687e-05,
+      "loss": 0.635,
+      "step": 8007
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 31.357988114948455,
+      "learning_rate": 1.3123190761533903e-05,
+      "loss": 0.6631,
+      "step": 8008
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 35.10681545704747,
+      "learning_rate": 1.3121588676451405e-05,
+      "loss": 0.735,
+      "step": 8009
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 39.32029412760117,
+      "learning_rate": 1.3119986502592747e-05,
+      "loss": 0.6836,
+      "step": 8010
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 28.856297263292024,
+      "learning_rate": 1.3118384240003498e-05,
+      "loss": 0.6566,
+      "step": 8011
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 47.7217459347106,
+      "learning_rate": 1.311678188872923e-05,
+      "loss": 0.8059,
+      "step": 8012
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 27.105864019902445,
+      "learning_rate": 1.3115179448815503e-05,
+      "loss": 0.5819,
+      "step": 8013
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 44.27071312323312,
+      "learning_rate": 1.3113576920307896e-05,
+      "loss": 0.6836,
+      "step": 8014
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 31.23359865079641,
+      "learning_rate": 1.3111974303251984e-05,
+      "loss": 0.6113,
+      "step": 8015
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 40.566667186942894,
+      "learning_rate": 1.3110371597693344e-05,
+      "loss": 0.6084,
+      "step": 8016
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 30.247505412215947,
+      "learning_rate": 1.3108768803677554e-05,
+      "loss": 0.6399,
+      "step": 8017
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.77958983703197,
+      "learning_rate": 1.3107165921250197e-05,
+      "loss": 0.603,
+      "step": 8018
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 35.82595714826464,
+      "learning_rate": 1.3105562950456857e-05,
+      "loss": 0.7127,
+      "step": 8019
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 37.97069728551647,
+      "learning_rate": 1.310395989134313e-05,
+      "loss": 0.687,
+      "step": 8020
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.2038176541795,
+      "learning_rate": 1.3102356743954594e-05,
+      "loss": 0.6472,
+      "step": 8021
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 33.06215577766628,
+      "learning_rate": 1.3100753508336852e-05,
+      "loss": 0.7033,
+      "step": 8022
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 37.329704551442205,
+      "learning_rate": 1.3099150184535487e-05,
+      "loss": 0.7105,
+      "step": 8023
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 57.98441751571573,
+      "learning_rate": 1.309754677259611e-05,
+      "loss": 0.5397,
+      "step": 8024
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 33.01429514050342,
+      "learning_rate": 1.3095943272564312e-05,
+      "loss": 0.6764,
+      "step": 8025
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.97917688690333,
+      "learning_rate": 1.3094339684485698e-05,
+      "loss": 0.6821,
+      "step": 8026
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 36.77666903792063,
+      "learning_rate": 1.3092736008405875e-05,
+      "loss": 0.7052,
+      "step": 8027
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.98207954007938,
+      "learning_rate": 1.3091132244370449e-05,
+      "loss": 0.6257,
+      "step": 8028
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 44.243110917895876,
+      "learning_rate": 1.3089528392425029e-05,
+      "loss": 0.8342,
+      "step": 8029
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 35.39862495933439,
+      "learning_rate": 1.3087924452615228e-05,
+      "loss": 0.6332,
+      "step": 8030
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 28.9138360023646,
+      "learning_rate": 1.3086320424986662e-05,
+      "loss": 0.6239,
+      "step": 8031
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 42.05188640270875,
+      "learning_rate": 1.308471630958495e-05,
+      "loss": 0.6807,
+      "step": 8032
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 31.96479045700783,
+      "learning_rate": 1.308311210645571e-05,
+      "loss": 0.6282,
+      "step": 8033
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.88592326101208,
+      "learning_rate": 1.3081507815644567e-05,
+      "loss": 0.6725,
+      "step": 8034
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.25507517390445,
+      "learning_rate": 1.3079903437197143e-05,
+      "loss": 0.7063,
+      "step": 8035
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 43.88662166949724,
+      "learning_rate": 1.3078298971159067e-05,
+      "loss": 0.7068,
+      "step": 8036
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.880191631900786,
+      "learning_rate": 1.307669441757597e-05,
+      "loss": 0.6807,
+      "step": 8037
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.387027165182374,
+      "learning_rate": 1.3075089776493481e-05,
+      "loss": 0.7248,
+      "step": 8038
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 28.924507192388212,
+      "learning_rate": 1.3073485047957242e-05,
+      "loss": 0.6124,
+      "step": 8039
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.59721903474591,
+      "learning_rate": 1.3071880232012881e-05,
+      "loss": 0.5956,
+      "step": 8040
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 29.23743507317701,
+      "learning_rate": 1.3070275328706047e-05,
+      "loss": 0.5654,
+      "step": 8041
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.117462289872776,
+      "learning_rate": 1.306867033808238e-05,
+      "loss": 0.6633,
+      "step": 8042
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 31.71229913818598,
+      "learning_rate": 1.306706526018752e-05,
+      "loss": 0.6492,
+      "step": 8043
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 33.50467759897073,
+      "learning_rate": 1.306546009506712e-05,
+      "loss": 0.574,
+      "step": 8044
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.52028699031496,
+      "learning_rate": 1.3063854842766828e-05,
+      "loss": 0.6575,
+      "step": 8045
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.40839946438173,
+      "learning_rate": 1.3062249503332294e-05,
+      "loss": 0.5869,
+      "step": 8046
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 40.139959594865886,
+      "learning_rate": 1.3060644076809178e-05,
+      "loss": 0.6875,
+      "step": 8047
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.13308549955219,
+      "learning_rate": 1.3059038563243134e-05,
+      "loss": 0.6794,
+      "step": 8048
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 61.15879467955932,
+      "learning_rate": 1.3057432962679823e-05,
+      "loss": 0.7729,
+      "step": 8049
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.96625759394739,
+      "learning_rate": 1.3055827275164908e-05,
+      "loss": 0.6877,
+      "step": 8050
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.09046819463141,
+      "learning_rate": 1.305422150074405e-05,
+      "loss": 0.6281,
+      "step": 8051
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.57786003180201,
+      "learning_rate": 1.3052615639462923e-05,
+      "loss": 0.699,
+      "step": 8052
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.66029775246484,
+      "learning_rate": 1.3051009691367191e-05,
+      "loss": 0.636,
+      "step": 8053
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.36979434590435,
+      "learning_rate": 1.304940365650253e-05,
+      "loss": 0.786,
+      "step": 8054
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 40.52773995005884,
+      "learning_rate": 1.3047797534914609e-05,
+      "loss": 0.8363,
+      "step": 8055
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 40.73302492890199,
+      "learning_rate": 1.3046191326649114e-05,
+      "loss": 0.7434,
+      "step": 8056
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.606138907127885,
+      "learning_rate": 1.3044585031751716e-05,
+      "loss": 0.7816,
+      "step": 8057
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.78614398379488,
+      "learning_rate": 1.3042978650268102e-05,
+      "loss": 0.6807,
+      "step": 8058
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.233697933903784,
+      "learning_rate": 1.3041372182243955e-05,
+      "loss": 0.7704,
+      "step": 8059
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 28.144010595393937,
+      "learning_rate": 1.3039765627724961e-05,
+      "loss": 0.6535,
+      "step": 8060
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 44.96155742413427,
+      "learning_rate": 1.3038158986756812e-05,
+      "loss": 0.7203,
+      "step": 8061
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 32.98479138666728,
+      "learning_rate": 1.30365522593852e-05,
+      "loss": 0.6723,
+      "step": 8062
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.342789054214336,
+      "learning_rate": 1.3034945445655813e-05,
+      "loss": 0.6431,
+      "step": 8063
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.47814265560963,
+      "learning_rate": 1.303333854561436e-05,
+      "loss": 0.612,
+      "step": 8064
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.72145200043733,
+      "learning_rate": 1.3031731559306528e-05,
+      "loss": 0.6614,
+      "step": 8065
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.38916387578139,
+      "learning_rate": 1.3030124486778029e-05,
+      "loss": 0.6866,
+      "step": 8066
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 42.210024659832605,
+      "learning_rate": 1.3028517328074557e-05,
+      "loss": 0.7028,
+      "step": 8067
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.91700899834282,
+      "learning_rate": 1.3026910083241825e-05,
+      "loss": 0.681,
+      "step": 8068
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.92881925694448,
+      "learning_rate": 1.3025302752325544e-05,
+      "loss": 0.6894,
+      "step": 8069
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 32.19106514824881,
+      "learning_rate": 1.302369533537142e-05,
+      "loss": 0.6779,
+      "step": 8070
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 47.659209042595926,
+      "learning_rate": 1.302208783242517e-05,
+      "loss": 0.7419,
+      "step": 8071
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.86032430918679,
+      "learning_rate": 1.302048024353251e-05,
+      "loss": 0.6616,
+      "step": 8072
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 33.14211929527927,
+      "learning_rate": 1.3018872568739158e-05,
+      "loss": 0.5684,
+      "step": 8073
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 36.88607298877476,
+      "learning_rate": 1.3017264808090834e-05,
+      "loss": 0.7226,
+      "step": 8074
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 45.05310393954284,
+      "learning_rate": 1.3015656961633267e-05,
+      "loss": 0.7656,
+      "step": 8075
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.61516803282448,
+      "learning_rate": 1.301404902941218e-05,
+      "loss": 0.6691,
+      "step": 8076
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.279212821582234,
+      "learning_rate": 1.3012441011473302e-05,
+      "loss": 0.6978,
+      "step": 8077
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 40.70346311110282,
+      "learning_rate": 1.3010832907862363e-05,
+      "loss": 0.7501,
+      "step": 8078
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.751027099779456,
+      "learning_rate": 1.30092247186251e-05,
+      "loss": 0.6324,
+      "step": 8079
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.41384104115561,
+      "learning_rate": 1.300761644380724e-05,
+      "loss": 0.7909,
+      "step": 8080
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 40.837635834756696,
+      "learning_rate": 1.3006008083454535e-05,
+      "loss": 0.7652,
+      "step": 8081
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 43.09818479583299,
+      "learning_rate": 1.3004399637612712e-05,
+      "loss": 0.6529,
+      "step": 8082
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 47.781347999447306,
+      "learning_rate": 1.3002791106327528e-05,
+      "loss": 0.8129,
+      "step": 8083
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.11746183836966,
+      "learning_rate": 1.3001182489644715e-05,
+      "loss": 0.6894,
+      "step": 8084
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 43.578527383734944,
+      "learning_rate": 1.2999573787610031e-05,
+      "loss": 0.6029,
+      "step": 8085
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.07756588077957,
+      "learning_rate": 1.2997965000269223e-05,
+      "loss": 0.6773,
+      "step": 8086
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.6533780152827,
+      "learning_rate": 1.2996356127668044e-05,
+      "loss": 0.6042,
+      "step": 8087
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.41886731228062,
+      "learning_rate": 1.2994747169852249e-05,
+      "loss": 0.7178,
+      "step": 8088
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 31.969382001087002,
+      "learning_rate": 1.2993138126867597e-05,
+      "loss": 0.6232,
+      "step": 8089
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 34.41673117085863,
+      "learning_rate": 1.2991528998759844e-05,
+      "loss": 0.7477,
+      "step": 8090
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.755138774821766,
+      "learning_rate": 1.2989919785574764e-05,
+      "loss": 0.6573,
+      "step": 8091
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 36.46149516573186,
+      "learning_rate": 1.2988310487358107e-05,
+      "loss": 0.6551,
+      "step": 8092
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 31.954243255671347,
+      "learning_rate": 1.298670110415565e-05,
+      "loss": 0.7597,
+      "step": 8093
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 33.8697121670231,
+      "learning_rate": 1.2985091636013163e-05,
+      "loss": 0.6975,
+      "step": 8094
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 41.03072782185661,
+      "learning_rate": 1.2983482082976412e-05,
+      "loss": 0.6672,
+      "step": 8095
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 36.1181577782795,
+      "learning_rate": 1.2981872445091179e-05,
+      "loss": 0.6744,
+      "step": 8096
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.738699714933176,
+      "learning_rate": 1.2980262722403238e-05,
+      "loss": 0.7019,
+      "step": 8097
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 33.5776339637493,
+      "learning_rate": 1.297865291495837e-05,
+      "loss": 0.7023,
+      "step": 8098
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 51.086567649799285,
+      "learning_rate": 1.2977043022802351e-05,
+      "loss": 0.7693,
+      "step": 8099
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 41.309607010627104,
+      "learning_rate": 1.2975433045980975e-05,
+      "loss": 0.7145,
+      "step": 8100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 30.78763586890178,
+      "learning_rate": 1.2973822984540018e-05,
+      "loss": 0.647,
+      "step": 8101
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 34.165950493556124,
+      "learning_rate": 1.2972212838525278e-05,
+      "loss": 0.682,
+      "step": 8102
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 45.069290723507486,
+      "learning_rate": 1.2970602607982546e-05,
+      "loss": 0.7389,
+      "step": 8103
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 39.432970890229356,
+      "learning_rate": 1.2968992292957609e-05,
+      "loss": 0.6821,
+      "step": 8104
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 32.845241625276415,
+      "learning_rate": 1.296738189349627e-05,
+      "loss": 0.6077,
+      "step": 8105
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 43.59307795180904,
+      "learning_rate": 1.2965771409644328e-05,
+      "loss": 0.7184,
+      "step": 8106
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 32.671642081891285,
+      "learning_rate": 1.2964160841447577e-05,
+      "loss": 0.688,
+      "step": 8107
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 57.67562674176493,
+      "learning_rate": 1.2962550188951832e-05,
+      "loss": 0.6752,
+      "step": 8108
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.88893791842426,
+      "learning_rate": 1.2960939452202888e-05,
+      "loss": 0.571,
+      "step": 8109
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 32.973839022650985,
+      "learning_rate": 1.2959328631246562e-05,
+      "loss": 0.6064,
+      "step": 8110
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 44.97938353554109,
+      "learning_rate": 1.2957717726128656e-05,
+      "loss": 0.706,
+      "step": 8111
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.846579853124915,
+      "learning_rate": 1.295610673689499e-05,
+      "loss": 0.6535,
+      "step": 8112
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 28.88400854224726,
+      "learning_rate": 1.295449566359138e-05,
+      "loss": 0.6829,
+      "step": 8113
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 30.19558979810187,
+      "learning_rate": 1.2952884506263641e-05,
+      "loss": 0.5853,
+      "step": 8114
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 31.77401164473861,
+      "learning_rate": 1.2951273264957591e-05,
+      "loss": 0.5288,
+      "step": 8115
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.23141867160936,
+      "learning_rate": 1.2949661939719059e-05,
+      "loss": 0.6803,
+      "step": 8116
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.50756666058658,
+      "learning_rate": 1.2948050530593863e-05,
+      "loss": 0.6067,
+      "step": 8117
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 41.33750026923203,
+      "learning_rate": 1.2946439037627838e-05,
+      "loss": 0.7188,
+      "step": 8118
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 57.43803202535236,
+      "learning_rate": 1.2944827460866809e-05,
+      "loss": 0.7362,
+      "step": 8119
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 33.9164771622246,
+      "learning_rate": 1.294321580035661e-05,
+      "loss": 0.5962,
+      "step": 8120
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 34.917105421998386,
+      "learning_rate": 1.2941604056143078e-05,
+      "loss": 0.6997,
+      "step": 8121
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 32.056740751059564,
+      "learning_rate": 1.2939992228272043e-05,
+      "loss": 0.6076,
+      "step": 8122
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 49.52735284341427,
+      "learning_rate": 1.2938380316789352e-05,
+      "loss": 0.6613,
+      "step": 8123
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 38.48320773449625,
+      "learning_rate": 1.293676832174084e-05,
+      "loss": 0.7376,
+      "step": 8124
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 45.77257795254406,
+      "learning_rate": 1.293515624317236e-05,
+      "loss": 0.6956,
+      "step": 8125
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 34.509794178760416,
+      "learning_rate": 1.293354408112975e-05,
+      "loss": 0.6668,
+      "step": 8126
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 34.24167480363437,
+      "learning_rate": 1.2931931835658864e-05,
+      "loss": 0.5377,
+      "step": 8127
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 34.763817935227195,
+      "learning_rate": 1.293031950680555e-05,
+      "loss": 0.6857,
+      "step": 8128
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 35.98338780309935,
+      "learning_rate": 1.2928707094615665e-05,
+      "loss": 0.6509,
+      "step": 8129
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 31.923067911781406,
+      "learning_rate": 1.2927094599135064e-05,
+      "loss": 0.6345,
+      "step": 8130
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 33.253110438973636,
+      "learning_rate": 1.2925482020409604e-05,
+      "loss": 0.6962,
+      "step": 8131
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 35.20818976122835,
+      "learning_rate": 1.2923869358485147e-05,
+      "loss": 0.6591,
+      "step": 8132
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 39.43105716038408,
+      "learning_rate": 1.2922256613407553e-05,
+      "loss": 0.6456,
+      "step": 8133
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 35.23308486437335,
+      "learning_rate": 1.292064378522269e-05,
+      "loss": 0.713,
+      "step": 8134
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 33.55123744010015,
+      "learning_rate": 1.2919030873976433e-05,
+      "loss": 0.6119,
+      "step": 8135
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 27.615190840093998,
+      "learning_rate": 1.291741787971464e-05,
+      "loss": 0.5912,
+      "step": 8136
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.3733829519021,
+      "learning_rate": 1.2915804802483192e-05,
+      "loss": 0.6855,
+      "step": 8137
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 32.39560823065919,
+      "learning_rate": 1.2914191642327958e-05,
+      "loss": 0.6208,
+      "step": 8138
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 40.966371806216245,
+      "learning_rate": 1.2912578399294822e-05,
+      "loss": 0.7116,
+      "step": 8139
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 33.87080262984773,
+      "learning_rate": 1.2910965073429657e-05,
+      "loss": 0.6897,
+      "step": 8140
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 34.8133110349134,
+      "learning_rate": 1.2909351664778352e-05,
+      "loss": 0.6421,
+      "step": 8141
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 38.161279895119385,
+      "learning_rate": 1.2907738173386783e-05,
+      "loss": 0.6462,
+      "step": 8142
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 37.263926213408844,
+      "learning_rate": 1.2906124599300842e-05,
+      "loss": 0.6794,
+      "step": 8143
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 29.219029866882483,
+      "learning_rate": 1.2904510942566421e-05,
+      "loss": 0.6064,
+      "step": 8144
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 33.648463695366246,
+      "learning_rate": 1.2902897203229406e-05,
+      "loss": 0.6365,
+      "step": 8145
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 37.629205535233865,
+      "learning_rate": 1.2901283381335691e-05,
+      "loss": 0.714,
+      "step": 8146
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 43.55716527622266,
+      "learning_rate": 1.2899669476931176e-05,
+      "loss": 0.6746,
+      "step": 8147
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 40.11483795803701,
+      "learning_rate": 1.2898055490061757e-05,
+      "loss": 0.6608,
+      "step": 8148
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 40.57994584326705,
+      "learning_rate": 1.2896441420773335e-05,
+      "loss": 0.6515,
+      "step": 8149
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.91881467233305,
+      "learning_rate": 1.2894827269111811e-05,
+      "loss": 0.7293,
+      "step": 8150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 42.21567147789123,
+      "learning_rate": 1.2893213035123096e-05,
+      "loss": 0.7968,
+      "step": 8151
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 39.80099647486993,
+      "learning_rate": 1.2891598718853096e-05,
+      "loss": 0.71,
+      "step": 8152
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 45.510586214466414,
+      "learning_rate": 1.2889984320347715e-05,
+      "loss": 0.6671,
+      "step": 8153
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 33.36629545805273,
+      "learning_rate": 1.2888369839652875e-05,
+      "loss": 0.6496,
+      "step": 8154
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.29559459882924,
+      "learning_rate": 1.2886755276814484e-05,
+      "loss": 0.6667,
+      "step": 8155
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 41.404364468810726,
+      "learning_rate": 1.2885140631878463e-05,
+      "loss": 0.7432,
+      "step": 8156
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 59.46940244395958,
+      "learning_rate": 1.2883525904890728e-05,
+      "loss": 0.6701,
+      "step": 8157
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 37.563293550204925,
+      "learning_rate": 1.2881911095897206e-05,
+      "loss": 0.7013,
+      "step": 8158
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 36.41063084583107,
+      "learning_rate": 1.2880296204943816e-05,
+      "loss": 0.6523,
+      "step": 8159
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 48.00803929547716,
+      "learning_rate": 1.287868123207649e-05,
+      "loss": 0.6792,
+      "step": 8160
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 44.777359643737164,
+      "learning_rate": 1.2877066177341148e-05,
+      "loss": 0.6976,
+      "step": 8161
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.589853155431065,
+      "learning_rate": 1.287545104078373e-05,
+      "loss": 0.5573,
+      "step": 8162
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 31.662383599035344,
+      "learning_rate": 1.2873835822450168e-05,
+      "loss": 0.6419,
+      "step": 8163
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 48.036791197019106,
+      "learning_rate": 1.2872220522386396e-05,
+      "loss": 0.7041,
+      "step": 8164
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 30.70290889518338,
+      "learning_rate": 1.2870605140638352e-05,
+      "loss": 0.6493,
+      "step": 8165
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.87271827937939,
+      "learning_rate": 1.2868989677251976e-05,
+      "loss": 0.6551,
+      "step": 8166
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.207843877568635,
+      "learning_rate": 1.2867374132273214e-05,
+      "loss": 0.7153,
+      "step": 8167
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 32.7093192958239,
+      "learning_rate": 1.2865758505748006e-05,
+      "loss": 0.6431,
+      "step": 8168
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.885654840110675,
+      "learning_rate": 1.2864142797722305e-05,
+      "loss": 0.6908,
+      "step": 8169
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 32.495083253586245,
+      "learning_rate": 1.2862527008242056e-05,
+      "loss": 0.6398,
+      "step": 8170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.153125620041,
+      "learning_rate": 1.2860911137353219e-05,
+      "loss": 0.6562,
+      "step": 8171
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 43.66009885441698,
+      "learning_rate": 1.2859295185101736e-05,
+      "loss": 0.6869,
+      "step": 8172
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 61.992290325115576,
+      "learning_rate": 1.2857679151533576e-05,
+      "loss": 0.6166,
+      "step": 8173
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.442093811925716,
+      "learning_rate": 1.2856063036694688e-05,
+      "loss": 0.6679,
+      "step": 8174
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.18689573456849,
+      "learning_rate": 1.2854446840631041e-05,
+      "loss": 0.6569,
+      "step": 8175
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.98706459730725,
+      "learning_rate": 1.2852830563388597e-05,
+      "loss": 0.6179,
+      "step": 8176
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 44.85787379119033,
+      "learning_rate": 1.2851214205013317e-05,
+      "loss": 0.7139,
+      "step": 8177
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 30.189585287917005,
+      "learning_rate": 1.2849597765551171e-05,
+      "loss": 0.6091,
+      "step": 8178
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.48991815664857,
+      "learning_rate": 1.2847981245048137e-05,
+      "loss": 0.7493,
+      "step": 8179
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.456199849874274,
+      "learning_rate": 1.284636464355018e-05,
+      "loss": 0.6443,
+      "step": 8180
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 42.85347532247233,
+      "learning_rate": 1.2844747961103279e-05,
+      "loss": 0.6441,
+      "step": 8181
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 42.07532801010103,
+      "learning_rate": 1.2843131197753408e-05,
+      "loss": 0.7008,
+      "step": 8182
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.68762328667254,
+      "learning_rate": 1.2841514353546549e-05,
+      "loss": 0.7403,
+      "step": 8183
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.07359238564624,
+      "learning_rate": 1.2839897428528687e-05,
+      "loss": 0.6788,
+      "step": 8184
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 40.37597458729039,
+      "learning_rate": 1.2838280422745801e-05,
+      "loss": 0.7109,
+      "step": 8185
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 29.843626597284157,
+      "learning_rate": 1.283666333624388e-05,
+      "loss": 0.5745,
+      "step": 8186
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 37.63195494438332,
+      "learning_rate": 1.2835046169068912e-05,
+      "loss": 0.6798,
+      "step": 8187
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 45.16757660942741,
+      "learning_rate": 1.283342892126689e-05,
+      "loss": 0.6802,
+      "step": 8188
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.530956267061256,
+      "learning_rate": 1.283181159288381e-05,
+      "loss": 0.6566,
+      "step": 8189
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 33.77861945031874,
+      "learning_rate": 1.2830194183965663e-05,
+      "loss": 0.7028,
+      "step": 8190
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 33.90311633628659,
+      "learning_rate": 1.2828576694558447e-05,
+      "loss": 0.6155,
+      "step": 8191
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 28.983727761005618,
+      "learning_rate": 1.2826959124708167e-05,
+      "loss": 0.648,
+      "step": 8192
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 44.543038531172385,
+      "learning_rate": 1.2825341474460824e-05,
+      "loss": 0.7074,
+      "step": 8193
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 50.948608549208025,
+      "learning_rate": 1.2823723743862422e-05,
+      "loss": 0.8322,
+      "step": 8194
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.64659124275702,
+      "learning_rate": 1.2822105932958964e-05,
+      "loss": 0.707,
+      "step": 8195
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.10905190310493,
+      "learning_rate": 1.282048804179647e-05,
+      "loss": 0.641,
+      "step": 8196
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 33.0767814604836,
+      "learning_rate": 1.2818870070420942e-05,
+      "loss": 0.6275,
+      "step": 8197
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 40.39112373501605,
+      "learning_rate": 1.2817252018878401e-05,
+      "loss": 0.7006,
+      "step": 8198
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.37901976833472,
+      "learning_rate": 1.2815633887214857e-05,
+      "loss": 0.6104,
+      "step": 8199
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.71249331965644,
+      "learning_rate": 1.2814015675476336e-05,
+      "loss": 0.5979,
+      "step": 8200
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.414374194655416,
+      "learning_rate": 1.2812397383708854e-05,
+      "loss": 0.687,
+      "step": 8201
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.63877791116131,
+      "learning_rate": 1.2810779011958437e-05,
+      "loss": 0.6347,
+      "step": 8202
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 37.002275216539886,
+      "learning_rate": 1.2809160560271108e-05,
+      "loss": 0.7599,
+      "step": 8203
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.81410556717997,
+      "learning_rate": 1.2807542028692894e-05,
+      "loss": 0.6554,
+      "step": 8204
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 30.28056568518113,
+      "learning_rate": 1.2805923417269828e-05,
+      "loss": 0.6489,
+      "step": 8205
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.5369502364321,
+      "learning_rate": 1.2804304726047943e-05,
+      "loss": 0.6034,
+      "step": 8206
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.78456345496038,
+      "learning_rate": 1.2802685955073271e-05,
+      "loss": 0.7299,
+      "step": 8207
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.97544321719198,
+      "learning_rate": 1.280106710439185e-05,
+      "loss": 0.6349,
+      "step": 8208
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.59139004755178,
+      "learning_rate": 1.279944817404972e-05,
+      "loss": 0.7487,
+      "step": 8209
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 32.97478433137923,
+      "learning_rate": 1.279782916409292e-05,
+      "loss": 0.7243,
+      "step": 8210
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 35.549100523647425,
+      "learning_rate": 1.2796210074567499e-05,
+      "loss": 0.7035,
+      "step": 8211
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 40.353021964224716,
+      "learning_rate": 1.2794590905519494e-05,
+      "loss": 0.6826,
+      "step": 8212
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 29.632789619853337,
+      "learning_rate": 1.2792971656994963e-05,
+      "loss": 0.6442,
+      "step": 8213
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 45.34937555394205,
+      "learning_rate": 1.2791352329039948e-05,
+      "loss": 0.7343,
+      "step": 8214
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 49.95669969088816,
+      "learning_rate": 1.278973292170051e-05,
+      "loss": 0.7286,
+      "step": 8215
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.038218952029055,
+      "learning_rate": 1.2788113435022696e-05,
+      "loss": 0.6723,
+      "step": 8216
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.67595337861778,
+      "learning_rate": 1.2786493869052568e-05,
+      "loss": 0.6873,
+      "step": 8217
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.07489860931943,
+      "learning_rate": 1.2784874223836184e-05,
+      "loss": 0.7132,
+      "step": 8218
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 47.87131087628621,
+      "learning_rate": 1.2783254499419607e-05,
+      "loss": 0.7571,
+      "step": 8219
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 46.049187306057696,
+      "learning_rate": 1.2781634695848899e-05,
+      "loss": 0.7139,
+      "step": 8220
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 42.7139444110875,
+      "learning_rate": 1.278001481317013e-05,
+      "loss": 0.7033,
+      "step": 8221
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 36.29244432079682,
+      "learning_rate": 1.2778394851429361e-05,
+      "loss": 0.5871,
+      "step": 8222
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.301878653235235,
+      "learning_rate": 1.2776774810672674e-05,
+      "loss": 0.7829,
+      "step": 8223
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 32.95500701189437,
+      "learning_rate": 1.277515469094613e-05,
+      "loss": 0.6312,
+      "step": 8224
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.60350619257461,
+      "learning_rate": 1.2773534492295813e-05,
+      "loss": 0.6972,
+      "step": 8225
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 41.4200039228385,
+      "learning_rate": 1.2771914214767798e-05,
+      "loss": 0.6966,
+      "step": 8226
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 46.49713461261385,
+      "learning_rate": 1.2770293858408166e-05,
+      "loss": 0.7288,
+      "step": 8227
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 40.61915391379849,
+      "learning_rate": 1.2768673423262992e-05,
+      "loss": 0.8137,
+      "step": 8228
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 94.60037102414817,
+      "learning_rate": 1.2767052909378367e-05,
+      "loss": 0.7168,
+      "step": 8229
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 39.61285045657742,
+      "learning_rate": 1.2765432316800379e-05,
+      "loss": 0.7639,
+      "step": 8230
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 31.217918144346772,
+      "learning_rate": 1.276381164557511e-05,
+      "loss": 0.6141,
+      "step": 8231
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 35.721367881071274,
+      "learning_rate": 1.2762190895748657e-05,
+      "loss": 0.6903,
+      "step": 8232
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 27.1300512963389,
+      "learning_rate": 1.2760570067367111e-05,
+      "loss": 0.5874,
+      "step": 8233
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 29.35798794818176,
+      "learning_rate": 1.2758949160476567e-05,
+      "loss": 0.6022,
+      "step": 8234
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.96406170097273,
+      "learning_rate": 1.2757328175123122e-05,
+      "loss": 0.7311,
+      "step": 8235
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 37.95559012211435,
+      "learning_rate": 1.2755707111352877e-05,
+      "loss": 0.7079,
+      "step": 8236
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 35.174528370117216,
+      "learning_rate": 1.2754085969211932e-05,
+      "loss": 0.7147,
+      "step": 8237
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 32.62533348468453,
+      "learning_rate": 1.2752464748746395e-05,
+      "loss": 0.6705,
+      "step": 8238
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.93394076115241,
+      "learning_rate": 1.275084345000237e-05,
+      "loss": 0.6179,
+      "step": 8239
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.4815628340279,
+      "learning_rate": 1.2749222073025968e-05,
+      "loss": 0.7437,
+      "step": 8240
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.693693154924006,
+      "learning_rate": 1.2747600617863296e-05,
+      "loss": 0.7014,
+      "step": 8241
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 41.75451745360259,
+      "learning_rate": 1.2745979084560474e-05,
+      "loss": 0.7274,
+      "step": 8242
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.325271783167736,
+      "learning_rate": 1.274435747316361e-05,
+      "loss": 0.6114,
+      "step": 8243
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 37.92315381522277,
+      "learning_rate": 1.2742735783718825e-05,
+      "loss": 0.589,
+      "step": 8244
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.268042312569904,
+      "learning_rate": 1.2741114016272239e-05,
+      "loss": 0.6866,
+      "step": 8245
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 42.07819989435364,
+      "learning_rate": 1.2739492170869973e-05,
+      "loss": 0.6588,
+      "step": 8246
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 40.19930212390139,
+      "learning_rate": 1.2737870247558152e-05,
+      "loss": 0.7053,
+      "step": 8247
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 38.73502185892316,
+      "learning_rate": 1.2736248246382906e-05,
+      "loss": 0.7707,
+      "step": 8248
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.87893370863212,
+      "learning_rate": 1.2734626167390356e-05,
+      "loss": 0.6654,
+      "step": 8249
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.96670700800641,
+      "learning_rate": 1.2733004010626639e-05,
+      "loss": 0.7243,
+      "step": 8250
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 76.53868598997238,
+      "learning_rate": 1.273138177613789e-05,
+      "loss": 0.7457,
+      "step": 8251
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 40.11609753220969,
+      "learning_rate": 1.2729759463970238e-05,
+      "loss": 0.7021,
+      "step": 8252
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 32.94816711230897,
+      "learning_rate": 1.2728137074169824e-05,
+      "loss": 0.5641,
+      "step": 8253
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.068382077285264,
+      "learning_rate": 1.2726514606782785e-05,
+      "loss": 0.6426,
+      "step": 8254
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 31.06062976533693,
+      "learning_rate": 1.2724892061855271e-05,
+      "loss": 0.641,
+      "step": 8255
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 52.38650941142038,
+      "learning_rate": 1.2723269439433416e-05,
+      "loss": 0.7557,
+      "step": 8256
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 33.79006603713186,
+      "learning_rate": 1.2721646739563378e-05,
+      "loss": 0.6781,
+      "step": 8257
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.4115960370663,
+      "learning_rate": 1.2720023962291292e-05,
+      "loss": 0.68,
+      "step": 8258
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 39.845690302259854,
+      "learning_rate": 1.2718401107663318e-05,
+      "loss": 0.6917,
+      "step": 8259
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.8471378066311,
+      "learning_rate": 1.2716778175725606e-05,
+      "loss": 0.7047,
+      "step": 8260
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 30.998947131443327,
+      "learning_rate": 1.2715155166524313e-05,
+      "loss": 0.5781,
+      "step": 8261
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 33.219093884386425,
+      "learning_rate": 1.2713532080105595e-05,
+      "loss": 0.5824,
+      "step": 8262
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 47.549467098246325,
+      "learning_rate": 1.2711908916515613e-05,
+      "loss": 0.6,
+      "step": 8263
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 39.178452878688965,
+      "learning_rate": 1.2710285675800527e-05,
+      "loss": 0.6425,
+      "step": 8264
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 35.35637549920388,
+      "learning_rate": 1.2708662358006504e-05,
+      "loss": 0.6761,
+      "step": 8265
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 46.772829458662784,
+      "learning_rate": 1.2707038963179704e-05,
+      "loss": 0.8209,
+      "step": 8266
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 35.56568675185412,
+      "learning_rate": 1.2705415491366303e-05,
+      "loss": 0.6584,
+      "step": 8267
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 37.65717162591256,
+      "learning_rate": 1.2703791942612462e-05,
+      "loss": 0.7353,
+      "step": 8268
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 31.254806881278842,
+      "learning_rate": 1.2702168316964365e-05,
+      "loss": 0.679,
+      "step": 8269
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 40.886420961263,
+      "learning_rate": 1.2700544614468181e-05,
+      "loss": 0.7215,
+      "step": 8270
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.2710745844519,
+      "learning_rate": 1.2698920835170088e-05,
+      "loss": 0.6911,
+      "step": 8271
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.339615379444595,
+      "learning_rate": 1.2697296979116265e-05,
+      "loss": 0.6642,
+      "step": 8272
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 38.67581218367251,
+      "learning_rate": 1.2695673046352893e-05,
+      "loss": 0.7587,
+      "step": 8273
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 32.84574963537483,
+      "learning_rate": 1.2694049036926159e-05,
+      "loss": 0.6012,
+      "step": 8274
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 47.613926268407,
+      "learning_rate": 1.2692424950882242e-05,
+      "loss": 0.7714,
+      "step": 8275
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.79322277823394,
+      "learning_rate": 1.2690800788267338e-05,
+      "loss": 0.7735,
+      "step": 8276
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 33.33019728372434,
+      "learning_rate": 1.2689176549127633e-05,
+      "loss": 0.6716,
+      "step": 8277
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 39.62961774688383,
+      "learning_rate": 1.2687552233509322e-05,
+      "loss": 0.7486,
+      "step": 8278
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 27.99205763603321,
+      "learning_rate": 1.2685927841458593e-05,
+      "loss": 0.6006,
+      "step": 8279
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.67995519570845,
+      "learning_rate": 1.2684303373021651e-05,
+      "loss": 0.7665,
+      "step": 8280
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 33.074230278340224,
+      "learning_rate": 1.268267882824469e-05,
+      "loss": 0.6156,
+      "step": 8281
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 43.52598988559344,
+      "learning_rate": 1.2681054207173915e-05,
+      "loss": 0.7694,
+      "step": 8282
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 40.72762585365333,
+      "learning_rate": 1.2679429509855521e-05,
+      "loss": 0.7125,
+      "step": 8283
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 30.699551797136127,
+      "learning_rate": 1.2677804736335726e-05,
+      "loss": 0.6499,
+      "step": 8284
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 32.73715158854645,
+      "learning_rate": 1.2676179886660725e-05,
+      "loss": 0.6188,
+      "step": 8285
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 32.74830354235163,
+      "learning_rate": 1.2674554960876737e-05,
+      "loss": 0.6273,
+      "step": 8286
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 37.06493721550513,
+      "learning_rate": 1.2672929959029969e-05,
+      "loss": 0.6253,
+      "step": 8287
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.42842047437217,
+      "learning_rate": 1.2671304881166638e-05,
+      "loss": 0.6213,
+      "step": 8288
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.30200226155669,
+      "learning_rate": 1.2669679727332957e-05,
+      "loss": 0.5763,
+      "step": 8289
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 40.58251097444522,
+      "learning_rate": 1.2668054497575147e-05,
+      "loss": 0.722,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.606127111770824,
+      "learning_rate": 1.2666429191939428e-05,
+      "loss": 0.6235,
+      "step": 8291
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 38.29219310441515,
+      "learning_rate": 1.2664803810472025e-05,
+      "loss": 0.673,
+      "step": 8292
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 39.70355591283053,
+      "learning_rate": 1.2663178353219154e-05,
+      "loss": 0.6394,
+      "step": 8293
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 30.702214805565024,
+      "learning_rate": 1.2661552820227053e-05,
+      "loss": 0.6586,
+      "step": 8294
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 38.85437273715738,
+      "learning_rate": 1.2659927211541946e-05,
+      "loss": 0.6896,
+      "step": 8295
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.24837234284883,
+      "learning_rate": 1.2658301527210066e-05,
+      "loss": 0.6437,
+      "step": 8296
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 29.540340029723005,
+      "learning_rate": 1.2656675767277647e-05,
+      "loss": 0.6242,
+      "step": 8297
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 50.94889841620252,
+      "learning_rate": 1.2655049931790919e-05,
+      "loss": 0.7783,
+      "step": 8298
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 32.77927324355221,
+      "learning_rate": 1.2653424020796128e-05,
+      "loss": 0.5921,
+      "step": 8299
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 34.67124370397342,
+      "learning_rate": 1.2651798034339506e-05,
+      "loss": 0.6376,
+      "step": 8300
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 36.277374185414025,
+      "learning_rate": 1.2650171972467304e-05,
+      "loss": 0.7879,
+      "step": 8301
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 43.251653231737194,
+      "learning_rate": 1.2648545835225758e-05,
+      "loss": 0.7578,
+      "step": 8302
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 34.63329060312915,
+      "learning_rate": 1.2646919622661119e-05,
+      "loss": 0.6287,
+      "step": 8303
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 34.32990925961066,
+      "learning_rate": 1.2645293334819635e-05,
+      "loss": 0.6556,
+      "step": 8304
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 31.579405545823835,
+      "learning_rate": 1.2643666971747555e-05,
+      "loss": 0.6431,
+      "step": 8305
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 43.22335093647146,
+      "learning_rate": 1.2642040533491133e-05,
+      "loss": 0.6959,
+      "step": 8306
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 41.01007250886156,
+      "learning_rate": 1.2640414020096626e-05,
+      "loss": 0.7393,
+      "step": 8307
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 32.68223995169186,
+      "learning_rate": 1.2638787431610285e-05,
+      "loss": 0.649,
+      "step": 8308
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 40.67143609494872,
+      "learning_rate": 1.2637160768078376e-05,
+      "loss": 0.7422,
+      "step": 8309
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 41.10266717391725,
+      "learning_rate": 1.2635534029547155e-05,
+      "loss": 0.678,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.74019244853604,
+      "learning_rate": 1.263390721606289e-05,
+      "loss": 0.6837,
+      "step": 8311
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 43.13610577426303,
+      "learning_rate": 1.2632280327671842e-05,
+      "loss": 0.6163,
+      "step": 8312
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 43.44027511459068,
+      "learning_rate": 1.2630653364420285e-05,
+      "loss": 0.8096,
+      "step": 8313
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 50.71862864406355,
+      "learning_rate": 1.2629026326354481e-05,
+      "loss": 0.6242,
+      "step": 8314
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 31.522614255836196,
+      "learning_rate": 1.262739921352071e-05,
+      "loss": 0.5757,
+      "step": 8315
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 39.80900741061273,
+      "learning_rate": 1.2625772025965241e-05,
+      "loss": 0.7658,
+      "step": 8316
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.88787803716531,
+      "learning_rate": 1.262414476373435e-05,
+      "loss": 0.7294,
+      "step": 8317
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 72.10471250394873,
+      "learning_rate": 1.2622517426874322e-05,
+      "loss": 0.6911,
+      "step": 8318
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.15736650177504,
+      "learning_rate": 1.2620890015431427e-05,
+      "loss": 0.697,
+      "step": 8319
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.78337931065182,
+      "learning_rate": 1.2619262529451954e-05,
+      "loss": 0.5923,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 39.313191284991284,
+      "learning_rate": 1.2617634968982189e-05,
+      "loss": 0.6881,
+      "step": 8321
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.321292064082236,
+      "learning_rate": 1.2616007334068417e-05,
+      "loss": 0.676,
+      "step": 8322
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.10305757547227,
+      "learning_rate": 1.2614379624756924e-05,
+      "loss": 0.7709,
+      "step": 8323
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.47343696123743,
+      "learning_rate": 1.2612751841094006e-05,
+      "loss": 0.6679,
+      "step": 8324
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.11227815535407,
+      "learning_rate": 1.2611123983125951e-05,
+      "loss": 0.6794,
+      "step": 8325
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.453404976490376,
+      "learning_rate": 1.2609496050899061e-05,
+      "loss": 0.7038,
+      "step": 8326
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.80804765195237,
+      "learning_rate": 1.2607868044459629e-05,
+      "loss": 0.647,
+      "step": 8327
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.070359544218974,
+      "learning_rate": 1.2606239963853957e-05,
+      "loss": 0.6441,
+      "step": 8328
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.76466950804662,
+      "learning_rate": 1.260461180912834e-05,
+      "loss": 0.7447,
+      "step": 8329
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 53.54040037604658,
+      "learning_rate": 1.260298358032909e-05,
+      "loss": 0.6989,
+      "step": 8330
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 40.071083425869446,
+      "learning_rate": 1.260135527750251e-05,
+      "loss": 0.6872,
+      "step": 8331
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 32.37966916942,
+      "learning_rate": 1.2599726900694909e-05,
+      "loss": 0.6039,
+      "step": 8332
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 31.520255724969672,
+      "learning_rate": 1.2598098449952594e-05,
+      "loss": 0.6027,
+      "step": 8333
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.968237269216274,
+      "learning_rate": 1.2596469925321877e-05,
+      "loss": 0.6687,
+      "step": 8334
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.33663663440284,
+      "learning_rate": 1.2594841326849075e-05,
+      "loss": 0.6984,
+      "step": 8335
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.08630100905427,
+      "learning_rate": 1.2593212654580507e-05,
+      "loss": 0.6924,
+      "step": 8336
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 45.42770449639311,
+      "learning_rate": 1.2591583908562483e-05,
+      "loss": 0.7044,
+      "step": 8337
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 41.43843262257255,
+      "learning_rate": 1.2589955088841332e-05,
+      "loss": 0.6494,
+      "step": 8338
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 44.869277467030386,
+      "learning_rate": 1.2588326195463373e-05,
+      "loss": 0.7081,
+      "step": 8339
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 42.52811558235821,
+      "learning_rate": 1.2586697228474931e-05,
+      "loss": 0.732,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 30.327515955159722,
+      "learning_rate": 1.2585068187922333e-05,
+      "loss": 0.6337,
+      "step": 8341
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 48.03968108618442,
+      "learning_rate": 1.2583439073851907e-05,
+      "loss": 0.6635,
+      "step": 8342
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 80.24066516619534,
+      "learning_rate": 1.2581809886309989e-05,
+      "loss": 0.6994,
+      "step": 8343
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 44.7330573257984,
+      "learning_rate": 1.25801806253429e-05,
+      "loss": 0.7212,
+      "step": 8344
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 40.6704570040927,
+      "learning_rate": 1.2578551290996993e-05,
+      "loss": 0.7547,
+      "step": 8345
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.29114724869248,
+      "learning_rate": 1.2576921883318589e-05,
+      "loss": 0.7179,
+      "step": 8346
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.08990971704239,
+      "learning_rate": 1.2575292402354036e-05,
+      "loss": 0.6622,
+      "step": 8347
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 29.282686847620525,
+      "learning_rate": 1.2573662848149674e-05,
+      "loss": 0.6927,
+      "step": 8348
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.974467746475604,
+      "learning_rate": 1.2572033220751844e-05,
+      "loss": 0.6306,
+      "step": 8349
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 35.52517217423919,
+      "learning_rate": 1.2570403520206898e-05,
+      "loss": 0.6982,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 31.379620654458474,
+      "learning_rate": 1.2568773746561179e-05,
+      "loss": 0.7131,
+      "step": 8351
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 28.563158501394465,
+      "learning_rate": 1.2567143899861031e-05,
+      "loss": 0.6387,
+      "step": 8352
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 61.48985511564196,
+      "learning_rate": 1.2565513980152822e-05,
+      "loss": 0.6164,
+      "step": 8353
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 33.02033123914219,
+      "learning_rate": 1.2563883987482887e-05,
+      "loss": 0.7104,
+      "step": 8354
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.63752907869447,
+      "learning_rate": 1.2562253921897598e-05,
+      "loss": 0.69,
+      "step": 8355
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 32.65539636910788,
+      "learning_rate": 1.2560623783443302e-05,
+      "loss": 0.6389,
+      "step": 8356
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 45.26136663103541,
+      "learning_rate": 1.2558993572166365e-05,
+      "loss": 0.7698,
+      "step": 8357
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 47.64158491279208,
+      "learning_rate": 1.255736328811315e-05,
+      "loss": 0.7443,
+      "step": 8358
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 40.89745395938799,
+      "learning_rate": 1.2555732931330015e-05,
+      "loss": 0.6684,
+      "step": 8359
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 39.908756643386674,
+      "learning_rate": 1.2554102501863333e-05,
+      "loss": 0.6638,
+      "step": 8360
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 29.284273307020438,
+      "learning_rate": 1.2552471999759469e-05,
+      "loss": 0.555,
+      "step": 8361
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 40.253504813763946,
+      "learning_rate": 1.2550841425064795e-05,
+      "loss": 0.7307,
+      "step": 8362
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 43.68681270511698,
+      "learning_rate": 1.254921077782568e-05,
+      "loss": 0.7418,
+      "step": 8363
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.30267144879527,
+      "learning_rate": 1.2547580058088507e-05,
+      "loss": 0.6382,
+      "step": 8364
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.89985668185836,
+      "learning_rate": 1.2545949265899645e-05,
+      "loss": 0.6854,
+      "step": 8365
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 32.90924430330749,
+      "learning_rate": 1.2544318401305476e-05,
+      "loss": 0.6656,
+      "step": 8366
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 29.806224910335974,
+      "learning_rate": 1.2542687464352378e-05,
+      "loss": 0.6377,
+      "step": 8367
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.65637889456278,
+      "learning_rate": 1.2541056455086737e-05,
+      "loss": 0.6805,
+      "step": 8368
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 40.138106659516325,
+      "learning_rate": 1.2539425373554937e-05,
+      "loss": 0.6921,
+      "step": 8369
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.259570653639095,
+      "learning_rate": 1.2537794219803368e-05,
+      "loss": 0.7144,
+      "step": 8370
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 31.021683834774628,
+      "learning_rate": 1.2536162993878409e-05,
+      "loss": 0.593,
+      "step": 8371
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.56252662577244,
+      "learning_rate": 1.2534531695826466e-05,
+      "loss": 0.6532,
+      "step": 8372
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.07555288557444,
+      "learning_rate": 1.2532900325693917e-05,
+      "loss": 0.6885,
+      "step": 8373
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 32.55346937397485,
+      "learning_rate": 1.2531268883527168e-05,
+      "loss": 0.6547,
+      "step": 8374
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.76930364352248,
+      "learning_rate": 1.2529637369372614e-05,
+      "loss": 0.6635,
+      "step": 8375
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 36.24024319721876,
+      "learning_rate": 1.2528005783276652e-05,
+      "loss": 0.6637,
+      "step": 8376
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 30.77478991265563,
+      "learning_rate": 1.2526374125285683e-05,
+      "loss": 0.6299,
+      "step": 8377
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 39.236474075589015,
+      "learning_rate": 1.2524742395446112e-05,
+      "loss": 0.6546,
+      "step": 8378
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 31.10084056040309,
+      "learning_rate": 1.2523110593804344e-05,
+      "loss": 0.5485,
+      "step": 8379
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 37.1295357640261,
+      "learning_rate": 1.252147872040679e-05,
+      "loss": 0.6722,
+      "step": 8380
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 42.32260889530904,
+      "learning_rate": 1.251984677529985e-05,
+      "loss": 0.6973,
+      "step": 8381
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 53.00128624812792,
+      "learning_rate": 1.2518214758529947e-05,
+      "loss": 0.7163,
+      "step": 8382
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 29.875187192912986,
+      "learning_rate": 1.251658267014349e-05,
+      "loss": 0.5487,
+      "step": 8383
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.090018409971336,
+      "learning_rate": 1.2514950510186892e-05,
+      "loss": 0.6469,
+      "step": 8384
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.16420859046754,
+      "learning_rate": 1.2513318278706573e-05,
+      "loss": 0.6447,
+      "step": 8385
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.153186769759294,
+      "learning_rate": 1.2511685975748948e-05,
+      "loss": 0.683,
+      "step": 8386
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.46562306740823,
+      "learning_rate": 1.251005360136045e-05,
+      "loss": 0.6478,
+      "step": 8387
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 42.488207426407726,
+      "learning_rate": 1.2508421155587492e-05,
+      "loss": 0.7161,
+      "step": 8388
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 39.25354722874568,
+      "learning_rate": 1.2506788638476506e-05,
+      "loss": 0.6557,
+      "step": 8389
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.632014073712256,
+      "learning_rate": 1.2505156050073914e-05,
+      "loss": 0.5987,
+      "step": 8390
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 32.816621089268835,
+      "learning_rate": 1.2503523390426153e-05,
+      "loss": 0.6413,
+      "step": 8391
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 42.24501977720282,
+      "learning_rate": 1.2501890659579649e-05,
+      "loss": 0.6323,
+      "step": 8392
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.063231725815754,
+      "learning_rate": 1.250025785758084e-05,
+      "loss": 0.7244,
+      "step": 8393
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.5140212301511,
+      "learning_rate": 1.249862498447616e-05,
+      "loss": 0.6623,
+      "step": 8394
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 42.11897513208706,
+      "learning_rate": 1.2496992040312045e-05,
+      "loss": 0.6897,
+      "step": 8395
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 46.510795545013764,
+      "learning_rate": 1.2495359025134939e-05,
+      "loss": 0.7868,
+      "step": 8396
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 41.411884169255636,
+      "learning_rate": 1.2493725938991283e-05,
+      "loss": 0.6383,
+      "step": 8397
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 39.512100828834186,
+      "learning_rate": 1.2492092781927517e-05,
+      "loss": 0.7115,
+      "step": 8398
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 36.26653521051728,
+      "learning_rate": 1.2490459553990095e-05,
+      "loss": 0.6992,
+      "step": 8399
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.48860881614354,
+      "learning_rate": 1.2488826255225455e-05,
+      "loss": 0.6378,
+      "step": 8400
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.24493409766305,
+      "learning_rate": 1.2487192885680053e-05,
+      "loss": 0.6224,
+      "step": 8401
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.725970625571584,
+      "learning_rate": 1.2485559445400343e-05,
+      "loss": 0.6273,
+      "step": 8402
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 31.020349368557362,
+      "learning_rate": 1.2483925934432774e-05,
+      "loss": 0.654,
+      "step": 8403
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.4421092487468,
+      "learning_rate": 1.2482292352823806e-05,
+      "loss": 0.6816,
+      "step": 8404
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 36.791543996368084,
+      "learning_rate": 1.2480658700619891e-05,
+      "loss": 0.7088,
+      "step": 8405
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.71690099456154,
+      "learning_rate": 1.24790249778675e-05,
+      "loss": 0.673,
+      "step": 8406
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 32.86874495856418,
+      "learning_rate": 1.2477391184613086e-05,
+      "loss": 0.7009,
+      "step": 8407
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 37.74086741952126,
+      "learning_rate": 1.2475757320903117e-05,
+      "loss": 0.7121,
+      "step": 8408
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 39.9052685144183,
+      "learning_rate": 1.2474123386784059e-05,
+      "loss": 0.6409,
+      "step": 8409
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 52.10015619221593,
+      "learning_rate": 1.2472489382302377e-05,
+      "loss": 0.6538,
+      "step": 8410
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 31.16945414309042,
+      "learning_rate": 1.2470855307504544e-05,
+      "loss": 0.6251,
+      "step": 8411
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.69148761595701,
+      "learning_rate": 1.2469221162437033e-05,
+      "loss": 0.6976,
+      "step": 8412
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.83411340585333,
+      "learning_rate": 1.2467586947146313e-05,
+      "loss": 0.7112,
+      "step": 8413
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 45.514942319905316,
+      "learning_rate": 1.2465952661678866e-05,
+      "loss": 0.642,
+      "step": 8414
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 36.29728532970559,
+      "learning_rate": 1.2464318306081167e-05,
+      "loss": 0.6264,
+      "step": 8415
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 32.59255208258167,
+      "learning_rate": 1.24626838803997e-05,
+      "loss": 0.6992,
+      "step": 8416
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 45.60225418295597,
+      "learning_rate": 1.2461049384680938e-05,
+      "loss": 0.6535,
+      "step": 8417
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.934125582603926,
+      "learning_rate": 1.2459414818971376e-05,
+      "loss": 0.6935,
+      "step": 8418
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 30.571914299428155,
+      "learning_rate": 1.2457780183317496e-05,
+      "loss": 0.5402,
+      "step": 8419
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 42.44765375660488,
+      "learning_rate": 1.2456145477765782e-05,
+      "loss": 0.6348,
+      "step": 8420
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.792566380806946,
+      "learning_rate": 1.245451070236273e-05,
+      "loss": 0.5818,
+      "step": 8421
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.77690867552308,
+      "learning_rate": 1.2452875857154827e-05,
+      "loss": 0.6624,
+      "step": 8422
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 28.631378096265017,
+      "learning_rate": 1.2451240942188569e-05,
+      "loss": 0.5858,
+      "step": 8423
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 35.984329680771644,
+      "learning_rate": 1.2449605957510456e-05,
+      "loss": 0.7181,
+      "step": 8424
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 46.68721307464635,
+      "learning_rate": 1.244797090316698e-05,
+      "loss": 0.6533,
+      "step": 8425
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.734660126894184,
+      "learning_rate": 1.2446335779204647e-05,
+      "loss": 0.6562,
+      "step": 8426
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 29.30988752765219,
+      "learning_rate": 1.2444700585669952e-05,
+      "loss": 0.6473,
+      "step": 8427
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 37.6898988431614,
+      "learning_rate": 1.2443065322609402e-05,
+      "loss": 0.7637,
+      "step": 8428
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 35.313122738773835,
+      "learning_rate": 1.2441429990069507e-05,
+      "loss": 0.5991,
+      "step": 8429
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 39.10162223026313,
+      "learning_rate": 1.2439794588096767e-05,
+      "loss": 0.7154,
+      "step": 8430
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.27336120965824,
+      "learning_rate": 1.24381591167377e-05,
+      "loss": 0.6794,
+      "step": 8431
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 36.17373962210267,
+      "learning_rate": 1.243652357603881e-05,
+      "loss": 0.6838,
+      "step": 8432
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.43129913876248,
+      "learning_rate": 1.243488796604662e-05,
+      "loss": 0.6531,
+      "step": 8433
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 43.70226382735815,
+      "learning_rate": 1.2433252286807635e-05,
+      "loss": 0.7376,
+      "step": 8434
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.79043678881197,
+      "learning_rate": 1.2431616538368383e-05,
+      "loss": 0.5707,
+      "step": 8435
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 42.442994117437834,
+      "learning_rate": 1.2429980720775376e-05,
+      "loss": 0.6937,
+      "step": 8436
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 46.14638474294991,
+      "learning_rate": 1.2428344834075142e-05,
+      "loss": 0.7682,
+      "step": 8437
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 34.61483092003777,
+      "learning_rate": 1.2426708878314197e-05,
+      "loss": 0.6538,
+      "step": 8438
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.02239445785914,
+      "learning_rate": 1.2425072853539076e-05,
+      "loss": 0.625,
+      "step": 8439
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.964439034303574,
+      "learning_rate": 1.2423436759796296e-05,
+      "loss": 0.6131,
+      "step": 8440
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 37.414420574814216,
+      "learning_rate": 1.2421800597132399e-05,
+      "loss": 0.6173,
+      "step": 8441
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 26.55385297388162,
+      "learning_rate": 1.2420164365593903e-05,
+      "loss": 0.5362,
+      "step": 8442
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 43.69660406524508,
+      "learning_rate": 1.2418528065227354e-05,
+      "loss": 0.7274,
+      "step": 8443
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 40.87350085481174,
+      "learning_rate": 1.2416891696079276e-05,
+      "loss": 0.7119,
+      "step": 8444
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 49.47382329685497,
+      "learning_rate": 1.2415255258196215e-05,
+      "loss": 0.8367,
+      "step": 8445
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 34.76473692027579,
+      "learning_rate": 1.2413618751624708e-05,
+      "loss": 0.6277,
+      "step": 8446
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 37.20929120586889,
+      "learning_rate": 1.2411982176411294e-05,
+      "loss": 0.6824,
+      "step": 8447
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 36.96210240681417,
+      "learning_rate": 1.2410345532602518e-05,
+      "loss": 0.6895,
+      "step": 8448
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.45810359521252,
+      "learning_rate": 1.2408708820244926e-05,
+      "loss": 0.6061,
+      "step": 8449
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 36.931447980772596,
+      "learning_rate": 1.2407072039385064e-05,
+      "loss": 0.6629,
+      "step": 8450
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.49198964878022,
+      "learning_rate": 1.2405435190069481e-05,
+      "loss": 0.6442,
+      "step": 8451
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 44.641800979436894,
+      "learning_rate": 1.2403798272344729e-05,
+      "loss": 0.6953,
+      "step": 8452
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 43.14804760482042,
+      "learning_rate": 1.240216128625736e-05,
+      "loss": 0.7299,
+      "step": 8453
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 40.31785436453584,
+      "learning_rate": 1.2400524231853929e-05,
+      "loss": 0.7237,
+      "step": 8454
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 41.29591394490406,
+      "learning_rate": 1.2398887109180992e-05,
+      "loss": 0.694,
+      "step": 8455
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.876990129014494,
+      "learning_rate": 1.2397249918285113e-05,
+      "loss": 0.6111,
+      "step": 8456
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.80339459352918,
+      "learning_rate": 1.2395612659212844e-05,
+      "loss": 0.622,
+      "step": 8457
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.15605209536009,
+      "learning_rate": 1.2393975332010757e-05,
+      "loss": 0.6539,
+      "step": 8458
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.75306944841211,
+      "learning_rate": 1.2392337936725408e-05,
+      "loss": 0.6013,
+      "step": 8459
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 28.88164631411296,
+      "learning_rate": 1.239070047340337e-05,
+      "loss": 0.5488,
+      "step": 8460
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 37.39493090604195,
+      "learning_rate": 1.2389062942091209e-05,
+      "loss": 0.679,
+      "step": 8461
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 39.76231786619635,
+      "learning_rate": 1.2387425342835492e-05,
+      "loss": 0.7331,
+      "step": 8462
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.638891224395433,
+      "learning_rate": 1.2385787675682799e-05,
+      "loss": 0.5288,
+      "step": 8463
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.61123227175548,
+      "learning_rate": 1.2384149940679697e-05,
+      "loss": 0.6124,
+      "step": 8464
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 112.34465565959708,
+      "learning_rate": 1.2382512137872769e-05,
+      "loss": 0.6168,
+      "step": 8465
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 41.55784663487872,
+      "learning_rate": 1.2380874267308586e-05,
+      "loss": 0.6918,
+      "step": 8466
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 36.44357858930565,
+      "learning_rate": 1.237923632903373e-05,
+      "loss": 0.6364,
+      "step": 8467
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.440494569159643,
+      "learning_rate": 1.2377598323094788e-05,
+      "loss": 0.5745,
+      "step": 8468
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 48.54337965380868,
+      "learning_rate": 1.2375960249538341e-05,
+      "loss": 0.684,
+      "step": 8469
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 43.53375964245654,
+      "learning_rate": 1.2374322108410974e-05,
+      "loss": 0.6992,
+      "step": 8470
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 39.50514672646929,
+      "learning_rate": 1.2372683899759274e-05,
+      "loss": 0.7034,
+      "step": 8471
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 28.940159718692932,
+      "learning_rate": 1.2371045623629834e-05,
+      "loss": 0.6446,
+      "step": 8472
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 40.69610911875044,
+      "learning_rate": 1.2369407280069241e-05,
+      "loss": 0.6043,
+      "step": 8473
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 34.832521458036496,
+      "learning_rate": 1.2367768869124091e-05,
+      "loss": 0.6127,
+      "step": 8474
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.262465542413814,
+      "learning_rate": 1.2366130390840982e-05,
+      "loss": 0.6471,
+      "step": 8475
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 38.20555845908102,
+      "learning_rate": 1.2364491845266506e-05,
+      "loss": 0.7036,
+      "step": 8476
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 55.682073532113485,
+      "learning_rate": 1.2362853232447267e-05,
+      "loss": 0.7734,
+      "step": 8477
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 69.49387561105046,
+      "learning_rate": 1.2361214552429863e-05,
+      "loss": 0.6094,
+      "step": 8478
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 34.673360100838934,
+      "learning_rate": 1.23595758052609e-05,
+      "loss": 0.6859,
+      "step": 8479
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 39.41483265452298,
+      "learning_rate": 1.235793699098698e-05,
+      "loss": 0.6423,
+      "step": 8480
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.59263288902734,
+      "learning_rate": 1.2356298109654712e-05,
+      "loss": 0.6694,
+      "step": 8481
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 49.63043137045636,
+      "learning_rate": 1.2354659161310704e-05,
+      "loss": 0.7534,
+      "step": 8482
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.8821649050997,
+      "learning_rate": 1.2353020146001568e-05,
+      "loss": 0.7057,
+      "step": 8483
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 50.713408667792315,
+      "learning_rate": 1.2351381063773913e-05,
+      "loss": 0.6639,
+      "step": 8484
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 46.559326593696625,
+      "learning_rate": 1.2349741914674361e-05,
+      "loss": 0.6933,
+      "step": 8485
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 36.10276223166095,
+      "learning_rate": 1.2348102698749518e-05,
+      "loss": 0.7536,
+      "step": 8486
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 41.029288858814326,
+      "learning_rate": 1.234646341604601e-05,
+      "loss": 0.6448,
+      "step": 8487
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.809404615654515,
+      "learning_rate": 1.2344824066610454e-05,
+      "loss": 0.6679,
+      "step": 8488
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 39.217750837509094,
+      "learning_rate": 1.2343184650489476e-05,
+      "loss": 0.7022,
+      "step": 8489
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.17381463761773,
+      "learning_rate": 1.2341545167729693e-05,
+      "loss": 0.6552,
+      "step": 8490
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.58670953027245,
+      "learning_rate": 1.2339905618377739e-05,
+      "loss": 0.6716,
+      "step": 8491
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 36.615356073957116,
+      "learning_rate": 1.2338266002480237e-05,
+      "loss": 0.6064,
+      "step": 8492
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 40.15030386870324,
+      "learning_rate": 1.2336626320083816e-05,
+      "loss": 0.7195,
+      "step": 8493
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 41.68169829390208,
+      "learning_rate": 1.233498657123511e-05,
+      "loss": 0.708,
+      "step": 8494
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.25138595422095,
+      "learning_rate": 1.2333346755980753e-05,
+      "loss": 0.6535,
+      "step": 8495
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 35.88569541494387,
+      "learning_rate": 1.233170687436738e-05,
+      "loss": 0.7033,
+      "step": 8496
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.0220231027698,
+      "learning_rate": 1.2330066926441626e-05,
+      "loss": 0.7539,
+      "step": 8497
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 39.9092826736454,
+      "learning_rate": 1.232842691225013e-05,
+      "loss": 0.718,
+      "step": 8498
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 30.187231204808704,
+      "learning_rate": 1.2326786831839536e-05,
+      "loss": 0.5917,
+      "step": 8499
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 35.792764724937626,
+      "learning_rate": 1.2325146685256489e-05,
+      "loss": 0.6764,
+      "step": 8500
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.56179104323554,
+      "learning_rate": 1.2323506472547626e-05,
+      "loss": 0.5707,
+      "step": 8501
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.93608229500433,
+      "learning_rate": 1.2321866193759602e-05,
+      "loss": 0.6702,
+      "step": 8502
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 33.529733042252964,
+      "learning_rate": 1.2320225848939059e-05,
+      "loss": 0.6606,
+      "step": 8503
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 60.491895026360005,
+      "learning_rate": 1.2318585438132654e-05,
+      "loss": 0.7109,
+      "step": 8504
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 35.219501998441416,
+      "learning_rate": 1.2316944961387028e-05,
+      "loss": 0.6657,
+      "step": 8505
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 44.79749525989299,
+      "learning_rate": 1.2315304418748848e-05,
+      "loss": 0.6684,
+      "step": 8506
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.233574965473874,
+      "learning_rate": 1.2313663810264762e-05,
+      "loss": 0.574,
+      "step": 8507
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 47.790546907423355,
+      "learning_rate": 1.2312023135981434e-05,
+      "loss": 0.6761,
+      "step": 8508
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 35.17214751169795,
+      "learning_rate": 1.231038239594552e-05,
+      "loss": 0.6633,
+      "step": 8509
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 36.67673734880743,
+      "learning_rate": 1.230874159020368e-05,
+      "loss": 0.7627,
+      "step": 8510
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.427276651615838,
+      "learning_rate": 1.2307100718802579e-05,
+      "loss": 0.7052,
+      "step": 8511
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 39.201065112746825,
+      "learning_rate": 1.2305459781788885e-05,
+      "loss": 0.6426,
+      "step": 8512
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.39270893879417,
+      "learning_rate": 1.2303818779209264e-05,
+      "loss": 0.7236,
+      "step": 8513
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 36.01946478708173,
+      "learning_rate": 1.2302177711110384e-05,
+      "loss": 0.6821,
+      "step": 8514
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 39.31549069683585,
+      "learning_rate": 1.2300536577538917e-05,
+      "loss": 0.758,
+      "step": 8515
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.122092051685144,
+      "learning_rate": 1.2298895378541536e-05,
+      "loss": 0.6991,
+      "step": 8516
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.5179139092373,
+      "learning_rate": 1.2297254114164914e-05,
+      "loss": 0.6088,
+      "step": 8517
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 36.758672424051895,
+      "learning_rate": 1.2295612784455728e-05,
+      "loss": 0.6522,
+      "step": 8518
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.70939275542474,
+      "learning_rate": 1.2293971389460659e-05,
+      "loss": 0.5986,
+      "step": 8519
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.466696165288646,
+      "learning_rate": 1.2292329929226385e-05,
+      "loss": 0.6353,
+      "step": 8520
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 29.455955266176094,
+      "learning_rate": 1.229068840379959e-05,
+      "loss": 0.5829,
+      "step": 8521
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.20708767784025,
+      "learning_rate": 1.2289046813226954e-05,
+      "loss": 0.6192,
+      "step": 8522
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 41.438133085313964,
+      "learning_rate": 1.2287405157555168e-05,
+      "loss": 0.6042,
+      "step": 8523
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.9675015342382,
+      "learning_rate": 1.2285763436830917e-05,
+      "loss": 0.7823,
+      "step": 8524
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.70869426059433,
+      "learning_rate": 1.2284121651100891e-05,
+      "loss": 0.6458,
+      "step": 8525
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 58.26580534723206,
+      "learning_rate": 1.228247980041178e-05,
+      "loss": 0.7025,
+      "step": 8526
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.008369081947805,
+      "learning_rate": 1.2280837884810282e-05,
+      "loss": 0.6207,
+      "step": 8527
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 29.862434422104414,
+      "learning_rate": 1.2279195904343084e-05,
+      "loss": 0.6392,
+      "step": 8528
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.23548181051903,
+      "learning_rate": 1.2277553859056894e-05,
+      "loss": 0.6148,
+      "step": 8529
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.39844426167643,
+      "learning_rate": 1.22759117489984e-05,
+      "loss": 0.6283,
+      "step": 8530
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.4876595589188,
+      "learning_rate": 1.227426957421431e-05,
+      "loss": 0.6065,
+      "step": 8531
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.65001188343869,
+      "learning_rate": 1.227262733475132e-05,
+      "loss": 0.6828,
+      "step": 8532
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 48.41893034330586,
+      "learning_rate": 1.2270985030656139e-05,
+      "loss": 0.7146,
+      "step": 8533
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.84205416892779,
+      "learning_rate": 1.2269342661975474e-05,
+      "loss": 0.6699,
+      "step": 8534
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 45.49532045414816,
+      "learning_rate": 1.2267700228756033e-05,
+      "loss": 0.7629,
+      "step": 8535
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.58396626471622,
+      "learning_rate": 1.2266057731044521e-05,
+      "loss": 0.6262,
+      "step": 8536
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.98762412613101,
+      "learning_rate": 1.2264415168887651e-05,
+      "loss": 0.6708,
+      "step": 8537
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 40.31198357835345,
+      "learning_rate": 1.226277254233214e-05,
+      "loss": 0.6597,
+      "step": 8538
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 40.83559846563983,
+      "learning_rate": 1.2261129851424703e-05,
+      "loss": 0.6549,
+      "step": 8539
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 33.20067401052443,
+      "learning_rate": 1.2259487096212055e-05,
+      "loss": 0.7732,
+      "step": 8540
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 62.110923937790574,
+      "learning_rate": 1.2257844276740916e-05,
+      "loss": 0.6485,
+      "step": 8541
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.5280186034445,
+      "learning_rate": 1.2256201393058006e-05,
+      "loss": 0.6287,
+      "step": 8542
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 30.054117457648662,
+      "learning_rate": 1.2254558445210048e-05,
+      "loss": 0.6342,
+      "step": 8543
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 40.45480552120702,
+      "learning_rate": 1.2252915433243768e-05,
+      "loss": 0.7014,
+      "step": 8544
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.24623476614918,
+      "learning_rate": 1.225127235720589e-05,
+      "loss": 0.6206,
+      "step": 8545
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 46.02642144065246,
+      "learning_rate": 1.2249629217143143e-05,
+      "loss": 0.6889,
+      "step": 8546
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 35.53857000231507,
+      "learning_rate": 1.2247986013102258e-05,
+      "loss": 0.647,
+      "step": 8547
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 33.78923390604104,
+      "learning_rate": 1.2246342745129964e-05,
+      "loss": 0.7143,
+      "step": 8548
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 39.37649987406776,
+      "learning_rate": 1.2244699413272998e-05,
+      "loss": 0.6891,
+      "step": 8549
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 32.99453323050282,
+      "learning_rate": 1.2243056017578095e-05,
+      "loss": 0.6447,
+      "step": 8550
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 42.86536939633339,
+      "learning_rate": 1.2241412558091988e-05,
+      "loss": 0.7729,
+      "step": 8551
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.19824252213977,
+      "learning_rate": 1.2239769034861423e-05,
+      "loss": 0.6785,
+      "step": 8552
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 35.15072962504244,
+      "learning_rate": 1.2238125447933134e-05,
+      "loss": 0.7223,
+      "step": 8553
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 39.06266755640333,
+      "learning_rate": 1.2236481797353865e-05,
+      "loss": 0.7289,
+      "step": 8554
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 31.4843596579086,
+      "learning_rate": 1.2234838083170362e-05,
+      "loss": 0.6256,
+      "step": 8555
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 39.027941194752536,
+      "learning_rate": 1.2233194305429375e-05,
+      "loss": 0.6981,
+      "step": 8556
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.31719801259053,
+      "learning_rate": 1.2231550464177646e-05,
+      "loss": 0.66,
+      "step": 8557
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.40558892093777,
+      "learning_rate": 1.2229906559461925e-05,
+      "loss": 0.5885,
+      "step": 8558
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.76315529339024,
+      "learning_rate": 1.222826259132897e-05,
+      "loss": 0.6827,
+      "step": 8559
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 42.05508236147832,
+      "learning_rate": 1.2226618559825529e-05,
+      "loss": 0.6656,
+      "step": 8560
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.455376256219814,
+      "learning_rate": 1.222497446499836e-05,
+      "loss": 0.7386,
+      "step": 8561
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 47.78276579543892,
+      "learning_rate": 1.2223330306894215e-05,
+      "loss": 0.7115,
+      "step": 8562
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 33.45680741804778,
+      "learning_rate": 1.222168608555986e-05,
+      "loss": 0.6772,
+      "step": 8563
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 33.48288280243454,
+      "learning_rate": 1.222004180104205e-05,
+      "loss": 0.6988,
+      "step": 8564
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 42.840861717714894,
+      "learning_rate": 1.2218397453387551e-05,
+      "loss": 0.8004,
+      "step": 8565
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 56.24133535831953,
+      "learning_rate": 1.2216753042643128e-05,
+      "loss": 0.752,
+      "step": 8566
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 41.822655557820525,
+      "learning_rate": 1.2215108568855545e-05,
+      "loss": 0.7325,
+      "step": 8567
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 43.872891633545755,
+      "learning_rate": 1.2213464032071567e-05,
+      "loss": 0.7182,
+      "step": 8568
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.655084669493647,
+      "learning_rate": 1.221181943233797e-05,
+      "loss": 0.6526,
+      "step": 8569
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.35925232002796,
+      "learning_rate": 1.221017476970152e-05,
+      "loss": 0.6435,
+      "step": 8570
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.440247164779734,
+      "learning_rate": 1.2208530044208995e-05,
+      "loss": 0.6658,
+      "step": 8571
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.810576332677826,
+      "learning_rate": 1.2206885255907163e-05,
+      "loss": 0.6111,
+      "step": 8572
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 31.341172021977197,
+      "learning_rate": 1.2205240404842811e-05,
+      "loss": 0.7094,
+      "step": 8573
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.136730738422855,
+      "learning_rate": 1.2203595491062707e-05,
+      "loss": 0.6243,
+      "step": 8574
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 31.134567337644295,
+      "learning_rate": 1.2201950514613638e-05,
+      "loss": 0.6363,
+      "step": 8575
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.816106974171426,
+      "learning_rate": 1.2200305475542385e-05,
+      "loss": 0.6572,
+      "step": 8576
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 44.58860105636747,
+      "learning_rate": 1.2198660373895731e-05,
+      "loss": 0.8108,
+      "step": 8577
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.56959273770708,
+      "learning_rate": 1.2197015209720462e-05,
+      "loss": 0.6616,
+      "step": 8578
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 39.4353209133834,
+      "learning_rate": 1.2195369983063368e-05,
+      "loss": 0.6368,
+      "step": 8579
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 47.08399775832393,
+      "learning_rate": 1.2193724693971235e-05,
+      "loss": 0.7129,
+      "step": 8580
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 35.10843885368017,
+      "learning_rate": 1.2192079342490851e-05,
+      "loss": 0.6336,
+      "step": 8581
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.3236050588356,
+      "learning_rate": 1.2190433928669015e-05,
+      "loss": 0.6471,
+      "step": 8582
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.8190065602988,
+      "learning_rate": 1.218878845255252e-05,
+      "loss": 0.6308,
+      "step": 8583
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.76466195302916,
+      "learning_rate": 1.2187142914188161e-05,
+      "loss": 0.7686,
+      "step": 8584
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 32.41494311030414,
+      "learning_rate": 1.218549731362274e-05,
+      "loss": 0.6098,
+      "step": 8585
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.73217904463392,
+      "learning_rate": 1.218385165090305e-05,
+      "loss": 0.6689,
+      "step": 8586
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 30.17067232900739,
+      "learning_rate": 1.2182205926075899e-05,
+      "loss": 0.6356,
+      "step": 8587
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.666810864521565,
+      "learning_rate": 1.2180560139188088e-05,
+      "loss": 0.639,
+      "step": 8588
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 44.398013232216236,
+      "learning_rate": 1.217891429028642e-05,
+      "loss": 0.6484,
+      "step": 8589
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 36.27924219140184,
+      "learning_rate": 1.2177268379417708e-05,
+      "loss": 0.7046,
+      "step": 8590
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 56.70762296461074,
+      "learning_rate": 1.2175622406628754e-05,
+      "loss": 0.645,
+      "step": 8591
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 41.73298505598655,
+      "learning_rate": 1.2173976371966372e-05,
+      "loss": 0.7162,
+      "step": 8592
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.81863411054017,
+      "learning_rate": 1.2172330275477374e-05,
+      "loss": 0.6569,
+      "step": 8593
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 32.38387780364145,
+      "learning_rate": 1.2170684117208573e-05,
+      "loss": 0.5722,
+      "step": 8594
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.89841520175115,
+      "learning_rate": 1.2169037897206787e-05,
+      "loss": 0.7147,
+      "step": 8595
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 30.643979901908825,
+      "learning_rate": 1.2167391615518831e-05,
+      "loss": 0.6606,
+      "step": 8596
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.989170559666718,
+      "learning_rate": 1.2165745272191524e-05,
+      "loss": 0.6444,
+      "step": 8597
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 39.630515232202185,
+      "learning_rate": 1.2164098867271694e-05,
+      "loss": 0.7696,
+      "step": 8598
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.84689029481944,
+      "learning_rate": 1.2162452400806151e-05,
+      "loss": 0.6874,
+      "step": 8599
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 42.06642079891067,
+      "learning_rate": 1.216080587284173e-05,
+      "loss": 0.6529,
+      "step": 8600
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 45.45027899200119,
+      "learning_rate": 1.2159159283425254e-05,
+      "loss": 0.7064,
+      "step": 8601
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.43610608417789,
+      "learning_rate": 1.2157512632603553e-05,
+      "loss": 0.6534,
+      "step": 8602
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 35.54127211461681,
+      "learning_rate": 1.215586592042345e-05,
+      "loss": 0.7531,
+      "step": 8603
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.967233518034767,
+      "learning_rate": 1.2154219146931786e-05,
+      "loss": 0.5705,
+      "step": 8604
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.12582966811843,
+      "learning_rate": 1.2152572312175388e-05,
+      "loss": 0.6068,
+      "step": 8605
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 41.21400039065965,
+      "learning_rate": 1.2150925416201091e-05,
+      "loss": 0.6537,
+      "step": 8606
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 37.75564282286966,
+      "learning_rate": 1.2149278459055737e-05,
+      "loss": 0.6571,
+      "step": 8607
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 38.394346259441186,
+      "learning_rate": 1.2147631440786156e-05,
+      "loss": 0.616,
+      "step": 8608
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 50.594715120983956,
+      "learning_rate": 1.2145984361439197e-05,
+      "loss": 0.8723,
+      "step": 8609
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 32.03310598067132,
+      "learning_rate": 1.2144337221061697e-05,
+      "loss": 0.6821,
+      "step": 8610
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.772751379089705,
+      "learning_rate": 1.2142690019700503e-05,
+      "loss": 0.6403,
+      "step": 8611
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.93469384747355,
+      "learning_rate": 1.2141042757402454e-05,
+      "loss": 0.5844,
+      "step": 8612
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 36.179985956201946,
+      "learning_rate": 1.2139395434214406e-05,
+      "loss": 0.6556,
+      "step": 8613
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 40.19556301027759,
+      "learning_rate": 1.2137748050183197e-05,
+      "loss": 0.6411,
+      "step": 8614
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 39.07477032928917,
+      "learning_rate": 1.2136100605355691e-05,
+      "loss": 0.78,
+      "step": 8615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 31.617349042166357,
+      "learning_rate": 1.213445309977873e-05,
+      "loss": 0.6947,
+      "step": 8616
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.10521882371619,
+      "learning_rate": 1.2132805533499172e-05,
+      "loss": 0.6778,
+      "step": 8617
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 25.684275663772976,
+      "learning_rate": 1.213115790656387e-05,
+      "loss": 0.5319,
+      "step": 8618
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.42826393909773,
+      "learning_rate": 1.2129510219019686e-05,
+      "loss": 0.6843,
+      "step": 8619
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.742893961466095,
+      "learning_rate": 1.2127862470913478e-05,
+      "loss": 0.6634,
+      "step": 8620
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 40.15666429476635,
+      "learning_rate": 1.2126214662292106e-05,
+      "loss": 0.6267,
+      "step": 8621
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.863886132215406,
+      "learning_rate": 1.2124566793202432e-05,
+      "loss": 0.676,
+      "step": 8622
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.35901954262314,
+      "learning_rate": 1.2122918863691322e-05,
+      "loss": 0.6535,
+      "step": 8623
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.303444054296783,
+      "learning_rate": 1.2121270873805638e-05,
+      "loss": 0.5618,
+      "step": 8624
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.97493165619017,
+      "learning_rate": 1.2119622823592254e-05,
+      "loss": 0.6627,
+      "step": 8625
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 38.212876661626154,
+      "learning_rate": 1.2117974713098038e-05,
+      "loss": 0.7113,
+      "step": 8626
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 31.657210121770667,
+      "learning_rate": 1.2116326542369859e-05,
+      "loss": 0.6303,
+      "step": 8627
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.83571414525879,
+      "learning_rate": 1.211467831145459e-05,
+      "loss": 0.7136,
+      "step": 8628
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 32.567346007267886,
+      "learning_rate": 1.2113030020399107e-05,
+      "loss": 0.6052,
+      "step": 8629
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.014566049775695,
+      "learning_rate": 1.2111381669250288e-05,
+      "loss": 0.6087,
+      "step": 8630
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 43.69449176423348,
+      "learning_rate": 1.2109733258055007e-05,
+      "loss": 0.6366,
+      "step": 8631
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.25161582544045,
+      "learning_rate": 1.210808478686015e-05,
+      "loss": 0.6737,
+      "step": 8632
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 36.401849256428164,
+      "learning_rate": 1.210643625571259e-05,
+      "loss": 0.7423,
+      "step": 8633
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 27.17736014993643,
+      "learning_rate": 1.2104787664659221e-05,
+      "loss": 0.5833,
+      "step": 8634
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 36.04333684428814,
+      "learning_rate": 1.2103139013746919e-05,
+      "loss": 0.6056,
+      "step": 8635
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.10079902244838,
+      "learning_rate": 1.2101490303022571e-05,
+      "loss": 0.6522,
+      "step": 8636
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 38.31941040097157,
+      "learning_rate": 1.2099841532533073e-05,
+      "loss": 0.6559,
+      "step": 8637
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 42.58294034553837,
+      "learning_rate": 1.2098192702325309e-05,
+      "loss": 0.7283,
+      "step": 8638
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 37.625030596774515,
+      "learning_rate": 1.209654381244617e-05,
+      "loss": 0.585,
+      "step": 8639
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.745750682009884,
+      "learning_rate": 1.2094894862942552e-05,
+      "loss": 0.6914,
+      "step": 8640
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 46.044191658322035,
+      "learning_rate": 1.2093245853861349e-05,
+      "loss": 0.7577,
+      "step": 8641
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 31.58594466706335,
+      "learning_rate": 1.2091596785249461e-05,
+      "loss": 0.6313,
+      "step": 8642
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 38.32375951228781,
+      "learning_rate": 1.208994765715378e-05,
+      "loss": 0.7112,
+      "step": 8643
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.61667524748826,
+      "learning_rate": 1.2088298469621214e-05,
+      "loss": 0.6759,
+      "step": 8644
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 44.48603596348323,
+      "learning_rate": 1.208664922269866e-05,
+      "loss": 0.6238,
+      "step": 8645
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.78288700813274,
+      "learning_rate": 1.2084999916433022e-05,
+      "loss": 0.7047,
+      "step": 8646
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 39.95131560201628,
+      "learning_rate": 1.2083350550871206e-05,
+      "loss": 0.6398,
+      "step": 8647
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.23272186959966,
+      "learning_rate": 1.2081701126060121e-05,
+      "loss": 0.6057,
+      "step": 8648
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 25.703579253382916,
+      "learning_rate": 1.2080051642046674e-05,
+      "loss": 0.6078,
+      "step": 8649
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 45.737504018620236,
+      "learning_rate": 1.2078402098877771e-05,
+      "loss": 0.8366,
+      "step": 8650
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.48874167557218,
+      "learning_rate": 1.2076752496600333e-05,
+      "loss": 0.7059,
+      "step": 8651
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.8610435981157,
+      "learning_rate": 1.2075102835261264e-05,
+      "loss": 0.6231,
+      "step": 8652
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 40.92516778586452,
+      "learning_rate": 1.2073453114907485e-05,
+      "loss": 0.6827,
+      "step": 8653
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 37.30534244301953,
+      "learning_rate": 1.2071803335585917e-05,
+      "loss": 0.6748,
+      "step": 8654
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.37300567936285,
+      "learning_rate": 1.207015349734347e-05,
+      "loss": 0.7155,
+      "step": 8655
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 43.44736378599273,
+      "learning_rate": 1.206850360022707e-05,
+      "loss": 0.5896,
+      "step": 8656
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 32.238653707993144,
+      "learning_rate": 1.2066853644283639e-05,
+      "loss": 0.6747,
+      "step": 8657
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 36.09658149082741,
+      "learning_rate": 1.2065203629560093e-05,
+      "loss": 0.7142,
+      "step": 8658
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 36.016850084499445,
+      "learning_rate": 1.2063553556103372e-05,
+      "loss": 0.6296,
+      "step": 8659
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 38.22902978184834,
+      "learning_rate": 1.206190342396039e-05,
+      "loss": 0.6261,
+      "step": 8660
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 39.62467690766321,
+      "learning_rate": 1.2060253233178086e-05,
+      "loss": 0.6075,
+      "step": 8661
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 37.33710509360541,
+      "learning_rate": 1.2058602983803378e-05,
+      "loss": 0.7381,
+      "step": 8662
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.73583571296005,
+      "learning_rate": 1.2056952675883208e-05,
+      "loss": 0.6224,
+      "step": 8663
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.396392557771996,
+      "learning_rate": 1.2055302309464509e-05,
+      "loss": 0.5822,
+      "step": 8664
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.07298859966028,
+      "learning_rate": 1.2053651884594215e-05,
+      "loss": 0.7252,
+      "step": 8665
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.10747097746053,
+      "learning_rate": 1.2052001401319262e-05,
+      "loss": 0.613,
+      "step": 8666
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 31.61510846970807,
+      "learning_rate": 1.205035085968659e-05,
+      "loss": 0.6742,
+      "step": 8667
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 43.137909988581775,
+      "learning_rate": 1.2048700259743136e-05,
+      "loss": 0.6405,
+      "step": 8668
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 46.51737040460403,
+      "learning_rate": 1.2047049601535847e-05,
+      "loss": 0.7857,
+      "step": 8669
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 34.878158542523614,
+      "learning_rate": 1.2045398885111665e-05,
+      "loss": 0.5496,
+      "step": 8670
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 38.65123674639643,
+      "learning_rate": 1.2043748110517536e-05,
+      "loss": 0.6538,
+      "step": 8671
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 36.96783220082153,
+      "learning_rate": 1.2042097277800406e-05,
+      "loss": 0.6646,
+      "step": 8672
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 33.49294712423388,
+      "learning_rate": 1.2040446387007222e-05,
+      "loss": 0.668,
+      "step": 8673
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 47.171367813118096,
+      "learning_rate": 1.203879543818494e-05,
+      "loss": 0.703,
+      "step": 8674
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 35.3364891634986,
+      "learning_rate": 1.2037144431380506e-05,
+      "loss": 0.603,
+      "step": 8675
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 47.826409953142445,
+      "learning_rate": 1.2035493366640879e-05,
+      "loss": 0.6181,
+      "step": 8676
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 36.57837356228724,
+      "learning_rate": 1.2033842244013006e-05,
+      "loss": 0.7158,
+      "step": 8677
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 40.868497622956625,
+      "learning_rate": 1.2032191063543855e-05,
+      "loss": 0.6356,
+      "step": 8678
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 41.87460506309906,
+      "learning_rate": 1.2030539825280373e-05,
+      "loss": 0.6397,
+      "step": 8679
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 37.90687256731859,
+      "learning_rate": 1.202888852926953e-05,
+      "loss": 0.6937,
+      "step": 8680
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 33.84384039376287,
+      "learning_rate": 1.2027237175558283e-05,
+      "loss": 0.5731,
+      "step": 8681
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 38.42644163019714,
+      "learning_rate": 1.2025585764193597e-05,
+      "loss": 0.6016,
+      "step": 8682
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 28.40198217042979,
+      "learning_rate": 1.2023934295222437e-05,
+      "loss": 0.6719,
+      "step": 8683
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 48.73327190768854,
+      "learning_rate": 1.202228276869177e-05,
+      "loss": 0.7423,
+      "step": 8684
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 39.46853103033292,
+      "learning_rate": 1.202063118464856e-05,
+      "loss": 0.6268,
+      "step": 8685
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 40.255683900394146,
+      "learning_rate": 1.2018979543139788e-05,
+      "loss": 0.6512,
+      "step": 8686
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.558081573833796,
+      "learning_rate": 1.2017327844212414e-05,
+      "loss": 0.6886,
+      "step": 8687
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 29.350643372023377,
+      "learning_rate": 1.2015676087913418e-05,
+      "loss": 0.6171,
+      "step": 8688
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 36.96932327071095,
+      "learning_rate": 1.2014024274289773e-05,
+      "loss": 0.7114,
+      "step": 8689
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 45.26674827153758,
+      "learning_rate": 1.2012372403388457e-05,
+      "loss": 0.683,
+      "step": 8690
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.311426576533826,
+      "learning_rate": 1.2010720475256446e-05,
+      "loss": 0.5899,
+      "step": 8691
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 37.47233147004716,
+      "learning_rate": 1.200906848994072e-05,
+      "loss": 0.735,
+      "step": 8692
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 40.57171476817743,
+      "learning_rate": 1.2007416447488263e-05,
+      "loss": 0.8123,
+      "step": 8693
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.602448195221942,
+      "learning_rate": 1.2005764347946053e-05,
+      "loss": 0.6748,
+      "step": 8694
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.82135105101665,
+      "learning_rate": 1.200411219136108e-05,
+      "loss": 0.6066,
+      "step": 8695
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 42.60293499379503,
+      "learning_rate": 1.2002459977780331e-05,
+      "loss": 0.6237,
+      "step": 8696
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 36.8556012998784,
+      "learning_rate": 1.200080770725079e-05,
+      "loss": 0.715,
+      "step": 8697
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 46.16285627812794,
+      "learning_rate": 1.1999155379819449e-05,
+      "loss": 0.6246,
+      "step": 8698
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 38.895657524497516,
+      "learning_rate": 1.1997502995533299e-05,
+      "loss": 0.634,
+      "step": 8699
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.59141859711847,
+      "learning_rate": 1.1995850554439332e-05,
+      "loss": 0.6252,
+      "step": 8700
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 41.27148738761964,
+      "learning_rate": 1.199419805658454e-05,
+      "loss": 0.6539,
+      "step": 8701
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 39.8196241361572,
+      "learning_rate": 1.1992545502015923e-05,
+      "loss": 0.6926,
+      "step": 8702
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 43.081935931231826,
+      "learning_rate": 1.1990892890780482e-05,
+      "loss": 0.6604,
+      "step": 8703
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 33.76925373305338,
+      "learning_rate": 1.1989240222925206e-05,
+      "loss": 0.6601,
+      "step": 8704
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.220029721651265,
+      "learning_rate": 1.1987587498497107e-05,
+      "loss": 0.6773,
+      "step": 8705
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 37.61945083479965,
+      "learning_rate": 1.1985934717543178e-05,
+      "loss": 0.7413,
+      "step": 8706
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 37.29559022874429,
+      "learning_rate": 1.1984281880110427e-05,
+      "loss": 0.633,
+      "step": 8707
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 35.509442641772836,
+      "learning_rate": 1.198262898624586e-05,
+      "loss": 0.5747,
+      "step": 8708
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.040406646160218,
+      "learning_rate": 1.1980976035996488e-05,
+      "loss": 0.6626,
+      "step": 8709
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 41.865325231433516,
+      "learning_rate": 1.1979323029409316e-05,
+      "loss": 0.7558,
+      "step": 8710
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 43.035595928576605,
+      "learning_rate": 1.1977669966531353e-05,
+      "loss": 0.6926,
+      "step": 8711
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.103896322892375,
+      "learning_rate": 1.197601684740961e-05,
+      "loss": 0.5852,
+      "step": 8712
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 40.527290921464065,
+      "learning_rate": 1.197436367209111e-05,
+      "loss": 0.6995,
+      "step": 8713
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.94250216131818,
+      "learning_rate": 1.1972710440622858e-05,
+      "loss": 0.6009,
+      "step": 8714
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 51.009693355883755,
+      "learning_rate": 1.1971057153051878e-05,
+      "loss": 0.7006,
+      "step": 8715
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 36.165161066681556,
+      "learning_rate": 1.1969403809425183e-05,
+      "loss": 0.5679,
+      "step": 8716
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.391210192651055,
+      "learning_rate": 1.1967750409789796e-05,
+      "loss": 0.6613,
+      "step": 8717
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 33.316153334924394,
+      "learning_rate": 1.196609695419274e-05,
+      "loss": 0.5747,
+      "step": 8718
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 38.7358061851906,
+      "learning_rate": 1.1964443442681036e-05,
+      "loss": 0.668,
+      "step": 8719
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 39.379665467714304,
+      "learning_rate": 1.196278987530171e-05,
+      "loss": 0.6911,
+      "step": 8720
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.16780232376585,
+      "learning_rate": 1.1961136252101786e-05,
+      "loss": 0.5994,
+      "step": 8721
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.27183686886737,
+      "learning_rate": 1.1959482573128298e-05,
+      "loss": 0.6761,
+      "step": 8722
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 33.70329437779225,
+      "learning_rate": 1.1957828838428269e-05,
+      "loss": 0.6625,
+      "step": 8723
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 43.02287015303589,
+      "learning_rate": 1.1956175048048734e-05,
+      "loss": 0.6316,
+      "step": 8724
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.30505330294941,
+      "learning_rate": 1.1954521202036726e-05,
+      "loss": 0.6086,
+      "step": 8725
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 39.238637260466334,
+      "learning_rate": 1.1952867300439276e-05,
+      "loss": 0.7123,
+      "step": 8726
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 45.43236996053101,
+      "learning_rate": 1.1951213343303425e-05,
+      "loss": 0.7048,
+      "step": 8727
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.361772523013926,
+      "learning_rate": 1.1949559330676209e-05,
+      "loss": 0.6565,
+      "step": 8728
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 39.02779099444207,
+      "learning_rate": 1.194790526260466e-05,
+      "loss": 0.6352,
+      "step": 8729
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 35.28092682034085,
+      "learning_rate": 1.1946251139135831e-05,
+      "loss": 0.6411,
+      "step": 8730
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.42210612520328,
+      "learning_rate": 1.1944596960316755e-05,
+      "loss": 0.5972,
+      "step": 8731
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 47.31689475547492,
+      "learning_rate": 1.1942942726194477e-05,
+      "loss": 0.7416,
+      "step": 8732
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 36.92823212759699,
+      "learning_rate": 1.194128843681605e-05,
+      "loss": 0.7376,
+      "step": 8733
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 33.109157397719436,
+      "learning_rate": 1.1939634092228511e-05,
+      "loss": 0.6508,
+      "step": 8734
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 27.875637149799164,
+      "learning_rate": 1.1937979692478915e-05,
+      "loss": 0.5909,
+      "step": 8735
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 33.120144851720255,
+      "learning_rate": 1.1936325237614312e-05,
+      "loss": 0.6282,
+      "step": 8736
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 38.7645350421785,
+      "learning_rate": 1.193467072768175e-05,
+      "loss": 0.7303,
+      "step": 8737
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 38.21144789076236,
+      "learning_rate": 1.1933016162728281e-05,
+      "loss": 0.6536,
+      "step": 8738
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.279064171160684,
+      "learning_rate": 1.1931361542800968e-05,
+      "loss": 0.6861,
+      "step": 8739
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.3179172407924,
+      "learning_rate": 1.192970686794686e-05,
+      "loss": 0.5855,
+      "step": 8740
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 28.361400620180223,
+      "learning_rate": 1.192805213821302e-05,
+      "loss": 0.5877,
+      "step": 8741
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 45.228720163606916,
+      "learning_rate": 1.1926397353646501e-05,
+      "loss": 0.6713,
+      "step": 8742
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.0594928084255,
+      "learning_rate": 1.1924742514294371e-05,
+      "loss": 0.6751,
+      "step": 8743
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.76040557892869,
+      "learning_rate": 1.1923087620203688e-05,
+      "loss": 0.6811,
+      "step": 8744
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 46.25962961316624,
+      "learning_rate": 1.1921432671421523e-05,
+      "loss": 0.6946,
+      "step": 8745
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 28.66635803446487,
+      "learning_rate": 1.1919777667994932e-05,
+      "loss": 0.5619,
+      "step": 8746
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.73556898530092,
+      "learning_rate": 1.191812260997099e-05,
+      "loss": 0.6414,
+      "step": 8747
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 48.33120251047348,
+      "learning_rate": 1.1916467497396759e-05,
+      "loss": 0.7271,
+      "step": 8748
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 39.95560604562888,
+      "learning_rate": 1.1914812330319318e-05,
+      "loss": 0.7002,
+      "step": 8749
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 30.283285885488333,
+      "learning_rate": 1.1913157108785731e-05,
+      "loss": 0.5793,
+      "step": 8750
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 46.36821180978157,
+      "learning_rate": 1.1911501832843077e-05,
+      "loss": 0.6658,
+      "step": 8751
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.97939462583327,
+      "learning_rate": 1.1909846502538429e-05,
+      "loss": 0.6322,
+      "step": 8752
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 32.8375859651141,
+      "learning_rate": 1.1908191117918864e-05,
+      "loss": 0.5677,
+      "step": 8753
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.428157403674135,
+      "learning_rate": 1.190653567903146e-05,
+      "loss": 0.7062,
+      "step": 8754
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.35892456129768,
+      "learning_rate": 1.1904880185923295e-05,
+      "loss": 0.6002,
+      "step": 8755
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 45.82184793512725,
+      "learning_rate": 1.190322463864145e-05,
+      "loss": 0.6683,
+      "step": 8756
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 41.294503564253546,
+      "learning_rate": 1.1901569037233012e-05,
+      "loss": 0.7253,
+      "step": 8757
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 32.12913232231028,
+      "learning_rate": 1.1899913381745062e-05,
+      "loss": 0.6564,
+      "step": 8758
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.41596806399222,
+      "learning_rate": 1.189825767222469e-05,
+      "loss": 0.7312,
+      "step": 8759
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 40.98976775163836,
+      "learning_rate": 1.1896601908718979e-05,
+      "loss": 0.665,
+      "step": 8760
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 40.04174394960649,
+      "learning_rate": 1.1894946091275014e-05,
+      "loss": 0.7271,
+      "step": 8761
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 42.38289093432222,
+      "learning_rate": 1.1893290219939899e-05,
+      "loss": 0.6572,
+      "step": 8762
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 39.33420374817792,
+      "learning_rate": 1.1891634294760713e-05,
+      "loss": 0.7527,
+      "step": 8763
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 42.41212696458916,
+      "learning_rate": 1.1889978315784557e-05,
+      "loss": 0.699,
+      "step": 8764
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.3445393670637,
+      "learning_rate": 1.1888322283058517e-05,
+      "loss": 0.6224,
+      "step": 8765
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 40.00459231300623,
+      "learning_rate": 1.1886666196629701e-05,
+      "loss": 0.6348,
+      "step": 8766
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 31.49812719098667,
+      "learning_rate": 1.1885010056545204e-05,
+      "loss": 0.5407,
+      "step": 8767
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 27.680412857926452,
+      "learning_rate": 1.1883353862852121e-05,
+      "loss": 0.5749,
+      "step": 8768
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.02382062863681,
+      "learning_rate": 1.1881697615597554e-05,
+      "loss": 0.604,
+      "step": 8769
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 30.640580635875153,
+      "learning_rate": 1.188004131482861e-05,
+      "loss": 0.5724,
+      "step": 8770
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 34.01639340136895,
+      "learning_rate": 1.187838496059239e-05,
+      "loss": 0.6666,
+      "step": 8771
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.625579173416185,
+      "learning_rate": 1.1876728552936e-05,
+      "loss": 0.6324,
+      "step": 8772
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 37.230205182343376,
+      "learning_rate": 1.1875072091906547e-05,
+      "loss": 0.6914,
+      "step": 8773
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 43.03685961902531,
+      "learning_rate": 1.1873415577551146e-05,
+      "loss": 0.7102,
+      "step": 8774
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 43.15164930721173,
+      "learning_rate": 1.1871759009916897e-05,
+      "loss": 0.7239,
+      "step": 8775
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 38.74538328909881,
+      "learning_rate": 1.1870102389050917e-05,
+      "loss": 0.636,
+      "step": 8776
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 39.63267245156065,
+      "learning_rate": 1.186844571500032e-05,
+      "loss": 0.6808,
+      "step": 8777
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.55897324429312,
+      "learning_rate": 1.1866788987812219e-05,
+      "loss": 0.5671,
+      "step": 8778
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 41.75458873726678,
+      "learning_rate": 1.1865132207533731e-05,
+      "loss": 0.607,
+      "step": 8779
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.362027554616766,
+      "learning_rate": 1.1863475374211974e-05,
+      "loss": 0.6753,
+      "step": 8780
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 45.20387573357192,
+      "learning_rate": 1.1861818487894072e-05,
+      "loss": 0.7888,
+      "step": 8781
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 42.529779906305464,
+      "learning_rate": 1.1860161548627137e-05,
+      "loss": 0.6473,
+      "step": 8782
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 37.23774062108067,
+      "learning_rate": 1.1858504556458294e-05,
+      "loss": 0.6898,
+      "step": 8783
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 37.27294858549076,
+      "learning_rate": 1.1856847511434673e-05,
+      "loss": 0.6908,
+      "step": 8784
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 32.552664224960225,
+      "learning_rate": 1.1855190413603392e-05,
+      "loss": 0.6006,
+      "step": 8785
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 32.249427357845924,
+      "learning_rate": 1.1853533263011583e-05,
+      "loss": 0.6872,
+      "step": 8786
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 38.196808993040555,
+      "learning_rate": 1.185187605970637e-05,
+      "loss": 0.6866,
+      "step": 8787
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 34.20182462531949,
+      "learning_rate": 1.1850218803734886e-05,
+      "loss": 0.6426,
+      "step": 8788
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 31.501143642887957,
+      "learning_rate": 1.1848561495144263e-05,
+      "loss": 0.6163,
+      "step": 8789
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 31.241279209190314,
+      "learning_rate": 1.184690413398163e-05,
+      "loss": 0.6026,
+      "step": 8790
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 38.050148330028414,
+      "learning_rate": 1.1845246720294129e-05,
+      "loss": 0.6003,
+      "step": 8791
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.48441877909623,
+      "learning_rate": 1.1843589254128884e-05,
+      "loss": 0.6121,
+      "step": 8792
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.611362889733115,
+      "learning_rate": 1.1841931735533043e-05,
+      "loss": 0.6195,
+      "step": 8793
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.8067787486938,
+      "learning_rate": 1.184027416455374e-05,
+      "loss": 0.6102,
+      "step": 8794
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.06855281732928,
+      "learning_rate": 1.1838616541238115e-05,
+      "loss": 0.667,
+      "step": 8795
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 35.401157689695985,
+      "learning_rate": 1.1836958865633315e-05,
+      "loss": 0.6175,
+      "step": 8796
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 37.175875274933716,
+      "learning_rate": 1.1835301137786476e-05,
+      "loss": 0.7004,
+      "step": 8797
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 29.79262433991474,
+      "learning_rate": 1.1833643357744747e-05,
+      "loss": 0.5851,
+      "step": 8798
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 39.12395033784189,
+      "learning_rate": 1.1831985525555274e-05,
+      "loss": 0.6092,
+      "step": 8799
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.3621880017553,
+      "learning_rate": 1.1830327641265202e-05,
+      "loss": 0.6539,
+      "step": 8800
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 34.5642783757705,
+      "learning_rate": 1.1828669704921685e-05,
+      "loss": 0.5784,
+      "step": 8801
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 36.2306260492658,
+      "learning_rate": 1.182701171657187e-05,
+      "loss": 0.6258,
+      "step": 8802
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 42.80595526686382,
+      "learning_rate": 1.1825353676262914e-05,
+      "loss": 0.7139,
+      "step": 8803
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 32.59100699289122,
+      "learning_rate": 1.1823695584041963e-05,
+      "loss": 0.6673,
+      "step": 8804
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 45.27073639521232,
+      "learning_rate": 1.1822037439956178e-05,
+      "loss": 0.8593,
+      "step": 8805
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 44.27337457461553,
+      "learning_rate": 1.1820379244052715e-05,
+      "loss": 0.6753,
+      "step": 8806
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 41.108119665601414,
+      "learning_rate": 1.1818720996378729e-05,
+      "loss": 0.655,
+      "step": 8807
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 43.41875384775378,
+      "learning_rate": 1.1817062696981384e-05,
+      "loss": 0.5795,
+      "step": 8808
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 33.943078336074116,
+      "learning_rate": 1.1815404345907837e-05,
+      "loss": 0.6849,
+      "step": 8809
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.38372518400422,
+      "learning_rate": 1.1813745943205254e-05,
+      "loss": 0.6434,
+      "step": 8810
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 44.17205724418836,
+      "learning_rate": 1.1812087488920798e-05,
+      "loss": 0.7465,
+      "step": 8811
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 41.127378910909734,
+      "learning_rate": 1.1810428983101632e-05,
+      "loss": 0.5832,
+      "step": 8812
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 41.55090941387585,
+      "learning_rate": 1.1808770425794927e-05,
+      "loss": 0.6529,
+      "step": 8813
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 41.20580568298257,
+      "learning_rate": 1.1807111817047846e-05,
+      "loss": 0.6871,
+      "step": 8814
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 32.215596900370414,
+      "learning_rate": 1.1805453156907562e-05,
+      "loss": 0.5993,
+      "step": 8815
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 37.62043554240992,
+      "learning_rate": 1.1803794445421251e-05,
+      "loss": 0.6456,
+      "step": 8816
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 45.35316739129968,
+      "learning_rate": 1.1802135682636076e-05,
+      "loss": 0.6759,
+      "step": 8817
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 28.324070681909763,
+      "learning_rate": 1.1800476868599222e-05,
+      "loss": 0.5376,
+      "step": 8818
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 40.90141615199687,
+      "learning_rate": 1.1798818003357853e-05,
+      "loss": 0.6378,
+      "step": 8819
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 34.21828704242109,
+      "learning_rate": 1.1797159086959156e-05,
+      "loss": 0.6704,
+      "step": 8820
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 34.33583443097821,
+      "learning_rate": 1.1795500119450305e-05,
+      "loss": 0.6297,
+      "step": 8821
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 31.757740023484526,
+      "learning_rate": 1.179384110087848e-05,
+      "loss": 0.6232,
+      "step": 8822
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.438244944231,
+      "learning_rate": 1.1792182031290867e-05,
+      "loss": 0.7188,
+      "step": 8823
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 40.4683596323399,
+      "learning_rate": 1.1790522910734638e-05,
+      "loss": 0.6246,
+      "step": 8824
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.08576794082849,
+      "learning_rate": 1.1788863739256992e-05,
+      "loss": 0.618,
+      "step": 8825
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 45.77514896816513,
+      "learning_rate": 1.1787204516905104e-05,
+      "loss": 0.6216,
+      "step": 8826
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 35.60469064537152,
+      "learning_rate": 1.1785545243726166e-05,
+      "loss": 0.5146,
+      "step": 8827
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 46.60178925809985,
+      "learning_rate": 1.1783885919767368e-05,
+      "loss": 0.8068,
+      "step": 8828
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 31.947363104977335,
+      "learning_rate": 1.1782226545075896e-05,
+      "loss": 0.6421,
+      "step": 8829
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 36.2164191087623,
+      "learning_rate": 1.1780567119698944e-05,
+      "loss": 0.6575,
+      "step": 8830
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 36.26987976334055,
+      "learning_rate": 1.1778907643683704e-05,
+      "loss": 0.6057,
+      "step": 8831
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 34.314729274185964,
+      "learning_rate": 1.1777248117077371e-05,
+      "loss": 0.681,
+      "step": 8832
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.46030710816821,
+      "learning_rate": 1.1775588539927142e-05,
+      "loss": 0.6348,
+      "step": 8833
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 35.4093974466129,
+      "learning_rate": 1.1773928912280213e-05,
+      "loss": 0.6429,
+      "step": 8834
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 36.71851348151468,
+      "learning_rate": 1.1772269234183786e-05,
+      "loss": 0.6516,
+      "step": 8835
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 37.97194082408101,
+      "learning_rate": 1.1770609505685056e-05,
+      "loss": 0.5829,
+      "step": 8836
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 39.97451318345416,
+      "learning_rate": 1.1768949726831228e-05,
+      "loss": 0.6849,
+      "step": 8837
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 34.63216686098741,
+      "learning_rate": 1.1767289897669505e-05,
+      "loss": 0.7318,
+      "step": 8838
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 46.92983763384821,
+      "learning_rate": 1.1765630018247089e-05,
+      "loss": 0.6789,
+      "step": 8839
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 30.079048772720792,
+      "learning_rate": 1.1763970088611192e-05,
+      "loss": 0.5493,
+      "step": 8840
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.642931513391154,
+      "learning_rate": 1.1762310108809017e-05,
+      "loss": 0.6911,
+      "step": 8841
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 40.82035163841014,
+      "learning_rate": 1.176065007888777e-05,
+      "loss": 0.7284,
+      "step": 8842
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 43.04092263609273,
+      "learning_rate": 1.1758989998894667e-05,
+      "loss": 0.6964,
+      "step": 8843
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 34.85477607993856,
+      "learning_rate": 1.1757329868876917e-05,
+      "loss": 0.6435,
+      "step": 8844
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 41.08767143227674,
+      "learning_rate": 1.1755669688881732e-05,
+      "loss": 0.56,
+      "step": 8845
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 37.2667630915526,
+      "learning_rate": 1.175400945895633e-05,
+      "loss": 0.6509,
+      "step": 8846
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 44.20235976699256,
+      "learning_rate": 1.1752349179147926e-05,
+      "loss": 0.7526,
+      "step": 8847
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 39.41218480154418,
+      "learning_rate": 1.1750688849503735e-05,
+      "loss": 0.6902,
+      "step": 8848
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 46.744958865859324,
+      "learning_rate": 1.1749028470070975e-05,
+      "loss": 0.6692,
+      "step": 8849
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 40.21523439681712,
+      "learning_rate": 1.1747368040896875e-05,
+      "loss": 0.7516,
+      "step": 8850
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 30.26470989996401,
+      "learning_rate": 1.1745707562028643e-05,
+      "loss": 0.5936,
+      "step": 8851
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.3241470082733,
+      "learning_rate": 1.1744047033513514e-05,
+      "loss": 0.6164,
+      "step": 8852
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 46.234389750847974,
+      "learning_rate": 1.1742386455398704e-05,
+      "loss": 0.6435,
+      "step": 8853
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 29.006207765656562,
+      "learning_rate": 1.1740725827731446e-05,
+      "loss": 0.5667,
+      "step": 8854
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 31.60704172209045,
+      "learning_rate": 1.1739065150558961e-05,
+      "loss": 0.6132,
+      "step": 8855
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 41.04584010765065,
+      "learning_rate": 1.1737404423928482e-05,
+      "loss": 0.7492,
+      "step": 8856
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 32.631911378802094,
+      "learning_rate": 1.1735743647887237e-05,
+      "loss": 0.6063,
+      "step": 8857
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 35.89535793839075,
+      "learning_rate": 1.1734082822482457e-05,
+      "loss": 0.6544,
+      "step": 8858
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 40.1594610154983,
+      "learning_rate": 1.1732421947761377e-05,
+      "loss": 0.7453,
+      "step": 8859
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 37.893911701293405,
+      "learning_rate": 1.173076102377123e-05,
+      "loss": 0.7076,
+      "step": 8860
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 32.55643283604445,
+      "learning_rate": 1.1729100050559252e-05,
+      "loss": 0.6364,
+      "step": 8861
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.01432295081318,
+      "learning_rate": 1.1727439028172682e-05,
+      "loss": 0.7502,
+      "step": 8862
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 39.91865370980555,
+      "learning_rate": 1.1725777956658752e-05,
+      "loss": 0.7177,
+      "step": 8863
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 36.58353232559156,
+      "learning_rate": 1.172411683606471e-05,
+      "loss": 0.6704,
+      "step": 8864
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 35.99299342133145,
+      "learning_rate": 1.1722455666437793e-05,
+      "loss": 0.6434,
+      "step": 8865
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 36.07407568225259,
+      "learning_rate": 1.1720794447825245e-05,
+      "loss": 0.7416,
+      "step": 8866
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 33.774916910087654,
+      "learning_rate": 1.171913318027431e-05,
+      "loss": 0.6436,
+      "step": 8867
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.26464948204947,
+      "learning_rate": 1.1717471863832231e-05,
+      "loss": 0.607,
+      "step": 8868
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 43.84456858615091,
+      "learning_rate": 1.1715810498546259e-05,
+      "loss": 0.6309,
+      "step": 8869
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.65244634460056,
+      "learning_rate": 1.1714149084463638e-05,
+      "loss": 0.7149,
+      "step": 8870
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.86550883746381,
+      "learning_rate": 1.1712487621631621e-05,
+      "loss": 0.6808,
+      "step": 8871
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 33.79071380990219,
+      "learning_rate": 1.1710826110097457e-05,
+      "loss": 0.6519,
+      "step": 8872
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.957118570619535,
+      "learning_rate": 1.17091645499084e-05,
+      "loss": 0.6879,
+      "step": 8873
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 122.7718019566479,
+      "learning_rate": 1.1707502941111704e-05,
+      "loss": 0.6765,
+      "step": 8874
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.68274972621415,
+      "learning_rate": 1.1705841283754622e-05,
+      "loss": 0.6472,
+      "step": 8875
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 88.33848182021407,
+      "learning_rate": 1.170417957788441e-05,
+      "loss": 0.6973,
+      "step": 8876
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 54.55092044843976,
+      "learning_rate": 1.1702517823548332e-05,
+      "loss": 0.6543,
+      "step": 8877
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 37.72698177784814,
+      "learning_rate": 1.1700856020793639e-05,
+      "loss": 0.6717,
+      "step": 8878
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 49.0751619170166,
+      "learning_rate": 1.1699194169667598e-05,
+      "loss": 0.6945,
+      "step": 8879
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 44.74112826053183,
+      "learning_rate": 1.1697532270217466e-05,
+      "loss": 0.646,
+      "step": 8880
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 41.75200834983096,
+      "learning_rate": 1.1695870322490512e-05,
+      "loss": 0.7625,
+      "step": 8881
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 45.819916732392734,
+      "learning_rate": 1.1694208326533997e-05,
+      "loss": 0.7667,
+      "step": 8882
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 42.21525831487561,
+      "learning_rate": 1.169254628239519e-05,
+      "loss": 0.601,
+      "step": 8883
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 46.252169707671406,
+      "learning_rate": 1.1690884190121356e-05,
+      "loss": 0.6932,
+      "step": 8884
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.55504362581405,
+      "learning_rate": 1.1689222049759765e-05,
+      "loss": 0.6886,
+      "step": 8885
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.56584333520741,
+      "learning_rate": 1.1687559861357685e-05,
+      "loss": 0.5788,
+      "step": 8886
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.70138371260137,
+      "learning_rate": 1.1685897624962392e-05,
+      "loss": 0.6705,
+      "step": 8887
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.57267834332826,
+      "learning_rate": 1.1684235340621155e-05,
+      "loss": 0.5959,
+      "step": 8888
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 28.1256883431018,
+      "learning_rate": 1.1682573008381252e-05,
+      "loss": 0.5174,
+      "step": 8889
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 41.288446173132535,
+      "learning_rate": 1.1680910628289956e-05,
+      "loss": 0.6032,
+      "step": 8890
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 64.56695668292093,
+      "learning_rate": 1.1679248200394546e-05,
+      "loss": 0.6553,
+      "step": 8891
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 46.59023481959545,
+      "learning_rate": 1.1677585724742298e-05,
+      "loss": 0.6572,
+      "step": 8892
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 39.85234447759754,
+      "learning_rate": 1.1675923201380493e-05,
+      "loss": 0.744,
+      "step": 8893
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 37.49320819067103,
+      "learning_rate": 1.1674260630356415e-05,
+      "loss": 0.621,
+      "step": 8894
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 34.27864737007022,
+      "learning_rate": 1.167259801171734e-05,
+      "loss": 0.5459,
+      "step": 8895
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.32763933448678,
+      "learning_rate": 1.1670935345510561e-05,
+      "loss": 0.6676,
+      "step": 8896
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 51.19022170715401,
+      "learning_rate": 1.1669272631783354e-05,
+      "loss": 0.6092,
+      "step": 8897
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 40.6357167531211,
+      "learning_rate": 1.1667609870583012e-05,
+      "loss": 0.6434,
+      "step": 8898
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 52.858740360465625,
+      "learning_rate": 1.1665947061956821e-05,
+      "loss": 0.6241,
+      "step": 8899
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 34.040351694421716,
+      "learning_rate": 1.166428420595207e-05,
+      "loss": 0.6955,
+      "step": 8900
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.25306960954061,
+      "learning_rate": 1.166262130261605e-05,
+      "loss": 0.6521,
+      "step": 8901
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 30.473848962834754,
+      "learning_rate": 1.1660958351996051e-05,
+      "loss": 0.6715,
+      "step": 8902
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 46.32721791915235,
+      "learning_rate": 1.1659295354139368e-05,
+      "loss": 0.7359,
+      "step": 8903
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.619166736537785,
+      "learning_rate": 1.1657632309093301e-05,
+      "loss": 0.6268,
+      "step": 8904
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 39.07830058244858,
+      "learning_rate": 1.1655969216905134e-05,
+      "loss": 0.654,
+      "step": 8905
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 37.47047799574424,
+      "learning_rate": 1.1654306077622176e-05,
+      "loss": 0.6458,
+      "step": 8906
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.64787067275054,
+      "learning_rate": 1.1652642891291717e-05,
+      "loss": 0.5996,
+      "step": 8907
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 34.2856775448239,
+      "learning_rate": 1.1650979657961063e-05,
+      "loss": 0.6597,
+      "step": 8908
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 42.355146151261124,
+      "learning_rate": 1.1649316377677513e-05,
+      "loss": 0.6631,
+      "step": 8909
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.710860011141584,
+      "learning_rate": 1.164765305048837e-05,
+      "loss": 0.6494,
+      "step": 8910
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 31.25478916091607,
+      "learning_rate": 1.1645989676440938e-05,
+      "loss": 0.6287,
+      "step": 8911
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 32.302466769845324,
+      "learning_rate": 1.1644326255582523e-05,
+      "loss": 0.6146,
+      "step": 8912
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.05083830248595,
+      "learning_rate": 1.164266278796043e-05,
+      "loss": 0.7029,
+      "step": 8913
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 46.65792404717943,
+      "learning_rate": 1.1640999273621969e-05,
+      "loss": 0.786,
+      "step": 8914
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 55.24721339694099,
+      "learning_rate": 1.1639335712614451e-05,
+      "loss": 0.6377,
+      "step": 8915
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.70558997817193,
+      "learning_rate": 1.1637672104985182e-05,
+      "loss": 0.5982,
+      "step": 8916
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 37.966378276830426,
+      "learning_rate": 1.1636008450781475e-05,
+      "loss": 0.6475,
+      "step": 8917
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 42.8822318281187,
+      "learning_rate": 1.1634344750050648e-05,
+      "loss": 0.7304,
+      "step": 8918
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 33.469556846830585,
+      "learning_rate": 1.163268100284001e-05,
+      "loss": 0.6356,
+      "step": 8919
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 43.07623499950211,
+      "learning_rate": 1.1631017209196878e-05,
+      "loss": 0.7388,
+      "step": 8920
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.84856568159495,
+      "learning_rate": 1.1629353369168574e-05,
+      "loss": 0.6632,
+      "step": 8921
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 32.7401248734574,
+      "learning_rate": 1.162768948280241e-05,
+      "loss": 0.7086,
+      "step": 8922
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 39.04911596882543,
+      "learning_rate": 1.1626025550145714e-05,
+      "loss": 0.6602,
+      "step": 8923
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.52633022990521,
+      "learning_rate": 1.16243615712458e-05,
+      "loss": 0.6614,
+      "step": 8924
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.27584253538708,
+      "learning_rate": 1.1622697546149992e-05,
+      "loss": 0.6092,
+      "step": 8925
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 55.63835508348971,
+      "learning_rate": 1.1621033474905617e-05,
+      "loss": 0.6254,
+      "step": 8926
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 41.6322562677525,
+      "learning_rate": 1.1619369357559998e-05,
+      "loss": 0.7476,
+      "step": 8927
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.625448190647155,
+      "learning_rate": 1.161770519416046e-05,
+      "loss": 0.6361,
+      "step": 8928
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 43.46105456709331,
+      "learning_rate": 1.1616040984754333e-05,
+      "loss": 0.7142,
+      "step": 8929
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 35.37370631544945,
+      "learning_rate": 1.1614376729388946e-05,
+      "loss": 0.6564,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 38.21671585446395,
+      "learning_rate": 1.161271242811163e-05,
+      "loss": 0.7686,
+      "step": 8931
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 52.55100481068996,
+      "learning_rate": 1.1611048080969715e-05,
+      "loss": 0.7055,
+      "step": 8932
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 42.877869326500154,
+      "learning_rate": 1.1609383688010536e-05,
+      "loss": 0.5263,
+      "step": 8933
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 41.93857631353928,
+      "learning_rate": 1.1607719249281426e-05,
+      "loss": 0.659,
+      "step": 8934
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 33.34406876344624,
+      "learning_rate": 1.1606054764829723e-05,
+      "loss": 0.6376,
+      "step": 8935
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 33.99761413169075,
+      "learning_rate": 1.1604390234702758e-05,
+      "loss": 0.6553,
+      "step": 8936
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.00144709985648,
+      "learning_rate": 1.1602725658947873e-05,
+      "loss": 0.6943,
+      "step": 8937
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.992694365273376,
+      "learning_rate": 1.1601061037612413e-05,
+      "loss": 0.543,
+      "step": 8938
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 53.103315824225184,
+      "learning_rate": 1.1599396370743707e-05,
+      "loss": 0.5926,
+      "step": 8939
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.00882307734281,
+      "learning_rate": 1.159773165838911e-05,
+      "loss": 0.7042,
+      "step": 8940
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 39.98991384897113,
+      "learning_rate": 1.1596066900595954e-05,
+      "loss": 0.6664,
+      "step": 8941
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 32.74970142213837,
+      "learning_rate": 1.1594402097411595e-05,
+      "loss": 0.5984,
+      "step": 8942
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.925682218765566,
+      "learning_rate": 1.159273724888337e-05,
+      "loss": 0.6578,
+      "step": 8943
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 45.59335351291797,
+      "learning_rate": 1.1591072355058629e-05,
+      "loss": 0.6348,
+      "step": 8944
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.045228317395672,
+      "learning_rate": 1.1589407415984721e-05,
+      "loss": 0.621,
+      "step": 8945
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 39.08358837497246,
+      "learning_rate": 1.1587742431708997e-05,
+      "loss": 0.7302,
+      "step": 8946
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 33.399363144210895,
+      "learning_rate": 1.1586077402278804e-05,
+      "loss": 0.5714,
+      "step": 8947
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.697699318666587,
+      "learning_rate": 1.1584412327741501e-05,
+      "loss": 0.595,
+      "step": 8948
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 53.35957861396513,
+      "learning_rate": 1.1582747208144435e-05,
+      "loss": 0.6028,
+      "step": 8949
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.20406662998417,
+      "learning_rate": 1.1581082043534968e-05,
+      "loss": 0.5965,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 46.86795106912187,
+      "learning_rate": 1.157941683396045e-05,
+      "loss": 0.7334,
+      "step": 8951
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 31.961236065104337,
+      "learning_rate": 1.1577751579468242e-05,
+      "loss": 0.5876,
+      "step": 8952
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 44.22833362153606,
+      "learning_rate": 1.1576086280105702e-05,
+      "loss": 0.737,
+      "step": 8953
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 38.98719609207238,
+      "learning_rate": 1.1574420935920192e-05,
+      "loss": 0.5999,
+      "step": 8954
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 36.750960986230105,
+      "learning_rate": 1.157275554695907e-05,
+      "loss": 0.6142,
+      "step": 8955
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 30.90633186352633,
+      "learning_rate": 1.1571090113269697e-05,
+      "loss": 0.5956,
+      "step": 8956
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 31.090237477562138,
+      "learning_rate": 1.1569424634899441e-05,
+      "loss": 0.6104,
+      "step": 8957
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 51.42067342702971,
+      "learning_rate": 1.156775911189567e-05,
+      "loss": 0.7477,
+      "step": 8958
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 35.844410491950796,
+      "learning_rate": 1.1566093544305747e-05,
+      "loss": 0.6434,
+      "step": 8959
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.15874180687189,
+      "learning_rate": 1.1564427932177037e-05,
+      "loss": 0.5589,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 32.36031462678805,
+      "learning_rate": 1.1562762275556916e-05,
+      "loss": 0.6263,
+      "step": 8961
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.798391407614965,
+      "learning_rate": 1.1561096574492745e-05,
+      "loss": 0.6092,
+      "step": 8962
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 43.38085812742171,
+      "learning_rate": 1.1559430829031905e-05,
+      "loss": 0.695,
+      "step": 8963
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 37.23023694888486,
+      "learning_rate": 1.155776503922176e-05,
+      "loss": 0.6496,
+      "step": 8964
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 35.44264207432496,
+      "learning_rate": 1.1556099205109694e-05,
+      "loss": 0.6128,
+      "step": 8965
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 37.337098452615336,
+      "learning_rate": 1.155443332674307e-05,
+      "loss": 0.6667,
+      "step": 8966
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 39.507739457192464,
+      "learning_rate": 1.1552767404169281e-05,
+      "loss": 0.5863,
+      "step": 8967
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 32.93073101449237,
+      "learning_rate": 1.1551101437435686e-05,
+      "loss": 0.5674,
+      "step": 8968
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 30.83327956495666,
+      "learning_rate": 1.1549435426589678e-05,
+      "loss": 0.6192,
+      "step": 8969
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 43.167094788520984,
+      "learning_rate": 1.1547769371678632e-05,
+      "loss": 0.7421,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 45.940514643356046,
+      "learning_rate": 1.1546103272749931e-05,
+      "loss": 0.5707,
+      "step": 8971
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 37.10360998406851,
+      "learning_rate": 1.1544437129850956e-05,
+      "loss": 0.6231,
+      "step": 8972
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 43.44801153433128,
+      "learning_rate": 1.1542770943029096e-05,
+      "loss": 0.6527,
+      "step": 8973
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.66999510656946,
+      "learning_rate": 1.1541104712331728e-05,
+      "loss": 0.7291,
+      "step": 8974
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.50706995961262,
+      "learning_rate": 1.1539438437806245e-05,
+      "loss": 0.7356,
+      "step": 8975
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.146594475678924,
+      "learning_rate": 1.1537772119500035e-05,
+      "loss": 0.6369,
+      "step": 8976
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.98107763428213,
+      "learning_rate": 1.1536105757460485e-05,
+      "loss": 0.672,
+      "step": 8977
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 38.3537736085515,
+      "learning_rate": 1.1534439351734986e-05,
+      "loss": 0.6283,
+      "step": 8978
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 33.65185750387507,
+      "learning_rate": 1.1532772902370929e-05,
+      "loss": 0.5882,
+      "step": 8979
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 45.471578230842056,
+      "learning_rate": 1.1531106409415706e-05,
+      "loss": 0.6789,
+      "step": 8980
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.922631745043354,
+      "learning_rate": 1.1529439872916712e-05,
+      "loss": 0.6442,
+      "step": 8981
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.33759331209202,
+      "learning_rate": 1.1527773292921347e-05,
+      "loss": 0.6481,
+      "step": 8982
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 30.01802184877573,
+      "learning_rate": 1.1526106669476999e-05,
+      "loss": 0.5631,
+      "step": 8983
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 36.42053731117156,
+      "learning_rate": 1.1524440002631073e-05,
+      "loss": 0.5199,
+      "step": 8984
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 36.659564038362326,
+      "learning_rate": 1.1522773292430963e-05,
+      "loss": 0.6971,
+      "step": 8985
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 32.46431596896356,
+      "learning_rate": 1.1521106538924071e-05,
+      "loss": 0.5754,
+      "step": 8986
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 35.4871877196339,
+      "learning_rate": 1.15194397421578e-05,
+      "loss": 0.6706,
+      "step": 8987
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 37.784745142260064,
+      "learning_rate": 1.1517772902179554e-05,
+      "loss": 0.6613,
+      "step": 8988
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 36.93497753405478,
+      "learning_rate": 1.1516106019036733e-05,
+      "loss": 0.6026,
+      "step": 8989
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 35.410604863973,
+      "learning_rate": 1.1514439092776741e-05,
+      "loss": 0.6423,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 38.738260718252974,
+      "learning_rate": 1.1512772123446988e-05,
+      "loss": 0.7375,
+      "step": 8991
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 39.3970617099107,
+      "learning_rate": 1.1511105111094884e-05,
+      "loss": 0.7221,
+      "step": 8992
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 38.05318674297263,
+      "learning_rate": 1.150943805576783e-05,
+      "loss": 0.6973,
+      "step": 8993
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 59.108970001390176,
+      "learning_rate": 1.1507770957513245e-05,
+      "loss": 0.6431,
+      "step": 8994
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.08821861077465,
+      "learning_rate": 1.1506103816378533e-05,
+      "loss": 0.6302,
+      "step": 8995
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 36.71332666481428,
+      "learning_rate": 1.1504436632411112e-05,
+      "loss": 0.7049,
+      "step": 8996
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 45.455952755383514,
+      "learning_rate": 1.150276940565839e-05,
+      "loss": 0.7009,
+      "step": 8997
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.84574172481428,
+      "learning_rate": 1.1501102136167788e-05,
+      "loss": 0.6731,
+      "step": 8998
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.2778963348203,
+      "learning_rate": 1.1499434823986719e-05,
+      "loss": 0.7686,
+      "step": 8999
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 32.180158018491184,
+      "learning_rate": 1.1497767469162598e-05,
+      "loss": 0.6254,
+      "step": 9000
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.823033273374094,
+      "learning_rate": 1.1496100071742849e-05,
+      "loss": 0.6364,
+      "step": 9001
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.56436343021937,
+      "learning_rate": 1.1494432631774888e-05,
+      "loss": 0.7072,
+      "step": 9002
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 35.90280531585608,
+      "learning_rate": 1.1492765149306137e-05,
+      "loss": 0.6548,
+      "step": 9003
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 35.79678238815523,
+      "learning_rate": 1.149109762438402e-05,
+      "loss": 0.6318,
+      "step": 9004
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.439755587631595,
+      "learning_rate": 1.1489430057055959e-05,
+      "loss": 0.7097,
+      "step": 9005
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 29.318847585095682,
+      "learning_rate": 1.1487762447369375e-05,
+      "loss": 0.5742,
+      "step": 9006
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.59220452483218,
+      "learning_rate": 1.1486094795371703e-05,
+      "loss": 0.6388,
+      "step": 9007
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 29.313224437846962,
+      "learning_rate": 1.1484427101110359e-05,
+      "loss": 0.5789,
+      "step": 9008
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.15066274241755,
+      "learning_rate": 1.148275936463278e-05,
+      "loss": 0.6185,
+      "step": 9009
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 44.22365557415294,
+      "learning_rate": 1.148109158598639e-05,
+      "loss": 0.7961,
+      "step": 9010
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.01414801247635,
+      "learning_rate": 1.1479423765218625e-05,
+      "loss": 0.6156,
+      "step": 9011
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 50.059088758625066,
+      "learning_rate": 1.1477755902376908e-05,
+      "loss": 0.7246,
+      "step": 9012
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 36.948284827392136,
+      "learning_rate": 1.147608799750868e-05,
+      "loss": 0.6946,
+      "step": 9013
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 35.48459058769175,
+      "learning_rate": 1.1474420050661374e-05,
+      "loss": 0.6897,
+      "step": 9014
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 46.205630258051805,
+      "learning_rate": 1.1472752061882426e-05,
+      "loss": 0.6952,
+      "step": 9015
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 36.22408496972589,
+      "learning_rate": 1.1471084031219268e-05,
+      "loss": 0.6506,
+      "step": 9016
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 31.585959110634747,
+      "learning_rate": 1.146941595871934e-05,
+      "loss": 0.6275,
+      "step": 9017
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.012593123278,
+      "learning_rate": 1.1467747844430082e-05,
+      "loss": 0.6214,
+      "step": 9018
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.16408568943484,
+      "learning_rate": 1.1466079688398937e-05,
+      "loss": 0.7324,
+      "step": 9019
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.53744794050307,
+      "learning_rate": 1.146441149067334e-05,
+      "loss": 0.6367,
+      "step": 9020
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 32.73063298507553,
+      "learning_rate": 1.1462743251300739e-05,
+      "loss": 0.639,
+      "step": 9021
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.04466203602782,
+      "learning_rate": 1.1461074970328574e-05,
+      "loss": 0.5834,
+      "step": 9022
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.10519437031822,
+      "learning_rate": 1.1459406647804295e-05,
+      "loss": 0.5984,
+      "step": 9023
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.03013276201962,
+      "learning_rate": 1.1457738283775339e-05,
+      "loss": 0.6276,
+      "step": 9024
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 39.9119854467086,
+      "learning_rate": 1.1456069878289161e-05,
+      "loss": 0.6509,
+      "step": 9025
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.09761572239793,
+      "learning_rate": 1.1454401431393212e-05,
+      "loss": 0.5845,
+      "step": 9026
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 28.406201134903007,
+      "learning_rate": 1.145273294313493e-05,
+      "loss": 0.5744,
+      "step": 9027
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.651258148674586,
+      "learning_rate": 1.1451064413561776e-05,
+      "loss": 0.6283,
+      "step": 9028
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 32.37395512061975,
+      "learning_rate": 1.1449395842721201e-05,
+      "loss": 0.5409,
+      "step": 9029
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.43191746530485,
+      "learning_rate": 1.1447727230660654e-05,
+      "loss": 0.6109,
+      "step": 9030
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 38.59367161793027,
+      "learning_rate": 1.1446058577427592e-05,
+      "loss": 0.754,
+      "step": 9031
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.907148656933856,
+      "learning_rate": 1.1444389883069471e-05,
+      "loss": 0.6579,
+      "step": 9032
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.28568561065515,
+      "learning_rate": 1.1442721147633744e-05,
+      "loss": 0.5974,
+      "step": 9033
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.80923164407334,
+      "learning_rate": 1.1441052371167876e-05,
+      "loss": 0.6497,
+      "step": 9034
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 48.4485242837376,
+      "learning_rate": 1.1439383553719315e-05,
+      "loss": 0.7371,
+      "step": 9035
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.42010937152467,
+      "learning_rate": 1.1437714695335534e-05,
+      "loss": 0.5944,
+      "step": 9036
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 46.32003844320516,
+      "learning_rate": 1.1436045796063983e-05,
+      "loss": 0.7077,
+      "step": 9037
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.54659073016011,
+      "learning_rate": 1.1434376855952136e-05,
+      "loss": 0.6589,
+      "step": 9038
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 32.06210919564011,
+      "learning_rate": 1.1432707875047445e-05,
+      "loss": 0.6596,
+      "step": 9039
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.66884472924991,
+      "learning_rate": 1.143103885339738e-05,
+      "loss": 0.619,
+      "step": 9040
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 44.04254602199778,
+      "learning_rate": 1.1429369791049409e-05,
+      "loss": 0.6975,
+      "step": 9041
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 37.05439270790193,
+      "learning_rate": 1.1427700688050998e-05,
+      "loss": 0.6587,
+      "step": 9042
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.92228051239948,
+      "learning_rate": 1.1426031544449614e-05,
+      "loss": 0.57,
+      "step": 9043
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 30.90527676244719,
+      "learning_rate": 1.1424362360292725e-05,
+      "loss": 0.5841,
+      "step": 9044
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.414634387527514,
+      "learning_rate": 1.1422693135627804e-05,
+      "loss": 0.6492,
+      "step": 9045
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.278038334862394,
+      "learning_rate": 1.1421023870502324e-05,
+      "loss": 0.5897,
+      "step": 9046
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 30.48301886939825,
+      "learning_rate": 1.1419354564963756e-05,
+      "loss": 0.5725,
+      "step": 9047
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 38.94032793506494,
+      "learning_rate": 1.1417685219059576e-05,
+      "loss": 0.6111,
+      "step": 9048
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 30.765877319200204,
+      "learning_rate": 1.1416015832837258e-05,
+      "loss": 0.5793,
+      "step": 9049
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 45.675691328365545,
+      "learning_rate": 1.1414346406344272e-05,
+      "loss": 0.6839,
+      "step": 9050
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 36.110038975710715,
+      "learning_rate": 1.1412676939628108e-05,
+      "loss": 0.6389,
+      "step": 9051
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 34.51961486675051,
+      "learning_rate": 1.1411007432736236e-05,
+      "loss": 0.5832,
+      "step": 9052
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.24713979110504,
+      "learning_rate": 1.1409337885716141e-05,
+      "loss": 0.6603,
+      "step": 9053
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 46.132376238692544,
+      "learning_rate": 1.1407668298615296e-05,
+      "loss": 0.6123,
+      "step": 9054
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 43.3305621502064,
+      "learning_rate": 1.1405998671481191e-05,
+      "loss": 0.6963,
+      "step": 9055
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 50.458632115750774,
+      "learning_rate": 1.1404329004361306e-05,
+      "loss": 0.5898,
+      "step": 9056
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 31.552074625478564,
+      "learning_rate": 1.1402659297303124e-05,
+      "loss": 0.649,
+      "step": 9057
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.27105178595466,
+      "learning_rate": 1.1400989550354133e-05,
+      "loss": 0.7114,
+      "step": 9058
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.352104066362855,
+      "learning_rate": 1.1399319763561821e-05,
+      "loss": 0.5812,
+      "step": 9059
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.626650789773787,
+      "learning_rate": 1.1397649936973672e-05,
+      "loss": 0.6523,
+      "step": 9060
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.625483671266736,
+      "learning_rate": 1.1395980070637175e-05,
+      "loss": 0.6313,
+      "step": 9061
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.682368502073714,
+      "learning_rate": 1.139431016459982e-05,
+      "loss": 0.6831,
+      "step": 9062
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 29.81021865976957,
+      "learning_rate": 1.1392640218909103e-05,
+      "loss": 0.6628,
+      "step": 9063
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.20569744692689,
+      "learning_rate": 1.1390970233612513e-05,
+      "loss": 0.6453,
+      "step": 9064
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 33.18776457801428,
+      "learning_rate": 1.138930020875754e-05,
+      "loss": 0.5936,
+      "step": 9065
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 44.62095474693049,
+      "learning_rate": 1.1387630144391685e-05,
+      "loss": 0.6036,
+      "step": 9066
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.65009715092105,
+      "learning_rate": 1.138596004056244e-05,
+      "loss": 0.6175,
+      "step": 9067
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 41.31087389413408,
+      "learning_rate": 1.1384289897317302e-05,
+      "loss": 0.6745,
+      "step": 9068
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 54.79912175080385,
+      "learning_rate": 1.1382619714703767e-05,
+      "loss": 0.6794,
+      "step": 9069
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 43.08198706603614,
+      "learning_rate": 1.1380949492769339e-05,
+      "loss": 0.6592,
+      "step": 9070
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.511306829988783,
+      "learning_rate": 1.1379279231561514e-05,
+      "loss": 0.5942,
+      "step": 9071
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 73.98910209595573,
+      "learning_rate": 1.1377608931127792e-05,
+      "loss": 0.6388,
+      "step": 9072
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 40.23989593255406,
+      "learning_rate": 1.137593859151568e-05,
+      "loss": 0.6542,
+      "step": 9073
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.93034828326009,
+      "learning_rate": 1.137426821277268e-05,
+      "loss": 0.6783,
+      "step": 9074
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 32.202922737119394,
+      "learning_rate": 1.1372597794946298e-05,
+      "loss": 0.5767,
+      "step": 9075
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 32.1395808946503,
+      "learning_rate": 1.1370927338084035e-05,
+      "loss": 0.5111,
+      "step": 9076
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 42.59713196569554,
+      "learning_rate": 1.1369256842233399e-05,
+      "loss": 0.7008,
+      "step": 9077
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.36919218739527,
+      "learning_rate": 1.1367586307441902e-05,
+      "loss": 0.5908,
+      "step": 9078
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.50415652763422,
+      "learning_rate": 1.136591573375705e-05,
+      "loss": 0.6503,
+      "step": 9079
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 33.04019843940742,
+      "learning_rate": 1.1364245121226354e-05,
+      "loss": 0.5564,
+      "step": 9080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.2830665336604,
+      "learning_rate": 1.1362574469897322e-05,
+      "loss": 0.5955,
+      "step": 9081
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.31411572799566,
+      "learning_rate": 1.136090377981747e-05,
+      "loss": 0.5992,
+      "step": 9082
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.20215299644578,
+      "learning_rate": 1.1359233051034316e-05,
+      "loss": 0.6028,
+      "step": 9083
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.20235572033841,
+      "learning_rate": 1.1357562283595363e-05,
+      "loss": 0.6529,
+      "step": 9084
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.837051798972254,
+      "learning_rate": 1.1355891477548136e-05,
+      "loss": 0.6381,
+      "step": 9085
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.436837885199612,
+      "learning_rate": 1.1354220632940147e-05,
+      "loss": 0.5504,
+      "step": 9086
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 54.2957397978965,
+      "learning_rate": 1.1352549749818916e-05,
+      "loss": 0.7093,
+      "step": 9087
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 47.612425633475304,
+      "learning_rate": 1.135087882823196e-05,
+      "loss": 0.6919,
+      "step": 9088
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.709003685562653,
+      "learning_rate": 1.1349207868226801e-05,
+      "loss": 0.5913,
+      "step": 9089
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.96774017167223,
+      "learning_rate": 1.1347536869850959e-05,
+      "loss": 0.6332,
+      "step": 9090
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 38.630991199480064,
+      "learning_rate": 1.134586583315196e-05,
+      "loss": 0.7198,
+      "step": 9091
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 45.73289483180359,
+      "learning_rate": 1.134419475817732e-05,
+      "loss": 0.5975,
+      "step": 9092
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 46.70097723393802,
+      "learning_rate": 1.1342523644974567e-05,
+      "loss": 0.7694,
+      "step": 9093
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 35.50455236460842,
+      "learning_rate": 1.134085249359123e-05,
+      "loss": 0.615,
+      "step": 9094
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 38.90520814167073,
+      "learning_rate": 1.1339181304074833e-05,
+      "loss": 0.7104,
+      "step": 9095
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.033598120475197,
+      "learning_rate": 1.13375100764729e-05,
+      "loss": 0.6064,
+      "step": 9096
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.08061298381553,
+      "learning_rate": 1.1335838810832966e-05,
+      "loss": 0.6512,
+      "step": 9097
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.01715010175349,
+      "learning_rate": 1.1334167507202553e-05,
+      "loss": 0.5853,
+      "step": 9098
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.19260601283726,
+      "learning_rate": 1.1332496165629201e-05,
+      "loss": 0.6639,
+      "step": 9099
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.373038731662085,
+      "learning_rate": 1.1330824786160437e-05,
+      "loss": 0.6377,
+      "step": 9100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 41.84543535287083,
+      "learning_rate": 1.1329153368843793e-05,
+      "loss": 0.7269,
+      "step": 9101
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 44.789265710518414,
+      "learning_rate": 1.1327481913726806e-05,
+      "loss": 0.7316,
+      "step": 9102
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 43.01952246299827,
+      "learning_rate": 1.1325810420857009e-05,
+      "loss": 0.748,
+      "step": 9103
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 46.43472109985964,
+      "learning_rate": 1.1324138890281937e-05,
+      "loss": 0.6129,
+      "step": 9104
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 49.074696350324224,
+      "learning_rate": 1.1322467322049135e-05,
+      "loss": 0.5914,
+      "step": 9105
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.893743077063746,
+      "learning_rate": 1.1320795716206132e-05,
+      "loss": 0.6911,
+      "step": 9106
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 42.76986242673545,
+      "learning_rate": 1.1319124072800471e-05,
+      "loss": 0.6673,
+      "step": 9107
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 38.42203525108002,
+      "learning_rate": 1.1317452391879695e-05,
+      "loss": 0.6001,
+      "step": 9108
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 47.113076856932196,
+      "learning_rate": 1.1315780673491344e-05,
+      "loss": 0.6673,
+      "step": 9109
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 38.72201617906855,
+      "learning_rate": 1.131410891768296e-05,
+      "loss": 0.6692,
+      "step": 9110
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 35.36538930694575,
+      "learning_rate": 1.1312437124502086e-05,
+      "loss": 0.628,
+      "step": 9111
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.759857730538286,
+      "learning_rate": 1.131076529399627e-05,
+      "loss": 0.6866,
+      "step": 9112
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 40.530868687302934,
+      "learning_rate": 1.1309093426213051e-05,
+      "loss": 0.6193,
+      "step": 9113
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 41.36579255515879,
+      "learning_rate": 1.1307421521199988e-05,
+      "loss": 0.7113,
+      "step": 9114
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 35.055435730387956,
+      "learning_rate": 1.1305749579004618e-05,
+      "loss": 0.6396,
+      "step": 9115
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.37480604000221,
+      "learning_rate": 1.1304077599674493e-05,
+      "loss": 0.546,
+      "step": 9116
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 34.54541943889983,
+      "learning_rate": 1.1302405583257163e-05,
+      "loss": 0.6647,
+      "step": 9117
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.782123308051883,
+      "learning_rate": 1.1300733529800183e-05,
+      "loss": 0.5291,
+      "step": 9118
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 40.75047206605231,
+      "learning_rate": 1.1299061439351102e-05,
+      "loss": 0.6686,
+      "step": 9119
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 32.65233621461437,
+      "learning_rate": 1.1297389311957472e-05,
+      "loss": 0.6273,
+      "step": 9120
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.78853152011997,
+      "learning_rate": 1.1295717147666848e-05,
+      "loss": 0.6285,
+      "step": 9121
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 35.52346609134165,
+      "learning_rate": 1.129404494652679e-05,
+      "loss": 0.6828,
+      "step": 9122
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.73898823889622,
+      "learning_rate": 1.1292372708584847e-05,
+      "loss": 0.6744,
+      "step": 9123
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 33.36950728047826,
+      "learning_rate": 1.1290700433888584e-05,
+      "loss": 0.6123,
+      "step": 9124
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.790854179311506,
+      "learning_rate": 1.1289028122485553e-05,
+      "loss": 0.6218,
+      "step": 9125
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 38.5540985031512,
+      "learning_rate": 1.1287355774423316e-05,
+      "loss": 0.6667,
+      "step": 9126
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.53571364347316,
+      "learning_rate": 1.1285683389749434e-05,
+      "loss": 0.6856,
+      "step": 9127
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 33.68818137306084,
+      "learning_rate": 1.128401096851147e-05,
+      "loss": 0.6221,
+      "step": 9128
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 32.81045338378754,
+      "learning_rate": 1.1282338510756985e-05,
+      "loss": 0.6947,
+      "step": 9129
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 36.507306882988416,
+      "learning_rate": 1.1280666016533542e-05,
+      "loss": 0.6573,
+      "step": 9130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 49.04985893338107,
+      "learning_rate": 1.1278993485888704e-05,
+      "loss": 0.5921,
+      "step": 9131
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.50225553000954,
+      "learning_rate": 1.1277320918870043e-05,
+      "loss": 0.6718,
+      "step": 9132
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 38.95548677839756,
+      "learning_rate": 1.1275648315525123e-05,
+      "loss": 0.6132,
+      "step": 9133
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.35915392892112,
+      "learning_rate": 1.127397567590151e-05,
+      "loss": 0.5792,
+      "step": 9134
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.1522197736039,
+      "learning_rate": 1.1272303000046774e-05,
+      "loss": 0.6083,
+      "step": 9135
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 34.55653954583146,
+      "learning_rate": 1.1270630288008484e-05,
+      "loss": 0.5983,
+      "step": 9136
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.053138911184732,
+      "learning_rate": 1.1268957539834213e-05,
+      "loss": 0.5945,
+      "step": 9137
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.1810192075275,
+      "learning_rate": 1.1267284755571528e-05,
+      "loss": 0.7402,
+      "step": 9138
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 49.56404735752196,
+      "learning_rate": 1.1265611935268013e-05,
+      "loss": 0.7606,
+      "step": 9139
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.30130012535356,
+      "learning_rate": 1.1263939078971229e-05,
+      "loss": 0.6268,
+      "step": 9140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.731574422081806,
+      "learning_rate": 1.1262266186728763e-05,
+      "loss": 0.602,
+      "step": 9141
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.069071668650764,
+      "learning_rate": 1.1260593258588179e-05,
+      "loss": 0.6687,
+      "step": 9142
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.62091880162121,
+      "learning_rate": 1.1258920294597063e-05,
+      "loss": 0.6026,
+      "step": 9143
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 39.52559223925701,
+      "learning_rate": 1.1257247294802988e-05,
+      "loss": 0.5281,
+      "step": 9144
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.8257736851909,
+      "learning_rate": 1.125557425925354e-05,
+      "loss": 0.711,
+      "step": 9145
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.136460499826576,
+      "learning_rate": 1.125390118799629e-05,
+      "loss": 0.6017,
+      "step": 9146
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.332058241766916,
+      "learning_rate": 1.1252228081078826e-05,
+      "loss": 0.6293,
+      "step": 9147
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.80899643865927,
+      "learning_rate": 1.1250554938548726e-05,
+      "loss": 0.7762,
+      "step": 9148
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.33496806021329,
+      "learning_rate": 1.1248881760453578e-05,
+      "loss": 0.5904,
+      "step": 9149
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 39.196228983799514,
+      "learning_rate": 1.124720854684096e-05,
+      "loss": 0.644,
+      "step": 9150
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 34.29542749083134,
+      "learning_rate": 1.1245535297758463e-05,
+      "loss": 0.7032,
+      "step": 9151
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.26336734098848,
+      "learning_rate": 1.124386201325367e-05,
+      "loss": 0.669,
+      "step": 9152
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 34.05565055085776,
+      "learning_rate": 1.124218869337417e-05,
+      "loss": 0.7018,
+      "step": 9153
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.94820538967874,
+      "learning_rate": 1.1240515338167548e-05,
+      "loss": 0.6267,
+      "step": 9154
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.95830556942075,
+      "learning_rate": 1.1238841947681399e-05,
+      "loss": 0.6019,
+      "step": 9155
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.8333899329383,
+      "learning_rate": 1.1237168521963307e-05,
+      "loss": 0.5949,
+      "step": 9156
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.64714700376534,
+      "learning_rate": 1.1235495061060864e-05,
+      "loss": 0.6493,
+      "step": 9157
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 48.04409569121783,
+      "learning_rate": 1.123382156502167e-05,
+      "loss": 0.753,
+      "step": 9158
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 32.070089665148686,
+      "learning_rate": 1.1232148033893307e-05,
+      "loss": 0.5899,
+      "step": 9159
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.161240159836918,
+      "learning_rate": 1.1230474467723376e-05,
+      "loss": 0.6078,
+      "step": 9160
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.947308892794645,
+      "learning_rate": 1.1228800866559472e-05,
+      "loss": 0.7197,
+      "step": 9161
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 32.049515658630305,
+      "learning_rate": 1.1227127230449191e-05,
+      "loss": 0.6692,
+      "step": 9162
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 33.90792717367354,
+      "learning_rate": 1.1225453559440128e-05,
+      "loss": 0.6312,
+      "step": 9163
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 41.89979002770228,
+      "learning_rate": 1.1223779853579884e-05,
+      "loss": 0.6425,
+      "step": 9164
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.66772062793697,
+      "learning_rate": 1.1222106112916053e-05,
+      "loss": 0.6664,
+      "step": 9165
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.51996969716274,
+      "learning_rate": 1.1220432337496244e-05,
+      "loss": 0.6593,
+      "step": 9166
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.188166083337734,
+      "learning_rate": 1.121875852736805e-05,
+      "loss": 0.6288,
+      "step": 9167
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.174481512582936,
+      "learning_rate": 1.1217084682579077e-05,
+      "loss": 0.6027,
+      "step": 9168
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 53.17833162487171,
+      "learning_rate": 1.1215410803176927e-05,
+      "loss": 0.6766,
+      "step": 9169
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 36.986036510720965,
+      "learning_rate": 1.1213736889209207e-05,
+      "loss": 0.6162,
+      "step": 9170
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 32.10177265951621,
+      "learning_rate": 1.1212062940723519e-05,
+      "loss": 0.6132,
+      "step": 9171
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.608184244080284,
+      "learning_rate": 1.1210388957767471e-05,
+      "loss": 0.6285,
+      "step": 9172
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 39.67057693599454,
+      "learning_rate": 1.1208714940388668e-05,
+      "loss": 0.7004,
+      "step": 9173
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.28081431298148,
+      "learning_rate": 1.1207040888634719e-05,
+      "loss": 0.5454,
+      "step": 9174
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 33.30568626457875,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.6815,
+      "step": 9175
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 36.992460719935444,
+      "learning_rate": 1.1203692682191819e-05,
+      "loss": 0.6285,
+      "step": 9176
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.52877652056325,
+      "learning_rate": 1.1202018527598091e-05,
+      "loss": 0.649,
+      "step": 9177
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.582327190043685,
+      "learning_rate": 1.120034433881966e-05,
+      "loss": 0.6281,
+      "step": 9178
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.09552493526933,
+      "learning_rate": 1.1198670115904136e-05,
+      "loss": 0.5965,
+      "step": 9179
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 51.05461136243844,
+      "learning_rate": 1.1196995858899138e-05,
+      "loss": 0.6642,
+      "step": 9180
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 46.328071015247716,
+      "learning_rate": 1.1195321567852275e-05,
+      "loss": 0.7302,
+      "step": 9181
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.53859720090588,
+      "learning_rate": 1.1193647242811165e-05,
+      "loss": 0.6464,
+      "step": 9182
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 35.34997994099956,
+      "learning_rate": 1.1191972883823433e-05,
+      "loss": 0.6927,
+      "step": 9183
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 39.09323076167112,
+      "learning_rate": 1.119029849093668e-05,
+      "loss": 0.5366,
+      "step": 9184
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 44.51069236465846,
+      "learning_rate": 1.118862406419854e-05,
+      "loss": 0.6704,
+      "step": 9185
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.23066937141799,
+      "learning_rate": 1.1186949603656624e-05,
+      "loss": 0.6211,
+      "step": 9186
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.20096736644859,
+      "learning_rate": 1.1185275109358558e-05,
+      "loss": 0.5901,
+      "step": 9187
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 43.66022093837123,
+      "learning_rate": 1.118360058135196e-05,
+      "loss": 0.6405,
+      "step": 9188
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.574018968629446,
+      "learning_rate": 1.1181926019684454e-05,
+      "loss": 0.5957,
+      "step": 9189
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.16815752855321,
+      "learning_rate": 1.1180251424403666e-05,
+      "loss": 0.6043,
+      "step": 9190
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 40.86045040079367,
+      "learning_rate": 1.1178576795557214e-05,
+      "loss": 0.7153,
+      "step": 9191
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 36.55096945918529,
+      "learning_rate": 1.117690213319273e-05,
+      "loss": 0.6761,
+      "step": 9192
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 40.710418246919446,
+      "learning_rate": 1.1175227437357835e-05,
+      "loss": 0.6848,
+      "step": 9193
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 30.52422515110645,
+      "learning_rate": 1.1173552708100162e-05,
+      "loss": 0.5676,
+      "step": 9194
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 38.9244974540017,
+      "learning_rate": 1.1171877945467333e-05,
+      "loss": 0.6401,
+      "step": 9195
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 43.03340102549182,
+      "learning_rate": 1.1170203149506984e-05,
+      "loss": 0.6656,
+      "step": 9196
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.37381764318584,
+      "learning_rate": 1.1168528320266743e-05,
+      "loss": 0.5967,
+      "step": 9197
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 48.33732531466411,
+      "learning_rate": 1.1166853457794238e-05,
+      "loss": 0.6806,
+      "step": 9198
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 44.20915870842794,
+      "learning_rate": 1.1165178562137103e-05,
+      "loss": 0.7367,
+      "step": 9199
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 36.90204487228622,
+      "learning_rate": 1.1163503633342971e-05,
+      "loss": 0.6407,
+      "step": 9200
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 46.995710258323854,
+      "learning_rate": 1.1161828671459475e-05,
+      "loss": 0.731,
+      "step": 9201
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 45.761533747717095,
+      "learning_rate": 1.1160153676534256e-05,
+      "loss": 0.7693,
+      "step": 9202
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.838470524305464,
+      "learning_rate": 1.115847864861494e-05,
+      "loss": 0.6099,
+      "step": 9203
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 39.57199792614651,
+      "learning_rate": 1.1156803587749172e-05,
+      "loss": 0.6677,
+      "step": 9204
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.46415994608511,
+      "learning_rate": 1.1155128493984588e-05,
+      "loss": 0.6125,
+      "step": 9205
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 48.21915780251621,
+      "learning_rate": 1.1153453367368824e-05,
+      "loss": 0.6503,
+      "step": 9206
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.18243826472842,
+      "learning_rate": 1.115177820794952e-05,
+      "loss": 0.6336,
+      "step": 9207
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.75911355795369,
+      "learning_rate": 1.1150103015774318e-05,
+      "loss": 0.5836,
+      "step": 9208
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.007837358273505,
+      "learning_rate": 1.1148427790890857e-05,
+      "loss": 0.6108,
+      "step": 9209
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 35.71282848491484,
+      "learning_rate": 1.1146752533346786e-05,
+      "loss": 0.6134,
+      "step": 9210
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 38.2294260648542,
+      "learning_rate": 1.1145077243189739e-05,
+      "loss": 0.7152,
+      "step": 9211
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.96296950588637,
+      "learning_rate": 1.114340192046737e-05,
+      "loss": 0.6951,
+      "step": 9212
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 48.6004350485776,
+      "learning_rate": 1.1141726565227313e-05,
+      "loss": 0.5605,
+      "step": 9213
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 39.19073482711268,
+      "learning_rate": 1.1140051177517224e-05,
+      "loss": 0.6219,
+      "step": 9214
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.77861783661277,
+      "learning_rate": 1.1138375757384747e-05,
+      "loss": 0.6447,
+      "step": 9215
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.30532245834581,
+      "learning_rate": 1.113670030487753e-05,
+      "loss": 0.6474,
+      "step": 9216
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 48.08930918991602,
+      "learning_rate": 1.1135024820043218e-05,
+      "loss": 0.6756,
+      "step": 9217
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 42.00660033815644,
+      "learning_rate": 1.1133349302929468e-05,
+      "loss": 0.6528,
+      "step": 9218
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 40.07962052378829,
+      "learning_rate": 1.1131673753583922e-05,
+      "loss": 0.6521,
+      "step": 9219
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 46.86884816809696,
+      "learning_rate": 1.112999817205424e-05,
+      "loss": 0.5681,
+      "step": 9220
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 38.091944215954406,
+      "learning_rate": 1.1128322558388071e-05,
+      "loss": 0.649,
+      "step": 9221
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.36936265147593,
+      "learning_rate": 1.1126646912633068e-05,
+      "loss": 0.6487,
+      "step": 9222
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 42.170069109937174,
+      "learning_rate": 1.1124971234836886e-05,
+      "loss": 0.6914,
+      "step": 9223
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 39.86959924566311,
+      "learning_rate": 1.1123295525047178e-05,
+      "loss": 0.61,
+      "step": 9224
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 30.860504512154947,
+      "learning_rate": 1.1121619783311606e-05,
+      "loss": 0.6061,
+      "step": 9225
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.088510638886504,
+      "learning_rate": 1.111994400967782e-05,
+      "loss": 0.5694,
+      "step": 9226
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 42.544127211307206,
+      "learning_rate": 1.1118268204193485e-05,
+      "loss": 0.6476,
+      "step": 9227
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 39.95169371546409,
+      "learning_rate": 1.1116592366906254e-05,
+      "loss": 0.666,
+      "step": 9228
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 35.08829734475216,
+      "learning_rate": 1.111491649786379e-05,
+      "loss": 0.5978,
+      "step": 9229
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 43.09030170212883,
+      "learning_rate": 1.1113240597113752e-05,
+      "loss": 0.6535,
+      "step": 9230
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 29.836016436859733,
+      "learning_rate": 1.1111564664703805e-05,
+      "loss": 0.6107,
+      "step": 9231
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.72921612344541,
+      "learning_rate": 1.1109888700681608e-05,
+      "loss": 0.6406,
+      "step": 9232
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.72813032976043,
+      "learning_rate": 1.1108212705094826e-05,
+      "loss": 0.656,
+      "step": 9233
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 24.981307379967113,
+      "learning_rate": 1.1106536677991122e-05,
+      "loss": 0.5505,
+      "step": 9234
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 36.875242442611004,
+      "learning_rate": 1.1104860619418162e-05,
+      "loss": 0.6142,
+      "step": 9235
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.273525233300695,
+      "learning_rate": 1.1103184529423612e-05,
+      "loss": 0.6232,
+      "step": 9236
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 42.934683636581184,
+      "learning_rate": 1.1101508408055141e-05,
+      "loss": 0.7833,
+      "step": 9237
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 40.64550268285414,
+      "learning_rate": 1.1099832255360412e-05,
+      "loss": 0.7015,
+      "step": 9238
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.94288735456886,
+      "learning_rate": 1.1098156071387101e-05,
+      "loss": 0.6206,
+      "step": 9239
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 42.003752336456884,
+      "learning_rate": 1.1096479856182872e-05,
+      "loss": 0.5933,
+      "step": 9240
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 43.824312307399204,
+      "learning_rate": 1.1094803609795398e-05,
+      "loss": 0.6233,
+      "step": 9241
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.912137919481985,
+      "learning_rate": 1.109312733227235e-05,
+      "loss": 0.5674,
+      "step": 9242
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 35.55257972633774,
+      "learning_rate": 1.10914510236614e-05,
+      "loss": 0.6504,
+      "step": 9243
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 36.73936052455818,
+      "learning_rate": 1.1089774684010219e-05,
+      "loss": 0.6476,
+      "step": 9244
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 40.16872169585037,
+      "learning_rate": 1.1088098313366484e-05,
+      "loss": 0.6244,
+      "step": 9245
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 37.24954245113501,
+      "learning_rate": 1.108642191177787e-05,
+      "loss": 0.6003,
+      "step": 9246
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.7785505596128,
+      "learning_rate": 1.1084745479292054e-05,
+      "loss": 0.6757,
+      "step": 9247
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.34092868813323,
+      "learning_rate": 1.108306901595671e-05,
+      "loss": 0.6305,
+      "step": 9248
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 34.75381709821958,
+      "learning_rate": 1.1081392521819518e-05,
+      "loss": 0.6649,
+      "step": 9249
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 31.91879329659786,
+      "learning_rate": 1.1079715996928156e-05,
+      "loss": 0.6086,
+      "step": 9250
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.75787120980612,
+      "learning_rate": 1.1078039441330298e-05,
+      "loss": 0.646,
+      "step": 9251
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 31.596959895254876,
+      "learning_rate": 1.1076362855073635e-05,
+      "loss": 0.5636,
+      "step": 9252
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 39.99423002708187,
+      "learning_rate": 1.1074686238205837e-05,
+      "loss": 0.5983,
+      "step": 9253
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 48.08174364160608,
+      "learning_rate": 1.1073009590774595e-05,
+      "loss": 0.6909,
+      "step": 9254
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.37036415999601,
+      "learning_rate": 1.1071332912827584e-05,
+      "loss": 0.5845,
+      "step": 9255
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.46031139620531,
+      "learning_rate": 1.1069656204412496e-05,
+      "loss": 0.5617,
+      "step": 9256
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.56058179019742,
+      "learning_rate": 1.1067979465577005e-05,
+      "loss": 0.6486,
+      "step": 9257
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 31.46199191883913,
+      "learning_rate": 1.106630269636881e-05,
+      "loss": 0.5675,
+      "step": 9258
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 45.94536515978666,
+      "learning_rate": 1.1064625896835585e-05,
+      "loss": 0.7242,
+      "step": 9259
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.23947539799663,
+      "learning_rate": 1.1062949067025024e-05,
+      "loss": 0.645,
+      "step": 9260
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.76175662003389,
+      "learning_rate": 1.1061272206984815e-05,
+      "loss": 0.6386,
+      "step": 9261
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.806107816283635,
+      "learning_rate": 1.1059595316762644e-05,
+      "loss": 0.739,
+      "step": 9262
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 60.53823927200107,
+      "learning_rate": 1.1057918396406197e-05,
+      "loss": 0.6562,
+      "step": 9263
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 39.53419770118181,
+      "learning_rate": 1.1056241445963177e-05,
+      "loss": 0.6768,
+      "step": 9264
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 41.02659122605362,
+      "learning_rate": 1.1054564465481263e-05,
+      "loss": 0.6743,
+      "step": 9265
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.58225268332948,
+      "learning_rate": 1.1052887455008157e-05,
+      "loss": 0.56,
+      "step": 9266
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 40.267986875221766,
+      "learning_rate": 1.1051210414591544e-05,
+      "loss": 0.6922,
+      "step": 9267
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.53365936931007,
+      "learning_rate": 1.1049533344279122e-05,
+      "loss": 0.6376,
+      "step": 9268
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.19232634277492,
+      "learning_rate": 1.1047856244118591e-05,
+      "loss": 0.6295,
+      "step": 9269
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 42.280986641277224,
+      "learning_rate": 1.1046179114157636e-05,
+      "loss": 0.6006,
+      "step": 9270
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.761886188822416,
+      "learning_rate": 1.1044501954443962e-05,
+      "loss": 0.6628,
+      "step": 9271
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 31.381584935470258,
+      "learning_rate": 1.104282476502526e-05,
+      "loss": 0.597,
+      "step": 9272
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 45.90653932401035,
+      "learning_rate": 1.1041147545949233e-05,
+      "loss": 0.7128,
+      "step": 9273
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 44.335349195164994,
+      "learning_rate": 1.103947029726358e-05,
+      "loss": 0.7423,
+      "step": 9274
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 33.75137982196959,
+      "learning_rate": 1.1037793019015999e-05,
+      "loss": 0.5884,
+      "step": 9275
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 33.05912801785697,
+      "learning_rate": 1.1036115711254191e-05,
+      "loss": 0.574,
+      "step": 9276
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 32.51157723510259,
+      "learning_rate": 1.103443837402586e-05,
+      "loss": 0.677,
+      "step": 9277
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 35.788326523163924,
+      "learning_rate": 1.1032761007378708e-05,
+      "loss": 0.6552,
+      "step": 9278
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 44.639995021734336,
+      "learning_rate": 1.1031083611360433e-05,
+      "loss": 0.631,
+      "step": 9279
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.82213898084294,
+      "learning_rate": 1.1029406186018745e-05,
+      "loss": 0.5808,
+      "step": 9280
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.332754232690185,
+      "learning_rate": 1.102772873140135e-05,
+      "loss": 0.6429,
+      "step": 9281
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 32.74331735906555,
+      "learning_rate": 1.1026051247555947e-05,
+      "loss": 0.6145,
+      "step": 9282
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 45.35569960381258,
+      "learning_rate": 1.1024373734530247e-05,
+      "loss": 0.6663,
+      "step": 9283
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.27936352851598,
+      "learning_rate": 1.102269619237196e-05,
+      "loss": 0.5604,
+      "step": 9284
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 30.847801858346504,
+      "learning_rate": 1.102101862112879e-05,
+      "loss": 0.5362,
+      "step": 9285
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 52.456950612083965,
+      "learning_rate": 1.1019341020848449e-05,
+      "loss": 0.666,
+      "step": 9286
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 32.862336127364856,
+      "learning_rate": 1.1017663391578644e-05,
+      "loss": 0.647,
+      "step": 9287
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.029203873591385,
+      "learning_rate": 1.1015985733367087e-05,
+      "loss": 0.6119,
+      "step": 9288
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.72193176119711,
+      "learning_rate": 1.1014308046261491e-05,
+      "loss": 0.628,
+      "step": 9289
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 43.28156496064203,
+      "learning_rate": 1.1012630330309567e-05,
+      "loss": 0.5989,
+      "step": 9290
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 50.67067397367838,
+      "learning_rate": 1.1010952585559029e-05,
+      "loss": 0.6535,
+      "step": 9291
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.54665926720997,
+      "learning_rate": 1.1009274812057592e-05,
+      "loss": 0.6264,
+      "step": 9292
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 30.06051957156022,
+      "learning_rate": 1.1007597009852968e-05,
+      "loss": 0.5784,
+      "step": 9293
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 30.363641129930983,
+      "learning_rate": 1.1005919178992876e-05,
+      "loss": 0.6648,
+      "step": 9294
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.29644344814315,
+      "learning_rate": 1.1004241319525028e-05,
+      "loss": 0.6674,
+      "step": 9295
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 42.05488658094685,
+      "learning_rate": 1.1002563431497151e-05,
+      "loss": 0.706,
+      "step": 9296
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 39.29021860706062,
+      "learning_rate": 1.1000885514956949e-05,
+      "loss": 0.596,
+      "step": 9297
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.361070343185986,
+      "learning_rate": 1.0999207569952155e-05,
+      "loss": 0.6891,
+      "step": 9298
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.851671062643305,
+      "learning_rate": 1.0997529596530477e-05,
+      "loss": 0.6509,
+      "step": 9299
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 42.52491142357914,
+      "learning_rate": 1.0995851594739642e-05,
+      "loss": 0.7106,
+      "step": 9300
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 65.14800083257882,
+      "learning_rate": 1.099417356462737e-05,
+      "loss": 0.6353,
+      "step": 9301
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.576216448181384,
+      "learning_rate": 1.0992495506241387e-05,
+      "loss": 0.6751,
+      "step": 9302
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 43.90483682876466,
+      "learning_rate": 1.0990817419629409e-05,
+      "loss": 0.616,
+      "step": 9303
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.79399865443721,
+      "learning_rate": 1.0989139304839165e-05,
+      "loss": 0.655,
+      "step": 9304
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.185962959760246,
+      "learning_rate": 1.0987461161918375e-05,
+      "loss": 0.6202,
+      "step": 9305
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 35.62302945024336,
+      "learning_rate": 1.0985782990914771e-05,
+      "loss": 0.6182,
+      "step": 9306
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 40.57920989303775,
+      "learning_rate": 1.098410479187607e-05,
+      "loss": 0.7284,
+      "step": 9307
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.95106388304604,
+      "learning_rate": 1.0982426564850009e-05,
+      "loss": 0.6177,
+      "step": 9308
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.10163378796187,
+      "learning_rate": 1.098074830988431e-05,
+      "loss": 0.5698,
+      "step": 9309
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 45.46837035345089,
+      "learning_rate": 1.0979070027026703e-05,
+      "loss": 0.8963,
+      "step": 9310
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.30134547127607,
+      "learning_rate": 1.0977391716324916e-05,
+      "loss": 0.6508,
+      "step": 9311
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.288184913709046,
+      "learning_rate": 1.097571337782668e-05,
+      "loss": 0.6242,
+      "step": 9312
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 32.54231129120574,
+      "learning_rate": 1.0974035011579731e-05,
+      "loss": 0.6174,
+      "step": 9313
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 35.46612624828357,
+      "learning_rate": 1.097235661763179e-05,
+      "loss": 0.659,
+      "step": 9314
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 40.23933769952292,
+      "learning_rate": 1.09706781960306e-05,
+      "loss": 0.6785,
+      "step": 9315
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.673875288954335,
+      "learning_rate": 1.0968999746823888e-05,
+      "loss": 0.5926,
+      "step": 9316
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 42.78415941999496,
+      "learning_rate": 1.0967321270059391e-05,
+      "loss": 0.6639,
+      "step": 9317
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 46.37655005730693,
+      "learning_rate": 1.0965642765784841e-05,
+      "loss": 0.6609,
+      "step": 9318
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.3379833996313,
+      "learning_rate": 1.0963964234047977e-05,
+      "loss": 0.685,
+      "step": 9319
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.209656002409154,
+      "learning_rate": 1.0962285674896534e-05,
+      "loss": 0.5609,
+      "step": 9320
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.001073759501786,
+      "learning_rate": 1.096060708837825e-05,
+      "loss": 0.6387,
+      "step": 9321
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 46.49672686942114,
+      "learning_rate": 1.0958928474540857e-05,
+      "loss": 0.6318,
+      "step": 9322
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 46.172381070069726,
+      "learning_rate": 1.0957249833432108e-05,
+      "loss": 0.6579,
+      "step": 9323
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.71606401512796,
+      "learning_rate": 1.0955571165099728e-05,
+      "loss": 0.6884,
+      "step": 9324
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.758055234291064,
+      "learning_rate": 1.0953892469591464e-05,
+      "loss": 0.6155,
+      "step": 9325
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.216688991123746,
+      "learning_rate": 1.0952213746955056e-05,
+      "loss": 0.695,
+      "step": 9326
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 32.82320307965114,
+      "learning_rate": 1.0950534997238246e-05,
+      "loss": 0.5931,
+      "step": 9327
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 39.84044202947615,
+      "learning_rate": 1.0948856220488779e-05,
+      "loss": 0.6167,
+      "step": 9328
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 46.54090807247273,
+      "learning_rate": 1.0947177416754396e-05,
+      "loss": 0.6162,
+      "step": 9329
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.81961122590373,
+      "learning_rate": 1.0945498586082838e-05,
+      "loss": 0.6548,
+      "step": 9330
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.59694689940551,
+      "learning_rate": 1.0943819728521854e-05,
+      "loss": 0.6535,
+      "step": 9331
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.036143017149264,
+      "learning_rate": 1.0942140844119194e-05,
+      "loss": 0.612,
+      "step": 9332
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.51795677682288,
+      "learning_rate": 1.0940461932922593e-05,
+      "loss": 0.598,
+      "step": 9333
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 43.41404603577171,
+      "learning_rate": 1.0938782994979808e-05,
+      "loss": 0.6856,
+      "step": 9334
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.41339059426719,
+      "learning_rate": 1.0937104030338584e-05,
+      "loss": 0.6468,
+      "step": 9335
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 40.108993962390905,
+      "learning_rate": 1.0935425039046668e-05,
+      "loss": 0.6494,
+      "step": 9336
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 30.611826208011834,
+      "learning_rate": 1.0933746021151813e-05,
+      "loss": 0.5408,
+      "step": 9337
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.37448971990263,
+      "learning_rate": 1.0932066976701767e-05,
+      "loss": 0.6004,
+      "step": 9338
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 43.78899643933011,
+      "learning_rate": 1.0930387905744278e-05,
+      "loss": 0.7305,
+      "step": 9339
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.08745968322831,
+      "learning_rate": 1.0928708808327107e-05,
+      "loss": 0.6317,
+      "step": 9340
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.28518022350391,
+      "learning_rate": 1.0927029684497995e-05,
+      "loss": 0.5948,
+      "step": 9341
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.10912681082738,
+      "learning_rate": 1.0925350534304706e-05,
+      "loss": 0.6098,
+      "step": 9342
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.705661489002246,
+      "learning_rate": 1.0923671357794986e-05,
+      "loss": 0.5668,
+      "step": 9343
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 32.680531913322575,
+      "learning_rate": 1.0921992155016594e-05,
+      "loss": 0.5767,
+      "step": 9344
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 41.14319404667998,
+      "learning_rate": 1.0920312926017285e-05,
+      "loss": 0.6165,
+      "step": 9345
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.3854129281427,
+      "learning_rate": 1.0918633670844813e-05,
+      "loss": 0.65,
+      "step": 9346
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 47.339771921956526,
+      "learning_rate": 1.0916954389546939e-05,
+      "loss": 0.6327,
+      "step": 9347
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.67516988401317,
+      "learning_rate": 1.0915275082171419e-05,
+      "loss": 0.7009,
+      "step": 9348
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 32.10943783941872,
+      "learning_rate": 1.0913595748766009e-05,
+      "loss": 0.633,
+      "step": 9349
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.159924241538526,
+      "learning_rate": 1.0911916389378473e-05,
+      "loss": 0.6371,
+      "step": 9350
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.15982651903262,
+      "learning_rate": 1.0910237004056563e-05,
+      "loss": 0.6347,
+      "step": 9351
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 39.02255623770318,
+      "learning_rate": 1.0908557592848048e-05,
+      "loss": 0.6794,
+      "step": 9352
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.86784985936136,
+      "learning_rate": 1.0906878155800686e-05,
+      "loss": 0.6404,
+      "step": 9353
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 30.594395277097604,
+      "learning_rate": 1.0905198692962241e-05,
+      "loss": 0.5296,
+      "step": 9354
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.24897588142213,
+      "learning_rate": 1.0903519204380475e-05,
+      "loss": 0.5731,
+      "step": 9355
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.323267292715,
+      "learning_rate": 1.090183969010315e-05,
+      "loss": 0.6118,
+      "step": 9356
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 44.890866885411555,
+      "learning_rate": 1.0900160150178034e-05,
+      "loss": 0.6955,
+      "step": 9357
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.41268436183262,
+      "learning_rate": 1.0898480584652887e-05,
+      "loss": 0.5995,
+      "step": 9358
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.23462752590182,
+      "learning_rate": 1.0896800993575482e-05,
+      "loss": 0.5592,
+      "step": 9359
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.01371627387158,
+      "learning_rate": 1.0895121376993576e-05,
+      "loss": 0.702,
+      "step": 9360
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.161615548826475,
+      "learning_rate": 1.0893441734954944e-05,
+      "loss": 0.6254,
+      "step": 9361
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 31.21068165546539,
+      "learning_rate": 1.0891762067507353e-05,
+      "loss": 0.6242,
+      "step": 9362
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 45.649732580164184,
+      "learning_rate": 1.089008237469857e-05,
+      "loss": 0.7406,
+      "step": 9363
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 39.71469292378013,
+      "learning_rate": 1.0888402656576364e-05,
+      "loss": 0.6473,
+      "step": 9364
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.68863840230825,
+      "learning_rate": 1.0886722913188508e-05,
+      "loss": 0.5406,
+      "step": 9365
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.80370560601305,
+      "learning_rate": 1.0885043144582765e-05,
+      "loss": 0.7038,
+      "step": 9366
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.71076121516271,
+      "learning_rate": 1.0883363350806922e-05,
+      "loss": 0.642,
+      "step": 9367
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.85240819986492,
+      "learning_rate": 1.0881683531908733e-05,
+      "loss": 0.5975,
+      "step": 9368
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.66776141471889,
+      "learning_rate": 1.0880003687935988e-05,
+      "loss": 0.6239,
+      "step": 9369
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.46580130310373,
+      "learning_rate": 1.0878323818936448e-05,
+      "loss": 0.6716,
+      "step": 9370
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 45.95971940556687,
+      "learning_rate": 1.0876643924957892e-05,
+      "loss": 0.6618,
+      "step": 9371
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.48581351404597,
+      "learning_rate": 1.0874964006048098e-05,
+      "loss": 0.59,
+      "step": 9372
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 44.35888912933191,
+      "learning_rate": 1.087328406225484e-05,
+      "loss": 0.6229,
+      "step": 9373
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.816828642792984,
+      "learning_rate": 1.0871604093625894e-05,
+      "loss": 0.6054,
+      "step": 9374
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 33.251688032395954,
+      "learning_rate": 1.0869924100209032e-05,
+      "loss": 0.6289,
+      "step": 9375
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.39612859086864,
+      "learning_rate": 1.0868244082052047e-05,
+      "loss": 0.6326,
+      "step": 9376
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.80119394250764,
+      "learning_rate": 1.0866564039202701e-05,
+      "loss": 0.6723,
+      "step": 9377
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 36.97979121294185,
+      "learning_rate": 1.0864883971708782e-05,
+      "loss": 0.6071,
+      "step": 9378
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.567406418137566,
+      "learning_rate": 1.0863203879618069e-05,
+      "loss": 0.5761,
+      "step": 9379
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 41.11836753571089,
+      "learning_rate": 1.0861523762978343e-05,
+      "loss": 0.5872,
+      "step": 9380
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 43.14651946377684,
+      "learning_rate": 1.0859843621837386e-05,
+      "loss": 0.7184,
+      "step": 9381
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 45.69915588880644,
+      "learning_rate": 1.085816345624298e-05,
+      "loss": 0.6201,
+      "step": 9382
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 41.2723324173031,
+      "learning_rate": 1.0856483266242903e-05,
+      "loss": 0.6584,
+      "step": 9383
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 39.50629706226811,
+      "learning_rate": 1.085480305188495e-05,
+      "loss": 0.6119,
+      "step": 9384
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 32.43913382529767,
+      "learning_rate": 1.0853122813216891e-05,
+      "loss": 0.5902,
+      "step": 9385
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 42.66691364959419,
+      "learning_rate": 1.0851442550286524e-05,
+      "loss": 0.6161,
+      "step": 9386
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 32.48155813036145,
+      "learning_rate": 1.0849762263141625e-05,
+      "loss": 0.5626,
+      "step": 9387
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 36.204128608737754,
+      "learning_rate": 1.0848081951829985e-05,
+      "loss": 0.637,
+      "step": 9388
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.21756383934363,
+      "learning_rate": 1.0846401616399392e-05,
+      "loss": 0.5966,
+      "step": 9389
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 42.62846185939679,
+      "learning_rate": 1.0844721256897633e-05,
+      "loss": 0.7958,
+      "step": 9390
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.036968693428726,
+      "learning_rate": 1.0843040873372493e-05,
+      "loss": 0.6621,
+      "step": 9391
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.3234409753389,
+      "learning_rate": 1.0841360465871765e-05,
+      "loss": 0.6236,
+      "step": 9392
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 30.814940383017237,
+      "learning_rate": 1.0839680034443234e-05,
+      "loss": 0.5948,
+      "step": 9393
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.656156708631244,
+      "learning_rate": 1.0837999579134699e-05,
+      "loss": 0.5362,
+      "step": 9394
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.155522436454774,
+      "learning_rate": 1.0836319099993944e-05,
+      "loss": 0.6723,
+      "step": 9395
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 27.285718956007518,
+      "learning_rate": 1.0834638597068763e-05,
+      "loss": 0.5859,
+      "step": 9396
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 50.78572122931361,
+      "learning_rate": 1.0832958070406949e-05,
+      "loss": 0.6448,
+      "step": 9397
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 33.54755921503372,
+      "learning_rate": 1.0831277520056296e-05,
+      "loss": 0.6415,
+      "step": 9398
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 50.708342727194264,
+      "learning_rate": 1.0829596946064595e-05,
+      "loss": 0.6879,
+      "step": 9399
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 40.46032749739291,
+      "learning_rate": 1.0827916348479641e-05,
+      "loss": 0.6689,
+      "step": 9400
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 33.508031739869374,
+      "learning_rate": 1.0826235727349235e-05,
+      "loss": 0.5885,
+      "step": 9401
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.502522552567996,
+      "learning_rate": 1.0824555082721161e-05,
+      "loss": 0.5853,
+      "step": 9402
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 32.35510105585933,
+      "learning_rate": 1.0822874414643229e-05,
+      "loss": 0.5489,
+      "step": 9403
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 53.229461362536824,
+      "learning_rate": 1.0821193723163228e-05,
+      "loss": 0.6617,
+      "step": 9404
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 36.3989308576192,
+      "learning_rate": 1.0819513008328957e-05,
+      "loss": 0.65,
+      "step": 9405
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 40.427859352538235,
+      "learning_rate": 1.0817832270188217e-05,
+      "loss": 0.61,
+      "step": 9406
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 38.91430996668916,
+      "learning_rate": 1.0816151508788804e-05,
+      "loss": 0.6116,
+      "step": 9407
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 52.89108772920622,
+      "learning_rate": 1.081447072417852e-05,
+      "loss": 0.6606,
+      "step": 9408
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.485841706943866,
+      "learning_rate": 1.0812789916405167e-05,
+      "loss": 0.5876,
+      "step": 9409
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 39.25524043980721,
+      "learning_rate": 1.081110908551654e-05,
+      "loss": 0.5426,
+      "step": 9410
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 43.16824607218093,
+      "learning_rate": 1.0809428231560451e-05,
+      "loss": 0.7282,
+      "step": 9411
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.011988931312,
+      "learning_rate": 1.0807747354584692e-05,
+      "loss": 0.6249,
+      "step": 9412
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 62.14471550231433,
+      "learning_rate": 1.0806066454637074e-05,
+      "loss": 0.6625,
+      "step": 9413
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 36.62195313119256,
+      "learning_rate": 1.0804385531765394e-05,
+      "loss": 0.642,
+      "step": 9414
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 30.290396532342143,
+      "learning_rate": 1.0802704586017463e-05,
+      "loss": 0.5399,
+      "step": 9415
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 38.4591386255421,
+      "learning_rate": 1.0801023617441082e-05,
+      "loss": 0.7315,
+      "step": 9416
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.94852142985641,
+      "learning_rate": 1.0799342626084057e-05,
+      "loss": 0.662,
+      "step": 9417
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 37.81776423104943,
+      "learning_rate": 1.0797661611994196e-05,
+      "loss": 0.6615,
+      "step": 9418
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 29.367890134678117,
+      "learning_rate": 1.0795980575219305e-05,
+      "loss": 0.6201,
+      "step": 9419
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.30372564630945,
+      "learning_rate": 1.0794299515807191e-05,
+      "loss": 0.5827,
+      "step": 9420
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 43.40636385579347,
+      "learning_rate": 1.0792618433805666e-05,
+      "loss": 0.6584,
+      "step": 9421
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 47.74415911157024,
+      "learning_rate": 1.0790937329262537e-05,
+      "loss": 0.6394,
+      "step": 9422
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 36.046712381420235,
+      "learning_rate": 1.0789256202225611e-05,
+      "loss": 0.6486,
+      "step": 9423
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 32.090903310992594,
+      "learning_rate": 1.07875750527427e-05,
+      "loss": 0.5981,
+      "step": 9424
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 65.82661125132223,
+      "learning_rate": 1.0785893880861616e-05,
+      "loss": 0.7765,
+      "step": 9425
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 30.588534774705884,
+      "learning_rate": 1.0784212686630171e-05,
+      "loss": 0.5656,
+      "step": 9426
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 32.18953266206883,
+      "learning_rate": 1.0782531470096171e-05,
+      "loss": 0.6298,
+      "step": 9427
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 37.428122060736946,
+      "learning_rate": 1.078085023130744e-05,
+      "loss": 0.6321,
+      "step": 9428
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 45.20351699042037,
+      "learning_rate": 1.0779168970311782e-05,
+      "loss": 0.7079,
+      "step": 9429
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 43.55422954233034,
+      "learning_rate": 1.0777487687157018e-05,
+      "loss": 0.6673,
+      "step": 9430
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.3300516406559,
+      "learning_rate": 1.0775806381890951e-05,
+      "loss": 0.6152,
+      "step": 9431
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 39.695221625863894,
+      "learning_rate": 1.0774125054561411e-05,
+      "loss": 0.5822,
+      "step": 9432
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 30.63144770108761,
+      "learning_rate": 1.0772443705216206e-05,
+      "loss": 0.5721,
+      "step": 9433
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 39.368701222109884,
+      "learning_rate": 1.0770762333903154e-05,
+      "loss": 0.6727,
+      "step": 9434
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 29.931817624438427,
+      "learning_rate": 1.0769080940670072e-05,
+      "loss": 0.6064,
+      "step": 9435
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 35.720156367425304,
+      "learning_rate": 1.0767399525564777e-05,
+      "loss": 0.6127,
+      "step": 9436
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 37.15296665824975,
+      "learning_rate": 1.0765718088635086e-05,
+      "loss": 0.6879,
+      "step": 9437
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 33.272732044918186,
+      "learning_rate": 1.0764036629928827e-05,
+      "loss": 0.6145,
+      "step": 9438
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 38.9057411375051,
+      "learning_rate": 1.0762355149493808e-05,
+      "loss": 0.6663,
+      "step": 9439
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 36.70888951140393,
+      "learning_rate": 1.0760673647377855e-05,
+      "loss": 0.612,
+      "step": 9440
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 32.67008259771862,
+      "learning_rate": 1.0758992123628792e-05,
+      "loss": 0.6842,
+      "step": 9441
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 40.5299045336484,
+      "learning_rate": 1.0757310578294433e-05,
+      "loss": 0.6226,
+      "step": 9442
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.82658356460245,
+      "learning_rate": 1.0755629011422605e-05,
+      "loss": 0.6421,
+      "step": 9443
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.277084422894845,
+      "learning_rate": 1.0753947423061131e-05,
+      "loss": 0.673,
+      "step": 9444
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.67129568263308,
+      "learning_rate": 1.0752265813257833e-05,
+      "loss": 0.5977,
+      "step": 9445
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 37.89695083511175,
+      "learning_rate": 1.0750584182060535e-05,
+      "loss": 0.7764,
+      "step": 9446
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 34.16653316208102,
+      "learning_rate": 1.0748902529517065e-05,
+      "loss": 0.5476,
+      "step": 9447
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 33.59815652024889,
+      "learning_rate": 1.074722085567524e-05,
+      "loss": 0.6041,
+      "step": 9448
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 42.55090413819553,
+      "learning_rate": 1.0745539160582894e-05,
+      "loss": 0.5979,
+      "step": 9449
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 40.889221423865514,
+      "learning_rate": 1.074385744428785e-05,
+      "loss": 0.6741,
+      "step": 9450
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 46.00481947758243,
+      "learning_rate": 1.0742175706837936e-05,
+      "loss": 0.6945,
+      "step": 9451
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 45.733683454148334,
+      "learning_rate": 1.0740493948280976e-05,
+      "loss": 0.6209,
+      "step": 9452
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 34.197848195706456,
+      "learning_rate": 1.0738812168664804e-05,
+      "loss": 0.62,
+      "step": 9453
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 29.27495804953888,
+      "learning_rate": 1.0737130368037245e-05,
+      "loss": 0.5577,
+      "step": 9454
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 39.68183517877759,
+      "learning_rate": 1.0735448546446132e-05,
+      "loss": 0.672,
+      "step": 9455
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 39.67312765829409,
+      "learning_rate": 1.073376670393929e-05,
+      "loss": 0.6925,
+      "step": 9456
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 31.004082293669693,
+      "learning_rate": 1.0732084840564556e-05,
+      "loss": 0.5293,
+      "step": 9457
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 36.12266739240913,
+      "learning_rate": 1.0730402956369753e-05,
+      "loss": 0.6474,
+      "step": 9458
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 37.354486018721836,
+      "learning_rate": 1.0728721051402718e-05,
+      "loss": 0.6029,
+      "step": 9459
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 40.92761253752035,
+      "learning_rate": 1.0727039125711283e-05,
+      "loss": 0.6284,
+      "step": 9460
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 37.020895792418806,
+      "learning_rate": 1.0725357179343282e-05,
+      "loss": 0.6327,
+      "step": 9461
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.25928901875201,
+      "learning_rate": 1.0723675212346546e-05,
+      "loss": 0.6668,
+      "step": 9462
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 43.791823260132894,
+      "learning_rate": 1.0721993224768908e-05,
+      "loss": 0.6631,
+      "step": 9463
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 43.67156708255645,
+      "learning_rate": 1.0720311216658205e-05,
+      "loss": 0.6937,
+      "step": 9464
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 39.944565874773474,
+      "learning_rate": 1.0718629188062275e-05,
+      "loss": 0.7127,
+      "step": 9465
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 35.355973185008054,
+      "learning_rate": 1.0716947139028953e-05,
+      "loss": 0.5818,
+      "step": 9466
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 39.146316814972806,
+      "learning_rate": 1.071526506960607e-05,
+      "loss": 0.7322,
+      "step": 9467
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.54772470127842,
+      "learning_rate": 1.071358297984147e-05,
+      "loss": 0.5791,
+      "step": 9468
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 37.28875550392067,
+      "learning_rate": 1.0711900869782983e-05,
+      "loss": 0.6083,
+      "step": 9469
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 36.20102735953123,
+      "learning_rate": 1.0710218739478457e-05,
+      "loss": 0.6884,
+      "step": 9470
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 40.62485997689399,
+      "learning_rate": 1.0708536588975721e-05,
+      "loss": 0.7435,
+      "step": 9471
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 37.979540581186946,
+      "learning_rate": 1.0706854418322625e-05,
+      "loss": 0.6028,
+      "step": 9472
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 54.25049295948562,
+      "learning_rate": 1.0705172227566996e-05,
+      "loss": 0.6342,
+      "step": 9473
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.74172332725613,
+      "learning_rate": 1.0703490016756688e-05,
+      "loss": 0.6153,
+      "step": 9474
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 54.770436923088376,
+      "learning_rate": 1.070180778593953e-05,
+      "loss": 0.7298,
+      "step": 9475
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 43.150476159223494,
+      "learning_rate": 1.0700125535163371e-05,
+      "loss": 0.6813,
+      "step": 9476
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.347514102118126,
+      "learning_rate": 1.0698443264476051e-05,
+      "loss": 0.6504,
+      "step": 9477
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.94023593579345,
+      "learning_rate": 1.0696760973925413e-05,
+      "loss": 0.6411,
+      "step": 9478
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 35.50105772994755,
+      "learning_rate": 1.0695078663559302e-05,
+      "loss": 0.6105,
+      "step": 9479
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 36.54049991686412,
+      "learning_rate": 1.0693396333425559e-05,
+      "loss": 0.5645,
+      "step": 9480
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.94026737060617,
+      "learning_rate": 1.0691713983572028e-05,
+      "loss": 0.5847,
+      "step": 9481
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.97456429362117,
+      "learning_rate": 1.069003161404656e-05,
+      "loss": 0.6429,
+      "step": 9482
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 33.99554048070686,
+      "learning_rate": 1.0688349224896991e-05,
+      "loss": 0.6243,
+      "step": 9483
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 47.22307334360393,
+      "learning_rate": 1.0686666816171179e-05,
+      "loss": 0.589,
+      "step": 9484
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 36.61498886940139,
+      "learning_rate": 1.0684984387916961e-05,
+      "loss": 0.6366,
+      "step": 9485
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 35.65526122411423,
+      "learning_rate": 1.068330194018219e-05,
+      "loss": 0.6822,
+      "step": 9486
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 35.59293541382233,
+      "learning_rate": 1.068161947301471e-05,
+      "loss": 0.6167,
+      "step": 9487
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 34.22264530533092,
+      "learning_rate": 1.067993698646237e-05,
+      "loss": 0.6535,
+      "step": 9488
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 33.74263727268577,
+      "learning_rate": 1.0678254480573023e-05,
+      "loss": 0.6277,
+      "step": 9489
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 31.25555163089375,
+      "learning_rate": 1.067657195539451e-05,
+      "loss": 0.6262,
+      "step": 9490
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.756094700329236,
+      "learning_rate": 1.0674889410974689e-05,
+      "loss": 0.6921,
+      "step": 9491
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 42.309150180040426,
+      "learning_rate": 1.0673206847361407e-05,
+      "loss": 0.71,
+      "step": 9492
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 43.195028165094286,
+      "learning_rate": 1.067152426460252e-05,
+      "loss": 0.626,
+      "step": 9493
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.23324050436138,
+      "learning_rate": 1.0669841662745874e-05,
+      "loss": 0.6716,
+      "step": 9494
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 37.656175681393044,
+      "learning_rate": 1.0668159041839321e-05,
+      "loss": 0.6774,
+      "step": 9495
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 34.86208399664296,
+      "learning_rate": 1.066647640193072e-05,
+      "loss": 0.6748,
+      "step": 9496
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 40.73045905063624,
+      "learning_rate": 1.0664793743067916e-05,
+      "loss": 0.6944,
+      "step": 9497
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 31.839186812198992,
+      "learning_rate": 1.0663111065298766e-05,
+      "loss": 0.5504,
+      "step": 9498
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 47.137204297417135,
+      "learning_rate": 1.066142836867113e-05,
+      "loss": 0.746,
+      "step": 9499
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.49256070519169,
+      "learning_rate": 1.0659745653232851e-05,
+      "loss": 0.6597,
+      "step": 9500
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 42.42508730589965,
+      "learning_rate": 1.06580629190318e-05,
+      "loss": 0.6424,
+      "step": 9501
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 36.26485570496785,
+      "learning_rate": 1.0656380166115818e-05,
+      "loss": 0.5928,
+      "step": 9502
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 35.56768838749033,
+      "learning_rate": 1.065469739453277e-05,
+      "loss": 0.6716,
+      "step": 9503
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 42.18111857999835,
+      "learning_rate": 1.0653014604330511e-05,
+      "loss": 0.5986,
+      "step": 9504
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 46.08915903278089,
+      "learning_rate": 1.06513317955569e-05,
+      "loss": 0.717,
+      "step": 9505
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 33.09400209685675,
+      "learning_rate": 1.0649648968259793e-05,
+      "loss": 0.5827,
+      "step": 9506
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.196774781470225,
+      "learning_rate": 1.0647966122487048e-05,
+      "loss": 0.6366,
+      "step": 9507
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.02223042710922,
+      "learning_rate": 1.0646283258286524e-05,
+      "loss": 0.6783,
+      "step": 9508
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.5369177345122,
+      "learning_rate": 1.0644600375706087e-05,
+      "loss": 0.635,
+      "step": 9509
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 28.460990468011786,
+      "learning_rate": 1.0642917474793591e-05,
+      "loss": 0.5975,
+      "step": 9510
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 34.80026394083515,
+      "learning_rate": 1.0641234555596898e-05,
+      "loss": 0.5748,
+      "step": 9511
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.49967527943845,
+      "learning_rate": 1.0639551618163869e-05,
+      "loss": 0.6577,
+      "step": 9512
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 45.00158706621455,
+      "learning_rate": 1.0637868662542364e-05,
+      "loss": 0.7467,
+      "step": 9513
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.89133199027163,
+      "learning_rate": 1.0636185688780253e-05,
+      "loss": 0.676,
+      "step": 9514
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 41.049896811496154,
+      "learning_rate": 1.0634502696925387e-05,
+      "loss": 0.6203,
+      "step": 9515
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 54.388071387622645,
+      "learning_rate": 1.063281968702564e-05,
+      "loss": 0.6427,
+      "step": 9516
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 35.50660022307036,
+      "learning_rate": 1.0631136659128867e-05,
+      "loss": 0.5979,
+      "step": 9517
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 34.44455848469023,
+      "learning_rate": 1.062945361328294e-05,
+      "loss": 0.5792,
+      "step": 9518
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 34.16734768878831,
+      "learning_rate": 1.0627770549535716e-05,
+      "loss": 0.6136,
+      "step": 9519
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 27.58777977629141,
+      "learning_rate": 1.0626087467935069e-05,
+      "loss": 0.5086,
+      "step": 9520
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 34.71819139472422,
+      "learning_rate": 1.062440436852886e-05,
+      "loss": 0.6637,
+      "step": 9521
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 40.70888196535795,
+      "learning_rate": 1.0622721251364955e-05,
+      "loss": 0.6693,
+      "step": 9522
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.07443097518465,
+      "learning_rate": 1.0621038116491224e-05,
+      "loss": 0.6646,
+      "step": 9523
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.74244439702195,
+      "learning_rate": 1.061935496395553e-05,
+      "loss": 0.6196,
+      "step": 9524
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.707117268267226,
+      "learning_rate": 1.0617671793805743e-05,
+      "loss": 0.6176,
+      "step": 9525
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 42.59216683762756,
+      "learning_rate": 1.0615988606089733e-05,
+      "loss": 0.706,
+      "step": 9526
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.9722060736726,
+      "learning_rate": 1.0614305400855367e-05,
+      "loss": 0.6338,
+      "step": 9527
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.71026463286707,
+      "learning_rate": 1.0612622178150515e-05,
+      "loss": 0.6384,
+      "step": 9528
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 37.528658505047325,
+      "learning_rate": 1.0610938938023047e-05,
+      "loss": 0.6424,
+      "step": 9529
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 35.419274946919835,
+      "learning_rate": 1.060925568052083e-05,
+      "loss": 0.6407,
+      "step": 9530
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 37.14446604906811,
+      "learning_rate": 1.0607572405691741e-05,
+      "loss": 0.6932,
+      "step": 9531
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.147866278579485,
+      "learning_rate": 1.0605889113583647e-05,
+      "loss": 0.6502,
+      "step": 9532
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.72008692766391,
+      "learning_rate": 1.0604205804244424e-05,
+      "loss": 0.6054,
+      "step": 9533
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 37.875081949806315,
+      "learning_rate": 1.0602522477721938e-05,
+      "loss": 0.6692,
+      "step": 9534
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.231253308132686,
+      "learning_rate": 1.0600839134064065e-05,
+      "loss": 0.5659,
+      "step": 9535
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.183646464982196,
+      "learning_rate": 1.0599155773318679e-05,
+      "loss": 0.5967,
+      "step": 9536
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 31.772617649356107,
+      "learning_rate": 1.0597472395533654e-05,
+      "loss": 0.6512,
+      "step": 9537
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 44.458054152270265,
+      "learning_rate": 1.0595789000756864e-05,
+      "loss": 0.682,
+      "step": 9538
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 37.46411354833726,
+      "learning_rate": 1.0594105589036182e-05,
+      "loss": 0.5869,
+      "step": 9539
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.57397781827092,
+      "learning_rate": 1.0592422160419484e-05,
+      "loss": 0.6138,
+      "step": 9540
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 32.11877847979415,
+      "learning_rate": 1.0590738714954652e-05,
+      "loss": 0.5712,
+      "step": 9541
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.906436337220434,
+      "learning_rate": 1.0589055252689549e-05,
+      "loss": 0.7028,
+      "step": 9542
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 46.8351321293529,
+      "learning_rate": 1.0587371773672064e-05,
+      "loss": 0.601,
+      "step": 9543
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 32.58616229473417,
+      "learning_rate": 1.0585688277950065e-05,
+      "loss": 0.5756,
+      "step": 9544
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.41582984344247,
+      "learning_rate": 1.058400476557144e-05,
+      "loss": 0.5867,
+      "step": 9545
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 31.872322489909685,
+      "learning_rate": 1.0582321236584057e-05,
+      "loss": 0.5221,
+      "step": 9546
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 39.90007846580763,
+      "learning_rate": 1.0580637691035799e-05,
+      "loss": 0.594,
+      "step": 9547
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 42.674601196454766,
+      "learning_rate": 1.0578954128974546e-05,
+      "loss": 0.6429,
+      "step": 9548
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 40.52908421761607,
+      "learning_rate": 1.0577270550448175e-05,
+      "loss": 0.746,
+      "step": 9549
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.315705508371096,
+      "learning_rate": 1.0575586955504568e-05,
+      "loss": 0.6172,
+      "step": 9550
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 29.229160840846642,
+      "learning_rate": 1.0573903344191603e-05,
+      "loss": 0.5556,
+      "step": 9551
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 32.676958357272326,
+      "learning_rate": 1.0572219716557163e-05,
+      "loss": 0.5274,
+      "step": 9552
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 41.195753512236536,
+      "learning_rate": 1.0570536072649132e-05,
+      "loss": 0.6287,
+      "step": 9553
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.116614374939424,
+      "learning_rate": 1.0568852412515388e-05,
+      "loss": 0.6622,
+      "step": 9554
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 42.44859839642798,
+      "learning_rate": 1.0567168736203811e-05,
+      "loss": 0.6714,
+      "step": 9555
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 31.466919210263754,
+      "learning_rate": 1.0565485043762289e-05,
+      "loss": 0.5234,
+      "step": 9556
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 41.22176262154327,
+      "learning_rate": 1.0563801335238701e-05,
+      "loss": 0.6065,
+      "step": 9557
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.826154412717116,
+      "learning_rate": 1.0562117610680938e-05,
+      "loss": 0.6061,
+      "step": 9558
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 44.67359590634734,
+      "learning_rate": 1.0560433870136872e-05,
+      "loss": 0.5802,
+      "step": 9559
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.61802578586735,
+      "learning_rate": 1.0558750113654404e-05,
+      "loss": 0.6033,
+      "step": 9560
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 47.60682036378815,
+      "learning_rate": 1.0557066341281402e-05,
+      "loss": 0.5793,
+      "step": 9561
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 37.9131939716007,
+      "learning_rate": 1.055538255306576e-05,
+      "loss": 0.685,
+      "step": 9562
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.04248883663877,
+      "learning_rate": 1.0553698749055365e-05,
+      "loss": 0.6167,
+      "step": 9563
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.9170925354056,
+      "learning_rate": 1.05520149292981e-05,
+      "loss": 0.6742,
+      "step": 9564
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.28058003053423,
+      "learning_rate": 1.0550331093841855e-05,
+      "loss": 0.6399,
+      "step": 9565
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 31.52232663301989,
+      "learning_rate": 1.0548647242734516e-05,
+      "loss": 0.534,
+      "step": 9566
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 36.11510632670548,
+      "learning_rate": 1.054696337602397e-05,
+      "loss": 0.481,
+      "step": 9567
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.14513313377229,
+      "learning_rate": 1.0545279493758103e-05,
+      "loss": 0.6588,
+      "step": 9568
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 64.3880761948911,
+      "learning_rate": 1.0543595595984806e-05,
+      "loss": 0.6511,
+      "step": 9569
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 34.8049109660466,
+      "learning_rate": 1.0541911682751971e-05,
+      "loss": 0.5736,
+      "step": 9570
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.238531319887468,
+      "learning_rate": 1.0540227754107483e-05,
+      "loss": 0.5984,
+      "step": 9571
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 32.69625963231554,
+      "learning_rate": 1.0538543810099236e-05,
+      "loss": 0.5724,
+      "step": 9572
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 32.02370006996198,
+      "learning_rate": 1.0536859850775117e-05,
+      "loss": 0.6126,
+      "step": 9573
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 41.45960767822577,
+      "learning_rate": 1.0535175876183017e-05,
+      "loss": 0.6532,
+      "step": 9574
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 29.61719406075532,
+      "learning_rate": 1.0533491886370828e-05,
+      "loss": 0.5039,
+      "step": 9575
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.797106508582864,
+      "learning_rate": 1.0531807881386442e-05,
+      "loss": 0.5671,
+      "step": 9576
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 41.03929516609971,
+      "learning_rate": 1.0530123861277752e-05,
+      "loss": 0.6585,
+      "step": 9577
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 39.480297021444244,
+      "learning_rate": 1.0528439826092649e-05,
+      "loss": 0.6677,
+      "step": 9578
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 50.49918279836429,
+      "learning_rate": 1.0526755775879024e-05,
+      "loss": 0.6807,
+      "step": 9579
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 33.598699274317525,
+      "learning_rate": 1.0525071710684777e-05,
+      "loss": 0.5768,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 37.122596675775945,
+      "learning_rate": 1.0523387630557797e-05,
+      "loss": 0.6223,
+      "step": 9581
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 42.863426592743906,
+      "learning_rate": 1.0521703535545977e-05,
+      "loss": 0.5901,
+      "step": 9582
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 36.43132428658883,
+      "learning_rate": 1.0520019425697217e-05,
+      "loss": 0.6598,
+      "step": 9583
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.247404839007345,
+      "learning_rate": 1.0518335301059404e-05,
+      "loss": 0.6544,
+      "step": 9584
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 38.14155945967369,
+      "learning_rate": 1.0516651161680443e-05,
+      "loss": 0.5344,
+      "step": 9585
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 33.607216212816745,
+      "learning_rate": 1.0514967007608221e-05,
+      "loss": 0.6694,
+      "step": 9586
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 44.2186216696235,
+      "learning_rate": 1.0513282838890642e-05,
+      "loss": 0.6759,
+      "step": 9587
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 43.17415274079749,
+      "learning_rate": 1.0511598655575594e-05,
+      "loss": 0.6737,
+      "step": 9588
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 38.17843908146553,
+      "learning_rate": 1.0509914457710986e-05,
+      "loss": 0.6562,
+      "step": 9589
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 52.01447046201223,
+      "learning_rate": 1.0508230245344707e-05,
+      "loss": 0.6035,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 36.29590213898883,
+      "learning_rate": 1.0506546018524656e-05,
+      "loss": 0.5972,
+      "step": 9591
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 32.82018135260712,
+      "learning_rate": 1.0504861777298732e-05,
+      "loss": 0.5345,
+      "step": 9592
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 58.52100441910952,
+      "learning_rate": 1.0503177521714836e-05,
+      "loss": 0.6366,
+      "step": 9593
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 28.995069795844547,
+      "learning_rate": 1.0501493251820864e-05,
+      "loss": 0.6596,
+      "step": 9594
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 29.75951605157213,
+      "learning_rate": 1.0499808967664717e-05,
+      "loss": 0.6175,
+      "step": 9595
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.94484712259701,
+      "learning_rate": 1.0498124669294296e-05,
+      "loss": 0.5673,
+      "step": 9596
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 42.71026993183248,
+      "learning_rate": 1.0496440356757502e-05,
+      "loss": 0.6233,
+      "step": 9597
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.561150540019014,
+      "learning_rate": 1.0494756030102236e-05,
+      "loss": 0.6296,
+      "step": 9598
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 36.71565658525119,
+      "learning_rate": 1.0493071689376396e-05,
+      "loss": 0.7041,
+      "step": 9599
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.38117922944453,
+      "learning_rate": 1.0491387334627884e-05,
+      "loss": 0.6822,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 39.07238492482685,
+      "learning_rate": 1.0489702965904605e-05,
+      "loss": 0.606,
+      "step": 9601
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.20321851484917,
+      "learning_rate": 1.0488018583254463e-05,
+      "loss": 0.5424,
+      "step": 9602
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 42.745046270030635,
+      "learning_rate": 1.0486334186725353e-05,
+      "loss": 0.5468,
+      "step": 9603
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 43.60788716887967,
+      "learning_rate": 1.0484649776365189e-05,
+      "loss": 0.6437,
+      "step": 9604
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 36.170922232617755,
+      "learning_rate": 1.0482965352221863e-05,
+      "loss": 0.614,
+      "step": 9605
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 33.25522209690958,
+      "learning_rate": 1.0481280914343289e-05,
+      "loss": 0.6154,
+      "step": 9606
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 32.87587394930634,
+      "learning_rate": 1.0479596462777363e-05,
+      "loss": 0.5962,
+      "step": 9607
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.02991385526928,
+      "learning_rate": 1.0477911997571998e-05,
+      "loss": 0.625,
+      "step": 9608
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 39.08046214219918,
+      "learning_rate": 1.0476227518775095e-05,
+      "loss": 0.7259,
+      "step": 9609
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 40.95573073811396,
+      "learning_rate": 1.0474543026434559e-05,
+      "loss": 0.6686,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.977148640151494,
+      "learning_rate": 1.0472858520598295e-05,
+      "loss": 0.5427,
+      "step": 9611
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 29.67398967710242,
+      "learning_rate": 1.0471174001314215e-05,
+      "loss": 0.5878,
+      "step": 9612
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.2006506424885,
+      "learning_rate": 1.0469489468630218e-05,
+      "loss": 0.6593,
+      "step": 9613
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 37.64719931709059,
+      "learning_rate": 1.0467804922594218e-05,
+      "loss": 0.6288,
+      "step": 9614
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 41.099398978217394,
+      "learning_rate": 1.0466120363254119e-05,
+      "loss": 0.7216,
+      "step": 9615
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.207148888949124,
+      "learning_rate": 1.0464435790657833e-05,
+      "loss": 0.5329,
+      "step": 9616
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 37.74829408125817,
+      "learning_rate": 1.0462751204853262e-05,
+      "loss": 0.7134,
+      "step": 9617
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.892610451000344,
+      "learning_rate": 1.0461066605888314e-05,
+      "loss": 0.669,
+      "step": 9618
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 44.312025056884416,
+      "learning_rate": 1.0459381993810906e-05,
+      "loss": 0.5989,
+      "step": 9619
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 39.639160703589816,
+      "learning_rate": 1.045769736866894e-05,
+      "loss": 0.6419,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 37.10994857505437,
+      "learning_rate": 1.0456012730510332e-05,
+      "loss": 0.622,
+      "step": 9621
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.474866559810636,
+      "learning_rate": 1.0454328079382985e-05,
+      "loss": 0.5794,
+      "step": 9622
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 46.39860216822246,
+      "learning_rate": 1.0452643415334816e-05,
+      "loss": 0.5798,
+      "step": 9623
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 42.62475518746424,
+      "learning_rate": 1.045095873841373e-05,
+      "loss": 0.6056,
+      "step": 9624
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 36.98369314437871,
+      "learning_rate": 1.0449274048667644e-05,
+      "loss": 0.5693,
+      "step": 9625
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 67.05184698327413,
+      "learning_rate": 1.0447589346144467e-05,
+      "loss": 0.7234,
+      "step": 9626
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 28.065329589580298,
+      "learning_rate": 1.044590463089211e-05,
+      "loss": 0.5421,
+      "step": 9627
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 33.0671214531636,
+      "learning_rate": 1.0444219902958483e-05,
+      "loss": 0.5708,
+      "step": 9628
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.69812999779245,
+      "learning_rate": 1.0442535162391508e-05,
+      "loss": 0.6179,
+      "step": 9629
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.934100877966166,
+      "learning_rate": 1.0440850409239085e-05,
+      "loss": 0.5903,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 38.88158294162361,
+      "learning_rate": 1.043916564354914e-05,
+      "loss": 0.6717,
+      "step": 9631
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 32.97307970082172,
+      "learning_rate": 1.0437480865369577e-05,
+      "loss": 0.5993,
+      "step": 9632
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 82.06223893832237,
+      "learning_rate": 1.0435796074748316e-05,
+      "loss": 0.6647,
+      "step": 9633
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 36.69678337620431,
+      "learning_rate": 1.0434111271733271e-05,
+      "loss": 0.601,
+      "step": 9634
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 46.25897691572536,
+      "learning_rate": 1.0432426456372356e-05,
+      "loss": 0.6156,
+      "step": 9635
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 48.20784169371985,
+      "learning_rate": 1.0430741628713482e-05,
+      "loss": 0.6026,
+      "step": 9636
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 38.964247944418815,
+      "learning_rate": 1.042905678880457e-05,
+      "loss": 0.6967,
+      "step": 9637
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.145205155139536,
+      "learning_rate": 1.0427371936693531e-05,
+      "loss": 0.612,
+      "step": 9638
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 49.08155580010281,
+      "learning_rate": 1.0425687072428288e-05,
+      "loss": 0.6538,
+      "step": 9639
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 39.467814379284675,
+      "learning_rate": 1.042400219605675e-05,
+      "loss": 0.6771,
+      "step": 9640
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 41.55717294155471,
+      "learning_rate": 1.0422317307626842e-05,
+      "loss": 0.574,
+      "step": 9641
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 34.976706307912025,
+      "learning_rate": 1.0420632407186475e-05,
+      "loss": 0.6266,
+      "step": 9642
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 58.44615374458011,
+      "learning_rate": 1.0418947494783567e-05,
+      "loss": 0.6471,
+      "step": 9643
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.20456218856361,
+      "learning_rate": 1.041726257046604e-05,
+      "loss": 0.6549,
+      "step": 9644
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 36.95949485670998,
+      "learning_rate": 1.0415577634281806e-05,
+      "loss": 0.616,
+      "step": 9645
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 39.29020965048377,
+      "learning_rate": 1.0413892686278791e-05,
+      "loss": 0.636,
+      "step": 9646
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.05334456141132,
+      "learning_rate": 1.0412207726504906e-05,
+      "loss": 0.6056,
+      "step": 9647
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.15659882775076,
+      "learning_rate": 1.041052275500808e-05,
+      "loss": 0.5527,
+      "step": 9648
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.92093392724464,
+      "learning_rate": 1.0408837771836222e-05,
+      "loss": 0.5426,
+      "step": 9649
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.338907318871875,
+      "learning_rate": 1.0407152777037258e-05,
+      "loss": 0.5782,
+      "step": 9650
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.889360407573086,
+      "learning_rate": 1.0405467770659107e-05,
+      "loss": 0.6853,
+      "step": 9651
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 28.49298080164184,
+      "learning_rate": 1.0403782752749692e-05,
+      "loss": 0.586,
+      "step": 9652
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 31.083871140485023,
+      "learning_rate": 1.040209772335693e-05,
+      "loss": 0.5503,
+      "step": 9653
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 38.8399175501075,
+      "learning_rate": 1.0400412682528745e-05,
+      "loss": 0.5587,
+      "step": 9654
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 32.40155064380762,
+      "learning_rate": 1.0398727630313055e-05,
+      "loss": 0.5985,
+      "step": 9655
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.57098180246493,
+      "learning_rate": 1.0397042566757789e-05,
+      "loss": 0.5701,
+      "step": 9656
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 34.22740493604902,
+      "learning_rate": 1.039535749191086e-05,
+      "loss": 0.6034,
+      "step": 9657
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.11843267658264,
+      "learning_rate": 1.0393672405820197e-05,
+      "loss": 0.6203,
+      "step": 9658
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 39.279428353325976,
+      "learning_rate": 1.0391987308533722e-05,
+      "loss": 0.5973,
+      "step": 9659
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 29.148875850509032,
+      "learning_rate": 1.0390302200099355e-05,
+      "loss": 0.4925,
+      "step": 9660
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 38.92297459124535,
+      "learning_rate": 1.0388617080565024e-05,
+      "loss": 0.5872,
+      "step": 9661
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.921118191438815,
+      "learning_rate": 1.0386931949978649e-05,
+      "loss": 0.578,
+      "step": 9662
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 41.14044830944977,
+      "learning_rate": 1.038524680838816e-05,
+      "loss": 0.6377,
+      "step": 9663
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.52759467872242,
+      "learning_rate": 1.038356165584147e-05,
+      "loss": 0.6723,
+      "step": 9664
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.52770537218748,
+      "learning_rate": 1.0381876492386516e-05,
+      "loss": 0.5998,
+      "step": 9665
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 33.808681748406244,
+      "learning_rate": 1.0380191318071215e-05,
+      "loss": 0.5924,
+      "step": 9666
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 39.31358408077562,
+      "learning_rate": 1.0378506132943497e-05,
+      "loss": 0.6209,
+      "step": 9667
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 32.56023715424927,
+      "learning_rate": 1.0376820937051286e-05,
+      "loss": 0.5228,
+      "step": 9668
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.469081887192246,
+      "learning_rate": 1.0375135730442507e-05,
+      "loss": 0.5572,
+      "step": 9669
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 43.57634176966875,
+      "learning_rate": 1.0373450513165089e-05,
+      "loss": 0.6483,
+      "step": 9670
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.66212425158444,
+      "learning_rate": 1.0371765285266957e-05,
+      "loss": 0.5571,
+      "step": 9671
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 38.869508410604205,
+      "learning_rate": 1.0370080046796034e-05,
+      "loss": 0.6746,
+      "step": 9672
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.419182481842185,
+      "learning_rate": 1.0368394797800256e-05,
+      "loss": 0.5779,
+      "step": 9673
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 42.68279999513091,
+      "learning_rate": 1.0366709538327542e-05,
+      "loss": 0.6234,
+      "step": 9674
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.33402664850894,
+      "learning_rate": 1.0365024268425826e-05,
+      "loss": 0.6089,
+      "step": 9675
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.33195460676544,
+      "learning_rate": 1.036333898814303e-05,
+      "loss": 0.6533,
+      "step": 9676
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.849210466064534,
+      "learning_rate": 1.0361653697527088e-05,
+      "loss": 0.6517,
+      "step": 9677
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 31.771608459939493,
+      "learning_rate": 1.0359968396625925e-05,
+      "loss": 0.5321,
+      "step": 9678
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 39.91079179732494,
+      "learning_rate": 1.0358283085487473e-05,
+      "loss": 0.6343,
+      "step": 9679
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.261256648488605,
+      "learning_rate": 1.0356597764159659e-05,
+      "loss": 0.6247,
+      "step": 9680
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 36.16896146900669,
+      "learning_rate": 1.0354912432690412e-05,
+      "loss": 0.5996,
+      "step": 9681
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.11959548179566,
+      "learning_rate": 1.0353227091127662e-05,
+      "loss": 0.6277,
+      "step": 9682
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 36.454566835486034,
+      "learning_rate": 1.0351541739519341e-05,
+      "loss": 0.6495,
+      "step": 9683
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 41.58870852760788,
+      "learning_rate": 1.034985637791338e-05,
+      "loss": 0.6222,
+      "step": 9684
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.865061287082376,
+      "learning_rate": 1.034817100635771e-05,
+      "loss": 0.6803,
+      "step": 9685
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.23737164122501,
+      "learning_rate": 1.0346485624900258e-05,
+      "loss": 0.6067,
+      "step": 9686
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 34.706807617428744,
+      "learning_rate": 1.0344800233588959e-05,
+      "loss": 0.5937,
+      "step": 9687
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 34.439114984489905,
+      "learning_rate": 1.0343114832471742e-05,
+      "loss": 0.6367,
+      "step": 9688
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.42203164558744,
+      "learning_rate": 1.0341429421596538e-05,
+      "loss": 0.6376,
+      "step": 9689
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.78056717808873,
+      "learning_rate": 1.0339744001011285e-05,
+      "loss": 0.7476,
+      "step": 9690
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 36.85003259753127,
+      "learning_rate": 1.0338058570763907e-05,
+      "loss": 0.6279,
+      "step": 9691
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 41.03485383290441,
+      "learning_rate": 1.0336373130902346e-05,
+      "loss": 0.5379,
+      "step": 9692
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 34.777440478556606,
+      "learning_rate": 1.0334687681474524e-05,
+      "loss": 0.5698,
+      "step": 9693
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.49508458009888,
+      "learning_rate": 1.0333002222528384e-05,
+      "loss": 0.5816,
+      "step": 9694
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 78.56067710588552,
+      "learning_rate": 1.0331316754111855e-05,
+      "loss": 0.5724,
+      "step": 9695
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 29.81349671894919,
+      "learning_rate": 1.032963127627287e-05,
+      "loss": 0.5837,
+      "step": 9696
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 55.52252239451682,
+      "learning_rate": 1.0327945789059366e-05,
+      "loss": 0.7009,
+      "step": 9697
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 49.2517884263766,
+      "learning_rate": 1.0326260292519274e-05,
+      "loss": 0.6649,
+      "step": 9698
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 41.34531151092645,
+      "learning_rate": 1.032457478670053e-05,
+      "loss": 0.6134,
+      "step": 9699
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 38.648554425267996,
+      "learning_rate": 1.0322889271651073e-05,
+      "loss": 0.6695,
+      "step": 9700
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 38.93235997154682,
+      "learning_rate": 1.0321203747418829e-05,
+      "loss": 0.6283,
+      "step": 9701
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.48300115859468,
+      "learning_rate": 1.031951821405174e-05,
+      "loss": 0.6289,
+      "step": 9702
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 38.0165289956766,
+      "learning_rate": 1.031783267159774e-05,
+      "loss": 0.6271,
+      "step": 9703
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 33.69618570322686,
+      "learning_rate": 1.0316147120104767e-05,
+      "loss": 0.5753,
+      "step": 9704
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.46782624646654,
+      "learning_rate": 1.0314461559620752e-05,
+      "loss": 0.577,
+      "step": 9705
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 62.912905508212866,
+      "learning_rate": 1.0312775990193635e-05,
+      "loss": 0.6964,
+      "step": 9706
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.43448791554632,
+      "learning_rate": 1.0311090411871354e-05,
+      "loss": 0.6001,
+      "step": 9707
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.77011409052978,
+      "learning_rate": 1.030940482470184e-05,
+      "loss": 0.5266,
+      "step": 9708
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 35.47633681638196,
+      "learning_rate": 1.0307719228733034e-05,
+      "loss": 0.6177,
+      "step": 9709
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 26.7760889851458,
+      "learning_rate": 1.0306033624012875e-05,
+      "loss": 0.5233,
+      "step": 9710
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.20346198636294,
+      "learning_rate": 1.0304348010589298e-05,
+      "loss": 0.6232,
+      "step": 9711
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 60.51361153914816,
+      "learning_rate": 1.0302662388510243e-05,
+      "loss": 0.5765,
+      "step": 9712
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 30.773255331444958,
+      "learning_rate": 1.0300976757823643e-05,
+      "loss": 0.536,
+      "step": 9713
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 39.86792943564227,
+      "learning_rate": 1.0299291118577445e-05,
+      "loss": 0.6571,
+      "step": 9714
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 39.31056684098433,
+      "learning_rate": 1.029760547081958e-05,
+      "loss": 0.6159,
+      "step": 9715
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 39.50879429447241,
+      "learning_rate": 1.0295919814597988e-05,
+      "loss": 0.6193,
+      "step": 9716
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 41.496174805504644,
+      "learning_rate": 1.0294234149960614e-05,
+      "loss": 0.6847,
+      "step": 9717
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 61.7067895353166,
+      "learning_rate": 1.0292548476955389e-05,
+      "loss": 0.7032,
+      "step": 9718
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 44.51308988540779,
+      "learning_rate": 1.0290862795630261e-05,
+      "loss": 0.637,
+      "step": 9719
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 43.74620369925393,
+      "learning_rate": 1.028917710603316e-05,
+      "loss": 0.6933,
+      "step": 9720
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 32.93449118478148,
+      "learning_rate": 1.0287491408212031e-05,
+      "loss": 0.5427,
+      "step": 9721
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.11584240903595,
+      "learning_rate": 1.0285805702214817e-05,
+      "loss": 0.6377,
+      "step": 9722
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 32.7178626026678,
+      "learning_rate": 1.0284119988089458e-05,
+      "loss": 0.553,
+      "step": 9723
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.886186220860544,
+      "learning_rate": 1.028243426588389e-05,
+      "loss": 0.7013,
+      "step": 9724
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 29.119687768574206,
+      "learning_rate": 1.0280748535646058e-05,
+      "loss": 0.5641,
+      "step": 9725
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 45.45715971893569,
+      "learning_rate": 1.0279062797423899e-05,
+      "loss": 0.7439,
+      "step": 9726
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.515708161197736,
+      "learning_rate": 1.0277377051265361e-05,
+      "loss": 0.6495,
+      "step": 9727
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 38.12704756194361,
+      "learning_rate": 1.027569129721838e-05,
+      "loss": 0.5792,
+      "step": 9728
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 31.863133762996156,
+      "learning_rate": 1.02740055353309e-05,
+      "loss": 0.6503,
+      "step": 9729
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.13465503024923,
+      "learning_rate": 1.0272319765650862e-05,
+      "loss": 0.6489,
+      "step": 9730
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.49619162278975,
+      "learning_rate": 1.0270633988226209e-05,
+      "loss": 0.6911,
+      "step": 9731
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 59.22173601739794,
+      "learning_rate": 1.0268948203104888e-05,
+      "loss": 0.5974,
+      "step": 9732
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 31.506706553221502,
+      "learning_rate": 1.026726241033483e-05,
+      "loss": 0.5518,
+      "step": 9733
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 30.98058847287115,
+      "learning_rate": 1.0265576609963994e-05,
+      "loss": 0.6138,
+      "step": 9734
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 35.542688687482325,
+      "learning_rate": 1.0263890802040308e-05,
+      "loss": 0.5526,
+      "step": 9735
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.886543240168166,
+      "learning_rate": 1.0262204986611726e-05,
+      "loss": 0.7098,
+      "step": 9736
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 48.88815172619841,
+      "learning_rate": 1.0260519163726183e-05,
+      "loss": 0.7067,
+      "step": 9737
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 43.739563634507185,
+      "learning_rate": 1.025883333343163e-05,
+      "loss": 0.5859,
+      "step": 9738
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 30.172446039565475,
+      "learning_rate": 1.025714749577601e-05,
+      "loss": 0.5595,
+      "step": 9739
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 32.56455687691284,
+      "learning_rate": 1.0255461650807264e-05,
+      "loss": 0.4961,
+      "step": 9740
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 32.862909606696675,
+      "learning_rate": 1.025377579857334e-05,
+      "loss": 0.5449,
+      "step": 9741
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.99667327165931,
+      "learning_rate": 1.0252089939122179e-05,
+      "loss": 0.5893,
+      "step": 9742
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.33186041622605,
+      "learning_rate": 1.0250404072501724e-05,
+      "loss": 0.6321,
+      "step": 9743
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 33.004832553314515,
+      "learning_rate": 1.024871819875993e-05,
+      "loss": 0.5758,
+      "step": 9744
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 39.895655472095854,
+      "learning_rate": 1.0247032317944731e-05,
+      "loss": 0.579,
+      "step": 9745
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 47.562590023762624,
+      "learning_rate": 1.0245346430104082e-05,
+      "loss": 0.686,
+      "step": 9746
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.92952657364906,
+      "learning_rate": 1.024366053528592e-05,
+      "loss": 0.6151,
+      "step": 9747
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 29.69562598396434,
+      "learning_rate": 1.0241974633538198e-05,
+      "loss": 0.5439,
+      "step": 9748
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 31.08931788768936,
+      "learning_rate": 1.0240288724908858e-05,
+      "loss": 0.6039,
+      "step": 9749
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 43.015462678990325,
+      "learning_rate": 1.0238602809445846e-05,
+      "loss": 0.7279,
+      "step": 9750
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 50.632299579835276,
+      "learning_rate": 1.023691688719711e-05,
+      "loss": 0.5353,
+      "step": 9751
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 32.43443963419017,
+      "learning_rate": 1.0235230958210596e-05,
+      "loss": 0.601,
+      "step": 9752
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.07346744883274,
+      "learning_rate": 1.0233545022534252e-05,
+      "loss": 0.4997,
+      "step": 9753
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 42.05091547610152,
+      "learning_rate": 1.0231859080216025e-05,
+      "loss": 0.6367,
+      "step": 9754
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 38.28958903824148,
+      "learning_rate": 1.023017313130386e-05,
+      "loss": 0.6562,
+      "step": 9755
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.824860168395226,
+      "learning_rate": 1.0228487175845707e-05,
+      "loss": 0.6745,
+      "step": 9756
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 36.85810977952231,
+      "learning_rate": 1.0226801213889512e-05,
+      "loss": 0.563,
+      "step": 9757
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.53776971709629,
+      "learning_rate": 1.0225115245483219e-05,
+      "loss": 0.6026,
+      "step": 9758
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 37.098293344361956,
+      "learning_rate": 1.0223429270674788e-05,
+      "loss": 0.6132,
+      "step": 9759
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 35.51224574582456,
+      "learning_rate": 1.0221743289512153e-05,
+      "loss": 0.6116,
+      "step": 9760
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 35.84278311252632,
+      "learning_rate": 1.0220057302043273e-05,
+      "loss": 0.6529,
+      "step": 9761
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.69862717801654,
+      "learning_rate": 1.0218371308316089e-05,
+      "loss": 0.7155,
+      "step": 9762
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 31.98009176417647,
+      "learning_rate": 1.0216685308378556e-05,
+      "loss": 0.5735,
+      "step": 9763
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 30.37155662453746,
+      "learning_rate": 1.0214999302278614e-05,
+      "loss": 0.6146,
+      "step": 9764
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 28.518113942674514,
+      "learning_rate": 1.0213313290064222e-05,
+      "loss": 0.5069,
+      "step": 9765
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 47.006500418561146,
+      "learning_rate": 1.0211627271783323e-05,
+      "loss": 0.6526,
+      "step": 9766
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 45.956480057471396,
+      "learning_rate": 1.0209941247483868e-05,
+      "loss": 0.6911,
+      "step": 9767
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 40.89733625530662,
+      "learning_rate": 1.0208255217213809e-05,
+      "loss": 0.6129,
+      "step": 9768
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.42454307458144,
+      "learning_rate": 1.0206569181021092e-05,
+      "loss": 0.611,
+      "step": 9769
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.910416696237192,
+      "learning_rate": 1.0204883138953666e-05,
+      "loss": 0.628,
+      "step": 9770
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 39.63348802877757,
+      "learning_rate": 1.0203197091059485e-05,
+      "loss": 0.6753,
+      "step": 9771
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.45199578221102,
+      "learning_rate": 1.02015110373865e-05,
+      "loss": 0.6468,
+      "step": 9772
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.19459669710106,
+      "learning_rate": 1.0199824977982658e-05,
+      "loss": 0.5714,
+      "step": 9773
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.92705706931522,
+      "learning_rate": 1.0198138912895907e-05,
+      "loss": 0.578,
+      "step": 9774
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 30.949385746913734,
+      "learning_rate": 1.0196452842174202e-05,
+      "loss": 0.5735,
+      "step": 9775
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.87246049768038,
+      "learning_rate": 1.0194766765865498e-05,
+      "loss": 0.5775,
+      "step": 9776
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.22840227428828,
+      "learning_rate": 1.0193080684017737e-05,
+      "loss": 0.6099,
+      "step": 9777
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.85377788851422,
+      "learning_rate": 1.0191394596678879e-05,
+      "loss": 0.5689,
+      "step": 9778
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 39.291842580922655,
+      "learning_rate": 1.0189708503896865e-05,
+      "loss": 0.6917,
+      "step": 9779
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.189261277792596,
+      "learning_rate": 1.0188022405719652e-05,
+      "loss": 0.5403,
+      "step": 9780
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.209414574584745,
+      "learning_rate": 1.0186336302195197e-05,
+      "loss": 0.6075,
+      "step": 9781
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 39.223569779005274,
+      "learning_rate": 1.0184650193371444e-05,
+      "loss": 0.5926,
+      "step": 9782
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 35.02610728128898,
+      "learning_rate": 1.0182964079296347e-05,
+      "loss": 0.5899,
+      "step": 9783
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 43.91032667409475,
+      "learning_rate": 1.0181277960017856e-05,
+      "loss": 0.547,
+      "step": 9784
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.94377312830383,
+      "learning_rate": 1.017959183558393e-05,
+      "loss": 0.6435,
+      "step": 9785
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.69018550296897,
+      "learning_rate": 1.0177905706042517e-05,
+      "loss": 0.6603,
+      "step": 9786
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 62.04282721191279,
+      "learning_rate": 1.0176219571441565e-05,
+      "loss": 0.7166,
+      "step": 9787
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.38792075251672,
+      "learning_rate": 1.0174533431829039e-05,
+      "loss": 0.6147,
+      "step": 9788
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.60355058871606,
+      "learning_rate": 1.0172847287252878e-05,
+      "loss": 0.5795,
+      "step": 9789
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.31758471947289,
+      "learning_rate": 1.0171161137761042e-05,
+      "loss": 0.5672,
+      "step": 9790
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.569593998129825,
+      "learning_rate": 1.0169474983401488e-05,
+      "loss": 0.6127,
+      "step": 9791
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 35.16459632642923,
+      "learning_rate": 1.016778882422216e-05,
+      "loss": 0.5913,
+      "step": 9792
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 35.90770736074063,
+      "learning_rate": 1.0166102660271018e-05,
+      "loss": 0.6866,
+      "step": 9793
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 33.958880541006394,
+      "learning_rate": 1.0164416491596014e-05,
+      "loss": 0.5822,
+      "step": 9794
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 42.586891134246684,
+      "learning_rate": 1.01627303182451e-05,
+      "loss": 0.6421,
+      "step": 9795
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 44.37464268500233,
+      "learning_rate": 1.016104414026623e-05,
+      "loss": 0.6851,
+      "step": 9796
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.375236293052815,
+      "learning_rate": 1.015935795770736e-05,
+      "loss": 0.6308,
+      "step": 9797
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 39.226343919744856,
+      "learning_rate": 1.0157671770616444e-05,
+      "loss": 0.5767,
+      "step": 9798
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.17112588791502,
+      "learning_rate": 1.0155985579041434e-05,
+      "loss": 0.6091,
+      "step": 9799
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.78799348741383,
+      "learning_rate": 1.0154299383030287e-05,
+      "loss": 0.511,
+      "step": 9800
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 42.486969122581996,
+      "learning_rate": 1.0152613182630953e-05,
+      "loss": 0.5918,
+      "step": 9801
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 41.70607883458834,
+      "learning_rate": 1.0150926977891388e-05,
+      "loss": 0.5766,
+      "step": 9802
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 34.53020965942269,
+      "learning_rate": 1.0149240768859554e-05,
+      "loss": 0.5345,
+      "step": 9803
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 35.89185065774469,
+      "learning_rate": 1.0147554555583394e-05,
+      "loss": 0.608,
+      "step": 9804
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 41.140045051818944,
+      "learning_rate": 1.0145868338110873e-05,
+      "loss": 0.6343,
+      "step": 9805
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 34.356902758244004,
+      "learning_rate": 1.0144182116489938e-05,
+      "loss": 0.6958,
+      "step": 9806
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.649047029702096,
+      "learning_rate": 1.0142495890768551e-05,
+      "loss": 0.6076,
+      "step": 9807
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 30.408498951386076,
+      "learning_rate": 1.0140809660994663e-05,
+      "loss": 0.5226,
+      "step": 9808
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 48.02986512827794,
+      "learning_rate": 1.0139123427216231e-05,
+      "loss": 0.6866,
+      "step": 9809
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 40.10590722340465,
+      "learning_rate": 1.013743718948121e-05,
+      "loss": 0.6672,
+      "step": 9810
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 39.630690830859876,
+      "learning_rate": 1.0135750947837558e-05,
+      "loss": 0.6303,
+      "step": 9811
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 33.0228827425167,
+      "learning_rate": 1.0134064702333225e-05,
+      "loss": 0.5836,
+      "step": 9812
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.75743073328091,
+      "learning_rate": 1.0132378453016171e-05,
+      "loss": 0.5822,
+      "step": 9813
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.95006353249842,
+      "learning_rate": 1.013069219993435e-05,
+      "loss": 0.5636,
+      "step": 9814
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 39.1055876406458,
+      "learning_rate": 1.0129005943135721e-05,
+      "loss": 0.6558,
+      "step": 9815
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.878148697559233,
+      "learning_rate": 1.012731968266824e-05,
+      "loss": 0.5852,
+      "step": 9816
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 33.838539166201116,
+      "learning_rate": 1.012563341857986e-05,
+      "loss": 0.5887,
+      "step": 9817
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 44.027557862605356,
+      "learning_rate": 1.0123947150918539e-05,
+      "loss": 0.667,
+      "step": 9818
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.70668052188332,
+      "learning_rate": 1.0122260879732231e-05,
+      "loss": 0.5817,
+      "step": 9819
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 45.463576719273675,
+      "learning_rate": 1.01205746050689e-05,
+      "loss": 0.5909,
+      "step": 9820
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.3352907544861,
+      "learning_rate": 1.0118888326976494e-05,
+      "loss": 0.6222,
+      "step": 9821
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 28.70265408764551,
+      "learning_rate": 1.0117202045502978e-05,
+      "loss": 0.509,
+      "step": 9822
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.535888046222404,
+      "learning_rate": 1.01155157606963e-05,
+      "loss": 0.5976,
+      "step": 9823
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.28300963067865,
+      "learning_rate": 1.0113829472604422e-05,
+      "loss": 0.648,
+      "step": 9824
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 31.3341869016476,
+      "learning_rate": 1.0112143181275302e-05,
+      "loss": 0.6455,
+      "step": 9825
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 48.41758865631964,
+      "learning_rate": 1.0110456886756894e-05,
+      "loss": 0.6932,
+      "step": 9826
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 35.02106478174187,
+      "learning_rate": 1.010877058909716e-05,
+      "loss": 0.643,
+      "step": 9827
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.22447671351198,
+      "learning_rate": 1.0107084288344052e-05,
+      "loss": 0.5682,
+      "step": 9828
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 98.69633008797244,
+      "learning_rate": 1.0105397984545524e-05,
+      "loss": 0.7167,
+      "step": 9829
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 31.075182741415027,
+      "learning_rate": 1.0103711677749548e-05,
+      "loss": 0.5541,
+      "step": 9830
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 32.27249330290509,
+      "learning_rate": 1.0102025368004066e-05,
+      "loss": 0.6035,
+      "step": 9831
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.785585824466132,
+      "learning_rate": 1.0100339055357048e-05,
+      "loss": 0.5699,
+      "step": 9832
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 46.08947758416416,
+      "learning_rate": 1.0098652739856441e-05,
+      "loss": 0.6613,
+      "step": 9833
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.06231670930575,
+      "learning_rate": 1.0096966421550209e-05,
+      "loss": 0.5389,
+      "step": 9834
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 54.94641812081478,
+      "learning_rate": 1.0095280100486309e-05,
+      "loss": 0.7005,
+      "step": 9835
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 41.76827081191088,
+      "learning_rate": 1.00935937767127e-05,
+      "loss": 0.6353,
+      "step": 9836
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 35.71907270924859,
+      "learning_rate": 1.0091907450277338e-05,
+      "loss": 0.5867,
+      "step": 9837
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 33.55777747389572,
+      "learning_rate": 1.0090221121228178e-05,
+      "loss": 0.6208,
+      "step": 9838
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 29.581580248902828,
+      "learning_rate": 1.0088534789613188e-05,
+      "loss": 0.5448,
+      "step": 9839
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 44.803341332643896,
+      "learning_rate": 1.0086848455480318e-05,
+      "loss": 0.5819,
+      "step": 9840
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 34.18802867668924,
+      "learning_rate": 1.0085162118877527e-05,
+      "loss": 0.5954,
+      "step": 9841
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 41.13938647534587,
+      "learning_rate": 1.0083475779852778e-05,
+      "loss": 0.5955,
+      "step": 9842
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 42.270823434358846,
+      "learning_rate": 1.0081789438454026e-05,
+      "loss": 0.6525,
+      "step": 9843
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 35.326016862168196,
+      "learning_rate": 1.0080103094729229e-05,
+      "loss": 0.6202,
+      "step": 9844
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 37.728569983755065,
+      "learning_rate": 1.007841674872635e-05,
+      "loss": 0.6216,
+      "step": 9845
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 31.8750459708088,
+      "learning_rate": 1.007673040049334e-05,
+      "loss": 0.5643,
+      "step": 9846
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 40.30802143857703,
+      "learning_rate": 1.0075044050078166e-05,
+      "loss": 0.6442,
+      "step": 9847
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.09331785086773,
+      "learning_rate": 1.0073357697528779e-05,
+      "loss": 0.6169,
+      "step": 9848
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 43.86265410371476,
+      "learning_rate": 1.0071671342893148e-05,
+      "loss": 0.6253,
+      "step": 9849
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 31.711959965234797,
+      "learning_rate": 1.0069984986219219e-05,
+      "loss": 0.5664,
+      "step": 9850
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 35.093931386178255,
+      "learning_rate": 1.0068298627554962e-05,
+      "loss": 0.5584,
+      "step": 9851
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 31.09035376398725,
+      "learning_rate": 1.0066612266948333e-05,
+      "loss": 0.5918,
+      "step": 9852
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 33.500605844599065,
+      "learning_rate": 1.0064925904447288e-05,
+      "loss": 0.5547,
+      "step": 9853
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.14983886419741,
+      "learning_rate": 1.006323954009979e-05,
+      "loss": 0.6685,
+      "step": 9854
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 44.598562471461726,
+      "learning_rate": 1.0061553173953794e-05,
+      "loss": 0.6152,
+      "step": 9855
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 33.45386687823166,
+      "learning_rate": 1.0059866806057263e-05,
+      "loss": 0.5499,
+      "step": 9856
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 35.768370763357,
+      "learning_rate": 1.0058180436458156e-05,
+      "loss": 0.5552,
+      "step": 9857
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 49.68099767016428,
+      "learning_rate": 1.0056494065204428e-05,
+      "loss": 0.6908,
+      "step": 9858
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 36.68098239745821,
+      "learning_rate": 1.0054807692344045e-05,
+      "loss": 0.6339,
+      "step": 9859
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 39.907505544389544,
+      "learning_rate": 1.0053121317924963e-05,
+      "loss": 0.682,
+      "step": 9860
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.55415045490761,
+      "learning_rate": 1.0051434941995142e-05,
+      "loss": 0.5714,
+      "step": 9861
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 37.99528318858091,
+      "learning_rate": 1.004974856460254e-05,
+      "loss": 0.6321,
+      "step": 9862
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 39.80081883910832,
+      "learning_rate": 1.0048062185795117e-05,
+      "loss": 0.6509,
+      "step": 9863
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 40.24767274821983,
+      "learning_rate": 1.0046375805620838e-05,
+      "loss": 0.6261,
+      "step": 9864
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 36.636172859053154,
+      "learning_rate": 1.004468942412765e-05,
+      "loss": 0.5901,
+      "step": 9865
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 37.62417379530373,
+      "learning_rate": 1.0043003041363531e-05,
+      "loss": 0.6298,
+      "step": 9866
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 32.30527298937575,
+      "learning_rate": 1.0041316657376426e-05,
+      "loss": 0.4577,
+      "step": 9867
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 33.637753071252945,
+      "learning_rate": 1.0039630272214299e-05,
+      "loss": 0.5648,
+      "step": 9868
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 30.369508870957834,
+      "learning_rate": 1.003794388592511e-05,
+      "loss": 0.5135,
+      "step": 9869
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 36.67143301256577,
+      "learning_rate": 1.0036257498556821e-05,
+      "loss": 0.639,
+      "step": 9870
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 47.31297458728127,
+      "learning_rate": 1.0034571110157388e-05,
+      "loss": 0.495,
+      "step": 9871
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 33.34797149791601,
+      "learning_rate": 1.0032884720774773e-05,
+      "loss": 0.5589,
+      "step": 9872
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 39.1546291798284,
+      "learning_rate": 1.0031198330456936e-05,
+      "loss": 0.623,
+      "step": 9873
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 34.24887689085568,
+      "learning_rate": 1.002951193925184e-05,
+      "loss": 0.6529,
+      "step": 9874
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.49065390695173,
+      "learning_rate": 1.0027825547207435e-05,
+      "loss": 0.6255,
+      "step": 9875
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.2635366945545,
+      "learning_rate": 1.0026139154371694e-05,
+      "loss": 0.5668,
+      "step": 9876
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 29.687066336024234,
+      "learning_rate": 1.0024452760792566e-05,
+      "loss": 0.5763,
+      "step": 9877
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 48.76946263489882,
+      "learning_rate": 1.0022766366518018e-05,
+      "loss": 0.5615,
+      "step": 9878
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.90443700220035,
+      "learning_rate": 1.0021079971596009e-05,
+      "loss": 0.6114,
+      "step": 9879
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 34.977474757148535,
+      "learning_rate": 1.0019393576074497e-05,
+      "loss": 0.5376,
+      "step": 9880
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 32.730322659379645,
+      "learning_rate": 1.0017707180001443e-05,
+      "loss": 0.6067,
+      "step": 9881
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 31.789671566974476,
+      "learning_rate": 1.0016020783424805e-05,
+      "loss": 0.5516,
+      "step": 9882
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 47.103716189988795,
+      "learning_rate": 1.001433438639255e-05,
+      "loss": 0.6771,
+      "step": 9883
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 33.02212099077735,
+      "learning_rate": 1.0012647988952628e-05,
+      "loss": 0.5706,
+      "step": 9884
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 42.05731372120125,
+      "learning_rate": 1.0010961591153008e-05,
+      "loss": 0.7215,
+      "step": 9885
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 35.1617646499481,
+      "learning_rate": 1.0009275193041645e-05,
+      "loss": 0.6314,
+      "step": 9886
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 39.02289277423004,
+      "learning_rate": 1.0007588794666503e-05,
+      "loss": 0.5756,
+      "step": 9887
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 39.673000083856614,
+      "learning_rate": 1.000590239607554e-05,
+      "loss": 0.6439,
+      "step": 9888
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 40.27407494825831,
+      "learning_rate": 1.0004215997316715e-05,
+      "loss": 0.632,
+      "step": 9889
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.030764969277264,
+      "learning_rate": 1.0002529598437988e-05,
+      "loss": 0.6329,
+      "step": 9890
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 40.13284468691298,
+      "learning_rate": 1.0000843199487325e-05,
+      "loss": 0.6557,
+      "step": 9891
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.84100498328179,
+      "learning_rate": 9.99915680051268e-06,
+      "loss": 0.5852,
+      "step": 9892
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 36.30662000851418,
+      "learning_rate": 9.997470401562015e-06,
+      "loss": 0.5992,
+      "step": 9893
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.96568493814405,
+      "learning_rate": 9.995784002683288e-06,
+      "loss": 0.5441,
+      "step": 9894
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.073104167416744,
+      "learning_rate": 9.994097603924462e-06,
+      "loss": 0.5624,
+      "step": 9895
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 34.55929248405696,
+      "learning_rate": 9.992411205333498e-06,
+      "loss": 0.5687,
+      "step": 9896
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 35.913979935815654,
+      "learning_rate": 9.990724806958358e-06,
+      "loss": 0.6839,
+      "step": 9897
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 42.53444636636495,
+      "learning_rate": 9.989038408846996e-06,
+      "loss": 0.6793,
+      "step": 9898
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 32.62015275279754,
+      "learning_rate": 9.987352011047374e-06,
+      "loss": 0.5985,
+      "step": 9899
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 32.648483331104536,
+      "learning_rate": 9.985665613607454e-06,
+      "loss": 0.5996,
+      "step": 9900
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 40.23493573140628,
+      "learning_rate": 9.983979216575195e-06,
+      "loss": 0.6174,
+      "step": 9901
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 35.531937940609254,
+      "learning_rate": 9.982292819998562e-06,
+      "loss": 0.5825,
+      "step": 9902
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 35.65336727622073,
+      "learning_rate": 9.980606423925506e-06,
+      "loss": 0.574,
+      "step": 9903
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 35.326593979160045,
+      "learning_rate": 9.978920028403995e-06,
+      "loss": 0.629,
+      "step": 9904
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 38.473757581375125,
+      "learning_rate": 9.977233633481984e-06,
+      "loss": 0.6587,
+      "step": 9905
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.481494603671656,
+      "learning_rate": 9.975547239207435e-06,
+      "loss": 0.6456,
+      "step": 9906
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 63.80738996917955,
+      "learning_rate": 9.973860845628311e-06,
+      "loss": 0.6679,
+      "step": 9907
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.62458253799864,
+      "learning_rate": 9.972174452792568e-06,
+      "loss": 0.5377,
+      "step": 9908
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 35.97056888518612,
+      "learning_rate": 9.970488060748164e-06,
+      "loss": 0.5437,
+      "step": 9909
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 39.095863394935904,
+      "learning_rate": 9.968801669543066e-06,
+      "loss": 0.6337,
+      "step": 9910
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 30.76030897997281,
+      "learning_rate": 9.967115279225228e-06,
+      "loss": 0.5273,
+      "step": 9911
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 34.93634946085372,
+      "learning_rate": 9.965428889842617e-06,
+      "loss": 0.688,
+      "step": 9912
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.1486928778491,
+      "learning_rate": 9.963742501443184e-06,
+      "loss": 0.5733,
+      "step": 9913
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 42.22443277668011,
+      "learning_rate": 9.962056114074893e-06,
+      "loss": 0.6422,
+      "step": 9914
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 27.860728172691243,
+      "learning_rate": 9.960369727785703e-06,
+      "loss": 0.5329,
+      "step": 9915
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.69799667694797,
+      "learning_rate": 9.958683342623579e-06,
+      "loss": 0.5724,
+      "step": 9916
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 31.74415624898973,
+      "learning_rate": 9.956996958636474e-06,
+      "loss": 0.5594,
+      "step": 9917
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 32.299915106044736,
+      "learning_rate": 9.955310575872351e-06,
+      "loss": 0.6272,
+      "step": 9918
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 36.218262047891386,
+      "learning_rate": 9.953624194379165e-06,
+      "loss": 0.5855,
+      "step": 9919
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 41.32064748229041,
+      "learning_rate": 9.951937814204884e-06,
+      "loss": 0.729,
+      "step": 9920
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 28.444191126628834,
+      "learning_rate": 9.950251435397466e-06,
+      "loss": 0.5574,
+      "step": 9921
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 31.807378221719617,
+      "learning_rate": 9.948565058004863e-06,
+      "loss": 0.5343,
+      "step": 9922
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 69.81793977176639,
+      "learning_rate": 9.94687868207504e-06,
+      "loss": 0.6393,
+      "step": 9923
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.57979961921655,
+      "learning_rate": 9.945192307655959e-06,
+      "loss": 0.6269,
+      "step": 9924
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 41.73023616387085,
+      "learning_rate": 9.943505934795572e-06,
+      "loss": 0.7028,
+      "step": 9925
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 36.97066274549347,
+      "learning_rate": 9.941819563541849e-06,
+      "loss": 0.6338,
+      "step": 9926
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 36.70190307479956,
+      "learning_rate": 9.940133193942742e-06,
+      "loss": 0.5408,
+      "step": 9927
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 36.13115712265908,
+      "learning_rate": 9.938446826046209e-06,
+      "loss": 0.5866,
+      "step": 9928
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.87228921751401,
+      "learning_rate": 9.936760459900215e-06,
+      "loss": 0.7126,
+      "step": 9929
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.81356603522643,
+      "learning_rate": 9.935074095552714e-06,
+      "loss": 0.6344,
+      "step": 9930
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.97218490573327,
+      "learning_rate": 9.933387733051672e-06,
+      "loss": 0.5985,
+      "step": 9931
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 41.4480579211153,
+      "learning_rate": 9.93170137244504e-06,
+      "loss": 0.6396,
+      "step": 9932
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 43.564059635960064,
+      "learning_rate": 9.930015013780783e-06,
+      "loss": 0.601,
+      "step": 9933
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 36.0810350588448,
+      "learning_rate": 9.928328657106855e-06,
+      "loss": 0.6024,
+      "step": 9934
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 34.52859242772212,
+      "learning_rate": 9.92664230247122e-06,
+      "loss": 0.5869,
+      "step": 9935
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 32.45172653160459,
+      "learning_rate": 9.924955949921839e-06,
+      "loss": 0.556,
+      "step": 9936
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 31.852335220855615,
+      "learning_rate": 9.923269599506664e-06,
+      "loss": 0.5521,
+      "step": 9937
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 38.869298308290205,
+      "learning_rate": 9.921583251273654e-06,
+      "loss": 0.6069,
+      "step": 9938
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.88984978625949,
+      "learning_rate": 9.919896905270772e-06,
+      "loss": 0.5416,
+      "step": 9939
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 43.809857102374174,
+      "learning_rate": 9.918210561545974e-06,
+      "loss": 0.698,
+      "step": 9940
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.83850618252684,
+      "learning_rate": 9.916524220147224e-06,
+      "loss": 0.6771,
+      "step": 9941
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 67.96149259385075,
+      "learning_rate": 9.914837881122474e-06,
+      "loss": 0.6742,
+      "step": 9942
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 47.69755618155021,
+      "learning_rate": 9.913151544519685e-06,
+      "loss": 0.6862,
+      "step": 9943
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 39.18900841299742,
+      "learning_rate": 9.911465210386813e-06,
+      "loss": 0.5507,
+      "step": 9944
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 35.78323331775082,
+      "learning_rate": 9.909778878771822e-06,
+      "loss": 0.6628,
+      "step": 9945
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 40.43352509939524,
+      "learning_rate": 9.908092549722667e-06,
+      "loss": 0.6608,
+      "step": 9946
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 40.20061858125907,
+      "learning_rate": 9.906406223287304e-06,
+      "loss": 0.5928,
+      "step": 9947
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 40.658552692481216,
+      "learning_rate": 9.904719899513693e-06,
+      "loss": 0.6315,
+      "step": 9948
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 39.22871238957335,
+      "learning_rate": 9.903033578449793e-06,
+      "loss": 0.661,
+      "step": 9949
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 34.213515625981174,
+      "learning_rate": 9.90134726014356e-06,
+      "loss": 0.6467,
+      "step": 9950
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.6103170556377,
+      "learning_rate": 9.899660944642957e-06,
+      "loss": 0.5793,
+      "step": 9951
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.72023143642788,
+      "learning_rate": 9.897974631995937e-06,
+      "loss": 0.527,
+      "step": 9952
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 34.18513594618343,
+      "learning_rate": 9.896288322250455e-06,
+      "loss": 0.6453,
+      "step": 9953
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 33.58838456355166,
+      "learning_rate": 9.894602015454476e-06,
+      "loss": 0.6036,
+      "step": 9954
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 38.401683638340074,
+      "learning_rate": 9.892915711655953e-06,
+      "loss": 0.5634,
+      "step": 9955
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 43.26531442824871,
+      "learning_rate": 9.891229410902846e-06,
+      "loss": 0.5829,
+      "step": 9956
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.75617877558444,
+      "learning_rate": 9.88954311324311e-06,
+      "loss": 0.6529,
+      "step": 9957
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.988006673233244,
+      "learning_rate": 9.887856818724702e-06,
+      "loss": 0.593,
+      "step": 9958
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.766855680135386,
+      "learning_rate": 9.88617052739558e-06,
+      "loss": 0.5299,
+      "step": 9959
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 38.49773367800388,
+      "learning_rate": 9.8844842393037e-06,
+      "loss": 0.5955,
+      "step": 9960
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 52.681321375747935,
+      "learning_rate": 9.882797954497028e-06,
+      "loss": 0.639,
+      "step": 9961
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 35.86813914258601,
+      "learning_rate": 9.881111673023509e-06,
+      "loss": 0.5876,
+      "step": 9962
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 30.1581516911466,
+      "learning_rate": 9.879425394931103e-06,
+      "loss": 0.5765,
+      "step": 9963
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.41772185148203,
+      "learning_rate": 9.877739120267769e-06,
+      "loss": 0.5633,
+      "step": 9964
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 31.35496844325837,
+      "learning_rate": 9.876052849081467e-06,
+      "loss": 0.5768,
+      "step": 9965
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 26.508125492207743,
+      "learning_rate": 9.874366581420144e-06,
+      "loss": 0.4971,
+      "step": 9966
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 42.931392679548004,
+      "learning_rate": 9.872680317331764e-06,
+      "loss": 0.6429,
+      "step": 9967
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.785604511095045,
+      "learning_rate": 9.87099405686428e-06,
+      "loss": 0.5818,
+      "step": 9968
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.0015986168099,
+      "learning_rate": 9.869307800065651e-06,
+      "loss": 0.6238,
+      "step": 9969
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 30.848349260204866,
+      "learning_rate": 9.867621546983834e-06,
+      "loss": 0.5663,
+      "step": 9970
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 44.48202272070674,
+      "learning_rate": 9.86593529766678e-06,
+      "loss": 0.6255,
+      "step": 9971
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 31.642570756024043,
+      "learning_rate": 9.864249052162447e-06,
+      "loss": 0.5523,
+      "step": 9972
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 43.136556624888826,
+      "learning_rate": 9.862562810518792e-06,
+      "loss": 0.5197,
+      "step": 9973
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 47.10752843777302,
+      "learning_rate": 9.860876572783772e-06,
+      "loss": 0.7032,
+      "step": 9974
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 32.98208801060797,
+      "learning_rate": 9.859190339005342e-06,
+      "loss": 0.553,
+      "step": 9975
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.549997192940275,
+      "learning_rate": 9.857504109231452e-06,
+      "loss": 0.5588,
+      "step": 9976
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.166946943934065,
+      "learning_rate": 9.855817883510063e-06,
+      "loss": 0.6071,
+      "step": 9977
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.447180283262064,
+      "learning_rate": 9.854131661889129e-06,
+      "loss": 0.5673,
+      "step": 9978
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.60476230404444,
+      "learning_rate": 9.852445444416606e-06,
+      "loss": 0.6559,
+      "step": 9979
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.25219492248607,
+      "learning_rate": 9.850759231140451e-06,
+      "loss": 0.5764,
+      "step": 9980
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.66791641777073,
+      "learning_rate": 9.849073022108613e-06,
+      "loss": 0.6163,
+      "step": 9981
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.05638052190256,
+      "learning_rate": 9.84738681736905e-06,
+      "loss": 0.6037,
+      "step": 9982
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.982617554301825,
+      "learning_rate": 9.845700616969718e-06,
+      "loss": 0.5753,
+      "step": 9983
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 51.98736130277894,
+      "learning_rate": 9.844014420958567e-06,
+      "loss": 0.7216,
+      "step": 9984
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 32.93867484789686,
+      "learning_rate": 9.84232822938356e-06,
+      "loss": 0.5544,
+      "step": 9985
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 36.080427843880194,
+      "learning_rate": 9.840642042292643e-06,
+      "loss": 0.5552,
+      "step": 9986
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 35.14564750673515,
+      "learning_rate": 9.838955859733773e-06,
+      "loss": 0.5987,
+      "step": 9987
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.4101782781526,
+      "learning_rate": 9.837269681754901e-06,
+      "loss": 0.6427,
+      "step": 9988
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 39.66903391487499,
+      "learning_rate": 9.83558350840399e-06,
+      "loss": 0.6716,
+      "step": 9989
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.05634838922296,
+      "learning_rate": 9.833897339728987e-06,
+      "loss": 0.6334,
+      "step": 9990
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 32.668756289615615,
+      "learning_rate": 9.832211175777841e-06,
+      "loss": 0.5617,
+      "step": 9991
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.043135102785335,
+      "learning_rate": 9.830525016598515e-06,
+      "loss": 0.5548,
+      "step": 9992
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 45.156720857904276,
+      "learning_rate": 9.82883886223896e-06,
+      "loss": 0.6046,
+      "step": 9993
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 44.127789779307236,
+      "learning_rate": 9.827152712747122e-06,
+      "loss": 0.6064,
+      "step": 9994
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 43.243685666590665,
+      "learning_rate": 9.825466568170966e-06,
+      "loss": 0.6776,
+      "step": 9995
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 32.03864575486851,
+      "learning_rate": 9.823780428558437e-06,
+      "loss": 0.6033,
+      "step": 9996
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 36.630224275180986,
+      "learning_rate": 9.822094293957486e-06,
+      "loss": 0.6341,
+      "step": 9997
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.203270724939024,
+      "learning_rate": 9.820408164416071e-06,
+      "loss": 0.7158,
+      "step": 9998
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 39.84580543751378,
+      "learning_rate": 9.818722039982145e-06,
+      "loss": 0.6221,
+      "step": 9999
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.20404532238281,
+      "learning_rate": 9.81703592070366e-06,
+      "loss": 0.6067,
+      "step": 10000
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 52.03545659103641,
+      "learning_rate": 9.81534980662856e-06,
+      "loss": 0.5857,
+      "step": 10001
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 39.649005444597755,
+      "learning_rate": 9.813663697804808e-06,
+      "loss": 0.6769,
+      "step": 10002
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 30.50446993482494,
+      "learning_rate": 9.811977594280348e-06,
+      "loss": 0.5957,
+      "step": 10003
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 31.672879606328685,
+      "learning_rate": 9.810291496103137e-06,
+      "loss": 0.5895,
+      "step": 10004
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 33.833934730828524,
+      "learning_rate": 9.808605403321128e-06,
+      "loss": 0.5603,
+      "step": 10005
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.01932294931282,
+      "learning_rate": 9.806919315982266e-06,
+      "loss": 0.602,
+      "step": 10006
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.94529788209611,
+      "learning_rate": 9.805233234134504e-06,
+      "loss": 0.639,
+      "step": 10007
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.222126817976374,
+      "learning_rate": 9.803547157825796e-06,
+      "loss": 0.5685,
+      "step": 10008
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.783826179372184,
+      "learning_rate": 9.801861087104093e-06,
+      "loss": 0.5832,
+      "step": 10009
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.053845804371754,
+      "learning_rate": 9.800175022017346e-06,
+      "loss": 0.6655,
+      "step": 10010
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 35.96301041122539,
+      "learning_rate": 9.798488962613503e-06,
+      "loss": 0.6262,
+      "step": 10011
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 40.93275355130106,
+      "learning_rate": 9.796802908940516e-06,
+      "loss": 0.6676,
+      "step": 10012
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 37.483289024710736,
+      "learning_rate": 9.795116861046334e-06,
+      "loss": 0.6358,
+      "step": 10013
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 38.04875700758783,
+      "learning_rate": 9.793430818978913e-06,
+      "loss": 0.6013,
+      "step": 10014
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 41.634904983172035,
+      "learning_rate": 9.791744782786196e-06,
+      "loss": 0.6425,
+      "step": 10015
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 48.249295733554774,
+      "learning_rate": 9.790058752516134e-06,
+      "loss": 0.6211,
+      "step": 10016
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 38.15720230338814,
+      "learning_rate": 9.788372728216679e-06,
+      "loss": 0.5908,
+      "step": 10017
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 33.78009749007174,
+      "learning_rate": 9.786686709935781e-06,
+      "loss": 0.586,
+      "step": 10018
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 35.59540147441629,
+      "learning_rate": 9.785000697721391e-06,
+      "loss": 0.5269,
+      "step": 10019
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.969232345515316,
+      "learning_rate": 9.783314691621451e-06,
+      "loss": 0.6846,
+      "step": 10020
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 32.157899564022145,
+      "learning_rate": 9.781628691683916e-06,
+      "loss": 0.6133,
+      "step": 10021
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.54377751195898,
+      "learning_rate": 9.77994269795673e-06,
+      "loss": 0.6197,
+      "step": 10022
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 40.24273348604349,
+      "learning_rate": 9.778256710487849e-06,
+      "loss": 0.6239,
+      "step": 10023
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 34.82259040144881,
+      "learning_rate": 9.776570729325217e-06,
+      "loss": 0.5302,
+      "step": 10024
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.156422041555395,
+      "learning_rate": 9.774884754516783e-06,
+      "loss": 0.5725,
+      "step": 10025
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 37.26610828462153,
+      "learning_rate": 9.773198786110492e-06,
+      "loss": 0.6543,
+      "step": 10026
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 33.95055538026601,
+      "learning_rate": 9.771512824154297e-06,
+      "loss": 0.5771,
+      "step": 10027
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 37.561734770915805,
+      "learning_rate": 9.76982686869614e-06,
+      "loss": 0.6791,
+      "step": 10028
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 30.56947727168869,
+      "learning_rate": 9.768140919783979e-06,
+      "loss": 0.5689,
+      "step": 10029
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.11881350613195,
+      "learning_rate": 9.766454977465751e-06,
+      "loss": 0.6098,
+      "step": 10030
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 48.00668535763956,
+      "learning_rate": 9.764769041789408e-06,
+      "loss": 0.6788,
+      "step": 10031
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 38.30192043977691,
+      "learning_rate": 9.763083112802891e-06,
+      "loss": 0.6441,
+      "step": 10032
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 40.938862732781004,
+      "learning_rate": 9.761397190554156e-06,
+      "loss": 0.5886,
+      "step": 10033
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 46.39271930247031,
+      "learning_rate": 9.759711275091149e-06,
+      "loss": 0.7619,
+      "step": 10034
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 37.01743027232813,
+      "learning_rate": 9.758025366461805e-06,
+      "loss": 0.6218,
+      "step": 10035
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 38.3111921712617,
+      "learning_rate": 9.756339464714081e-06,
+      "loss": 0.6347,
+      "step": 10036
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.90280820413552,
+      "learning_rate": 9.754653569895922e-06,
+      "loss": 0.6005,
+      "step": 10037
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 38.27607289398341,
+      "learning_rate": 9.752967682055269e-06,
+      "loss": 0.5936,
+      "step": 10038
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.569228963524445,
+      "learning_rate": 9.751281801240075e-06,
+      "loss": 0.5977,
+      "step": 10039
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 40.691805457923294,
+      "learning_rate": 9.749595927498277e-06,
+      "loss": 0.5514,
+      "step": 10040
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 33.59502611797128,
+      "learning_rate": 9.747910060877824e-06,
+      "loss": 0.5723,
+      "step": 10041
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 44.30641508627265,
+      "learning_rate": 9.746224201426663e-06,
+      "loss": 0.6191,
+      "step": 10042
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 28.96355924221073,
+      "learning_rate": 9.744538349192736e-06,
+      "loss": 0.5343,
+      "step": 10043
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.80558801665564,
+      "learning_rate": 9.742852504223995e-06,
+      "loss": 0.6599,
+      "step": 10044
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 33.41774978967275,
+      "learning_rate": 9.741166666568371e-06,
+      "loss": 0.4993,
+      "step": 10045
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.9034907390524,
+      "learning_rate": 9.739480836273819e-06,
+      "loss": 0.5869,
+      "step": 10046
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.3222767640702,
+      "learning_rate": 9.737795013388277e-06,
+      "loss": 0.6692,
+      "step": 10047
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.93053307754695,
+      "learning_rate": 9.736109197959693e-06,
+      "loss": 0.5782,
+      "step": 10048
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 43.22991639678269,
+      "learning_rate": 9.734423390036011e-06,
+      "loss": 0.75,
+      "step": 10049
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 34.75917345856418,
+      "learning_rate": 9.732737589665171e-06,
+      "loss": 0.5995,
+      "step": 10050
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.26769682929348,
+      "learning_rate": 9.731051796895116e-06,
+      "loss": 0.5397,
+      "step": 10051
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 66.40100266492387,
+      "learning_rate": 9.729366011773793e-06,
+      "loss": 0.6413,
+      "step": 10052
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.70520547027034,
+      "learning_rate": 9.727680234349138e-06,
+      "loss": 0.6746,
+      "step": 10053
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 41.18782618716196,
+      "learning_rate": 9.725994464669103e-06,
+      "loss": 0.618,
+      "step": 10054
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.3582932769186,
+      "learning_rate": 9.724308702781625e-06,
+      "loss": 0.6571,
+      "step": 10055
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 41.422365137493436,
+      "learning_rate": 9.722622948734644e-06,
+      "loss": 0.6688,
+      "step": 10056
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.313344679707484,
+      "learning_rate": 9.720937202576101e-06,
+      "loss": 0.5847,
+      "step": 10057
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 37.789981155002174,
+      "learning_rate": 9.719251464353944e-06,
+      "loss": 0.5695,
+      "step": 10058
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 32.3365942553534,
+      "learning_rate": 9.717565734116114e-06,
+      "loss": 0.5326,
+      "step": 10059
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 60.88979477838874,
+      "learning_rate": 9.715880011910545e-06,
+      "loss": 0.6481,
+      "step": 10060
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 34.823344217852345,
+      "learning_rate": 9.714194297785184e-06,
+      "loss": 0.5828,
+      "step": 10061
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.906878415151,
+      "learning_rate": 9.71250859178797e-06,
+      "loss": 0.636,
+      "step": 10062
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 45.006079040634916,
+      "learning_rate": 9.710822893966845e-06,
+      "loss": 0.7268,
+      "step": 10063
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 37.590291917767985,
+      "learning_rate": 9.709137204369746e-06,
+      "loss": 0.5807,
+      "step": 10064
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 38.99679900445571,
+      "learning_rate": 9.707451523044614e-06,
+      "loss": 0.6806,
+      "step": 10065
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 32.12787819003123,
+      "learning_rate": 9.705765850039388e-06,
+      "loss": 0.579,
+      "step": 10066
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.83319788723778,
+      "learning_rate": 9.704080185402012e-06,
+      "loss": 0.583,
+      "step": 10067
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.317678129273915,
+      "learning_rate": 9.702394529180424e-06,
+      "loss": 0.5766,
+      "step": 10068
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 31.767948026288934,
+      "learning_rate": 9.70070888142256e-06,
+      "loss": 0.6064,
+      "step": 10069
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 31.080369242115793,
+      "learning_rate": 9.699023242176358e-06,
+      "loss": 0.5522,
+      "step": 10070
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.2488751434872,
+      "learning_rate": 9.69733761148976e-06,
+      "loss": 0.5579,
+      "step": 10071
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 33.721251199633684,
+      "learning_rate": 9.695651989410702e-06,
+      "loss": 0.5244,
+      "step": 10072
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 34.867331536790644,
+      "learning_rate": 9.693966375987128e-06,
+      "loss": 0.541,
+      "step": 10073
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.308282620398444,
+      "learning_rate": 9.692280771266969e-06,
+      "loss": 0.6128,
+      "step": 10074
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.254450360425885,
+      "learning_rate": 9.690595175298164e-06,
+      "loss": 0.5654,
+      "step": 10075
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 44.09896047668824,
+      "learning_rate": 9.68890958812865e-06,
+      "loss": 0.6764,
+      "step": 10076
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 31.356430504689744,
+      "learning_rate": 9.687224009806366e-06,
+      "loss": 0.5823,
+      "step": 10077
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 41.081551726298315,
+      "learning_rate": 9.685538440379253e-06,
+      "loss": 0.527,
+      "step": 10078
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 42.56372790588969,
+      "learning_rate": 9.683852879895236e-06,
+      "loss": 0.664,
+      "step": 10079
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 37.948476105401504,
+      "learning_rate": 9.682167328402261e-06,
+      "loss": 0.6138,
+      "step": 10080
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 41.72810408060976,
+      "learning_rate": 9.680481785948263e-06,
+      "loss": 0.6736,
+      "step": 10081
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 36.16730155951559,
+      "learning_rate": 9.678796252581171e-06,
+      "loss": 0.5709,
+      "step": 10082
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.27393976500865,
+      "learning_rate": 9.677110728348932e-06,
+      "loss": 0.5668,
+      "step": 10083
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 39.42990210202452,
+      "learning_rate": 9.675425213299471e-06,
+      "loss": 0.5869,
+      "step": 10084
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 36.84764862141348,
+      "learning_rate": 9.673739707480727e-06,
+      "loss": 0.6747,
+      "step": 10085
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 35.64532145086755,
+      "learning_rate": 9.672054210940638e-06,
+      "loss": 0.6031,
+      "step": 10086
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 45.43410908509512,
+      "learning_rate": 9.670368723727131e-06,
+      "loss": 0.5831,
+      "step": 10087
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 44.365999969318715,
+      "learning_rate": 9.66868324588815e-06,
+      "loss": 0.6056,
+      "step": 10088
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 42.027283534540366,
+      "learning_rate": 9.66699777747162e-06,
+      "loss": 0.7475,
+      "step": 10089
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.958137102733,
+      "learning_rate": 9.665312318525478e-06,
+      "loss": 0.5968,
+      "step": 10090
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 41.49432412365286,
+      "learning_rate": 9.663626869097657e-06,
+      "loss": 0.6698,
+      "step": 10091
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 34.1487904296734,
+      "learning_rate": 9.661941429236094e-06,
+      "loss": 0.5696,
+      "step": 10092
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.413563794725,
+      "learning_rate": 9.66025599898872e-06,
+      "loss": 0.5926,
+      "step": 10093
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 36.225251243804685,
+      "learning_rate": 9.658570578403465e-06,
+      "loss": 0.618,
+      "step": 10094
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.73791806542262,
+      "learning_rate": 9.656885167528261e-06,
+      "loss": 0.5896,
+      "step": 10095
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 35.703946968483265,
+      "learning_rate": 9.655199766411044e-06,
+      "loss": 0.5846,
+      "step": 10096
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 43.40763691278638,
+      "learning_rate": 9.653514375099742e-06,
+      "loss": 0.6969,
+      "step": 10097
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 41.060766052842745,
+      "learning_rate": 9.651828993642293e-06,
+      "loss": 0.5501,
+      "step": 10098
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 45.15441442012565,
+      "learning_rate": 9.650143622086621e-06,
+      "loss": 0.6479,
+      "step": 10099
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 45.954340918991086,
+      "learning_rate": 9.64845826048066e-06,
+      "loss": 0.686,
+      "step": 10100
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 41.56222425670506,
+      "learning_rate": 9.646772908872338e-06,
+      "loss": 0.7043,
+      "step": 10101
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 42.67745096602167,
+      "learning_rate": 9.64508756730959e-06,
+      "loss": 0.6775,
+      "step": 10102
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 36.78473870740319,
+      "learning_rate": 9.643402235840346e-06,
+      "loss": 0.6395,
+      "step": 10103
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.76452959289669,
+      "learning_rate": 9.641716914512532e-06,
+      "loss": 0.5938,
+      "step": 10104
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 35.20181499537789,
+      "learning_rate": 9.640031603374078e-06,
+      "loss": 0.5567,
+      "step": 10105
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 40.906442663614754,
+      "learning_rate": 9.638346302472916e-06,
+      "loss": 0.6378,
+      "step": 10106
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 32.97493403502627,
+      "learning_rate": 9.636661011856971e-06,
+      "loss": 0.5566,
+      "step": 10107
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 40.12130894425428,
+      "learning_rate": 9.63497573157418e-06,
+      "loss": 0.6474,
+      "step": 10108
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.06571150543893,
+      "learning_rate": 9.633290461672463e-06,
+      "loss": 0.5344,
+      "step": 10109
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.558598922109226,
+      "learning_rate": 9.631605202199748e-06,
+      "loss": 0.6359,
+      "step": 10110
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.84842927449529,
+      "learning_rate": 9.629919953203966e-06,
+      "loss": 0.57,
+      "step": 10111
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 45.58161797803222,
+      "learning_rate": 9.628234714733048e-06,
+      "loss": 0.6378,
+      "step": 10112
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.79428189323093,
+      "learning_rate": 9.626549486834916e-06,
+      "loss": 0.6085,
+      "step": 10113
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.747882250288214,
+      "learning_rate": 9.624864269557495e-06,
+      "loss": 0.5395,
+      "step": 10114
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 35.133569500151715,
+      "learning_rate": 9.623179062948716e-06,
+      "loss": 0.558,
+      "step": 10115
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.891343845392385,
+      "learning_rate": 9.621493867056505e-06,
+      "loss": 0.6279,
+      "step": 10116
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.13253229981151,
+      "learning_rate": 9.619808681928788e-06,
+      "loss": 0.6035,
+      "step": 10117
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 32.7337755680964,
+      "learning_rate": 9.618123507613487e-06,
+      "loss": 0.5846,
+      "step": 10118
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.582051618711205,
+      "learning_rate": 9.616438344158533e-06,
+      "loss": 0.6735,
+      "step": 10119
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.557684315030972,
+      "learning_rate": 9.614753191611846e-06,
+      "loss": 0.5338,
+      "step": 10120
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 40.78018483491873,
+      "learning_rate": 9.613068050021353e-06,
+      "loss": 0.6173,
+      "step": 10121
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 44.39530290704161,
+      "learning_rate": 9.61138291943498e-06,
+      "loss": 0.5699,
+      "step": 10122
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 46.63853091550071,
+      "learning_rate": 9.609697799900647e-06,
+      "loss": 0.6102,
+      "step": 10123
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 35.500589567312396,
+      "learning_rate": 9.608012691466281e-06,
+      "loss": 0.6085,
+      "step": 10124
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 36.3633260641622,
+      "learning_rate": 9.606327594179806e-06,
+      "loss": 0.6162,
+      "step": 10125
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 31.303069630517058,
+      "learning_rate": 9.604642508089142e-06,
+      "loss": 0.5871,
+      "step": 10126
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 40.89000348678149,
+      "learning_rate": 9.602957433242218e-06,
+      "loss": 0.6712,
+      "step": 10127
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 36.90072561742042,
+      "learning_rate": 9.601272369686948e-06,
+      "loss": 0.5771,
+      "step": 10128
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.2258391399094,
+      "learning_rate": 9.599587317471259e-06,
+      "loss": 0.6564,
+      "step": 10129
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 42.087752710075996,
+      "learning_rate": 9.597902276643074e-06,
+      "loss": 0.6015,
+      "step": 10130
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 30.284298145974883,
+      "learning_rate": 9.59621724725031e-06,
+      "loss": 0.5455,
+      "step": 10131
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 34.89912620191093,
+      "learning_rate": 9.594532229340898e-06,
+      "loss": 0.5234,
+      "step": 10132
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.19273056622923,
+      "learning_rate": 9.592847222962744e-06,
+      "loss": 0.6327,
+      "step": 10133
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.858301777547,
+      "learning_rate": 9.591162228163781e-06,
+      "loss": 0.6279,
+      "step": 10134
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.458949590892814,
+      "learning_rate": 9.589477244991924e-06,
+      "loss": 0.5765,
+      "step": 10135
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.391225852235436,
+      "learning_rate": 9.587792273495095e-06,
+      "loss": 0.5847,
+      "step": 10136
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.55490533582612,
+      "learning_rate": 9.586107313721214e-06,
+      "loss": 0.613,
+      "step": 10137
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.28897846915553,
+      "learning_rate": 9.584422365718197e-06,
+      "loss": 0.6308,
+      "step": 10138
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.560292482147325,
+      "learning_rate": 9.582737429533964e-06,
+      "loss": 0.5474,
+      "step": 10139
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 49.24655481574879,
+      "learning_rate": 9.581052505216434e-06,
+      "loss": 0.6475,
+      "step": 10140
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.644681209223165,
+      "learning_rate": 9.579367592813526e-06,
+      "loss": 0.6595,
+      "step": 10141
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 44.43642606658123,
+      "learning_rate": 9.577682692373161e-06,
+      "loss": 0.776,
+      "step": 10142
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 34.48342864479868,
+      "learning_rate": 9.575997803943251e-06,
+      "loss": 0.5708,
+      "step": 10143
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 43.57844853546088,
+      "learning_rate": 9.574312927571717e-06,
+      "loss": 0.5695,
+      "step": 10144
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 41.63239465504891,
+      "learning_rate": 9.572628063306469e-06,
+      "loss": 0.6429,
+      "step": 10145
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 34.70384313122878,
+      "learning_rate": 9.570943211195434e-06,
+      "loss": 0.6162,
+      "step": 10146
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.41517242423007,
+      "learning_rate": 9.569258371286523e-06,
+      "loss": 0.6363,
+      "step": 10147
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 28.214085350739307,
+      "learning_rate": 9.567573543627649e-06,
+      "loss": 0.5913,
+      "step": 10148
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 32.94422688639294,
+      "learning_rate": 9.565888728266732e-06,
+      "loss": 0.6222,
+      "step": 10149
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 33.631267247162796,
+      "learning_rate": 9.564203925251685e-06,
+      "loss": 0.5873,
+      "step": 10150
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.416410777032866,
+      "learning_rate": 9.562519134630423e-06,
+      "loss": 0.6953,
+      "step": 10151
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 29.951736465769578,
+      "learning_rate": 9.560834356450864e-06,
+      "loss": 0.5694,
+      "step": 10152
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 34.22021466260318,
+      "learning_rate": 9.559149590760917e-06,
+      "loss": 0.6141,
+      "step": 10153
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 31.131382088724646,
+      "learning_rate": 9.557464837608496e-06,
+      "loss": 0.6189,
+      "step": 10154
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 32.37470342065772,
+      "learning_rate": 9.555780097041517e-06,
+      "loss": 0.4999,
+      "step": 10155
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 37.82067327203208,
+      "learning_rate": 9.554095369107892e-06,
+      "loss": 0.6013,
+      "step": 10156
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 49.31596122072564,
+      "learning_rate": 9.55241065385554e-06,
+      "loss": 0.5419,
+      "step": 10157
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 36.64060544336769,
+      "learning_rate": 9.55072595133236e-06,
+      "loss": 0.6647,
+      "step": 10158
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 31.110963291335505,
+      "learning_rate": 9.549041261586273e-06,
+      "loss": 0.5435,
+      "step": 10159
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 31.334077127908248,
+      "learning_rate": 9.547356584665185e-06,
+      "loss": 0.5845,
+      "step": 10160
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 36.08020111362036,
+      "learning_rate": 9.545671920617018e-06,
+      "loss": 0.6039,
+      "step": 10161
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 32.42429891326894,
+      "learning_rate": 9.543987269489673e-06,
+      "loss": 0.5411,
+      "step": 10162
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 42.202945374124575,
+      "learning_rate": 9.542302631331063e-06,
+      "loss": 0.6633,
+      "step": 10163
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 38.081966377005045,
+      "learning_rate": 9.540618006189096e-06,
+      "loss": 0.6532,
+      "step": 10164
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 44.838667647357994,
+      "learning_rate": 9.538933394111687e-06,
+      "loss": 0.6648,
+      "step": 10165
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 34.88735252986125,
+      "learning_rate": 9.537248795146745e-06,
+      "loss": 0.571,
+      "step": 10166
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 32.785516964843694,
+      "learning_rate": 9.535564209342172e-06,
+      "loss": 0.5767,
+      "step": 10167
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.66020400996999,
+      "learning_rate": 9.533879636745883e-06,
+      "loss": 0.6882,
+      "step": 10168
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 36.22409236105232,
+      "learning_rate": 9.532195077405784e-06,
+      "loss": 0.6186,
+      "step": 10169
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 33.2681677560059,
+      "learning_rate": 9.53051053136978e-06,
+      "loss": 0.582,
+      "step": 10170
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 37.281975265060886,
+      "learning_rate": 9.528825998685788e-06,
+      "loss": 0.6313,
+      "step": 10171
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 40.8256616969776,
+      "learning_rate": 9.527141479401708e-06,
+      "loss": 0.6119,
+      "step": 10172
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 39.50210143788791,
+      "learning_rate": 9.525456973565443e-06,
+      "loss": 0.5515,
+      "step": 10173
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 30.293581367953585,
+      "learning_rate": 9.523772481224908e-06,
+      "loss": 0.5507,
+      "step": 10174
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 36.29372792060171,
+      "learning_rate": 9.522088002428003e-06,
+      "loss": 0.5856,
+      "step": 10175
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 31.9329771471135,
+      "learning_rate": 9.520403537222642e-06,
+      "loss": 0.5547,
+      "step": 10176
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 47.09500895983567,
+      "learning_rate": 9.518719085656716e-06,
+      "loss": 0.6728,
+      "step": 10177
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 42.50821075505288,
+      "learning_rate": 9.51703464777814e-06,
+      "loss": 0.6974,
+      "step": 10178
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 41.25480068188227,
+      "learning_rate": 9.515350223634815e-06,
+      "loss": 0.6195,
+      "step": 10179
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 33.756211285003005,
+      "learning_rate": 9.513665813274647e-06,
+      "loss": 0.5615,
+      "step": 10180
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 37.99755058425256,
+      "learning_rate": 9.511981416745542e-06,
+      "loss": 0.5415,
+      "step": 10181
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.78665310767317,
+      "learning_rate": 9.510297034095398e-06,
+      "loss": 0.6776,
+      "step": 10182
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 33.54549180880693,
+      "learning_rate": 9.508612665372117e-06,
+      "loss": 0.5805,
+      "step": 10183
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 57.606931274001184,
+      "learning_rate": 9.506928310623608e-06,
+      "loss": 0.6198,
+      "step": 10184
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 31.77909041037504,
+      "learning_rate": 9.505243969897766e-06,
+      "loss": 0.511,
+      "step": 10185
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 53.48191797094102,
+      "learning_rate": 9.5035596432425e-06,
+      "loss": 0.6976,
+      "step": 10186
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.48503377664199,
+      "learning_rate": 9.501875330705706e-06,
+      "loss": 0.6726,
+      "step": 10187
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 39.725538254458876,
+      "learning_rate": 9.500191032335286e-06,
+      "loss": 0.6945,
+      "step": 10188
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 44.49431425911124,
+      "learning_rate": 9.498506748179137e-06,
+      "loss": 0.6839,
+      "step": 10189
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 38.828075717457246,
+      "learning_rate": 9.496822478285167e-06,
+      "loss": 0.6451,
+      "step": 10190
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 44.613726792124396,
+      "learning_rate": 9.495138222701273e-06,
+      "loss": 0.6906,
+      "step": 10191
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 46.2641808240303,
+      "learning_rate": 9.493453981475348e-06,
+      "loss": 0.6918,
+      "step": 10192
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 38.93830696237212,
+      "learning_rate": 9.491769754655298e-06,
+      "loss": 0.6106,
+      "step": 10193
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 73.82623794646764,
+      "learning_rate": 9.490085542289016e-06,
+      "loss": 0.7139,
+      "step": 10194
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.594798036508486,
+      "learning_rate": 9.488401344424404e-06,
+      "loss": 0.731,
+      "step": 10195
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 34.59462091853797,
+      "learning_rate": 9.486717161109363e-06,
+      "loss": 0.5777,
+      "step": 10196
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 35.57236527573651,
+      "learning_rate": 9.485032992391782e-06,
+      "loss": 0.6109,
+      "step": 10197
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 33.762664696928326,
+      "learning_rate": 9.48334883831956e-06,
+      "loss": 0.583,
+      "step": 10198
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 39.052470624089814,
+      "learning_rate": 9.481664698940598e-06,
+      "loss": 0.6621,
+      "step": 10199
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.5570443950043,
+      "learning_rate": 9.479980574302785e-06,
+      "loss": 0.5973,
+      "step": 10200
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 34.994003422523456,
+      "learning_rate": 9.478296464454028e-06,
+      "loss": 0.6149,
+      "step": 10201
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 42.77855097614127,
+      "learning_rate": 9.476612369442207e-06,
+      "loss": 0.5803,
+      "step": 10202
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 38.40526178204463,
+      "learning_rate": 9.474928289315224e-06,
+      "loss": 0.5923,
+      "step": 10203
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 42.27191022300536,
+      "learning_rate": 9.473244224120974e-06,
+      "loss": 0.7492,
+      "step": 10204
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 34.99126496986475,
+      "learning_rate": 9.471560173907353e-06,
+      "loss": 0.6393,
+      "step": 10205
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 41.03975151292593,
+      "learning_rate": 9.469876138722252e-06,
+      "loss": 0.5938,
+      "step": 10206
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 38.01413997347632,
+      "learning_rate": 9.46819211861356e-06,
+      "loss": 0.6255,
+      "step": 10207
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 36.44640746930058,
+      "learning_rate": 9.466508113629174e-06,
+      "loss": 0.603,
+      "step": 10208
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 39.62999697143539,
+      "learning_rate": 9.464824123816986e-06,
+      "loss": 0.5767,
+      "step": 10209
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 29.63828856771807,
+      "learning_rate": 9.463140149224888e-06,
+      "loss": 0.5935,
+      "step": 10210
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 41.846579264456906,
+      "learning_rate": 9.461456189900767e-06,
+      "loss": 0.6181,
+      "step": 10211
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.784932575873874,
+      "learning_rate": 9.459772245892518e-06,
+      "loss": 0.5784,
+      "step": 10212
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.711183887698322,
+      "learning_rate": 9.458088317248032e-06,
+      "loss": 0.5818,
+      "step": 10213
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 47.25130770907152,
+      "learning_rate": 9.456404404015194e-06,
+      "loss": 0.6229,
+      "step": 10214
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.37598418548693,
+      "learning_rate": 9.454720506241902e-06,
+      "loss": 0.5964,
+      "step": 10215
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 51.58802206312762,
+      "learning_rate": 9.453036623976036e-06,
+      "loss": 0.6238,
+      "step": 10216
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 30.62970316059024,
+      "learning_rate": 9.451352757265488e-06,
+      "loss": 0.6023,
+      "step": 10217
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 37.153408220989036,
+      "learning_rate": 9.449668906158149e-06,
+      "loss": 0.6428,
+      "step": 10218
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 52.187574799992305,
+      "learning_rate": 9.4479850707019e-06,
+      "loss": 0.682,
+      "step": 10219
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.27611397680031,
+      "learning_rate": 9.446301250944641e-06,
+      "loss": 0.5543,
+      "step": 10220
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 43.11821577916469,
+      "learning_rate": 9.444617446934244e-06,
+      "loss": 0.626,
+      "step": 10221
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 43.922025210063644,
+      "learning_rate": 9.442933658718603e-06,
+      "loss": 0.7133,
+      "step": 10222
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.543714080602648,
+      "learning_rate": 9.4412498863456e-06,
+      "loss": 0.574,
+      "step": 10223
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 37.17211014431519,
+      "learning_rate": 9.439566129863126e-06,
+      "loss": 0.5448,
+      "step": 10224
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.2843422678057,
+      "learning_rate": 9.437882389319067e-06,
+      "loss": 0.578,
+      "step": 10225
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.98620825738495,
+      "learning_rate": 9.436198664761302e-06,
+      "loss": 0.6129,
+      "step": 10226
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.059512165759095,
+      "learning_rate": 9.434514956237714e-06,
+      "loss": 0.5901,
+      "step": 10227
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.1809277911632,
+      "learning_rate": 9.43283126379619e-06,
+      "loss": 0.582,
+      "step": 10228
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 30.966942858584964,
+      "learning_rate": 9.431147587484614e-06,
+      "loss": 0.5596,
+      "step": 10229
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.73516984448466,
+      "learning_rate": 9.429463927350872e-06,
+      "loss": 0.5503,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.399225282527205,
+      "learning_rate": 9.427780283442838e-06,
+      "loss": 0.5689,
+      "step": 10231
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 45.653938762382246,
+      "learning_rate": 9.4260966558084e-06,
+      "loss": 0.6885,
+      "step": 10232
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 36.53972572114623,
+      "learning_rate": 9.424413044495435e-06,
+      "loss": 0.585,
+      "step": 10233
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 33.34098579940203,
+      "learning_rate": 9.422729449551828e-06,
+      "loss": 0.5249,
+      "step": 10234
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.201399352042465,
+      "learning_rate": 9.42104587102546e-06,
+      "loss": 0.5666,
+      "step": 10235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.15504443664322,
+      "learning_rate": 9.419362308964203e-06,
+      "loss": 0.643,
+      "step": 10236
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 39.78265782312671,
+      "learning_rate": 9.417678763415948e-06,
+      "loss": 0.6366,
+      "step": 10237
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 39.210304949838616,
+      "learning_rate": 9.415995234428563e-06,
+      "loss": 0.6597,
+      "step": 10238
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 44.063093949772046,
+      "learning_rate": 9.414311722049935e-06,
+      "loss": 0.6585,
+      "step": 10239
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.30736340068893,
+      "learning_rate": 9.41262822632794e-06,
+      "loss": 0.588,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 40.47018473081323,
+      "learning_rate": 9.410944747310454e-06,
+      "loss": 0.6607,
+      "step": 10241
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.48132255994303,
+      "learning_rate": 9.409261285045352e-06,
+      "loss": 0.6026,
+      "step": 10242
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 42.39040768965221,
+      "learning_rate": 9.407577839580516e-06,
+      "loss": 0.6872,
+      "step": 10243
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.022028879287593,
+      "learning_rate": 9.405894410963817e-06,
+      "loss": 0.5906,
+      "step": 10244
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 40.322219431012954,
+      "learning_rate": 9.404210999243141e-06,
+      "loss": 0.6425,
+      "step": 10245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.9954893432335,
+      "learning_rate": 9.402527604466347e-06,
+      "loss": 0.5217,
+      "step": 10246
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.808037291745585,
+      "learning_rate": 9.400844226681324e-06,
+      "loss": 0.6866,
+      "step": 10247
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 37.27869190351344,
+      "learning_rate": 9.399160865935936e-06,
+      "loss": 0.4802,
+      "step": 10248
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.920986381631586,
+      "learning_rate": 9.397477522278064e-06,
+      "loss": 0.5487,
+      "step": 10249
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 36.47880868538206,
+      "learning_rate": 9.395794195755581e-06,
+      "loss": 0.505,
+      "step": 10250
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.25898507769916,
+      "learning_rate": 9.394110886416355e-06,
+      "loss": 0.5246,
+      "step": 10251
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 34.3115774284254,
+      "learning_rate": 9.39242759430826e-06,
+      "loss": 0.5465,
+      "step": 10252
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 34.047792689301104,
+      "learning_rate": 9.390744319479171e-06,
+      "loss": 0.5414,
+      "step": 10253
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 45.87238672541336,
+      "learning_rate": 9.389061061976958e-06,
+      "loss": 0.7427,
+      "step": 10254
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 33.884770451804336,
+      "learning_rate": 9.387377821849489e-06,
+      "loss": 0.519,
+      "step": 10255
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.0380870813901,
+      "learning_rate": 9.385694599144636e-06,
+      "loss": 0.5515,
+      "step": 10256
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 45.919221856572065,
+      "learning_rate": 9.38401139391027e-06,
+      "loss": 0.6029,
+      "step": 10257
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 34.99824953470268,
+      "learning_rate": 9.382328206194259e-06,
+      "loss": 0.5884,
+      "step": 10258
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 34.16133013933528,
+      "learning_rate": 9.380645036044473e-06,
+      "loss": 0.6555,
+      "step": 10259
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 41.776875483031766,
+      "learning_rate": 9.37896188350878e-06,
+      "loss": 0.6835,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 34.07617058062227,
+      "learning_rate": 9.377278748635046e-06,
+      "loss": 0.5866,
+      "step": 10261
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 35.692373975101255,
+      "learning_rate": 9.375595631471143e-06,
+      "loss": 0.6094,
+      "step": 10262
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 41.04625866111025,
+      "learning_rate": 9.373912532064931e-06,
+      "loss": 0.6415,
+      "step": 10263
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.14486482614039,
+      "learning_rate": 9.372229450464287e-06,
+      "loss": 0.6449,
+      "step": 10264
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 39.051286989268206,
+      "learning_rate": 9.370546386717065e-06,
+      "loss": 0.697,
+      "step": 10265
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 37.34293917809181,
+      "learning_rate": 9.368863340871137e-06,
+      "loss": 0.5916,
+      "step": 10266
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 34.68708928526621,
+      "learning_rate": 9.367180312974364e-06,
+      "loss": 0.5292,
+      "step": 10267
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 40.32703539777867,
+      "learning_rate": 9.365497303074615e-06,
+      "loss": 0.6305,
+      "step": 10268
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.66649835569138,
+      "learning_rate": 9.363814311219754e-06,
+      "loss": 0.5975,
+      "step": 10269
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 33.44893670979451,
+      "learning_rate": 9.36213133745764e-06,
+      "loss": 0.5329,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.02633980681384,
+      "learning_rate": 9.360448381836134e-06,
+      "loss": 0.59,
+      "step": 10271
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.90603568432058,
+      "learning_rate": 9.358765444403107e-06,
+      "loss": 0.5374,
+      "step": 10272
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.85169979615613,
+      "learning_rate": 9.35708252520641e-06,
+      "loss": 0.5386,
+      "step": 10273
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.34322146025786,
+      "learning_rate": 9.355399624293917e-06,
+      "loss": 0.7259,
+      "step": 10274
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 31.311968838875032,
+      "learning_rate": 9.353716741713477e-06,
+      "loss": 0.5887,
+      "step": 10275
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.858904762389486,
+      "learning_rate": 9.352033877512957e-06,
+      "loss": 0.5575,
+      "step": 10276
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 35.47284658024092,
+      "learning_rate": 9.35035103174021e-06,
+      "loss": 0.5621,
+      "step": 10277
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 42.118618982747826,
+      "learning_rate": 9.348668204443103e-06,
+      "loss": 0.6784,
+      "step": 10278
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.36433368072972,
+      "learning_rate": 9.346985395669494e-06,
+      "loss": 0.5065,
+      "step": 10279
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.17185667579206,
+      "learning_rate": 9.345302605467232e-06,
+      "loss": 0.5612,
+      "step": 10280
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 47.776949347785546,
+      "learning_rate": 9.343619833884186e-06,
+      "loss": 0.6621,
+      "step": 10281
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 43.57097539453888,
+      "learning_rate": 9.341937080968204e-06,
+      "loss": 0.6613,
+      "step": 10282
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.70064010004503,
+      "learning_rate": 9.340254346767149e-06,
+      "loss": 0.6021,
+      "step": 10283
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 66.33453515016795,
+      "learning_rate": 9.338571631328877e-06,
+      "loss": 0.6024,
+      "step": 10284
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.132439596365444,
+      "learning_rate": 9.336888934701238e-06,
+      "loss": 0.5814,
+      "step": 10285
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.73056287630722,
+      "learning_rate": 9.335206256932088e-06,
+      "loss": 0.6029,
+      "step": 10286
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 40.31664798302998,
+      "learning_rate": 9.333523598069286e-06,
+      "loss": 0.626,
+      "step": 10287
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.916646731374904,
+      "learning_rate": 9.331840958160679e-06,
+      "loss": 0.6765,
+      "step": 10288
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 35.30716341592526,
+      "learning_rate": 9.330158337254131e-06,
+      "loss": 0.5536,
+      "step": 10289
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.58616233085776,
+      "learning_rate": 9.328475735397483e-06,
+      "loss": 0.6393,
+      "step": 10290
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.12457777390352,
+      "learning_rate": 9.326793152638594e-06,
+      "loss": 0.6333,
+      "step": 10291
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 41.04036663662211,
+      "learning_rate": 9.325110589025311e-06,
+      "loss": 0.6,
+      "step": 10292
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.55429776969663,
+      "learning_rate": 9.323428044605491e-06,
+      "loss": 0.593,
+      "step": 10293
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.285818345687574,
+      "learning_rate": 9.321745519426984e-06,
+      "loss": 0.5607,
+      "step": 10294
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.58088709777985,
+      "learning_rate": 9.320063013537634e-06,
+      "loss": 0.6352,
+      "step": 10295
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.091883711631475,
+      "learning_rate": 9.318380526985293e-06,
+      "loss": 0.5507,
+      "step": 10296
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 43.54466336112187,
+      "learning_rate": 9.316698059817814e-06,
+      "loss": 0.5708,
+      "step": 10297
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.33843629452972,
+      "learning_rate": 9.315015612083039e-06,
+      "loss": 0.5508,
+      "step": 10298
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 43.109437235240804,
+      "learning_rate": 9.313333183828824e-06,
+      "loss": 0.5854,
+      "step": 10299
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 42.05522393811591,
+      "learning_rate": 9.31165077510301e-06,
+      "loss": 0.5902,
+      "step": 10300
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.09647790891408,
+      "learning_rate": 9.309968385953445e-06,
+      "loss": 0.5719,
+      "step": 10301
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 46.2408945038512,
+      "learning_rate": 9.308286016427974e-06,
+      "loss": 0.6858,
+      "step": 10302
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.14982651512327,
+      "learning_rate": 9.306603666574446e-06,
+      "loss": 0.54,
+      "step": 10303
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.78392912180578,
+      "learning_rate": 9.304921336440704e-06,
+      "loss": 0.5757,
+      "step": 10304
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 43.152470184911685,
+      "learning_rate": 9.303239026074589e-06,
+      "loss": 0.633,
+      "step": 10305
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.28893844515122,
+      "learning_rate": 9.301556735523952e-06,
+      "loss": 0.5536,
+      "step": 10306
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 49.55136743423597,
+      "learning_rate": 9.29987446483663e-06,
+      "loss": 0.6777,
+      "step": 10307
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 35.67848582089067,
+      "learning_rate": 9.298192214060476e-06,
+      "loss": 0.596,
+      "step": 10308
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 40.88218567475379,
+      "learning_rate": 9.296509983243319e-06,
+      "loss": 0.5242,
+      "step": 10309
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.56364160369396,
+      "learning_rate": 9.294827772433006e-06,
+      "loss": 0.6437,
+      "step": 10310
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.88032753641456,
+      "learning_rate": 9.293145581677378e-06,
+      "loss": 0.5775,
+      "step": 10311
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 35.67997807400505,
+      "learning_rate": 9.291463411024279e-06,
+      "loss": 0.5798,
+      "step": 10312
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 35.105473293509164,
+      "learning_rate": 9.289781260521548e-06,
+      "loss": 0.6103,
+      "step": 10313
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 47.38262005520804,
+      "learning_rate": 9.288099130217018e-06,
+      "loss": 0.7275,
+      "step": 10314
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.97644690739047,
+      "learning_rate": 9.286417020158533e-06,
+      "loss": 0.5483,
+      "step": 10315
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.66200929631827,
+      "learning_rate": 9.284734930393931e-06,
+      "loss": 0.5311,
+      "step": 10316
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.139951052505815,
+      "learning_rate": 9.283052860971049e-06,
+      "loss": 0.5528,
+      "step": 10317
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.50990596269268,
+      "learning_rate": 9.281370811937726e-06,
+      "loss": 0.561,
+      "step": 10318
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 44.87057149175849,
+      "learning_rate": 9.279688783341796e-06,
+      "loss": 0.6268,
+      "step": 10319
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.46041061964969,
+      "learning_rate": 9.278006775231097e-06,
+      "loss": 0.6208,
+      "step": 10320
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.09101473807231,
+      "learning_rate": 9.276324787653458e-06,
+      "loss": 0.4967,
+      "step": 10321
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.23433104443566,
+      "learning_rate": 9.274642820656722e-06,
+      "loss": 0.647,
+      "step": 10322
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.270280227341274,
+      "learning_rate": 9.272960874288722e-06,
+      "loss": 0.572,
+      "step": 10323
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.455385815031,
+      "learning_rate": 9.271278948597285e-06,
+      "loss": 0.5606,
+      "step": 10324
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 43.48085613042184,
+      "learning_rate": 9.269597043630252e-06,
+      "loss": 0.5489,
+      "step": 10325
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.38184483678676,
+      "learning_rate": 9.267915159435447e-06,
+      "loss": 0.5846,
+      "step": 10326
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.802055545834286,
+      "learning_rate": 9.266233296060712e-06,
+      "loss": 0.5661,
+      "step": 10327
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.76360072880671,
+      "learning_rate": 9.264551453553872e-06,
+      "loss": 0.5818,
+      "step": 10328
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 34.82356387735172,
+      "learning_rate": 9.262869631962758e-06,
+      "loss": 0.6062,
+      "step": 10329
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.222528848554184,
+      "learning_rate": 9.261187831335197e-06,
+      "loss": 0.6923,
+      "step": 10330
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 43.40865102324463,
+      "learning_rate": 9.259506051719025e-06,
+      "loss": 0.6225,
+      "step": 10331
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.796247923558504,
+      "learning_rate": 9.257824293162066e-06,
+      "loss": 0.5872,
+      "step": 10332
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.354545339680186,
+      "learning_rate": 9.256142555712156e-06,
+      "loss": 0.5933,
+      "step": 10333
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 30.186579009039463,
+      "learning_rate": 9.25446083941711e-06,
+      "loss": 0.6023,
+      "step": 10334
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.703448171425244,
+      "learning_rate": 9.252779144324763e-06,
+      "loss": 0.6577,
+      "step": 10335
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 30.75101020750818,
+      "learning_rate": 9.251097470482938e-06,
+      "loss": 0.5295,
+      "step": 10336
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.80009239559855,
+      "learning_rate": 9.249415817939465e-06,
+      "loss": 0.5961,
+      "step": 10337
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.888759285523996,
+      "learning_rate": 9.24773418674217e-06,
+      "loss": 0.5867,
+      "step": 10338
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 41.34435253900842,
+      "learning_rate": 9.246052576938874e-06,
+      "loss": 0.6559,
+      "step": 10339
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.33961415943465,
+      "learning_rate": 9.244370988577396e-06,
+      "loss": 0.562,
+      "step": 10340
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 52.371723313814435,
+      "learning_rate": 9.24268942170557e-06,
+      "loss": 0.5671,
+      "step": 10341
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.387830457240135,
+      "learning_rate": 9.24100787637121e-06,
+      "loss": 0.7085,
+      "step": 10342
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 37.301724441848556,
+      "learning_rate": 9.239326352622146e-06,
+      "loss": 0.655,
+      "step": 10343
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.90867213574316,
+      "learning_rate": 9.237644850506196e-06,
+      "loss": 0.5749,
+      "step": 10344
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 37.83846097947628,
+      "learning_rate": 9.235963370071176e-06,
+      "loss": 0.6418,
+      "step": 10345
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 28.820401943085574,
+      "learning_rate": 9.234281911364914e-06,
+      "loss": 0.4942,
+      "step": 10346
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 38.14420377763618,
+      "learning_rate": 9.232600474435226e-06,
+      "loss": 0.6157,
+      "step": 10347
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.571458887551636,
+      "learning_rate": 9.230919059329934e-06,
+      "loss": 0.678,
+      "step": 10348
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 39.00065157419385,
+      "learning_rate": 9.229237666096848e-06,
+      "loss": 0.5979,
+      "step": 10349
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.21698409771014,
+      "learning_rate": 9.227556294783797e-06,
+      "loss": 0.6371,
+      "step": 10350
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.26054991600947,
+      "learning_rate": 9.22587494543859e-06,
+      "loss": 0.6413,
+      "step": 10351
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 44.44217218444987,
+      "learning_rate": 9.224193618109052e-06,
+      "loss": 0.6077,
+      "step": 10352
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 32.14856401235502,
+      "learning_rate": 9.222512312842988e-06,
+      "loss": 0.5705,
+      "step": 10353
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.34436342413393,
+      "learning_rate": 9.220831029688222e-06,
+      "loss": 0.5615,
+      "step": 10354
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.662680344471575,
+      "learning_rate": 9.219149768692563e-06,
+      "loss": 0.5634,
+      "step": 10355
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 31.99916886949043,
+      "learning_rate": 9.217468529903829e-06,
+      "loss": 0.5268,
+      "step": 10356
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.01090076175788,
+      "learning_rate": 9.215787313369836e-06,
+      "loss": 0.5451,
+      "step": 10357
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.68921401976793,
+      "learning_rate": 9.214106119138388e-06,
+      "loss": 0.6623,
+      "step": 10358
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 34.997092616714966,
+      "learning_rate": 9.212424947257302e-06,
+      "loss": 0.6319,
+      "step": 10359
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.18851892441769,
+      "learning_rate": 9.210743797774392e-06,
+      "loss": 0.6457,
+      "step": 10360
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 41.50507411040728,
+      "learning_rate": 9.209062670737465e-06,
+      "loss": 0.6698,
+      "step": 10361
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 50.00281724145517,
+      "learning_rate": 9.207381566194336e-06,
+      "loss": 0.6314,
+      "step": 10362
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 34.74469612660074,
+      "learning_rate": 9.20570048419281e-06,
+      "loss": 0.5355,
+      "step": 10363
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 37.769190565106086,
+      "learning_rate": 9.2040194247807e-06,
+      "loss": 0.6598,
+      "step": 10364
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 34.202071095659015,
+      "learning_rate": 9.202338388005807e-06,
+      "loss": 0.5916,
+      "step": 10365
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 38.89006147823148,
+      "learning_rate": 9.200657373915946e-06,
+      "loss": 0.5585,
+      "step": 10366
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 46.67831694949196,
+      "learning_rate": 9.198976382558924e-06,
+      "loss": 0.6054,
+      "step": 10367
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 27.25273848256151,
+      "learning_rate": 9.19729541398254e-06,
+      "loss": 0.6154,
+      "step": 10368
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.080125311264744,
+      "learning_rate": 9.19561446823461e-06,
+      "loss": 0.5632,
+      "step": 10369
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.31362790444585,
+      "learning_rate": 9.19393354536293e-06,
+      "loss": 0.6376,
+      "step": 10370
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.775509501398666,
+      "learning_rate": 9.19225264541531e-06,
+      "loss": 0.7064,
+      "step": 10371
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 52.575098943141064,
+      "learning_rate": 9.190571768439554e-06,
+      "loss": 0.636,
+      "step": 10372
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 39.352375893803966,
+      "learning_rate": 9.188890914483462e-06,
+      "loss": 0.6484,
+      "step": 10373
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 37.55111879899226,
+      "learning_rate": 9.187210083594835e-06,
+      "loss": 0.5609,
+      "step": 10374
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.28965150263327,
+      "learning_rate": 9.185529275821481e-06,
+      "loss": 0.6538,
+      "step": 10375
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 43.710422308973435,
+      "learning_rate": 9.183848491211196e-06,
+      "loss": 0.6629,
+      "step": 10376
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.6851326579808,
+      "learning_rate": 9.182167729811788e-06,
+      "loss": 0.6406,
+      "step": 10377
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.51413587627695,
+      "learning_rate": 9.180486991671046e-06,
+      "loss": 0.5429,
+      "step": 10378
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 41.487217023090736,
+      "learning_rate": 9.178806276836777e-06,
+      "loss": 0.6969,
+      "step": 10379
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.85991149368044,
+      "learning_rate": 9.177125585356773e-06,
+      "loss": 0.534,
+      "step": 10380
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.25268254178865,
+      "learning_rate": 9.175444917278839e-06,
+      "loss": 0.5246,
+      "step": 10381
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 34.71667086936521,
+      "learning_rate": 9.173764272650771e-06,
+      "loss": 0.6146,
+      "step": 10382
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.329510309013834,
+      "learning_rate": 9.172083651520362e-06,
+      "loss": 0.5775,
+      "step": 10383
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.40828061314284,
+      "learning_rate": 9.170403053935408e-06,
+      "loss": 0.6388,
+      "step": 10384
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 29.595966699192232,
+      "learning_rate": 9.168722479943707e-06,
+      "loss": 0.5528,
+      "step": 10385
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.06348944418981,
+      "learning_rate": 9.167041929593051e-06,
+      "loss": 0.5192,
+      "step": 10386
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.38713697545732,
+      "learning_rate": 9.16536140293124e-06,
+      "loss": 0.6184,
+      "step": 10387
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 32.91259952157777,
+      "learning_rate": 9.16368090000606e-06,
+      "loss": 0.5756,
+      "step": 10388
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.820805365624594,
+      "learning_rate": 9.162000420865303e-06,
+      "loss": 0.6013,
+      "step": 10389
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.8308082425873,
+      "learning_rate": 9.160319965556765e-06,
+      "loss": 0.6747,
+      "step": 10390
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.98503396147665,
+      "learning_rate": 9.158639534128239e-06,
+      "loss": 0.6592,
+      "step": 10391
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 44.25570930675382,
+      "learning_rate": 9.156959126627512e-06,
+      "loss": 0.6348,
+      "step": 10392
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.48874664581577,
+      "learning_rate": 9.155278743102372e-06,
+      "loss": 0.6457,
+      "step": 10393
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 31.84958106383749,
+      "learning_rate": 9.15359838360061e-06,
+      "loss": 0.5198,
+      "step": 10394
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.833085306920864,
+      "learning_rate": 9.151918048170015e-06,
+      "loss": 0.6105,
+      "step": 10395
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 48.00504155306095,
+      "learning_rate": 9.150237736858376e-06,
+      "loss": 0.6766,
+      "step": 10396
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.83035968332778,
+      "learning_rate": 9.148557449713481e-06,
+      "loss": 0.5417,
+      "step": 10397
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.42932082742466,
+      "learning_rate": 9.146877186783112e-06,
+      "loss": 0.5381,
+      "step": 10398
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 38.678797296132004,
+      "learning_rate": 9.145196948115053e-06,
+      "loss": 0.6076,
+      "step": 10399
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 38.844187414111,
+      "learning_rate": 9.143516733757097e-06,
+      "loss": 0.6969,
+      "step": 10400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 34.07667461109028,
+      "learning_rate": 9.141836543757025e-06,
+      "loss": 0.5289,
+      "step": 10401
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 34.453579411353694,
+      "learning_rate": 9.140156378162619e-06,
+      "loss": 0.6085,
+      "step": 10402
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 36.126741116550214,
+      "learning_rate": 9.138476237021659e-06,
+      "loss": 0.5775,
+      "step": 10403
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 41.01850366572205,
+      "learning_rate": 9.136796120381933e-06,
+      "loss": 0.5907,
+      "step": 10404
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.492757580489865,
+      "learning_rate": 9.13511602829122e-06,
+      "loss": 0.6479,
+      "step": 10405
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 36.132785129846056,
+      "learning_rate": 9.133435960797304e-06,
+      "loss": 0.6448,
+      "step": 10406
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 30.065526100172864,
+      "learning_rate": 9.131755917947958e-06,
+      "loss": 0.5283,
+      "step": 10407
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.033378338414565,
+      "learning_rate": 9.13007589979097e-06,
+      "loss": 0.6037,
+      "step": 10408
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 36.930740438274846,
+      "learning_rate": 9.12839590637411e-06,
+      "loss": 0.6646,
+      "step": 10409
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.201536678643556,
+      "learning_rate": 9.126715937745163e-06,
+      "loss": 0.5584,
+      "step": 10410
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 32.43919002121482,
+      "learning_rate": 9.125035993951907e-06,
+      "loss": 0.5656,
+      "step": 10411
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 37.71363224432569,
+      "learning_rate": 9.12335607504211e-06,
+      "loss": 0.5041,
+      "step": 10412
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.294926839227905,
+      "learning_rate": 9.121676181063556e-06,
+      "loss": 0.5393,
+      "step": 10413
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 34.219473934789654,
+      "learning_rate": 9.119996312064014e-06,
+      "loss": 0.5273,
+      "step": 10414
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 34.72095413490655,
+      "learning_rate": 9.118316468091267e-06,
+      "loss": 0.5152,
+      "step": 10415
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.25146714993283,
+      "learning_rate": 9.116636649193085e-06,
+      "loss": 0.6075,
+      "step": 10416
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.677272214300395,
+      "learning_rate": 9.114956855417236e-06,
+      "loss": 0.6456,
+      "step": 10417
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 32.22857074560606,
+      "learning_rate": 9.113277086811495e-06,
+      "loss": 0.5786,
+      "step": 10418
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 34.545577015813684,
+      "learning_rate": 9.11159734342364e-06,
+      "loss": 0.6288,
+      "step": 10419
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.212780758839045,
+      "learning_rate": 9.109917625301432e-06,
+      "loss": 0.6476,
+      "step": 10420
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 37.81768525250776,
+      "learning_rate": 9.108237932492652e-06,
+      "loss": 0.6035,
+      "step": 10421
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 36.706447153781454,
+      "learning_rate": 9.106558265045058e-06,
+      "loss": 0.6111,
+      "step": 10422
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 34.51006788397441,
+      "learning_rate": 9.104878623006427e-06,
+      "loss": 0.613,
+      "step": 10423
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 43.17410521626509,
+      "learning_rate": 9.103199006424521e-06,
+      "loss": 0.5785,
+      "step": 10424
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 55.57836131125277,
+      "learning_rate": 9.101519415347113e-06,
+      "loss": 0.6544,
+      "step": 10425
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.78740188485865,
+      "learning_rate": 9.099839849821971e-06,
+      "loss": 0.5942,
+      "step": 10426
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.67125281416012,
+      "learning_rate": 9.098160309896852e-06,
+      "loss": 0.5921,
+      "step": 10427
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 43.05171235142597,
+      "learning_rate": 9.096480795619527e-06,
+      "loss": 0.5719,
+      "step": 10428
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.55781064558426,
+      "learning_rate": 9.09480130703776e-06,
+      "loss": 0.6096,
+      "step": 10429
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 34.77779446665113,
+      "learning_rate": 9.093121844199312e-06,
+      "loss": 0.6234,
+      "step": 10430
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 36.90606200051613,
+      "learning_rate": 9.091442407151955e-06,
+      "loss": 0.6471,
+      "step": 10431
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.14641337402511,
+      "learning_rate": 9.08976299594344e-06,
+      "loss": 0.609,
+      "step": 10432
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 30.52557207019813,
+      "learning_rate": 9.088083610621532e-06,
+      "loss": 0.58,
+      "step": 10433
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.05254426353295,
+      "learning_rate": 9.086404251233993e-06,
+      "loss": 0.602,
+      "step": 10434
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 42.248550669511026,
+      "learning_rate": 9.084724917828585e-06,
+      "loss": 0.6915,
+      "step": 10435
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 43.77833445042322,
+      "learning_rate": 9.083045610453065e-06,
+      "loss": 0.6268,
+      "step": 10436
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.10720480309282,
+      "learning_rate": 9.081366329155188e-06,
+      "loss": 0.5606,
+      "step": 10437
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.23467145086987,
+      "learning_rate": 9.079687073982719e-06,
+      "loss": 0.6584,
+      "step": 10438
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 42.27634422335284,
+      "learning_rate": 9.078007844983406e-06,
+      "loss": 0.5726,
+      "step": 10439
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.31430770691639,
+      "learning_rate": 9.076328642205015e-06,
+      "loss": 0.6236,
+      "step": 10440
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 41.77619056625981,
+      "learning_rate": 9.074649465695298e-06,
+      "loss": 0.6046,
+      "step": 10441
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 30.101344007050074,
+      "learning_rate": 9.072970315502008e-06,
+      "loss": 0.5603,
+      "step": 10442
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.12557970450502,
+      "learning_rate": 9.071291191672896e-06,
+      "loss": 0.5454,
+      "step": 10443
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 46.95128838474562,
+      "learning_rate": 9.069612094255722e-06,
+      "loss": 0.6737,
+      "step": 10444
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 37.84274300732548,
+      "learning_rate": 9.067933023298234e-06,
+      "loss": 0.5844,
+      "step": 10445
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 37.45519579341619,
+      "learning_rate": 9.066253978848192e-06,
+      "loss": 0.598,
+      "step": 10446
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 42.69728114274103,
+      "learning_rate": 9.064574960953334e-06,
+      "loss": 0.5666,
+      "step": 10447
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.37450408304469,
+      "learning_rate": 9.06289596966142e-06,
+      "loss": 0.5637,
+      "step": 10448
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.14269163725214,
+      "learning_rate": 9.061217005020194e-06,
+      "loss": 0.6406,
+      "step": 10449
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 36.858194705036325,
+      "learning_rate": 9.05953806707741e-06,
+      "loss": 0.5943,
+      "step": 10450
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.857266821630695,
+      "learning_rate": 9.057859155880811e-06,
+      "loss": 0.5495,
+      "step": 10451
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.59741717233743,
+      "learning_rate": 9.05618027147815e-06,
+      "loss": 0.6244,
+      "step": 10452
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 36.11910479203509,
+      "learning_rate": 9.054501413917163e-06,
+      "loss": 0.5753,
+      "step": 10453
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 41.52575338370356,
+      "learning_rate": 9.052822583245608e-06,
+      "loss": 0.6293,
+      "step": 10454
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.23516798997152,
+      "learning_rate": 9.051143779511226e-06,
+      "loss": 0.6484,
+      "step": 10455
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 32.945187939993446,
+      "learning_rate": 9.049465002761756e-06,
+      "loss": 0.4783,
+      "step": 10456
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.4101545744094,
+      "learning_rate": 9.047786253044945e-06,
+      "loss": 0.6653,
+      "step": 10457
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 31.28452477857969,
+      "learning_rate": 9.046107530408537e-06,
+      "loss": 0.5419,
+      "step": 10458
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.40232977912526,
+      "learning_rate": 9.044428834900274e-06,
+      "loss": 0.6496,
+      "step": 10459
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 38.5331408306668,
+      "learning_rate": 9.042750166567898e-06,
+      "loss": 0.6757,
+      "step": 10460
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.80616179562087,
+      "learning_rate": 9.041071525459145e-06,
+      "loss": 0.5478,
+      "step": 10461
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 31.08779788532194,
+      "learning_rate": 9.039392911621754e-06,
+      "loss": 0.5896,
+      "step": 10462
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.93882647775504,
+      "learning_rate": 9.03771432510347e-06,
+      "loss": 0.7419,
+      "step": 10463
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 55.4550745651711,
+      "learning_rate": 9.036035765952023e-06,
+      "loss": 0.6432,
+      "step": 10464
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 33.659297068453164,
+      "learning_rate": 9.034357234215164e-06,
+      "loss": 0.6452,
+      "step": 10465
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 37.80628141861309,
+      "learning_rate": 9.032678729940614e-06,
+      "loss": 0.618,
+      "step": 10466
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.07819463415213,
+      "learning_rate": 9.031000253176117e-06,
+      "loss": 0.5596,
+      "step": 10467
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.30661423354122,
+      "learning_rate": 9.029321803969402e-06,
+      "loss": 0.5435,
+      "step": 10468
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.74718250647568,
+      "learning_rate": 9.02764338236821e-06,
+      "loss": 0.5795,
+      "step": 10469
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 54.02481813466372,
+      "learning_rate": 9.025964988420274e-06,
+      "loss": 0.6734,
+      "step": 10470
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.51356180841191,
+      "learning_rate": 9.024286622173323e-06,
+      "loss": 0.5507,
+      "step": 10471
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.89551980223082,
+      "learning_rate": 9.022608283675086e-06,
+      "loss": 0.5706,
+      "step": 10472
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 43.06145235179705,
+      "learning_rate": 9.0209299729733e-06,
+      "loss": 0.7404,
+      "step": 10473
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 35.66999108464927,
+      "learning_rate": 9.019251690115692e-06,
+      "loss": 0.5875,
+      "step": 10474
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 30.877738943119784,
+      "learning_rate": 9.017573435149994e-06,
+      "loss": 0.5708,
+      "step": 10475
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 46.98087194467495,
+      "learning_rate": 9.015895208123933e-06,
+      "loss": 0.6229,
+      "step": 10476
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.11132600152146,
+      "learning_rate": 9.014217009085232e-06,
+      "loss": 0.5807,
+      "step": 10477
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.350352719113545,
+      "learning_rate": 9.012538838081626e-06,
+      "loss": 0.6391,
+      "step": 10478
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.09917108249831,
+      "learning_rate": 9.010860695160839e-06,
+      "loss": 0.5552,
+      "step": 10479
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.6468680031473,
+      "learning_rate": 9.009182580370596e-06,
+      "loss": 0.5849,
+      "step": 10480
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.739316541233975,
+      "learning_rate": 9.007504493758616e-06,
+      "loss": 0.591,
+      "step": 10481
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.99238000301351,
+      "learning_rate": 9.005826435372631e-06,
+      "loss": 0.6004,
+      "step": 10482
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 36.70634032903619,
+      "learning_rate": 9.004148405260358e-06,
+      "loss": 0.6193,
+      "step": 10483
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.83985656402726,
+      "learning_rate": 9.002470403469525e-06,
+      "loss": 0.5601,
+      "step": 10484
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.14741199290018,
+      "learning_rate": 9.000792430047852e-06,
+      "loss": 0.5789,
+      "step": 10485
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 31.740599331287196,
+      "learning_rate": 8.999114485043053e-06,
+      "loss": 0.5649,
+      "step": 10486
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 40.07984837922144,
+      "learning_rate": 8.997436568502852e-06,
+      "loss": 0.608,
+      "step": 10487
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 41.666707911788876,
+      "learning_rate": 8.995758680474972e-06,
+      "loss": 0.6201,
+      "step": 10488
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.38674341977579,
+      "learning_rate": 8.994080821007124e-06,
+      "loss": 0.5504,
+      "step": 10489
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.7561354332229,
+      "learning_rate": 8.992402990147035e-06,
+      "loss": 0.6058,
+      "step": 10490
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 40.7849120414181,
+      "learning_rate": 8.99072518794241e-06,
+      "loss": 0.5622,
+      "step": 10491
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.35988277868374,
+      "learning_rate": 8.989047414440973e-06,
+      "loss": 0.6088,
+      "step": 10492
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.24722790178521,
+      "learning_rate": 8.987369669690433e-06,
+      "loss": 0.6461,
+      "step": 10493
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.38614831787595,
+      "learning_rate": 8.98569195373851e-06,
+      "loss": 0.5611,
+      "step": 10494
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.80462310755859,
+      "learning_rate": 8.984014266632916e-06,
+      "loss": 0.5969,
+      "step": 10495
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.91524523244538,
+      "learning_rate": 8.98233660842136e-06,
+      "loss": 0.5782,
+      "step": 10496
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 36.6169925926235,
+      "learning_rate": 8.980658979151556e-06,
+      "loss": 0.5861,
+      "step": 10497
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 35.676103138325146,
+      "learning_rate": 8.978981378871213e-06,
+      "loss": 0.6445,
+      "step": 10498
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 35.20239676963809,
+      "learning_rate": 8.977303807628046e-06,
+      "loss": 0.5323,
+      "step": 10499
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.473649551713535,
+      "learning_rate": 8.975626265469755e-06,
+      "loss": 0.5627,
+      "step": 10500
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 42.8135631512808,
+      "learning_rate": 8.973948752444057e-06,
+      "loss": 0.5624,
+      "step": 10501
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 43.66925420659751,
+      "learning_rate": 8.972271268598653e-06,
+      "loss": 0.6194,
+      "step": 10502
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.04377029611381,
+      "learning_rate": 8.970593813981257e-06,
+      "loss": 0.6015,
+      "step": 10503
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.03751996076784,
+      "learning_rate": 8.96891638863957e-06,
+      "loss": 0.5885,
+      "step": 10504
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 36.12561148155671,
+      "learning_rate": 8.967238992621297e-06,
+      "loss": 0.6117,
+      "step": 10505
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.06268721238918,
+      "learning_rate": 8.965561625974142e-06,
+      "loss": 0.593,
+      "step": 10506
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 31.883741605008815,
+      "learning_rate": 8.96388428874581e-06,
+      "loss": 0.5448,
+      "step": 10507
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 49.54845749064085,
+      "learning_rate": 8.962206980984001e-06,
+      "loss": 0.5776,
+      "step": 10508
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 29.74667074103508,
+      "learning_rate": 8.960529702736425e-06,
+      "loss": 0.5222,
+      "step": 10509
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 33.56589962880947,
+      "learning_rate": 8.95885245405077e-06,
+      "loss": 0.6011,
+      "step": 10510
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.783659968893495,
+      "learning_rate": 8.957175234974744e-06,
+      "loss": 0.6359,
+      "step": 10511
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 36.13017996770919,
+      "learning_rate": 8.955498045556041e-06,
+      "loss": 0.5851,
+      "step": 10512
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 40.869994040708804,
+      "learning_rate": 8.953820885842366e-06,
+      "loss": 0.6321,
+      "step": 10513
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 44.89684060361539,
+      "learning_rate": 8.952143755881416e-06,
+      "loss": 0.6638,
+      "step": 10514
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 40.99669945620611,
+      "learning_rate": 8.95046665572088e-06,
+      "loss": 0.6604,
+      "step": 10515
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.98130067695279,
+      "learning_rate": 8.948789585408457e-06,
+      "loss": 0.5853,
+      "step": 10516
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.431911453404595,
+      "learning_rate": 8.947112544991846e-06,
+      "loss": 0.5582,
+      "step": 10517
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 35.9957481429781,
+      "learning_rate": 8.945435534518737e-06,
+      "loss": 0.5755,
+      "step": 10518
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.25169631912723,
+      "learning_rate": 8.943758554036828e-06,
+      "loss": 0.5831,
+      "step": 10519
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 38.390009075932824,
+      "learning_rate": 8.942081603593805e-06,
+      "loss": 0.5919,
+      "step": 10520
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 41.958045969221786,
+      "learning_rate": 8.94040468323736e-06,
+      "loss": 0.5832,
+      "step": 10521
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 29.235250642386127,
+      "learning_rate": 8.93872779301519e-06,
+      "loss": 0.4969,
+      "step": 10522
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 41.34688734980305,
+      "learning_rate": 8.93705093297498e-06,
+      "loss": 0.5545,
+      "step": 10523
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 35.494275823675814,
+      "learning_rate": 8.93537410316442e-06,
+      "loss": 0.6088,
+      "step": 10524
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 42.63019362884187,
+      "learning_rate": 8.933697303631195e-06,
+      "loss": 0.6167,
+      "step": 10525
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.970793240795125,
+      "learning_rate": 8.932020534422997e-06,
+      "loss": 0.5055,
+      "step": 10526
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.19377458023794,
+      "learning_rate": 8.930343795587508e-06,
+      "loss": 0.5796,
+      "step": 10527
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 35.268989412643336,
+      "learning_rate": 8.928667087172417e-06,
+      "loss": 0.5321,
+      "step": 10528
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.910200568855714,
+      "learning_rate": 8.92699040922541e-06,
+      "loss": 0.6419,
+      "step": 10529
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 36.39046818923414,
+      "learning_rate": 8.925313761794166e-06,
+      "loss": 0.6245,
+      "step": 10530
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 37.69468982241538,
+      "learning_rate": 8.92363714492637e-06,
+      "loss": 0.619,
+      "step": 10531
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 40.99834748461817,
+      "learning_rate": 8.921960558669704e-06,
+      "loss": 0.5966,
+      "step": 10532
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.38180434449871,
+      "learning_rate": 8.920284003071846e-06,
+      "loss": 0.6888,
+      "step": 10533
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.45062641470044,
+      "learning_rate": 8.918607478180487e-06,
+      "loss": 0.5505,
+      "step": 10534
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 41.51937930784275,
+      "learning_rate": 8.916930984043291e-06,
+      "loss": 0.6002,
+      "step": 10535
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.28163398082993,
+      "learning_rate": 8.915254520707948e-06,
+      "loss": 0.5791,
+      "step": 10536
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 61.63421423468038,
+      "learning_rate": 8.913578088222129e-06,
+      "loss": 0.651,
+      "step": 10537
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 42.476173953380076,
+      "learning_rate": 8.911901686633516e-06,
+      "loss": 0.6204,
+      "step": 10538
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.84351508709789,
+      "learning_rate": 8.910225315989786e-06,
+      "loss": 0.5571,
+      "step": 10539
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 40.44376366346397,
+      "learning_rate": 8.908548976338604e-06,
+      "loss": 0.6349,
+      "step": 10540
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.86107011039488,
+      "learning_rate": 8.906872667727652e-06,
+      "loss": 0.5555,
+      "step": 10541
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 40.39761662723522,
+      "learning_rate": 8.905196390204604e-06,
+      "loss": 0.6216,
+      "step": 10542
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 40.67665545067157,
+      "learning_rate": 8.903520143817128e-06,
+      "loss": 0.7146,
+      "step": 10543
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 46.22691245282627,
+      "learning_rate": 8.901843928612902e-06,
+      "loss": 0.6672,
+      "step": 10544
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 43.50401792112015,
+      "learning_rate": 8.90016774463959e-06,
+      "loss": 0.596,
+      "step": 10545
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.59520164356273,
+      "learning_rate": 8.89849159194486e-06,
+      "loss": 0.5484,
+      "step": 10546
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 36.638450307505046,
+      "learning_rate": 8.89681547057639e-06,
+      "loss": 0.6322,
+      "step": 10547
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.51759604117168,
+      "learning_rate": 8.895139380581843e-06,
+      "loss": 0.7084,
+      "step": 10548
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 31.955179617563044,
+      "learning_rate": 8.893463322008885e-06,
+      "loss": 0.5228,
+      "step": 10549
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.766575174098506,
+      "learning_rate": 8.891787294905178e-06,
+      "loss": 0.621,
+      "step": 10550
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.74231910208844,
+      "learning_rate": 8.890111299318396e-06,
+      "loss": 0.5326,
+      "step": 10551
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.37509611636509,
+      "learning_rate": 8.888435335296197e-06,
+      "loss": 0.5382,
+      "step": 10552
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.85525842143553,
+      "learning_rate": 8.886759402886253e-06,
+      "loss": 0.6648,
+      "step": 10553
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 42.45252360738259,
+      "learning_rate": 8.885083502136214e-06,
+      "loss": 0.6652,
+      "step": 10554
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 39.88756294914976,
+      "learning_rate": 8.883407633093751e-06,
+      "loss": 0.6256,
+      "step": 10555
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 36.342499397389616,
+      "learning_rate": 8.881731795806517e-06,
+      "loss": 0.6211,
+      "step": 10556
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.3820933114962,
+      "learning_rate": 8.88005599032218e-06,
+      "loss": 0.5835,
+      "step": 10557
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.61024008275582,
+      "learning_rate": 8.878380216688399e-06,
+      "loss": 0.555,
+      "step": 10558
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 30.5278116240375,
+      "learning_rate": 8.876704474952826e-06,
+      "loss": 0.5624,
+      "step": 10559
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 41.206205189982995,
+      "learning_rate": 8.875028765163117e-06,
+      "loss": 0.5836,
+      "step": 10560
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.726194625070704,
+      "learning_rate": 8.873353087366936e-06,
+      "loss": 0.728,
+      "step": 10561
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 43.95349491965873,
+      "learning_rate": 8.87167744161193e-06,
+      "loss": 0.5834,
+      "step": 10562
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.80070857273102,
+      "learning_rate": 8.870001827945764e-06,
+      "loss": 0.5389,
+      "step": 10563
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.849199004114986,
+      "learning_rate": 8.86832624641608e-06,
+      "loss": 0.6313,
+      "step": 10564
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.16959086364392,
+      "learning_rate": 8.866650697070535e-06,
+      "loss": 0.5935,
+      "step": 10565
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 42.45567157451857,
+      "learning_rate": 8.864975179956784e-06,
+      "loss": 0.5776,
+      "step": 10566
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.81644025899392,
+      "learning_rate": 8.863299695122473e-06,
+      "loss": 0.5361,
+      "step": 10567
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 39.440966304181536,
+      "learning_rate": 8.861624242615258e-06,
+      "loss": 0.6446,
+      "step": 10568
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 42.00305485968945,
+      "learning_rate": 8.859948822482777e-06,
+      "loss": 0.5944,
+      "step": 10569
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.64286267086883,
+      "learning_rate": 8.85827343477269e-06,
+      "loss": 0.5847,
+      "step": 10570
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 42.088430678860874,
+      "learning_rate": 8.856598079532633e-06,
+      "loss": 0.5972,
+      "step": 10571
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 37.87426204403122,
+      "learning_rate": 8.854922756810263e-06,
+      "loss": 0.5769,
+      "step": 10572
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.72551078361017,
+      "learning_rate": 8.85324746665322e-06,
+      "loss": 0.5686,
+      "step": 10573
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 41.94609025336099,
+      "learning_rate": 8.851572209109147e-06,
+      "loss": 0.6059,
+      "step": 10574
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.51787267001141,
+      "learning_rate": 8.849896984225685e-06,
+      "loss": 0.6037,
+      "step": 10575
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.42039564774421,
+      "learning_rate": 8.848221792050484e-06,
+      "loss": 0.5332,
+      "step": 10576
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 36.201078529124366,
+      "learning_rate": 8.846546632631178e-06,
+      "loss": 0.535,
+      "step": 10577
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.792281651261,
+      "learning_rate": 8.844871506015417e-06,
+      "loss": 0.6102,
+      "step": 10578
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 39.43017183584201,
+      "learning_rate": 8.84319641225083e-06,
+      "loss": 0.6808,
+      "step": 10579
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.7382267225948,
+      "learning_rate": 8.841521351385061e-06,
+      "loss": 0.6237,
+      "step": 10580
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 30.767046324532096,
+      "learning_rate": 8.839846323465745e-06,
+      "loss": 0.5241,
+      "step": 10581
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 38.623097956222146,
+      "learning_rate": 8.838171328540524e-06,
+      "loss": 0.7014,
+      "step": 10582
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.63886598345978,
+      "learning_rate": 8.836496366657032e-06,
+      "loss": 0.5748,
+      "step": 10583
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.43319945978123,
+      "learning_rate": 8.834821437862899e-06,
+      "loss": 0.573,
+      "step": 10584
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 39.27873053774795,
+      "learning_rate": 8.833146542205765e-06,
+      "loss": 0.5731,
+      "step": 10585
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 43.647707049346685,
+      "learning_rate": 8.831471679733262e-06,
+      "loss": 0.5877,
+      "step": 10586
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 39.2356111794339,
+      "learning_rate": 8.829796850493016e-06,
+      "loss": 0.4883,
+      "step": 10587
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 36.42178054725317,
+      "learning_rate": 8.828122054532669e-06,
+      "loss": 0.5413,
+      "step": 10588
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 53.19540030575409,
+      "learning_rate": 8.826447291899842e-06,
+      "loss": 0.6531,
+      "step": 10589
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.772793550372626,
+      "learning_rate": 8.824772562642166e-06,
+      "loss": 0.6148,
+      "step": 10590
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 35.09327775564426,
+      "learning_rate": 8.823097866807272e-06,
+      "loss": 0.5877,
+      "step": 10591
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.93996340064063,
+      "learning_rate": 8.82142320444279e-06,
+      "loss": 0.5588,
+      "step": 10592
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.87622641352758,
+      "learning_rate": 8.81974857559634e-06,
+      "loss": 0.5886,
+      "step": 10593
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 40.358767308095494,
+      "learning_rate": 8.818073980315547e-06,
+      "loss": 0.6208,
+      "step": 10594
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.2877230335575,
+      "learning_rate": 8.816399418648041e-06,
+      "loss": 0.5631,
+      "step": 10595
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 25.802720842941902,
+      "learning_rate": 8.814724890641444e-06,
+      "loss": 0.5199,
+      "step": 10596
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 34.56402196516622,
+      "learning_rate": 8.81305039634338e-06,
+      "loss": 0.5999,
+      "step": 10597
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 32.74401456431474,
+      "learning_rate": 8.811375935801463e-06,
+      "loss": 0.5327,
+      "step": 10598
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 29.95901704723383,
+      "learning_rate": 8.809701509063322e-06,
+      "loss": 0.4904,
+      "step": 10599
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.28185096736107,
+      "learning_rate": 8.808027116176572e-06,
+      "loss": 0.6044,
+      "step": 10600
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.02991310847966,
+      "learning_rate": 8.806352757188835e-06,
+      "loss": 0.5128,
+      "step": 10601
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.81028301581049,
+      "learning_rate": 8.80467843214773e-06,
+      "loss": 0.6023,
+      "step": 10602
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 39.85142772526883,
+      "learning_rate": 8.803004141100869e-06,
+      "loss": 0.6236,
+      "step": 10603
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 41.837882629824016,
+      "learning_rate": 8.801329884095866e-06,
+      "loss": 0.5573,
+      "step": 10604
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 48.9729516783493,
+      "learning_rate": 8.799655661180344e-06,
+      "loss": 0.6613,
+      "step": 10605
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 40.03978723977274,
+      "learning_rate": 8.79798147240191e-06,
+      "loss": 0.6558,
+      "step": 10606
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 32.31631719447163,
+      "learning_rate": 8.796307317808183e-06,
+      "loss": 0.4607,
+      "step": 10607
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 44.59419316564146,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.6633,
+      "step": 10608
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.16912191489403,
+      "learning_rate": 8.792959111365285e-06,
+      "loss": 0.5817,
+      "step": 10609
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.204399526041584,
+      "learning_rate": 8.791285059611335e-06,
+      "loss": 0.6238,
+      "step": 10610
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.01834149159615,
+      "learning_rate": 8.789611042232534e-06,
+      "loss": 0.573,
+      "step": 10611
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.00865125682779,
+      "learning_rate": 8.787937059276486e-06,
+      "loss": 0.5239,
+      "step": 10612
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.924929350034446,
+      "learning_rate": 8.786263110790796e-06,
+      "loss": 0.536,
+      "step": 10613
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 43.43909604001399,
+      "learning_rate": 8.784589196823075e-06,
+      "loss": 0.5496,
+      "step": 10614
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 49.350319461890024,
+      "learning_rate": 8.782915317420924e-06,
+      "loss": 0.6562,
+      "step": 10615
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 43.11416845989164,
+      "learning_rate": 8.781241472631952e-06,
+      "loss": 0.5685,
+      "step": 10616
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 34.40119515826827,
+      "learning_rate": 8.779567662503761e-06,
+      "loss": 0.5198,
+      "step": 10617
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.75067372068256,
+      "learning_rate": 8.77789388708395e-06,
+      "loss": 0.5333,
+      "step": 10618
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.50530358111001,
+      "learning_rate": 8.77622014642012e-06,
+      "loss": 0.4868,
+      "step": 10619
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.56807811087383,
+      "learning_rate": 8.774546440559874e-06,
+      "loss": 0.5722,
+      "step": 10620
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.64443906477889,
+      "learning_rate": 8.77287276955081e-06,
+      "loss": 0.5965,
+      "step": 10621
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.792403771566306,
+      "learning_rate": 8.771199133440533e-06,
+      "loss": 0.6187,
+      "step": 10622
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 46.18695623467469,
+      "learning_rate": 8.769525532276627e-06,
+      "loss": 0.5993,
+      "step": 10623
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.16803064206088,
+      "learning_rate": 8.767851966106696e-06,
+      "loss": 0.6008,
+      "step": 10624
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 40.91614261895016,
+      "learning_rate": 8.766178434978334e-06,
+      "loss": 0.6166,
+      "step": 10625
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 42.79224857088199,
+      "learning_rate": 8.764504938939137e-06,
+      "loss": 0.6054,
+      "step": 10626
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 33.28965373432971,
+      "learning_rate": 8.762831478036698e-06,
+      "loss": 0.6419,
+      "step": 10627
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 34.9753951064691,
+      "learning_rate": 8.761158052318605e-06,
+      "loss": 0.5536,
+      "step": 10628
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.06201188733942,
+      "learning_rate": 8.759484661832454e-06,
+      "loss": 0.6095,
+      "step": 10629
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 39.30680743183259,
+      "learning_rate": 8.757811306625833e-06,
+      "loss": 0.5514,
+      "step": 10630
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 39.24879444603454,
+      "learning_rate": 8.75613798674633e-06,
+      "loss": 0.5472,
+      "step": 10631
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 39.870664169495356,
+      "learning_rate": 8.754464702241539e-06,
+      "loss": 0.6245,
+      "step": 10632
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 40.696222615514436,
+      "learning_rate": 8.752791453159041e-06,
+      "loss": 0.6786,
+      "step": 10633
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 59.35063752134867,
+      "learning_rate": 8.751118239546424e-06,
+      "loss": 0.5992,
+      "step": 10634
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 29.491062782516984,
+      "learning_rate": 8.749445061451274e-06,
+      "loss": 0.5143,
+      "step": 10635
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 40.7751582919708,
+      "learning_rate": 8.747771918921176e-06,
+      "loss": 0.6261,
+      "step": 10636
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 55.85967814756711,
+      "learning_rate": 8.746098812003714e-06,
+      "loss": 0.5656,
+      "step": 10637
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 33.32989041588248,
+      "learning_rate": 8.744425740746464e-06,
+      "loss": 0.6461,
+      "step": 10638
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 31.4663081580755,
+      "learning_rate": 8.742752705197013e-06,
+      "loss": 0.5376,
+      "step": 10639
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.44959496690385,
+      "learning_rate": 8.741079705402939e-06,
+      "loss": 0.5922,
+      "step": 10640
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 44.61595811253935,
+      "learning_rate": 8.739406741411826e-06,
+      "loss": 0.6325,
+      "step": 10641
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 32.9184651358327,
+      "learning_rate": 8.737733813271244e-06,
+      "loss": 0.5227,
+      "step": 10642
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.66283860180124,
+      "learning_rate": 8.736060921028773e-06,
+      "loss": 0.5589,
+      "step": 10643
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.83282375649418,
+      "learning_rate": 8.73438806473199e-06,
+      "loss": 0.553,
+      "step": 10644
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.339014188792135,
+      "learning_rate": 8.732715244428472e-06,
+      "loss": 0.6211,
+      "step": 10645
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 65.2835488286705,
+      "learning_rate": 8.731042460165793e-06,
+      "loss": 0.6294,
+      "step": 10646
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 27.96722425422915,
+      "learning_rate": 8.72936971199152e-06,
+      "loss": 0.5595,
+      "step": 10647
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.61129719523925,
+      "learning_rate": 8.72769699995323e-06,
+      "loss": 0.591,
+      "step": 10648
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.30602315962198,
+      "learning_rate": 8.726024324098494e-06,
+      "loss": 0.5808,
+      "step": 10649
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 41.95076415169603,
+      "learning_rate": 8.724351684474878e-06,
+      "loss": 0.6479,
+      "step": 10650
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.33952800667281,
+      "learning_rate": 8.72267908112996e-06,
+      "loss": 0.5475,
+      "step": 10651
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 34.19018482077964,
+      "learning_rate": 8.721006514111299e-06,
+      "loss": 0.5367,
+      "step": 10652
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.48684063105338,
+      "learning_rate": 8.719333983466462e-06,
+      "loss": 0.5882,
+      "step": 10653
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 45.29552047404352,
+      "learning_rate": 8.71766148924302e-06,
+      "loss": 0.63,
+      "step": 10654
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 41.640641810429884,
+      "learning_rate": 8.715989031488534e-06,
+      "loss": 0.583,
+      "step": 10655
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 37.67575817750152,
+      "learning_rate": 8.71431661025057e-06,
+      "loss": 0.5712,
+      "step": 10656
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 44.308635072336585,
+      "learning_rate": 8.712644225576687e-06,
+      "loss": 0.6214,
+      "step": 10657
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 41.83825033225652,
+      "learning_rate": 8.710971877514452e-06,
+      "loss": 0.7033,
+      "step": 10658
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 34.19389380409264,
+      "learning_rate": 8.709299566111418e-06,
+      "loss": 0.5273,
+      "step": 10659
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 48.69559799739796,
+      "learning_rate": 8.707627291415153e-06,
+      "loss": 0.7458,
+      "step": 10660
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 43.13033680166751,
+      "learning_rate": 8.705955053473215e-06,
+      "loss": 0.6112,
+      "step": 10661
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 44.43479025367913,
+      "learning_rate": 8.704282852333154e-06,
+      "loss": 0.5977,
+      "step": 10662
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 34.290033632207994,
+      "learning_rate": 8.702610688042531e-06,
+      "loss": 0.5678,
+      "step": 10663
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 38.91550053082102,
+      "learning_rate": 8.700938560648901e-06,
+      "loss": 0.6242,
+      "step": 10664
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 46.02363964289609,
+      "learning_rate": 8.699266470199817e-06,
+      "loss": 0.6029,
+      "step": 10665
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 39.39768500425448,
+      "learning_rate": 8.697594416742842e-06,
+      "loss": 0.6984,
+      "step": 10666
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 37.09367088752929,
+      "learning_rate": 8.695922400325512e-06,
+      "loss": 0.606,
+      "step": 10667
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 32.44447162805214,
+      "learning_rate": 8.694250420995387e-06,
+      "loss": 0.5665,
+      "step": 10668
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 35.02560712164251,
+      "learning_rate": 8.692578478800015e-06,
+      "loss": 0.5787,
+      "step": 10669
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 44.840056298486004,
+      "learning_rate": 8.690906573786949e-06,
+      "loss": 0.6658,
+      "step": 10670
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 35.627195742349684,
+      "learning_rate": 8.689234706003735e-06,
+      "loss": 0.6139,
+      "step": 10671
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 44.92844489576,
+      "learning_rate": 8.687562875497915e-06,
+      "loss": 0.6601,
+      "step": 10672
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.005914321497436,
+      "learning_rate": 8.685891082317043e-06,
+      "loss": 0.6513,
+      "step": 10673
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.856573169678885,
+      "learning_rate": 8.684219326508657e-06,
+      "loss": 0.594,
+      "step": 10674
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 32.49622810923857,
+      "learning_rate": 8.682547608120305e-06,
+      "loss": 0.5467,
+      "step": 10675
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 30.716157541869766,
+      "learning_rate": 8.68087592719953e-06,
+      "loss": 0.5787,
+      "step": 10676
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.650404542899494,
+      "learning_rate": 8.679204283793872e-06,
+      "loss": 0.5529,
+      "step": 10677
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 39.50039980782604,
+      "learning_rate": 8.677532677950868e-06,
+      "loss": 0.597,
+      "step": 10678
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 38.05273145528703,
+      "learning_rate": 8.675861109718064e-06,
+      "loss": 0.5496,
+      "step": 10679
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 50.813585342508674,
+      "learning_rate": 8.674189579142993e-06,
+      "loss": 0.6716,
+      "step": 10680
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 30.545507305776248,
+      "learning_rate": 8.672518086273199e-06,
+      "loss": 0.5405,
+      "step": 10681
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 35.32369817898365,
+      "learning_rate": 8.67084663115621e-06,
+      "loss": 0.5478,
+      "step": 10682
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.266170481730995,
+      "learning_rate": 8.669175213839566e-06,
+      "loss": 0.6299,
+      "step": 10683
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.86792201842591,
+      "learning_rate": 8.6675038343708e-06,
+      "loss": 0.5905,
+      "step": 10684
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 38.08600325515157,
+      "learning_rate": 8.665832492797447e-06,
+      "loss": 0.5797,
+      "step": 10685
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.371740390334143,
+      "learning_rate": 8.664161189167039e-06,
+      "loss": 0.5556,
+      "step": 10686
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.23019217578376,
+      "learning_rate": 8.662489923527104e-06,
+      "loss": 0.5583,
+      "step": 10687
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 33.50438425924609,
+      "learning_rate": 8.66081869592517e-06,
+      "loss": 0.5711,
+      "step": 10688
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 37.700424325260016,
+      "learning_rate": 8.659147506408771e-06,
+      "loss": 0.5663,
+      "step": 10689
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.75839064518724,
+      "learning_rate": 8.657476355025436e-06,
+      "loss": 0.6995,
+      "step": 10690
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 41.41708913327948,
+      "learning_rate": 8.655805241822683e-06,
+      "loss": 0.6378,
+      "step": 10691
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.765723386611434,
+      "learning_rate": 8.654134166848045e-06,
+      "loss": 0.6257,
+      "step": 10692
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 34.702814415955,
+      "learning_rate": 8.652463130149044e-06,
+      "loss": 0.644,
+      "step": 10693
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 39.21811136888407,
+      "learning_rate": 8.6507921317732e-06,
+      "loss": 0.6449,
+      "step": 10694
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 49.2190661562499,
+      "learning_rate": 8.649121171768045e-06,
+      "loss": 0.6094,
+      "step": 10695
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 37.89636766785065,
+      "learning_rate": 8.64745025018109e-06,
+      "loss": 0.6808,
+      "step": 10696
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.87843351496321,
+      "learning_rate": 8.645779367059856e-06,
+      "loss": 0.6441,
+      "step": 10697
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 42.71713636653322,
+      "learning_rate": 8.644108522451868e-06,
+      "loss": 0.6599,
+      "step": 10698
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 33.87520136192693,
+      "learning_rate": 8.642437716404639e-06,
+      "loss": 0.5637,
+      "step": 10699
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 35.88961258798459,
+      "learning_rate": 8.64076694896569e-06,
+      "loss": 0.5604,
+      "step": 10700
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 47.308943379209246,
+      "learning_rate": 8.63909622018253e-06,
+      "loss": 0.6327,
+      "step": 10701
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 48.201934720619384,
+      "learning_rate": 8.63742553010268e-06,
+      "loss": 0.6757,
+      "step": 10702
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 34.56523621956711,
+      "learning_rate": 8.635754878773647e-06,
+      "loss": 0.557,
+      "step": 10703
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 30.273338500814692,
+      "learning_rate": 8.63408426624295e-06,
+      "loss": 0.4921,
+      "step": 10704
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 37.25290433710954,
+      "learning_rate": 8.632413692558101e-06,
+      "loss": 0.5839,
+      "step": 10705
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 38.02353915562864,
+      "learning_rate": 8.630743157766603e-06,
+      "loss": 0.5975,
+      "step": 10706
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.42642508502823,
+      "learning_rate": 8.629072661915969e-06,
+      "loss": 0.5571,
+      "step": 10707
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 45.63030519864528,
+      "learning_rate": 8.627402205053705e-06,
+      "loss": 0.6182,
+      "step": 10708
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.21367764768615,
+      "learning_rate": 8.62573178722732e-06,
+      "loss": 0.5701,
+      "step": 10709
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 54.6042268915971,
+      "learning_rate": 8.624061408484323e-06,
+      "loss": 0.6052,
+      "step": 10710
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 38.376075021305034,
+      "learning_rate": 8.62239106887221e-06,
+      "loss": 0.6294,
+      "step": 10711
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.929937221221664,
+      "learning_rate": 8.620720768438491e-06,
+      "loss": 0.5379,
+      "step": 10712
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 38.76084522330595,
+      "learning_rate": 8.619050507230664e-06,
+      "loss": 0.5042,
+      "step": 10713
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 29.253104108361427,
+      "learning_rate": 8.617380285296235e-06,
+      "loss": 0.4614,
+      "step": 10714
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.963283050751855,
+      "learning_rate": 8.615710102682704e-06,
+      "loss": 0.5987,
+      "step": 10715
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.730114976358664,
+      "learning_rate": 8.614039959437563e-06,
+      "loss": 0.5811,
+      "step": 10716
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.68531069161385,
+      "learning_rate": 8.612369855608318e-06,
+      "loss": 0.5734,
+      "step": 10717
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 35.32916552476191,
+      "learning_rate": 8.610699791242463e-06,
+      "loss": 0.5558,
+      "step": 10718
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 41.230303685857635,
+      "learning_rate": 8.609029766387489e-06,
+      "loss": 0.6855,
+      "step": 10719
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.82079168449121,
+      "learning_rate": 8.6073597810909e-06,
+      "loss": 0.5882,
+      "step": 10720
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.72200063723394,
+      "learning_rate": 8.60568983540018e-06,
+      "loss": 0.629,
+      "step": 10721
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 41.279105339391066,
+      "learning_rate": 8.604019929362826e-06,
+      "loss": 0.6387,
+      "step": 10722
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 36.895612612982596,
+      "learning_rate": 8.602350063026332e-06,
+      "loss": 0.6372,
+      "step": 10723
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 35.93888567036037,
+      "learning_rate": 8.600680236438182e-06,
+      "loss": 0.5398,
+      "step": 10724
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 39.3882276266255,
+      "learning_rate": 8.59901044964587e-06,
+      "loss": 0.6128,
+      "step": 10725
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 38.7028949316438,
+      "learning_rate": 8.597340702696879e-06,
+      "loss": 0.5595,
+      "step": 10726
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 42.72504171656251,
+      "learning_rate": 8.595670995638697e-06,
+      "loss": 0.6461,
+      "step": 10727
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.47865950534777,
+      "learning_rate": 8.59400132851881e-06,
+      "loss": 0.5623,
+      "step": 10728
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 38.665817748128646,
+      "learning_rate": 8.592331701384704e-06,
+      "loss": 0.6083,
+      "step": 10729
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 30.40259715894506,
+      "learning_rate": 8.590662114283864e-06,
+      "loss": 0.4938,
+      "step": 10730
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.41438514653484,
+      "learning_rate": 8.588992567263767e-06,
+      "loss": 0.553,
+      "step": 10731
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 40.78328416216253,
+      "learning_rate": 8.587323060371893e-06,
+      "loss": 0.6516,
+      "step": 10732
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 45.57714199801658,
+      "learning_rate": 8.585653593655728e-06,
+      "loss": 0.6924,
+      "step": 10733
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 36.60104726390059,
+      "learning_rate": 8.583984167162744e-06,
+      "loss": 0.5957,
+      "step": 10734
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 39.78190886138431,
+      "learning_rate": 8.582314780940426e-06,
+      "loss": 0.6025,
+      "step": 10735
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.01452734626155,
+      "learning_rate": 8.580645435036246e-06,
+      "loss": 0.5209,
+      "step": 10736
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 40.5290525159813,
+      "learning_rate": 8.578976129497678e-06,
+      "loss": 0.5807,
+      "step": 10737
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.75068769396983,
+      "learning_rate": 8.577306864372196e-06,
+      "loss": 0.6586,
+      "step": 10738
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.50345562476828,
+      "learning_rate": 8.575637639707279e-06,
+      "loss": 0.5861,
+      "step": 10739
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.71221593219409,
+      "learning_rate": 8.573968455550393e-06,
+      "loss": 0.5949,
+      "step": 10740
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.95913149809488,
+      "learning_rate": 8.572299311949005e-06,
+      "loss": 0.6178,
+      "step": 10741
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.882093350813705,
+      "learning_rate": 8.570630208950593e-06,
+      "loss": 0.5637,
+      "step": 10742
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 41.590072939332515,
+      "learning_rate": 8.568961146602623e-06,
+      "loss": 0.6092,
+      "step": 10743
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.383835595768346,
+      "learning_rate": 8.567292124952562e-06,
+      "loss": 0.5501,
+      "step": 10744
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 31.026251410483553,
+      "learning_rate": 8.565623144047869e-06,
+      "loss": 0.4952,
+      "step": 10745
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.84041306288163,
+      "learning_rate": 8.563954203936018e-06,
+      "loss": 0.5568,
+      "step": 10746
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 33.90104348776601,
+      "learning_rate": 8.562285304664468e-06,
+      "loss": 0.5565,
+      "step": 10747
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 40.57316267586176,
+      "learning_rate": 8.560616446280685e-06,
+      "loss": 0.5581,
+      "step": 10748
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 39.43562239576908,
+      "learning_rate": 8.558947628832131e-06,
+      "loss": 0.5713,
+      "step": 10749
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 36.69759074158517,
+      "learning_rate": 8.55727885236626e-06,
+      "loss": 0.6738,
+      "step": 10750
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.174171911279515,
+      "learning_rate": 8.555610116930532e-06,
+      "loss": 0.6249,
+      "step": 10751
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 44.97706497704192,
+      "learning_rate": 8.55394142257241e-06,
+      "loss": 0.5469,
+      "step": 10752
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 45.644238598937804,
+      "learning_rate": 8.552272769339346e-06,
+      "loss": 0.6488,
+      "step": 10753
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 38.476067002877095,
+      "learning_rate": 8.550604157278804e-06,
+      "loss": 0.5488,
+      "step": 10754
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 38.277488480095876,
+      "learning_rate": 8.548935586438226e-06,
+      "loss": 0.5757,
+      "step": 10755
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 43.01715216539427,
+      "learning_rate": 8.547267056865072e-06,
+      "loss": 0.5613,
+      "step": 10756
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 39.43749326533579,
+      "learning_rate": 8.545598568606791e-06,
+      "loss": 0.5041,
+      "step": 10757
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 36.72987494901001,
+      "learning_rate": 8.54393012171084e-06,
+      "loss": 0.5875,
+      "step": 10758
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 39.79773218784144,
+      "learning_rate": 8.542261716224664e-06,
+      "loss": 0.5261,
+      "step": 10759
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 33.93681423976736,
+      "learning_rate": 8.54059335219571e-06,
+      "loss": 0.5068,
+      "step": 10760
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.47612604321007,
+      "learning_rate": 8.538925029671428e-06,
+      "loss": 0.5363,
+      "step": 10761
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.19083193733888,
+      "learning_rate": 8.537256748699264e-06,
+      "loss": 0.5799,
+      "step": 10762
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 42.65545831860882,
+      "learning_rate": 8.53558850932666e-06,
+      "loss": 0.5884,
+      "step": 10763
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.725403278715696,
+      "learning_rate": 8.533920311601068e-06,
+      "loss": 0.5831,
+      "step": 10764
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 36.970832024421675,
+      "learning_rate": 8.53225215556992e-06,
+      "loss": 0.6026,
+      "step": 10765
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 33.72846435175403,
+      "learning_rate": 8.530584041280661e-06,
+      "loss": 0.563,
+      "step": 10766
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 36.88792245377998,
+      "learning_rate": 8.528915968780735e-06,
+      "loss": 0.5444,
+      "step": 10767
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 33.59531859528281,
+      "learning_rate": 8.527247938117577e-06,
+      "loss": 0.6264,
+      "step": 10768
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 30.689779340697164,
+      "learning_rate": 8.525579949338631e-06,
+      "loss": 0.59,
+      "step": 10769
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 35.02060478849949,
+      "learning_rate": 8.523912002491321e-06,
+      "loss": 0.6876,
+      "step": 10770
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 28.270884916970168,
+      "learning_rate": 8.522244097623095e-06,
+      "loss": 0.4894,
+      "step": 10771
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 35.809062945950956,
+      "learning_rate": 8.520576234781379e-06,
+      "loss": 0.5773,
+      "step": 10772
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 48.3664546092139,
+      "learning_rate": 8.518908414013612e-06,
+      "loss": 0.7397,
+      "step": 10773
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.50960977447954,
+      "learning_rate": 8.517240635367225e-06,
+      "loss": 0.5673,
+      "step": 10774
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.60862388166186,
+      "learning_rate": 8.515572898889645e-06,
+      "loss": 0.5499,
+      "step": 10775
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.400164760134885,
+      "learning_rate": 8.513905204628302e-06,
+      "loss": 0.6304,
+      "step": 10776
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 43.03488843851309,
+      "learning_rate": 8.512237552630625e-06,
+      "loss": 0.6476,
+      "step": 10777
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 34.65584320091028,
+      "learning_rate": 8.510569942944043e-06,
+      "loss": 0.5815,
+      "step": 10778
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 31.71756518175871,
+      "learning_rate": 8.508902375615982e-06,
+      "loss": 0.6137,
+      "step": 10779
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 39.57630292732105,
+      "learning_rate": 8.507234850693864e-06,
+      "loss": 0.6091,
+      "step": 10780
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.66602082810033,
+      "learning_rate": 8.505567368225116e-06,
+      "loss": 0.5648,
+      "step": 10781
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 52.32624931539028,
+      "learning_rate": 8.503899928257151e-06,
+      "loss": 0.739,
+      "step": 10782
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 30.206260056654813,
+      "learning_rate": 8.502232530837404e-06,
+      "loss": 0.5488,
+      "step": 10783
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.321742106481125,
+      "learning_rate": 8.500565176013286e-06,
+      "loss": 0.6287,
+      "step": 10784
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 43.120524613731334,
+      "learning_rate": 8.498897863832215e-06,
+      "loss": 0.5221,
+      "step": 10785
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 40.853515333203106,
+      "learning_rate": 8.497230594341612e-06,
+      "loss": 0.5085,
+      "step": 10786
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.85293636761582,
+      "learning_rate": 8.495563367588893e-06,
+      "loss": 0.5606,
+      "step": 10787
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 48.44888027263309,
+      "learning_rate": 8.493896183621474e-06,
+      "loss": 0.6284,
+      "step": 10788
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.349497613758096,
+      "learning_rate": 8.49222904248676e-06,
+      "loss": 0.5911,
+      "step": 10789
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.3751947097242,
+      "learning_rate": 8.490561944232173e-06,
+      "loss": 0.5518,
+      "step": 10790
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 36.93597728153102,
+      "learning_rate": 8.48889488890512e-06,
+      "loss": 0.6006,
+      "step": 10791
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.93538778275995,
+      "learning_rate": 8.487227876553012e-06,
+      "loss": 0.5671,
+      "step": 10792
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 31.92549741995257,
+      "learning_rate": 8.485560907223264e-06,
+      "loss": 0.5007,
+      "step": 10793
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 29.862096908965732,
+      "learning_rate": 8.483893980963274e-06,
+      "loss": 0.5658,
+      "step": 10794
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 51.130256128484724,
+      "learning_rate": 8.48222709782045e-06,
+      "loss": 0.6323,
+      "step": 10795
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 34.152046346302136,
+      "learning_rate": 8.480560257842201e-06,
+      "loss": 0.5386,
+      "step": 10796
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 50.952086268754,
+      "learning_rate": 8.478893461075929e-06,
+      "loss": 0.6671,
+      "step": 10797
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 31.001789314405855,
+      "learning_rate": 8.477226707569042e-06,
+      "loss": 0.519,
+      "step": 10798
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 38.95852569219797,
+      "learning_rate": 8.475559997368932e-06,
+      "loss": 0.5835,
+      "step": 10799
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 29.698219343712697,
+      "learning_rate": 8.473893330523004e-06,
+      "loss": 0.5038,
+      "step": 10800
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 36.784319361442975,
+      "learning_rate": 8.472226707078655e-06,
+      "loss": 0.5997,
+      "step": 10801
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 35.514554834584715,
+      "learning_rate": 8.470560127083288e-06,
+      "loss": 0.5606,
+      "step": 10802
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 47.22912104817148,
+      "learning_rate": 8.468893590584299e-06,
+      "loss": 0.5932,
+      "step": 10803
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 42.44484572116181,
+      "learning_rate": 8.467227097629075e-06,
+      "loss": 0.6264,
+      "step": 10804
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 38.139209644478086,
+      "learning_rate": 8.465560648265017e-06,
+      "loss": 0.5272,
+      "step": 10805
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 28.33758816093256,
+      "learning_rate": 8.463894242539518e-06,
+      "loss": 0.5139,
+      "step": 10806
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 40.993691255597454,
+      "learning_rate": 8.462227880499967e-06,
+      "loss": 0.5786,
+      "step": 10807
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 40.043639776829245,
+      "learning_rate": 8.460561562193758e-06,
+      "loss": 0.503,
+      "step": 10808
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 38.13016731702209,
+      "learning_rate": 8.458895287668277e-06,
+      "loss": 0.6112,
+      "step": 10809
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 36.113576680074125,
+      "learning_rate": 8.457229056970908e-06,
+      "loss": 0.5642,
+      "step": 10810
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 44.17321008878265,
+      "learning_rate": 8.455562870149046e-06,
+      "loss": 0.5755,
+      "step": 10811
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 39.93747016595195,
+      "learning_rate": 8.453896727250072e-06,
+      "loss": 0.5741,
+      "step": 10812
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 39.3134430255193,
+      "learning_rate": 8.452230628321373e-06,
+      "loss": 0.5333,
+      "step": 10813
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 39.008883277625536,
+      "learning_rate": 8.450564573410324e-06,
+      "loss": 0.6358,
+      "step": 10814
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 38.805493820473245,
+      "learning_rate": 8.448898562564316e-06,
+      "loss": 0.567,
+      "step": 10815
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.421812614118146,
+      "learning_rate": 8.447232595830724e-06,
+      "loss": 0.5414,
+      "step": 10816
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 36.36226648784837,
+      "learning_rate": 8.445566673256928e-06,
+      "loss": 0.5202,
+      "step": 10817
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.46171953103004,
+      "learning_rate": 8.443900794890311e-06,
+      "loss": 0.6495,
+      "step": 10818
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 40.83099574709563,
+      "learning_rate": 8.442234960778242e-06,
+      "loss": 0.6002,
+      "step": 10819
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 27.114093156668755,
+      "learning_rate": 8.440569170968098e-06,
+      "loss": 0.5008,
+      "step": 10820
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 36.16285674878134,
+      "learning_rate": 8.438903425507257e-06,
+      "loss": 0.5498,
+      "step": 10821
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 36.83803953158176,
+      "learning_rate": 8.437237724443086e-06,
+      "loss": 0.5522,
+      "step": 10822
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 38.80861272710635,
+      "learning_rate": 8.435572067822964e-06,
+      "loss": 0.5828,
+      "step": 10823
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 41.329639751748104,
+      "learning_rate": 8.433906455694256e-06,
+      "loss": 0.633,
+      "step": 10824
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 49.46742463158815,
+      "learning_rate": 8.432240888104331e-06,
+      "loss": 0.588,
+      "step": 10825
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 34.832934171193514,
+      "learning_rate": 8.430575365100557e-06,
+      "loss": 0.5531,
+      "step": 10826
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 32.603898605284726,
+      "learning_rate": 8.428909886730303e-06,
+      "loss": 0.5339,
+      "step": 10827
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.21587948495506,
+      "learning_rate": 8.427244453040937e-06,
+      "loss": 0.6038,
+      "step": 10828
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.36933500039688,
+      "learning_rate": 8.425579064079811e-06,
+      "loss": 0.587,
+      "step": 10829
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 35.750176429313186,
+      "learning_rate": 8.423913719894301e-06,
+      "loss": 0.5315,
+      "step": 10830
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.57114621539354,
+      "learning_rate": 8.42224842053176e-06,
+      "loss": 0.6733,
+      "step": 10831
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 56.881892501376186,
+      "learning_rate": 8.420583166039551e-06,
+      "loss": 0.6139,
+      "step": 10832
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.69140668166893,
+      "learning_rate": 8.418917956465037e-06,
+      "loss": 0.5301,
+      "step": 10833
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 35.98251310460251,
+      "learning_rate": 8.417252791855566e-06,
+      "loss": 0.5908,
+      "step": 10834
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 29.873063539448605,
+      "learning_rate": 8.4155876722585e-06,
+      "loss": 0.4678,
+      "step": 10835
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 42.75651359014027,
+      "learning_rate": 8.413922597721197e-06,
+      "loss": 0.638,
+      "step": 10836
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 39.91548640126227,
+      "learning_rate": 8.41225756829101e-06,
+      "loss": 0.586,
+      "step": 10837
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 41.63099665591186,
+      "learning_rate": 8.410592584015284e-06,
+      "loss": 0.6433,
+      "step": 10838
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 50.65968695152658,
+      "learning_rate": 8.408927644941373e-06,
+      "loss": 0.5653,
+      "step": 10839
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 34.60886221477814,
+      "learning_rate": 8.407262751116633e-06,
+      "loss": 0.5142,
+      "step": 10840
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 38.277667934439755,
+      "learning_rate": 8.405597902588407e-06,
+      "loss": 0.5615,
+      "step": 10841
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 51.557355330465455,
+      "learning_rate": 8.403933099404047e-06,
+      "loss": 0.6421,
+      "step": 10842
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 52.092048035973335,
+      "learning_rate": 8.402268341610895e-06,
+      "loss": 0.6274,
+      "step": 10843
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.977724156037404,
+      "learning_rate": 8.400603629256294e-06,
+      "loss": 0.6248,
+      "step": 10844
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 41.37099418091482,
+      "learning_rate": 8.39893896238759e-06,
+      "loss": 0.604,
+      "step": 10845
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 40.3903920622552,
+      "learning_rate": 8.397274341052128e-06,
+      "loss": 0.5897,
+      "step": 10846
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.83759352600315,
+      "learning_rate": 8.395609765297249e-06,
+      "loss": 0.5311,
+      "step": 10847
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 34.62418768374019,
+      "learning_rate": 8.393945235170283e-06,
+      "loss": 0.5177,
+      "step": 10848
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.22488407862479,
+      "learning_rate": 8.392280750718577e-06,
+      "loss": 0.5792,
+      "step": 10849
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 35.26871315039725,
+      "learning_rate": 8.390616311989468e-06,
+      "loss": 0.5538,
+      "step": 10850
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.90143016479604,
+      "learning_rate": 8.388951919030287e-06,
+      "loss": 0.5542,
+      "step": 10851
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 43.750003618512686,
+      "learning_rate": 8.387287571888373e-06,
+      "loss": 0.613,
+      "step": 10852
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.46526014987333,
+      "learning_rate": 8.385623270611058e-06,
+      "loss": 0.6452,
+      "step": 10853
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.90019177447567,
+      "learning_rate": 8.383959015245669e-06,
+      "loss": 0.5329,
+      "step": 10854
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.95348659604607,
+      "learning_rate": 8.382294805839543e-06,
+      "loss": 0.5845,
+      "step": 10855
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.94093584974825,
+      "learning_rate": 8.380630642440006e-06,
+      "loss": 0.636,
+      "step": 10856
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.45631768071719,
+      "learning_rate": 8.378966525094388e-06,
+      "loss": 0.5847,
+      "step": 10857
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.826204524555735,
+      "learning_rate": 8.377302453850012e-06,
+      "loss": 0.596,
+      "step": 10858
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.69355605258586,
+      "learning_rate": 8.375638428754204e-06,
+      "loss": 0.6869,
+      "step": 10859
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 30.48424833458141,
+      "learning_rate": 8.373974449854288e-06,
+      "loss": 0.4825,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 35.45534259960388,
+      "learning_rate": 8.37231051719759e-06,
+      "loss": 0.5468,
+      "step": 10861
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 54.07717075514494,
+      "learning_rate": 8.37064663083143e-06,
+      "loss": 0.6277,
+      "step": 10862
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.91559222097539,
+      "learning_rate": 8.368982790803124e-06,
+      "loss": 0.5773,
+      "step": 10863
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.28594117578397,
+      "learning_rate": 8.367318997159992e-06,
+      "loss": 0.5787,
+      "step": 10864
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 34.801271202859475,
+      "learning_rate": 8.365655249949355e-06,
+      "loss": 0.5206,
+      "step": 10865
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 39.37414231350176,
+      "learning_rate": 8.363991549218525e-06,
+      "loss": 0.5869,
+      "step": 10866
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.18515454282033,
+      "learning_rate": 8.362327895014821e-06,
+      "loss": 0.5284,
+      "step": 10867
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 30.52684818190183,
+      "learning_rate": 8.360664287385552e-06,
+      "loss": 0.5131,
+      "step": 10868
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 35.11946519502597,
+      "learning_rate": 8.359000726378033e-06,
+      "loss": 0.57,
+      "step": 10869
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.45001950341812,
+      "learning_rate": 8.35733721203957e-06,
+      "loss": 0.5584,
+      "step": 10870
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 49.83676776352852,
+      "learning_rate": 8.355673744417479e-06,
+      "loss": 0.6096,
+      "step": 10871
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.39480534378971,
+      "learning_rate": 8.354010323559065e-06,
+      "loss": 0.5385,
+      "step": 10872
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.67366498185551,
+      "learning_rate": 8.352346949511632e-06,
+      "loss": 0.5438,
+      "step": 10873
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.2774275245492,
+      "learning_rate": 8.35068362232249e-06,
+      "loss": 0.5251,
+      "step": 10874
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 46.30578621245778,
+      "learning_rate": 8.34902034203894e-06,
+      "loss": 0.6999,
+      "step": 10875
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.950480583283905,
+      "learning_rate": 8.347357108708284e-06,
+      "loss": 0.5578,
+      "step": 10876
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.58274569634516,
+      "learning_rate": 8.345693922377829e-06,
+      "loss": 0.5633,
+      "step": 10877
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 31.809625713103514,
+      "learning_rate": 8.344030783094869e-06,
+      "loss": 0.5472,
+      "step": 10878
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 40.336706827626706,
+      "learning_rate": 8.342367690906702e-06,
+      "loss": 0.5464,
+      "step": 10879
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.689279333557465,
+      "learning_rate": 8.340704645860632e-06,
+      "loss": 0.5578,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.62940952531147,
+      "learning_rate": 8.339041648003952e-06,
+      "loss": 0.6661,
+      "step": 10881
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.052610258806446,
+      "learning_rate": 8.337378697383956e-06,
+      "loss": 0.6895,
+      "step": 10882
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 45.147637920755145,
+      "learning_rate": 8.335715794047933e-06,
+      "loss": 0.6645,
+      "step": 10883
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.589854994523606,
+      "learning_rate": 8.33405293804318e-06,
+      "loss": 0.6411,
+      "step": 10884
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 34.50887860893957,
+      "learning_rate": 8.332390129416988e-06,
+      "loss": 0.6277,
+      "step": 10885
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.80603279736244,
+      "learning_rate": 8.330727368216647e-06,
+      "loss": 0.5852,
+      "step": 10886
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 30.257099094938784,
+      "learning_rate": 8.329064654489442e-06,
+      "loss": 0.5391,
+      "step": 10887
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 31.20075719232451,
+      "learning_rate": 8.327401988282661e-06,
+      "loss": 0.5134,
+      "step": 10888
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.73768350675673,
+      "learning_rate": 8.325739369643588e-06,
+      "loss": 0.6419,
+      "step": 10889
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.31192784059581,
+      "learning_rate": 8.324076798619508e-06,
+      "loss": 0.5579,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 39.17548452027785,
+      "learning_rate": 8.322414275257707e-06,
+      "loss": 0.6179,
+      "step": 10891
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.10323386547899,
+      "learning_rate": 8.320751799605458e-06,
+      "loss": 0.5577,
+      "step": 10892
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 35.51144030076508,
+      "learning_rate": 8.319089371710048e-06,
+      "loss": 0.608,
+      "step": 10893
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.4851365832949,
+      "learning_rate": 8.317426991618751e-06,
+      "loss": 0.5565,
+      "step": 10894
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 42.121082502956355,
+      "learning_rate": 8.315764659378845e-06,
+      "loss": 0.6946,
+      "step": 10895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 42.199422680263964,
+      "learning_rate": 8.314102375037611e-06,
+      "loss": 0.6465,
+      "step": 10896
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 42.40572473233016,
+      "learning_rate": 8.312440138642319e-06,
+      "loss": 0.5292,
+      "step": 10897
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.5737713419023,
+      "learning_rate": 8.310777950240238e-06,
+      "loss": 0.6106,
+      "step": 10898
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.52786588045784,
+      "learning_rate": 8.309115809878646e-06,
+      "loss": 0.5592,
+      "step": 10899
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.89945277865343,
+      "learning_rate": 8.307453717604811e-06,
+      "loss": 0.6469,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 34.63208241105138,
+      "learning_rate": 8.305791673466006e-06,
+      "loss": 0.5469,
+      "step": 10901
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.70237072262474,
+      "learning_rate": 8.304129677509491e-06,
+      "loss": 0.5508,
+      "step": 10902
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 39.32086103038911,
+      "learning_rate": 8.302467729782535e-06,
+      "loss": 0.6344,
+      "step": 10903
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.03927689642967,
+      "learning_rate": 8.300805830332404e-06,
+      "loss": 0.4945,
+      "step": 10904
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.5772090839898,
+      "learning_rate": 8.299143979206363e-06,
+      "loss": 0.6149,
+      "step": 10905
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 30.365138595939378,
+      "learning_rate": 8.297482176451675e-06,
+      "loss": 0.5693,
+      "step": 10906
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 37.792047382331084,
+      "learning_rate": 8.295820422115594e-06,
+      "loss": 0.5471,
+      "step": 10907
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 40.97619629324452,
+      "learning_rate": 8.294158716245381e-06,
+      "loss": 0.5883,
+      "step": 10908
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.214158179511465,
+      "learning_rate": 8.2924970588883e-06,
+      "loss": 0.558,
+      "step": 10909
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 40.101381558237286,
+      "learning_rate": 8.290835450091601e-06,
+      "loss": 0.5921,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 40.49713154336283,
+      "learning_rate": 8.289173889902546e-06,
+      "loss": 0.6207,
+      "step": 10911
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 36.87277747099818,
+      "learning_rate": 8.287512378368384e-06,
+      "loss": 0.57,
+      "step": 10912
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 33.006240861483676,
+      "learning_rate": 8.285850915536367e-06,
+      "loss": 0.5373,
+      "step": 10913
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 39.37970492394338,
+      "learning_rate": 8.284189501453745e-06,
+      "loss": 0.5676,
+      "step": 10914
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.36117133073609,
+      "learning_rate": 8.28252813616777e-06,
+      "loss": 0.5575,
+      "step": 10915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.95402088400238,
+      "learning_rate": 8.280866819725695e-06,
+      "loss": 0.4622,
+      "step": 10916
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 34.307159101964515,
+      "learning_rate": 8.279205552174758e-06,
+      "loss": 0.555,
+      "step": 10917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.914257829604345,
+      "learning_rate": 8.27754433356221e-06,
+      "loss": 0.5747,
+      "step": 10918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 49.65706642012676,
+      "learning_rate": 8.275883163935292e-06,
+      "loss": 0.6149,
+      "step": 10919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.76997621665038,
+      "learning_rate": 8.274222043341247e-06,
+      "loss": 0.5539,
+      "step": 10920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 36.08830419304359,
+      "learning_rate": 8.272560971827323e-06,
+      "loss": 0.584,
+      "step": 10921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 56.67613854999934,
+      "learning_rate": 8.270899949440751e-06,
+      "loss": 0.707,
+      "step": 10922
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.87001291811198,
+      "learning_rate": 8.269238976228771e-06,
+      "loss": 0.6488,
+      "step": 10923
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 39.31610439618624,
+      "learning_rate": 8.267578052238624e-06,
+      "loss": 0.6144,
+      "step": 10924
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 41.04752060375583,
+      "learning_rate": 8.265917177517545e-06,
+      "loss": 0.5277,
+      "step": 10925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 46.62455468054618,
+      "learning_rate": 8.264256352112768e-06,
+      "loss": 0.6319,
+      "step": 10926
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.304222267795474,
+      "learning_rate": 8.262595576071521e-06,
+      "loss": 0.5674,
+      "step": 10927
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.60670757145295,
+      "learning_rate": 8.260934849441042e-06,
+      "loss": 0.6052,
+      "step": 10928
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 39.86622129267815,
+      "learning_rate": 8.259274172268556e-06,
+      "loss": 0.5909,
+      "step": 10929
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 117.56171772892795,
+      "learning_rate": 8.2576135446013e-06,
+      "loss": 0.6991,
+      "step": 10930
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 48.718008110363144,
+      "learning_rate": 8.255952966486491e-06,
+      "loss": 0.6471,
+      "step": 10931
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 34.08333768510363,
+      "learning_rate": 8.25429243797136e-06,
+      "loss": 0.5589,
+      "step": 10932
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 44.88565741836149,
+      "learning_rate": 8.252631959103129e-06,
+      "loss": 0.6603,
+      "step": 10933
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.340052289926156,
+      "learning_rate": 8.250971529929026e-06,
+      "loss": 0.5622,
+      "step": 10934
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.620375821164515,
+      "learning_rate": 8.249311150496271e-06,
+      "loss": 0.6374,
+      "step": 10935
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 42.111629571521775,
+      "learning_rate": 8.247650820852078e-06,
+      "loss": 0.6156,
+      "step": 10936
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 41.024443986554836,
+      "learning_rate": 8.245990541043672e-06,
+      "loss": 0.6103,
+      "step": 10937
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 33.10979629903847,
+      "learning_rate": 8.24433031111827e-06,
+      "loss": 0.5323,
+      "step": 10938
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 31.32247394383787,
+      "learning_rate": 8.242670131123085e-06,
+      "loss": 0.5771,
+      "step": 10939
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 33.58212834382228,
+      "learning_rate": 8.241010001105338e-06,
+      "loss": 0.6567,
+      "step": 10940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 38.1960373683099,
+      "learning_rate": 8.239349921112235e-06,
+      "loss": 0.5857,
+      "step": 10941
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 31.12966582934121,
+      "learning_rate": 8.237689891190988e-06,
+      "loss": 0.5294,
+      "step": 10942
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.71981347014159,
+      "learning_rate": 8.236029911388811e-06,
+      "loss": 0.5859,
+      "step": 10943
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 40.58681308363606,
+      "learning_rate": 8.234369981752913e-06,
+      "loss": 0.6413,
+      "step": 10944
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 34.127425404526434,
+      "learning_rate": 8.232710102330502e-06,
+      "loss": 0.6174,
+      "step": 10945
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.145024983840166,
+      "learning_rate": 8.231050273168776e-06,
+      "loss": 0.611,
+      "step": 10946
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.485843578102244,
+      "learning_rate": 8.229390494314949e-06,
+      "loss": 0.5426,
+      "step": 10947
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 39.990236053280455,
+      "learning_rate": 8.227730765816216e-06,
+      "loss": 0.5619,
+      "step": 10948
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 47.82927019803879,
+      "learning_rate": 8.226071087719789e-06,
+      "loss": 0.6776,
+      "step": 10949
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 33.01292399719501,
+      "learning_rate": 8.224411460072863e-06,
+      "loss": 0.5828,
+      "step": 10950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 38.02920445228244,
+      "learning_rate": 8.222751882922632e-06,
+      "loss": 0.678,
+      "step": 10951
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 47.43816886902742,
+      "learning_rate": 8.221092356316298e-06,
+      "loss": 0.6688,
+      "step": 10952
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.37151231353593,
+      "learning_rate": 8.21943288030106e-06,
+      "loss": 0.4933,
+      "step": 10953
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 46.77302261303612,
+      "learning_rate": 8.217773454924105e-06,
+      "loss": 0.596,
+      "step": 10954
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.84186673776556,
+      "learning_rate": 8.216114080232635e-06,
+      "loss": 0.5724,
+      "step": 10955
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 39.19092043257393,
+      "learning_rate": 8.214454756273836e-06,
+      "loss": 0.5456,
+      "step": 10956
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.99284957366229,
+      "learning_rate": 8.2127954830949e-06,
+      "loss": 0.6241,
+      "step": 10957
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 44.793495127891056,
+      "learning_rate": 8.21113626074301e-06,
+      "loss": 0.5625,
+      "step": 10958
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.166751731396026,
+      "learning_rate": 8.20947708926536e-06,
+      "loss": 0.6105,
+      "step": 10959
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.31246868571015,
+      "learning_rate": 8.20781796870914e-06,
+      "loss": 0.6106,
+      "step": 10960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.78420042849956,
+      "learning_rate": 8.20615889912152e-06,
+      "loss": 0.5045,
+      "step": 10961
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 43.79873052851968,
+      "learning_rate": 8.204499880549699e-06,
+      "loss": 0.6455,
+      "step": 10962
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 43.31623448817637,
+      "learning_rate": 8.202840913040847e-06,
+      "loss": 0.737,
+      "step": 10963
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 43.15128744266789,
+      "learning_rate": 8.201181996642147e-06,
+      "loss": 0.5963,
+      "step": 10964
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 40.78467282952358,
+      "learning_rate": 8.199523131400783e-06,
+      "loss": 0.65,
+      "step": 10965
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 50.27619514163961,
+      "learning_rate": 8.197864317363926e-06,
+      "loss": 0.571,
+      "step": 10966
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 45.666249244875175,
+      "learning_rate": 8.19620555457875e-06,
+      "loss": 0.6114,
+      "step": 10967
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 41.65600721513149,
+      "learning_rate": 8.194546843092438e-06,
+      "loss": 0.7051,
+      "step": 10968
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 36.982988113166236,
+      "learning_rate": 8.192888182952155e-06,
+      "loss": 0.5792,
+      "step": 10969
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.60667419879051,
+      "learning_rate": 8.191229574205078e-06,
+      "loss": 0.5014,
+      "step": 10970
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 31.761438678392867,
+      "learning_rate": 8.18957101689837e-06,
+      "loss": 0.5372,
+      "step": 10971
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 41.965379717402115,
+      "learning_rate": 8.187912511079205e-06,
+      "loss": 0.693,
+      "step": 10972
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 40.55202780967274,
+      "learning_rate": 8.186254056794747e-06,
+      "loss": 0.6866,
+      "step": 10973
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 41.49806203108912,
+      "learning_rate": 8.184595654092161e-06,
+      "loss": 0.5917,
+      "step": 10974
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 36.245348118865145,
+      "learning_rate": 8.182937303018619e-06,
+      "loss": 0.5805,
+      "step": 10975
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 37.47475693174071,
+      "learning_rate": 8.181279003621274e-06,
+      "loss": 0.6153,
+      "step": 10976
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 43.3661464593011,
+      "learning_rate": 8.179620755947287e-06,
+      "loss": 0.637,
+      "step": 10977
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 36.79869736615366,
+      "learning_rate": 8.177962560043824e-06,
+      "loss": 0.6116,
+      "step": 10978
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.7224303054796,
+      "learning_rate": 8.17630441595804e-06,
+      "loss": 0.5714,
+      "step": 10979
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.76256624193666,
+      "learning_rate": 8.17464632373709e-06,
+      "loss": 0.582,
+      "step": 10980
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 39.82041553635913,
+      "learning_rate": 8.172988283428132e-06,
+      "loss": 0.6542,
+      "step": 10981
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 33.12834168978517,
+      "learning_rate": 8.171330295078318e-06,
+      "loss": 0.5833,
+      "step": 10982
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.29621297241629,
+      "learning_rate": 8.1696723587348e-06,
+      "loss": 0.6479,
+      "step": 10983
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 29.64758176511012,
+      "learning_rate": 8.168014474444731e-06,
+      "loss": 0.5186,
+      "step": 10984
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 31.979637372821983,
+      "learning_rate": 8.166356642255258e-06,
+      "loss": 0.5285,
+      "step": 10985
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 42.2505797707545,
+      "learning_rate": 8.164698862213527e-06,
+      "loss": 0.6515,
+      "step": 10986
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.64856559888361,
+      "learning_rate": 8.16304113436669e-06,
+      "loss": 0.599,
+      "step": 10987
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 30.577554088286135,
+      "learning_rate": 8.161383458761887e-06,
+      "loss": 0.5461,
+      "step": 10988
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 40.58080918310447,
+      "learning_rate": 8.159725835446266e-06,
+      "loss": 0.6426,
+      "step": 10989
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 45.094637726566695,
+      "learning_rate": 8.15806826446696e-06,
+      "loss": 0.4873,
+      "step": 10990
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 31.540228549284016,
+      "learning_rate": 8.156410745871119e-06,
+      "loss": 0.5232,
+      "step": 10991
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 32.24578897943309,
+      "learning_rate": 8.154753279705875e-06,
+      "loss": 0.5448,
+      "step": 10992
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.699594163986944,
+      "learning_rate": 8.15309586601837e-06,
+      "loss": 0.5635,
+      "step": 10993
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 32.83264084794119,
+      "learning_rate": 8.151438504855742e-06,
+      "loss": 0.5257,
+      "step": 10994
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 42.32555383003326,
+      "learning_rate": 8.149781196265117e-06,
+      "loss": 0.5733,
+      "step": 10995
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.53623127655819,
+      "learning_rate": 8.148123940293632e-06,
+      "loss": 0.5535,
+      "step": 10996
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.720839021444164,
+      "learning_rate": 8.146466736988422e-06,
+      "loss": 0.7269,
+      "step": 10997
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 39.311023241698074,
+      "learning_rate": 8.144809586396608e-06,
+      "loss": 0.5842,
+      "step": 10998
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.91482154562726,
+      "learning_rate": 8.143152488565332e-06,
+      "loss": 0.608,
+      "step": 10999
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 42.62322073660082,
+      "learning_rate": 8.141495443541708e-06,
+      "loss": 0.6102,
+      "step": 11000
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.112002584669376,
+      "learning_rate": 8.139838451372868e-06,
+      "loss": 0.4598,
+      "step": 11001
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 42.84606934550155,
+      "learning_rate": 8.138181512105931e-06,
+      "loss": 0.6787,
+      "step": 11002
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 48.02127428272997,
+      "learning_rate": 8.136524625788026e-06,
+      "loss": 0.6187,
+      "step": 11003
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 33.32632249696282,
+      "learning_rate": 8.134867792466272e-06,
+      "loss": 0.528,
+      "step": 11004
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 40.18374282810744,
+      "learning_rate": 8.133211012187783e-06,
+      "loss": 0.5306,
+      "step": 11005
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 43.047965189288234,
+      "learning_rate": 8.131554284999683e-06,
+      "loss": 0.6086,
+      "step": 11006
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 38.14612143112773,
+      "learning_rate": 8.129897610949086e-06,
+      "loss": 0.5824,
+      "step": 11007
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 43.917263742956784,
+      "learning_rate": 8.128240990083104e-06,
+      "loss": 0.6214,
+      "step": 11008
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 44.203479060200905,
+      "learning_rate": 8.12658442244886e-06,
+      "loss": 0.6728,
+      "step": 11009
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 31.635779277710885,
+      "learning_rate": 8.124927908093455e-06,
+      "loss": 0.5331,
+      "step": 11010
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 38.219139455836846,
+      "learning_rate": 8.123271447064001e-06,
+      "loss": 0.5499,
+      "step": 11011
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 41.308282385840805,
+      "learning_rate": 8.121615039407613e-06,
+      "loss": 0.548,
+      "step": 11012
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 50.88810261796153,
+      "learning_rate": 8.119958685171392e-06,
+      "loss": 0.6485,
+      "step": 11013
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 44.65290716431684,
+      "learning_rate": 8.11830238440245e-06,
+      "loss": 0.666,
+      "step": 11014
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 42.09248048812143,
+      "learning_rate": 8.116646137147884e-06,
+      "loss": 0.6411,
+      "step": 11015
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 36.690506488623924,
+      "learning_rate": 8.114989943454801e-06,
+      "loss": 0.5951,
+      "step": 11016
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 28.581783533945323,
+      "learning_rate": 8.113333803370297e-06,
+      "loss": 0.4771,
+      "step": 11017
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.92909084051136,
+      "learning_rate": 8.111677716941481e-06,
+      "loss": 0.5821,
+      "step": 11018
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 39.11419994540545,
+      "learning_rate": 8.110021684215448e-06,
+      "loss": 0.6588,
+      "step": 11019
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.74996088354345,
+      "learning_rate": 8.10836570523929e-06,
+      "loss": 0.5735,
+      "step": 11020
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.52683797499247,
+      "learning_rate": 8.106709780060103e-06,
+      "loss": 0.577,
+      "step": 11021
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.04932623908108,
+      "learning_rate": 8.105053908724985e-06,
+      "loss": 0.5179,
+      "step": 11022
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 50.78982609560779,
+      "learning_rate": 8.103398091281023e-06,
+      "loss": 0.6198,
+      "step": 11023
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 43.33188553953006,
+      "learning_rate": 8.101742327775312e-06,
+      "loss": 0.6544,
+      "step": 11024
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 34.88760119265974,
+      "learning_rate": 8.10008661825494e-06,
+      "loss": 0.4959,
+      "step": 11025
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 34.69627839779488,
+      "learning_rate": 8.09843096276699e-06,
+      "loss": 0.529,
+      "step": 11026
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 31.749093264300317,
+      "learning_rate": 8.096775361358552e-06,
+      "loss": 0.4893,
+      "step": 11027
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 36.019762012589325,
+      "learning_rate": 8.095119814076711e-06,
+      "loss": 0.5991,
+      "step": 11028
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 36.055615427427604,
+      "learning_rate": 8.093464320968547e-06,
+      "loss": 0.4592,
+      "step": 11029
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 34.91211051546108,
+      "learning_rate": 8.09180888208114e-06,
+      "loss": 0.6307,
+      "step": 11030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 52.18115012090447,
+      "learning_rate": 8.090153497461573e-06,
+      "loss": 0.5792,
+      "step": 11031
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 38.55852828755975,
+      "learning_rate": 8.088498167156926e-06,
+      "loss": 0.6498,
+      "step": 11032
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 41.888892963576744,
+      "learning_rate": 8.086842891214274e-06,
+      "loss": 0.5951,
+      "step": 11033
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 38.32781799417388,
+      "learning_rate": 8.085187669680687e-06,
+      "loss": 0.5938,
+      "step": 11034
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 30.45670384085144,
+      "learning_rate": 8.083532502603243e-06,
+      "loss": 0.4918,
+      "step": 11035
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 45.558627525251,
+      "learning_rate": 8.081877390029013e-06,
+      "loss": 0.6938,
+      "step": 11036
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 44.97297894579015,
+      "learning_rate": 8.08022233200507e-06,
+      "loss": 0.6712,
+      "step": 11037
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.866898926142404,
+      "learning_rate": 8.078567328578482e-06,
+      "loss": 0.5458,
+      "step": 11038
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 39.8003088162772,
+      "learning_rate": 8.076912379796314e-06,
+      "loss": 0.6502,
+      "step": 11039
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 64.11802882324302,
+      "learning_rate": 8.07525748570563e-06,
+      "loss": 0.6375,
+      "step": 11040
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 32.26122699533201,
+      "learning_rate": 8.0736026463535e-06,
+      "loss": 0.5071,
+      "step": 11041
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 36.51274089661938,
+      "learning_rate": 8.071947861786982e-06,
+      "loss": 0.6468,
+      "step": 11042
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 34.642108209264684,
+      "learning_rate": 8.070293132053143e-06,
+      "loss": 0.5792,
+      "step": 11043
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 55.249514461127234,
+      "learning_rate": 8.068638457199037e-06,
+      "loss": 0.5277,
+      "step": 11044
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 37.50270188452665,
+      "learning_rate": 8.066983837271722e-06,
+      "loss": 0.5414,
+      "step": 11045
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 37.927098628439786,
+      "learning_rate": 8.065329272318255e-06,
+      "loss": 0.5714,
+      "step": 11046
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 36.692808010957734,
+      "learning_rate": 8.063674762385691e-06,
+      "loss": 0.6148,
+      "step": 11047
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 41.0844399642982,
+      "learning_rate": 8.06202030752109e-06,
+      "loss": 0.6218,
+      "step": 11048
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 53.42293028725006,
+      "learning_rate": 8.06036590777149e-06,
+      "loss": 0.6023,
+      "step": 11049
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 35.05613098505186,
+      "learning_rate": 8.058711563183955e-06,
+      "loss": 0.5588,
+      "step": 11050
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 39.840877157124424,
+      "learning_rate": 8.057057273805525e-06,
+      "loss": 0.6176,
+      "step": 11051
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 39.82684973631452,
+      "learning_rate": 8.055403039683247e-06,
+      "loss": 0.5663,
+      "step": 11052
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.31144907726856,
+      "learning_rate": 8.053748860864174e-06,
+      "loss": 0.5789,
+      "step": 11053
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.26451627532997,
+      "learning_rate": 8.052094737395343e-06,
+      "loss": 0.4767,
+      "step": 11054
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 34.443463309390516,
+      "learning_rate": 8.050440669323796e-06,
+      "loss": 0.5553,
+      "step": 11055
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 36.861631397945594,
+      "learning_rate": 8.048786656696579e-06,
+      "loss": 0.6125,
+      "step": 11056
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 42.037530950194004,
+      "learning_rate": 8.047132699560725e-06,
+      "loss": 0.6205,
+      "step": 11057
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 53.774429035037464,
+      "learning_rate": 8.04547879796328e-06,
+      "loss": 0.6061,
+      "step": 11058
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 40.79512564839761,
+      "learning_rate": 8.04382495195127e-06,
+      "loss": 0.6963,
+      "step": 11059
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 47.988562895243724,
+      "learning_rate": 8.042171161571734e-06,
+      "loss": 0.5826,
+      "step": 11060
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.41671156966812,
+      "learning_rate": 8.040517426871703e-06,
+      "loss": 0.5024,
+      "step": 11061
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 42.28691546368487,
+      "learning_rate": 8.038863747898214e-06,
+      "loss": 0.6369,
+      "step": 11062
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 48.98789640238512,
+      "learning_rate": 8.037210124698294e-06,
+      "loss": 0.6459,
+      "step": 11063
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 44.073391037782365,
+      "learning_rate": 8.03555655731897e-06,
+      "loss": 0.555,
+      "step": 11064
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 46.27509281328437,
+      "learning_rate": 8.033903045807262e-06,
+      "loss": 0.5851,
+      "step": 11065
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 39.09714775048243,
+      "learning_rate": 8.032249590210206e-06,
+      "loss": 0.6155,
+      "step": 11066
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 36.50600392135079,
+      "learning_rate": 8.030596190574818e-06,
+      "loss": 0.5156,
+      "step": 11067
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.557132408943986,
+      "learning_rate": 8.028942846948126e-06,
+      "loss": 0.5467,
+      "step": 11068
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 36.13957559832479,
+      "learning_rate": 8.027289559377145e-06,
+      "loss": 0.5483,
+      "step": 11069
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 41.67578994560669,
+      "learning_rate": 8.025636327908895e-06,
+      "loss": 0.4903,
+      "step": 11070
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.24062212817534,
+      "learning_rate": 8.02398315259039e-06,
+      "loss": 0.6144,
+      "step": 11071
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.2679814713048,
+      "learning_rate": 8.02233003346865e-06,
+      "loss": 0.4971,
+      "step": 11072
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 35.05122806626065,
+      "learning_rate": 8.02067697059069e-06,
+      "loss": 0.5547,
+      "step": 11073
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.456230461304465,
+      "learning_rate": 8.019023964003513e-06,
+      "loss": 0.564,
+      "step": 11074
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 51.48030887459969,
+      "learning_rate": 8.01737101375414e-06,
+      "loss": 0.6382,
+      "step": 11075
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.231938448562055,
+      "learning_rate": 8.015718119889576e-06,
+      "loss": 0.5948,
+      "step": 11076
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 40.017827495837054,
+      "learning_rate": 8.014065282456829e-06,
+      "loss": 0.5417,
+      "step": 11077
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 40.575977255380884,
+      "learning_rate": 8.0124125015029e-06,
+      "loss": 0.5871,
+      "step": 11078
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.895433523582525,
+      "learning_rate": 8.010759777074796e-06,
+      "loss": 0.5341,
+      "step": 11079
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.497021361204844,
+      "learning_rate": 8.009107109219522e-06,
+      "loss": 0.5347,
+      "step": 11080
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 35.94458054491881,
+      "learning_rate": 8.007454497984078e-06,
+      "loss": 0.6085,
+      "step": 11081
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 32.96107591381609,
+      "learning_rate": 8.005801943415462e-06,
+      "loss": 0.5363,
+      "step": 11082
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 42.17870863689948,
+      "learning_rate": 8.004149445560675e-06,
+      "loss": 0.5433,
+      "step": 11083
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 33.78485407910824,
+      "learning_rate": 8.002497004466703e-06,
+      "loss": 0.5817,
+      "step": 11084
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 48.40063286852267,
+      "learning_rate": 8.000844620180553e-06,
+      "loss": 0.7202,
+      "step": 11085
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.88444694040878,
+      "learning_rate": 7.999192292749209e-06,
+      "loss": 0.5223,
+      "step": 11086
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 31.448065747394892,
+      "learning_rate": 7.997540022219672e-06,
+      "loss": 0.5043,
+      "step": 11087
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 42.558808537746955,
+      "learning_rate": 7.995887808638921e-06,
+      "loss": 0.5567,
+      "step": 11088
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 43.71474687410293,
+      "learning_rate": 7.99423565205395e-06,
+      "loss": 0.5989,
+      "step": 11089
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 36.82970530923137,
+      "learning_rate": 7.99258355251174e-06,
+      "loss": 0.594,
+      "step": 11090
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 39.303160548097395,
+      "learning_rate": 7.990931510059281e-06,
+      "loss": 0.543,
+      "step": 11091
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 35.48247516943257,
+      "learning_rate": 7.989279524743559e-06,
+      "loss": 0.5402,
+      "step": 11092
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 42.241939467048866,
+      "learning_rate": 7.987627596611546e-06,
+      "loss": 0.5687,
+      "step": 11093
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 39.57820740520634,
+      "learning_rate": 7.985975725710229e-06,
+      "loss": 0.618,
+      "step": 11094
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 32.16961037222874,
+      "learning_rate": 7.984323912086584e-06,
+      "loss": 0.5497,
+      "step": 11095
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 40.928111075668134,
+      "learning_rate": 7.982672155787586e-06,
+      "loss": 0.58,
+      "step": 11096
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 30.12769916915236,
+      "learning_rate": 7.981020456860215e-06,
+      "loss": 0.4592,
+      "step": 11097
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 35.49602183362716,
+      "learning_rate": 7.979368815351441e-06,
+      "loss": 0.5637,
+      "step": 11098
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 32.51616633421506,
+      "learning_rate": 7.977717231308233e-06,
+      "loss": 0.5381,
+      "step": 11099
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 31.20916257545154,
+      "learning_rate": 7.976065704777566e-06,
+      "loss": 0.5576,
+      "step": 11100
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 37.035409086998115,
+      "learning_rate": 7.974414235806403e-06,
+      "loss": 0.5782,
+      "step": 11101
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.177292185268925,
+      "learning_rate": 7.972762824441722e-06,
+      "loss": 0.5893,
+      "step": 11102
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 46.38333063027368,
+      "learning_rate": 7.971111470730474e-06,
+      "loss": 0.6861,
+      "step": 11103
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 45.74488256318686,
+      "learning_rate": 7.96946017471963e-06,
+      "loss": 0.6298,
+      "step": 11104
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.329531053271474,
+      "learning_rate": 7.967808936456149e-06,
+      "loss": 0.6046,
+      "step": 11105
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 31.29496610764891,
+      "learning_rate": 7.966157755986994e-06,
+      "loss": 0.4977,
+      "step": 11106
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 85.21173502156338,
+      "learning_rate": 7.964506633359128e-06,
+      "loss": 0.6119,
+      "step": 11107
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.5549188380928,
+      "learning_rate": 7.962855568619499e-06,
+      "loss": 0.5108,
+      "step": 11108
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.85300959644695,
+      "learning_rate": 7.961204561815063e-06,
+      "loss": 0.5814,
+      "step": 11109
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 45.34023457418944,
+      "learning_rate": 7.95955361299278e-06,
+      "loss": 0.5985,
+      "step": 11110
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.05379590220337,
+      "learning_rate": 7.957902722199596e-06,
+      "loss": 0.5825,
+      "step": 11111
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.33493755666756,
+      "learning_rate": 7.956251889482467e-06,
+      "loss": 0.5717,
+      "step": 11112
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 36.72357415507921,
+      "learning_rate": 7.954601114888338e-06,
+      "loss": 0.5899,
+      "step": 11113
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 42.47022739336052,
+      "learning_rate": 7.952950398464156e-06,
+      "loss": 0.5337,
+      "step": 11114
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 41.93670726885335,
+      "learning_rate": 7.951299740256865e-06,
+      "loss": 0.611,
+      "step": 11115
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 44.25145468772297,
+      "learning_rate": 7.949649140313413e-06,
+      "loss": 0.6404,
+      "step": 11116
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.190807291185926,
+      "learning_rate": 7.947998598680743e-06,
+      "loss": 0.5832,
+      "step": 11117
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 33.94687025141619,
+      "learning_rate": 7.946348115405788e-06,
+      "loss": 0.5302,
+      "step": 11118
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 45.45731321088241,
+      "learning_rate": 7.944697690535493e-06,
+      "loss": 0.5799,
+      "step": 11119
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.71687802522409,
+      "learning_rate": 7.943047324116793e-06,
+      "loss": 0.6351,
+      "step": 11120
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.788132822313514,
+      "learning_rate": 7.941397016196622e-06,
+      "loss": 0.6361,
+      "step": 11121
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.818077007150336,
+      "learning_rate": 7.939746766821921e-06,
+      "loss": 0.6015,
+      "step": 11122
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.864528339322476,
+      "learning_rate": 7.938096576039613e-06,
+      "loss": 0.5236,
+      "step": 11123
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 34.282478414832084,
+      "learning_rate": 7.93644644389663e-06,
+      "loss": 0.5298,
+      "step": 11124
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.07440902215799,
+      "learning_rate": 7.934796370439907e-06,
+      "loss": 0.5518,
+      "step": 11125
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 35.820942727371865,
+      "learning_rate": 7.933146355716368e-06,
+      "loss": 0.5412,
+      "step": 11126
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.48056333369913,
+      "learning_rate": 7.931496399772935e-06,
+      "loss": 0.5346,
+      "step": 11127
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 42.90993511436409,
+      "learning_rate": 7.929846502656533e-06,
+      "loss": 0.6721,
+      "step": 11128
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 43.36894785737916,
+      "learning_rate": 7.928196664414088e-06,
+      "loss": 0.6387,
+      "step": 11129
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 49.90136513782524,
+      "learning_rate": 7.926546885092515e-06,
+      "loss": 0.5209,
+      "step": 11130
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 43.8937998757669,
+      "learning_rate": 7.92489716473874e-06,
+      "loss": 0.6039,
+      "step": 11131
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.060277136254946,
+      "learning_rate": 7.923247503399674e-06,
+      "loss": 0.5772,
+      "step": 11132
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 41.81825968783181,
+      "learning_rate": 7.921597901122234e-06,
+      "loss": 0.5123,
+      "step": 11133
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 45.63788125039022,
+      "learning_rate": 7.919948357953332e-06,
+      "loss": 0.675,
+      "step": 11134
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 53.877979028036194,
+      "learning_rate": 7.918298873939882e-06,
+      "loss": 0.6205,
+      "step": 11135
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.1695002651824,
+      "learning_rate": 7.916649449128799e-06,
+      "loss": 0.6522,
+      "step": 11136
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.29832679086814,
+      "learning_rate": 7.915000083566981e-06,
+      "loss": 0.6172,
+      "step": 11137
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 42.25703588848656,
+      "learning_rate": 7.913350777301342e-06,
+      "loss": 0.6245,
+      "step": 11138
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 35.70519738391036,
+      "learning_rate": 7.911701530378789e-06,
+      "loss": 0.6203,
+      "step": 11139
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 50.440478727358915,
+      "learning_rate": 7.910052342846219e-06,
+      "loss": 0.6556,
+      "step": 11140
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.64074817077483,
+      "learning_rate": 7.908403214750544e-06,
+      "loss": 0.6009,
+      "step": 11141
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.26221162415142,
+      "learning_rate": 7.906754146138655e-06,
+      "loss": 0.5213,
+      "step": 11142
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 41.37617945070436,
+      "learning_rate": 7.905105137057451e-06,
+      "loss": 0.5694,
+      "step": 11143
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 35.30903159299585,
+      "learning_rate": 7.903456187553833e-06,
+      "loss": 0.5256,
+      "step": 11144
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 30.666594674105383,
+      "learning_rate": 7.901807297674693e-06,
+      "loss": 0.5087,
+      "step": 11145
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.427200615353506,
+      "learning_rate": 7.900158467466932e-06,
+      "loss": 0.5613,
+      "step": 11146
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.24874606678179,
+      "learning_rate": 7.89850969697743e-06,
+      "loss": 0.5945,
+      "step": 11147
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.69892426210565,
+      "learning_rate": 7.896860986253086e-06,
+      "loss": 0.5598,
+      "step": 11148
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 42.54920057650742,
+      "learning_rate": 7.89521233534078e-06,
+      "loss": 0.6248,
+      "step": 11149
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.96746592176507,
+      "learning_rate": 7.89356374428741e-06,
+      "loss": 0.5772,
+      "step": 11150
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.948078704217174,
+      "learning_rate": 7.891915213139855e-06,
+      "loss": 0.5766,
+      "step": 11151
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.54954500900511,
+      "learning_rate": 7.890266741944995e-06,
+      "loss": 0.5833,
+      "step": 11152
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 31.593887253339584,
+      "learning_rate": 7.888618330749715e-06,
+      "loss": 0.5284,
+      "step": 11153
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.683230766928496,
+      "learning_rate": 7.886969979600894e-06,
+      "loss": 0.6636,
+      "step": 11154
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.154195752274674,
+      "learning_rate": 7.885321688545412e-06,
+      "loss": 0.5812,
+      "step": 11155
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 36.46239752795938,
+      "learning_rate": 7.883673457630144e-06,
+      "loss": 0.5361,
+      "step": 11156
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 41.4194203228268,
+      "learning_rate": 7.882025286901965e-06,
+      "loss": 0.643,
+      "step": 11157
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 49.54422482729433,
+      "learning_rate": 7.880377176407749e-06,
+      "loss": 0.5338,
+      "step": 11158
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 33.830824179148024,
+      "learning_rate": 7.878729126194362e-06,
+      "loss": 0.5852,
+      "step": 11159
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.7091518783395,
+      "learning_rate": 7.877081136308681e-06,
+      "loss": 0.6523,
+      "step": 11160
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.231202046242124,
+      "learning_rate": 7.875433206797573e-06,
+      "loss": 0.5859,
+      "step": 11161
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.20863383073413,
+      "learning_rate": 7.873785337707897e-06,
+      "loss": 0.5836,
+      "step": 11162
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.21661716583351,
+      "learning_rate": 7.872137529086525e-06,
+      "loss": 0.4773,
+      "step": 11163
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 60.50708168135156,
+      "learning_rate": 7.870489780980315e-06,
+      "loss": 0.5391,
+      "step": 11164
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 35.85796979423943,
+      "learning_rate": 7.86884209343613e-06,
+      "loss": 0.6307,
+      "step": 11165
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 42.84168310765504,
+      "learning_rate": 7.867194466500831e-06,
+      "loss": 0.6964,
+      "step": 11166
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 36.1596311407666,
+      "learning_rate": 7.865546900221274e-06,
+      "loss": 0.5399,
+      "step": 11167
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 30.821082291459902,
+      "learning_rate": 7.863899394644312e-06,
+      "loss": 0.4824,
+      "step": 11168
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.75227093348777,
+      "learning_rate": 7.862251949816803e-06,
+      "loss": 0.542,
+      "step": 11169
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 37.656469858881074,
+      "learning_rate": 7.860604565785596e-06,
+      "loss": 0.5682,
+      "step": 11170
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.408255376171915,
+      "learning_rate": 7.85895724259755e-06,
+      "loss": 0.5446,
+      "step": 11171
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 45.40739991319679,
+      "learning_rate": 7.857309980299502e-06,
+      "loss": 0.6305,
+      "step": 11172
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 48.17340025589021,
+      "learning_rate": 7.855662778938306e-06,
+      "loss": 0.6676,
+      "step": 11173
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.00752121126657,
+      "learning_rate": 7.854015638560804e-06,
+      "loss": 0.56,
+      "step": 11174
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 35.66845344178205,
+      "learning_rate": 7.852368559213846e-06,
+      "loss": 0.5674,
+      "step": 11175
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 50.15594657259513,
+      "learning_rate": 7.850721540944268e-06,
+      "loss": 0.5889,
+      "step": 11176
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.80984499477694,
+      "learning_rate": 7.849074583798912e-06,
+      "loss": 0.6218,
+      "step": 11177
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 42.508159084270794,
+      "learning_rate": 7.847427687824615e-06,
+      "loss": 0.5172,
+      "step": 11178
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.12832160300598,
+      "learning_rate": 7.845780853068217e-06,
+      "loss": 0.5414,
+      "step": 11179
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 34.26628435007578,
+      "learning_rate": 7.844134079576553e-06,
+      "loss": 0.544,
+      "step": 11180
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.74747036091078,
+      "learning_rate": 7.84248736739645e-06,
+      "loss": 0.5742,
+      "step": 11181
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.15855490106516,
+      "learning_rate": 7.840840716574748e-06,
+      "loss": 0.544,
+      "step": 11182
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.592349363604534,
+      "learning_rate": 7.839194127158271e-06,
+      "loss": 0.6009,
+      "step": 11183
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.4149747959763,
+      "learning_rate": 7.837547599193849e-06,
+      "loss": 0.5345,
+      "step": 11184
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.81116341394936,
+      "learning_rate": 7.835901132728311e-06,
+      "loss": 0.5652,
+      "step": 11185
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.36877790876204,
+      "learning_rate": 7.834254727808477e-06,
+      "loss": 0.6291,
+      "step": 11186
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 50.140187547771944,
+      "learning_rate": 7.83260838448117e-06,
+      "loss": 0.6843,
+      "step": 11187
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.960297755981955,
+      "learning_rate": 7.830962102793214e-06,
+      "loss": 0.6117,
+      "step": 11188
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 46.259265301515136,
+      "learning_rate": 7.829315882791426e-06,
+      "loss": 0.6679,
+      "step": 11189
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 44.72859649164499,
+      "learning_rate": 7.827669724522632e-06,
+      "loss": 0.6271,
+      "step": 11190
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.30386232079021,
+      "learning_rate": 7.826023628033631e-06,
+      "loss": 0.5467,
+      "step": 11191
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 41.90486358351829,
+      "learning_rate": 7.824377593371248e-06,
+      "loss": 0.6672,
+      "step": 11192
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.33706170208727,
+      "learning_rate": 7.822731620582294e-06,
+      "loss": 0.5673,
+      "step": 11193
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 45.51482595730402,
+      "learning_rate": 7.821085709713581e-06,
+      "loss": 0.7105,
+      "step": 11194
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.04793371379539,
+      "learning_rate": 7.819439860811915e-06,
+      "loss": 0.5463,
+      "step": 11195
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 43.34275544448128,
+      "learning_rate": 7.817794073924104e-06,
+      "loss": 0.5236,
+      "step": 11196
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 34.80639807953368,
+      "learning_rate": 7.816148349096951e-06,
+      "loss": 0.5581,
+      "step": 11197
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.668798481976964,
+      "learning_rate": 7.814502686377263e-06,
+      "loss": 0.6358,
+      "step": 11198
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.05676730388731,
+      "learning_rate": 7.812857085811837e-06,
+      "loss": 0.6551,
+      "step": 11199
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.1175244592275,
+      "learning_rate": 7.811211547447483e-06,
+      "loss": 0.5172,
+      "step": 11200
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 33.176883712878656,
+      "learning_rate": 7.809566071330987e-06,
+      "loss": 0.5858,
+      "step": 11201
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 42.243571413094145,
+      "learning_rate": 7.807920657509152e-06,
+      "loss": 0.6067,
+      "step": 11202
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.43355564042645,
+      "learning_rate": 7.80627530602877e-06,
+      "loss": 0.5399,
+      "step": 11203
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.64480971718038,
+      "learning_rate": 7.804630016936636e-06,
+      "loss": 0.6121,
+      "step": 11204
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 43.59454786329363,
+      "learning_rate": 7.802984790279542e-06,
+      "loss": 0.6035,
+      "step": 11205
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 43.39422267220687,
+      "learning_rate": 7.801339626104272e-06,
+      "loss": 0.5602,
+      "step": 11206
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.936312299761894,
+      "learning_rate": 7.799694524457616e-06,
+      "loss": 0.5986,
+      "step": 11207
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 35.383217147026066,
+      "learning_rate": 7.798049485386365e-06,
+      "loss": 0.5473,
+      "step": 11208
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 43.811187659184455,
+      "learning_rate": 7.796404508937295e-06,
+      "loss": 0.7014,
+      "step": 11209
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 44.888848832254716,
+      "learning_rate": 7.794759595157194e-06,
+      "loss": 0.589,
+      "step": 11210
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.81252015798993,
+      "learning_rate": 7.79311474409284e-06,
+      "loss": 0.5304,
+      "step": 11211
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 35.00703302293306,
+      "learning_rate": 7.791469955791009e-06,
+      "loss": 0.5652,
+      "step": 11212
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 47.05544723614372,
+      "learning_rate": 7.789825230298481e-06,
+      "loss": 0.5894,
+      "step": 11213
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.92981121961684,
+      "learning_rate": 7.788180567662031e-06,
+      "loss": 0.5972,
+      "step": 11214
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 41.89056634383059,
+      "learning_rate": 7.786535967928437e-06,
+      "loss": 0.6165,
+      "step": 11215
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 44.4664555143312,
+      "learning_rate": 7.784891431144459e-06,
+      "loss": 0.5575,
+      "step": 11216
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.01502023260007,
+      "learning_rate": 7.783246957356876e-06,
+      "loss": 0.6083,
+      "step": 11217
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 47.10183102253894,
+      "learning_rate": 7.781602546612447e-06,
+      "loss": 0.6235,
+      "step": 11218
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.195961901470774,
+      "learning_rate": 7.779958198957951e-06,
+      "loss": 0.5541,
+      "step": 11219
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 35.85478562052234,
+      "learning_rate": 7.778313914440143e-06,
+      "loss": 0.6296,
+      "step": 11220
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 39.342984962448675,
+      "learning_rate": 7.776669693105786e-06,
+      "loss": 0.6522,
+      "step": 11221
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 35.758887106658975,
+      "learning_rate": 7.775025535001643e-06,
+      "loss": 0.5316,
+      "step": 11222
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.66769710131634,
+      "learning_rate": 7.773381440174473e-06,
+      "loss": 0.6189,
+      "step": 11223
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 34.31783066704273,
+      "learning_rate": 7.771737408671034e-06,
+      "loss": 0.5007,
+      "step": 11224
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 42.94939112747941,
+      "learning_rate": 7.770093440538076e-06,
+      "loss": 0.6185,
+      "step": 11225
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 40.36699221833076,
+      "learning_rate": 7.768449535822357e-06,
+      "loss": 0.5133,
+      "step": 11226
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 33.334730667787376,
+      "learning_rate": 7.766805694570629e-06,
+      "loss": 0.5734,
+      "step": 11227
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.264004650916405,
+      "learning_rate": 7.76516191682964e-06,
+      "loss": 0.6114,
+      "step": 11228
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 39.769615473750335,
+      "learning_rate": 7.76351820264614e-06,
+      "loss": 0.6561,
+      "step": 11229
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 41.4848718818175,
+      "learning_rate": 7.761874552066873e-06,
+      "loss": 0.5605,
+      "step": 11230
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 39.737845031929915,
+      "learning_rate": 7.760230965138582e-06,
+      "loss": 0.5674,
+      "step": 11231
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 45.12123587011901,
+      "learning_rate": 7.758587441908014e-06,
+      "loss": 0.6551,
+      "step": 11232
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 45.13773492769404,
+      "learning_rate": 7.756943982421907e-06,
+      "loss": 0.5914,
+      "step": 11233
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 34.6557539929324,
+      "learning_rate": 7.755300586727007e-06,
+      "loss": 0.6232,
+      "step": 11234
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 34.01110635304479,
+      "learning_rate": 7.753657254870038e-06,
+      "loss": 0.5296,
+      "step": 11235
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 48.3314980514714,
+      "learning_rate": 7.752013986897747e-06,
+      "loss": 0.7722,
+      "step": 11236
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 39.97188171692053,
+      "learning_rate": 7.750370782856858e-06,
+      "loss": 0.5473,
+      "step": 11237
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.267209937773366,
+      "learning_rate": 7.748727642794111e-06,
+      "loss": 0.6398,
+      "step": 11238
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.479317134565086,
+      "learning_rate": 7.747084566756237e-06,
+      "loss": 0.5634,
+      "step": 11239
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.8624272623287,
+      "learning_rate": 7.745441554789956e-06,
+      "loss": 0.5851,
+      "step": 11240
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 44.933177830335524,
+      "learning_rate": 7.743798606941997e-06,
+      "loss": 0.5832,
+      "step": 11241
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 41.228875253478606,
+      "learning_rate": 7.742155723259086e-06,
+      "loss": 0.6266,
+      "step": 11242
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 42.824262139332234,
+      "learning_rate": 7.740512903787945e-06,
+      "loss": 0.5913,
+      "step": 11243
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 34.881372354986965,
+      "learning_rate": 7.738870148575299e-06,
+      "loss": 0.5065,
+      "step": 11244
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 45.97514182645029,
+      "learning_rate": 7.737227457667861e-06,
+      "loss": 0.6238,
+      "step": 11245
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 35.15853863196792,
+      "learning_rate": 7.735584831112352e-06,
+      "loss": 0.5328,
+      "step": 11246
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 43.90343502428581,
+      "learning_rate": 7.733942268955482e-06,
+      "loss": 0.5429,
+      "step": 11247
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 42.753478550067754,
+      "learning_rate": 7.732299771243972e-06,
+      "loss": 0.5458,
+      "step": 11248
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 33.22174830645338,
+      "learning_rate": 7.73065733802453e-06,
+      "loss": 0.5077,
+      "step": 11249
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 39.07135970694151,
+      "learning_rate": 7.729014969343863e-06,
+      "loss": 0.5967,
+      "step": 11250
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.71579788526158,
+      "learning_rate": 7.727372665248682e-06,
+      "loss": 0.6175,
+      "step": 11251
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 42.51406277355792,
+      "learning_rate": 7.725730425785694e-06,
+      "loss": 0.57,
+      "step": 11252
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 38.25337380757678,
+      "learning_rate": 7.724088251001602e-06,
+      "loss": 0.6064,
+      "step": 11253
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 43.36806461173619,
+      "learning_rate": 7.722446140943111e-06,
+      "loss": 0.6623,
+      "step": 11254
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 33.44890680850362,
+      "learning_rate": 7.720804095656918e-06,
+      "loss": 0.5527,
+      "step": 11255
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 44.921686990095694,
+      "learning_rate": 7.71916211518972e-06,
+      "loss": 0.6534,
+      "step": 11256
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.073094310652664,
+      "learning_rate": 7.717520199588222e-06,
+      "loss": 0.5916,
+      "step": 11257
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.739997999512404,
+      "learning_rate": 7.71587834889911e-06,
+      "loss": 0.5322,
+      "step": 11258
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 34.82932096321571,
+      "learning_rate": 7.714236563169088e-06,
+      "loss": 0.5277,
+      "step": 11259
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 47.55348807687199,
+      "learning_rate": 7.712594842444836e-06,
+      "loss": 0.6102,
+      "step": 11260
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 58.59979214452393,
+      "learning_rate": 7.710953186773048e-06,
+      "loss": 0.6634,
+      "step": 11261
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 41.31367292484877,
+      "learning_rate": 7.709311596200412e-06,
+      "loss": 0.6388,
+      "step": 11262
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 41.41067048282407,
+      "learning_rate": 7.707670070773616e-06,
+      "loss": 0.5789,
+      "step": 11263
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 46.769632832413585,
+      "learning_rate": 7.706028610539345e-06,
+      "loss": 0.6045,
+      "step": 11264
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.0504433020395,
+      "learning_rate": 7.704387215544276e-06,
+      "loss": 0.5281,
+      "step": 11265
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 44.092478706098206,
+      "learning_rate": 7.70274588583509e-06,
+      "loss": 0.6796,
+      "step": 11266
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 39.707535622954296,
+      "learning_rate": 7.701104621458467e-06,
+      "loss": 0.6089,
+      "step": 11267
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 38.03575848551893,
+      "learning_rate": 7.69946342246109e-06,
+      "loss": 0.5502,
+      "step": 11268
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 34.95303587546175,
+      "learning_rate": 7.697822288889617e-06,
+      "loss": 0.5008,
+      "step": 11269
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 39.806932122879715,
+      "learning_rate": 7.69618122079074e-06,
+      "loss": 0.5743,
+      "step": 11270
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 41.231291166675014,
+      "learning_rate": 7.694540218211117e-06,
+      "loss": 0.6207,
+      "step": 11271
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 49.56006730352178,
+      "learning_rate": 7.692899281197421e-06,
+      "loss": 0.6396,
+      "step": 11272
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 133.01357276755638,
+      "learning_rate": 7.691258409796324e-06,
+      "loss": 0.5771,
+      "step": 11273
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 38.09726081172781,
+      "learning_rate": 7.689617604054487e-06,
+      "loss": 0.5673,
+      "step": 11274
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 37.00866404466294,
+      "learning_rate": 7.68797686401857e-06,
+      "loss": 0.6087,
+      "step": 11275
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 41.95775044265844,
+      "learning_rate": 7.68633618973524e-06,
+      "loss": 0.6409,
+      "step": 11276
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 38.48172580523349,
+      "learning_rate": 7.684695581251153e-06,
+      "loss": 0.5402,
+      "step": 11277
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 33.34759214229081,
+      "learning_rate": 7.683055038612977e-06,
+      "loss": 0.5545,
+      "step": 11278
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.23386060663667,
+      "learning_rate": 7.681414561867353e-06,
+      "loss": 0.558,
+      "step": 11279
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 30.97668026376205,
+      "learning_rate": 7.679774151060945e-06,
+      "loss": 0.5347,
+      "step": 11280
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 35.51235582348601,
+      "learning_rate": 7.6781338062404e-06,
+      "loss": 0.6695,
+      "step": 11281
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 51.24034545490896,
+      "learning_rate": 7.676493527452374e-06,
+      "loss": 0.5774,
+      "step": 11282
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 36.03634443438293,
+      "learning_rate": 7.674853314743516e-06,
+      "loss": 0.5587,
+      "step": 11283
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 34.36228338668006,
+      "learning_rate": 7.673213168160465e-06,
+      "loss": 0.5491,
+      "step": 11284
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 34.991451115151364,
+      "learning_rate": 7.671573087749871e-06,
+      "loss": 0.5839,
+      "step": 11285
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 51.050071894102814,
+      "learning_rate": 7.669933073558377e-06,
+      "loss": 0.5228,
+      "step": 11286
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 33.97935540672912,
+      "learning_rate": 7.668293125632621e-06,
+      "loss": 0.5282,
+      "step": 11287
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 39.765944731800325,
+      "learning_rate": 7.66665324401925e-06,
+      "loss": 0.5745,
+      "step": 11288
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.95414036638595,
+      "learning_rate": 7.665013428764892e-06,
+      "loss": 0.5981,
+      "step": 11289
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 40.14057594495295,
+      "learning_rate": 7.663373679916187e-06,
+      "loss": 0.5576,
+      "step": 11290
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 35.524889276829946,
+      "learning_rate": 7.661733997519765e-06,
+      "loss": 0.5876,
+      "step": 11291
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 38.41006659655669,
+      "learning_rate": 7.660094381622265e-06,
+      "loss": 0.6082,
+      "step": 11292
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 36.26509303894794,
+      "learning_rate": 7.658454832270312e-06,
+      "loss": 0.5227,
+      "step": 11293
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 34.219795615189504,
+      "learning_rate": 7.656815349510528e-06,
+      "loss": 0.5441,
+      "step": 11294
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 44.67729054915311,
+      "learning_rate": 7.65517593338955e-06,
+      "loss": 0.6391,
+      "step": 11295
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 43.428061289979716,
+      "learning_rate": 7.653536583953993e-06,
+      "loss": 0.6014,
+      "step": 11296
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 37.226790186695084,
+      "learning_rate": 7.651897301250484e-06,
+      "loss": 0.5689,
+      "step": 11297
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 26.994342173629946,
+      "learning_rate": 7.650258085325646e-06,
+      "loss": 0.5005,
+      "step": 11298
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 35.9081844044487,
+      "learning_rate": 7.64861893622609e-06,
+      "loss": 0.5357,
+      "step": 11299
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.34082571734411,
+      "learning_rate": 7.646979853998434e-06,
+      "loss": 0.58,
+      "step": 11300
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.87750604810103,
+      "learning_rate": 7.645340838689297e-06,
+      "loss": 0.5594,
+      "step": 11301
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 39.24009238165151,
+      "learning_rate": 7.643701890345288e-06,
+      "loss": 0.6127,
+      "step": 11302
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.97388962857033,
+      "learning_rate": 7.642063009013025e-06,
+      "loss": 0.5152,
+      "step": 11303
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.87088741661393,
+      "learning_rate": 7.640424194739102e-06,
+      "loss": 0.5691,
+      "step": 11304
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.31855526138165,
+      "learning_rate": 7.63878544757014e-06,
+      "loss": 0.5584,
+      "step": 11305
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 42.64671440717619,
+      "learning_rate": 7.637146767552735e-06,
+      "loss": 0.6764,
+      "step": 11306
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.72529452824498,
+      "learning_rate": 7.635508154733494e-06,
+      "loss": 0.5521,
+      "step": 11307
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.2069827037905,
+      "learning_rate": 7.633869609159023e-06,
+      "loss": 0.5617,
+      "step": 11308
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 31.420928182046914,
+      "learning_rate": 7.632231130875912e-06,
+      "loss": 0.5063,
+      "step": 11309
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 41.18546913014088,
+      "learning_rate": 7.63059271993076e-06,
+      "loss": 0.6628,
+      "step": 11310
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.632493198132934,
+      "learning_rate": 7.628954376370169e-06,
+      "loss": 0.5396,
+      "step": 11311
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.394400105189305,
+      "learning_rate": 7.627316100240726e-06,
+      "loss": 0.6153,
+      "step": 11312
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 42.670946572762574,
+      "learning_rate": 7.625677891589029e-06,
+      "loss": 0.5803,
+      "step": 11313
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 35.652995717471846,
+      "learning_rate": 7.6240397504616615e-06,
+      "loss": 0.586,
+      "step": 11314
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 35.87807515219548,
+      "learning_rate": 7.622401676905214e-06,
+      "loss": 0.5434,
+      "step": 11315
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.244941007145876,
+      "learning_rate": 7.6207636709662694e-06,
+      "loss": 0.4487,
+      "step": 11316
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.4474735045018,
+      "learning_rate": 7.619125732691419e-06,
+      "loss": 0.554,
+      "step": 11317
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.52772853208358,
+      "learning_rate": 7.617487862127238e-06,
+      "loss": 0.5778,
+      "step": 11318
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 38.68087615712179,
+      "learning_rate": 7.615850059320306e-06,
+      "loss": 0.4792,
+      "step": 11319
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 39.70754032437264,
+      "learning_rate": 7.614212324317205e-06,
+      "loss": 0.6572,
+      "step": 11320
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.81222772092787,
+      "learning_rate": 7.612574657164509e-06,
+      "loss": 0.6085,
+      "step": 11321
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 40.529151168045864,
+      "learning_rate": 7.610937057908798e-06,
+      "loss": 0.5074,
+      "step": 11322
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 41.13123283577158,
+      "learning_rate": 7.609299526596634e-06,
+      "loss": 0.537,
+      "step": 11323
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 49.33650308920595,
+      "learning_rate": 7.607662063274595e-06,
+      "loss": 0.6376,
+      "step": 11324
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 49.86787647152061,
+      "learning_rate": 7.606024667989246e-06,
+      "loss": 0.5932,
+      "step": 11325
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 45.36296161974905,
+      "learning_rate": 7.604387340787156e-06,
+      "loss": 0.585,
+      "step": 11326
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 35.20587224691641,
+      "learning_rate": 7.6027500817148915e-06,
+      "loss": 0.5476,
+      "step": 11327
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.08004936419267,
+      "learning_rate": 7.6011128908190105e-06,
+      "loss": 0.552,
+      "step": 11328
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.44869043092385,
+      "learning_rate": 7.599475768146072e-06,
+      "loss": 0.6253,
+      "step": 11329
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 41.819029020314474,
+      "learning_rate": 7.5978387137426425e-06,
+      "loss": 0.6465,
+      "step": 11330
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 35.614990240232544,
+      "learning_rate": 7.59620172765527e-06,
+      "loss": 0.5664,
+      "step": 11331
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 39.64476454084785,
+      "learning_rate": 7.5945648099305206e-06,
+      "loss": 0.6344,
+      "step": 11332
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 39.62455331087295,
+      "learning_rate": 7.592927960614939e-06,
+      "loss": 0.5875,
+      "step": 11333
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.27641744571913,
+      "learning_rate": 7.591291179755077e-06,
+      "loss": 0.5053,
+      "step": 11334
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.32152966753049,
+      "learning_rate": 7.5896544673974825e-06,
+      "loss": 0.5703,
+      "step": 11335
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.59611949714243,
+      "learning_rate": 7.588017823588707e-06,
+      "loss": 0.5201,
+      "step": 11336
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 46.391244099778284,
+      "learning_rate": 7.586381248375296e-06,
+      "loss": 0.5391,
+      "step": 11337
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.2160081919991,
+      "learning_rate": 7.584744741803787e-06,
+      "loss": 0.5362,
+      "step": 11338
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 50.869527639614724,
+      "learning_rate": 7.583108303920726e-06,
+      "loss": 0.5462,
+      "step": 11339
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 38.123963157437885,
+      "learning_rate": 7.581471934772649e-06,
+      "loss": 0.5529,
+      "step": 11340
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 35.197898446279964,
+      "learning_rate": 7.579835634406097e-06,
+      "loss": 0.6248,
+      "step": 11341
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.6939660961889,
+      "learning_rate": 7.578199402867607e-06,
+      "loss": 0.5692,
+      "step": 11342
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.67087931762922,
+      "learning_rate": 7.576563240203707e-06,
+      "loss": 0.6096,
+      "step": 11343
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 32.16283859359358,
+      "learning_rate": 7.574927146460928e-06,
+      "loss": 0.4569,
+      "step": 11344
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 43.41558396984038,
+      "learning_rate": 7.573291121685805e-06,
+      "loss": 0.6247,
+      "step": 11345
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.58286649191764,
+      "learning_rate": 7.571655165924861e-06,
+      "loss": 0.5718,
+      "step": 11346
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 42.09145673524391,
+      "learning_rate": 7.570019279224629e-06,
+      "loss": 0.5577,
+      "step": 11347
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 35.36461827084928,
+      "learning_rate": 7.568383461631621e-06,
+      "loss": 0.5314,
+      "step": 11348
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 44.5115978364863,
+      "learning_rate": 7.566747713192368e-06,
+      "loss": 0.5835,
+      "step": 11349
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 49.229665518272775,
+      "learning_rate": 7.5651120339533836e-06,
+      "loss": 0.5368,
+      "step": 11350
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 32.211155308022434,
+      "learning_rate": 7.56347642396119e-06,
+      "loss": 0.5136,
+      "step": 11351
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 48.47377646261037,
+      "learning_rate": 7.5618408832623044e-06,
+      "loss": 0.6282,
+      "step": 11352
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 32.841277326137565,
+      "learning_rate": 7.5602054119032364e-06,
+      "loss": 0.4708,
+      "step": 11353
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 31.187257475282667,
+      "learning_rate": 7.558570009930497e-06,
+      "loss": 0.5635,
+      "step": 11354
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 42.344407181373214,
+      "learning_rate": 7.556934677390599e-06,
+      "loss": 0.5381,
+      "step": 11355
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 38.00300488386737,
+      "learning_rate": 7.555299414330048e-06,
+      "loss": 0.5468,
+      "step": 11356
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 36.13476628150484,
+      "learning_rate": 7.553664220795357e-06,
+      "loss": 0.5066,
+      "step": 11357
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 40.61087922219797,
+      "learning_rate": 7.5520290968330226e-06,
+      "loss": 0.5922,
+      "step": 11358
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 40.74319676496004,
+      "learning_rate": 7.550394042489547e-06,
+      "loss": 0.6819,
+      "step": 11359
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 31.29565653188333,
+      "learning_rate": 7.54875905781143e-06,
+      "loss": 0.4653,
+      "step": 11360
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 38.819505882366215,
+      "learning_rate": 7.547124142845174e-06,
+      "loss": 0.6437,
+      "step": 11361
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 41.52909617583401,
+      "learning_rate": 7.545489297637275e-06,
+      "loss": 0.5821,
+      "step": 11362
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 39.51793486215855,
+      "learning_rate": 7.54385452223422e-06,
+      "loss": 0.6161,
+      "step": 11363
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.601501123714556,
+      "learning_rate": 7.542219816682508e-06,
+      "loss": 0.5728,
+      "step": 11364
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.14815525979391,
+      "learning_rate": 7.5405851810286236e-06,
+      "loss": 0.6186,
+      "step": 11365
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 37.716561289496944,
+      "learning_rate": 7.538950615319065e-06,
+      "loss": 0.5878,
+      "step": 11366
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.71446265867434,
+      "learning_rate": 7.537316119600305e-06,
+      "loss": 0.6424,
+      "step": 11367
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 37.25073469320727,
+      "learning_rate": 7.535681693918836e-06,
+      "loss": 0.5495,
+      "step": 11368
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 37.33943365202871,
+      "learning_rate": 7.534047338321135e-06,
+      "loss": 0.5973,
+      "step": 11369
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.98577536625068,
+      "learning_rate": 7.5324130528536885e-06,
+      "loss": 0.509,
+      "step": 11370
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 33.245059793641985,
+      "learning_rate": 7.530778837562973e-06,
+      "loss": 0.6012,
+      "step": 11371
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.49728941808017,
+      "learning_rate": 7.5291446924954604e-06,
+      "loss": 0.5428,
+      "step": 11372
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.98122883747131,
+      "learning_rate": 7.527510617697627e-06,
+      "loss": 0.5742,
+      "step": 11373
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 41.66860826101152,
+      "learning_rate": 7.525876613215944e-06,
+      "loss": 0.6065,
+      "step": 11374
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.866300752386955,
+      "learning_rate": 7.524242679096884e-06,
+      "loss": 0.4965,
+      "step": 11375
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 32.787550350765734,
+      "learning_rate": 7.522608815386916e-06,
+      "loss": 0.5342,
+      "step": 11376
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.02347376429606,
+      "learning_rate": 7.520975022132503e-06,
+      "loss": 0.6053,
+      "step": 11377
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 36.60605312878694,
+      "learning_rate": 7.5193412993801096e-06,
+      "loss": 0.5187,
+      "step": 11378
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.74147324270705,
+      "learning_rate": 7.517707647176197e-06,
+      "loss": 0.5216,
+      "step": 11379
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 47.204993040870015,
+      "learning_rate": 7.5160740655672285e-06,
+      "loss": 0.7032,
+      "step": 11380
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 41.49771442209243,
+      "learning_rate": 7.514440554599662e-06,
+      "loss": 0.5365,
+      "step": 11381
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.940773941894875,
+      "learning_rate": 7.5128071143199485e-06,
+      "loss": 0.5041,
+      "step": 11382
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 33.217638949943876,
+      "learning_rate": 7.511173744774548e-06,
+      "loss": 0.539,
+      "step": 11383
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.128785058390136,
+      "learning_rate": 7.509540446009909e-06,
+      "loss": 0.4676,
+      "step": 11384
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.163622565442495,
+      "learning_rate": 7.5079072180724834e-06,
+      "loss": 0.5272,
+      "step": 11385
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.184288788930814,
+      "learning_rate": 7.506274061008721e-06,
+      "loss": 0.6011,
+      "step": 11386
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.181592306381354,
+      "learning_rate": 7.504640974865065e-06,
+      "loss": 0.6144,
+      "step": 11387
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.13588542314966,
+      "learning_rate": 7.503007959687956e-06,
+      "loss": 0.5478,
+      "step": 11388
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 31.97578233502966,
+      "learning_rate": 7.501375015523843e-06,
+      "loss": 0.469,
+      "step": 11389
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.56028253051537,
+      "learning_rate": 7.4997421424191595e-06,
+      "loss": 0.6507,
+      "step": 11390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 45.24749571893951,
+      "learning_rate": 7.498109340420354e-06,
+      "loss": 0.5788,
+      "step": 11391
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.08789721944626,
+      "learning_rate": 7.496476609573851e-06,
+      "loss": 0.5196,
+      "step": 11392
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 46.04781088704781,
+      "learning_rate": 7.494843949926088e-06,
+      "loss": 0.6094,
+      "step": 11393
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 39.05378004478489,
+      "learning_rate": 7.493211361523496e-06,
+      "loss": 0.6009,
+      "step": 11394
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 48.081203445199804,
+      "learning_rate": 7.49157884441251e-06,
+      "loss": 0.5266,
+      "step": 11395
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 42.460219379097296,
+      "learning_rate": 7.489946398639555e-06,
+      "loss": 0.6567,
+      "step": 11396
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 41.27639757293746,
+      "learning_rate": 7.488314024251055e-06,
+      "loss": 0.5597,
+      "step": 11397
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 31.77405423498384,
+      "learning_rate": 7.486681721293432e-06,
+      "loss": 0.5289,
+      "step": 11398
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.245930480141006,
+      "learning_rate": 7.485049489813112e-06,
+      "loss": 0.6063,
+      "step": 11399
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.811883931078725,
+      "learning_rate": 7.483417329856513e-06,
+      "loss": 0.5424,
+      "step": 11400
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.99400490741018,
+      "learning_rate": 7.481785241470055e-06,
+      "loss": 0.5578,
+      "step": 11401
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 44.288705202263955,
+      "learning_rate": 7.48015322470015e-06,
+      "loss": 0.6201,
+      "step": 11402
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 36.16518424258404,
+      "learning_rate": 7.478521279593213e-06,
+      "loss": 0.6014,
+      "step": 11403
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 43.22374473887035,
+      "learning_rate": 7.476889406195656e-06,
+      "loss": 0.5547,
+      "step": 11404
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.0835724398539,
+      "learning_rate": 7.4752576045538894e-06,
+      "loss": 0.5459,
+      "step": 11405
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 48.95531217898462,
+      "learning_rate": 7.473625874714322e-06,
+      "loss": 0.6587,
+      "step": 11406
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.22916653045512,
+      "learning_rate": 7.471994216723352e-06,
+      "loss": 0.547,
+      "step": 11407
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 41.409738999686304,
+      "learning_rate": 7.47036263062739e-06,
+      "loss": 0.5669,
+      "step": 11408
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.88912260264024,
+      "learning_rate": 7.4687311164728315e-06,
+      "loss": 0.5202,
+      "step": 11409
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 36.70685964708445,
+      "learning_rate": 7.467099674306083e-06,
+      "loss": 0.5595,
+      "step": 11410
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 33.774626482176586,
+      "learning_rate": 7.46546830417354e-06,
+      "loss": 0.5391,
+      "step": 11411
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.384674561296016,
+      "learning_rate": 7.463837006121593e-06,
+      "loss": 0.5804,
+      "step": 11412
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 37.8366953025418,
+      "learning_rate": 7.462205780196637e-06,
+      "loss": 0.5794,
+      "step": 11413
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.541837213843415,
+      "learning_rate": 7.460574626445065e-06,
+      "loss": 0.5811,
+      "step": 11414
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 50.148725930457296,
+      "learning_rate": 7.458943544913266e-06,
+      "loss": 0.7313,
+      "step": 11415
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.78338592314412,
+      "learning_rate": 7.457312535647627e-06,
+      "loss": 0.5398,
+      "step": 11416
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 44.09952547895007,
+      "learning_rate": 7.455681598694529e-06,
+      "loss": 0.6621,
+      "step": 11417
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 37.47790730417209,
+      "learning_rate": 7.454050734100358e-06,
+      "loss": 0.6242,
+      "step": 11418
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 46.25609893279347,
+      "learning_rate": 7.452419941911495e-06,
+      "loss": 0.4845,
+      "step": 11419
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 49.077139649311164,
+      "learning_rate": 7.450789222174322e-06,
+      "loss": 0.5718,
+      "step": 11420
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 32.358975439768415,
+      "learning_rate": 7.449158574935209e-06,
+      "loss": 0.5617,
+      "step": 11421
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.25481765033922,
+      "learning_rate": 7.4475280002405355e-06,
+      "loss": 0.5352,
+      "step": 11422
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 35.28166348170275,
+      "learning_rate": 7.445897498136671e-06,
+      "loss": 0.5066,
+      "step": 11423
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.08339626813456,
+      "learning_rate": 7.444267068669988e-06,
+      "loss": 0.5237,
+      "step": 11424
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 38.619609999877824,
+      "learning_rate": 7.442636711886857e-06,
+      "loss": 0.5783,
+      "step": 11425
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 32.488714934657395,
+      "learning_rate": 7.441006427833638e-06,
+      "loss": 0.5374,
+      "step": 11426
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.85132109150741,
+      "learning_rate": 7.439376216556702e-06,
+      "loss": 0.5788,
+      "step": 11427
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 46.711651812073185,
+      "learning_rate": 7.437746078102406e-06,
+      "loss": 0.5666,
+      "step": 11428
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 37.40405004369207,
+      "learning_rate": 7.4361160125171135e-06,
+      "loss": 0.5414,
+      "step": 11429
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 37.79377411844258,
+      "learning_rate": 7.4344860198471845e-06,
+      "loss": 0.4715,
+      "step": 11430
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.51050507901634,
+      "learning_rate": 7.432856100138971e-06,
+      "loss": 0.5589,
+      "step": 11431
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 33.87787942859808,
+      "learning_rate": 7.431226253438826e-06,
+      "loss": 0.5767,
+      "step": 11432
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.06145414954788,
+      "learning_rate": 7.4295964797931045e-06,
+      "loss": 0.5012,
+      "step": 11433
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.99910563886182,
+      "learning_rate": 7.427966779248155e-06,
+      "loss": 0.5966,
+      "step": 11434
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 45.33526045487466,
+      "learning_rate": 7.4263371518503314e-06,
+      "loss": 0.5299,
+      "step": 11435
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.09438260739676,
+      "learning_rate": 7.424707597645967e-06,
+      "loss": 0.5413,
+      "step": 11436
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 34.961020766857025,
+      "learning_rate": 7.423078116681415e-06,
+      "loss": 0.5059,
+      "step": 11437
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.85768511105779,
+      "learning_rate": 7.42144870900301e-06,
+      "loss": 0.5869,
+      "step": 11438
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.656011692843244,
+      "learning_rate": 7.419819374657099e-06,
+      "loss": 0.5299,
+      "step": 11439
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 35.50027987880254,
+      "learning_rate": 7.418190113690018e-06,
+      "loss": 0.5391,
+      "step": 11440
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 39.56016786158812,
+      "learning_rate": 7.416560926148096e-06,
+      "loss": 0.5383,
+      "step": 11441
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 39.14018013219751,
+      "learning_rate": 7.414931812077669e-06,
+      "loss": 0.5896,
+      "step": 11442
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.778772593143266,
+      "learning_rate": 7.413302771525071e-06,
+      "loss": 0.573,
+      "step": 11443
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 33.70926542765579,
+      "learning_rate": 7.4116738045366275e-06,
+      "loss": 0.4804,
+      "step": 11444
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.87360053594613,
+      "learning_rate": 7.410044911158671e-06,
+      "loss": 0.5516,
+      "step": 11445
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 65.98152481850067,
+      "learning_rate": 7.408416091437519e-06,
+      "loss": 0.5426,
+      "step": 11446
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 42.22052420354875,
+      "learning_rate": 7.406787345419496e-06,
+      "loss": 0.6024,
+      "step": 11447
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.84539946535354,
+      "learning_rate": 7.405158673150925e-06,
+      "loss": 0.5661,
+      "step": 11448
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.407119390541595,
+      "learning_rate": 7.403530074678124e-06,
+      "loss": 0.5298,
+      "step": 11449
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 44.048719932454745,
+      "learning_rate": 7.401901550047412e-06,
+      "loss": 0.56,
+      "step": 11450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 41.47468892326299,
+      "learning_rate": 7.400273099305095e-06,
+      "loss": 0.5187,
+      "step": 11451
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 39.04002924850611,
+      "learning_rate": 7.398644722497492e-06,
+      "loss": 0.501,
+      "step": 11452
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.41713447764984,
+      "learning_rate": 7.39701641967091e-06,
+      "loss": 0.5298,
+      "step": 11453
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 44.89494607790687,
+      "learning_rate": 7.39538819087166e-06,
+      "loss": 0.6752,
+      "step": 11454
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 37.699007482137304,
+      "learning_rate": 7.393760036146049e-06,
+      "loss": 0.6548,
+      "step": 11455
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.76437975667031,
+      "learning_rate": 7.392131955540375e-06,
+      "loss": 0.5661,
+      "step": 11456
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 34.7548240668199,
+      "learning_rate": 7.39050394910094e-06,
+      "loss": 0.5512,
+      "step": 11457
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.15208764544535,
+      "learning_rate": 7.388876016874049e-06,
+      "loss": 0.5701,
+      "step": 11458
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 50.53684671910535,
+      "learning_rate": 7.3872481589059955e-06,
+      "loss": 0.634,
+      "step": 11459
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 37.50002974826905,
+      "learning_rate": 7.3856203752430815e-06,
+      "loss": 0.6399,
+      "step": 11460
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 41.54231869323931,
+      "learning_rate": 7.383992665931587e-06,
+      "loss": 0.5439,
+      "step": 11461
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.62722976263582,
+      "learning_rate": 7.382365031017815e-06,
+      "loss": 0.5253,
+      "step": 11462
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.601504409686825,
+      "learning_rate": 7.380737470548047e-06,
+      "loss": 0.4877,
+      "step": 11463
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 43.048860103300555,
+      "learning_rate": 7.379109984568578e-06,
+      "loss": 0.5877,
+      "step": 11464
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 39.035911242023445,
+      "learning_rate": 7.377482573125685e-06,
+      "loss": 0.5638,
+      "step": 11465
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 31.52558456728981,
+      "learning_rate": 7.375855236265653e-06,
+      "loss": 0.5159,
+      "step": 11466
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.88224150683144,
+      "learning_rate": 7.374227974034762e-06,
+      "loss": 0.4978,
+      "step": 11467
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 39.014845153201136,
+      "learning_rate": 7.372600786479292e-06,
+      "loss": 0.5825,
+      "step": 11468
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 35.34913537561115,
+      "learning_rate": 7.370973673645523e-06,
+      "loss": 0.5104,
+      "step": 11469
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 42.479675684048296,
+      "learning_rate": 7.369346635579719e-06,
+      "loss": 0.5679,
+      "step": 11470
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 36.43141807324979,
+      "learning_rate": 7.36771967232816e-06,
+      "loss": 0.5693,
+      "step": 11471
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 44.35369434645727,
+      "learning_rate": 7.366092783937112e-06,
+      "loss": 0.6666,
+      "step": 11472
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.855439221555386,
+      "learning_rate": 7.364465970452846e-06,
+      "loss": 0.6442,
+      "step": 11473
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 35.84913027491808,
+      "learning_rate": 7.362839231921629e-06,
+      "loss": 0.4833,
+      "step": 11474
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 44.50788947686826,
+      "learning_rate": 7.36121256838972e-06,
+      "loss": 0.5122,
+      "step": 11475
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 34.04657046710122,
+      "learning_rate": 7.359585979903378e-06,
+      "loss": 0.504,
+      "step": 11476
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 33.20352572210509,
+      "learning_rate": 7.357959466508869e-06,
+      "loss": 0.5214,
+      "step": 11477
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 41.88440546337632,
+      "learning_rate": 7.356333028252445e-06,
+      "loss": 0.6164,
+      "step": 11478
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 46.130871122567,
+      "learning_rate": 7.35470666518037e-06,
+      "loss": 0.5775,
+      "step": 11479
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 37.03698127272611,
+      "learning_rate": 7.353080377338884e-06,
+      "loss": 0.5371,
+      "step": 11480
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 40.727030526199044,
+      "learning_rate": 7.351454164774246e-06,
+      "loss": 0.5158,
+      "step": 11481
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 28.566449779141927,
+      "learning_rate": 7.349828027532697e-06,
+      "loss": 0.4778,
+      "step": 11482
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.30279037131475,
+      "learning_rate": 7.348201965660493e-06,
+      "loss": 0.5607,
+      "step": 11483
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 35.663495649029954,
+      "learning_rate": 7.346575979203876e-06,
+      "loss": 0.4521,
+      "step": 11484
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 29.159558761617795,
+      "learning_rate": 7.344950068209085e-06,
+      "loss": 0.4961,
+      "step": 11485
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 47.05598122890738,
+      "learning_rate": 7.343324232722358e-06,
+      "loss": 0.6138,
+      "step": 11486
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 42.30790937136273,
+      "learning_rate": 7.341698472789937e-06,
+      "loss": 0.519,
+      "step": 11487
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 30.76598290912833,
+      "learning_rate": 7.340072788458054e-06,
+      "loss": 0.487,
+      "step": 11488
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 35.31241734925447,
+      "learning_rate": 7.33844717977295e-06,
+      "loss": 0.5183,
+      "step": 11489
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.754350467239604,
+      "learning_rate": 7.336821646780848e-06,
+      "loss": 0.5902,
+      "step": 11490
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 43.68513183287438,
+      "learning_rate": 7.33519618952798e-06,
+      "loss": 0.6529,
+      "step": 11491
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.97929366207793,
+      "learning_rate": 7.333570808060575e-06,
+      "loss": 0.4973,
+      "step": 11492
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 49.47507627692374,
+      "learning_rate": 7.331945502424856e-06,
+      "loss": 0.6351,
+      "step": 11493
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 40.57657464627397,
+      "learning_rate": 7.330320272667048e-06,
+      "loss": 0.6313,
+      "step": 11494
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.201764326021916,
+      "learning_rate": 7.328695118833366e-06,
+      "loss": 0.5321,
+      "step": 11495
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 37.91604484565232,
+      "learning_rate": 7.327070040970034e-06,
+      "loss": 0.5697,
+      "step": 11496
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 38.39756099286524,
+      "learning_rate": 7.325445039123264e-06,
+      "loss": 0.5714,
+      "step": 11497
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.74339681456597,
+      "learning_rate": 7.323820113339275e-06,
+      "loss": 0.5392,
+      "step": 11498
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 47.11580635567733,
+      "learning_rate": 7.32219526366428e-06,
+      "loss": 0.6425,
+      "step": 11499
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.737392465470535,
+      "learning_rate": 7.320570490144481e-06,
+      "loss": 0.5962,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 41.40195041429703,
+      "learning_rate": 7.318945792826089e-06,
+      "loss": 0.6617,
+      "step": 11501
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 32.22861465855934,
+      "learning_rate": 7.317321171755312e-06,
+      "loss": 0.5176,
+      "step": 11502
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 44.6626164334197,
+      "learning_rate": 7.315696626978349e-06,
+      "loss": 0.6607,
+      "step": 11503
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.52994401003315,
+      "learning_rate": 7.31407215854141e-06,
+      "loss": 0.5674,
+      "step": 11504
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 43.45830027443984,
+      "learning_rate": 7.312447766490682e-06,
+      "loss": 0.6477,
+      "step": 11505
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 40.092003175497666,
+      "learning_rate": 7.310823450872368e-06,
+      "loss": 0.4985,
+      "step": 11506
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 35.92485963519236,
+      "learning_rate": 7.309199211732662e-06,
+      "loss": 0.4906,
+      "step": 11507
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 38.69198691766664,
+      "learning_rate": 7.307575049117758e-06,
+      "loss": 0.5529,
+      "step": 11508
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 33.169365628489814,
+      "learning_rate": 7.305950963073845e-06,
+      "loss": 0.4823,
+      "step": 11509
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 40.353863772921464,
+      "learning_rate": 7.30432695364711e-06,
+      "loss": 0.6261,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.24592382302501,
+      "learning_rate": 7.3027030208837365e-06,
+      "loss": 0.52,
+      "step": 11511
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 37.887442918065545,
+      "learning_rate": 7.301079164829914e-06,
+      "loss": 0.5933,
+      "step": 11512
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 68.30802083316455,
+      "learning_rate": 7.299455385531824e-06,
+      "loss": 0.5436,
+      "step": 11513
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.764092860375136,
+      "learning_rate": 7.297831683035638e-06,
+      "loss": 0.493,
+      "step": 11514
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 44.64672390109955,
+      "learning_rate": 7.296208057387539e-06,
+      "loss": 0.6734,
+      "step": 11515
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.944656156508536,
+      "learning_rate": 7.294584508633702e-06,
+      "loss": 0.553,
+      "step": 11516
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 49.79538508407568,
+      "learning_rate": 7.292961036820299e-06,
+      "loss": 0.621,
+      "step": 11517
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 42.59393654752366,
+      "learning_rate": 7.291337641993503e-06,
+      "loss": 0.5764,
+      "step": 11518
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 33.74895742713422,
+      "learning_rate": 7.289714324199477e-06,
+      "loss": 0.5412,
+      "step": 11519
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.74668458391669,
+      "learning_rate": 7.28809108348439e-06,
+      "loss": 0.6082,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 42.373647212487846,
+      "learning_rate": 7.2864679198944065e-06,
+      "loss": 0.5745,
+      "step": 11521
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 40.30174305840298,
+      "learning_rate": 7.284844833475687e-06,
+      "loss": 0.5839,
+      "step": 11522
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 37.30221846313771,
+      "learning_rate": 7.2832218242743976e-06,
+      "loss": 0.5755,
+      "step": 11523
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.75679696456475,
+      "learning_rate": 7.281598892336685e-06,
+      "loss": 0.4852,
+      "step": 11524
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.03907799868157,
+      "learning_rate": 7.279976037708712e-06,
+      "loss": 0.5068,
+      "step": 11525
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.64825101441878,
+      "learning_rate": 7.278353260436626e-06,
+      "loss": 0.5386,
+      "step": 11526
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.7945129219533,
+      "learning_rate": 7.276730560566583e-06,
+      "loss": 0.6158,
+      "step": 11527
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 47.528540188542706,
+      "learning_rate": 7.275107938144732e-06,
+      "loss": 0.7283,
+      "step": 11528
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.88084850483845,
+      "learning_rate": 7.273485393217217e-06,
+      "loss": 0.5404,
+      "step": 11529
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.21625327867702,
+      "learning_rate": 7.2718629258301786e-06,
+      "loss": 0.5229,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 43.782804348546215,
+      "learning_rate": 7.270240536029765e-06,
+      "loss": 0.6323,
+      "step": 11531
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.29660740907239,
+      "learning_rate": 7.268618223862112e-06,
+      "loss": 0.5614,
+      "step": 11532
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 38.85145729920569,
+      "learning_rate": 7.266995989373362e-06,
+      "loss": 0.5859,
+      "step": 11533
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 40.46688945131287,
+      "learning_rate": 7.265373832609647e-06,
+      "loss": 0.5682,
+      "step": 11534
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.659137280406945,
+      "learning_rate": 7.263751753617096e-06,
+      "loss": 0.5758,
+      "step": 11535
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 41.85573600509272,
+      "learning_rate": 7.262129752441849e-06,
+      "loss": 0.6264,
+      "step": 11536
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 43.52217022145224,
+      "learning_rate": 7.26050782913003e-06,
+      "loss": 0.6515,
+      "step": 11537
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 38.77083819206709,
+      "learning_rate": 7.258885983727767e-06,
+      "loss": 0.5679,
+      "step": 11538
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 37.73503033876495,
+      "learning_rate": 7.257264216281179e-06,
+      "loss": 0.6122,
+      "step": 11539
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 37.320590380136615,
+      "learning_rate": 7.255642526836395e-06,
+      "loss": 0.5869,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 48.67598829201651,
+      "learning_rate": 7.254020915439528e-06,
+      "loss": 0.6897,
+      "step": 11541
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 35.20920284747509,
+      "learning_rate": 7.252399382136703e-06,
+      "loss": 0.596,
+      "step": 11542
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.493017699466606,
+      "learning_rate": 7.250777926974035e-06,
+      "loss": 0.5208,
+      "step": 11543
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 35.587500280575206,
+      "learning_rate": 7.2491565499976335e-06,
+      "loss": 0.544,
+      "step": 11544
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 33.773983268201896,
+      "learning_rate": 7.247535251253606e-06,
+      "loss": 0.4758,
+      "step": 11545
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 38.47510832386495,
+      "learning_rate": 7.245914030788069e-06,
+      "loss": 0.5799,
+      "step": 11546
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.87631556863569,
+      "learning_rate": 7.2442928886471246e-06,
+      "loss": 0.5397,
+      "step": 11547
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 41.8922377962076,
+      "learning_rate": 7.242671824876884e-06,
+      "loss": 0.5938,
+      "step": 11548
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 42.58800575889834,
+      "learning_rate": 7.241050839523437e-06,
+      "loss": 0.6757,
+      "step": 11549
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 32.68088511972572,
+      "learning_rate": 7.239429932632892e-06,
+      "loss": 0.5091,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.80661110657935,
+      "learning_rate": 7.237809104251343e-06,
+      "loss": 0.5719,
+      "step": 11551
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 43.644192643688505,
+      "learning_rate": 7.23618835442489e-06,
+      "loss": 0.6077,
+      "step": 11552
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 38.89402116757302,
+      "learning_rate": 7.234567683199624e-06,
+      "loss": 0.5533,
+      "step": 11553
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.32969796661496,
+      "learning_rate": 7.232947090621633e-06,
+      "loss": 0.6334,
+      "step": 11554
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 43.460824837710895,
+      "learning_rate": 7.23132657673701e-06,
+      "loss": 0.6123,
+      "step": 11555
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.942501619423076,
+      "learning_rate": 7.2297061415918394e-06,
+      "loss": 0.621,
+      "step": 11556
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.78983592457162,
+      "learning_rate": 7.228085785232206e-06,
+      "loss": 0.5638,
+      "step": 11557
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.53496387096188,
+      "learning_rate": 7.226465507704189e-06,
+      "loss": 0.5945,
+      "step": 11558
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.743752245978584,
+      "learning_rate": 7.224845309053872e-06,
+      "loss": 0.5314,
+      "step": 11559
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.803006721626936,
+      "learning_rate": 7.2232251893273295e-06,
+      "loss": 0.6279,
+      "step": 11560
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 35.09321849281233,
+      "learning_rate": 7.2216051485706385e-06,
+      "loss": 0.5056,
+      "step": 11561
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 34.450023444047325,
+      "learning_rate": 7.219985186829877e-06,
+      "loss": 0.5939,
+      "step": 11562
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.53166964039469,
+      "learning_rate": 7.2183653041511045e-06,
+      "loss": 0.5498,
+      "step": 11563
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.09376046316813,
+      "learning_rate": 7.216745500580396e-06,
+      "loss": 0.5286,
+      "step": 11564
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 40.53494614076783,
+      "learning_rate": 7.215125776163818e-06,
+      "loss": 0.5955,
+      "step": 11565
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.08155186611277,
+      "learning_rate": 7.213506130947434e-06,
+      "loss": 0.5228,
+      "step": 11566
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.886095585680025,
+      "learning_rate": 7.21188656497731e-06,
+      "loss": 0.6014,
+      "step": 11567
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.301291895154584,
+      "learning_rate": 7.210267078299495e-06,
+      "loss": 0.5521,
+      "step": 11568
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.18182323360822,
+      "learning_rate": 7.208647670960055e-06,
+      "loss": 0.5067,
+      "step": 11569
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 45.27913470754036,
+      "learning_rate": 7.2070283430050405e-06,
+      "loss": 0.6371,
+      "step": 11570
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 44.74130533240274,
+      "learning_rate": 7.205409094480506e-06,
+      "loss": 0.6087,
+      "step": 11571
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.21841056841434,
+      "learning_rate": 7.203789925432507e-06,
+      "loss": 0.6406,
+      "step": 11572
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.149054234683476,
+      "learning_rate": 7.202170835907083e-06,
+      "loss": 0.5131,
+      "step": 11573
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 41.15006350376796,
+      "learning_rate": 7.200551825950281e-06,
+      "loss": 0.5839,
+      "step": 11574
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 42.82228432945269,
+      "learning_rate": 7.198932895608153e-06,
+      "loss": 0.5729,
+      "step": 11575
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 42.66480725428927,
+      "learning_rate": 7.197314044926729e-06,
+      "loss": 0.539,
+      "step": 11576
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 42.20948549802515,
+      "learning_rate": 7.1956952739520605e-06,
+      "loss": 0.5767,
+      "step": 11577
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 32.39718528371568,
+      "learning_rate": 7.194076582730174e-06,
+      "loss": 0.5479,
+      "step": 11578
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.04800076544388,
+      "learning_rate": 7.192457971307107e-06,
+      "loss": 0.5378,
+      "step": 11579
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 41.44293035192284,
+      "learning_rate": 7.190839439728896e-06,
+      "loss": 0.6094,
+      "step": 11580
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 46.69016828150846,
+      "learning_rate": 7.1892209880415665e-06,
+      "loss": 0.4813,
+      "step": 11581
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 37.796469791506404,
+      "learning_rate": 7.18760261629115e-06,
+      "loss": 0.6499,
+      "step": 11582
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 35.983190933603645,
+      "learning_rate": 7.185984324523667e-06,
+      "loss": 0.5301,
+      "step": 11583
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 37.498062843515264,
+      "learning_rate": 7.184366112785144e-06,
+      "loss": 0.5373,
+      "step": 11584
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 35.0578164867214,
+      "learning_rate": 7.1827479811216e-06,
+      "loss": 0.5575,
+      "step": 11585
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.436192650000244,
+      "learning_rate": 7.181129929579058e-06,
+      "loss": 0.5393,
+      "step": 11586
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 37.68395070938694,
+      "learning_rate": 7.179511958203535e-06,
+      "loss": 0.4877,
+      "step": 11587
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 37.59398019391976,
+      "learning_rate": 7.177894067041038e-06,
+      "loss": 0.5839,
+      "step": 11588
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.60547909810335,
+      "learning_rate": 7.176276256137582e-06,
+      "loss": 0.6644,
+      "step": 11589
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 41.85509986590282,
+      "learning_rate": 7.174658525539179e-06,
+      "loss": 0.6273,
+      "step": 11590
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 30.053823769525703,
+      "learning_rate": 7.173040875291832e-06,
+      "loss": 0.5476,
+      "step": 11591
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 31.584712704621097,
+      "learning_rate": 7.171423305441556e-06,
+      "loss": 0.5016,
+      "step": 11592
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 33.27037889834689,
+      "learning_rate": 7.169805816034341e-06,
+      "loss": 0.4762,
+      "step": 11593
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.594461119399476,
+      "learning_rate": 7.168188407116194e-06,
+      "loss": 0.6043,
+      "step": 11594
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.76706688262415,
+      "learning_rate": 7.166571078733109e-06,
+      "loss": 0.5641,
+      "step": 11595
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.6352349429616,
+      "learning_rate": 7.164953830931089e-06,
+      "loss": 0.5829,
+      "step": 11596
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.168856299972354,
+      "learning_rate": 7.1633366637561245e-06,
+      "loss": 0.542,
+      "step": 11597
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 34.1657936709609,
+      "learning_rate": 7.161719577254203e-06,
+      "loss": 0.5664,
+      "step": 11598
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.18471553159564,
+      "learning_rate": 7.160102571471317e-06,
+      "loss": 0.5289,
+      "step": 11599
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.95797517331285,
+      "learning_rate": 7.158485646453452e-06,
+      "loss": 0.5256,
+      "step": 11600
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 45.497663343605765,
+      "learning_rate": 7.156868802246593e-06,
+      "loss": 0.5676,
+      "step": 11601
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 37.37138249359596,
+      "learning_rate": 7.155252038896725e-06,
+      "loss": 0.5304,
+      "step": 11602
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.30029590801046,
+      "learning_rate": 7.153635356449824e-06,
+      "loss": 0.51,
+      "step": 11603
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 30.62411563141337,
+      "learning_rate": 7.152018754951864e-06,
+      "loss": 0.5131,
+      "step": 11604
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.558076082328775,
+      "learning_rate": 7.150402234448829e-06,
+      "loss": 0.5679,
+      "step": 11605
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 37.653077044896925,
+      "learning_rate": 7.1487857949866875e-06,
+      "loss": 0.4972,
+      "step": 11606
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 34.261476083769665,
+      "learning_rate": 7.14716943661141e-06,
+      "loss": 0.5878,
+      "step": 11607
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.234696263318064,
+      "learning_rate": 7.145553159368962e-06,
+      "loss": 0.5147,
+      "step": 11608
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 36.52640922002814,
+      "learning_rate": 7.143936963305314e-06,
+      "loss": 0.5358,
+      "step": 11609
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 48.89294144289682,
+      "learning_rate": 7.142320848466427e-06,
+      "loss": 0.5241,
+      "step": 11610
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.7349653181417,
+      "learning_rate": 7.140704814898268e-06,
+      "loss": 0.5533,
+      "step": 11611
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 41.0481861545473,
+      "learning_rate": 7.1390888626467855e-06,
+      "loss": 0.5711,
+      "step": 11612
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.53193951660394,
+      "learning_rate": 7.137472991757945e-06,
+      "loss": 0.5035,
+      "step": 11613
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 33.43888914484809,
+      "learning_rate": 7.1358572022776965e-06,
+      "loss": 0.5976,
+      "step": 11614
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 41.69983510046539,
+      "learning_rate": 7.134241494251994e-06,
+      "loss": 0.5479,
+      "step": 11615
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 41.58842669212136,
+      "learning_rate": 7.1326258677267916e-06,
+      "loss": 0.5466,
+      "step": 11616
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 38.718286913303054,
+      "learning_rate": 7.131010322748029e-06,
+      "loss": 0.4733,
+      "step": 11617
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 43.264799584766905,
+      "learning_rate": 7.129394859361652e-06,
+      "loss": 0.5587,
+      "step": 11618
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.988837317725284,
+      "learning_rate": 7.127779477613608e-06,
+      "loss": 0.5685,
+      "step": 11619
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 54.594013215660375,
+      "learning_rate": 7.126164177549832e-06,
+      "loss": 0.6004,
+      "step": 11620
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.178755871551814,
+      "learning_rate": 7.124548959216272e-06,
+      "loss": 0.5638,
+      "step": 11621
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.84555050057804,
+      "learning_rate": 7.122933822658855e-06,
+      "loss": 0.5274,
+      "step": 11622
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.69126197918973,
+      "learning_rate": 7.121318767923514e-06,
+      "loss": 0.5177,
+      "step": 11623
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.038961246644156,
+      "learning_rate": 7.1197037950561855e-06,
+      "loss": 0.5351,
+      "step": 11624
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.7065581644101,
+      "learning_rate": 7.118088904102798e-06,
+      "loss": 0.548,
+      "step": 11625
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.9801326740569,
+      "learning_rate": 7.116474095109276e-06,
+      "loss": 0.5308,
+      "step": 11626
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.41583083139004,
+      "learning_rate": 7.11485936812154e-06,
+      "loss": 0.5376,
+      "step": 11627
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 41.88582939065408,
+      "learning_rate": 7.113244723185519e-06,
+      "loss": 0.5002,
+      "step": 11628
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.62256638816775,
+      "learning_rate": 7.111630160347127e-06,
+      "loss": 0.6178,
+      "step": 11629
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 46.02676317421349,
+      "learning_rate": 7.110015679652286e-06,
+      "loss": 0.5445,
+      "step": 11630
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.37512469734313,
+      "learning_rate": 7.1084012811469105e-06,
+      "loss": 0.5309,
+      "step": 11631
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.94033527058452,
+      "learning_rate": 7.106786964876909e-06,
+      "loss": 0.6712,
+      "step": 11632
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.96855072854648,
+      "learning_rate": 7.10517273088819e-06,
+      "loss": 0.6583,
+      "step": 11633
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 38.4940197746435,
+      "learning_rate": 7.103558579226668e-06,
+      "loss": 0.5846,
+      "step": 11634
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 43.47577976231671,
+      "learning_rate": 7.101944509938244e-06,
+      "loss": 0.6069,
+      "step": 11635
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 43.467931919366485,
+      "learning_rate": 7.100330523068829e-06,
+      "loss": 0.6003,
+      "step": 11636
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.86753622362054,
+      "learning_rate": 7.098716618664312e-06,
+      "loss": 0.5315,
+      "step": 11637
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 34.835169826278126,
+      "learning_rate": 7.097102796770598e-06,
+      "loss": 0.5056,
+      "step": 11638
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.8932337612031,
+      "learning_rate": 7.0954890574335814e-06,
+      "loss": 0.5943,
+      "step": 11639
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.90848325863385,
+      "learning_rate": 7.0938754006991584e-06,
+      "loss": 0.5109,
+      "step": 11640
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 34.42087010783883,
+      "learning_rate": 7.092261826613221e-06,
+      "loss": 0.4584,
+      "step": 11641
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 41.13020630900447,
+      "learning_rate": 7.0906483352216525e-06,
+      "loss": 0.5851,
+      "step": 11642
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 29.177472252360676,
+      "learning_rate": 7.0890349265703455e-06,
+      "loss": 0.5146,
+      "step": 11643
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.063351840077154,
+      "learning_rate": 7.0874216007051826e-06,
+      "loss": 0.6076,
+      "step": 11644
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 46.18255238236972,
+      "learning_rate": 7.0858083576720416e-06,
+      "loss": 0.5322,
+      "step": 11645
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 41.11500924798861,
+      "learning_rate": 7.084195197516812e-06,
+      "loss": 0.5805,
+      "step": 11646
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.365719352907085,
+      "learning_rate": 7.082582120285363e-06,
+      "loss": 0.5038,
+      "step": 11647
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 44.01833151288405,
+      "learning_rate": 7.080969126023569e-06,
+      "loss": 0.6247,
+      "step": 11648
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 47.33241101942391,
+      "learning_rate": 7.079356214777309e-06,
+      "loss": 0.6238,
+      "step": 11649
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 46.34755562033099,
+      "learning_rate": 7.077743386592448e-06,
+      "loss": 0.6089,
+      "step": 11650
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 38.091088553282326,
+      "learning_rate": 7.07613064151486e-06,
+      "loss": 0.5648,
+      "step": 11651
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.12783090264441,
+      "learning_rate": 7.0745179795904005e-06,
+      "loss": 0.5565,
+      "step": 11652
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.0931349659452,
+      "learning_rate": 7.07290540086494e-06,
+      "loss": 0.5291,
+      "step": 11653
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.29314006321151,
+      "learning_rate": 7.071292905384336e-06,
+      "loss": 0.5594,
+      "step": 11654
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 38.28753359236133,
+      "learning_rate": 7.069680493194455e-06,
+      "loss": 0.4855,
+      "step": 11655
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 51.36995701971292,
+      "learning_rate": 7.068068164341139e-06,
+      "loss": 0.606,
+      "step": 11656
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 40.55351074839555,
+      "learning_rate": 7.066455918870254e-06,
+      "loss": 0.6169,
+      "step": 11657
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 75.19981905504534,
+      "learning_rate": 7.064843756827641e-06,
+      "loss": 0.5966,
+      "step": 11658
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 43.99533407169657,
+      "learning_rate": 7.06323167825916e-06,
+      "loss": 0.6255,
+      "step": 11659
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.90168698577372,
+      "learning_rate": 7.0616196832106535e-06,
+      "loss": 0.5318,
+      "step": 11660
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.58924689485234,
+      "learning_rate": 7.060007771727961e-06,
+      "loss": 0.7086,
+      "step": 11661
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.3117709797277,
+      "learning_rate": 7.058395943856926e-06,
+      "loss": 0.6272,
+      "step": 11662
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.287014885835625,
+      "learning_rate": 7.056784199643392e-06,
+      "loss": 0.5649,
+      "step": 11663
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.01377206202232,
+      "learning_rate": 7.05517253913319e-06,
+      "loss": 0.6234,
+      "step": 11664
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 38.24238498780923,
+      "learning_rate": 7.053560962372166e-06,
+      "loss": 0.5659,
+      "step": 11665
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 34.90702920383802,
+      "learning_rate": 7.051949469406139e-06,
+      "loss": 0.4745,
+      "step": 11666
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 30.980280176112363,
+      "learning_rate": 7.0503380602809435e-06,
+      "loss": 0.5241,
+      "step": 11667
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 41.29707922363589,
+      "learning_rate": 7.04872673504241e-06,
+      "loss": 0.6158,
+      "step": 11668
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.893916389184476,
+      "learning_rate": 7.047115493736363e-06,
+      "loss": 0.5306,
+      "step": 11669
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 32.94675692240706,
+      "learning_rate": 7.045504336408625e-06,
+      "loss": 0.5451,
+      "step": 11670
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.19836470306435,
+      "learning_rate": 7.043893263105012e-06,
+      "loss": 0.5717,
+      "step": 11671
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.61097804624843,
+      "learning_rate": 7.042282273871346e-06,
+      "loss": 0.5246,
+      "step": 11672
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.48421448335747,
+      "learning_rate": 7.04067136875344e-06,
+      "loss": 0.5465,
+      "step": 11673
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 49.25070781973683,
+      "learning_rate": 7.039060547797112e-06,
+      "loss": 0.5741,
+      "step": 11674
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.706756201934084,
+      "learning_rate": 7.037449811048172e-06,
+      "loss": 0.559,
+      "step": 11675
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.84824778849702,
+      "learning_rate": 7.035839158552424e-06,
+      "loss": 0.5536,
+      "step": 11676
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.50496745614249,
+      "learning_rate": 7.034228590355674e-06,
+      "loss": 0.5385,
+      "step": 11677
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 34.12848642027057,
+      "learning_rate": 7.03261810650373e-06,
+      "loss": 0.5313,
+      "step": 11678
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 42.731011278178336,
+      "learning_rate": 7.03100770704239e-06,
+      "loss": 0.5596,
+      "step": 11679
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.55624625289289,
+      "learning_rate": 7.029397392017461e-06,
+      "loss": 0.551,
+      "step": 11680
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.60236946031053,
+      "learning_rate": 7.027787161474724e-06,
+      "loss": 0.516,
+      "step": 11681
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 38.09887995300088,
+      "learning_rate": 7.026177015459984e-06,
+      "loss": 0.469,
+      "step": 11682
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 41.39100563850649,
+      "learning_rate": 7.024566954019028e-06,
+      "loss": 0.598,
+      "step": 11683
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 31.5473983290458,
+      "learning_rate": 7.0229569771976505e-06,
+      "loss": 0.5202,
+      "step": 11684
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 34.54085835514605,
+      "learning_rate": 7.021347085041636e-06,
+      "loss": 0.5389,
+      "step": 11685
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.15186770995192,
+      "learning_rate": 7.019737277596763e-06,
+      "loss": 0.587,
+      "step": 11686
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.51670295027802,
+      "learning_rate": 7.018127554908823e-06,
+      "loss": 0.5577,
+      "step": 11687
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 44.74666108296666,
+      "learning_rate": 7.016517917023588e-06,
+      "loss": 0.5679,
+      "step": 11688
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 45.871750537218276,
+      "learning_rate": 7.014908363986839e-06,
+      "loss": 0.6098,
+      "step": 11689
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 43.79696492768717,
+      "learning_rate": 7.013298895844352e-06,
+      "loss": 0.6218,
+      "step": 11690
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 31.328660419103016,
+      "learning_rate": 7.011689512641896e-06,
+      "loss": 0.463,
+      "step": 11691
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 40.65507462522943,
+      "learning_rate": 7.01008021442524e-06,
+      "loss": 0.6019,
+      "step": 11692
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 31.9363666564753,
+      "learning_rate": 7.008471001240156e-06,
+      "loss": 0.6024,
+      "step": 11693
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 33.85507023392923,
+      "learning_rate": 7.006861873132406e-06,
+      "loss": 0.4915,
+      "step": 11694
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 53.87018402595026,
+      "learning_rate": 7.005252830147756e-06,
+      "loss": 0.5914,
+      "step": 11695
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 35.396670487450336,
+      "learning_rate": 7.00364387233196e-06,
+      "loss": 0.4391,
+      "step": 11696
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 38.249885586960154,
+      "learning_rate": 7.00203499973078e-06,
+      "loss": 0.5745,
+      "step": 11697
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 40.02931724784204,
+      "learning_rate": 7.00042621238997e-06,
+      "loss": 0.536,
+      "step": 11698
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 41.2224253388922,
+      "learning_rate": 6.998817510355285e-06,
+      "loss": 0.5734,
+      "step": 11699
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 40.795419878158846,
+      "learning_rate": 6.997208893672478e-06,
+      "loss": 0.5261,
+      "step": 11700
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 38.185201263541686,
+      "learning_rate": 6.9956003623872895e-06,
+      "loss": 0.505,
+      "step": 11701
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 36.26280351976608,
+      "learning_rate": 6.993991916545468e-06,
+      "loss": 0.5713,
+      "step": 11702
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.27734632750887,
+      "learning_rate": 6.99238355619276e-06,
+      "loss": 0.5206,
+      "step": 11703
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 42.12663992001706,
+      "learning_rate": 6.990775281374907e-06,
+      "loss": 0.5334,
+      "step": 11704
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 38.37662779744751,
+      "learning_rate": 6.989167092137639e-06,
+      "loss": 0.5738,
+      "step": 11705
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 42.78403579190711,
+      "learning_rate": 6.987558988526702e-06,
+      "loss": 0.6212,
+      "step": 11706
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 39.660357407610455,
+      "learning_rate": 6.985950970587823e-06,
+      "loss": 0.6135,
+      "step": 11707
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 38.36910821374936,
+      "learning_rate": 6.984343038366733e-06,
+      "loss": 0.5593,
+      "step": 11708
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 40.814221624774895,
+      "learning_rate": 6.982735191909168e-06,
+      "loss": 0.5006,
+      "step": 11709
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 41.754809119731874,
+      "learning_rate": 6.981127431260846e-06,
+      "loss": 0.6079,
+      "step": 11710
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 40.16929059078116,
+      "learning_rate": 6.9795197564674935e-06,
+      "loss": 0.5651,
+      "step": 11711
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 43.1161034460383,
+      "learning_rate": 6.977912167574833e-06,
+      "loss": 0.5351,
+      "step": 11712
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 36.46371788080766,
+      "learning_rate": 6.9763046646285836e-06,
+      "loss": 0.6008,
+      "step": 11713
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 42.534479134970056,
+      "learning_rate": 6.974697247674462e-06,
+      "loss": 0.5476,
+      "step": 11714
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.61241975955131,
+      "learning_rate": 6.973089916758177e-06,
+      "loss": 0.5187,
+      "step": 11715
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 34.10271251089589,
+      "learning_rate": 6.971482671925446e-06,
+      "loss": 0.5761,
+      "step": 11716
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 44.124487147174065,
+      "learning_rate": 6.969875513221974e-06,
+      "loss": 0.5344,
+      "step": 11717
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 41.86670516687089,
+      "learning_rate": 6.968268440693472e-06,
+      "loss": 0.5026,
+      "step": 11718
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 36.94207653669048,
+      "learning_rate": 6.966661454385644e-06,
+      "loss": 0.569,
+      "step": 11719
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 43.374368432143875,
+      "learning_rate": 6.965054554344188e-06,
+      "loss": 0.6042,
+      "step": 11720
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 34.012428519003166,
+      "learning_rate": 6.963447740614804e-06,
+      "loss": 0.577,
+      "step": 11721
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 41.18055176194366,
+      "learning_rate": 6.96184101324319e-06,
+      "loss": 0.5168,
+      "step": 11722
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 38.45183326693753,
+      "learning_rate": 6.960234372275039e-06,
+      "loss": 0.6004,
+      "step": 11723
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 38.661006446748836,
+      "learning_rate": 6.958627817756052e-06,
+      "loss": 0.4681,
+      "step": 11724
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 45.585603081556066,
+      "learning_rate": 6.957021349731902e-06,
+      "loss": 0.5908,
+      "step": 11725
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.929174936369016,
+      "learning_rate": 6.955414968248289e-06,
+      "loss": 0.5196,
+      "step": 11726
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 47.071331254373504,
+      "learning_rate": 6.953808673350889e-06,
+      "loss": 0.5619,
+      "step": 11727
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 47.69243439309681,
+      "learning_rate": 6.952202465085391e-06,
+      "loss": 0.6462,
+      "step": 11728
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 36.49758197665734,
+      "learning_rate": 6.950596343497475e-06,
+      "loss": 0.5429,
+      "step": 11729
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 35.624637447988555,
+      "learning_rate": 6.9489903086328106e-06,
+      "loss": 0.4796,
+      "step": 11730
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.33130265926248,
+      "learning_rate": 6.94738436053708e-06,
+      "loss": 0.4779,
+      "step": 11731
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 36.69795918559198,
+      "learning_rate": 6.945778499255951e-06,
+      "loss": 0.5862,
+      "step": 11732
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.725024967931084,
+      "learning_rate": 6.944172724835093e-06,
+      "loss": 0.4731,
+      "step": 11733
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.896872638140195,
+      "learning_rate": 6.94256703732018e-06,
+      "loss": 0.6396,
+      "step": 11734
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 35.26122322380902,
+      "learning_rate": 6.94096143675687e-06,
+      "loss": 0.5464,
+      "step": 11735
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 44.3861027971706,
+      "learning_rate": 6.939355923190823e-06,
+      "loss": 0.647,
+      "step": 11736
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 33.7829883738651,
+      "learning_rate": 6.9377504966677075e-06,
+      "loss": 0.5489,
+      "step": 11737
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 34.07541757678999,
+      "learning_rate": 6.936145157233176e-06,
+      "loss": 0.554,
+      "step": 11738
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 40.17152915043072,
+      "learning_rate": 6.934539904932887e-06,
+      "loss": 0.5924,
+      "step": 11739
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 32.64082777545591,
+      "learning_rate": 6.9329347398124825e-06,
+      "loss": 0.5227,
+      "step": 11740
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 39.373807120040595,
+      "learning_rate": 6.9313296619176255e-06,
+      "loss": 0.6015,
+      "step": 11741
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 35.29304721403451,
+      "learning_rate": 6.929724671293954e-06,
+      "loss": 0.5735,
+      "step": 11742
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.00999202998234,
+      "learning_rate": 6.928119767987118e-06,
+      "loss": 0.6119,
+      "step": 11743
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 39.713584624053574,
+      "learning_rate": 6.926514952042762e-06,
+      "loss": 0.5037,
+      "step": 11744
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.88884251757332,
+      "learning_rate": 6.9249102235065205e-06,
+      "loss": 0.5876,
+      "step": 11745
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 41.57766445840965,
+      "learning_rate": 6.923305582424031e-06,
+      "loss": 0.5847,
+      "step": 11746
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 37.67014044206109,
+      "learning_rate": 6.9217010288409335e-06,
+      "loss": 0.5614,
+      "step": 11747
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.87994006959833,
+      "learning_rate": 6.920096562802856e-06,
+      "loss": 0.5797,
+      "step": 11748
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 36.65595024422104,
+      "learning_rate": 6.918492184355434e-06,
+      "loss": 0.5485,
+      "step": 11749
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 46.07249885638153,
+      "learning_rate": 6.9168878935442905e-06,
+      "loss": 0.5729,
+      "step": 11750
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 35.080331969979454,
+      "learning_rate": 6.915283690415051e-06,
+      "loss": 0.5421,
+      "step": 11751
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 42.395545929755485,
+      "learning_rate": 6.913679575013338e-06,
+      "loss": 0.5855,
+      "step": 11752
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 36.837606896671545,
+      "learning_rate": 6.9120755473847755e-06,
+      "loss": 0.5602,
+      "step": 11753
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 32.62414535046125,
+      "learning_rate": 6.9104716075749755e-06,
+      "loss": 0.5024,
+      "step": 11754
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 31.728316231591556,
+      "learning_rate": 6.908867755629555e-06,
+      "loss": 0.5181,
+      "step": 11755
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.48688262082549,
+      "learning_rate": 6.907263991594129e-06,
+      "loss": 0.5432,
+      "step": 11756
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.560285734647756,
+      "learning_rate": 6.905660315514304e-06,
+      "loss": 0.5462,
+      "step": 11757
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.96128372753624,
+      "learning_rate": 6.904056727435694e-06,
+      "loss": 0.5662,
+      "step": 11758
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 35.778118826114024,
+      "learning_rate": 6.9024532274038935e-06,
+      "loss": 0.5105,
+      "step": 11759
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 36.87903336271836,
+      "learning_rate": 6.900849815464515e-06,
+      "loss": 0.5969,
+      "step": 11760
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 29.522927232639518,
+      "learning_rate": 6.8992464916631515e-06,
+      "loss": 0.4877,
+      "step": 11761
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 49.57663269393873,
+      "learning_rate": 6.897643256045407e-06,
+      "loss": 0.5901,
+      "step": 11762
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.816179914986435,
+      "learning_rate": 6.896040108656874e-06,
+      "loss": 0.5263,
+      "step": 11763
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 40.46644851738629,
+      "learning_rate": 6.894437049543144e-06,
+      "loss": 0.5637,
+      "step": 11764
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 36.67683877333442,
+      "learning_rate": 6.892834078749805e-06,
+      "loss": 0.5305,
+      "step": 11765
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 41.59205269735982,
+      "learning_rate": 6.89123119632245e-06,
+      "loss": 0.5107,
+      "step": 11766
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 42.55324909880979,
+      "learning_rate": 6.8896284023066585e-06,
+      "loss": 0.5059,
+      "step": 11767
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 35.890333514094856,
+      "learning_rate": 6.888025696748021e-06,
+      "loss": 0.4794,
+      "step": 11768
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.984986975450326,
+      "learning_rate": 6.886423079692106e-06,
+      "loss": 0.4906,
+      "step": 11769
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.00394864020667,
+      "learning_rate": 6.884820551184501e-06,
+      "loss": 0.5593,
+      "step": 11770
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 39.94202224069906,
+      "learning_rate": 6.883218111270774e-06,
+      "loss": 0.556,
+      "step": 11771
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.12919721187925,
+      "learning_rate": 6.8816157599965025e-06,
+      "loss": 0.5084,
+      "step": 11772
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 49.67668848993325,
+      "learning_rate": 6.880013497407257e-06,
+      "loss": 0.6365,
+      "step": 11773
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 33.203818610632105,
+      "learning_rate": 6.878411323548599e-06,
+      "loss": 0.5137,
+      "step": 11774
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 43.5528316495861,
+      "learning_rate": 6.8768092384660975e-06,
+      "loss": 0.623,
+      "step": 11775
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.276734448765595,
+      "learning_rate": 6.875207242205316e-06,
+      "loss": 0.5655,
+      "step": 11776
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 42.75037856519963,
+      "learning_rate": 6.8736053348118106e-06,
+      "loss": 0.6137,
+      "step": 11777
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 31.707557424775864,
+      "learning_rate": 6.8720035163311425e-06,
+      "loss": 0.4953,
+      "step": 11778
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 49.264728406108894,
+      "learning_rate": 6.870401786808865e-06,
+      "loss": 0.5726,
+      "step": 11779
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.1858612058536,
+      "learning_rate": 6.868800146290526e-06,
+      "loss": 0.596,
+      "step": 11780
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.322080525395315,
+      "learning_rate": 6.867198594821683e-06,
+      "loss": 0.5885,
+      "step": 11781
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.99477048361247,
+      "learning_rate": 6.8655971324478764e-06,
+      "loss": 0.5149,
+      "step": 11782
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 43.40448974379763,
+      "learning_rate": 6.863995759214659e-06,
+      "loss": 0.6164,
+      "step": 11783
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.140467244008555,
+      "learning_rate": 6.862394475167562e-06,
+      "loss": 0.5062,
+      "step": 11784
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 58.819467373057456,
+      "learning_rate": 6.860793280352132e-06,
+      "loss": 0.6339,
+      "step": 11785
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.61517900155492,
+      "learning_rate": 6.859192174813905e-06,
+      "loss": 0.574,
+      "step": 11786
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 36.141309125587576,
+      "learning_rate": 6.857591158598415e-06,
+      "loss": 0.5481,
+      "step": 11787
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 39.09033698987487,
+      "learning_rate": 6.855990231751197e-06,
+      "loss": 0.5028,
+      "step": 11788
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 46.903180951503124,
+      "learning_rate": 6.854389394317776e-06,
+      "loss": 0.5364,
+      "step": 11789
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 30.713910580242032,
+      "learning_rate": 6.852788646343678e-06,
+      "loss": 0.5076,
+      "step": 11790
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 41.76174265937809,
+      "learning_rate": 6.851187987874432e-06,
+      "loss": 0.5898,
+      "step": 11791
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 40.853095873012066,
+      "learning_rate": 6.849587418955554e-06,
+      "loss": 0.5746,
+      "step": 11792
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 40.503436084085685,
+      "learning_rate": 6.847986939632574e-06,
+      "loss": 0.4909,
+      "step": 11793
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.364429832287435,
+      "learning_rate": 6.846386549950998e-06,
+      "loss": 0.5258,
+      "step": 11794
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 39.74550286231058,
+      "learning_rate": 6.844786249956343e-06,
+      "loss": 0.5255,
+      "step": 11795
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.097572710658866,
+      "learning_rate": 6.8431860396941205e-06,
+      "loss": 0.4549,
+      "step": 11796
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.709105784406034,
+      "learning_rate": 6.841585919209842e-06,
+      "loss": 0.4841,
+      "step": 11797
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.0818701008922,
+      "learning_rate": 6.839985888549015e-06,
+      "loss": 0.4817,
+      "step": 11798
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 48.851908533849496,
+      "learning_rate": 6.838385947757136e-06,
+      "loss": 0.5708,
+      "step": 11799
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.91282548660938,
+      "learning_rate": 6.836786096879713e-06,
+      "loss": 0.5141,
+      "step": 11800
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.94667352690292,
+      "learning_rate": 6.8351863359622424e-06,
+      "loss": 0.5537,
+      "step": 11801
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 32.18695002706341,
+      "learning_rate": 6.833586665050225e-06,
+      "loss": 0.4531,
+      "step": 11802
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 44.66512255288954,
+      "learning_rate": 6.831987084189144e-06,
+      "loss": 0.58,
+      "step": 11803
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.24547082327273,
+      "learning_rate": 6.830387593424501e-06,
+      "loss": 0.5465,
+      "step": 11804
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 32.542644810028726,
+      "learning_rate": 6.828788192801779e-06,
+      "loss": 0.5058,
+      "step": 11805
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 35.16638093910631,
+      "learning_rate": 6.827188882366467e-06,
+      "loss": 0.4857,
+      "step": 11806
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 41.435296904857076,
+      "learning_rate": 6.825589662164049e-06,
+      "loss": 0.5642,
+      "step": 11807
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 50.32635236269458,
+      "learning_rate": 6.823990532240001e-06,
+      "loss": 0.5548,
+      "step": 11808
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.341047578821474,
+      "learning_rate": 6.822391492639804e-06,
+      "loss": 0.526,
+      "step": 11809
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 38.52839287454276,
+      "learning_rate": 6.820792543408935e-06,
+      "loss": 0.4688,
+      "step": 11810
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.65198092915181,
+      "learning_rate": 6.819193684592864e-06,
+      "loss": 0.5117,
+      "step": 11811
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.44755479541047,
+      "learning_rate": 6.817594916237071e-06,
+      "loss": 0.609,
+      "step": 11812
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.01971091097015,
+      "learning_rate": 6.815996238387011e-06,
+      "loss": 0.5467,
+      "step": 11813
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 44.68912773635727,
+      "learning_rate": 6.814397651088156e-06,
+      "loss": 0.5409,
+      "step": 11814
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 35.2332002532747,
+      "learning_rate": 6.812799154385967e-06,
+      "loss": 0.5074,
+      "step": 11815
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 44.36724570583389,
+      "learning_rate": 6.811200748325908e-06,
+      "loss": 0.6443,
+      "step": 11816
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 40.96444744734313,
+      "learning_rate": 6.8096024329534376e-06,
+      "loss": 0.5262,
+      "step": 11817
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 46.40695594224962,
+      "learning_rate": 6.808004208314001e-06,
+      "loss": 0.6009,
+      "step": 11818
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 39.57789633737947,
+      "learning_rate": 6.806406074453061e-06,
+      "loss": 0.5325,
+      "step": 11819
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.27713248651147,
+      "learning_rate": 6.804808031416064e-06,
+      "loss": 0.5217,
+      "step": 11820
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 33.72895102915989,
+      "learning_rate": 6.803210079248454e-06,
+      "loss": 0.5069,
+      "step": 11821
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.93284565460043,
+      "learning_rate": 6.8016122179956855e-06,
+      "loss": 0.6044,
+      "step": 11822
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 52.61072985607085,
+      "learning_rate": 6.800014447703191e-06,
+      "loss": 0.609,
+      "step": 11823
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.73528445580357,
+      "learning_rate": 6.798416768416412e-06,
+      "loss": 0.7347,
+      "step": 11824
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 35.97431522934447,
+      "learning_rate": 6.7968191801807894e-06,
+      "loss": 0.536,
+      "step": 11825
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 40.01825176705149,
+      "learning_rate": 6.795221683041756e-06,
+      "loss": 0.5702,
+      "step": 11826
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 35.620207889224396,
+      "learning_rate": 6.793624277044745e-06,
+      "loss": 0.4806,
+      "step": 11827
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 53.22536112948167,
+      "learning_rate": 6.792026962235178e-06,
+      "loss": 0.6363,
+      "step": 11828
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 39.21536899294313,
+      "learning_rate": 6.790429738658493e-06,
+      "loss": 0.5688,
+      "step": 11829
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 33.615183946752204,
+      "learning_rate": 6.788832606360104e-06,
+      "loss": 0.5592,
+      "step": 11830
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 32.073921604566394,
+      "learning_rate": 6.787235565385439e-06,
+      "loss": 0.4673,
+      "step": 11831
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.71658257438587,
+      "learning_rate": 6.785638615779919e-06,
+      "loss": 0.5253,
+      "step": 11832
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 38.747067211285696,
+      "learning_rate": 6.784041757588954e-06,
+      "loss": 0.5304,
+      "step": 11833
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 41.90151719257554,
+      "learning_rate": 6.782444990857957e-06,
+      "loss": 0.5931,
+      "step": 11834
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 33.743603778429744,
+      "learning_rate": 6.780848315632344e-06,
+      "loss": 0.5149,
+      "step": 11835
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 31.954525591832887,
+      "learning_rate": 6.779251731957522e-06,
+      "loss": 0.5635,
+      "step": 11836
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 32.912032007931586,
+      "learning_rate": 6.7776552398788995e-06,
+      "loss": 0.5364,
+      "step": 11837
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 40.39073720949624,
+      "learning_rate": 6.776058839441875e-06,
+      "loss": 0.5608,
+      "step": 11838
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 38.16050934508618,
+      "learning_rate": 6.77446253069185e-06,
+      "loss": 0.587,
+      "step": 11839
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 46.73433249881997,
+      "learning_rate": 6.772866313674223e-06,
+      "loss": 0.5633,
+      "step": 11840
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 37.41794117616808,
+      "learning_rate": 6.771270188434392e-06,
+      "loss": 0.5443,
+      "step": 11841
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 32.85482846603235,
+      "learning_rate": 6.7696741550177505e-06,
+      "loss": 0.514,
+      "step": 11842
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 40.5717301145833,
+      "learning_rate": 6.768078213469682e-06,
+      "loss": 0.5254,
+      "step": 11843
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.40528668624789,
+      "learning_rate": 6.76648236383558e-06,
+      "loss": 0.5238,
+      "step": 11844
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.59977820738311,
+      "learning_rate": 6.764886606160828e-06,
+      "loss": 0.5712,
+      "step": 11845
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 38.78335821921142,
+      "learning_rate": 6.763290940490807e-06,
+      "loss": 0.5372,
+      "step": 11846
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.823397262149186,
+      "learning_rate": 6.761695366870902e-06,
+      "loss": 0.5728,
+      "step": 11847
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.81774371158902,
+      "learning_rate": 6.760099885346485e-06,
+      "loss": 0.5389,
+      "step": 11848
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.71814655081447,
+      "learning_rate": 6.758504495962929e-06,
+      "loss": 0.6431,
+      "step": 11849
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 41.331098634865796,
+      "learning_rate": 6.756909198765611e-06,
+      "loss": 0.5943,
+      "step": 11850
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 39.350189945412026,
+      "learning_rate": 6.755313993799901e-06,
+      "loss": 0.5916,
+      "step": 11851
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.71269592829672,
+      "learning_rate": 6.75371888111116e-06,
+      "loss": 0.5317,
+      "step": 11852
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 33.920214028549175,
+      "learning_rate": 6.7521238607447536e-06,
+      "loss": 0.4987,
+      "step": 11853
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.50916191059305,
+      "learning_rate": 6.750528932746047e-06,
+      "loss": 0.5717,
+      "step": 11854
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 32.874340152282876,
+      "learning_rate": 6.7489340971603935e-06,
+      "loss": 0.4966,
+      "step": 11855
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 51.3552519254807,
+      "learning_rate": 6.747339354033158e-06,
+      "loss": 0.5813,
+      "step": 11856
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 40.79838210338773,
+      "learning_rate": 6.745744703409685e-06,
+      "loss": 0.5797,
+      "step": 11857
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 38.60431311604115,
+      "learning_rate": 6.74415014533533e-06,
+      "loss": 0.4862,
+      "step": 11858
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 48.406413635213745,
+      "learning_rate": 6.742555679855438e-06,
+      "loss": 0.5901,
+      "step": 11859
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.06161104711134,
+      "learning_rate": 6.740961307015359e-06,
+      "loss": 0.4839,
+      "step": 11860
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 39.338035583607805,
+      "learning_rate": 6.739367026860435e-06,
+      "loss": 0.5118,
+      "step": 11861
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 44.151192555678406,
+      "learning_rate": 6.737772839436003e-06,
+      "loss": 0.5714,
+      "step": 11862
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 45.733116930205405,
+      "learning_rate": 6.736178744787403e-06,
+      "loss": 0.5647,
+      "step": 11863
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 35.38363664020615,
+      "learning_rate": 6.734584742959972e-06,
+      "loss": 0.6332,
+      "step": 11864
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 44.46182428661056,
+      "learning_rate": 6.732990833999038e-06,
+      "loss": 0.5915,
+      "step": 11865
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.21282035658129,
+      "learning_rate": 6.731397017949937e-06,
+      "loss": 0.6212,
+      "step": 11866
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 34.67752534228065,
+      "learning_rate": 6.7298032948579925e-06,
+      "loss": 0.5183,
+      "step": 11867
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 37.410609438229336,
+      "learning_rate": 6.728209664768525e-06,
+      "loss": 0.5126,
+      "step": 11868
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 39.048455409763726,
+      "learning_rate": 6.726616127726863e-06,
+      "loss": 0.5941,
+      "step": 11869
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 37.28626509107182,
+      "learning_rate": 6.725022683778326e-06,
+      "loss": 0.6118,
+      "step": 11870
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.44208738405153,
+      "learning_rate": 6.723429332968227e-06,
+      "loss": 0.6134,
+      "step": 11871
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.492474871095816,
+      "learning_rate": 6.721836075341879e-06,
+      "loss": 0.5313,
+      "step": 11872
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.64011572054846,
+      "learning_rate": 6.720242910944596e-06,
+      "loss": 0.502,
+      "step": 11873
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 40.69778999319965,
+      "learning_rate": 6.718649839821686e-06,
+      "loss": 0.5882,
+      "step": 11874
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 35.29856381169349,
+      "learning_rate": 6.717056862018456e-06,
+      "loss": 0.551,
+      "step": 11875
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 45.06029386582872,
+      "learning_rate": 6.71546397758021e-06,
+      "loss": 0.5627,
+      "step": 11876
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.23278735457495,
+      "learning_rate": 6.713871186552246e-06,
+      "loss": 0.4845,
+      "step": 11877
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 32.70420044626575,
+      "learning_rate": 6.7122784889798596e-06,
+      "loss": 0.5736,
+      "step": 11878
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.819414582503555,
+      "learning_rate": 6.710685884908352e-06,
+      "loss": 0.5778,
+      "step": 11879
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 36.585568432898356,
+      "learning_rate": 6.709093374383012e-06,
+      "loss": 0.5065,
+      "step": 11880
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 33.33421143130207,
+      "learning_rate": 6.7075009574491355e-06,
+      "loss": 0.4855,
+      "step": 11881
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 43.92801550326215,
+      "learning_rate": 6.7059086341520045e-06,
+      "loss": 0.4858,
+      "step": 11882
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 37.17731855764639,
+      "learning_rate": 6.704316404536904e-06,
+      "loss": 0.5871,
+      "step": 11883
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 49.414984102965484,
+      "learning_rate": 6.702724268649114e-06,
+      "loss": 0.626,
+      "step": 11884
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.60825626215441,
+      "learning_rate": 6.70113222653392e-06,
+      "loss": 0.5402,
+      "step": 11885
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 47.31406149057473,
+      "learning_rate": 6.699540278236598e-06,
+      "loss": 0.5767,
+      "step": 11886
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.82327249396495,
+      "learning_rate": 6.697948423802416e-06,
+      "loss": 0.6059,
+      "step": 11887
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 44.1974445197625,
+      "learning_rate": 6.696356663276651e-06,
+      "loss": 0.5932,
+      "step": 11888
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.246646684248304,
+      "learning_rate": 6.694764996704569e-06,
+      "loss": 0.5375,
+      "step": 11889
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 36.282246736147094,
+      "learning_rate": 6.6931734241314365e-06,
+      "loss": 0.5266,
+      "step": 11890
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 52.50880266597445,
+      "learning_rate": 6.69158194560252e-06,
+      "loss": 0.6338,
+      "step": 11891
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 42.43596516054219,
+      "learning_rate": 6.689990561163075e-06,
+      "loss": 0.5341,
+      "step": 11892
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.9054116216205,
+      "learning_rate": 6.688399270858362e-06,
+      "loss": 0.5815,
+      "step": 11893
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 36.12592453747006,
+      "learning_rate": 6.686808074733638e-06,
+      "loss": 0.554,
+      "step": 11894
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 41.04800784054706,
+      "learning_rate": 6.6852169728341565e-06,
+      "loss": 0.5377,
+      "step": 11895
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.83829738681142,
+      "learning_rate": 6.683625965205162e-06,
+      "loss": 0.46,
+      "step": 11896
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 49.092905135627916,
+      "learning_rate": 6.6820350518919044e-06,
+      "loss": 0.6341,
+      "step": 11897
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 33.10387774670084,
+      "learning_rate": 6.680444232939629e-06,
+      "loss": 0.4993,
+      "step": 11898
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 34.230298292725315,
+      "learning_rate": 6.6788535083935764e-06,
+      "loss": 0.5317,
+      "step": 11899
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 32.63559657872288,
+      "learning_rate": 6.677262878298993e-06,
+      "loss": 0.5234,
+      "step": 11900
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 67.51478270676468,
+      "learning_rate": 6.675672342701107e-06,
+      "loss": 0.587,
+      "step": 11901
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 45.78638292512603,
+      "learning_rate": 6.674081901645153e-06,
+      "loss": 0.5199,
+      "step": 11902
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 51.991923516546976,
+      "learning_rate": 6.672491555176363e-06,
+      "loss": 0.613,
+      "step": 11903
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.70797453318108,
+      "learning_rate": 6.6709013033399695e-06,
+      "loss": 0.5226,
+      "step": 11904
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 41.644425931159915,
+      "learning_rate": 6.6693111461811975e-06,
+      "loss": 0.5761,
+      "step": 11905
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 36.91626284222719,
+      "learning_rate": 6.6677210837452645e-06,
+      "loss": 0.5183,
+      "step": 11906
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.66409583669828,
+      "learning_rate": 6.6661311160773945e-06,
+      "loss": 0.5641,
+      "step": 11907
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 35.787803722695664,
+      "learning_rate": 6.664541243222806e-06,
+      "loss": 0.567,
+      "step": 11908
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.11458344779118,
+      "learning_rate": 6.662951465226713e-06,
+      "loss": 0.4502,
+      "step": 11909
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.23646510352583,
+      "learning_rate": 6.66136178213433e-06,
+      "loss": 0.5814,
+      "step": 11910
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 43.00260255112436,
+      "learning_rate": 6.659772193990865e-06,
+      "loss": 0.5482,
+      "step": 11911
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 46.15591449396078,
+      "learning_rate": 6.658182700841521e-06,
+      "loss": 0.596,
+      "step": 11912
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 34.74243973747013,
+      "learning_rate": 6.656593302731509e-06,
+      "loss": 0.529,
+      "step": 11913
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 32.96333684730556,
+      "learning_rate": 6.655003999706027e-06,
+      "loss": 0.5682,
+      "step": 11914
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 37.303919580354915,
+      "learning_rate": 6.653414791810277e-06,
+      "loss": 0.5439,
+      "step": 11915
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 49.07175508907311,
+      "learning_rate": 6.6518256790894484e-06,
+      "loss": 0.6005,
+      "step": 11916
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 49.33997056534179,
+      "learning_rate": 6.6502366615887415e-06,
+      "loss": 0.5356,
+      "step": 11917
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 46.07711805379777,
+      "learning_rate": 6.648647739353342e-06,
+      "loss": 0.5807,
+      "step": 11918
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 32.91592827866098,
+      "learning_rate": 6.6470589124284414e-06,
+      "loss": 0.5109,
+      "step": 11919
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 45.89743082963544,
+      "learning_rate": 6.645470180859227e-06,
+      "loss": 0.6081,
+      "step": 11920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.62759604155548,
+      "learning_rate": 6.643881544690876e-06,
+      "loss": 0.618,
+      "step": 11921
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.1775229168088,
+      "learning_rate": 6.6422930039685695e-06,
+      "loss": 0.5469,
+      "step": 11922
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 35.94208445291084,
+      "learning_rate": 6.640704558737487e-06,
+      "loss": 0.5462,
+      "step": 11923
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.001011162136614,
+      "learning_rate": 6.639116209042798e-06,
+      "loss": 0.5016,
+      "step": 11924
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 47.46156619813306,
+      "learning_rate": 6.637527954929685e-06,
+      "loss": 0.5676,
+      "step": 11925
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 43.39684607938236,
+      "learning_rate": 6.635939796443306e-06,
+      "loss": 0.6026,
+      "step": 11926
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 34.97572085365051,
+      "learning_rate": 6.634351733628832e-06,
+      "loss": 0.4719,
+      "step": 11927
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.39499620583742,
+      "learning_rate": 6.632763766531423e-06,
+      "loss": 0.5527,
+      "step": 11928
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 48.07171968049256,
+      "learning_rate": 6.631175895196245e-06,
+      "loss": 0.6446,
+      "step": 11929
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 46.346200221002285,
+      "learning_rate": 6.629588119668458e-06,
+      "loss": 0.502,
+      "step": 11930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 41.07710525738602,
+      "learning_rate": 6.628000439993207e-06,
+      "loss": 0.5683,
+      "step": 11931
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.706316263808816,
+      "learning_rate": 6.626412856215653e-06,
+      "loss": 0.6091,
+      "step": 11932
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.029709308352,
+      "learning_rate": 6.624825368380944e-06,
+      "loss": 0.5895,
+      "step": 11933
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 42.45497928707384,
+      "learning_rate": 6.623237976534225e-06,
+      "loss": 0.5847,
+      "step": 11934
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 33.0256821117953,
+      "learning_rate": 6.6216506807206475e-06,
+      "loss": 0.481,
+      "step": 11935
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.4715879527771,
+      "learning_rate": 6.620063480985346e-06,
+      "loss": 0.5779,
+      "step": 11936
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.11098186439655,
+      "learning_rate": 6.61847637737346e-06,
+      "loss": 0.5996,
+      "step": 11937
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 32.1671779218046,
+      "learning_rate": 6.6168893699301286e-06,
+      "loss": 0.5043,
+      "step": 11938
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.63427325624182,
+      "learning_rate": 6.6153024587004855e-06,
+      "loss": 0.5497,
+      "step": 11939
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 55.47417488326881,
+      "learning_rate": 6.613715643729661e-06,
+      "loss": 0.523,
+      "step": 11940
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.7645461443038,
+      "learning_rate": 6.612128925062781e-06,
+      "loss": 0.574,
+      "step": 11941
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.67783671653,
+      "learning_rate": 6.610542302744973e-06,
+      "loss": 0.5572,
+      "step": 11942
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.84704815526237,
+      "learning_rate": 6.608955776821357e-06,
+      "loss": 0.4801,
+      "step": 11943
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.818466243707206,
+      "learning_rate": 6.60736934733706e-06,
+      "loss": 0.5013,
+      "step": 11944
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.439754797614704,
+      "learning_rate": 6.605783014337192e-06,
+      "loss": 0.5198,
+      "step": 11945
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 46.5824076037928,
+      "learning_rate": 6.6041967778668695e-06,
+      "loss": 0.6286,
+      "step": 11946
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 39.82665883624733,
+      "learning_rate": 6.602610637971201e-06,
+      "loss": 0.5706,
+      "step": 11947
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 45.699823118739815,
+      "learning_rate": 6.601024594695302e-06,
+      "loss": 0.5315,
+      "step": 11948
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.30208910538414,
+      "learning_rate": 6.599438648084277e-06,
+      "loss": 0.5691,
+      "step": 11949
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.134085445028425,
+      "learning_rate": 6.597852798183222e-06,
+      "loss": 0.5862,
+      "step": 11950
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 43.137954662428456,
+      "learning_rate": 6.596267045037247e-06,
+      "loss": 0.5509,
+      "step": 11951
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 33.84262426164866,
+      "learning_rate": 6.594681388691445e-06,
+      "loss": 0.5021,
+      "step": 11952
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 33.20763401426329,
+      "learning_rate": 6.593095829190911e-06,
+      "loss": 0.4897,
+      "step": 11953
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 35.76553764015951,
+      "learning_rate": 6.591510366580742e-06,
+      "loss": 0.5878,
+      "step": 11954
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 32.74507704708991,
+      "learning_rate": 6.589925000906023e-06,
+      "loss": 0.5693,
+      "step": 11955
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 36.971937727357044,
+      "learning_rate": 6.588339732211839e-06,
+      "loss": 0.5477,
+      "step": 11956
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.10344143728822,
+      "learning_rate": 6.586754560543281e-06,
+      "loss": 0.5572,
+      "step": 11957
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.99036481629391,
+      "learning_rate": 6.585169485945425e-06,
+      "loss": 0.4795,
+      "step": 11958
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 44.32339057474762,
+      "learning_rate": 6.583584508463354e-06,
+      "loss": 0.6469,
+      "step": 11959
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 46.339037140469095,
+      "learning_rate": 6.581999628142137e-06,
+      "loss": 0.5381,
+      "step": 11960
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.99882711209913,
+      "learning_rate": 6.580414845026853e-06,
+      "loss": 0.6061,
+      "step": 11961
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.0082411455324,
+      "learning_rate": 6.578830159162568e-06,
+      "loss": 0.4105,
+      "step": 11962
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 36.00846058052696,
+      "learning_rate": 6.577245570594355e-06,
+      "loss": 0.5657,
+      "step": 11963
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 36.88586110010063,
+      "learning_rate": 6.575661079367277e-06,
+      "loss": 0.4883,
+      "step": 11964
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.81844666677052,
+      "learning_rate": 6.574076685526392e-06,
+      "loss": 0.4695,
+      "step": 11965
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 28.403865592434936,
+      "learning_rate": 6.572492389116761e-06,
+      "loss": 0.4489,
+      "step": 11966
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 36.435578627663276,
+      "learning_rate": 6.570908190183442e-06,
+      "loss": 0.545,
+      "step": 11967
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 42.17540310111344,
+      "learning_rate": 6.569324088771488e-06,
+      "loss": 0.5437,
+      "step": 11968
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 37.621040078822865,
+      "learning_rate": 6.5677400849259535e-06,
+      "loss": 0.5862,
+      "step": 11969
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 36.9829679671937,
+      "learning_rate": 6.566156178691882e-06,
+      "loss": 0.537,
+      "step": 11970
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 41.79074788184967,
+      "learning_rate": 6.564572370114318e-06,
+      "loss": 0.6102,
+      "step": 11971
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 45.50128407814789,
+      "learning_rate": 6.562988659238305e-06,
+      "loss": 0.5635,
+      "step": 11972
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 42.222485267645176,
+      "learning_rate": 6.561405046108886e-06,
+      "loss": 0.4162,
+      "step": 11973
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 39.63689689863115,
+      "learning_rate": 6.5598215307710975e-06,
+      "loss": 0.5747,
+      "step": 11974
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 69.53769182517834,
+      "learning_rate": 6.55823811326997e-06,
+      "loss": 0.5774,
+      "step": 11975
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 33.98959861861803,
+      "learning_rate": 6.5566547936505375e-06,
+      "loss": 0.4985,
+      "step": 11976
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.15355518177299,
+      "learning_rate": 6.5550715719578295e-06,
+      "loss": 0.4794,
+      "step": 11977
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 39.1553447615705,
+      "learning_rate": 6.5534884482368665e-06,
+      "loss": 0.5421,
+      "step": 11978
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.790431460160534,
+      "learning_rate": 6.551905422532683e-06,
+      "loss": 0.5294,
+      "step": 11979
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.220854568211664,
+      "learning_rate": 6.550322494890289e-06,
+      "loss": 0.5587,
+      "step": 11980
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.60578350699277,
+      "learning_rate": 6.548739665354702e-06,
+      "loss": 0.504,
+      "step": 11981
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.592463046399686,
+      "learning_rate": 6.547156933970943e-06,
+      "loss": 0.5173,
+      "step": 11982
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 32.732075987863425,
+      "learning_rate": 6.545574300784021e-06,
+      "loss": 0.4772,
+      "step": 11983
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 42.71809834316022,
+      "learning_rate": 6.543991765838946e-06,
+      "loss": 0.6038,
+      "step": 11984
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 35.806280711196294,
+      "learning_rate": 6.542409329180721e-06,
+      "loss": 0.4921,
+      "step": 11985
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 33.29006527489508,
+      "learning_rate": 6.540826990854353e-06,
+      "loss": 0.4676,
+      "step": 11986
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 53.62641980885877,
+      "learning_rate": 6.53924475090484e-06,
+      "loss": 0.5958,
+      "step": 11987
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 58.05438388400613,
+      "learning_rate": 6.537662609377184e-06,
+      "loss": 0.4734,
+      "step": 11988
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 47.235496661628936,
+      "learning_rate": 6.53608056631638e-06,
+      "loss": 0.5442,
+      "step": 11989
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.55445603343381,
+      "learning_rate": 6.534498621767418e-06,
+      "loss": 0.5181,
+      "step": 11990
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.660471797501145,
+      "learning_rate": 6.5329167757752835e-06,
+      "loss": 0.6037,
+      "step": 11991
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.52247514143519,
+      "learning_rate": 6.531335028384971e-06,
+      "loss": 0.6287,
+      "step": 11992
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.094328174162044,
+      "learning_rate": 6.529753379641462e-06,
+      "loss": 0.5863,
+      "step": 11993
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 32.593225868085256,
+      "learning_rate": 6.528171829589734e-06,
+      "loss": 0.5054,
+      "step": 11994
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 32.664718792865585,
+      "learning_rate": 6.52659037827477e-06,
+      "loss": 0.5016,
+      "step": 11995
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 34.708371301304325,
+      "learning_rate": 6.525009025741543e-06,
+      "loss": 0.5718,
+      "step": 11996
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 35.822901012612824,
+      "learning_rate": 6.523427772035024e-06,
+      "loss": 0.5607,
+      "step": 11997
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 37.81172329128573,
+      "learning_rate": 6.521846617200191e-06,
+      "loss": 0.5202,
+      "step": 11998
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.20957729839851,
+      "learning_rate": 6.520265561282004e-06,
+      "loss": 0.5387,
+      "step": 11999
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 55.123251924024245,
+      "learning_rate": 6.5186846043254246e-06,
+      "loss": 0.5448,
+      "step": 12000
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 37.94230028856839,
+      "learning_rate": 6.5171037463754224e-06,
+      "loss": 0.5463,
+      "step": 12001
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 41.0996346528227,
+      "learning_rate": 6.515522987476952e-06,
+      "loss": 0.5941,
+      "step": 12002
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 40.16579499558562,
+      "learning_rate": 6.513942327674971e-06,
+      "loss": 0.4951,
+      "step": 12003
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 46.53890034722754,
+      "learning_rate": 6.512361767014427e-06,
+      "loss": 0.5936,
+      "step": 12004
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.185114788849226,
+      "learning_rate": 6.510781305540277e-06,
+      "loss": 0.5601,
+      "step": 12005
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.700466623504674,
+      "learning_rate": 6.509200943297463e-06,
+      "loss": 0.5545,
+      "step": 12006
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 42.07906854976103,
+      "learning_rate": 6.507620680330933e-06,
+      "loss": 0.5484,
+      "step": 12007
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 49.04377772288039,
+      "learning_rate": 6.50604051668563e-06,
+      "loss": 0.52,
+      "step": 12008
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 42.530237582431326,
+      "learning_rate": 6.504460452406489e-06,
+      "loss": 0.5736,
+      "step": 12009
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 42.7555052129335,
+      "learning_rate": 6.502880487538448e-06,
+      "loss": 0.5254,
+      "step": 12010
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 44.24460460615066,
+      "learning_rate": 6.50130062212644e-06,
+      "loss": 0.6237,
+      "step": 12011
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 40.510899369311645,
+      "learning_rate": 6.499720856215395e-06,
+      "loss": 0.5291,
+      "step": 12012
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 28.825095474965106,
+      "learning_rate": 6.498141189850243e-06,
+      "loss": 0.4767,
+      "step": 12013
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.635327715029085,
+      "learning_rate": 6.496561623075907e-06,
+      "loss": 0.6006,
+      "step": 12014
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.3301165654664,
+      "learning_rate": 6.494982155937308e-06,
+      "loss": 0.5606,
+      "step": 12015
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 44.22531480887389,
+      "learning_rate": 6.493402788479365e-06,
+      "loss": 0.6432,
+      "step": 12016
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 36.02079149372119,
+      "learning_rate": 6.491823520746996e-06,
+      "loss": 0.4996,
+      "step": 12017
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 50.643766880703446,
+      "learning_rate": 6.4902443527851175e-06,
+      "loss": 0.5902,
+      "step": 12018
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.98144399114133,
+      "learning_rate": 6.488665284638631e-06,
+      "loss": 0.5736,
+      "step": 12019
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 43.10428679228106,
+      "learning_rate": 6.487086316352453e-06,
+      "loss": 0.5866,
+      "step": 12020
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 41.781587013020996,
+      "learning_rate": 6.4855074479714845e-06,
+      "loss": 0.5817,
+      "step": 12021
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 45.33233833915703,
+      "learning_rate": 6.483928679540627e-06,
+      "loss": 0.6188,
+      "step": 12022
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 41.75997861519993,
+      "learning_rate": 6.482350011104785e-06,
+      "loss": 0.4919,
+      "step": 12023
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 43.85052608622427,
+      "learning_rate": 6.480771442708847e-06,
+      "loss": 0.5178,
+      "step": 12024
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.38443854063583,
+      "learning_rate": 6.47919297439771e-06,
+      "loss": 0.559,
+      "step": 12025
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 33.20066416518454,
+      "learning_rate": 6.477614606216267e-06,
+      "loss": 0.519,
+      "step": 12026
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.94763955124185,
+      "learning_rate": 6.476036338209404e-06,
+      "loss": 0.4909,
+      "step": 12027
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 31.425945995176374,
+      "learning_rate": 6.474458170422009e-06,
+      "loss": 0.4594,
+      "step": 12028
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 41.110085609178014,
+      "learning_rate": 6.4728801028989556e-06,
+      "loss": 0.63,
+      "step": 12029
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.2444564803542,
+      "learning_rate": 6.471302135685131e-06,
+      "loss": 0.5924,
+      "step": 12030
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 43.5913207768509,
+      "learning_rate": 6.4697242688254095e-06,
+      "loss": 0.5592,
+      "step": 12031
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 48.32920569794224,
+      "learning_rate": 6.468146502364665e-06,
+      "loss": 0.6558,
+      "step": 12032
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 44.5384620581868,
+      "learning_rate": 6.46656883634777e-06,
+      "loss": 0.5554,
+      "step": 12033
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 40.581130200351254,
+      "learning_rate": 6.464991270819587e-06,
+      "loss": 0.5797,
+      "step": 12034
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 35.89051397005012,
+      "learning_rate": 6.463413805824985e-06,
+      "loss": 0.5154,
+      "step": 12035
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.09458580696153,
+      "learning_rate": 6.4618364414088264e-06,
+      "loss": 0.5844,
+      "step": 12036
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.26300738327561,
+      "learning_rate": 6.460259177615966e-06,
+      "loss": 0.5462,
+      "step": 12037
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.29729673021735,
+      "learning_rate": 6.458682014491271e-06,
+      "loss": 0.5237,
+      "step": 12038
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.94705680442322,
+      "learning_rate": 6.4571049520795836e-06,
+      "loss": 0.4725,
+      "step": 12039
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.704229924389224,
+      "learning_rate": 6.45552799042576e-06,
+      "loss": 0.5779,
+      "step": 12040
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 31.896790483657448,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.5188,
+      "step": 12041
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.158502956888114,
+      "learning_rate": 6.452374369571088e-06,
+      "loss": 0.5883,
+      "step": 12042
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 36.63177880577657,
+      "learning_rate": 6.4507977104599275e-06,
+      "loss": 0.5342,
+      "step": 12043
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 34.784206535487286,
+      "learning_rate": 6.449221152286001e-06,
+      "loss": 0.4599,
+      "step": 12044
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.56783144855849,
+      "learning_rate": 6.447644695094151e-06,
+      "loss": 0.5677,
+      "step": 12045
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 32.93591796289521,
+      "learning_rate": 6.446068338929208e-06,
+      "loss": 0.5085,
+      "step": 12046
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 36.84633140159747,
+      "learning_rate": 6.444492083836004e-06,
+      "loss": 0.5148,
+      "step": 12047
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.984572189101215,
+      "learning_rate": 6.442915929859361e-06,
+      "loss": 0.5045,
+      "step": 12048
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.768832700187005,
+      "learning_rate": 6.44133987704411e-06,
+      "loss": 0.5037,
+      "step": 12049
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 47.6106002344483,
+      "learning_rate": 6.43976392543507e-06,
+      "loss": 0.6727,
+      "step": 12050
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 36.7014050427829,
+      "learning_rate": 6.438188075077062e-06,
+      "loss": 0.5094,
+      "step": 12051
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.454470537481946,
+      "learning_rate": 6.436612326014904e-06,
+      "loss": 0.4724,
+      "step": 12052
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.504144404596516,
+      "learning_rate": 6.435036678293405e-06,
+      "loss": 0.4723,
+      "step": 12053
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.65671327786701,
+      "learning_rate": 6.433461131957376e-06,
+      "loss": 0.5017,
+      "step": 12054
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.98414764995569,
+      "learning_rate": 6.431885687051628e-06,
+      "loss": 0.5564,
+      "step": 12055
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 42.508854594372174,
+      "learning_rate": 6.43031034362096e-06,
+      "loss": 0.6522,
+      "step": 12056
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.26896205571091,
+      "learning_rate": 6.428735101710184e-06,
+      "loss": 0.5364,
+      "step": 12057
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 46.47370859047877,
+      "learning_rate": 6.427159961364089e-06,
+      "loss": 0.5461,
+      "step": 12058
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.62191558299929,
+      "learning_rate": 6.425584922627476e-06,
+      "loss": 0.5297,
+      "step": 12059
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.15768956956114,
+      "learning_rate": 6.424009985545134e-06,
+      "loss": 0.5398,
+      "step": 12060
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 44.71795683449936,
+      "learning_rate": 6.422435150161859e-06,
+      "loss": 0.5927,
+      "step": 12061
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.12355625029714,
+      "learning_rate": 6.4208604165224366e-06,
+      "loss": 0.534,
+      "step": 12062
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 45.040661100419285,
+      "learning_rate": 6.419285784671645e-06,
+      "loss": 0.5534,
+      "step": 12063
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.050738146224774,
+      "learning_rate": 6.417711254654275e-06,
+      "loss": 0.5585,
+      "step": 12064
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.40477887465987,
+      "learning_rate": 6.4161368265151e-06,
+      "loss": 0.5193,
+      "step": 12065
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 36.31285673481662,
+      "learning_rate": 6.414562500298896e-06,
+      "loss": 0.5684,
+      "step": 12066
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 41.20623419223403,
+      "learning_rate": 6.412988276050441e-06,
+      "loss": 0.5237,
+      "step": 12067
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.77560013395954,
+      "learning_rate": 6.411414153814499e-06,
+      "loss": 0.5617,
+      "step": 12068
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 41.13447570046556,
+      "learning_rate": 6.409840133635837e-06,
+      "loss": 0.536,
+      "step": 12069
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 46.99888650356351,
+      "learning_rate": 6.408266215559225e-06,
+      "loss": 0.6421,
+      "step": 12070
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.985787714143925,
+      "learning_rate": 6.406692399629418e-06,
+      "loss": 0.5826,
+      "step": 12071
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 41.4495288311324,
+      "learning_rate": 6.405118685891181e-06,
+      "loss": 0.5928,
+      "step": 12072
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.87348129458283,
+      "learning_rate": 6.403545074389261e-06,
+      "loss": 0.5244,
+      "step": 12073
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 34.17236093350911,
+      "learning_rate": 6.401971565168416e-06,
+      "loss": 0.4708,
+      "step": 12074
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 46.96044432197547,
+      "learning_rate": 6.400398158273395e-06,
+      "loss": 0.6058,
+      "step": 12075
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 39.255848536923345,
+      "learning_rate": 6.398824853748946e-06,
+      "loss": 0.5463,
+      "step": 12076
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 40.52633000367562,
+      "learning_rate": 6.397251651639813e-06,
+      "loss": 0.6312,
+      "step": 12077
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.69453745597731,
+      "learning_rate": 6.395678551990735e-06,
+      "loss": 0.5602,
+      "step": 12078
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 40.48500139465179,
+      "learning_rate": 6.394105554846447e-06,
+      "loss": 0.6261,
+      "step": 12079
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.37180066552917,
+      "learning_rate": 6.3925326602516904e-06,
+      "loss": 0.517,
+      "step": 12080
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.519436021592156,
+      "learning_rate": 6.390959868251192e-06,
+      "loss": 0.5847,
+      "step": 12081
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 40.20392474710779,
+      "learning_rate": 6.389387178889688e-06,
+      "loss": 0.4957,
+      "step": 12082
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 40.906547947462236,
+      "learning_rate": 6.3878145922118985e-06,
+      "loss": 0.5931,
+      "step": 12083
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.84501092386712,
+      "learning_rate": 6.3862421082625495e-06,
+      "loss": 0.5244,
+      "step": 12084
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.823164058642824,
+      "learning_rate": 6.3846697270863576e-06,
+      "loss": 0.5958,
+      "step": 12085
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 41.6783247434428,
+      "learning_rate": 6.383097448728046e-06,
+      "loss": 0.5218,
+      "step": 12086
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.62357334827352,
+      "learning_rate": 6.3815252732323304e-06,
+      "loss": 0.5374,
+      "step": 12087
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.75340180498076,
+      "learning_rate": 6.379953200643915e-06,
+      "loss": 0.6049,
+      "step": 12088
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.82040321396851,
+      "learning_rate": 6.378381231007514e-06,
+      "loss": 0.5281,
+      "step": 12089
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 33.79275433255481,
+      "learning_rate": 6.376809364367833e-06,
+      "loss": 0.5034,
+      "step": 12090
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.363304036410135,
+      "learning_rate": 6.375237600769575e-06,
+      "loss": 0.4554,
+      "step": 12091
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.59225388908763,
+      "learning_rate": 6.373665940257436e-06,
+      "loss": 0.5819,
+      "step": 12092
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 34.98214809661372,
+      "learning_rate": 6.37209438287612e-06,
+      "loss": 0.5662,
+      "step": 12093
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 39.4511752109478,
+      "learning_rate": 6.370522928670313e-06,
+      "loss": 0.5026,
+      "step": 12094
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 50.42432986992907,
+      "learning_rate": 6.368951577684713e-06,
+      "loss": 0.5435,
+      "step": 12095
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.51103847346641,
+      "learning_rate": 6.367380329964009e-06,
+      "loss": 0.5507,
+      "step": 12096
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.30037761617256,
+      "learning_rate": 6.36580918555288e-06,
+      "loss": 0.5273,
+      "step": 12097
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 34.07097590972066,
+      "learning_rate": 6.364238144496011e-06,
+      "loss": 0.602,
+      "step": 12098
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.19280808976777,
+      "learning_rate": 6.362667206838085e-06,
+      "loss": 0.5024,
+      "step": 12099
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 41.431779714027286,
+      "learning_rate": 6.361096372623773e-06,
+      "loss": 0.519,
+      "step": 12100
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 49.20576707876023,
+      "learning_rate": 6.359525641897755e-06,
+      "loss": 0.6327,
+      "step": 12101
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 44.84990442025537,
+      "learning_rate": 6.357955014704698e-06,
+      "loss": 0.6776,
+      "step": 12102
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.506784391248154,
+      "learning_rate": 6.356384491089267e-06,
+      "loss": 0.6275,
+      "step": 12103
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 39.70972606635538,
+      "learning_rate": 6.354814071096129e-06,
+      "loss": 0.5285,
+      "step": 12104
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 46.07288793668887,
+      "learning_rate": 6.353243754769948e-06,
+      "loss": 0.6534,
+      "step": 12105
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.54255976217833,
+      "learning_rate": 6.351673542155385e-06,
+      "loss": 0.5691,
+      "step": 12106
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 36.185537966969406,
+      "learning_rate": 6.350103433297087e-06,
+      "loss": 0.5204,
+      "step": 12107
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 47.73052623859892,
+      "learning_rate": 6.348533428239714e-06,
+      "loss": 0.5934,
+      "step": 12108
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 44.935178587989355,
+      "learning_rate": 6.346963527027914e-06,
+      "loss": 0.5572,
+      "step": 12109
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 40.095660871013074,
+      "learning_rate": 6.345393729706333e-06,
+      "loss": 0.5477,
+      "step": 12110
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 54.906605701455916,
+      "learning_rate": 6.343824036319622e-06,
+      "loss": 0.6942,
+      "step": 12111
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.68735863408651,
+      "learning_rate": 6.3422544469124116e-06,
+      "loss": 0.5721,
+      "step": 12112
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 39.83934329328512,
+      "learning_rate": 6.340684961529344e-06,
+      "loss": 0.5382,
+      "step": 12113
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 45.42001483048987,
+      "learning_rate": 6.3391155802150585e-06,
+      "loss": 0.536,
+      "step": 12114
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.87083420840806,
+      "learning_rate": 6.337546303014182e-06,
+      "loss": 0.56,
+      "step": 12115
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 39.77056984725058,
+      "learning_rate": 6.3359771299713515e-06,
+      "loss": 0.548,
+      "step": 12116
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 31.81947330049135,
+      "learning_rate": 6.334408061131182e-06,
+      "loss": 0.5084,
+      "step": 12117
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.18241974335583,
+      "learning_rate": 6.332839096538306e-06,
+      "loss": 0.6007,
+      "step": 12118
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 44.658173163975576,
+      "learning_rate": 6.331270236237337e-06,
+      "loss": 0.5811,
+      "step": 12119
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.32751164526797,
+      "learning_rate": 6.3297014802729e-06,
+      "loss": 0.5353,
+      "step": 12120
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.23808448297408,
+      "learning_rate": 6.328132828689607e-06,
+      "loss": 0.5602,
+      "step": 12121
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.435144035895235,
+      "learning_rate": 6.326564281532066e-06,
+      "loss": 0.5187,
+      "step": 12122
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 34.88152027363758,
+      "learning_rate": 6.324995838844887e-06,
+      "loss": 0.5703,
+      "step": 12123
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 37.96037276672858,
+      "learning_rate": 6.3234275006726785e-06,
+      "loss": 0.5379,
+      "step": 12124
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 36.468769802424816,
+      "learning_rate": 6.321859267060039e-06,
+      "loss": 0.5857,
+      "step": 12125
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 41.92426037906549,
+      "learning_rate": 6.320291138051574e-06,
+      "loss": 0.5526,
+      "step": 12126
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.727078483341074,
+      "learning_rate": 6.318723113691874e-06,
+      "loss": 0.5713,
+      "step": 12127
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 29.93990274873482,
+      "learning_rate": 6.317155194025536e-06,
+      "loss": 0.4292,
+      "step": 12128
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 38.38890263098216,
+      "learning_rate": 6.315587379097148e-06,
+      "loss": 0.5068,
+      "step": 12129
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.674222094740458,
+      "learning_rate": 6.314019668951301e-06,
+      "loss": 0.4087,
+      "step": 12130
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.23505479249305,
+      "learning_rate": 6.312452063632581e-06,
+      "loss": 0.6756,
+      "step": 12131
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 54.12794749139059,
+      "learning_rate": 6.310884563185562e-06,
+      "loss": 0.6439,
+      "step": 12132
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 35.362232458586504,
+      "learning_rate": 6.309317167654832e-06,
+      "loss": 0.6423,
+      "step": 12133
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 55.25645138975995,
+      "learning_rate": 6.307749877084961e-06,
+      "loss": 0.6095,
+      "step": 12134
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 32.43953198663291,
+      "learning_rate": 6.306182691520522e-06,
+      "loss": 0.5042,
+      "step": 12135
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 42.41876426798832,
+      "learning_rate": 6.304615611006091e-06,
+      "loss": 0.5602,
+      "step": 12136
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 41.29216694199992,
+      "learning_rate": 6.303048635586226e-06,
+      "loss": 0.5513,
+      "step": 12137
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 34.24006366344465,
+      "learning_rate": 6.301481765305495e-06,
+      "loss": 0.5334,
+      "step": 12138
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 37.29238098390985,
+      "learning_rate": 6.2999150002084605e-06,
+      "loss": 0.4999,
+      "step": 12139
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 44.29202478622298,
+      "learning_rate": 6.298348340339681e-06,
+      "loss": 0.5717,
+      "step": 12140
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 37.89223188945311,
+      "learning_rate": 6.296781785743708e-06,
+      "loss": 0.5552,
+      "step": 12141
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.94620982464022,
+      "learning_rate": 6.295215336465094e-06,
+      "loss": 0.4875,
+      "step": 12142
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 44.66862707082449,
+      "learning_rate": 6.293648992548389e-06,
+      "loss": 0.5648,
+      "step": 12143
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 34.557240539528294,
+      "learning_rate": 6.292082754038135e-06,
+      "loss": 0.5058,
+      "step": 12144
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 44.651212922054135,
+      "learning_rate": 6.2905166209788845e-06,
+      "loss": 0.5174,
+      "step": 12145
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 46.1265563133698,
+      "learning_rate": 6.288950593415168e-06,
+      "loss": 0.6229,
+      "step": 12146
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.90136585755105,
+      "learning_rate": 6.2873846713915276e-06,
+      "loss": 0.5333,
+      "step": 12147
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.361026673972525,
+      "learning_rate": 6.285818854952492e-06,
+      "loss": 0.4888,
+      "step": 12148
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 39.80473852332142,
+      "learning_rate": 6.284253144142598e-06,
+      "loss": 0.494,
+      "step": 12149
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 34.33017969385017,
+      "learning_rate": 6.282687539006371e-06,
+      "loss": 0.4533,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 53.497934907647924,
+      "learning_rate": 6.2811220395883346e-06,
+      "loss": 0.6214,
+      "step": 12151
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.67921524206435,
+      "learning_rate": 6.279556645933011e-06,
+      "loss": 0.5699,
+      "step": 12152
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 43.56356891212764,
+      "learning_rate": 6.277991358084923e-06,
+      "loss": 0.5509,
+      "step": 12153
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 35.88590815317829,
+      "learning_rate": 6.276426176088579e-06,
+      "loss": 0.5099,
+      "step": 12154
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 33.0968660899379,
+      "learning_rate": 6.274861099988502e-06,
+      "loss": 0.4724,
+      "step": 12155
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 35.1625852242748,
+      "learning_rate": 6.273296129829194e-06,
+      "loss": 0.489,
+      "step": 12156
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.5169493631354,
+      "learning_rate": 6.271731265655161e-06,
+      "loss": 0.5495,
+      "step": 12157
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 34.00316650707824,
+      "learning_rate": 6.270166507510913e-06,
+      "loss": 0.4778,
+      "step": 12158
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 48.297586387052704,
+      "learning_rate": 6.268601855440944e-06,
+      "loss": 0.5505,
+      "step": 12159
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.202382691309886,
+      "learning_rate": 6.267037309489763e-06,
+      "loss": 0.5533,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.72086182710099,
+      "learning_rate": 6.265472869701851e-06,
+      "loss": 0.4448,
+      "step": 12161
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 41.03422372912271,
+      "learning_rate": 6.2639085361217075e-06,
+      "loss": 0.5295,
+      "step": 12162
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 47.14607949163838,
+      "learning_rate": 6.262344308793818e-06,
+      "loss": 0.6404,
+      "step": 12163
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 34.888384520314666,
+      "learning_rate": 6.2607801877626715e-06,
+      "loss": 0.4954,
+      "step": 12164
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.27594892894006,
+      "learning_rate": 6.259216173072751e-06,
+      "loss": 0.5553,
+      "step": 12165
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.836336281358335,
+      "learning_rate": 6.257652264768532e-06,
+      "loss": 0.5136,
+      "step": 12166
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 37.88615667364935,
+      "learning_rate": 6.256088462894492e-06,
+      "loss": 0.5725,
+      "step": 12167
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 42.33455751869266,
+      "learning_rate": 6.254524767495108e-06,
+      "loss": 0.6187,
+      "step": 12168
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 42.8186308042142,
+      "learning_rate": 6.252961178614847e-06,
+      "loss": 0.4944,
+      "step": 12169
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 42.13866620007385,
+      "learning_rate": 6.251397696298181e-06,
+      "loss": 0.5272,
+      "step": 12170
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 45.846891863709786,
+      "learning_rate": 6.24983432058957e-06,
+      "loss": 0.5834,
+      "step": 12171
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 43.058964621555816,
+      "learning_rate": 6.2482710515334775e-06,
+      "loss": 0.5776,
+      "step": 12172
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.79866174111257,
+      "learning_rate": 6.246707889174359e-06,
+      "loss": 0.6064,
+      "step": 12173
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 42.062787382325936,
+      "learning_rate": 6.245144833556675e-06,
+      "loss": 0.5664,
+      "step": 12174
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.760687387914054,
+      "learning_rate": 6.243581884724879e-06,
+      "loss": 0.5519,
+      "step": 12175
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 43.505230956328774,
+      "learning_rate": 6.242019042723411e-06,
+      "loss": 0.5334,
+      "step": 12176
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.768530575598284,
+      "learning_rate": 6.240456307596725e-06,
+      "loss": 0.4747,
+      "step": 12177
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.98859957202069,
+      "learning_rate": 6.238893679389262e-06,
+      "loss": 0.5214,
+      "step": 12178
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.164806253298025,
+      "learning_rate": 6.237331158145461e-06,
+      "loss": 0.5632,
+      "step": 12179
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.182427764022606,
+      "learning_rate": 6.2357687439097645e-06,
+      "loss": 0.5064,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 34.09688131007018,
+      "learning_rate": 6.234206436726602e-06,
+      "loss": 0.5151,
+      "step": 12181
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 44.08894655597914,
+      "learning_rate": 6.232644236640403e-06,
+      "loss": 0.4552,
+      "step": 12182
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 39.989656440182465,
+      "learning_rate": 6.231082143695599e-06,
+      "loss": 0.5475,
+      "step": 12183
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.16091565499077,
+      "learning_rate": 6.229520157936617e-06,
+      "loss": 0.5413,
+      "step": 12184
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.73713372571343,
+      "learning_rate": 6.2279582794078755e-06,
+      "loss": 0.5261,
+      "step": 12185
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 44.16597575900855,
+      "learning_rate": 6.226396508153789e-06,
+      "loss": 0.6837,
+      "step": 12186
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.129184391822,
+      "learning_rate": 6.2248348442187825e-06,
+      "loss": 0.5201,
+      "step": 12187
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 46.649531489449274,
+      "learning_rate": 6.223273287647262e-06,
+      "loss": 0.5953,
+      "step": 12188
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 44.252847790003045,
+      "learning_rate": 6.221711838483644e-06,
+      "loss": 0.5252,
+      "step": 12189
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.43252861618211,
+      "learning_rate": 6.220150496772328e-06,
+      "loss": 0.529,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.190809012861585,
+      "learning_rate": 6.2185892625577215e-06,
+      "loss": 0.5225,
+      "step": 12191
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 41.4501917290481,
+      "learning_rate": 6.217028135884222e-06,
+      "loss": 0.5602,
+      "step": 12192
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 41.857327612302214,
+      "learning_rate": 6.215467116796232e-06,
+      "loss": 0.5072,
+      "step": 12193
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 39.798512131948726,
+      "learning_rate": 6.213906205338145e-06,
+      "loss": 0.5622,
+      "step": 12194
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 43.64919513995959,
+      "learning_rate": 6.212345401554347e-06,
+      "loss": 0.5928,
+      "step": 12195
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 43.52701539791556,
+      "learning_rate": 6.2107847054892325e-06,
+      "loss": 0.5367,
+      "step": 12196
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 45.83898347649571,
+      "learning_rate": 6.209224117187184e-06,
+      "loss": 0.6421,
+      "step": 12197
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.18334922433527,
+      "learning_rate": 6.207663636692581e-06,
+      "loss": 0.5565,
+      "step": 12198
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 44.51055157325037,
+      "learning_rate": 6.2061032640498095e-06,
+      "loss": 0.5028,
+      "step": 12199
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.29212881628125,
+      "learning_rate": 6.204542999303242e-06,
+      "loss": 0.6383,
+      "step": 12200
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 32.48222905456663,
+      "learning_rate": 6.202982842497247e-06,
+      "loss": 0.4684,
+      "step": 12201
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 43.4448670431485,
+      "learning_rate": 6.201422793676203e-06,
+      "loss": 0.559,
+      "step": 12202
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 33.76717640403215,
+      "learning_rate": 6.1998628528844685e-06,
+      "loss": 0.5618,
+      "step": 12203
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 36.58881146459332,
+      "learning_rate": 6.198303020166418e-06,
+      "loss": 0.5022,
+      "step": 12204
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 38.39263992083365,
+      "learning_rate": 6.196743295566401e-06,
+      "loss": 0.4738,
+      "step": 12205
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.02826355197787,
+      "learning_rate": 6.195183679128781e-06,
+      "loss": 0.5694,
+      "step": 12206
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 34.45355085944467,
+      "learning_rate": 6.19362417089791e-06,
+      "loss": 0.4788,
+      "step": 12207
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 32.8661216524935,
+      "learning_rate": 6.1920647709181425e-06,
+      "loss": 0.4736,
+      "step": 12208
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 35.696061311625115,
+      "learning_rate": 6.190505479233826e-06,
+      "loss": 0.5027,
+      "step": 12209
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 43.298365357711155,
+      "learning_rate": 6.188946295889304e-06,
+      "loss": 0.5835,
+      "step": 12210
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.71575880426689,
+      "learning_rate": 6.18738722092892e-06,
+      "loss": 0.5186,
+      "step": 12211
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 58.744508559687425,
+      "learning_rate": 6.185828254397012e-06,
+      "loss": 0.4656,
+      "step": 12212
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.74493511188896,
+      "learning_rate": 6.184269396337916e-06,
+      "loss": 0.5212,
+      "step": 12213
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.75549161802401,
+      "learning_rate": 6.182710646795972e-06,
+      "loss": 0.5473,
+      "step": 12214
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 44.0636236006954,
+      "learning_rate": 6.181152005815501e-06,
+      "loss": 0.5628,
+      "step": 12215
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.742405189525094,
+      "learning_rate": 6.179593473440833e-06,
+      "loss": 0.5457,
+      "step": 12216
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 41.50602054587094,
+      "learning_rate": 6.178035049716289e-06,
+      "loss": 0.5365,
+      "step": 12217
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.01893352677413,
+      "learning_rate": 6.1764767346861955e-06,
+      "loss": 0.53,
+      "step": 12218
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 37.711750876042046,
+      "learning_rate": 6.174918528394868e-06,
+      "loss": 0.5639,
+      "step": 12219
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 46.93099674527881,
+      "learning_rate": 6.173360430886617e-06,
+      "loss": 0.5466,
+      "step": 12220
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 44.106424098504355,
+      "learning_rate": 6.171802442205759e-06,
+      "loss": 0.4885,
+      "step": 12221
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.73116255546618,
+      "learning_rate": 6.170244562396599e-06,
+      "loss": 0.586,
+      "step": 12222
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.26133688614866,
+      "learning_rate": 6.1686867915034425e-06,
+      "loss": 0.5161,
+      "step": 12223
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.894615109851685,
+      "learning_rate": 6.167129129570595e-06,
+      "loss": 0.5301,
+      "step": 12224
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 34.00833867351961,
+      "learning_rate": 6.165571576642351e-06,
+      "loss": 0.4866,
+      "step": 12225
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.85993918618351,
+      "learning_rate": 6.164014132763008e-06,
+      "loss": 0.5525,
+      "step": 12226
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.812137852828634,
+      "learning_rate": 6.162456797976859e-06,
+      "loss": 0.5983,
+      "step": 12227
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 44.27447927503703,
+      "learning_rate": 6.1608995723281925e-06,
+      "loss": 0.6177,
+      "step": 12228
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 41.40861412352854,
+      "learning_rate": 6.159342455861304e-06,
+      "loss": 0.5055,
+      "step": 12229
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 29.61427706599101,
+      "learning_rate": 6.157785448620461e-06,
+      "loss": 0.5005,
+      "step": 12230
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 41.081476329491665,
+      "learning_rate": 6.156228550649954e-06,
+      "loss": 0.6057,
+      "step": 12231
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.42998584747079,
+      "learning_rate": 6.154671761994058e-06,
+      "loss": 0.5655,
+      "step": 12232
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.959272680768166,
+      "learning_rate": 6.153115082697052e-06,
+      "loss": 0.5738,
+      "step": 12233
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 35.75575071712051,
+      "learning_rate": 6.1515585128032e-06,
+      "loss": 0.497,
+      "step": 12234
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.71257166677413,
+      "learning_rate": 6.150002052356772e-06,
+      "loss": 0.5639,
+      "step": 12235
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 37.8985073802196,
+      "learning_rate": 6.148445701402032e-06,
+      "loss": 0.541,
+      "step": 12236
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.65231517151491,
+      "learning_rate": 6.1468894599832455e-06,
+      "loss": 0.4813,
+      "step": 12237
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 32.33395433484995,
+      "learning_rate": 6.145333328144669e-06,
+      "loss": 0.4985,
+      "step": 12238
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 35.26735574838406,
+      "learning_rate": 6.143777305930556e-06,
+      "loss": 0.4636,
+      "step": 12239
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 33.36152561411356,
+      "learning_rate": 6.142221393385161e-06,
+      "loss": 0.4812,
+      "step": 12240
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 36.112585541546935,
+      "learning_rate": 6.140665590552735e-06,
+      "loss": 0.4798,
+      "step": 12241
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 46.55097367142158,
+      "learning_rate": 6.139109897477517e-06,
+      "loss": 0.5701,
+      "step": 12242
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 34.782880444968185,
+      "learning_rate": 6.137554314203762e-06,
+      "loss": 0.514,
+      "step": 12243
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.24458653481411,
+      "learning_rate": 6.135998840775699e-06,
+      "loss": 0.5493,
+      "step": 12244
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 50.16526880671904,
+      "learning_rate": 6.134443477237566e-06,
+      "loss": 0.5949,
+      "step": 12245
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 58.42103836019603,
+      "learning_rate": 6.132888223633602e-06,
+      "loss": 0.6476,
+      "step": 12246
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 43.38437867802417,
+      "learning_rate": 6.131333080008033e-06,
+      "loss": 0.5255,
+      "step": 12247
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.826802074287386,
+      "learning_rate": 6.129778046405095e-06,
+      "loss": 0.5233,
+      "step": 12248
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 38.879176182299936,
+      "learning_rate": 6.128223122868997e-06,
+      "loss": 0.5717,
+      "step": 12249
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.11812061981519,
+      "learning_rate": 6.126668309443973e-06,
+      "loss": 0.5796,
+      "step": 12250
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 45.65820876455723,
+      "learning_rate": 6.125113606174233e-06,
+      "loss": 0.5331,
+      "step": 12251
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 41.03369698705499,
+      "learning_rate": 6.123559013103997e-06,
+      "loss": 0.5641,
+      "step": 12252
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 36.48092018123436,
+      "learning_rate": 6.122004530277476e-06,
+      "loss": 0.5358,
+      "step": 12253
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 46.14144959247426,
+      "learning_rate": 6.1204501577388775e-06,
+      "loss": 0.4923,
+      "step": 12254
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 37.4399988744707,
+      "learning_rate": 6.118895895532405e-06,
+      "loss": 0.4695,
+      "step": 12255
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 37.39149034441687,
+      "learning_rate": 6.117341743702265e-06,
+      "loss": 0.4791,
+      "step": 12256
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 37.85255261711755,
+      "learning_rate": 6.115787702292651e-06,
+      "loss": 0.4671,
+      "step": 12257
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 44.616153545419124,
+      "learning_rate": 6.1142337713477685e-06,
+      "loss": 0.5377,
+      "step": 12258
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 33.95036365429441,
+      "learning_rate": 6.112679950911802e-06,
+      "loss": 0.4329,
+      "step": 12259
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 31.71246187616026,
+      "learning_rate": 6.111126241028942e-06,
+      "loss": 0.4702,
+      "step": 12260
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 36.54745409548714,
+      "learning_rate": 6.1095726417433765e-06,
+      "loss": 0.5143,
+      "step": 12261
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 34.93586522066771,
+      "learning_rate": 6.10801915309929e-06,
+      "loss": 0.5117,
+      "step": 12262
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 49.234400684733764,
+      "learning_rate": 6.106465775140865e-06,
+      "loss": 0.6089,
+      "step": 12263
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 32.501208682692905,
+      "learning_rate": 6.104912507912272e-06,
+      "loss": 0.5122,
+      "step": 12264
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.949772114628644,
+      "learning_rate": 6.1033593514576895e-06,
+      "loss": 0.5245,
+      "step": 12265
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.38072957246148,
+      "learning_rate": 6.101806305821288e-06,
+      "loss": 0.5159,
+      "step": 12266
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 45.90502898979539,
+      "learning_rate": 6.100253371047233e-06,
+      "loss": 0.5785,
+      "step": 12267
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.52996705851568,
+      "learning_rate": 6.098700547179695e-06,
+      "loss": 0.5337,
+      "step": 12268
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.415299194730785,
+      "learning_rate": 6.0971478342628285e-06,
+      "loss": 0.4971,
+      "step": 12269
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.53105776544615,
+      "learning_rate": 6.095595232340794e-06,
+      "loss": 0.5001,
+      "step": 12270
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.159980953929555,
+      "learning_rate": 6.094042741457749e-06,
+      "loss": 0.58,
+      "step": 12271
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.99621272664088,
+      "learning_rate": 6.0924903616578394e-06,
+      "loss": 0.5742,
+      "step": 12272
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.72179472915497,
+      "learning_rate": 6.090938092985226e-06,
+      "loss": 0.5828,
+      "step": 12273
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.3651765473263,
+      "learning_rate": 6.08938593548404e-06,
+      "loss": 0.5693,
+      "step": 12274
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 46.103744126820025,
+      "learning_rate": 6.087833889198431e-06,
+      "loss": 0.6398,
+      "step": 12275
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 42.82282272968214,
+      "learning_rate": 6.086281954172537e-06,
+      "loss": 0.525,
+      "step": 12276
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 46.79564348998119,
+      "learning_rate": 6.084730130450497e-06,
+      "loss": 0.5911,
+      "step": 12277
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.75459686830181,
+      "learning_rate": 6.083178418076443e-06,
+      "loss": 0.6332,
+      "step": 12278
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.110559793083034,
+      "learning_rate": 6.081626817094502e-06,
+      "loss": 0.5041,
+      "step": 12279
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.52333808105819,
+      "learning_rate": 6.0800753275488e-06,
+      "loss": 0.5798,
+      "step": 12280
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.700332456631095,
+      "learning_rate": 6.078523949483465e-06,
+      "loss": 0.5686,
+      "step": 12281
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 35.497526599791584,
+      "learning_rate": 6.076972682942617e-06,
+      "loss": 0.5146,
+      "step": 12282
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.51401577135634,
+      "learning_rate": 6.075421527970366e-06,
+      "loss": 0.5642,
+      "step": 12283
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.74673557377332,
+      "learning_rate": 6.073870484610833e-06,
+      "loss": 0.4928,
+      "step": 12284
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.50682356106145,
+      "learning_rate": 6.072319552908128e-06,
+      "loss": 0.5334,
+      "step": 12285
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 31.101673462360047,
+      "learning_rate": 6.070768732906353e-06,
+      "loss": 0.4731,
+      "step": 12286
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 46.002124830596124,
+      "learning_rate": 6.069218024649622e-06,
+      "loss": 0.5241,
+      "step": 12287
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.4238289451716,
+      "learning_rate": 6.0676674281820294e-06,
+      "loss": 0.5132,
+      "step": 12288
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.67522015517329,
+      "learning_rate": 6.066116943547673e-06,
+      "loss": 0.5185,
+      "step": 12289
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 42.04155034344361,
+      "learning_rate": 6.064566570790651e-06,
+      "loss": 0.529,
+      "step": 12290
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 48.363798998486246,
+      "learning_rate": 6.063016309955051e-06,
+      "loss": 0.5954,
+      "step": 12291
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 42.92464301727922,
+      "learning_rate": 6.06146616108497e-06,
+      "loss": 0.5059,
+      "step": 12292
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.198387299338464,
+      "learning_rate": 6.059916124224483e-06,
+      "loss": 0.5439,
+      "step": 12293
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 44.9778921604325,
+      "learning_rate": 6.058366199417676e-06,
+      "loss": 0.6669,
+      "step": 12294
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 34.7181268356255,
+      "learning_rate": 6.056816386708628e-06,
+      "loss": 0.4582,
+      "step": 12295
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 44.93989060559644,
+      "learning_rate": 6.055266686141418e-06,
+      "loss": 0.5619,
+      "step": 12296
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.016579741041745,
+      "learning_rate": 6.053717097760117e-06,
+      "loss": 0.5229,
+      "step": 12297
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 35.03700124744224,
+      "learning_rate": 6.052167621608791e-06,
+      "loss": 0.5472,
+      "step": 12298
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 40.04162872278253,
+      "learning_rate": 6.050618257731507e-06,
+      "loss": 0.5047,
+      "step": 12299
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 50.7847160682354,
+      "learning_rate": 6.049069006172331e-06,
+      "loss": 0.5477,
+      "step": 12300
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 44.05314094222233,
+      "learning_rate": 6.04751986697532e-06,
+      "loss": 0.5523,
+      "step": 12301
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.26506051182623,
+      "learning_rate": 6.045970840184534e-06,
+      "loss": 0.4963,
+      "step": 12302
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.14565876706464,
+      "learning_rate": 6.044421925844024e-06,
+      "loss": 0.5252,
+      "step": 12303
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 50.61449954548105,
+      "learning_rate": 6.0428731239978415e-06,
+      "loss": 0.5515,
+      "step": 12304
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 50.324675362720015,
+      "learning_rate": 6.041324434690028e-06,
+      "loss": 0.5658,
+      "step": 12305
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.178996199095764,
+      "learning_rate": 6.0397758579646335e-06,
+      "loss": 0.4897,
+      "step": 12306
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 43.2309935294818,
+      "learning_rate": 6.038227393865701e-06,
+      "loss": 0.5494,
+      "step": 12307
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.30148127047772,
+      "learning_rate": 6.036679042437259e-06,
+      "loss": 0.5163,
+      "step": 12308
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 35.6522222178095,
+      "learning_rate": 6.035130803723349e-06,
+      "loss": 0.4763,
+      "step": 12309
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.17415791213843,
+      "learning_rate": 6.033582677767997e-06,
+      "loss": 0.5655,
+      "step": 12310
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.24480789612871,
+      "learning_rate": 6.032034664615236e-06,
+      "loss": 0.4797,
+      "step": 12311
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.025531691056955,
+      "learning_rate": 6.030486764309089e-06,
+      "loss": 0.4557,
+      "step": 12312
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 40.08981419881583,
+      "learning_rate": 6.028938976893575e-06,
+      "loss": 0.5231,
+      "step": 12313
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.01485084557002,
+      "learning_rate": 6.027391302412711e-06,
+      "loss": 0.5882,
+      "step": 12314
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.356035836727365,
+      "learning_rate": 6.025843740910518e-06,
+      "loss": 0.461,
+      "step": 12315
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 37.42371752190898,
+      "learning_rate": 6.024296292431e-06,
+      "loss": 0.5164,
+      "step": 12316
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 34.7379558877862,
+      "learning_rate": 6.022748957018176e-06,
+      "loss": 0.4086,
+      "step": 12317
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 39.25127960354601,
+      "learning_rate": 6.02120173471604e-06,
+      "loss": 0.5759,
+      "step": 12318
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 43.1217154178062,
+      "learning_rate": 6.0196546255686016e-06,
+      "loss": 0.5513,
+      "step": 12319
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 48.23853447677991,
+      "learning_rate": 6.018107629619854e-06,
+      "loss": 0.5977,
+      "step": 12320
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 38.89973556029061,
+      "learning_rate": 6.0165607469137984e-06,
+      "loss": 0.5868,
+      "step": 12321
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 47.362330239727434,
+      "learning_rate": 6.015013977494429e-06,
+      "loss": 0.5407,
+      "step": 12322
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 32.74388344262956,
+      "learning_rate": 6.013467321405726e-06,
+      "loss": 0.4196,
+      "step": 12323
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 50.345105655539214,
+      "learning_rate": 6.011920778691679e-06,
+      "loss": 0.6263,
+      "step": 12324
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 44.774989540358725,
+      "learning_rate": 6.010374349396274e-06,
+      "loss": 0.5275,
+      "step": 12325
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.318045057566664,
+      "learning_rate": 6.008828033563487e-06,
+      "loss": 0.4849,
+      "step": 12326
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 30.78018477479849,
+      "learning_rate": 6.0072818312373e-06,
+      "loss": 0.4626,
+      "step": 12327
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 46.81806179870446,
+      "learning_rate": 6.005735742461681e-06,
+      "loss": 0.5719,
+      "step": 12328
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 62.273166244810035,
+      "learning_rate": 6.004189767280599e-06,
+      "loss": 0.6023,
+      "step": 12329
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 41.2269767608851,
+      "learning_rate": 6.002643905738022e-06,
+      "loss": 0.5079,
+      "step": 12330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 36.206316060250096,
+      "learning_rate": 6.0010981578779175e-06,
+      "loss": 0.4782,
+      "step": 12331
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 49.111061661946145,
+      "learning_rate": 5.9995525237442395e-06,
+      "loss": 0.5861,
+      "step": 12332
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.587492944618226,
+      "learning_rate": 5.9980070033809455e-06,
+      "loss": 0.5356,
+      "step": 12333
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 46.91738110953127,
+      "learning_rate": 5.996461596831995e-06,
+      "loss": 0.5302,
+      "step": 12334
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 38.919652991036784,
+      "learning_rate": 5.9949163041413315e-06,
+      "loss": 0.5588,
+      "step": 12335
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 37.06752932424319,
+      "learning_rate": 5.9933711253529115e-06,
+      "loss": 0.519,
+      "step": 12336
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 41.853998637048036,
+      "learning_rate": 5.991826060510666e-06,
+      "loss": 0.5471,
+      "step": 12337
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 39.630988423642755,
+      "learning_rate": 5.990281109658545e-06,
+      "loss": 0.5241,
+      "step": 12338
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 43.020584064466554,
+      "learning_rate": 5.988736272840483e-06,
+      "loss": 0.589,
+      "step": 12339
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 49.74001326077249,
+      "learning_rate": 5.987191550100415e-06,
+      "loss": 0.5557,
+      "step": 12340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 44.283961706377696,
+      "learning_rate": 5.985646941482275e-06,
+      "loss": 0.5834,
+      "step": 12341
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 33.44250857551934,
+      "learning_rate": 5.984102447029986e-06,
+      "loss": 0.4771,
+      "step": 12342
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 38.690926635788976,
+      "learning_rate": 5.9825580667874715e-06,
+      "loss": 0.5747,
+      "step": 12343
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 40.22248889347027,
+      "learning_rate": 5.981013800798657e-06,
+      "loss": 0.5111,
+      "step": 12344
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 100.1652110645152,
+      "learning_rate": 5.9794696491074576e-06,
+      "loss": 0.5456,
+      "step": 12345
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 53.51192154815627,
+      "learning_rate": 5.977925611757794e-06,
+      "loss": 0.5451,
+      "step": 12346
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 33.69427976876404,
+      "learning_rate": 5.976381688793571e-06,
+      "loss": 0.5724,
+      "step": 12347
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 36.81174832533545,
+      "learning_rate": 5.974837880258697e-06,
+      "loss": 0.5212,
+      "step": 12348
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 36.80382875150403,
+      "learning_rate": 5.973294186197079e-06,
+      "loss": 0.5717,
+      "step": 12349
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.76606112273007,
+      "learning_rate": 5.971750606652621e-06,
+      "loss": 0.4909,
+      "step": 12350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 38.10759837321003,
+      "learning_rate": 5.97020714166922e-06,
+      "loss": 0.5623,
+      "step": 12351
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 36.324422635335814,
+      "learning_rate": 5.9686637912907675e-06,
+      "loss": 0.4754,
+      "step": 12352
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 34.108434775128856,
+      "learning_rate": 5.96712055556116e-06,
+      "loss": 0.5233,
+      "step": 12353
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 33.12237539150038,
+      "learning_rate": 5.965577434524283e-06,
+      "loss": 0.4604,
+      "step": 12354
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 45.61184229618336,
+      "learning_rate": 5.964034428224026e-06,
+      "loss": 0.6054,
+      "step": 12355
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.95509965589174,
+      "learning_rate": 5.9624915367042705e-06,
+      "loss": 0.491,
+      "step": 12356
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 42.90817205870821,
+      "learning_rate": 5.960948760008891e-06,
+      "loss": 0.5931,
+      "step": 12357
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 43.00534587719687,
+      "learning_rate": 5.959406098181766e-06,
+      "loss": 0.5892,
+      "step": 12358
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 40.306317048942034,
+      "learning_rate": 5.9578635512667686e-06,
+      "loss": 0.4705,
+      "step": 12359
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 32.174003260185664,
+      "learning_rate": 5.9563211193077644e-06,
+      "loss": 0.4902,
+      "step": 12360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.35642945229067,
+      "learning_rate": 5.9547788023486296e-06,
+      "loss": 0.5316,
+      "step": 12361
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 30.354080594546666,
+      "learning_rate": 5.953236600433212e-06,
+      "loss": 0.4907,
+      "step": 12362
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 44.3906990246691,
+      "learning_rate": 5.951694513605381e-06,
+      "loss": 0.5479,
+      "step": 12363
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 39.070777900973525,
+      "learning_rate": 5.950152541908988e-06,
+      "loss": 0.5491,
+      "step": 12364
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 41.04600225420368,
+      "learning_rate": 5.94861068538789e-06,
+      "loss": 0.5888,
+      "step": 12365
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 40.29729895426036,
+      "learning_rate": 5.947068944085934e-06,
+      "loss": 0.6238,
+      "step": 12366
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 46.857306938162395,
+      "learning_rate": 5.9455273180469666e-06,
+      "loss": 0.6623,
+      "step": 12367
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 37.152392666010115,
+      "learning_rate": 5.943985807314827e-06,
+      "loss": 0.5527,
+      "step": 12368
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 34.47415910335682,
+      "learning_rate": 5.942444411933361e-06,
+      "loss": 0.4996,
+      "step": 12369
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.94133652152839,
+      "learning_rate": 5.940903131946401e-06,
+      "loss": 0.4797,
+      "step": 12370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.65234858158732,
+      "learning_rate": 5.939361967397785e-06,
+      "loss": 0.5154,
+      "step": 12371
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 40.971323664271644,
+      "learning_rate": 5.937820918331336e-06,
+      "loss": 0.5358,
+      "step": 12372
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.987091200462054,
+      "learning_rate": 5.9362799847908845e-06,
+      "loss": 0.4444,
+      "step": 12373
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 34.09218867264481,
+      "learning_rate": 5.934739166820252e-06,
+      "loss": 0.4815,
+      "step": 12374
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 37.97351495175872,
+      "learning_rate": 5.9331984644632615e-06,
+      "loss": 0.6047,
+      "step": 12375
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 39.194839721042484,
+      "learning_rate": 5.931657877763728e-06,
+      "loss": 0.5582,
+      "step": 12376
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 70.98147393692025,
+      "learning_rate": 5.930117406765462e-06,
+      "loss": 0.5181,
+      "step": 12377
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 60.21794529344031,
+      "learning_rate": 5.928577051512277e-06,
+      "loss": 0.577,
+      "step": 12378
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 43.24370755192018,
+      "learning_rate": 5.927036812047978e-06,
+      "loss": 0.5628,
+      "step": 12379
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 46.27822976879404,
+      "learning_rate": 5.925496688416374e-06,
+      "loss": 0.5361,
+      "step": 12380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 79.99212560046804,
+      "learning_rate": 5.923956680661257e-06,
+      "loss": 0.6734,
+      "step": 12381
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 51.47910443709639,
+      "learning_rate": 5.922416788826429e-06,
+      "loss": 0.6462,
+      "step": 12382
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 42.14791161577664,
+      "learning_rate": 5.920877012955679e-06,
+      "loss": 0.5156,
+      "step": 12383
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.30381116393557,
+      "learning_rate": 5.919337353092802e-06,
+      "loss": 0.526,
+      "step": 12384
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 41.99122156637956,
+      "learning_rate": 5.917797809281587e-06,
+      "loss": 0.5269,
+      "step": 12385
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 30.738869288515048,
+      "learning_rate": 5.916258381565811e-06,
+      "loss": 0.4518,
+      "step": 12386
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 36.379252047612056,
+      "learning_rate": 5.914719069989257e-06,
+      "loss": 0.5124,
+      "step": 12387
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 34.09670079763372,
+      "learning_rate": 5.913179874595705e-06,
+      "loss": 0.4789,
+      "step": 12388
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 45.265248822145956,
+      "learning_rate": 5.911640795428923e-06,
+      "loss": 0.5068,
+      "step": 12389
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 41.316525706078565,
+      "learning_rate": 5.910101832532691e-06,
+      "loss": 0.5779,
+      "step": 12390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.71982565343983,
+      "learning_rate": 5.908562985950768e-06,
+      "loss": 0.4764,
+      "step": 12391
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.39829294753997,
+      "learning_rate": 5.907024255726919e-06,
+      "loss": 0.5626,
+      "step": 12392
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 43.86083979754574,
+      "learning_rate": 5.905485641904904e-06,
+      "loss": 0.5582,
+      "step": 12393
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.05881409545897,
+      "learning_rate": 5.903947144528486e-06,
+      "loss": 0.5123,
+      "step": 12394
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 40.75921049347337,
+      "learning_rate": 5.902408763641416e-06,
+      "loss": 0.5557,
+      "step": 12395
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 35.63934959728479,
+      "learning_rate": 5.9008704992874384e-06,
+      "loss": 0.5195,
+      "step": 12396
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 50.303750958517206,
+      "learning_rate": 5.89933235151031e-06,
+      "loss": 0.5764,
+      "step": 12397
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 36.05463786178594,
+      "learning_rate": 5.897794320353768e-06,
+      "loss": 0.4893,
+      "step": 12398
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.34631024127474,
+      "learning_rate": 5.896256405861558e-06,
+      "loss": 0.4516,
+      "step": 12399
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.13947461191329,
+      "learning_rate": 5.8947186080774165e-06,
+      "loss": 0.5239,
+      "step": 12400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 43.68152608125142,
+      "learning_rate": 5.8931809270450746e-06,
+      "loss": 0.5514,
+      "step": 12401
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 44.347688463734116,
+      "learning_rate": 5.8916433628082626e-06,
+      "loss": 0.6167,
+      "step": 12402
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 34.55903009997471,
+      "learning_rate": 5.890105915410712e-06,
+      "loss": 0.4684,
+      "step": 12403
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.81884168448855,
+      "learning_rate": 5.8885685848961424e-06,
+      "loss": 0.5427,
+      "step": 12404
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.84376764633686,
+      "learning_rate": 5.8870313713082845e-06,
+      "loss": 0.4932,
+      "step": 12405
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 32.59441802966898,
+      "learning_rate": 5.885494274690842e-06,
+      "loss": 0.454,
+      "step": 12406
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 53.21072021135327,
+      "learning_rate": 5.883957295087537e-06,
+      "loss": 0.5653,
+      "step": 12407
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.093087467617536,
+      "learning_rate": 5.882420432542077e-06,
+      "loss": 0.4966,
+      "step": 12408
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.09065795133328,
+      "learning_rate": 5.8808836870981725e-06,
+      "loss": 0.5395,
+      "step": 12409
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 34.51485852227579,
+      "learning_rate": 5.87934705879953e-06,
+      "loss": 0.5104,
+      "step": 12410
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 32.17804863135204,
+      "learning_rate": 5.8778105476898415e-06,
+      "loss": 0.4767,
+      "step": 12411
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.21666408326162,
+      "learning_rate": 5.87627415381281e-06,
+      "loss": 0.5465,
+      "step": 12412
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 36.97621063703343,
+      "learning_rate": 5.874737877212131e-06,
+      "loss": 0.4933,
+      "step": 12413
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 40.688946573703085,
+      "learning_rate": 5.873201717931492e-06,
+      "loss": 0.5198,
+      "step": 12414
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 40.11653403220751,
+      "learning_rate": 5.871665676014584e-06,
+      "loss": 0.5157,
+      "step": 12415
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 48.2750108706019,
+      "learning_rate": 5.870129751505089e-06,
+      "loss": 0.6018,
+      "step": 12416
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 45.55015885569303,
+      "learning_rate": 5.868593944446688e-06,
+      "loss": 0.6205,
+      "step": 12417
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 38.63214410062865,
+      "learning_rate": 5.867058254883056e-06,
+      "loss": 0.5569,
+      "step": 12418
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.74622713138073,
+      "learning_rate": 5.8655226828578715e-06,
+      "loss": 0.5356,
+      "step": 12419
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 47.91221611860428,
+      "learning_rate": 5.863987228414805e-06,
+      "loss": 0.6016,
+      "step": 12420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.824931433092424,
+      "learning_rate": 5.8624518915975185e-06,
+      "loss": 0.6156,
+      "step": 12421
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 32.51601100621018,
+      "learning_rate": 5.860916672449682e-06,
+      "loss": 0.5162,
+      "step": 12422
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.00975322084009,
+      "learning_rate": 5.859381571014951e-06,
+      "loss": 0.5037,
+      "step": 12423
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 44.72622085907235,
+      "learning_rate": 5.857846587336989e-06,
+      "loss": 0.6016,
+      "step": 12424
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.90275708403188,
+      "learning_rate": 5.856311721459449e-06,
+      "loss": 0.5552,
+      "step": 12425
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 40.75133792751987,
+      "learning_rate": 5.854776973425978e-06,
+      "loss": 0.6326,
+      "step": 12426
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 36.81394162941626,
+      "learning_rate": 5.853242343280222e-06,
+      "loss": 0.5588,
+      "step": 12427
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 41.16238053238182,
+      "learning_rate": 5.85170783106583e-06,
+      "loss": 0.5082,
+      "step": 12428
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 83.23907569714869,
+      "learning_rate": 5.8501734368264425e-06,
+      "loss": 0.4567,
+      "step": 12429
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.22549080842501,
+      "learning_rate": 5.848639160605694e-06,
+      "loss": 0.4814,
+      "step": 12430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.04240692565709,
+      "learning_rate": 5.847105002447218e-06,
+      "loss": 0.565,
+      "step": 12431
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 36.96591446837602,
+      "learning_rate": 5.845570962394647e-06,
+      "loss": 0.457,
+      "step": 12432
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 33.59281864738211,
+      "learning_rate": 5.8440370404916035e-06,
+      "loss": 0.5086,
+      "step": 12433
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 47.25693915139209,
+      "learning_rate": 5.842503236781722e-06,
+      "loss": 0.6417,
+      "step": 12434
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 34.851966912001814,
+      "learning_rate": 5.840969551308614e-06,
+      "loss": 0.5307,
+      "step": 12435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.39290417104048,
+      "learning_rate": 5.839435984115899e-06,
+      "loss": 0.5379,
+      "step": 12436
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 41.48739938002542,
+      "learning_rate": 5.8379025352471905e-06,
+      "loss": 0.6143,
+      "step": 12437
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 44.26705869971222,
+      "learning_rate": 5.836369204746097e-06,
+      "loss": 0.5675,
+      "step": 12438
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.418875178990845,
+      "learning_rate": 5.834835992656232e-06,
+      "loss": 0.5157,
+      "step": 12439
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 38.71161026047671,
+      "learning_rate": 5.833302899021191e-06,
+      "loss": 0.4775,
+      "step": 12440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 36.63518881417359,
+      "learning_rate": 5.831769923884579e-06,
+      "loss": 0.5358,
+      "step": 12441
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 43.265153512421165,
+      "learning_rate": 5.830237067289993e-06,
+      "loss": 0.5962,
+      "step": 12442
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.786008390482074,
+      "learning_rate": 5.828704329281024e-06,
+      "loss": 0.5514,
+      "step": 12443
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 34.34785231438109,
+      "learning_rate": 5.827171709901267e-06,
+      "loss": 0.5086,
+      "step": 12444
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 41.77734533080766,
+      "learning_rate": 5.825639209194302e-06,
+      "loss": 0.5322,
+      "step": 12445
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.08575953518007,
+      "learning_rate": 5.824106827203719e-06,
+      "loss": 0.5161,
+      "step": 12446
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.375619309887554,
+      "learning_rate": 5.822574563973091e-06,
+      "loss": 0.5247,
+      "step": 12447
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.59475249697595,
+      "learning_rate": 5.8210424195460005e-06,
+      "loss": 0.6144,
+      "step": 12448
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 45.13317055846594,
+      "learning_rate": 5.8195103939660214e-06,
+      "loss": 0.6208,
+      "step": 12449
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 43.536790789540255,
+      "learning_rate": 5.817978487276722e-06,
+      "loss": 0.5331,
+      "step": 12450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 34.65510745009712,
+      "learning_rate": 5.816446699521663e-06,
+      "loss": 0.5273,
+      "step": 12451
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 45.73724450589714,
+      "learning_rate": 5.814915030744414e-06,
+      "loss": 0.5754,
+      "step": 12452
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 35.37641201975455,
+      "learning_rate": 5.813383480988533e-06,
+      "loss": 0.5347,
+      "step": 12453
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 48.04759981430549,
+      "learning_rate": 5.811852050297579e-06,
+      "loss": 0.5457,
+      "step": 12454
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.93695944034886,
+      "learning_rate": 5.810320738715104e-06,
+      "loss": 0.5353,
+      "step": 12455
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 43.50756447889696,
+      "learning_rate": 5.808789546284652e-06,
+      "loss": 0.5698,
+      "step": 12456
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 34.15479174238154,
+      "learning_rate": 5.807258473049774e-06,
+      "loss": 0.5023,
+      "step": 12457
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 40.69425729521218,
+      "learning_rate": 5.805727519054017e-06,
+      "loss": 0.543,
+      "step": 12458
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.03360522563444,
+      "learning_rate": 5.8041966843409135e-06,
+      "loss": 0.423,
+      "step": 12459
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.161576873542515,
+      "learning_rate": 5.802665968954e-06,
+      "loss": 0.5502,
+      "step": 12460
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.08303017889041,
+      "learning_rate": 5.801135372936809e-06,
+      "loss": 0.508,
+      "step": 12461
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 38.04430219344467,
+      "learning_rate": 5.7996048963328775e-06,
+      "loss": 0.5417,
+      "step": 12462
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 40.24577671308623,
+      "learning_rate": 5.798074539185721e-06,
+      "loss": 0.5719,
+      "step": 12463
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 43.15555995138118,
+      "learning_rate": 5.79654430153887e-06,
+      "loss": 0.5188,
+      "step": 12464
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 35.727527053995736,
+      "learning_rate": 5.7950141834358365e-06,
+      "loss": 0.4367,
+      "step": 12465
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.24506203356985,
+      "learning_rate": 5.793484184920139e-06,
+      "loss": 0.4918,
+      "step": 12466
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.20867722122057,
+      "learning_rate": 5.7919543060352965e-06,
+      "loss": 0.617,
+      "step": 12467
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.13524262045568,
+      "learning_rate": 5.790424546824806e-06,
+      "loss": 0.5098,
+      "step": 12468
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 44.6681341931307,
+      "learning_rate": 5.788894907332184e-06,
+      "loss": 0.5394,
+      "step": 12469
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.56011459850677,
+      "learning_rate": 5.787365387600923e-06,
+      "loss": 0.5384,
+      "step": 12470
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.66499265290048,
+      "learning_rate": 5.78583598767453e-06,
+      "loss": 0.5673,
+      "step": 12471
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.528493711725176,
+      "learning_rate": 5.784306707596492e-06,
+      "loss": 0.5276,
+      "step": 12472
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 30.960181493430483,
+      "learning_rate": 5.782777547410305e-06,
+      "loss": 0.4636,
+      "step": 12473
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 44.33371511392961,
+      "learning_rate": 5.781248507159463e-06,
+      "loss": 0.5666,
+      "step": 12474
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.351023525929435,
+      "learning_rate": 5.7797195868874445e-06,
+      "loss": 0.5307,
+      "step": 12475
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 34.74430527782888,
+      "learning_rate": 5.778190786637729e-06,
+      "loss": 0.4409,
+      "step": 12476
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 41.57161517468887,
+      "learning_rate": 5.776662106453797e-06,
+      "loss": 0.524,
+      "step": 12477
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.15175603701092,
+      "learning_rate": 5.775133546379128e-06,
+      "loss": 0.4772,
+      "step": 12478
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.15923384056268,
+      "learning_rate": 5.773605106457185e-06,
+      "loss": 0.5397,
+      "step": 12479
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 44.473682014788416,
+      "learning_rate": 5.7720767867314464e-06,
+      "loss": 0.5619,
+      "step": 12480
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.04408679621413,
+      "learning_rate": 5.7705485872453645e-06,
+      "loss": 0.5812,
+      "step": 12481
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.05543860668807,
+      "learning_rate": 5.769020508042408e-06,
+      "loss": 0.5204,
+      "step": 12482
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 32.64782593047744,
+      "learning_rate": 5.7674925491660365e-06,
+      "loss": 0.467,
+      "step": 12483
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 45.37871162219792,
+      "learning_rate": 5.765964710659702e-06,
+      "loss": 0.5351,
+      "step": 12484
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.58384158334428,
+      "learning_rate": 5.764436992566849e-06,
+      "loss": 0.5384,
+      "step": 12485
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.95857701800037,
+      "learning_rate": 5.762909394930931e-06,
+      "loss": 0.495,
+      "step": 12486
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 45.47307255441889,
+      "learning_rate": 5.761381917795394e-06,
+      "loss": 0.5754,
+      "step": 12487
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 48.49382117328025,
+      "learning_rate": 5.759854561203677e-06,
+      "loss": 0.5723,
+      "step": 12488
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 35.3088260764957,
+      "learning_rate": 5.7583273251992115e-06,
+      "loss": 0.571,
+      "step": 12489
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 40.866778471396024,
+      "learning_rate": 5.756800209825435e-06,
+      "loss": 0.5041,
+      "step": 12490
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.36059303164775,
+      "learning_rate": 5.75527321512578e-06,
+      "loss": 0.5234,
+      "step": 12491
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 39.45003941622217,
+      "learning_rate": 5.753746341143674e-06,
+      "loss": 0.4676,
+      "step": 12492
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 39.88934050366032,
+      "learning_rate": 5.752219587922538e-06,
+      "loss": 0.5169,
+      "step": 12493
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 43.18397363213539,
+      "learning_rate": 5.75069295550579e-06,
+      "loss": 0.5058,
+      "step": 12494
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.246998558864547,
+      "learning_rate": 5.749166443936847e-06,
+      "loss": 0.4537,
+      "step": 12495
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.38841517127699,
+      "learning_rate": 5.747640053259127e-06,
+      "loss": 0.498,
+      "step": 12496
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 29.982947160002052,
+      "learning_rate": 5.746113783516034e-06,
+      "loss": 0.4602,
+      "step": 12497
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 43.02628936653179,
+      "learning_rate": 5.74458763475098e-06,
+      "loss": 0.5241,
+      "step": 12498
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.0534566981229,
+      "learning_rate": 5.743061607007359e-06,
+      "loss": 0.4968,
+      "step": 12499
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 36.785786908801555,
+      "learning_rate": 5.741535700328581e-06,
+      "loss": 0.5208,
+      "step": 12500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 38.94596994470681,
+      "learning_rate": 5.740009914758032e-06,
+      "loss": 0.4852,
+      "step": 12501
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 43.36185214278146,
+      "learning_rate": 5.738484250339109e-06,
+      "loss": 0.6251,
+      "step": 12502
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.005267963544824,
+      "learning_rate": 5.7369587071152055e-06,
+      "loss": 0.4558,
+      "step": 12503
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 39.73981460497327,
+      "learning_rate": 5.735433285129699e-06,
+      "loss": 0.5971,
+      "step": 12504
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 40.82388248584346,
+      "learning_rate": 5.733907984425979e-06,
+      "loss": 0.5381,
+      "step": 12505
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 41.63986859233983,
+      "learning_rate": 5.732382805047416e-06,
+      "loss": 0.5491,
+      "step": 12506
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 45.400060322381165,
+      "learning_rate": 5.730857747037389e-06,
+      "loss": 0.4906,
+      "step": 12507
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 43.643242292798455,
+      "learning_rate": 5.729332810439274e-06,
+      "loss": 0.5586,
+      "step": 12508
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.175976353454,
+      "learning_rate": 5.727807995296437e-06,
+      "loss": 0.5214,
+      "step": 12509
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 41.82304256542191,
+      "learning_rate": 5.7262833016522366e-06,
+      "loss": 0.5674,
+      "step": 12510
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 34.84143520586426,
+      "learning_rate": 5.72475872955004e-06,
+      "loss": 0.4572,
+      "step": 12511
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 41.428896912813606,
+      "learning_rate": 5.723234279033207e-06,
+      "loss": 0.5584,
+      "step": 12512
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 35.611653926649694,
+      "learning_rate": 5.721709950145089e-06,
+      "loss": 0.4764,
+      "step": 12513
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.67473803747441,
+      "learning_rate": 5.720185742929034e-06,
+      "loss": 0.5491,
+      "step": 12514
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 37.55774900822471,
+      "learning_rate": 5.718661657428393e-06,
+      "loss": 0.4316,
+      "step": 12515
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.65348243520357,
+      "learning_rate": 5.717137693686509e-06,
+      "loss": 0.5366,
+      "step": 12516
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 40.53307434893886,
+      "learning_rate": 5.7156138517467285e-06,
+      "loss": 0.4929,
+      "step": 12517
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 47.47013958422058,
+      "learning_rate": 5.714090131652385e-06,
+      "loss": 0.5337,
+      "step": 12518
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 44.30585566853864,
+      "learning_rate": 5.712566533446805e-06,
+      "loss": 0.5019,
+      "step": 12519
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.49842688510108,
+      "learning_rate": 5.711043057173326e-06,
+      "loss": 0.5318,
+      "step": 12520
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.98527718376874,
+      "learning_rate": 5.709519702875277e-06,
+      "loss": 0.5633,
+      "step": 12521
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.40811382392921,
+      "learning_rate": 5.707996470595977e-06,
+      "loss": 0.4585,
+      "step": 12522
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 48.434557297440634,
+      "learning_rate": 5.706473360378743e-06,
+      "loss": 0.5146,
+      "step": 12523
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.582492628037215,
+      "learning_rate": 5.704950372266895e-06,
+      "loss": 0.4755,
+      "step": 12524
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 34.37886822870365,
+      "learning_rate": 5.70342750630375e-06,
+      "loss": 0.5291,
+      "step": 12525
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 41.24507982214627,
+      "learning_rate": 5.70190476253261e-06,
+      "loss": 0.6035,
+      "step": 12526
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.06687562978012,
+      "learning_rate": 5.700382140996787e-06,
+      "loss": 0.5758,
+      "step": 12527
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.761033542548965,
+      "learning_rate": 5.698859641739578e-06,
+      "loss": 0.5204,
+      "step": 12528
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.216461162509404,
+      "learning_rate": 5.697337264804283e-06,
+      "loss": 0.5217,
+      "step": 12529
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 41.019809737575656,
+      "learning_rate": 5.695815010234204e-06,
+      "loss": 0.4839,
+      "step": 12530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.89027114038955,
+      "learning_rate": 5.694292878072625e-06,
+      "loss": 0.5453,
+      "step": 12531
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 46.13379164907479,
+      "learning_rate": 5.6927708683628415e-06,
+      "loss": 0.5995,
+      "step": 12532
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 28.909203954752897,
+      "learning_rate": 5.6912489811481295e-06,
+      "loss": 0.4293,
+      "step": 12533
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 34.55220419205436,
+      "learning_rate": 5.689727216471782e-06,
+      "loss": 0.5038,
+      "step": 12534
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 34.54188751078013,
+      "learning_rate": 5.688205574377066e-06,
+      "loss": 0.4863,
+      "step": 12535
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.65244238770691,
+      "learning_rate": 5.686684054907261e-06,
+      "loss": 0.5152,
+      "step": 12536
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 41.29064563036731,
+      "learning_rate": 5.685162658105643e-06,
+      "loss": 0.5025,
+      "step": 12537
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 40.75724789749551,
+      "learning_rate": 5.683641384015475e-06,
+      "loss": 0.5511,
+      "step": 12538
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.02680729403859,
+      "learning_rate": 5.682120232680015e-06,
+      "loss": 0.4609,
+      "step": 12539
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.20050655510959,
+      "learning_rate": 5.6805992041425315e-06,
+      "loss": 0.6017,
+      "step": 12540
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.72043098038593,
+      "learning_rate": 5.679078298446279e-06,
+      "loss": 0.4456,
+      "step": 12541
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.3298762163378,
+      "learning_rate": 5.677557515634517e-06,
+      "loss": 0.5199,
+      "step": 12542
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.02970392663672,
+      "learning_rate": 5.67603685575049e-06,
+      "loss": 0.5086,
+      "step": 12543
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.153018974559124,
+      "learning_rate": 5.674516318837442e-06,
+      "loss": 0.505,
+      "step": 12544
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 31.168151816683753,
+      "learning_rate": 5.6729959049386185e-06,
+      "loss": 0.4714,
+      "step": 12545
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 44.80447591716692,
+      "learning_rate": 5.6714756140972645e-06,
+      "loss": 0.5162,
+      "step": 12546
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.48270697228597,
+      "learning_rate": 5.669955446356612e-06,
+      "loss": 0.5612,
+      "step": 12547
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.90548411975629,
+      "learning_rate": 5.668435401759891e-06,
+      "loss": 0.4709,
+      "step": 12548
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 33.09843275171633,
+      "learning_rate": 5.666915480350332e-06,
+      "loss": 0.4946,
+      "step": 12549
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.20675232321576,
+      "learning_rate": 5.665395682171166e-06,
+      "loss": 0.5719,
+      "step": 12550
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.47637109824889,
+      "learning_rate": 5.6638760072656075e-06,
+      "loss": 0.4941,
+      "step": 12551
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.82091668684253,
+      "learning_rate": 5.662356455676882e-06,
+      "loss": 0.5046,
+      "step": 12552
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 34.06270726342095,
+      "learning_rate": 5.660837027448198e-06,
+      "loss": 0.5007,
+      "step": 12553
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 48.1345126175766,
+      "learning_rate": 5.65931772262277e-06,
+      "loss": 0.608,
+      "step": 12554
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.82476758337126,
+      "learning_rate": 5.657798541243812e-06,
+      "loss": 0.4847,
+      "step": 12555
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.3942156331285,
+      "learning_rate": 5.656279483354519e-06,
+      "loss": 0.5596,
+      "step": 12556
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 52.13725846499182,
+      "learning_rate": 5.6547605489981e-06,
+      "loss": 0.5512,
+      "step": 12557
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 34.0537274674497,
+      "learning_rate": 5.653241738217745e-06,
+      "loss": 0.5038,
+      "step": 12558
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 56.60771162857022,
+      "learning_rate": 5.651723051056657e-06,
+      "loss": 0.4981,
+      "step": 12559
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 40.45283062040588,
+      "learning_rate": 5.650204487558016e-06,
+      "loss": 0.5594,
+      "step": 12560
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.29929583215914,
+      "learning_rate": 5.648686047765017e-06,
+      "loss": 0.4986,
+      "step": 12561
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.92971004146127,
+      "learning_rate": 5.647167731720844e-06,
+      "loss": 0.5114,
+      "step": 12562
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 38.4787948533982,
+      "learning_rate": 5.645649539468675e-06,
+      "loss": 0.5053,
+      "step": 12563
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.46186916589069,
+      "learning_rate": 5.644131471051681e-06,
+      "loss": 0.5634,
+      "step": 12564
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 31.28714667275819,
+      "learning_rate": 5.642613526513041e-06,
+      "loss": 0.4879,
+      "step": 12565
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 43.83808077581665,
+      "learning_rate": 5.641095705895923e-06,
+      "loss": 0.5627,
+      "step": 12566
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 43.17894343781935,
+      "learning_rate": 5.639578009243496e-06,
+      "loss": 0.5102,
+      "step": 12567
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 48.52461574624386,
+      "learning_rate": 5.638060436598922e-06,
+      "loss": 0.5486,
+      "step": 12568
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 36.791994167193664,
+      "learning_rate": 5.636542988005351e-06,
+      "loss": 0.6064,
+      "step": 12569
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 42.408828572620905,
+      "learning_rate": 5.635025663505946e-06,
+      "loss": 0.6152,
+      "step": 12570
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.1016967557723,
+      "learning_rate": 5.633508463143862e-06,
+      "loss": 0.5666,
+      "step": 12571
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 28.936950216982222,
+      "learning_rate": 5.631991386962243e-06,
+      "loss": 0.4241,
+      "step": 12572
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.48888148554419,
+      "learning_rate": 5.6304744350042295e-06,
+      "loss": 0.4147,
+      "step": 12573
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.897785988906186,
+      "learning_rate": 5.628957607312967e-06,
+      "loss": 0.5941,
+      "step": 12574
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 40.300868010697926,
+      "learning_rate": 5.627440903931598e-06,
+      "loss": 0.5342,
+      "step": 12575
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 34.38453774348257,
+      "learning_rate": 5.62592432490325e-06,
+      "loss": 0.4836,
+      "step": 12576
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 57.55775678147149,
+      "learning_rate": 5.624407870271052e-06,
+      "loss": 0.6936,
+      "step": 12577
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.140418728648484,
+      "learning_rate": 5.622891540078135e-06,
+      "loss": 0.5422,
+      "step": 12578
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.172195229752376,
+      "learning_rate": 5.621375334367622e-06,
+      "loss": 0.5303,
+      "step": 12579
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 46.53453234796511,
+      "learning_rate": 5.619859253182638e-06,
+      "loss": 0.5383,
+      "step": 12580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 35.66989359488737,
+      "learning_rate": 5.618343296566293e-06,
+      "loss": 0.5336,
+      "step": 12581
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 37.55104647230943,
+      "learning_rate": 5.616827464561698e-06,
+      "loss": 0.5557,
+      "step": 12582
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.23501899339594,
+      "learning_rate": 5.615311757211965e-06,
+      "loss": 0.5419,
+      "step": 12583
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 37.288537412375476,
+      "learning_rate": 5.613796174560207e-06,
+      "loss": 0.5255,
+      "step": 12584
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.16272417988518,
+      "learning_rate": 5.612280716649514e-06,
+      "loss": 0.5409,
+      "step": 12585
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.32312429444864,
+      "learning_rate": 5.6107653835229954e-06,
+      "loss": 0.4397,
+      "step": 12586
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.880573823730785,
+      "learning_rate": 5.609250175223737e-06,
+      "loss": 0.5375,
+      "step": 12587
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 44.24159500342365,
+      "learning_rate": 5.607735091794839e-06,
+      "loss": 0.5936,
+      "step": 12588
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.16064809974156,
+      "learning_rate": 5.606220133279383e-06,
+      "loss": 0.5626,
+      "step": 12589
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.610038820611145,
+      "learning_rate": 5.604705299720455e-06,
+      "loss": 0.5576,
+      "step": 12590
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.0038771536102,
+      "learning_rate": 5.603190591161141e-06,
+      "loss": 0.4931,
+      "step": 12591
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.66813714097053,
+      "learning_rate": 5.601676007644511e-06,
+      "loss": 0.5387,
+      "step": 12592
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.60097197184191,
+      "learning_rate": 5.600161549213647e-06,
+      "loss": 0.537,
+      "step": 12593
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 36.01997920461983,
+      "learning_rate": 5.598647215911609e-06,
+      "loss": 0.4824,
+      "step": 12594
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.55872815127108,
+      "learning_rate": 5.59713300778147e-06,
+      "loss": 0.5762,
+      "step": 12595
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 32.556246122175814,
+      "learning_rate": 5.595618924866298e-06,
+      "loss": 0.4745,
+      "step": 12596
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 40.21765697805017,
+      "learning_rate": 5.594104967209146e-06,
+      "loss": 0.5326,
+      "step": 12597
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 40.06843804973007,
+      "learning_rate": 5.592591134853067e-06,
+      "loss": 0.5588,
+      "step": 12598
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.811180862418944,
+      "learning_rate": 5.591077427841118e-06,
+      "loss": 0.5681,
+      "step": 12599
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 42.12525212017193,
+      "learning_rate": 5.5895638462163536e-06,
+      "loss": 0.5451,
+      "step": 12600
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 52.82673978526719,
+      "learning_rate": 5.588050390021812e-06,
+      "loss": 0.6018,
+      "step": 12601
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 58.117608156154866,
+      "learning_rate": 5.586537059300532e-06,
+      "loss": 0.5058,
+      "step": 12602
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 33.18961006396381,
+      "learning_rate": 5.585023854095557e-06,
+      "loss": 0.5076,
+      "step": 12603
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 33.487387752594024,
+      "learning_rate": 5.5835107744499205e-06,
+      "loss": 0.4546,
+      "step": 12604
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 34.94133014190897,
+      "learning_rate": 5.581997820406659e-06,
+      "loss": 0.5302,
+      "step": 12605
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.84640252652429,
+      "learning_rate": 5.580484992008795e-06,
+      "loss": 0.5468,
+      "step": 12606
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 42.19693848885466,
+      "learning_rate": 5.5789722892993466e-06,
+      "loss": 0.5399,
+      "step": 12607
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 37.6315371958733,
+      "learning_rate": 5.577459712321341e-06,
+      "loss": 0.4697,
+      "step": 12608
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 37.90269113236191,
+      "learning_rate": 5.575947261117798e-06,
+      "loss": 0.5241,
+      "step": 12609
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 40.91712249973283,
+      "learning_rate": 5.574434935731723e-06,
+      "loss": 0.4843,
+      "step": 12610
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.24919308328637,
+      "learning_rate": 5.572922736206135e-06,
+      "loss": 0.5905,
+      "step": 12611
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 33.42506395617738,
+      "learning_rate": 5.571410662584029e-06,
+      "loss": 0.4799,
+      "step": 12612
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 35.82644310658117,
+      "learning_rate": 5.5698987149084174e-06,
+      "loss": 0.5087,
+      "step": 12613
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.779617924462904,
+      "learning_rate": 5.568386893222291e-06,
+      "loss": 0.4918,
+      "step": 12614
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 34.318383725139064,
+      "learning_rate": 5.5668751975686485e-06,
+      "loss": 0.4967,
+      "step": 12615
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 34.67465191620346,
+      "learning_rate": 5.565363627990485e-06,
+      "loss": 0.4995,
+      "step": 12616
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 34.82702817768551,
+      "learning_rate": 5.563852184530784e-06,
+      "loss": 0.4929,
+      "step": 12617
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 41.849599564709116,
+      "learning_rate": 5.562340867232535e-06,
+      "loss": 0.4781,
+      "step": 12618
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 55.50792740575721,
+      "learning_rate": 5.560829676138712e-06,
+      "loss": 0.6399,
+      "step": 12619
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 41.24181701528436,
+      "learning_rate": 5.559318611292299e-06,
+      "loss": 0.5211,
+      "step": 12620
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.748330450867524,
+      "learning_rate": 5.557807672736264e-06,
+      "loss": 0.4302,
+      "step": 12621
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.759633932700815,
+      "learning_rate": 5.556296860513584e-06,
+      "loss": 0.5544,
+      "step": 12622
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.50751374357687,
+      "learning_rate": 5.554786174667217e-06,
+      "loss": 0.4937,
+      "step": 12623
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.705250273423154,
+      "learning_rate": 5.553275615240132e-06,
+      "loss": 0.512,
+      "step": 12624
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 40.66470732457058,
+      "learning_rate": 5.551765182275292e-06,
+      "loss": 0.4671,
+      "step": 12625
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.258569620147846,
+      "learning_rate": 5.550254875815646e-06,
+      "loss": 0.5615,
+      "step": 12626
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.438972560516596,
+      "learning_rate": 5.548744695904145e-06,
+      "loss": 0.511,
+      "step": 12627
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 44.394878142076095,
+      "learning_rate": 5.5472346425837405e-06,
+      "loss": 0.6167,
+      "step": 12628
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 48.663075952578076,
+      "learning_rate": 5.545724715897381e-06,
+      "loss": 0.6158,
+      "step": 12629
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 44.612031168784256,
+      "learning_rate": 5.544214915888006e-06,
+      "loss": 0.6433,
+      "step": 12630
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.720407590645674,
+      "learning_rate": 5.542705242598552e-06,
+      "loss": 0.4429,
+      "step": 12631
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.24437744329393,
+      "learning_rate": 5.54119569607195e-06,
+      "loss": 0.5712,
+      "step": 12632
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.42266335130407,
+      "learning_rate": 5.539686276351135e-06,
+      "loss": 0.521,
+      "step": 12633
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.089758250673924,
+      "learning_rate": 5.538176983479036e-06,
+      "loss": 0.4478,
+      "step": 12634
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 44.143679347407726,
+      "learning_rate": 5.536667817498573e-06,
+      "loss": 0.4935,
+      "step": 12635
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.65156847688472,
+      "learning_rate": 5.535158778452664e-06,
+      "loss": 0.4835,
+      "step": 12636
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 41.575867509536494,
+      "learning_rate": 5.533649866384226e-06,
+      "loss": 0.546,
+      "step": 12637
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 41.62958755071676,
+      "learning_rate": 5.532141081336177e-06,
+      "loss": 0.5834,
+      "step": 12638
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.259148335787906,
+      "learning_rate": 5.530632423351421e-06,
+      "loss": 0.5007,
+      "step": 12639
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 37.36402646172573,
+      "learning_rate": 5.5291238924728654e-06,
+      "loss": 0.495,
+      "step": 12640
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.149559321676854,
+      "learning_rate": 5.5276154887434075e-06,
+      "loss": 0.5087,
+      "step": 12641
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 42.5314470464555,
+      "learning_rate": 5.526107212205949e-06,
+      "loss": 0.4789,
+      "step": 12642
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 48.11445185011628,
+      "learning_rate": 5.524599062903388e-06,
+      "loss": 0.5815,
+      "step": 12643
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.25296897481836,
+      "learning_rate": 5.523091040878608e-06,
+      "loss": 0.5635,
+      "step": 12644
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 43.41713706870364,
+      "learning_rate": 5.521583146174503e-06,
+      "loss": 0.5903,
+      "step": 12645
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 44.67900160162793,
+      "learning_rate": 5.5200753788339515e-06,
+      "loss": 0.5446,
+      "step": 12646
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 44.31267606218708,
+      "learning_rate": 5.518567738899838e-06,
+      "loss": 0.5504,
+      "step": 12647
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 51.536896363321716,
+      "learning_rate": 5.517060226415032e-06,
+      "loss": 0.6335,
+      "step": 12648
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 47.07340197289438,
+      "learning_rate": 5.515552841422412e-06,
+      "loss": 0.5646,
+      "step": 12649
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 33.862831097769345,
+      "learning_rate": 5.514045583964848e-06,
+      "loss": 0.4346,
+      "step": 12650
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.45438473947016,
+      "learning_rate": 5.512538454085206e-06,
+      "loss": 0.5008,
+      "step": 12651
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 41.88834661472297,
+      "learning_rate": 5.51103145182634e-06,
+      "loss": 0.5264,
+      "step": 12652
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.29748669066726,
+      "learning_rate": 5.509524577231114e-06,
+      "loss": 0.5443,
+      "step": 12653
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 45.96322300642893,
+      "learning_rate": 5.50801783034238e-06,
+      "loss": 0.5297,
+      "step": 12654
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 36.5038503222907,
+      "learning_rate": 5.506511211202997e-06,
+      "loss": 0.5042,
+      "step": 12655
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 36.54669043929398,
+      "learning_rate": 5.505004719855806e-06,
+      "loss": 0.5587,
+      "step": 12656
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.318186797970284,
+      "learning_rate": 5.503498356343648e-06,
+      "loss": 0.4947,
+      "step": 12657
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 38.25643613059875,
+      "learning_rate": 5.501992120709367e-06,
+      "loss": 0.5552,
+      "step": 12658
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 44.2988519567181,
+      "learning_rate": 5.500486012995801e-06,
+      "loss": 0.5251,
+      "step": 12659
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.186732760947,
+      "learning_rate": 5.498980033245782e-06,
+      "loss": 0.509,
+      "step": 12660
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 37.75494765391085,
+      "learning_rate": 5.4974741815021336e-06,
+      "loss": 0.4735,
+      "step": 12661
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 33.23274376887339,
+      "learning_rate": 5.4959684578076855e-06,
+      "loss": 0.4353,
+      "step": 12662
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 49.2882481542658,
+      "learning_rate": 5.494462862205263e-06,
+      "loss": 0.6561,
+      "step": 12663
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 37.0136037285138,
+      "learning_rate": 5.492957394737677e-06,
+      "loss": 0.5743,
+      "step": 12664
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 36.41095827297936,
+      "learning_rate": 5.49145205544775e-06,
+      "loss": 0.4983,
+      "step": 12665
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.94279433843997,
+      "learning_rate": 5.4899468443782864e-06,
+      "loss": 0.517,
+      "step": 12666
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 34.97098832044766,
+      "learning_rate": 5.488441761572096e-06,
+      "loss": 0.4605,
+      "step": 12667
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 42.58068758191236,
+      "learning_rate": 5.486936807071986e-06,
+      "loss": 0.5067,
+      "step": 12668
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.49698900263219,
+      "learning_rate": 5.485431980920753e-06,
+      "loss": 0.5461,
+      "step": 12669
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 37.627642709812335,
+      "learning_rate": 5.4839272831611905e-06,
+      "loss": 0.5738,
+      "step": 12670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 38.179516536859886,
+      "learning_rate": 5.482422713836094e-06,
+      "loss": 0.5326,
+      "step": 12671
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 44.169435491083114,
+      "learning_rate": 5.480918272988256e-06,
+      "loss": 0.533,
+      "step": 12672
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 43.53807535221221,
+      "learning_rate": 5.4794139606604556e-06,
+      "loss": 0.5765,
+      "step": 12673
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.54396841108857,
+      "learning_rate": 5.477909776895481e-06,
+      "loss": 0.5064,
+      "step": 12674
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 37.99924362207842,
+      "learning_rate": 5.476405721736104e-06,
+      "loss": 0.5735,
+      "step": 12675
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 42.150008795070974,
+      "learning_rate": 5.474901795225105e-06,
+      "loss": 0.5928,
+      "step": 12676
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.20118246058867,
+      "learning_rate": 5.473397997405249e-06,
+      "loss": 0.5861,
+      "step": 12677
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.82694278002528,
+      "learning_rate": 5.471894328319305e-06,
+      "loss": 0.5584,
+      "step": 12678
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 48.40622836778032,
+      "learning_rate": 5.470390788010042e-06,
+      "loss": 0.6291,
+      "step": 12679
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 42.61323537150054,
+      "learning_rate": 5.4688873765202114e-06,
+      "loss": 0.6076,
+      "step": 12680
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 36.33919073422499,
+      "learning_rate": 5.467384093892576e-06,
+      "loss": 0.6165,
+      "step": 12681
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 36.163765959431416,
+      "learning_rate": 5.465880940169881e-06,
+      "loss": 0.5121,
+      "step": 12682
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.36812956073258,
+      "learning_rate": 5.464377915394882e-06,
+      "loss": 0.529,
+      "step": 12683
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 43.200673858477636,
+      "learning_rate": 5.4628750196103245e-06,
+      "loss": 0.4984,
+      "step": 12684
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 42.867248857696545,
+      "learning_rate": 5.461372252858948e-06,
+      "loss": 0.5576,
+      "step": 12685
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 42.40244748433628,
+      "learning_rate": 5.459869615183484e-06,
+      "loss": 0.467,
+      "step": 12686
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 35.087682557862685,
+      "learning_rate": 5.458367106626674e-06,
+      "loss": 0.4961,
+      "step": 12687
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.351733136332506,
+      "learning_rate": 5.456864727231246e-06,
+      "loss": 0.5133,
+      "step": 12688
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 46.0354698976458,
+      "learning_rate": 5.455362477039935e-06,
+      "loss": 0.5375,
+      "step": 12689
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 35.97918931481392,
+      "learning_rate": 5.453860356095448e-06,
+      "loss": 0.4722,
+      "step": 12690
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 38.65303770023529,
+      "learning_rate": 5.452358364440515e-06,
+      "loss": 0.5037,
+      "step": 12691
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.166283884968166,
+      "learning_rate": 5.450856502117849e-06,
+      "loss": 0.5204,
+      "step": 12692
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 33.49147736615395,
+      "learning_rate": 5.449354769170168e-06,
+      "loss": 0.5018,
+      "step": 12693
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 37.660352415379734,
+      "learning_rate": 5.447853165640173e-06,
+      "loss": 0.4917,
+      "step": 12694
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 35.4656836037047,
+      "learning_rate": 5.4463516915705684e-06,
+      "loss": 0.4943,
+      "step": 12695
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 30.51816314673632,
+      "learning_rate": 5.444850347004058e-06,
+      "loss": 0.4554,
+      "step": 12696
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 33.58927341531068,
+      "learning_rate": 5.443349131983343e-06,
+      "loss": 0.4828,
+      "step": 12697
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 49.781017597467795,
+      "learning_rate": 5.441848046551108e-06,
+      "loss": 0.6264,
+      "step": 12698
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 36.63791756784064,
+      "learning_rate": 5.440347090750053e-06,
+      "loss": 0.5296,
+      "step": 12699
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 41.50735910499379,
+      "learning_rate": 5.438846264622857e-06,
+      "loss": 0.5253,
+      "step": 12700
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 39.5388309371731,
+      "learning_rate": 5.437345568212207e-06,
+      "loss": 0.4632,
+      "step": 12701
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 34.897798667715655,
+      "learning_rate": 5.435845001560775e-06,
+      "loss": 0.4959,
+      "step": 12702
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 33.985159604891955,
+      "learning_rate": 5.434344564711244e-06,
+      "loss": 0.504,
+      "step": 12703
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 42.397494188748404,
+      "learning_rate": 5.432844257706285e-06,
+      "loss": 0.5316,
+      "step": 12704
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 40.59450895320912,
+      "learning_rate": 5.431344080588561e-06,
+      "loss": 0.5395,
+      "step": 12705
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 43.025195283795696,
+      "learning_rate": 5.4298440334007415e-06,
+      "loss": 0.5787,
+      "step": 12706
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 43.46718915746461,
+      "learning_rate": 5.42834411618548e-06,
+      "loss": 0.5402,
+      "step": 12707
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 47.47730332261573,
+      "learning_rate": 5.426844328985439e-06,
+      "loss": 0.5462,
+      "step": 12708
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 39.97734157033481,
+      "learning_rate": 5.4253446718432724e-06,
+      "loss": 0.503,
+      "step": 12709
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.07360048977303,
+      "learning_rate": 5.423845144801627e-06,
+      "loss": 0.5324,
+      "step": 12710
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 38.39198910653901,
+      "learning_rate": 5.422345747903146e-06,
+      "loss": 0.4954,
+      "step": 12711
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.02088492795447,
+      "learning_rate": 5.420846481190473e-06,
+      "loss": 0.4638,
+      "step": 12712
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 56.74014425190125,
+      "learning_rate": 5.41934734470625e-06,
+      "loss": 0.5788,
+      "step": 12713
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.1231165706799,
+      "learning_rate": 5.417848338493114e-06,
+      "loss": 0.4981,
+      "step": 12714
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.68725057783187,
+      "learning_rate": 5.416349462593684e-06,
+      "loss": 0.5169,
+      "step": 12715
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.724318295586,
+      "learning_rate": 5.414850717050593e-06,
+      "loss": 0.478,
+      "step": 12716
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.045805768296404,
+      "learning_rate": 5.413352101906466e-06,
+      "loss": 0.4598,
+      "step": 12717
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 44.565174755488655,
+      "learning_rate": 5.411853617203926e-06,
+      "loss": 0.5904,
+      "step": 12718
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.2509242301718,
+      "learning_rate": 5.410355262985585e-06,
+      "loss": 0.5011,
+      "step": 12719
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.18611432586613,
+      "learning_rate": 5.408857039294052e-06,
+      "loss": 0.5256,
+      "step": 12720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 49.4622434673545,
+      "learning_rate": 5.407358946171939e-06,
+      "loss": 0.6245,
+      "step": 12721
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 39.887594772101956,
+      "learning_rate": 5.405860983661854e-06,
+      "loss": 0.5071,
+      "step": 12722
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 32.950523776636594,
+      "learning_rate": 5.404363151806397e-06,
+      "loss": 0.5296,
+      "step": 12723
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 44.47724664491268,
+      "learning_rate": 5.402865450648158e-06,
+      "loss": 0.6382,
+      "step": 12724
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.13898846363201,
+      "learning_rate": 5.401367880229737e-06,
+      "loss": 0.473,
+      "step": 12725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 38.608210114984296,
+      "learning_rate": 5.399870440593726e-06,
+      "loss": 0.4831,
+      "step": 12726
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 28.72038787388612,
+      "learning_rate": 5.3983731317827075e-06,
+      "loss": 0.4057,
+      "step": 12727
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.287542546498955,
+      "learning_rate": 5.396875953839267e-06,
+      "loss": 0.4985,
+      "step": 12728
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 40.59779852808519,
+      "learning_rate": 5.3953789068059785e-06,
+      "loss": 0.4886,
+      "step": 12729
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 35.93894624536626,
+      "learning_rate": 5.3938819907254204e-06,
+      "loss": 0.5126,
+      "step": 12730
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 34.45190068705719,
+      "learning_rate": 5.392385205640167e-06,
+      "loss": 0.4538,
+      "step": 12731
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 47.14399581252384,
+      "learning_rate": 5.390888551592779e-06,
+      "loss": 0.5914,
+      "step": 12732
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.2634738254812,
+      "learning_rate": 5.3893920286258285e-06,
+      "loss": 0.6216,
+      "step": 12733
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 35.986229912094615,
+      "learning_rate": 5.387895636781868e-06,
+      "loss": 0.4807,
+      "step": 12734
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.838940430261886,
+      "learning_rate": 5.386399376103462e-06,
+      "loss": 0.5244,
+      "step": 12735
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 33.47074219217851,
+      "learning_rate": 5.3849032466331555e-06,
+      "loss": 0.4572,
+      "step": 12736
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.31476447168447,
+      "learning_rate": 5.3834072484134995e-06,
+      "loss": 0.4874,
+      "step": 12737
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.59755636062257,
+      "learning_rate": 5.381911381487044e-06,
+      "loss": 0.4362,
+      "step": 12738
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 53.72248837975879,
+      "learning_rate": 5.380415645896329e-06,
+      "loss": 0.5171,
+      "step": 12739
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 45.90986858983096,
+      "learning_rate": 5.378920041683886e-06,
+      "loss": 0.4733,
+      "step": 12740
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 43.92570793176268,
+      "learning_rate": 5.377424568892253e-06,
+      "loss": 0.5333,
+      "step": 12741
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.82212972442989,
+      "learning_rate": 5.375929227563963e-06,
+      "loss": 0.5013,
+      "step": 12742
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.6914810249755,
+      "learning_rate": 5.374434017741543e-06,
+      "loss": 0.573,
+      "step": 12743
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 38.099674746734536,
+      "learning_rate": 5.372938939467514e-06,
+      "loss": 0.5363,
+      "step": 12744
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.32448147900179,
+      "learning_rate": 5.371443992784391e-06,
+      "loss": 0.5428,
+      "step": 12745
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 38.09950107788458,
+      "learning_rate": 5.3699491777346935e-06,
+      "loss": 0.4849,
+      "step": 12746
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 34.47492911391685,
+      "learning_rate": 5.3684544943609375e-06,
+      "loss": 0.4449,
+      "step": 12747
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 51.17417892744059,
+      "learning_rate": 5.366959942705625e-06,
+      "loss": 0.5648,
+      "step": 12748
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.92781246904899,
+      "learning_rate": 5.3654655228112574e-06,
+      "loss": 0.5111,
+      "step": 12749
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 48.086413626401495,
+      "learning_rate": 5.36397123472034e-06,
+      "loss": 0.5665,
+      "step": 12750
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 40.186515831650546,
+      "learning_rate": 5.362477078475372e-06,
+      "loss": 0.5759,
+      "step": 12751
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 34.20242010847963,
+      "learning_rate": 5.36098305411884e-06,
+      "loss": 0.4957,
+      "step": 12752
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.693770840270766,
+      "learning_rate": 5.35948916169324e-06,
+      "loss": 0.5324,
+      "step": 12753
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 39.494269359739704,
+      "learning_rate": 5.357995401241049e-06,
+      "loss": 0.4892,
+      "step": 12754
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 35.92706441542357,
+      "learning_rate": 5.3565017728047545e-06,
+      "loss": 0.4366,
+      "step": 12755
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 38.0833848491017,
+      "learning_rate": 5.355008276426836e-06,
+      "loss": 0.4896,
+      "step": 12756
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 48.83952921998199,
+      "learning_rate": 5.353514912149761e-06,
+      "loss": 0.5412,
+      "step": 12757
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.224088136617624,
+      "learning_rate": 5.3520216800160085e-06,
+      "loss": 0.5395,
+      "step": 12758
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 49.432812517941876,
+      "learning_rate": 5.350528580068035e-06,
+      "loss": 0.4637,
+      "step": 12759
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.247532231558694,
+      "learning_rate": 5.349035612348314e-06,
+      "loss": 0.531,
+      "step": 12760
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.57508061819762,
+      "learning_rate": 5.347542776899295e-06,
+      "loss": 0.5054,
+      "step": 12761
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 47.855713373259114,
+      "learning_rate": 5.346050073763438e-06,
+      "loss": 0.5016,
+      "step": 12762
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.57729828767769,
+      "learning_rate": 5.344557502983198e-06,
+      "loss": 0.5262,
+      "step": 12763
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.76486950899954,
+      "learning_rate": 5.343065064601021e-06,
+      "loss": 0.5389,
+      "step": 12764
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 32.320366600974296,
+      "learning_rate": 5.3415727586593435e-06,
+      "loss": 0.4261,
+      "step": 12765
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 39.65808346525287,
+      "learning_rate": 5.340080585200612e-06,
+      "loss": 0.5532,
+      "step": 12766
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.071578927515844,
+      "learning_rate": 5.338588544267267e-06,
+      "loss": 0.5009,
+      "step": 12767
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 49.4767231206005,
+      "learning_rate": 5.3370966359017325e-06,
+      "loss": 0.5247,
+      "step": 12768
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 31.94947021773458,
+      "learning_rate": 5.335604860146446e-06,
+      "loss": 0.4521,
+      "step": 12769
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.86112657437162,
+      "learning_rate": 5.334113217043826e-06,
+      "loss": 0.5236,
+      "step": 12770
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 43.846300329640655,
+      "learning_rate": 5.332621706636295e-06,
+      "loss": 0.5682,
+      "step": 12771
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 39.27156689242898,
+      "learning_rate": 5.331130328966276e-06,
+      "loss": 0.5323,
+      "step": 12772
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 33.75698485680811,
+      "learning_rate": 5.329639084076181e-06,
+      "loss": 0.4731,
+      "step": 12773
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 36.75536900181101,
+      "learning_rate": 5.3281479720084125e-06,
+      "loss": 0.4489,
+      "step": 12774
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 39.43892087054436,
+      "learning_rate": 5.326656992805384e-06,
+      "loss": 0.5125,
+      "step": 12775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 33.4300787417631,
+      "learning_rate": 5.325166146509497e-06,
+      "loss": 0.4932,
+      "step": 12776
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 54.24738818794526,
+      "learning_rate": 5.323675433163158e-06,
+      "loss": 0.5585,
+      "step": 12777
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 40.08323764128706,
+      "learning_rate": 5.322184852808745e-06,
+      "loss": 0.5792,
+      "step": 12778
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 42.9177144536871,
+      "learning_rate": 5.32069440548866e-06,
+      "loss": 0.5017,
+      "step": 12779
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 40.393419360189206,
+      "learning_rate": 5.319204091245287e-06,
+      "loss": 0.5025,
+      "step": 12780
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 48.1020213023627,
+      "learning_rate": 5.317713910121016e-06,
+      "loss": 0.5092,
+      "step": 12781
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 36.09341575831103,
+      "learning_rate": 5.316223862158223e-06,
+      "loss": 0.4739,
+      "step": 12782
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 47.36965668169544,
+      "learning_rate": 5.31473394739928e-06,
+      "loss": 0.6193,
+      "step": 12783
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 33.793591118014525,
+      "learning_rate": 5.313244165886562e-06,
+      "loss": 0.4739,
+      "step": 12784
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 43.850610795763835,
+      "learning_rate": 5.311754517662442e-06,
+      "loss": 0.5129,
+      "step": 12785
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 56.62915453290984,
+      "learning_rate": 5.310265002769277e-06,
+      "loss": 0.4434,
+      "step": 12786
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 43.14179340066631,
+      "learning_rate": 5.308775621249435e-06,
+      "loss": 0.5511,
+      "step": 12787
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 37.42969434431949,
+      "learning_rate": 5.3072863731452674e-06,
+      "loss": 0.5203,
+      "step": 12788
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 34.06097692531703,
+      "learning_rate": 5.305797258499134e-06,
+      "loss": 0.5191,
+      "step": 12789
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 48.5283093565574,
+      "learning_rate": 5.304308277353376e-06,
+      "loss": 0.4873,
+      "step": 12790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 42.2785894201067,
+      "learning_rate": 5.3028194297503445e-06,
+      "loss": 0.5578,
+      "step": 12791
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 35.74424217113397,
+      "learning_rate": 5.301330715732385e-06,
+      "loss": 0.5212,
+      "step": 12792
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 40.037923335547674,
+      "learning_rate": 5.299842135341825e-06,
+      "loss": 0.5064,
+      "step": 12793
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 49.28323294956501,
+      "learning_rate": 5.29835368862101e-06,
+      "loss": 0.6172,
+      "step": 12794
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 46.430773390335254,
+      "learning_rate": 5.296865375612264e-06,
+      "loss": 0.5444,
+      "step": 12795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 36.40433835753226,
+      "learning_rate": 5.295377196357914e-06,
+      "loss": 0.5202,
+      "step": 12796
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 33.82135970410333,
+      "learning_rate": 5.293889150900287e-06,
+      "loss": 0.5083,
+      "step": 12797
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 38.03510667653304,
+      "learning_rate": 5.2924012392817014e-06,
+      "loss": 0.4861,
+      "step": 12798
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 42.056480549489216,
+      "learning_rate": 5.290913461544466e-06,
+      "loss": 0.5787,
+      "step": 12799
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 38.62394204280269,
+      "learning_rate": 5.289425817730897e-06,
+      "loss": 0.4775,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 34.510876951977586,
+      "learning_rate": 5.287938307883302e-06,
+      "loss": 0.5232,
+      "step": 12801
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 35.92191287151131,
+      "learning_rate": 5.286450932043994e-06,
+      "loss": 0.4713,
+      "step": 12802
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 31.105796148274692,
+      "learning_rate": 5.284963690255254e-06,
+      "loss": 0.4394,
+      "step": 12803
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 39.63858312656554,
+      "learning_rate": 5.28347658255939e-06,
+      "loss": 0.5343,
+      "step": 12804
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 47.007753209537434,
+      "learning_rate": 5.281989608998693e-06,
+      "loss": 0.5301,
+      "step": 12805
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 35.76627256108015,
+      "learning_rate": 5.280502769615456e-06,
+      "loss": 0.4529,
+      "step": 12806
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 40.05848795162163,
+      "learning_rate": 5.279016064451959e-06,
+      "loss": 0.5607,
+      "step": 12807
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 49.66130125210635,
+      "learning_rate": 5.277529493550478e-06,
+      "loss": 0.5325,
+      "step": 12808
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 41.04594573387396,
+      "learning_rate": 5.276043056953297e-06,
+      "loss": 0.5128,
+      "step": 12809
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 46.06688775395026,
+      "learning_rate": 5.274556754702691e-06,
+      "loss": 0.6061,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 37.974299630442665,
+      "learning_rate": 5.273070586840925e-06,
+      "loss": 0.4911,
+      "step": 12811
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 34.44218770849917,
+      "learning_rate": 5.271584553410267e-06,
+      "loss": 0.5149,
+      "step": 12812
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 44.260522916646394,
+      "learning_rate": 5.270098654452977e-06,
+      "loss": 0.5166,
+      "step": 12813
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 45.052724134999934,
+      "learning_rate": 5.268612890011318e-06,
+      "loss": 0.5077,
+      "step": 12814
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 39.278832630560636,
+      "learning_rate": 5.267127260127536e-06,
+      "loss": 0.4877,
+      "step": 12815
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 31.78368527769824,
+      "learning_rate": 5.26564176484389e-06,
+      "loss": 0.4527,
+      "step": 12816
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 33.81029593922359,
+      "learning_rate": 5.264156404202618e-06,
+      "loss": 0.4348,
+      "step": 12817
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 38.99391800577809,
+      "learning_rate": 5.262671178245968e-06,
+      "loss": 0.5123,
+      "step": 12818
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 42.8676700053896,
+      "learning_rate": 5.261186087016183e-06,
+      "loss": 0.4888,
+      "step": 12819
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 52.064732516301454,
+      "learning_rate": 5.259701130555489e-06,
+      "loss": 0.5718,
+      "step": 12820
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 41.755497107978925,
+      "learning_rate": 5.2582163089061255e-06,
+      "loss": 0.4714,
+      "step": 12821
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 43.05508653029789,
+      "learning_rate": 5.256731622110314e-06,
+      "loss": 0.5102,
+      "step": 12822
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 52.0850701289196,
+      "learning_rate": 5.2552470702102835e-06,
+      "loss": 0.5122,
+      "step": 12823
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 42.664892339817584,
+      "learning_rate": 5.253762653248245e-06,
+      "loss": 0.5008,
+      "step": 12824
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 38.8590712403135,
+      "learning_rate": 5.252278371266422e-06,
+      "loss": 0.5279,
+      "step": 12825
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 32.62571709674226,
+      "learning_rate": 5.250794224307028e-06,
+      "loss": 0.4177,
+      "step": 12826
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 36.51719540706956,
+      "learning_rate": 5.249310212412269e-06,
+      "loss": 0.4518,
+      "step": 12827
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 46.34377777037043,
+      "learning_rate": 5.247826335624344e-06,
+      "loss": 0.5081,
+      "step": 12828
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 44.96415718806475,
+      "learning_rate": 5.246342593985458e-06,
+      "loss": 0.6117,
+      "step": 12829
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 43.57896378764111,
+      "learning_rate": 5.244858987537808e-06,
+      "loss": 0.5357,
+      "step": 12830
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 43.64562086115842,
+      "learning_rate": 5.243375516323591e-06,
+      "loss": 0.5304,
+      "step": 12831
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 46.30724290464714,
+      "learning_rate": 5.241892180384991e-06,
+      "loss": 0.5954,
+      "step": 12832
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 35.36400087164365,
+      "learning_rate": 5.240408979764191e-06,
+      "loss": 0.5052,
+      "step": 12833
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 45.6159655365996,
+      "learning_rate": 5.2389259145033744e-06,
+      "loss": 0.5327,
+      "step": 12834
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 38.68144634749875,
+      "learning_rate": 5.237442984644724e-06,
+      "loss": 0.4598,
+      "step": 12835
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 40.733571505916835,
+      "learning_rate": 5.235960190230409e-06,
+      "loss": 0.5303,
+      "step": 12836
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.324453737327524,
+      "learning_rate": 5.234477531302595e-06,
+      "loss": 0.5189,
+      "step": 12837
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 34.47628811732029,
+      "learning_rate": 5.232995007903453e-06,
+      "loss": 0.4893,
+      "step": 12838
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.95986884359625,
+      "learning_rate": 5.231512620075143e-06,
+      "loss": 0.4831,
+      "step": 12839
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.92963388013158,
+      "learning_rate": 5.23003036785983e-06,
+      "loss": 0.5744,
+      "step": 12840
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 47.46568596338904,
+      "learning_rate": 5.228548251299663e-06,
+      "loss": 0.6002,
+      "step": 12841
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 37.695775197406185,
+      "learning_rate": 5.227066270436788e-06,
+      "loss": 0.5014,
+      "step": 12842
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 38.24880864588083,
+      "learning_rate": 5.225584425313357e-06,
+      "loss": 0.554,
+      "step": 12843
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 37.67742894170334,
+      "learning_rate": 5.224102715971515e-06,
+      "loss": 0.5436,
+      "step": 12844
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 31.593050321685308,
+      "learning_rate": 5.222621142453394e-06,
+      "loss": 0.4445,
+      "step": 12845
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 50.66773648084164,
+      "learning_rate": 5.221139704801137e-06,
+      "loss": 0.6251,
+      "step": 12846
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 42.40539515342785,
+      "learning_rate": 5.219658403056867e-06,
+      "loss": 0.4584,
+      "step": 12847
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 34.64897174272808,
+      "learning_rate": 5.218177237262721e-06,
+      "loss": 0.4254,
+      "step": 12848
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 55.72323158199799,
+      "learning_rate": 5.216696207460812e-06,
+      "loss": 0.5115,
+      "step": 12849
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 41.21780303029241,
+      "learning_rate": 5.215215313693265e-06,
+      "loss": 0.5222,
+      "step": 12850
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 42.89929758577244,
+      "learning_rate": 5.213734556002199e-06,
+      "loss": 0.4625,
+      "step": 12851
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 41.74585526845239,
+      "learning_rate": 5.212253934429723e-06,
+      "loss": 0.599,
+      "step": 12852
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 38.154470803709174,
+      "learning_rate": 5.210773449017939e-06,
+      "loss": 0.4424,
+      "step": 12853
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 30.70155265218144,
+      "learning_rate": 5.209293099808959e-06,
+      "loss": 0.4631,
+      "step": 12854
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 53.54102902855286,
+      "learning_rate": 5.207812886844879e-06,
+      "loss": 0.5592,
+      "step": 12855
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.467826084504175,
+      "learning_rate": 5.2063328101678004e-06,
+      "loss": 0.5291,
+      "step": 12856
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 42.895366855448366,
+      "learning_rate": 5.204852869819814e-06,
+      "loss": 0.6009,
+      "step": 12857
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 39.36263050579824,
+      "learning_rate": 5.203373065843003e-06,
+      "loss": 0.4712,
+      "step": 12858
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 51.36383969381018,
+      "learning_rate": 5.201893398279454e-06,
+      "loss": 0.5315,
+      "step": 12859
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 39.78346374537222,
+      "learning_rate": 5.2004138671712555e-06,
+      "loss": 0.5278,
+      "step": 12860
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 48.45594126941151,
+      "learning_rate": 5.198934472560479e-06,
+      "loss": 0.5428,
+      "step": 12861
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 52.69900145064236,
+      "learning_rate": 5.197455214489193e-06,
+      "loss": 0.5425,
+      "step": 12862
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 41.4427301089511,
+      "learning_rate": 5.195976092999472e-06,
+      "loss": 0.5148,
+      "step": 12863
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 52.64559506341141,
+      "learning_rate": 5.194497108133381e-06,
+      "loss": 0.6827,
+      "step": 12864
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 43.39599772822166,
+      "learning_rate": 5.193018259932987e-06,
+      "loss": 0.5459,
+      "step": 12865
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 38.7742341278901,
+      "learning_rate": 5.191539548440336e-06,
+      "loss": 0.5016,
+      "step": 12866
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 46.70346873754491,
+      "learning_rate": 5.190060973697486e-06,
+      "loss": 0.5565,
+      "step": 12867
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 33.55490686702409,
+      "learning_rate": 5.188582535746489e-06,
+      "loss": 0.4988,
+      "step": 12868
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 41.737255069897124,
+      "learning_rate": 5.187104234629394e-06,
+      "loss": 0.5759,
+      "step": 12869
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.45432781849165,
+      "learning_rate": 5.185626070388239e-06,
+      "loss": 0.5274,
+      "step": 12870
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.92177962517448,
+      "learning_rate": 5.184148043065058e-06,
+      "loss": 0.456,
+      "step": 12871
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.23288357236787,
+      "learning_rate": 5.182670152701889e-06,
+      "loss": 0.4552,
+      "step": 12872
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.41382593398479,
+      "learning_rate": 5.181192399340768e-06,
+      "loss": 0.4745,
+      "step": 12873
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 38.266672577231695,
+      "learning_rate": 5.179714783023711e-06,
+      "loss": 0.475,
+      "step": 12874
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 44.31497853695766,
+      "learning_rate": 5.17823730379275e-06,
+      "loss": 0.5275,
+      "step": 12875
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 49.36222196931995,
+      "learning_rate": 5.1767599616898965e-06,
+      "loss": 0.4727,
+      "step": 12876
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 39.80074919889243,
+      "learning_rate": 5.175282756757172e-06,
+      "loss": 0.54,
+      "step": 12877
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 75.7902079968405,
+      "learning_rate": 5.1738056890365775e-06,
+      "loss": 0.5855,
+      "step": 12878
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 37.59720069985787,
+      "learning_rate": 5.1723287585701285e-06,
+      "loss": 0.5262,
+      "step": 12879
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 41.22356624388695,
+      "learning_rate": 5.1708519653998275e-06,
+      "loss": 0.4747,
+      "step": 12880
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 32.76779217845377,
+      "learning_rate": 5.169375309567669e-06,
+      "loss": 0.4477,
+      "step": 12881
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 43.14183063749784,
+      "learning_rate": 5.167898791115654e-06,
+      "loss": 0.5389,
+      "step": 12882
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 42.209366755937666,
+      "learning_rate": 5.166422410085767e-06,
+      "loss": 0.4914,
+      "step": 12883
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 48.2380222811255,
+      "learning_rate": 5.16494616652e-06,
+      "loss": 0.5611,
+      "step": 12884
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 33.28186666069579,
+      "learning_rate": 5.163470060460338e-06,
+      "loss": 0.4737,
+      "step": 12885
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.602418542253716,
+      "learning_rate": 5.161994091948759e-06,
+      "loss": 0.4459,
+      "step": 12886
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 39.007288105100706,
+      "learning_rate": 5.160518261027234e-06,
+      "loss": 0.4052,
+      "step": 12887
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 34.17818382225335,
+      "learning_rate": 5.159042567737739e-06,
+      "loss": 0.4999,
+      "step": 12888
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 39.98474928280839,
+      "learning_rate": 5.157567012122241e-06,
+      "loss": 0.5142,
+      "step": 12889
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 42.599292948613076,
+      "learning_rate": 5.1560915942227115e-06,
+      "loss": 0.5397,
+      "step": 12890
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 43.88342879373464,
+      "learning_rate": 5.154616314081098e-06,
+      "loss": 0.5493,
+      "step": 12891
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.48240925559777,
+      "learning_rate": 5.15314117173936e-06,
+      "loss": 0.4678,
+      "step": 12892
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 44.22166044544271,
+      "learning_rate": 5.151666167239452e-06,
+      "loss": 0.505,
+      "step": 12893
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 45.15914958369385,
+      "learning_rate": 5.150191300623326e-06,
+      "loss": 0.5395,
+      "step": 12894
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 37.312360887891145,
+      "learning_rate": 5.1487165719329225e-06,
+      "loss": 0.4925,
+      "step": 12895
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 39.31634922317182,
+      "learning_rate": 5.147241981210178e-06,
+      "loss": 0.5381,
+      "step": 12896
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 37.89276508383292,
+      "learning_rate": 5.145767528497032e-06,
+      "loss": 0.44,
+      "step": 12897
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.978466546679854,
+      "learning_rate": 5.1442932138354225e-06,
+      "loss": 0.5075,
+      "step": 12898
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 29.445500884107556,
+      "learning_rate": 5.142819037267271e-06,
+      "loss": 0.4567,
+      "step": 12899
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 41.00122640101362,
+      "learning_rate": 5.1413449988345055e-06,
+      "loss": 0.5696,
+      "step": 12900
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.15464251468322,
+      "learning_rate": 5.139871098579045e-06,
+      "loss": 0.449,
+      "step": 12901
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 36.83899239324994,
+      "learning_rate": 5.138397336542809e-06,
+      "loss": 0.4409,
+      "step": 12902
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 36.08494052141533,
+      "learning_rate": 5.136923712767706e-06,
+      "loss": 0.4504,
+      "step": 12903
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 44.093833550171944,
+      "learning_rate": 5.1354502272956486e-06,
+      "loss": 0.5032,
+      "step": 12904
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 34.06315424579544,
+      "learning_rate": 5.133976880168543e-06,
+      "loss": 0.4112,
+      "step": 12905
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.70395984695206,
+      "learning_rate": 5.132503671428286e-06,
+      "loss": 0.5005,
+      "step": 12906
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 41.90420238774924,
+      "learning_rate": 5.13103060111678e-06,
+      "loss": 0.5374,
+      "step": 12907
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 33.51481697869128,
+      "learning_rate": 5.129557669275913e-06,
+      "loss": 0.4634,
+      "step": 12908
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 44.81047541142899,
+      "learning_rate": 5.128084875947579e-06,
+      "loss": 0.5227,
+      "step": 12909
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.73930541815453,
+      "learning_rate": 5.126612221173659e-06,
+      "loss": 0.5038,
+      "step": 12910
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.01085631865959,
+      "learning_rate": 5.125139704996038e-06,
+      "loss": 0.473,
+      "step": 12911
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.808227435177066,
+      "learning_rate": 5.123667327456591e-06,
+      "loss": 0.5669,
+      "step": 12912
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 34.027028872018676,
+      "learning_rate": 5.12219508859719e-06,
+      "loss": 0.4549,
+      "step": 12913
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 36.28679081732683,
+      "learning_rate": 5.1207229884597135e-06,
+      "loss": 0.476,
+      "step": 12914
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 47.185219594661326,
+      "learning_rate": 5.119251027086019e-06,
+      "loss": 0.566,
+      "step": 12915
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 41.97074547381663,
+      "learning_rate": 5.117779204517967e-06,
+      "loss": 0.4778,
+      "step": 12916
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 43.90231494578477,
+      "learning_rate": 5.116307520797419e-06,
+      "loss": 0.5086,
+      "step": 12917
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.03519084391838,
+      "learning_rate": 5.114835975966228e-06,
+      "loss": 0.4618,
+      "step": 12918
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.73293299257134,
+      "learning_rate": 5.113364570066248e-06,
+      "loss": 0.487,
+      "step": 12919
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 33.55328793254757,
+      "learning_rate": 5.11189330313932e-06,
+      "loss": 0.4402,
+      "step": 12920
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.277505051230904,
+      "learning_rate": 5.110422175227284e-06,
+      "loss": 0.5496,
+      "step": 12921
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 46.83610970129024,
+      "learning_rate": 5.1089511863719824e-06,
+      "loss": 0.5293,
+      "step": 12922
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 51.96517514098277,
+      "learning_rate": 5.10748033661525e-06,
+      "loss": 0.6025,
+      "step": 12923
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 34.6288265771697,
+      "learning_rate": 5.106009625998913e-06,
+      "loss": 0.4498,
+      "step": 12924
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 50.08243188771425,
+      "learning_rate": 5.104539054564799e-06,
+      "loss": 0.5569,
+      "step": 12925
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 38.89544464874714,
+      "learning_rate": 5.103068622354729e-06,
+      "loss": 0.4434,
+      "step": 12926
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.0924519903905,
+      "learning_rate": 5.101598329410522e-06,
+      "loss": 0.4463,
+      "step": 12927
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 43.76398648262879,
+      "learning_rate": 5.100128175773998e-06,
+      "loss": 0.5534,
+      "step": 12928
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.913686178600855,
+      "learning_rate": 5.098658161486962e-06,
+      "loss": 0.4718,
+      "step": 12929
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 52.69894172268209,
+      "learning_rate": 5.097188286591217e-06,
+      "loss": 0.5689,
+      "step": 12930
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 45.77367127478132,
+      "learning_rate": 5.095718551128569e-06,
+      "loss": 0.5128,
+      "step": 12931
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.291408536347795,
+      "learning_rate": 5.09424895514082e-06,
+      "loss": 0.5793,
+      "step": 12932
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.966425724280626,
+      "learning_rate": 5.092779498669758e-06,
+      "loss": 0.5213,
+      "step": 12933
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.714726955608995,
+      "learning_rate": 5.091310181757178e-06,
+      "loss": 0.5535,
+      "step": 12934
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 50.28995011894375,
+      "learning_rate": 5.089841004444864e-06,
+      "loss": 0.5347,
+      "step": 12935
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 36.22665098573146,
+      "learning_rate": 5.088371966774601e-06,
+      "loss": 0.4723,
+      "step": 12936
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.93336848706931,
+      "learning_rate": 5.086903068788163e-06,
+      "loss": 0.4464,
+      "step": 12937
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 42.80846378485257,
+      "learning_rate": 5.085434310527329e-06,
+      "loss": 0.4581,
+      "step": 12938
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.55239136137884,
+      "learning_rate": 5.083965692033871e-06,
+      "loss": 0.566,
+      "step": 12939
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 38.274987884908256,
+      "learning_rate": 5.0824972133495535e-06,
+      "loss": 0.5701,
+      "step": 12940
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 35.39532317919733,
+      "learning_rate": 5.081028874516134e-06,
+      "loss": 0.5256,
+      "step": 12941
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.06898365991362,
+      "learning_rate": 5.0795606755753745e-06,
+      "loss": 0.5103,
+      "step": 12942
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.2097903618498,
+      "learning_rate": 5.0780926165690325e-06,
+      "loss": 0.5462,
+      "step": 12943
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 35.41186939099484,
+      "learning_rate": 5.076624697538861e-06,
+      "loss": 0.4843,
+      "step": 12944
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 31.133373486144446,
+      "learning_rate": 5.075156918526604e-06,
+      "loss": 0.4488,
+      "step": 12945
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 44.70079052031944,
+      "learning_rate": 5.073689279573999e-06,
+      "loss": 0.5191,
+      "step": 12946
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.34688558781828,
+      "learning_rate": 5.072221780722788e-06,
+      "loss": 0.5262,
+      "step": 12947
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 43.724249373242294,
+      "learning_rate": 5.070754422014711e-06,
+      "loss": 0.5628,
+      "step": 12948
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.05614842428613,
+      "learning_rate": 5.069287203491495e-06,
+      "loss": 0.4514,
+      "step": 12949
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.313227584109235,
+      "learning_rate": 5.0678201251948625e-06,
+      "loss": 0.5411,
+      "step": 12950
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 43.37347149423144,
+      "learning_rate": 5.066353187166542e-06,
+      "loss": 0.5409,
+      "step": 12951
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 46.40889874743667,
+      "learning_rate": 5.064886389448249e-06,
+      "loss": 0.4619,
+      "step": 12952
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.56760237912559,
+      "learning_rate": 5.063419732081704e-06,
+      "loss": 0.499,
+      "step": 12953
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 35.22022289929645,
+      "learning_rate": 5.061953215108614e-06,
+      "loss": 0.4729,
+      "step": 12954
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 43.206589847852115,
+      "learning_rate": 5.060486838570682e-06,
+      "loss": 0.5871,
+      "step": 12955
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 31.901142178539306,
+      "learning_rate": 5.059020602509616e-06,
+      "loss": 0.435,
+      "step": 12956
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 44.06810326627135,
+      "learning_rate": 5.057554506967116e-06,
+      "loss": 0.5024,
+      "step": 12957
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 47.842670077164975,
+      "learning_rate": 5.056088551984876e-06,
+      "loss": 0.5868,
+      "step": 12958
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 49.950563748742006,
+      "learning_rate": 5.0546227376045796e-06,
+      "loss": 0.5468,
+      "step": 12959
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 41.72279106884794,
+      "learning_rate": 5.053157063867922e-06,
+      "loss": 0.5856,
+      "step": 12960
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.17816831281011,
+      "learning_rate": 5.051691530816586e-06,
+      "loss": 0.5515,
+      "step": 12961
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 71.6864334931731,
+      "learning_rate": 5.050226138492245e-06,
+      "loss": 0.5381,
+      "step": 12962
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.09885553675331,
+      "learning_rate": 5.048760886936581e-06,
+      "loss": 0.4472,
+      "step": 12963
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 31.92507739442303,
+      "learning_rate": 5.0472957761912565e-06,
+      "loss": 0.4017,
+      "step": 12964
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 45.21679139657304,
+      "learning_rate": 5.0458308062979466e-06,
+      "loss": 0.5094,
+      "step": 12965
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.61034897145584,
+      "learning_rate": 5.0443659772983064e-06,
+      "loss": 0.4763,
+      "step": 12966
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 51.272579875878066,
+      "learning_rate": 5.042901289234001e-06,
+      "loss": 0.4923,
+      "step": 12967
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.31484144094476,
+      "learning_rate": 5.041436742146685e-06,
+      "loss": 0.51,
+      "step": 12968
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.56901121890226,
+      "learning_rate": 5.039972336078004e-06,
+      "loss": 0.5057,
+      "step": 12969
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.82055371207498,
+      "learning_rate": 5.038508071069612e-06,
+      "loss": 0.5672,
+      "step": 12970
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 40.76246872391564,
+      "learning_rate": 5.037043947163145e-06,
+      "loss": 0.5431,
+      "step": 12971
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.493948850208085,
+      "learning_rate": 5.035579964400244e-06,
+      "loss": 0.5056,
+      "step": 12972
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.58746250205549,
+      "learning_rate": 5.034116122822547e-06,
+      "loss": 0.5411,
+      "step": 12973
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 42.525220478614344,
+      "learning_rate": 5.032652422471685e-06,
+      "loss": 0.4725,
+      "step": 12974
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.108938745658186,
+      "learning_rate": 5.031188863389278e-06,
+      "loss": 0.5217,
+      "step": 12975
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.54696418936027,
+      "learning_rate": 5.029725445616953e-06,
+      "loss": 0.4973,
+      "step": 12976
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 42.96915580974031,
+      "learning_rate": 5.028262169196329e-06,
+      "loss": 0.5405,
+      "step": 12977
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 37.42859791869918,
+      "learning_rate": 5.0267990341690275e-06,
+      "loss": 0.5085,
+      "step": 12978
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 37.42097663517818,
+      "learning_rate": 5.025336040576645e-06,
+      "loss": 0.4334,
+      "step": 12979
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 54.95762226704215,
+      "learning_rate": 5.023873188460795e-06,
+      "loss": 0.5744,
+      "step": 12980
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.796470593892366,
+      "learning_rate": 5.02241047786308e-06,
+      "loss": 0.5313,
+      "step": 12981
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 36.09768245462012,
+      "learning_rate": 5.020947908825104e-06,
+      "loss": 0.5248,
+      "step": 12982
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 40.452085046230664,
+      "learning_rate": 5.019485481388457e-06,
+      "loss": 0.5347,
+      "step": 12983
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 45.50747104238347,
+      "learning_rate": 5.018023195594726e-06,
+      "loss": 0.5498,
+      "step": 12984
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 40.76085280781277,
+      "learning_rate": 5.0165610514855005e-06,
+      "loss": 0.4983,
+      "step": 12985
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 43.14702340177393,
+      "learning_rate": 5.0150990491023655e-06,
+      "loss": 0.5794,
+      "step": 12986
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.9996357905985,
+      "learning_rate": 5.013637188486895e-06,
+      "loss": 0.4805,
+      "step": 12987
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 42.72948807524656,
+      "learning_rate": 5.0121754696806705e-06,
+      "loss": 0.4269,
+      "step": 12988
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.01198767128943,
+      "learning_rate": 5.010713892725253e-06,
+      "loss": 0.5546,
+      "step": 12989
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.649727979614696,
+      "learning_rate": 5.009252457662215e-06,
+      "loss": 0.4518,
+      "step": 12990
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 28.76132471699841,
+      "learning_rate": 5.00779116453312e-06,
+      "loss": 0.4354,
+      "step": 12991
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 50.507266357490224,
+      "learning_rate": 5.00633001337952e-06,
+      "loss": 0.5051,
+      "step": 12992
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 33.54140958075398,
+      "learning_rate": 5.004869004242978e-06,
+      "loss": 0.4834,
+      "step": 12993
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.00724615790499,
+      "learning_rate": 5.0034081371650355e-06,
+      "loss": 0.4759,
+      "step": 12994
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 45.022914467083204,
+      "learning_rate": 5.001947412187246e-06,
+      "loss": 0.4627,
+      "step": 12995
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.8731334691879,
+      "learning_rate": 5.000486829351145e-06,
+      "loss": 0.4472,
+      "step": 12996
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 42.25588259853909,
+      "learning_rate": 4.999026388698274e-06,
+      "loss": 0.5586,
+      "step": 12997
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.13768203373362,
+      "learning_rate": 4.9975660902701695e-06,
+      "loss": 0.4611,
+      "step": 12998
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 37.85880013036122,
+      "learning_rate": 4.996105934108361e-06,
+      "loss": 0.4242,
+      "step": 12999
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.734990834045114,
+      "learning_rate": 4.9946459202543675e-06,
+      "loss": 0.4502,
+      "step": 13000
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 48.263735263985524,
+      "learning_rate": 4.993186048749717e-06,
+      "loss": 0.4445,
+      "step": 13001
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 43.58668328984087,
+      "learning_rate": 4.991726319635925e-06,
+      "loss": 0.5209,
+      "step": 13002
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 42.83041267728822,
+      "learning_rate": 4.990266732954515e-06,
+      "loss": 0.5278,
+      "step": 13003
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 31.294089571887714,
+      "learning_rate": 4.98880728874698e-06,
+      "loss": 0.4139,
+      "step": 13004
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 54.434347768761754,
+      "learning_rate": 4.987347987054835e-06,
+      "loss": 0.577,
+      "step": 13005
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 55.69631261485451,
+      "learning_rate": 4.9858888279195815e-06,
+      "loss": 0.566,
+      "step": 13006
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 42.79622349655775,
+      "learning_rate": 4.984429811382721e-06,
+      "loss": 0.5575,
+      "step": 13007
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.452261119487,
+      "learning_rate": 4.982970937485741e-06,
+      "loss": 0.5131,
+      "step": 13008
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 37.83397676884224,
+      "learning_rate": 4.981512206270129e-06,
+      "loss": 0.5085,
+      "step": 13009
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 40.76759790374271,
+      "learning_rate": 4.980053617777375e-06,
+      "loss": 0.54,
+      "step": 13010
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 40.188892660734645,
+      "learning_rate": 4.978595172048963e-06,
+      "loss": 0.5379,
+      "step": 13011
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 46.579127012043074,
+      "learning_rate": 4.977136869126366e-06,
+      "loss": 0.591,
+      "step": 13012
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.93648756710988,
+      "learning_rate": 4.975678709051056e-06,
+      "loss": 0.498,
+      "step": 13013
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 34.078618468883704,
+      "learning_rate": 4.974220691864503e-06,
+      "loss": 0.4784,
+      "step": 13014
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 47.41107589208159,
+      "learning_rate": 4.972762817608174e-06,
+      "loss": 0.5149,
+      "step": 13015
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 34.454480694245355,
+      "learning_rate": 4.971305086323534e-06,
+      "loss": 0.455,
+      "step": 13016
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.37744565362155,
+      "learning_rate": 4.969847498052035e-06,
+      "loss": 0.5471,
+      "step": 13017
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 50.975985071412346,
+      "learning_rate": 4.968390052835127e-06,
+      "loss": 0.5123,
+      "step": 13018
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.948364695301414,
+      "learning_rate": 4.966932750714262e-06,
+      "loss": 0.4642,
+      "step": 13019
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 51.557778594281636,
+      "learning_rate": 4.96547559173089e-06,
+      "loss": 0.5887,
+      "step": 13020
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 44.24314160457454,
+      "learning_rate": 4.964018575926442e-06,
+      "loss": 0.4409,
+      "step": 13021
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 41.84394487103418,
+      "learning_rate": 4.962561703342364e-06,
+      "loss": 0.4794,
+      "step": 13022
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 46.98367729214472,
+      "learning_rate": 4.961104974020081e-06,
+      "loss": 0.557,
+      "step": 13023
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 36.7558049366209,
+      "learning_rate": 4.959648388001028e-06,
+      "loss": 0.5089,
+      "step": 13024
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 39.12963882343945,
+      "learning_rate": 4.958191945326624e-06,
+      "loss": 0.5206,
+      "step": 13025
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 41.32891923397245,
+      "learning_rate": 4.9567356460382895e-06,
+      "loss": 0.5736,
+      "step": 13026
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.18251568845568,
+      "learning_rate": 4.955279490177448e-06,
+      "loss": 0.5114,
+      "step": 13027
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 38.21516744458425,
+      "learning_rate": 4.953823477785507e-06,
+      "loss": 0.4976,
+      "step": 13028
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 43.40115185892276,
+      "learning_rate": 4.952367608903871e-06,
+      "loss": 0.5462,
+      "step": 13029
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 50.00665546944656,
+      "learning_rate": 4.9509118835739465e-06,
+      "loss": 0.5746,
+      "step": 13030
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 37.19742319439729,
+      "learning_rate": 4.949456301837134e-06,
+      "loss": 0.5019,
+      "step": 13031
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 33.50742208049264,
+      "learning_rate": 4.948000863734834e-06,
+      "loss": 0.4177,
+      "step": 13032
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 35.79678046331823,
+      "learning_rate": 4.946545569308436e-06,
+      "loss": 0.4841,
+      "step": 13033
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 36.31422563021098,
+      "learning_rate": 4.94509041859932e-06,
+      "loss": 0.4692,
+      "step": 13034
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 40.99780548723938,
+      "learning_rate": 4.943635411648875e-06,
+      "loss": 0.4776,
+      "step": 13035
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 40.27776479625858,
+      "learning_rate": 4.942180548498485e-06,
+      "loss": 0.5015,
+      "step": 13036
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 44.04145395317185,
+      "learning_rate": 4.940725829189523e-06,
+      "loss": 0.5879,
+      "step": 13037
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 38.19906651132731,
+      "learning_rate": 4.939271253763354e-06,
+      "loss": 0.4794,
+      "step": 13038
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 37.74044807482949,
+      "learning_rate": 4.93781682226135e-06,
+      "loss": 0.515,
+      "step": 13039
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 40.61754503953063,
+      "learning_rate": 4.936362534724874e-06,
+      "loss": 0.4743,
+      "step": 13040
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 56.291387289988364,
+      "learning_rate": 4.93490839119529e-06,
+      "loss": 0.472,
+      "step": 13041
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 48.11247042751419,
+      "learning_rate": 4.933454391713948e-06,
+      "loss": 0.46,
+      "step": 13042
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 42.451484631636134,
+      "learning_rate": 4.932000536322196e-06,
+      "loss": 0.5324,
+      "step": 13043
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 41.93373810770375,
+      "learning_rate": 4.930546825061383e-06,
+      "loss": 0.5193,
+      "step": 13044
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 43.452468016649824,
+      "learning_rate": 4.929093257972857e-06,
+      "loss": 0.5038,
+      "step": 13045
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 45.497785786839415,
+      "learning_rate": 4.927639835097948e-06,
+      "loss": 0.5263,
+      "step": 13046
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 35.452922680926044,
+      "learning_rate": 4.926186556478e-06,
+      "loss": 0.4896,
+      "step": 13047
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 34.68254819012169,
+      "learning_rate": 4.924733422154333e-06,
+      "loss": 0.4638,
+      "step": 13048
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 37.19971920543235,
+      "learning_rate": 4.9232804321682835e-06,
+      "loss": 0.4957,
+      "step": 13049
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 42.73683883008284,
+      "learning_rate": 4.921827586561164e-06,
+      "loss": 0.5455,
+      "step": 13050
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 37.5475348939663,
+      "learning_rate": 4.9203748853742986e-06,
+      "loss": 0.4349,
+      "step": 13051
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 44.23793049452213,
+      "learning_rate": 4.918922328649003e-06,
+      "loss": 0.4815,
+      "step": 13052
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 56.48172226371003,
+      "learning_rate": 4.917469916426584e-06,
+      "loss": 0.689,
+      "step": 13053
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 42.33827359303967,
+      "learning_rate": 4.916017648748342e-06,
+      "loss": 0.4817,
+      "step": 13054
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 48.34716865174705,
+      "learning_rate": 4.914565525655587e-06,
+      "loss": 0.5846,
+      "step": 13055
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 43.122498724450494,
+      "learning_rate": 4.913113547189615e-06,
+      "loss": 0.4327,
+      "step": 13056
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 37.08179446395649,
+      "learning_rate": 4.911661713391716e-06,
+      "loss": 0.473,
+      "step": 13057
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 46.86493438478429,
+      "learning_rate": 4.9102100243031835e-06,
+      "loss": 0.5423,
+      "step": 13058
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 40.13148808535673,
+      "learning_rate": 4.9087584799652976e-06,
+      "loss": 0.4963,
+      "step": 13059
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 46.47734317710609,
+      "learning_rate": 4.907307080419341e-06,
+      "loss": 0.542,
+      "step": 13060
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 50.007650190655255,
+      "learning_rate": 4.905855825706597e-06,
+      "loss": 0.5823,
+      "step": 13061
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 40.439329179667354,
+      "learning_rate": 4.904404715868333e-06,
+      "loss": 0.5314,
+      "step": 13062
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 34.03126025309202,
+      "learning_rate": 4.902953750945814e-06,
+      "loss": 0.4511,
+      "step": 13063
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 42.69948783205535,
+      "learning_rate": 4.90150293098031e-06,
+      "loss": 0.522,
+      "step": 13064
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 50.62529552043797,
+      "learning_rate": 4.900052256013078e-06,
+      "loss": 0.6319,
+      "step": 13065
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 42.635547680921064,
+      "learning_rate": 4.898601726085385e-06,
+      "loss": 0.527,
+      "step": 13066
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 38.85545632759188,
+      "learning_rate": 4.897151341238468e-06,
+      "loss": 0.4663,
+      "step": 13067
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 46.857955641459355,
+      "learning_rate": 4.89570110151358e-06,
+      "loss": 0.492,
+      "step": 13068
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 33.449768229395794,
+      "learning_rate": 4.894251006951968e-06,
+      "loss": 0.4478,
+      "step": 13069
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 53.73348131293531,
+      "learning_rate": 4.892801057594874e-06,
+      "loss": 0.5815,
+      "step": 13070
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 36.93518039113476,
+      "learning_rate": 4.891351253483529e-06,
+      "loss": 0.5394,
+      "step": 13071
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 39.62140950924549,
+      "learning_rate": 4.8899015946591645e-06,
+      "loss": 0.4902,
+      "step": 13072
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 49.657646543671355,
+      "learning_rate": 4.888452081163007e-06,
+      "loss": 0.529,
+      "step": 13073
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 38.6705433132653,
+      "learning_rate": 4.8870027130362885e-06,
+      "loss": 0.4502,
+      "step": 13074
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 46.53970506041036,
+      "learning_rate": 4.8855534903202165e-06,
+      "loss": 0.5233,
+      "step": 13075
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 43.6519029073622,
+      "learning_rate": 4.884104413056015e-06,
+      "loss": 0.526,
+      "step": 13076
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 39.35539223637491,
+      "learning_rate": 4.882655481284888e-06,
+      "loss": 0.495,
+      "step": 13077
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 40.92073827725434,
+      "learning_rate": 4.881206695048046e-06,
+      "loss": 0.5118,
+      "step": 13078
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 46.009518820210026,
+      "learning_rate": 4.879758054386694e-06,
+      "loss": 0.5394,
+      "step": 13079
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 44.410451605703244,
+      "learning_rate": 4.8783095593420234e-06,
+      "loss": 0.5406,
+      "step": 13080
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 44.71405930779531,
+      "learning_rate": 4.876861209955238e-06,
+      "loss": 0.5259,
+      "step": 13081
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 51.59472603108645,
+      "learning_rate": 4.875413006267518e-06,
+      "loss": 0.6271,
+      "step": 13082
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 49.08321024172795,
+      "learning_rate": 4.873964948320059e-06,
+      "loss": 0.54,
+      "step": 13083
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 49.85774937689992,
+      "learning_rate": 4.8725170361540345e-06,
+      "loss": 0.5188,
+      "step": 13084
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 41.881610337224224,
+      "learning_rate": 4.871069269810626e-06,
+      "loss": 0.4742,
+      "step": 13085
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 65.4106102668813,
+      "learning_rate": 4.869621649331009e-06,
+      "loss": 0.4581,
+      "step": 13086
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 48.758461452640276,
+      "learning_rate": 4.868174174756353e-06,
+      "loss": 0.5395,
+      "step": 13087
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 36.743643914926245,
+      "learning_rate": 4.866726846127817e-06,
+      "loss": 0.3817,
+      "step": 13088
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 50.38868679843535,
+      "learning_rate": 4.865279663486567e-06,
+      "loss": 0.5821,
+      "step": 13089
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 33.15304792296153,
+      "learning_rate": 4.863832626873759e-06,
+      "loss": 0.4667,
+      "step": 13090
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 32.366570358414634,
+      "learning_rate": 4.862385736330555e-06,
+      "loss": 0.4376,
+      "step": 13091
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 44.81771914097598,
+      "learning_rate": 4.860938991898088e-06,
+      "loss": 0.5437,
+      "step": 13092
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 42.645767329577545,
+      "learning_rate": 4.859492393617509e-06,
+      "loss": 0.467,
+      "step": 13093
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.582894419164184,
+      "learning_rate": 4.8580459415299585e-06,
+      "loss": 0.4968,
+      "step": 13094
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 35.685647138072966,
+      "learning_rate": 4.856599635676578e-06,
+      "loss": 0.4557,
+      "step": 13095
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.60632463041726,
+      "learning_rate": 4.8551534760984954e-06,
+      "loss": 0.4509,
+      "step": 13096
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 45.71213374107198,
+      "learning_rate": 4.853707462836834e-06,
+      "loss": 0.4966,
+      "step": 13097
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 38.71061752910774,
+      "learning_rate": 4.852261595932721e-06,
+      "loss": 0.4682,
+      "step": 13098
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 36.75856932814611,
+      "learning_rate": 4.850815875427283e-06,
+      "loss": 0.4694,
+      "step": 13099
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.009436254172904,
+      "learning_rate": 4.849370301361623e-06,
+      "loss": 0.5624,
+      "step": 13100
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.71726536198627,
+      "learning_rate": 4.8479248737768644e-06,
+      "loss": 0.513,
+      "step": 13101
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.84531415048576,
+      "learning_rate": 4.846479592714104e-06,
+      "loss": 0.4876,
+      "step": 13102
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.49704992770368,
+      "learning_rate": 4.845034458214449e-06,
+      "loss": 0.4832,
+      "step": 13103
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 43.29194772768006,
+      "learning_rate": 4.8435894703190026e-06,
+      "loss": 0.5306,
+      "step": 13104
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.82413895628124,
+      "learning_rate": 4.842144629068854e-06,
+      "loss": 0.4588,
+      "step": 13105
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.49497686836438,
+      "learning_rate": 4.8406999345050934e-06,
+      "loss": 0.5702,
+      "step": 13106
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.77930559471474,
+      "learning_rate": 4.839255386668806e-06,
+      "loss": 0.5755,
+      "step": 13107
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 31.470699996157354,
+      "learning_rate": 4.8378109856010825e-06,
+      "loss": 0.4136,
+      "step": 13108
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 38.782244359869814,
+      "learning_rate": 4.8363667313429896e-06,
+      "loss": 0.4848,
+      "step": 13109
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 39.2160233602248,
+      "learning_rate": 4.83492262393561e-06,
+      "loss": 0.5241,
+      "step": 13110
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 43.774374880158845,
+      "learning_rate": 4.833478663420007e-06,
+      "loss": 0.5062,
+      "step": 13111
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.57308137478785,
+      "learning_rate": 4.832034849837252e-06,
+      "loss": 0.5104,
+      "step": 13112
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 47.55017220067834,
+      "learning_rate": 4.830591183228398e-06,
+      "loss": 0.5288,
+      "step": 13113
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 34.45549185098861,
+      "learning_rate": 4.829147663634507e-06,
+      "loss": 0.4317,
+      "step": 13114
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 45.771201489784836,
+      "learning_rate": 4.827704291096635e-06,
+      "loss": 0.4886,
+      "step": 13115
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 36.76876441855426,
+      "learning_rate": 4.826261065655828e-06,
+      "loss": 0.4581,
+      "step": 13116
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.01636894395479,
+      "learning_rate": 4.824817987353126e-06,
+      "loss": 0.5399,
+      "step": 13117
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.37688256501481,
+      "learning_rate": 4.823375056229573e-06,
+      "loss": 0.5605,
+      "step": 13118
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 42.5388386113604,
+      "learning_rate": 4.821932272326204e-06,
+      "loss": 0.5293,
+      "step": 13119
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 36.11664399850612,
+      "learning_rate": 4.820489635684057e-06,
+      "loss": 0.5018,
+      "step": 13120
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 43.6725644707208,
+      "learning_rate": 4.819047146344154e-06,
+      "loss": 0.4204,
+      "step": 13121
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 39.65349080472214,
+      "learning_rate": 4.817604804347517e-06,
+      "loss": 0.4661,
+      "step": 13122
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 44.24843916322361,
+      "learning_rate": 4.816162609735167e-06,
+      "loss": 0.5177,
+      "step": 13123
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.262927396128944,
+      "learning_rate": 4.814720562548123e-06,
+      "loss": 0.5219,
+      "step": 13124
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 52.539437731002074,
+      "learning_rate": 4.8132786628273945e-06,
+      "loss": 0.5514,
+      "step": 13125
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 47.3202604734712,
+      "learning_rate": 4.811836910613982e-06,
+      "loss": 0.4642,
+      "step": 13126
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.14736185574578,
+      "learning_rate": 4.810395305948892e-06,
+      "loss": 0.475,
+      "step": 13127
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 45.410040422189304,
+      "learning_rate": 4.8089538488731244e-06,
+      "loss": 0.5342,
+      "step": 13128
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 42.609076588071474,
+      "learning_rate": 4.807512539427677e-06,
+      "loss": 0.5211,
+      "step": 13129
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 44.77570068252022,
+      "learning_rate": 4.806071377653534e-06,
+      "loss": 0.4874,
+      "step": 13130
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 34.630930175846565,
+      "learning_rate": 4.80463036359168e-06,
+      "loss": 0.4031,
+      "step": 13131
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.0729779554423,
+      "learning_rate": 4.803189497283099e-06,
+      "loss": 0.4057,
+      "step": 13132
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 38.6250732260692,
+      "learning_rate": 4.801748778768772e-06,
+      "loss": 0.5044,
+      "step": 13133
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 49.69621024889848,
+      "learning_rate": 4.800308208089666e-06,
+      "loss": 0.5073,
+      "step": 13134
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 35.49555024614295,
+      "learning_rate": 4.798867785286756e-06,
+      "loss": 0.5264,
+      "step": 13135
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 49.86846153882469,
+      "learning_rate": 4.797427510401001e-06,
+      "loss": 0.5215,
+      "step": 13136
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 42.681100703581755,
+      "learning_rate": 4.795987383473366e-06,
+      "loss": 0.5008,
+      "step": 13137
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 48.17480522446981,
+      "learning_rate": 4.794547404544802e-06,
+      "loss": 0.5341,
+      "step": 13138
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 43.004014093979535,
+      "learning_rate": 4.793107573656266e-06,
+      "loss": 0.5495,
+      "step": 13139
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.33410601252598,
+      "learning_rate": 4.791667890848708e-06,
+      "loss": 0.4757,
+      "step": 13140
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 44.55128341318208,
+      "learning_rate": 4.790228356163065e-06,
+      "loss": 0.4773,
+      "step": 13141
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.32194613849446,
+      "learning_rate": 4.788788969640283e-06,
+      "loss": 0.4246,
+      "step": 13142
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.76672575375872,
+      "learning_rate": 4.787349731321291e-06,
+      "loss": 0.5525,
+      "step": 13143
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 42.699232365790394,
+      "learning_rate": 4.785910641247022e-06,
+      "loss": 0.4318,
+      "step": 13144
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.89319965227785,
+      "learning_rate": 4.78447169945841e-06,
+      "loss": 0.4836,
+      "step": 13145
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 49.51205005783198,
+      "learning_rate": 4.78303290599637e-06,
+      "loss": 0.4558,
+      "step": 13146
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 47.54606020477184,
+      "learning_rate": 4.78159426090182e-06,
+      "loss": 0.4609,
+      "step": 13147
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 39.291278857760986,
+      "learning_rate": 4.780155764215676e-06,
+      "loss": 0.3781,
+      "step": 13148
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 33.43327275231395,
+      "learning_rate": 4.7787174159788514e-06,
+      "loss": 0.4531,
+      "step": 13149
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 46.154786651700086,
+      "learning_rate": 4.7772792162322505e-06,
+      "loss": 0.5232,
+      "step": 13150
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 56.57483072495769,
+      "learning_rate": 4.775841165016769e-06,
+      "loss": 0.4999,
+      "step": 13151
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 37.940838615030415,
+      "learning_rate": 4.774403262373308e-06,
+      "loss": 0.5084,
+      "step": 13152
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 39.45629692697234,
+      "learning_rate": 4.772965508342763e-06,
+      "loss": 0.4595,
+      "step": 13153
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 41.136722325984216,
+      "learning_rate": 4.771527902966028e-06,
+      "loss": 0.5079,
+      "step": 13154
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 39.47189089614926,
+      "learning_rate": 4.7700904462839735e-06,
+      "loss": 0.4631,
+      "step": 13155
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 32.59874443491513,
+      "learning_rate": 4.7686531383374865e-06,
+      "loss": 0.3765,
+      "step": 13156
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 42.69252932901409,
+      "learning_rate": 4.767215979167445e-06,
+      "loss": 0.4967,
+      "step": 13157
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 47.29143678158358,
+      "learning_rate": 4.765778968814724e-06,
+      "loss": 0.5902,
+      "step": 13158
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 36.42748045857119,
+      "learning_rate": 4.764342107320186e-06,
+      "loss": 0.4536,
+      "step": 13159
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 39.15860201141433,
+      "learning_rate": 4.762905394724691e-06,
+      "loss": 0.4551,
+      "step": 13160
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 38.473349707608286,
+      "learning_rate": 4.761468831069106e-06,
+      "loss": 0.4587,
+      "step": 13161
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 38.06395290213818,
+      "learning_rate": 4.7600324163942845e-06,
+      "loss": 0.4868,
+      "step": 13162
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 43.837742924839766,
+      "learning_rate": 4.758596150741073e-06,
+      "loss": 0.5681,
+      "step": 13163
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 40.0326534266442,
+      "learning_rate": 4.757160034150324e-06,
+      "loss": 0.5657,
+      "step": 13164
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 48.65490231273764,
+      "learning_rate": 4.7557240666628735e-06,
+      "loss": 0.5514,
+      "step": 13165
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 31.72928781834844,
+      "learning_rate": 4.754288248319563e-06,
+      "loss": 0.4497,
+      "step": 13166
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 44.911298968391336,
+      "learning_rate": 4.752852579161229e-06,
+      "loss": 0.5065,
+      "step": 13167
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 41.214998208975764,
+      "learning_rate": 4.751417059228695e-06,
+      "loss": 0.5019,
+      "step": 13168
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 53.506529561103264,
+      "learning_rate": 4.749981688562792e-06,
+      "loss": 0.5422,
+      "step": 13169
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 37.24664405772282,
+      "learning_rate": 4.748546467204336e-06,
+      "loss": 0.4333,
+      "step": 13170
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 47.595876700061986,
+      "learning_rate": 4.747111395194149e-06,
+      "loss": 0.5798,
+      "step": 13171
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 40.259327488913904,
+      "learning_rate": 4.7456764725730385e-06,
+      "loss": 0.4586,
+      "step": 13172
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 45.06499562118613,
+      "learning_rate": 4.744241699381816e-06,
+      "loss": 0.5058,
+      "step": 13173
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 39.52689363502488,
+      "learning_rate": 4.742807075661288e-06,
+      "loss": 0.4893,
+      "step": 13174
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 40.31742721135681,
+      "learning_rate": 4.741372601452251e-06,
+      "loss": 0.5063,
+      "step": 13175
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 30.791042849803617,
+      "learning_rate": 4.739938276795498e-06,
+      "loss": 0.4452,
+      "step": 13176
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 40.24365399368904,
+      "learning_rate": 4.738504101731824e-06,
+      "loss": 0.5151,
+      "step": 13177
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 40.03262336275044,
+      "learning_rate": 4.737070076302014e-06,
+      "loss": 0.4513,
+      "step": 13178
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 46.565743869716556,
+      "learning_rate": 4.735636200546859e-06,
+      "loss": 0.4997,
+      "step": 13179
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 50.463150639718066,
+      "learning_rate": 4.734202474507126e-06,
+      "loss": 0.6076,
+      "step": 13180
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 43.37805074082611,
+      "learning_rate": 4.7327688982235916e-06,
+      "loss": 0.4979,
+      "step": 13181
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 45.03144913681653,
+      "learning_rate": 4.731335471737031e-06,
+      "loss": 0.5584,
+      "step": 13182
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 51.970557910257995,
+      "learning_rate": 4.72990219508821e-06,
+      "loss": 0.5777,
+      "step": 13183
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 35.27605982904414,
+      "learning_rate": 4.728469068317887e-06,
+      "loss": 0.466,
+      "step": 13184
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 53.69744530022903,
+      "learning_rate": 4.727036091466816e-06,
+      "loss": 0.5491,
+      "step": 13185
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 89.67682953737066,
+      "learning_rate": 4.7256032645757555e-06,
+      "loss": 0.6422,
+      "step": 13186
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 47.33091228913183,
+      "learning_rate": 4.724170587685456e-06,
+      "loss": 0.5374,
+      "step": 13187
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 33.12559395023687,
+      "learning_rate": 4.722738060836654e-06,
+      "loss": 0.4101,
+      "step": 13188
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 38.58023834311119,
+      "learning_rate": 4.7213056840701e-06,
+      "loss": 0.4784,
+      "step": 13189
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 39.318584629281666,
+      "learning_rate": 4.719873457426519e-06,
+      "loss": 0.493,
+      "step": 13190
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 35.99617279093732,
+      "learning_rate": 4.718441380946649e-06,
+      "loss": 0.4815,
+      "step": 13191
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 43.59350270847199,
+      "learning_rate": 4.717009454671221e-06,
+      "loss": 0.4554,
+      "step": 13192
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 44.41484871925838,
+      "learning_rate": 4.715577678640948e-06,
+      "loss": 0.5045,
+      "step": 13193
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 38.01732751159611,
+      "learning_rate": 4.714146052896559e-06,
+      "loss": 0.5203,
+      "step": 13194
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 41.94589756489929,
+      "learning_rate": 4.7127145774787604e-06,
+      "loss": 0.5694,
+      "step": 13195
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 34.12242732638185,
+      "learning_rate": 4.71128325242827e-06,
+      "loss": 0.4563,
+      "step": 13196
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 36.45564662225411,
+      "learning_rate": 4.709852077785787e-06,
+      "loss": 0.4529,
+      "step": 13197
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 36.77603583702074,
+      "learning_rate": 4.708421053592019e-06,
+      "loss": 0.4309,
+      "step": 13198
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 41.809606956672766,
+      "learning_rate": 4.706990179887658e-06,
+      "loss": 0.5352,
+      "step": 13199
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 37.65820931999979,
+      "learning_rate": 4.705559456713403e-06,
+      "loss": 0.4017,
+      "step": 13200
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 37.81872417516525,
+      "learning_rate": 4.704128884109936e-06,
+      "loss": 0.4945,
+      "step": 13201
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 40.36438675502368,
+      "learning_rate": 4.702698462117946e-06,
+      "loss": 0.4473,
+      "step": 13202
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 44.47420253285299,
+      "learning_rate": 4.701268190778117e-06,
+      "loss": 0.5015,
+      "step": 13203
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 35.37577636999084,
+      "learning_rate": 4.6998380701311196e-06,
+      "loss": 0.4227,
+      "step": 13204
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 41.23819312270239,
+      "learning_rate": 4.6984081002176256e-06,
+      "loss": 0.4975,
+      "step": 13205
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 45.30724328165628,
+      "learning_rate": 4.696978281078302e-06,
+      "loss": 0.5468,
+      "step": 13206
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 46.971062801877444,
+      "learning_rate": 4.695548612753814e-06,
+      "loss": 0.6874,
+      "step": 13207
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 42.74606129978764,
+      "learning_rate": 4.694119095284825e-06,
+      "loss": 0.4771,
+      "step": 13208
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 50.54713325257636,
+      "learning_rate": 4.6926897287119845e-06,
+      "loss": 0.5191,
+      "step": 13209
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 39.922713089896845,
+      "learning_rate": 4.6912605130759396e-06,
+      "loss": 0.4854,
+      "step": 13210
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 39.94443038721476,
+      "learning_rate": 4.68983144841734e-06,
+      "loss": 0.4988,
+      "step": 13211
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 41.697535069085696,
+      "learning_rate": 4.688402534776832e-06,
+      "loss": 0.4378,
+      "step": 13212
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 36.69399398959103,
+      "learning_rate": 4.686973772195048e-06,
+      "loss": 0.5124,
+      "step": 13213
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 36.305001572920325,
+      "learning_rate": 4.685545160712619e-06,
+      "loss": 0.4398,
+      "step": 13214
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 42.15623381369063,
+      "learning_rate": 4.684116700370176e-06,
+      "loss": 0.4488,
+      "step": 13215
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 35.18479674611941,
+      "learning_rate": 4.682688391208345e-06,
+      "loss": 0.3999,
+      "step": 13216
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 34.67214477145302,
+      "learning_rate": 4.681260233267749e-06,
+      "loss": 0.4797,
+      "step": 13217
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 45.489658134066424,
+      "learning_rate": 4.679832226589001e-06,
+      "loss": 0.4673,
+      "step": 13218
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 49.175411585688714,
+      "learning_rate": 4.6784043712127084e-06,
+      "loss": 0.4536,
+      "step": 13219
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 50.03634748634864,
+      "learning_rate": 4.676976667179482e-06,
+      "loss": 0.4714,
+      "step": 13220
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 38.92062283419428,
+      "learning_rate": 4.675549114529929e-06,
+      "loss": 0.5549,
+      "step": 13221
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.54343770475679,
+      "learning_rate": 4.6741217133046415e-06,
+      "loss": 0.4368,
+      "step": 13222
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 44.464435735373925,
+      "learning_rate": 4.67269446354422e-06,
+      "loss": 0.506,
+      "step": 13223
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.88134898549895,
+      "learning_rate": 4.671267365289247e-06,
+      "loss": 0.5305,
+      "step": 13224
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.938484546965164,
+      "learning_rate": 4.669840418580318e-06,
+      "loss": 0.512,
+      "step": 13225
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 50.76346895166255,
+      "learning_rate": 4.668413623458006e-06,
+      "loss": 0.6781,
+      "step": 13226
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.25462384173696,
+      "learning_rate": 4.666986979962891e-06,
+      "loss": 0.4839,
+      "step": 13227
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.77486151739926,
+      "learning_rate": 4.6655604881355495e-06,
+      "loss": 0.499,
+      "step": 13228
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 48.38855854040398,
+      "learning_rate": 4.664134148016545e-06,
+      "loss": 0.4657,
+      "step": 13229
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 32.86630564840202,
+      "learning_rate": 4.662707959646446e-06,
+      "loss": 0.4219,
+      "step": 13230
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.433436350220795,
+      "learning_rate": 4.661281923065808e-06,
+      "loss": 0.4855,
+      "step": 13231
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 48.24521343424756,
+      "learning_rate": 4.659856038315188e-06,
+      "loss": 0.5505,
+      "step": 13232
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 42.028600355391944,
+      "learning_rate": 4.658430305435143e-06,
+      "loss": 0.6227,
+      "step": 13233
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 44.02206998256413,
+      "learning_rate": 4.657004724466216e-06,
+      "loss": 0.54,
+      "step": 13234
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.92241475207044,
+      "learning_rate": 4.655579295448944e-06,
+      "loss": 0.4731,
+      "step": 13235
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 36.3188372874684,
+      "learning_rate": 4.654154018423871e-06,
+      "loss": 0.4266,
+      "step": 13236
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 47.11617103761725,
+      "learning_rate": 4.652728893431534e-06,
+      "loss": 0.4786,
+      "step": 13237
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 47.319100230740666,
+      "learning_rate": 4.651303920512459e-06,
+      "loss": 0.456,
+      "step": 13238
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 34.913488919831856,
+      "learning_rate": 4.649879099707168e-06,
+      "loss": 0.4564,
+      "step": 13239
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.97442631931699,
+      "learning_rate": 4.648454431056186e-06,
+      "loss": 0.4926,
+      "step": 13240
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 36.96040486734306,
+      "learning_rate": 4.64702991460003e-06,
+      "loss": 0.4498,
+      "step": 13241
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 42.2717025815361,
+      "learning_rate": 4.645605550379214e-06,
+      "loss": 0.4033,
+      "step": 13242
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.039286457629295,
+      "learning_rate": 4.644181338434245e-06,
+      "loss": 0.4801,
+      "step": 13243
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 45.13419970519487,
+      "learning_rate": 4.642757278805622e-06,
+      "loss": 0.4968,
+      "step": 13244
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 33.85193049435296,
+      "learning_rate": 4.641333371533848e-06,
+      "loss": 0.4442,
+      "step": 13245
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 43.10039026499674,
+      "learning_rate": 4.6399096166594215e-06,
+      "loss": 0.4752,
+      "step": 13246
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.17624617649683,
+      "learning_rate": 4.638486014222831e-06,
+      "loss": 0.4745,
+      "step": 13247
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 45.90954042591456,
+      "learning_rate": 4.6370625642645565e-06,
+      "loss": 0.5313,
+      "step": 13248
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 45.729765695922204,
+      "learning_rate": 4.635639266825086e-06,
+      "loss": 0.5025,
+      "step": 13249
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 47.96619562561057,
+      "learning_rate": 4.634216121944901e-06,
+      "loss": 0.5137,
+      "step": 13250
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 38.05166093306591,
+      "learning_rate": 4.632793129664466e-06,
+      "loss": 0.4721,
+      "step": 13251
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.649318837280546,
+      "learning_rate": 4.63137029002426e-06,
+      "loss": 0.4642,
+      "step": 13252
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.97372605854007,
+      "learning_rate": 4.629947603064737e-06,
+      "loss": 0.5151,
+      "step": 13253
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.86852780050417,
+      "learning_rate": 4.6285250688263625e-06,
+      "loss": 0.4825,
+      "step": 13254
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 51.78932126576609,
+      "learning_rate": 4.627102687349599e-06,
+      "loss": 0.6175,
+      "step": 13255
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.19461508892123,
+      "learning_rate": 4.625680458674886e-06,
+      "loss": 0.5132,
+      "step": 13256
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 46.79475258422783,
+      "learning_rate": 4.624258382842681e-06,
+      "loss": 0.4957,
+      "step": 13257
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.63265354153808,
+      "learning_rate": 4.622836459893419e-06,
+      "loss": 0.423,
+      "step": 13258
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 49.32091604631213,
+      "learning_rate": 4.621414689867547e-06,
+      "loss": 0.5719,
+      "step": 13259
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.6170597894398,
+      "learning_rate": 4.619993072805491e-06,
+      "loss": 0.4616,
+      "step": 13260
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 39.84747851111213,
+      "learning_rate": 4.618571608747685e-06,
+      "loss": 0.4344,
+      "step": 13261
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.759071643712076,
+      "learning_rate": 4.617150297734557e-06,
+      "loss": 0.4656,
+      "step": 13262
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.68448370199598,
+      "learning_rate": 4.615729139806527e-06,
+      "loss": 0.5448,
+      "step": 13263
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 39.28033075038491,
+      "learning_rate": 4.614308135004006e-06,
+      "loss": 0.4548,
+      "step": 13264
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 39.509168725688575,
+      "learning_rate": 4.612887283367411e-06,
+      "loss": 0.4447,
+      "step": 13265
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.06207374548332,
+      "learning_rate": 4.611466584937151e-06,
+      "loss": 0.5137,
+      "step": 13266
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 45.49418172210893,
+      "learning_rate": 4.610046039753636e-06,
+      "loss": 0.536,
+      "step": 13267
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 36.999834775555776,
+      "learning_rate": 4.608625647857251e-06,
+      "loss": 0.4969,
+      "step": 13268
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.361229222229795,
+      "learning_rate": 4.6072054092884e-06,
+      "loss": 0.4422,
+      "step": 13269
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 42.6592862679982,
+      "learning_rate": 4.60578532408747e-06,
+      "loss": 0.5134,
+      "step": 13270
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 56.26484589380272,
+      "learning_rate": 4.604365392294856e-06,
+      "loss": 0.5446,
+      "step": 13271
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.570693901158364,
+      "learning_rate": 4.602945613950933e-06,
+      "loss": 0.438,
+      "step": 13272
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 42.70013224628168,
+      "learning_rate": 4.6015259890960765e-06,
+      "loss": 0.4324,
+      "step": 13273
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 39.98182368961914,
+      "learning_rate": 4.600106517770662e-06,
+      "loss": 0.5263,
+      "step": 13274
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 33.1801260893819,
+      "learning_rate": 4.598687200015063e-06,
+      "loss": 0.4447,
+      "step": 13275
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 50.96107828862639,
+      "learning_rate": 4.597268035869636e-06,
+      "loss": 0.5092,
+      "step": 13276
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 43.42803903639691,
+      "learning_rate": 4.59584902537475e-06,
+      "loss": 0.5899,
+      "step": 13277
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 46.23690203186952,
+      "learning_rate": 4.594430168570753e-06,
+      "loss": 0.5449,
+      "step": 13278
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 40.42952024858314,
+      "learning_rate": 4.593011465498e-06,
+      "loss": 0.4478,
+      "step": 13279
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 44.06159754701457,
+      "learning_rate": 4.591592916196841e-06,
+      "loss": 0.4767,
+      "step": 13280
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 46.52756939659459,
+      "learning_rate": 4.590174520707612e-06,
+      "loss": 0.5188,
+      "step": 13281
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 38.735308437018055,
+      "learning_rate": 4.58875627907066e-06,
+      "loss": 0.468,
+      "step": 13282
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 46.347175502413776,
+      "learning_rate": 4.587338191326308e-06,
+      "loss": 0.4994,
+      "step": 13283
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 45.656506056593024,
+      "learning_rate": 4.585920257514897e-06,
+      "loss": 0.5092,
+      "step": 13284
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 38.31039614602805,
+      "learning_rate": 4.584502477676742e-06,
+      "loss": 0.5359,
+      "step": 13285
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 48.2642489527651,
+      "learning_rate": 4.583084851852169e-06,
+      "loss": 0.5137,
+      "step": 13286
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 51.09538870233856,
+      "learning_rate": 4.581667380081497e-06,
+      "loss": 0.5675,
+      "step": 13287
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 31.20596921333374,
+      "learning_rate": 4.580250062405036e-06,
+      "loss": 0.4322,
+      "step": 13288
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 42.06640789645859,
+      "learning_rate": 4.578832898863088e-06,
+      "loss": 0.4595,
+      "step": 13289
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 45.54926411560855,
+      "learning_rate": 4.577415889495962e-06,
+      "loss": 0.4336,
+      "step": 13290
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 42.78176799875757,
+      "learning_rate": 4.575999034343957e-06,
+      "loss": 0.5113,
+      "step": 13291
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 37.71743193111471,
+      "learning_rate": 4.5745823334473685e-06,
+      "loss": 0.4519,
+      "step": 13292
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 48.07732723195322,
+      "learning_rate": 4.573165786846485e-06,
+      "loss": 0.5339,
+      "step": 13293
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 43.533518042703164,
+      "learning_rate": 4.571749394581588e-06,
+      "loss": 0.4725,
+      "step": 13294
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 64.93864584380887,
+      "learning_rate": 4.570333156692963e-06,
+      "loss": 0.5407,
+      "step": 13295
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 38.733682919405794,
+      "learning_rate": 4.568917073220891e-06,
+      "loss": 0.4473,
+      "step": 13296
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 35.52463401791146,
+      "learning_rate": 4.567501144205639e-06,
+      "loss": 0.4197,
+      "step": 13297
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 46.29067405760405,
+      "learning_rate": 4.566085369687475e-06,
+      "loss": 0.4545,
+      "step": 13298
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 34.35245174586962,
+      "learning_rate": 4.564669749706663e-06,
+      "loss": 0.4071,
+      "step": 13299
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 38.46918495669854,
+      "learning_rate": 4.563254284303468e-06,
+      "loss": 0.4656,
+      "step": 13300
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 42.39809031403231,
+      "learning_rate": 4.56183897351814e-06,
+      "loss": 0.4904,
+      "step": 13301
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.34942220087454,
+      "learning_rate": 4.560423817390927e-06,
+      "loss": 0.5219,
+      "step": 13302
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 39.52827199341157,
+      "learning_rate": 4.559008815962078e-06,
+      "loss": 0.5503,
+      "step": 13303
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 45.589289441145105,
+      "learning_rate": 4.557593969271836e-06,
+      "loss": 0.5694,
+      "step": 13304
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 34.59308236295441,
+      "learning_rate": 4.556179277360442e-06,
+      "loss": 0.4198,
+      "step": 13305
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 43.0481167373595,
+      "learning_rate": 4.554764740268124e-06,
+      "loss": 0.5137,
+      "step": 13306
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 45.32360047748261,
+      "learning_rate": 4.553350358035108e-06,
+      "loss": 0.462,
+      "step": 13307
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 43.49457509644662,
+      "learning_rate": 4.551936130701622e-06,
+      "loss": 0.5247,
+      "step": 13308
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 41.12888836667794,
+      "learning_rate": 4.550522058307888e-06,
+      "loss": 0.4457,
+      "step": 13309
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 51.3608867328775,
+      "learning_rate": 4.549108140894117e-06,
+      "loss": 0.5859,
+      "step": 13310
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 37.242453514767135,
+      "learning_rate": 4.547694378500523e-06,
+      "loss": 0.4105,
+      "step": 13311
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 47.23318302771793,
+      "learning_rate": 4.54628077116731e-06,
+      "loss": 0.5513,
+      "step": 13312
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.04115828523155,
+      "learning_rate": 4.544867318934684e-06,
+      "loss": 0.4705,
+      "step": 13313
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 37.52069873492669,
+      "learning_rate": 4.543454021842836e-06,
+      "loss": 0.488,
+      "step": 13314
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 41.91191548361687,
+      "learning_rate": 4.5420408799319646e-06,
+      "loss": 0.5413,
+      "step": 13315
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 43.75328403953766,
+      "learning_rate": 4.540627893242261e-06,
+      "loss": 0.506,
+      "step": 13316
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.70031565747648,
+      "learning_rate": 4.539215061813903e-06,
+      "loss": 0.5337,
+      "step": 13317
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.71791394327599,
+      "learning_rate": 4.537802385687078e-06,
+      "loss": 0.5315,
+      "step": 13318
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 53.29793751735876,
+      "learning_rate": 4.5363898649019545e-06,
+      "loss": 0.5268,
+      "step": 13319
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.10341684313391,
+      "learning_rate": 4.534977499498706e-06,
+      "loss": 0.4759,
+      "step": 13320
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.53896370987039,
+      "learning_rate": 4.533565289517506e-06,
+      "loss": 0.5219,
+      "step": 13321
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 54.72846268616288,
+      "learning_rate": 4.5321532349985095e-06,
+      "loss": 0.5018,
+      "step": 13322
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 33.564097737619285,
+      "learning_rate": 4.530741335981874e-06,
+      "loss": 0.4193,
+      "step": 13323
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 36.10770183600584,
+      "learning_rate": 4.529329592507755e-06,
+      "loss": 0.4057,
+      "step": 13324
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.099251439754525,
+      "learning_rate": 4.527918004616305e-06,
+      "loss": 0.5337,
+      "step": 13325
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 38.9672063844524,
+      "learning_rate": 4.5265065723476675e-06,
+      "loss": 0.4789,
+      "step": 13326
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 37.915818028157915,
+      "learning_rate": 4.525095295741976e-06,
+      "loss": 0.452,
+      "step": 13327
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 36.268870417624925,
+      "learning_rate": 4.523684174839372e-06,
+      "loss": 0.4769,
+      "step": 13328
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.00916826031262,
+      "learning_rate": 4.5222732096799875e-06,
+      "loss": 0.5222,
+      "step": 13329
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 39.64206667373429,
+      "learning_rate": 4.520862400303951e-06,
+      "loss": 0.4804,
+      "step": 13330
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 37.451974694352145,
+      "learning_rate": 4.519451746751383e-06,
+      "loss": 0.4493,
+      "step": 13331
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 39.757429184417894,
+      "learning_rate": 4.518041249062398e-06,
+      "loss": 0.5107,
+      "step": 13332
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 38.049385523753195,
+      "learning_rate": 4.516630907277113e-06,
+      "loss": 0.4689,
+      "step": 13333
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 32.64536374080367,
+      "learning_rate": 4.5152207214356406e-06,
+      "loss": 0.4321,
+      "step": 13334
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 39.46580424608265,
+      "learning_rate": 4.513810691578079e-06,
+      "loss": 0.4954,
+      "step": 13335
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.715405599828784,
+      "learning_rate": 4.512400817744535e-06,
+      "loss": 0.5405,
+      "step": 13336
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 48.348424373583384,
+      "learning_rate": 4.510991099975098e-06,
+      "loss": 0.573,
+      "step": 13337
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 45.167090443292324,
+      "learning_rate": 4.509581538309867e-06,
+      "loss": 0.5708,
+      "step": 13338
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 49.74546599694094,
+      "learning_rate": 4.508172132788923e-06,
+      "loss": 0.5636,
+      "step": 13339
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.13214587307287,
+      "learning_rate": 4.506762883452349e-06,
+      "loss": 0.543,
+      "step": 13340
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 39.987723394797214,
+      "learning_rate": 4.50535379034023e-06,
+      "loss": 0.4869,
+      "step": 13341
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 36.64889484771517,
+      "learning_rate": 4.503944853492631e-06,
+      "loss": 0.4467,
+      "step": 13342
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 45.66958725713726,
+      "learning_rate": 4.502536072949628e-06,
+      "loss": 0.53,
+      "step": 13343
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.313738704741404,
+      "learning_rate": 4.5011274487512806e-06,
+      "loss": 0.5095,
+      "step": 13344
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 51.83836609498837,
+      "learning_rate": 4.4997189809376554e-06,
+      "loss": 0.5459,
+      "step": 13345
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 41.59547949791428,
+      "learning_rate": 4.498310669548801e-06,
+      "loss": 0.4435,
+      "step": 13346
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.93542345910129,
+      "learning_rate": 4.496902514624775e-06,
+      "loss": 0.4477,
+      "step": 13347
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 40.00589258778959,
+      "learning_rate": 4.495494516205621e-06,
+      "loss": 0.5241,
+      "step": 13348
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.92870406665723,
+      "learning_rate": 4.494086674331382e-06,
+      "loss": 0.4621,
+      "step": 13349
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 48.15910086093034,
+      "learning_rate": 4.492678989042099e-06,
+      "loss": 0.5239,
+      "step": 13350
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 33.75967809937784,
+      "learning_rate": 4.491271460377806e-06,
+      "loss": 0.395,
+      "step": 13351
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 37.61594949905416,
+      "learning_rate": 4.489864088378526e-06,
+      "loss": 0.4601,
+      "step": 13352
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 35.44175823133169,
+      "learning_rate": 4.488456873084288e-06,
+      "loss": 0.4341,
+      "step": 13353
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.24901493514681,
+      "learning_rate": 4.487049814535112e-06,
+      "loss": 0.4235,
+      "step": 13354
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 43.43744995882733,
+      "learning_rate": 4.4856429127710224e-06,
+      "loss": 0.4611,
+      "step": 13355
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 37.68142751859761,
+      "learning_rate": 4.484236167832015e-06,
+      "loss": 0.5026,
+      "step": 13356
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 37.111995538640976,
+      "learning_rate": 4.482829579758103e-06,
+      "loss": 0.4643,
+      "step": 13357
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 47.1501405068627,
+      "learning_rate": 4.481423148589292e-06,
+      "loss": 0.5055,
+      "step": 13358
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 42.03696471436961,
+      "learning_rate": 4.4800168743655814e-06,
+      "loss": 0.5193,
+      "step": 13359
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.563367249514904,
+      "learning_rate": 4.478610757126962e-06,
+      "loss": 0.5014,
+      "step": 13360
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 44.60533304425175,
+      "learning_rate": 4.477204796913419e-06,
+      "loss": 0.4404,
+      "step": 13361
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 44.37480009329456,
+      "learning_rate": 4.475798993764941e-06,
+      "loss": 0.5319,
+      "step": 13362
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.97128279734483,
+      "learning_rate": 4.474393347721511e-06,
+      "loss": 0.5091,
+      "step": 13363
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 52.570844017467735,
+      "learning_rate": 4.472987858823099e-06,
+      "loss": 0.514,
+      "step": 13364
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.887990462167146,
+      "learning_rate": 4.471582527109683e-06,
+      "loss": 0.533,
+      "step": 13365
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.45110863973452,
+      "learning_rate": 4.470177352621221e-06,
+      "loss": 0.5032,
+      "step": 13366
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.3243174855987,
+      "learning_rate": 4.468772335397681e-06,
+      "loss": 0.48,
+      "step": 13367
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 43.5005804384903,
+      "learning_rate": 4.467367475479023e-06,
+      "loss": 0.5813,
+      "step": 13368
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.99404065110616,
+      "learning_rate": 4.465962772905195e-06,
+      "loss": 0.5322,
+      "step": 13369
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.797248806477285,
+      "learning_rate": 4.464558227716152e-06,
+      "loss": 0.492,
+      "step": 13370
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 43.390756858924796,
+      "learning_rate": 4.463153839951829e-06,
+      "loss": 0.4596,
+      "step": 13371
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 53.874192437582735,
+      "learning_rate": 4.461749609652179e-06,
+      "loss": 0.5572,
+      "step": 13372
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.9365839341358,
+      "learning_rate": 4.460345536857124e-06,
+      "loss": 0.5082,
+      "step": 13373
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 38.709186878518665,
+      "learning_rate": 4.458941621606602e-06,
+      "loss": 0.5687,
+      "step": 13374
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 52.29717985348708,
+      "learning_rate": 4.457537863940543e-06,
+      "loss": 0.5448,
+      "step": 13375
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.0224212712403,
+      "learning_rate": 4.4561342638988645e-06,
+      "loss": 0.502,
+      "step": 13376
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.807011149293714,
+      "learning_rate": 4.4547308215214815e-06,
+      "loss": 0.5294,
+      "step": 13377
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.21327929301076,
+      "learning_rate": 4.453327536848309e-06,
+      "loss": 0.554,
+      "step": 13378
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.85285429700339,
+      "learning_rate": 4.451924409919257e-06,
+      "loss": 0.5051,
+      "step": 13379
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 36.19253162460384,
+      "learning_rate": 4.450521440774233e-06,
+      "loss": 0.4517,
+      "step": 13380
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 43.805950714669486,
+      "learning_rate": 4.449118629453133e-06,
+      "loss": 0.5033,
+      "step": 13381
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.58401759271594,
+      "learning_rate": 4.447715975995848e-06,
+      "loss": 0.497,
+      "step": 13382
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.319620075535646,
+      "learning_rate": 4.446313480442272e-06,
+      "loss": 0.4412,
+      "step": 13383
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.96208569104996,
+      "learning_rate": 4.444911142832297e-06,
+      "loss": 0.4543,
+      "step": 13384
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 39.47277774933657,
+      "learning_rate": 4.4435089632058e-06,
+      "loss": 0.5074,
+      "step": 13385
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 42.85995755580237,
+      "learning_rate": 4.442106941602652e-06,
+      "loss": 0.5112,
+      "step": 13386
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 35.66602580588053,
+      "learning_rate": 4.440705078062732e-06,
+      "loss": 0.4451,
+      "step": 13387
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 36.489246884485645,
+      "learning_rate": 4.4393033726259116e-06,
+      "loss": 0.5105,
+      "step": 13388
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.67001448723653,
+      "learning_rate": 4.437901825332046e-06,
+      "loss": 0.5328,
+      "step": 13389
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 39.328219908098504,
+      "learning_rate": 4.436500436221003e-06,
+      "loss": 0.4829,
+      "step": 13390
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 39.29775163950589,
+      "learning_rate": 4.4350992053326295e-06,
+      "loss": 0.4373,
+      "step": 13391
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.08657017873532,
+      "learning_rate": 4.433698132706779e-06,
+      "loss": 0.5028,
+      "step": 13392
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 42.66117757205682,
+      "learning_rate": 4.4322972183833e-06,
+      "loss": 0.508,
+      "step": 13393
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 39.795621216415405,
+      "learning_rate": 4.430896462402033e-06,
+      "loss": 0.4684,
+      "step": 13394
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.8334486243214,
+      "learning_rate": 4.429495864802808e-06,
+      "loss": 0.4813,
+      "step": 13395
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 52.52352697366801,
+      "learning_rate": 4.428095425625462e-06,
+      "loss": 0.5134,
+      "step": 13396
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 45.738478851477936,
+      "learning_rate": 4.426695144909826e-06,
+      "loss": 0.5127,
+      "step": 13397
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 38.518805614321735,
+      "learning_rate": 4.425295022695716e-06,
+      "loss": 0.4979,
+      "step": 13398
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 42.31383233952084,
+      "learning_rate": 4.423895059022959e-06,
+      "loss": 0.4421,
+      "step": 13399
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 35.1572521422923,
+      "learning_rate": 4.42249525393136e-06,
+      "loss": 0.4056,
+      "step": 13400
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.141564381596936,
+      "learning_rate": 4.421095607460738e-06,
+      "loss": 0.4289,
+      "step": 13401
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.028541166566484,
+      "learning_rate": 4.419696119650889e-06,
+      "loss": 0.5399,
+      "step": 13402
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 38.88898222087075,
+      "learning_rate": 4.418296790541618e-06,
+      "loss": 0.4386,
+      "step": 13403
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 37.839293597898276,
+      "learning_rate": 4.4168976201727255e-06,
+      "loss": 0.4571,
+      "step": 13404
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 39.83629742504851,
+      "learning_rate": 4.415498608583993e-06,
+      "loss": 0.4686,
+      "step": 13405
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 36.33145223495972,
+      "learning_rate": 4.414099755815219e-06,
+      "loss": 0.4194,
+      "step": 13406
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 48.983206922119884,
+      "learning_rate": 4.412701061906176e-06,
+      "loss": 0.5832,
+      "step": 13407
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 41.5136777100237,
+      "learning_rate": 4.411302526896646e-06,
+      "loss": 0.4499,
+      "step": 13408
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.94410143325677,
+      "learning_rate": 4.409904150826408e-06,
+      "loss": 0.4505,
+      "step": 13409
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 40.33540208820632,
+      "learning_rate": 4.408505933735223e-06,
+      "loss": 0.5519,
+      "step": 13410
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 50.6766553865289,
+      "learning_rate": 4.407107875662857e-06,
+      "loss": 0.569,
+      "step": 13411
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 46.743050976772544,
+      "learning_rate": 4.405709976649069e-06,
+      "loss": 0.532,
+      "step": 13412
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 38.01267769422962,
+      "learning_rate": 4.404312236733623e-06,
+      "loss": 0.4896,
+      "step": 13413
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 38.85931647111866,
+      "learning_rate": 4.402914655956262e-06,
+      "loss": 0.5053,
+      "step": 13414
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 36.641335899760655,
+      "learning_rate": 4.401517234356731e-06,
+      "loss": 0.4315,
+      "step": 13415
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.85206018678734,
+      "learning_rate": 4.400119971974774e-06,
+      "loss": 0.5094,
+      "step": 13416
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 37.54593576470148,
+      "learning_rate": 4.398722868850131e-06,
+      "loss": 0.4039,
+      "step": 13417
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 47.98390450428602,
+      "learning_rate": 4.3973259250225355e-06,
+      "loss": 0.566,
+      "step": 13418
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 37.11797739006228,
+      "learning_rate": 4.395929140531713e-06,
+      "loss": 0.4703,
+      "step": 13419
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.084171221987475,
+      "learning_rate": 4.3945325154173834e-06,
+      "loss": 0.4952,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 33.345873206484704,
+      "learning_rate": 4.39313604971927e-06,
+      "loss": 0.4304,
+      "step": 13421
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.59087539021172,
+      "learning_rate": 4.391739743477092e-06,
+      "loss": 0.445,
+      "step": 13422
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 56.79872033480953,
+      "learning_rate": 4.390343596730551e-06,
+      "loss": 0.6146,
+      "step": 13423
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.196145416616254,
+      "learning_rate": 4.38894760951936e-06,
+      "loss": 0.4821,
+      "step": 13424
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 34.47483797512473,
+      "learning_rate": 4.387551781883213e-06,
+      "loss": 0.4287,
+      "step": 13425
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 49.471404553228915,
+      "learning_rate": 4.386156113861814e-06,
+      "loss": 0.53,
+      "step": 13426
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.319987273306836,
+      "learning_rate": 4.384760605494847e-06,
+      "loss": 0.4775,
+      "step": 13427
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.7505076265604,
+      "learning_rate": 4.383365256822003e-06,
+      "loss": 0.4976,
+      "step": 13428
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.822403712164956,
+      "learning_rate": 4.3819700678829705e-06,
+      "loss": 0.4975,
+      "step": 13429
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 39.632443742711544,
+      "learning_rate": 4.380575038717419e-06,
+      "loss": 0.4372,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 46.53853855987597,
+      "learning_rate": 4.37918016936503e-06,
+      "loss": 0.5277,
+      "step": 13431
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.13887134562254,
+      "learning_rate": 4.377785459865463e-06,
+      "loss": 0.5034,
+      "step": 13432
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 35.431128458920526,
+      "learning_rate": 4.376390910258391e-06,
+      "loss": 0.4372,
+      "step": 13433
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 32.792916784648604,
+      "learning_rate": 4.374996520583474e-06,
+      "loss": 0.3784,
+      "step": 13434
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.55704460847894,
+      "learning_rate": 4.373602290880367e-06,
+      "loss": 0.5152,
+      "step": 13435
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 48.64420902173349,
+      "learning_rate": 4.372208221188715e-06,
+      "loss": 0.4662,
+      "step": 13436
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 33.70728834785591,
+      "learning_rate": 4.370814311548168e-06,
+      "loss": 0.4804,
+      "step": 13437
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 48.182135722502764,
+      "learning_rate": 4.3694205619983744e-06,
+      "loss": 0.5551,
+      "step": 13438
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 62.48238715911079,
+      "learning_rate": 4.368026972578965e-06,
+      "loss": 0.5865,
+      "step": 13439
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 51.74365086698614,
+      "learning_rate": 4.366633543329571e-06,
+      "loss": 0.5385,
+      "step": 13440
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 41.07809558517663,
+      "learning_rate": 4.365240274289824e-06,
+      "loss": 0.4778,
+      "step": 13441
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 38.75384695277153,
+      "learning_rate": 4.363847165499347e-06,
+      "loss": 0.4261,
+      "step": 13442
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.994151968947385,
+      "learning_rate": 4.362454216997761e-06,
+      "loss": 0.5054,
+      "step": 13443
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 51.142943518809986,
+      "learning_rate": 4.361061428824682e-06,
+      "loss": 0.5381,
+      "step": 13444
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 47.165862224287864,
+      "learning_rate": 4.359668801019713e-06,
+      "loss": 0.4778,
+      "step": 13445
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 39.238100641422626,
+      "learning_rate": 4.358276333622463e-06,
+      "loss": 0.5348,
+      "step": 13446
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 38.551779482548625,
+      "learning_rate": 4.356884026672537e-06,
+      "loss": 0.4928,
+      "step": 13447
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.371693446222615,
+      "learning_rate": 4.35549188020953e-06,
+      "loss": 0.5097,
+      "step": 13448
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.35530279204995,
+      "learning_rate": 4.354099894273027e-06,
+      "loss": 0.5156,
+      "step": 13449
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 36.075952192708165,
+      "learning_rate": 4.352708068902621e-06,
+      "loss": 0.5045,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.35303578378355,
+      "learning_rate": 4.351316404137898e-06,
+      "loss": 0.5185,
+      "step": 13451
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 51.9682176921694,
+      "learning_rate": 4.349924900018427e-06,
+      "loss": 0.4958,
+      "step": 13452
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.62342860483926,
+      "learning_rate": 4.3485335565837905e-06,
+      "loss": 0.4679,
+      "step": 13453
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 39.3909932338355,
+      "learning_rate": 4.347142373873551e-06,
+      "loss": 0.4886,
+      "step": 13454
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.43395502391381,
+      "learning_rate": 4.345751351927275e-06,
+      "loss": 0.4805,
+      "step": 13455
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 51.626624287321555,
+      "learning_rate": 4.344360490784526e-06,
+      "loss": 0.5535,
+      "step": 13456
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 41.592150496201945,
+      "learning_rate": 4.342969790484853e-06,
+      "loss": 0.5027,
+      "step": 13457
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.0872981048219,
+      "learning_rate": 4.341579251067815e-06,
+      "loss": 0.4472,
+      "step": 13458
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 48.97601645672306,
+      "learning_rate": 4.3401888725729465e-06,
+      "loss": 0.6336,
+      "step": 13459
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 46.92624207764711,
+      "learning_rate": 4.338798655039802e-06,
+      "loss": 0.525,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 39.49114968948295,
+      "learning_rate": 4.3374085985079075e-06,
+      "loss": 0.5162,
+      "step": 13461
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.182368720675655,
+      "learning_rate": 4.336018703016799e-06,
+      "loss": 0.5478,
+      "step": 13462
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 38.31034020439372,
+      "learning_rate": 4.33462896860601e-06,
+      "loss": 0.4919,
+      "step": 13463
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.12405558118678,
+      "learning_rate": 4.3332393953150574e-06,
+      "loss": 0.4974,
+      "step": 13464
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 40.23354499627424,
+      "learning_rate": 4.331849983183459e-06,
+      "loss": 0.5037,
+      "step": 13465
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.3763964459491,
+      "learning_rate": 4.330460732250732e-06,
+      "loss": 0.4659,
+      "step": 13466
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 46.59464581841766,
+      "learning_rate": 4.329071642556384e-06,
+      "loss": 0.5225,
+      "step": 13467
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 38.14493197979189,
+      "learning_rate": 4.327682714139925e-06,
+      "loss": 0.4464,
+      "step": 13468
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.191570368956356,
+      "learning_rate": 4.326293947040851e-06,
+      "loss": 0.399,
+      "step": 13469
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 41.13550863531823,
+      "learning_rate": 4.324905341298655e-06,
+      "loss": 0.5065,
+      "step": 13470
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.0183911503555,
+      "learning_rate": 4.323516896952832e-06,
+      "loss": 0.5456,
+      "step": 13471
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.01776681397501,
+      "learning_rate": 4.3221286140428695e-06,
+      "loss": 0.5002,
+      "step": 13472
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 41.47463786476976,
+      "learning_rate": 4.32074049260825e-06,
+      "loss": 0.5122,
+      "step": 13473
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 31.980409593482012,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.4854,
+      "step": 13474
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 58.039269640413366,
+      "learning_rate": 4.317964734322928e-06,
+      "loss": 0.5205,
+      "step": 13475
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.71351250421633,
+      "learning_rate": 4.316577097551176e-06,
+      "loss": 0.4969,
+      "step": 13476
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 46.656589108235686,
+      "learning_rate": 4.315189622412642e-06,
+      "loss": 0.4993,
+      "step": 13477
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 41.43065099946391,
+      "learning_rate": 4.313802308946794e-06,
+      "loss": 0.4203,
+      "step": 13478
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 43.07753119545432,
+      "learning_rate": 4.312415157193078e-06,
+      "loss": 0.4872,
+      "step": 13479
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.12917992151368,
+      "learning_rate": 4.3110281671909484e-06,
+      "loss": 0.435,
+      "step": 13480
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 46.021898219629264,
+      "learning_rate": 4.309641338979853e-06,
+      "loss": 0.4893,
+      "step": 13481
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 50.897383206295935,
+      "learning_rate": 4.308254672599225e-06,
+      "loss": 0.5133,
+      "step": 13482
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 36.62882524505605,
+      "learning_rate": 4.306868168088508e-06,
+      "loss": 0.4143,
+      "step": 13483
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 46.05886881373067,
+      "learning_rate": 4.305481825487128e-06,
+      "loss": 0.47,
+      "step": 13484
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 43.76667856012654,
+      "learning_rate": 4.304095644834516e-06,
+      "loss": 0.4837,
+      "step": 13485
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 36.04427604086372,
+      "learning_rate": 4.302709626170089e-06,
+      "loss": 0.4101,
+      "step": 13486
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 40.59408940972855,
+      "learning_rate": 4.30132376953327e-06,
+      "loss": 0.5822,
+      "step": 13487
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 43.97965496478234,
+      "learning_rate": 4.299938074963465e-06,
+      "loss": 0.5029,
+      "step": 13488
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 39.75317242396666,
+      "learning_rate": 4.298552542500093e-06,
+      "loss": 0.4936,
+      "step": 13489
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 43.49374963535901,
+      "learning_rate": 4.297167172182546e-06,
+      "loss": 0.556,
+      "step": 13490
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.83839462610613,
+      "learning_rate": 4.295781964050229e-06,
+      "loss": 0.4529,
+      "step": 13491
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 44.13320097749769,
+      "learning_rate": 4.2943969181425395e-06,
+      "loss": 0.5174,
+      "step": 13492
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 46.22830353192269,
+      "learning_rate": 4.2930120344988605e-06,
+      "loss": 0.5179,
+      "step": 13493
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 44.507354771471896,
+      "learning_rate": 4.291627313158584e-06,
+      "loss": 0.5413,
+      "step": 13494
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.955519656540595,
+      "learning_rate": 4.290242754161084e-06,
+      "loss": 0.4833,
+      "step": 13495
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 43.215715093516685,
+      "learning_rate": 4.28885835754574e-06,
+      "loss": 0.4642,
+      "step": 13496
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 35.78243102077749,
+      "learning_rate": 4.287474123351928e-06,
+      "loss": 0.4669,
+      "step": 13497
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 45.89540456028133,
+      "learning_rate": 4.286090051619008e-06,
+      "loss": 0.5168,
+      "step": 13498
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 40.08020793119905,
+      "learning_rate": 4.284706142386342e-06,
+      "loss": 0.4675,
+      "step": 13499
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 36.89470521109398,
+      "learning_rate": 4.2833223956932916e-06,
+      "loss": 0.4786,
+      "step": 13500
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 45.9992002801209,
+      "learning_rate": 4.2819388115792095e-06,
+      "loss": 0.6161,
+      "step": 13501
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.78743730362699,
+      "learning_rate": 4.280555390083443e-06,
+      "loss": 0.4477,
+      "step": 13502
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 60.33946453821634,
+      "learning_rate": 4.279172131245332e-06,
+      "loss": 0.4911,
+      "step": 13503
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 41.685341809182184,
+      "learning_rate": 4.27778903510422e-06,
+      "loss": 0.4549,
+      "step": 13504
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.72110717222045,
+      "learning_rate": 4.2764061016994386e-06,
+      "loss": 0.4635,
+      "step": 13505
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.83321893926601,
+      "learning_rate": 4.275023331070324e-06,
+      "loss": 0.4762,
+      "step": 13506
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 36.690894887750495,
+      "learning_rate": 4.2736407232561975e-06,
+      "loss": 0.435,
+      "step": 13507
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 45.79073693208337,
+      "learning_rate": 4.272258278296374e-06,
+      "loss": 0.4391,
+      "step": 13508
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.06206405096314,
+      "learning_rate": 4.270875996230176e-06,
+      "loss": 0.4301,
+      "step": 13509
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 53.78794655123588,
+      "learning_rate": 4.269493877096915e-06,
+      "loss": 0.5043,
+      "step": 13510
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 35.054603564657334,
+      "learning_rate": 4.268111920935893e-06,
+      "loss": 0.4988,
+      "step": 13511
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 60.90358189649213,
+      "learning_rate": 4.266730127786419e-06,
+      "loss": 0.5495,
+      "step": 13512
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 49.18490271151593,
+      "learning_rate": 4.265348497687784e-06,
+      "loss": 0.4959,
+      "step": 13513
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.46237171916109,
+      "learning_rate": 4.263967030679284e-06,
+      "loss": 0.4237,
+      "step": 13514
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.9742219286448,
+      "learning_rate": 4.262585726800204e-06,
+      "loss": 0.5131,
+      "step": 13515
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.870295758652,
+      "learning_rate": 4.26120458608983e-06,
+      "loss": 0.5096,
+      "step": 13516
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 39.39692523814209,
+      "learning_rate": 4.259823608587443e-06,
+      "loss": 0.5071,
+      "step": 13517
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.392091834085846,
+      "learning_rate": 4.258442794332311e-06,
+      "loss": 0.5314,
+      "step": 13518
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 40.485411995460026,
+      "learning_rate": 4.2570621433637116e-06,
+      "loss": 0.3948,
+      "step": 13519
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 40.00595971055651,
+      "learning_rate": 4.255681655720901e-06,
+      "loss": 0.4541,
+      "step": 13520
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.564630261857246,
+      "learning_rate": 4.254301331443145e-06,
+      "loss": 0.517,
+      "step": 13521
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 46.436603388595444,
+      "learning_rate": 4.252921170569701e-06,
+      "loss": 0.5864,
+      "step": 13522
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 30.99982294678118,
+      "learning_rate": 4.251541173139816e-06,
+      "loss": 0.4313,
+      "step": 13523
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 35.75531825942192,
+      "learning_rate": 4.250161339192734e-06,
+      "loss": 0.517,
+      "step": 13524
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.78246389802688,
+      "learning_rate": 4.2487816687677006e-06,
+      "loss": 0.4755,
+      "step": 13525
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.51230424721657,
+      "learning_rate": 4.247402161903956e-06,
+      "loss": 0.4995,
+      "step": 13526
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.09850291242913,
+      "learning_rate": 4.246022818640728e-06,
+      "loss": 0.4931,
+      "step": 13527
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 40.855396440286796,
+      "learning_rate": 4.244643639017242e-06,
+      "loss": 0.5189,
+      "step": 13528
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.582055437854336,
+      "learning_rate": 4.243264623072725e-06,
+      "loss": 0.4817,
+      "step": 13529
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.58433819517885,
+      "learning_rate": 4.241885770846392e-06,
+      "loss": 0.4589,
+      "step": 13530
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 36.765006161080564,
+      "learning_rate": 4.240507082377465e-06,
+      "loss": 0.4882,
+      "step": 13531
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 51.497010998419285,
+      "learning_rate": 4.239128557705146e-06,
+      "loss": 0.5804,
+      "step": 13532
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.83520882042866,
+      "learning_rate": 4.237750196868639e-06,
+      "loss": 0.479,
+      "step": 13533
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.89241459592341,
+      "learning_rate": 4.236371999907144e-06,
+      "loss": 0.431,
+      "step": 13534
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.448806155505835,
+      "learning_rate": 4.234993966859862e-06,
+      "loss": 0.4997,
+      "step": 13535
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 40.45471605551325,
+      "learning_rate": 4.233616097765979e-06,
+      "loss": 0.4733,
+      "step": 13536
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 33.123057765521395,
+      "learning_rate": 4.232238392664677e-06,
+      "loss": 0.4143,
+      "step": 13537
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 34.86374022451513,
+      "learning_rate": 4.2308608515951414e-06,
+      "loss": 0.4694,
+      "step": 13538
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 42.81874096900534,
+      "learning_rate": 4.229483474596552e-06,
+      "loss": 0.4844,
+      "step": 13539
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.888342006032076,
+      "learning_rate": 4.228106261708073e-06,
+      "loss": 0.4487,
+      "step": 13540
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 40.64347988310501,
+      "learning_rate": 4.22672921296888e-06,
+      "loss": 0.5148,
+      "step": 13541
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 42.378494459006866,
+      "learning_rate": 4.225352328418126e-06,
+      "loss": 0.4623,
+      "step": 13542
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 49.015648240827815,
+      "learning_rate": 4.223975608094975e-06,
+      "loss": 0.5451,
+      "step": 13543
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 42.93662660868208,
+      "learning_rate": 4.222599052038581e-06,
+      "loss": 0.4779,
+      "step": 13544
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.52538545953841,
+      "learning_rate": 4.221222660288087e-06,
+      "loss": 0.4712,
+      "step": 13545
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 48.585398544509445,
+      "learning_rate": 4.219846432882644e-06,
+      "loss": 0.561,
+      "step": 13546
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 40.5753114567318,
+      "learning_rate": 4.218470369861384e-06,
+      "loss": 0.535,
+      "step": 13547
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 41.66608489139031,
+      "learning_rate": 4.217094471263447e-06,
+      "loss": 0.4918,
+      "step": 13548
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 38.69089024838516,
+      "learning_rate": 4.2157187371279585e-06,
+      "loss": 0.4862,
+      "step": 13549
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.10905195262546,
+      "learning_rate": 4.214343167494044e-06,
+      "loss": 0.4627,
+      "step": 13550
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 48.59558574484029,
+      "learning_rate": 4.2129677624008294e-06,
+      "loss": 0.5523,
+      "step": 13551
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 40.26967691352055,
+      "learning_rate": 4.211592521887426e-06,
+      "loss": 0.4875,
+      "step": 13552
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 37.402584431722964,
+      "learning_rate": 4.2102174459929424e-06,
+      "loss": 0.4924,
+      "step": 13553
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 48.677275733668196,
+      "learning_rate": 4.2088425347564866e-06,
+      "loss": 0.4736,
+      "step": 13554
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 40.77875492458058,
+      "learning_rate": 4.207467788217162e-06,
+      "loss": 0.4997,
+      "step": 13555
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 42.51684639216207,
+      "learning_rate": 4.206093206414069e-06,
+      "loss": 0.4965,
+      "step": 13556
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 38.70592369460693,
+      "learning_rate": 4.204718789386295e-06,
+      "loss": 0.4788,
+      "step": 13557
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 50.6963281481231,
+      "learning_rate": 4.203344537172925e-06,
+      "loss": 0.5264,
+      "step": 13558
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.74874008544384,
+      "learning_rate": 4.201970449813045e-06,
+      "loss": 0.4497,
+      "step": 13559
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.28744110353297,
+      "learning_rate": 4.200596527345738e-06,
+      "loss": 0.4892,
+      "step": 13560
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 51.171266061202715,
+      "learning_rate": 4.199222769810074e-06,
+      "loss": 0.6443,
+      "step": 13561
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.10002315193138,
+      "learning_rate": 4.197849177245117e-06,
+      "loss": 0.4706,
+      "step": 13562
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 47.44113566924915,
+      "learning_rate": 4.196475749689934e-06,
+      "loss": 0.4924,
+      "step": 13563
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 45.15909184419961,
+      "learning_rate": 4.19510248718359e-06,
+      "loss": 0.4558,
+      "step": 13564
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.556584879183525,
+      "learning_rate": 4.1937293897651324e-06,
+      "loss": 0.4521,
+      "step": 13565
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 51.646159433266575,
+      "learning_rate": 4.192356457473617e-06,
+      "loss": 0.5425,
+      "step": 13566
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.108390070114,
+      "learning_rate": 4.190983690348084e-06,
+      "loss": 0.4351,
+      "step": 13567
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 42.814616297185616,
+      "learning_rate": 4.189611088427576e-06,
+      "loss": 0.465,
+      "step": 13568
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.87913321999492,
+      "learning_rate": 4.188238651751134e-06,
+      "loss": 0.49,
+      "step": 13569
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 38.52965615372303,
+      "learning_rate": 4.186866380357782e-06,
+      "loss": 0.5292,
+      "step": 13570
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.09598209563031,
+      "learning_rate": 4.1854942742865514e-06,
+      "loss": 0.5873,
+      "step": 13571
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 34.270212300402434,
+      "learning_rate": 4.184122333576459e-06,
+      "loss": 0.4563,
+      "step": 13572
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.11559979519679,
+      "learning_rate": 4.1827505582665295e-06,
+      "loss": 0.476,
+      "step": 13573
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.34820353339633,
+      "learning_rate": 4.181378948395767e-06,
+      "loss": 0.422,
+      "step": 13574
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 45.899524998544486,
+      "learning_rate": 4.180007504003183e-06,
+      "loss": 0.5487,
+      "step": 13575
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.10584691120104,
+      "learning_rate": 4.178636225127786e-06,
+      "loss": 0.4396,
+      "step": 13576
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 45.777615191321125,
+      "learning_rate": 4.177265111808568e-06,
+      "loss": 0.5126,
+      "step": 13577
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.748277396746985,
+      "learning_rate": 4.175894164084521e-06,
+      "loss": 0.447,
+      "step": 13578
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 45.122841833606685,
+      "learning_rate": 4.174523381994637e-06,
+      "loss": 0.5378,
+      "step": 13579
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 40.37602542307272,
+      "learning_rate": 4.1731527655779e-06,
+      "loss": 0.4625,
+      "step": 13580
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 41.83243631766484,
+      "learning_rate": 4.171782314873294e-06,
+      "loss": 0.4579,
+      "step": 13581
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 40.71018187400505,
+      "learning_rate": 4.170412029919788e-06,
+      "loss": 0.4868,
+      "step": 13582
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 42.50573516183913,
+      "learning_rate": 4.169041910756352e-06,
+      "loss": 0.4883,
+      "step": 13583
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 33.711191240507205,
+      "learning_rate": 4.167671957421952e-06,
+      "loss": 0.4437,
+      "step": 13584
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.93315066309339,
+      "learning_rate": 4.166302169955553e-06,
+      "loss": 0.4262,
+      "step": 13585
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.99116408970357,
+      "learning_rate": 4.164932548396108e-06,
+      "loss": 0.4583,
+      "step": 13586
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 39.107914634076494,
+      "learning_rate": 4.163563092782564e-06,
+      "loss": 0.4779,
+      "step": 13587
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 39.33223289159979,
+      "learning_rate": 4.162193803153872e-06,
+      "loss": 0.4884,
+      "step": 13588
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.806936705044414,
+      "learning_rate": 4.160824679548977e-06,
+      "loss": 0.4789,
+      "step": 13589
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 35.48542788113878,
+      "learning_rate": 4.1594557220068125e-06,
+      "loss": 0.4385,
+      "step": 13590
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 42.98632719875426,
+      "learning_rate": 4.158086930566307e-06,
+      "loss": 0.4541,
+      "step": 13591
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 53.41328689170535,
+      "learning_rate": 4.15671830526639e-06,
+      "loss": 0.5332,
+      "step": 13592
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 38.66017990365972,
+      "learning_rate": 4.155349846145988e-06,
+      "loss": 0.4453,
+      "step": 13593
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.50976385733999,
+      "learning_rate": 4.153981553244021e-06,
+      "loss": 0.4879,
+      "step": 13594
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 46.91453683376725,
+      "learning_rate": 4.152613426599398e-06,
+      "loss": 0.545,
+      "step": 13595
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 39.88978079990705,
+      "learning_rate": 4.151245466251025e-06,
+      "loss": 0.4411,
+      "step": 13596
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 41.201312630307385,
+      "learning_rate": 4.14987767223781e-06,
+      "loss": 0.4761,
+      "step": 13597
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.70382955109584,
+      "learning_rate": 4.148510044598655e-06,
+      "loss": 0.4123,
+      "step": 13598
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 50.56629974837302,
+      "learning_rate": 4.147142583372448e-06,
+      "loss": 0.4532,
+      "step": 13599
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 39.81654497854278,
+      "learning_rate": 4.145775288598085e-06,
+      "loss": 0.4854,
+      "step": 13600
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 34.154074230505685,
+      "learning_rate": 4.1444081603144445e-06,
+      "loss": 0.3803,
+      "step": 13601
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 47.870997428408046,
+      "learning_rate": 4.1430411985604145e-06,
+      "loss": 0.4605,
+      "step": 13602
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 43.6263963098784,
+      "learning_rate": 4.141674403374864e-06,
+      "loss": 0.4829,
+      "step": 13603
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 37.80567213472294,
+      "learning_rate": 4.1403077747966646e-06,
+      "loss": 0.4861,
+      "step": 13604
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.20452586656949,
+      "learning_rate": 4.13894131286469e-06,
+      "loss": 0.551,
+      "step": 13605
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 37.50395721854835,
+      "learning_rate": 4.13757501761779e-06,
+      "loss": 0.4149,
+      "step": 13606
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 53.26310997528399,
+      "learning_rate": 4.136208889094832e-06,
+      "loss": 0.6341,
+      "step": 13607
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.28117917051613,
+      "learning_rate": 4.1348429273346595e-06,
+      "loss": 0.518,
+      "step": 13608
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 35.01410718400304,
+      "learning_rate": 4.133477132376122e-06,
+      "loss": 0.3984,
+      "step": 13609
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 40.3353597779529,
+      "learning_rate": 4.132111504258067e-06,
+      "loss": 0.4558,
+      "step": 13610
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 39.07994537201326,
+      "learning_rate": 4.130746043019329e-06,
+      "loss": 0.4674,
+      "step": 13611
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 34.29218100922602,
+      "learning_rate": 4.129380748698737e-06,
+      "loss": 0.4085,
+      "step": 13612
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.91054227218307,
+      "learning_rate": 4.128015621335121e-06,
+      "loss": 0.4617,
+      "step": 13613
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 54.62862419571141,
+      "learning_rate": 4.12665066096731e-06,
+      "loss": 0.5714,
+      "step": 13614
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 44.95653840790135,
+      "learning_rate": 4.1252858676341175e-06,
+      "loss": 0.4814,
+      "step": 13615
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 41.73810899169677,
+      "learning_rate": 4.123921241374355e-06,
+      "loss": 0.5363,
+      "step": 13616
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.215323212339676,
+      "learning_rate": 4.1225567822268365e-06,
+      "loss": 0.5087,
+      "step": 13617
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 33.07197404053799,
+      "learning_rate": 4.121192490230363e-06,
+      "loss": 0.3978,
+      "step": 13618
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 46.913352647123,
+      "learning_rate": 4.11982836542374e-06,
+      "loss": 0.5688,
+      "step": 13619
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.08881491549031,
+      "learning_rate": 4.118464407845759e-06,
+      "loss": 0.5257,
+      "step": 13620
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 39.00986651791107,
+      "learning_rate": 4.117100617535207e-06,
+      "loss": 0.4503,
+      "step": 13621
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 32.93520489827157,
+      "learning_rate": 4.11573699453087e-06,
+      "loss": 0.4229,
+      "step": 13622
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 36.49360230957083,
+      "learning_rate": 4.114373538871535e-06,
+      "loss": 0.4618,
+      "step": 13623
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 41.43312908727359,
+      "learning_rate": 4.1130102505959715e-06,
+      "loss": 0.4972,
+      "step": 13624
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 41.03630338696782,
+      "learning_rate": 4.111647129742954e-06,
+      "loss": 0.4834,
+      "step": 13625
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 45.55722186961732,
+      "learning_rate": 4.110284176351245e-06,
+      "loss": 0.4401,
+      "step": 13626
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 44.52116750515288,
+      "learning_rate": 4.108921390459612e-06,
+      "loss": 0.48,
+      "step": 13627
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 47.99913174618796,
+      "learning_rate": 4.107558772106805e-06,
+      "loss": 0.5414,
+      "step": 13628
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 42.984070200309176,
+      "learning_rate": 4.106196321331581e-06,
+      "loss": 0.4927,
+      "step": 13629
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 44.29959465168627,
+      "learning_rate": 4.104834038172687e-06,
+      "loss": 0.473,
+      "step": 13630
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 55.07513645586381,
+      "learning_rate": 4.103471922668862e-06,
+      "loss": 0.5435,
+      "step": 13631
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 45.582239492962124,
+      "learning_rate": 4.10210997485885e-06,
+      "loss": 0.5161,
+      "step": 13632
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 38.64940733547737,
+      "learning_rate": 4.100748194781376e-06,
+      "loss": 0.4658,
+      "step": 13633
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 36.13934185027809,
+      "learning_rate": 4.099386582475175e-06,
+      "loss": 0.4344,
+      "step": 13634
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 40.92633972631635,
+      "learning_rate": 4.0980251379789655e-06,
+      "loss": 0.4736,
+      "step": 13635
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 36.921509575054515,
+      "learning_rate": 4.096663861331472e-06,
+      "loss": 0.4502,
+      "step": 13636
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 36.783136084465156,
+      "learning_rate": 4.095302752571402e-06,
+      "loss": 0.5113,
+      "step": 13637
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 38.03861760451031,
+      "learning_rate": 4.093941811737466e-06,
+      "loss": 0.4927,
+      "step": 13638
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 38.83223795607951,
+      "learning_rate": 4.092581038868375e-06,
+      "loss": 0.4543,
+      "step": 13639
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 87.88221951994308,
+      "learning_rate": 4.0912204340028224e-06,
+      "loss": 0.4907,
+      "step": 13640
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 45.66789575417101,
+      "learning_rate": 4.089859997179502e-06,
+      "loss": 0.5981,
+      "step": 13641
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 42.255125474121726,
+      "learning_rate": 4.088499728437104e-06,
+      "loss": 0.5021,
+      "step": 13642
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.0506675586492,
+      "learning_rate": 4.087139627814317e-06,
+      "loss": 0.4704,
+      "step": 13643
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 46.696088437568974,
+      "learning_rate": 4.085779695349823e-06,
+      "loss": 0.5187,
+      "step": 13644
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 45.8622303790248,
+      "learning_rate": 4.084419931082295e-06,
+      "loss": 0.5324,
+      "step": 13645
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 42.2382468376277,
+      "learning_rate": 4.083060335050401e-06,
+      "loss": 0.4929,
+      "step": 13646
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 52.235680653112155,
+      "learning_rate": 4.0817009072928085e-06,
+      "loss": 0.5386,
+      "step": 13647
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 51.07629099972724,
+      "learning_rate": 4.080341647848184e-06,
+      "loss": 0.569,
+      "step": 13648
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 33.450646060497846,
+      "learning_rate": 4.07898255675518e-06,
+      "loss": 0.398,
+      "step": 13649
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 35.16372309963579,
+      "learning_rate": 4.077623634052445e-06,
+      "loss": 0.5128,
+      "step": 13650
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 40.6136497193925,
+      "learning_rate": 4.076264879778629e-06,
+      "loss": 0.5223,
+      "step": 13651
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 42.43637248916475,
+      "learning_rate": 4.074906293972378e-06,
+      "loss": 0.4314,
+      "step": 13652
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 38.87441225619518,
+      "learning_rate": 4.073547876672323e-06,
+      "loss": 0.4855,
+      "step": 13653
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 33.873004636379115,
+      "learning_rate": 4.072189627917102e-06,
+      "loss": 0.4368,
+      "step": 13654
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 39.72548213903825,
+      "learning_rate": 4.070831547745338e-06,
+      "loss": 0.4222,
+      "step": 13655
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 48.606362951205575,
+      "learning_rate": 4.0694736361956555e-06,
+      "loss": 0.5605,
+      "step": 13656
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 38.727986038514,
+      "learning_rate": 4.068115893306678e-06,
+      "loss": 0.5159,
+      "step": 13657
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 42.38038869932468,
+      "learning_rate": 4.06675831911701e-06,
+      "loss": 0.5595,
+      "step": 13658
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 44.33349498322542,
+      "learning_rate": 4.0654009136652685e-06,
+      "loss": 0.5978,
+      "step": 13659
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 46.2530880077568,
+      "learning_rate": 4.06404367699005e-06,
+      "loss": 0.5049,
+      "step": 13660
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 40.58360497359609,
+      "learning_rate": 4.0626866091299595e-06,
+      "loss": 0.508,
+      "step": 13661
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 38.805879037380016,
+      "learning_rate": 4.061329710123586e-06,
+      "loss": 0.4522,
+      "step": 13662
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 52.38703414211063,
+      "learning_rate": 4.059972980009522e-06,
+      "loss": 0.5106,
+      "step": 13663
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 39.30435792879201,
+      "learning_rate": 4.058616418826355e-06,
+      "loss": 0.4564,
+      "step": 13664
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 43.3760846301959,
+      "learning_rate": 4.05726002661266e-06,
+      "loss": 0.4792,
+      "step": 13665
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 33.01760434429185,
+      "learning_rate": 4.055903803407011e-06,
+      "loss": 0.3921,
+      "step": 13666
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 37.83519220021802,
+      "learning_rate": 4.05454774924798e-06,
+      "loss": 0.4579,
+      "step": 13667
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 41.24215385492003,
+      "learning_rate": 4.0531918641741344e-06,
+      "loss": 0.4963,
+      "step": 13668
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 46.934698795893524,
+      "learning_rate": 4.051836148224035e-06,
+      "loss": 0.5732,
+      "step": 13669
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.29020739798455,
+      "learning_rate": 4.050480601436237e-06,
+      "loss": 0.44,
+      "step": 13670
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 50.03269414110105,
+      "learning_rate": 4.049125223849287e-06,
+      "loss": 0.5688,
+      "step": 13671
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 53.6922127561474,
+      "learning_rate": 4.047770015501734e-06,
+      "loss": 0.4861,
+      "step": 13672
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 32.53759504400346,
+      "learning_rate": 4.046414976432124e-06,
+      "loss": 0.4095,
+      "step": 13673
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.39221567075612,
+      "learning_rate": 4.045060106678989e-06,
+      "loss": 0.4179,
+      "step": 13674
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 44.43048982548135,
+      "learning_rate": 4.043705406280857e-06,
+      "loss": 0.5481,
+      "step": 13675
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 41.017941829096465,
+      "learning_rate": 4.04235087527626e-06,
+      "loss": 0.4923,
+      "step": 13676
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 39.97060849877328,
+      "learning_rate": 4.040996513703721e-06,
+      "loss": 0.4733,
+      "step": 13677
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 46.99467582678495,
+      "learning_rate": 4.039642321601753e-06,
+      "loss": 0.532,
+      "step": 13678
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 62.50501611798338,
+      "learning_rate": 4.0382882990088735e-06,
+      "loss": 0.4343,
+      "step": 13679
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 47.45420753266675,
+      "learning_rate": 4.036934445963584e-06,
+      "loss": 0.5369,
+      "step": 13680
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 32.72456677239678,
+      "learning_rate": 4.035580762504391e-06,
+      "loss": 0.4193,
+      "step": 13681
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 39.36862213918244,
+      "learning_rate": 4.034227248669794e-06,
+      "loss": 0.4719,
+      "step": 13682
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 48.10648671543343,
+      "learning_rate": 4.032873904498286e-06,
+      "loss": 0.5215,
+      "step": 13683
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 35.88344550991523,
+      "learning_rate": 4.031520730028348e-06,
+      "loss": 0.4648,
+      "step": 13684
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 38.033225519909,
+      "learning_rate": 4.030167725298472e-06,
+      "loss": 0.5256,
+      "step": 13685
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 40.000737117889685,
+      "learning_rate": 4.028814890347134e-06,
+      "loss": 0.4652,
+      "step": 13686
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 27.424976149612114,
+      "learning_rate": 4.027462225212806e-06,
+      "loss": 0.4095,
+      "step": 13687
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.17053315558683,
+      "learning_rate": 4.026109729933962e-06,
+      "loss": 0.4638,
+      "step": 13688
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 41.08652499834825,
+      "learning_rate": 4.024757404549058e-06,
+      "loss": 0.4545,
+      "step": 13689
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 42.54862082629851,
+      "learning_rate": 4.023405249096561e-06,
+      "loss": 0.4081,
+      "step": 13690
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 41.684131017208706,
+      "learning_rate": 4.022053263614921e-06,
+      "loss": 0.4582,
+      "step": 13691
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 56.43968581722523,
+      "learning_rate": 4.0207014481425875e-06,
+      "loss": 0.4138,
+      "step": 13692
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 49.58935237008542,
+      "learning_rate": 4.019349802718011e-06,
+      "loss": 0.5236,
+      "step": 13693
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 39.653831152648785,
+      "learning_rate": 4.017998327379626e-06,
+      "loss": 0.5107,
+      "step": 13694
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 44.11675040526898,
+      "learning_rate": 4.01664702216587e-06,
+      "loss": 0.5328,
+      "step": 13695
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 42.02726601363792,
+      "learning_rate": 4.015295887115169e-06,
+      "loss": 0.4996,
+      "step": 13696
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 38.264534657070186,
+      "learning_rate": 4.013944922265953e-06,
+      "loss": 0.4221,
+      "step": 13697
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 49.02782841566651,
+      "learning_rate": 4.012594127656646e-06,
+      "loss": 0.5431,
+      "step": 13698
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 41.46600275778224,
+      "learning_rate": 4.011243503325658e-06,
+      "loss": 0.5028,
+      "step": 13699
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 45.6989385164386,
+      "learning_rate": 4.009893049311399e-06,
+      "loss": 0.5571,
+      "step": 13700
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 33.985022569822576,
+      "learning_rate": 4.008542765652277e-06,
+      "loss": 0.4025,
+      "step": 13701
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 42.671938890194916,
+      "learning_rate": 4.007192652386695e-06,
+      "loss": 0.5563,
+      "step": 13702
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 44.04118163684743,
+      "learning_rate": 4.005842709553053e-06,
+      "loss": 0.4684,
+      "step": 13703
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.02872381503587,
+      "learning_rate": 4.0044929371897325e-06,
+      "loss": 0.51,
+      "step": 13704
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 48.88409198283275,
+      "learning_rate": 4.003143335335124e-06,
+      "loss": 0.4638,
+      "step": 13705
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 39.253228449618724,
+      "learning_rate": 4.0017939040276136e-06,
+      "loss": 0.4849,
+      "step": 13706
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 34.626688341093086,
+      "learning_rate": 4.000444643305577e-06,
+      "loss": 0.4721,
+      "step": 13707
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 37.82354246585035,
+      "learning_rate": 3.999095553207386e-06,
+      "loss": 0.4409,
+      "step": 13708
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 42.921095800391925,
+      "learning_rate": 3.997746633771403e-06,
+      "loss": 0.4765,
+      "step": 13709
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 38.900398583441756,
+      "learning_rate": 3.996397885035995e-06,
+      "loss": 0.437,
+      "step": 13710
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 41.234981279444085,
+      "learning_rate": 3.9950493070395235e-06,
+      "loss": 0.4397,
+      "step": 13711
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.186554546468784,
+      "learning_rate": 3.993700899820332e-06,
+      "loss": 0.4697,
+      "step": 13712
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.469240845377804,
+      "learning_rate": 3.9923526634167775e-06,
+      "loss": 0.4222,
+      "step": 13713
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.31695308334977,
+      "learning_rate": 3.991004597867195e-06,
+      "loss": 0.4678,
+      "step": 13714
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 45.478477554354505,
+      "learning_rate": 3.989656703209931e-06,
+      "loss": 0.5527,
+      "step": 13715
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 45.22178448590305,
+      "learning_rate": 3.9883089794833094e-06,
+      "loss": 0.4781,
+      "step": 13716
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 46.9797352972811,
+      "learning_rate": 3.986961426725665e-06,
+      "loss": 0.5325,
+      "step": 13717
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 40.94709353839532,
+      "learning_rate": 3.985614044975323e-06,
+      "loss": 0.5076,
+      "step": 13718
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 42.003179747695725,
+      "learning_rate": 3.9842668342705956e-06,
+      "loss": 0.476,
+      "step": 13719
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.1997233426101,
+      "learning_rate": 3.982919794649805e-06,
+      "loss": 0.5008,
+      "step": 13720
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 37.16967451678657,
+      "learning_rate": 3.98157292615125e-06,
+      "loss": 0.4036,
+      "step": 13721
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 35.428630694491396,
+      "learning_rate": 3.980226228813242e-06,
+      "loss": 0.4603,
+      "step": 13722
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 40.20939625865267,
+      "learning_rate": 3.978879702674081e-06,
+      "loss": 0.456,
+      "step": 13723
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 38.68076606653636,
+      "learning_rate": 3.97753334777206e-06,
+      "loss": 0.4523,
+      "step": 13724
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 44.07411447675143,
+      "learning_rate": 3.976187164145463e-06,
+      "loss": 0.4661,
+      "step": 13725
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 44.30487802054538,
+      "learning_rate": 3.97484115183258e-06,
+      "loss": 0.4946,
+      "step": 13726
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 29.330705560670587,
+      "learning_rate": 3.97349531087169e-06,
+      "loss": 0.4197,
+      "step": 13727
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 47.742921399415465,
+      "learning_rate": 3.972149641301075e-06,
+      "loss": 0.5214,
+      "step": 13728
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 36.096134708116196,
+      "learning_rate": 3.97080414315899e-06,
+      "loss": 0.4668,
+      "step": 13729
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 32.08853111066719,
+      "learning_rate": 3.969458816483709e-06,
+      "loss": 0.4197,
+      "step": 13730
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 45.503017430282796,
+      "learning_rate": 3.968113661313492e-06,
+      "loss": 0.5021,
+      "step": 13731
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 45.82769648250475,
+      "learning_rate": 3.966768677686597e-06,
+      "loss": 0.4848,
+      "step": 13732
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.2075838221399,
+      "learning_rate": 3.9654238656412715e-06,
+      "loss": 0.5722,
+      "step": 13733
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.988745429827,
+      "learning_rate": 3.9640792252157564e-06,
+      "loss": 0.4432,
+      "step": 13734
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 33.98835573230996,
+      "learning_rate": 3.962734756448299e-06,
+      "loss": 0.42,
+      "step": 13735
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 50.78019328265011,
+      "learning_rate": 3.961390459377136e-06,
+      "loss": 0.5855,
+      "step": 13736
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.664715418057064,
+      "learning_rate": 3.960046334040496e-06,
+      "loss": 0.4013,
+      "step": 13737
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 46.84849106486872,
+      "learning_rate": 3.958702380476602e-06,
+      "loss": 0.5153,
+      "step": 13738
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.10149227383523,
+      "learning_rate": 3.957358598723677e-06,
+      "loss": 0.4812,
+      "step": 13739
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.950282148655894,
+      "learning_rate": 3.956014988819944e-06,
+      "loss": 0.5237,
+      "step": 13740
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.59968654459888,
+      "learning_rate": 3.954671550803604e-06,
+      "loss": 0.4347,
+      "step": 13741
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.85139510587848,
+      "learning_rate": 3.953328284712873e-06,
+      "loss": 0.5655,
+      "step": 13742
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 50.09100580819766,
+      "learning_rate": 3.951985190585944e-06,
+      "loss": 0.5489,
+      "step": 13743
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 46.71612886456559,
+      "learning_rate": 3.950642268461019e-06,
+      "loss": 0.4966,
+      "step": 13744
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 64.42464614448284,
+      "learning_rate": 3.9492995183762925e-06,
+      "loss": 0.58,
+      "step": 13745
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.1044112010618,
+      "learning_rate": 3.947956940369944e-06,
+      "loss": 0.49,
+      "step": 13746
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.83944047229789,
+      "learning_rate": 3.946614534480164e-06,
+      "loss": 0.4981,
+      "step": 13747
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 46.35590207855391,
+      "learning_rate": 3.945272300745122e-06,
+      "loss": 0.5107,
+      "step": 13748
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.22998628609628,
+      "learning_rate": 3.943930239202996e-06,
+      "loss": 0.5111,
+      "step": 13749
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.3538026186712,
+      "learning_rate": 3.94258834989195e-06,
+      "loss": 0.4669,
+      "step": 13750
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 37.976226805156934,
+      "learning_rate": 3.941246632850146e-06,
+      "loss": 0.4349,
+      "step": 13751
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 52.198469286066334,
+      "learning_rate": 3.939905088115749e-06,
+      "loss": 0.5249,
+      "step": 13752
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.95490099488266,
+      "learning_rate": 3.938563715726904e-06,
+      "loss": 0.506,
+      "step": 13753
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 37.32082326506382,
+      "learning_rate": 3.937222515721759e-06,
+      "loss": 0.4871,
+      "step": 13754
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.367306148666096,
+      "learning_rate": 3.935881488138459e-06,
+      "loss": 0.4316,
+      "step": 13755
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.49651972413709,
+      "learning_rate": 3.934540633015141e-06,
+      "loss": 0.523,
+      "step": 13756
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 39.296624912808525,
+      "learning_rate": 3.933199950389944e-06,
+      "loss": 0.437,
+      "step": 13757
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 39.75692045800579,
+      "learning_rate": 3.931859440300991e-06,
+      "loss": 0.4595,
+      "step": 13758
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 36.47280263733492,
+      "learning_rate": 3.930519102786402e-06,
+      "loss": 0.464,
+      "step": 13759
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 39.10760070641011,
+      "learning_rate": 3.9291789378843e-06,
+      "loss": 0.4678,
+      "step": 13760
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 48.17664893811938,
+      "learning_rate": 3.927838945632799e-06,
+      "loss": 0.5189,
+      "step": 13761
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.88548833942825,
+      "learning_rate": 3.926499126070008e-06,
+      "loss": 0.4506,
+      "step": 13762
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 39.105900087722226,
+      "learning_rate": 3.925159479234025e-06,
+      "loss": 0.4854,
+      "step": 13763
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 37.29156820117305,
+      "learning_rate": 3.923820005162954e-06,
+      "loss": 0.4242,
+      "step": 13764
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 35.060034986534006,
+      "learning_rate": 3.92248070389489e-06,
+      "loss": 0.4998,
+      "step": 13765
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.280605256260245,
+      "learning_rate": 3.9211415754679164e-06,
+      "loss": 0.5531,
+      "step": 13766
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.10532466397687,
+      "learning_rate": 3.919802619920122e-06,
+      "loss": 0.4803,
+      "step": 13767
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.50487407335852,
+      "learning_rate": 3.9184638372895835e-06,
+      "loss": 0.5034,
+      "step": 13768
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 47.60950440156664,
+      "learning_rate": 3.9171252276143745e-06,
+      "loss": 0.456,
+      "step": 13769
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.14785428496593,
+      "learning_rate": 3.9157867909325684e-06,
+      "loss": 0.4586,
+      "step": 13770
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.67218963104772,
+      "learning_rate": 3.914448527282225e-06,
+      "loss": 0.5005,
+      "step": 13771
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 35.55337013011921,
+      "learning_rate": 3.9131104367014085e-06,
+      "loss": 0.4658,
+      "step": 13772
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 37.071433754817434,
+      "learning_rate": 3.9117725192281664e-06,
+      "loss": 0.4689,
+      "step": 13773
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 48.09558532266665,
+      "learning_rate": 3.910434774900555e-06,
+      "loss": 0.4869,
+      "step": 13774
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 46.99315863134615,
+      "learning_rate": 3.909097203756615e-06,
+      "loss": 0.4503,
+      "step": 13775
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 28.472683296708883,
+      "learning_rate": 3.907759805834387e-06,
+      "loss": 0.3846,
+      "step": 13776
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.66247885766739,
+      "learning_rate": 3.906422581171908e-06,
+      "loss": 0.4797,
+      "step": 13777
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 47.441151343988274,
+      "learning_rate": 3.905085529807208e-06,
+      "loss": 0.5403,
+      "step": 13778
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 46.005983227875035,
+      "learning_rate": 3.903748651778306e-06,
+      "loss": 0.5016,
+      "step": 13779
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 52.367545673978555,
+      "learning_rate": 3.902411947123226e-06,
+      "loss": 0.5511,
+      "step": 13780
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.955076369042295,
+      "learning_rate": 3.901075415879987e-06,
+      "loss": 0.5157,
+      "step": 13781
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.552342554703785,
+      "learning_rate": 3.8997390580865915e-06,
+      "loss": 0.4894,
+      "step": 13782
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 35.74209822846818,
+      "learning_rate": 3.898402873781052e-06,
+      "loss": 0.4128,
+      "step": 13783
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 35.08474705642301,
+      "learning_rate": 3.897066863001363e-06,
+      "loss": 0.4132,
+      "step": 13784
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 52.34180428148211,
+      "learning_rate": 3.895731025785521e-06,
+      "loss": 0.5217,
+      "step": 13785
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 43.16361776407622,
+      "learning_rate": 3.894395362171523e-06,
+      "loss": 0.5109,
+      "step": 13786
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.91069407157543,
+      "learning_rate": 3.893059872197347e-06,
+      "loss": 0.4435,
+      "step": 13787
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.14641175161513,
+      "learning_rate": 3.8917245559009735e-06,
+      "loss": 0.4806,
+      "step": 13788
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 41.08501155467786,
+      "learning_rate": 3.89038941332038e-06,
+      "loss": 0.4346,
+      "step": 13789
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 36.65536377624949,
+      "learning_rate": 3.889054444493537e-06,
+      "loss": 0.46,
+      "step": 13790
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 44.62352507154655,
+      "learning_rate": 3.887719649458419e-06,
+      "loss": 0.5294,
+      "step": 13791
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.7487907127089,
+      "learning_rate": 3.8863850282529715e-06,
+      "loss": 0.434,
+      "step": 13792
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 47.04040876919306,
+      "learning_rate": 3.8850505809151574e-06,
+      "loss": 0.5136,
+      "step": 13793
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.83592732941884,
+      "learning_rate": 3.883716307482928e-06,
+      "loss": 0.432,
+      "step": 13794
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 45.202919720560395,
+      "learning_rate": 3.882382207994231e-06,
+      "loss": 0.414,
+      "step": 13795
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 35.49457052511324,
+      "learning_rate": 3.881048282487005e-06,
+      "loss": 0.4464,
+      "step": 13796
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.80760142285162,
+      "learning_rate": 3.879714530999185e-06,
+      "loss": 0.5513,
+      "step": 13797
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 40.18757345806901,
+      "learning_rate": 3.878380953568702e-06,
+      "loss": 0.4665,
+      "step": 13798
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 43.170211837271786,
+      "learning_rate": 3.877047550233488e-06,
+      "loss": 0.5437,
+      "step": 13799
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.10054259860112,
+      "learning_rate": 3.875714321031455e-06,
+      "loss": 0.4295,
+      "step": 13800
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 42.404438755988785,
+      "learning_rate": 3.874381266000529e-06,
+      "loss": 0.4811,
+      "step": 13801
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.17724557227506,
+      "learning_rate": 3.873048385178613e-06,
+      "loss": 0.5104,
+      "step": 13802
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 39.83710331547246,
+      "learning_rate": 3.8717156786036194e-06,
+      "loss": 0.4473,
+      "step": 13803
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 37.90793273474,
+      "learning_rate": 3.870383146313444e-06,
+      "loss": 0.4598,
+      "step": 13804
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 48.660574694492006,
+      "learning_rate": 3.869050788345985e-06,
+      "loss": 0.5496,
+      "step": 13805
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 49.89599330560554,
+      "learning_rate": 3.86771860473914e-06,
+      "loss": 0.4688,
+      "step": 13806
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 47.83637990066325,
+      "learning_rate": 3.866386595530787e-06,
+      "loss": 0.5235,
+      "step": 13807
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 52.510860695752825,
+      "learning_rate": 3.865054760758813e-06,
+      "loss": 0.5708,
+      "step": 13808
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 41.06964567070146,
+      "learning_rate": 3.863723100461091e-06,
+      "loss": 0.4516,
+      "step": 13809
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 45.52873919756952,
+      "learning_rate": 3.8623916146754935e-06,
+      "loss": 0.4499,
+      "step": 13810
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 45.0534664363641,
+      "learning_rate": 3.861060303439892e-06,
+      "loss": 0.5093,
+      "step": 13811
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 35.84329530981294,
+      "learning_rate": 3.859729166792144e-06,
+      "loss": 0.4221,
+      "step": 13812
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 50.28002407080227,
+      "learning_rate": 3.858398204770104e-06,
+      "loss": 0.547,
+      "step": 13813
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 44.53328095740438,
+      "learning_rate": 3.8570674174116256e-06,
+      "loss": 0.4632,
+      "step": 13814
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 43.823328897459476,
+      "learning_rate": 3.855736804754556e-06,
+      "loss": 0.5047,
+      "step": 13815
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 44.87107027002907,
+      "learning_rate": 3.854406366836745e-06,
+      "loss": 0.463,
+      "step": 13816
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 44.07621125965672,
+      "learning_rate": 3.853076103696015e-06,
+      "loss": 0.4757,
+      "step": 13817
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 44.956229984740915,
+      "learning_rate": 3.851746015370204e-06,
+      "loss": 0.5298,
+      "step": 13818
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 35.813240389801535,
+      "learning_rate": 3.8504161018971395e-06,
+      "loss": 0.4581,
+      "step": 13819
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 38.707930795132086,
+      "learning_rate": 3.849086363314648e-06,
+      "loss": 0.4636,
+      "step": 13820
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 43.64469624072226,
+      "learning_rate": 3.847756799660543e-06,
+      "loss": 0.5313,
+      "step": 13821
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 41.44644039823281,
+      "learning_rate": 3.84642741097263e-06,
+      "loss": 0.4875,
+      "step": 13822
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 45.83353483704863,
+      "learning_rate": 3.845098197288724e-06,
+      "loss": 0.4694,
+      "step": 13823
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 36.88431059435721,
+      "learning_rate": 3.843769158646626e-06,
+      "loss": 0.4707,
+      "step": 13824
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 37.390504211154166,
+      "learning_rate": 3.842440295084133e-06,
+      "loss": 0.4644,
+      "step": 13825
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 41.718832583113524,
+      "learning_rate": 3.841111606639032e-06,
+      "loss": 0.4568,
+      "step": 13826
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 37.99068238188996,
+      "learning_rate": 3.839783093349114e-06,
+      "loss": 0.443,
+      "step": 13827
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 36.89418278898646,
+      "learning_rate": 3.838454755252166e-06,
+      "loss": 0.5054,
+      "step": 13828
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 42.46806524088191,
+      "learning_rate": 3.837126592385956e-06,
+      "loss": 0.5044,
+      "step": 13829
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 41.2773195253837,
+      "learning_rate": 3.8357986047882635e-06,
+      "loss": 0.541,
+      "step": 13830
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 44.676485552901795,
+      "learning_rate": 3.8344707924968494e-06,
+      "loss": 0.52,
+      "step": 13831
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 36.87367614050734,
+      "learning_rate": 3.833143155549479e-06,
+      "loss": 0.4312,
+      "step": 13832
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 35.73437151056096,
+      "learning_rate": 3.831815693983915e-06,
+      "loss": 0.5226,
+      "step": 13833
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 42.1568036166771,
+      "learning_rate": 3.830488407837899e-06,
+      "loss": 0.4943,
+      "step": 13834
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.37386837835623,
+      "learning_rate": 3.829161297149187e-06,
+      "loss": 0.5134,
+      "step": 13835
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 40.53215169381563,
+      "learning_rate": 3.8278343619555155e-06,
+      "loss": 0.454,
+      "step": 13836
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 48.62566313499795,
+      "learning_rate": 3.826507602294628e-06,
+      "loss": 0.5537,
+      "step": 13837
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 39.63423369250951,
+      "learning_rate": 3.825181018204248e-06,
+      "loss": 0.4809,
+      "step": 13838
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 45.22222991397073,
+      "learning_rate": 3.823854609722108e-06,
+      "loss": 0.502,
+      "step": 13839
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 43.7426039847434,
+      "learning_rate": 3.8225283768859345e-06,
+      "loss": 0.5153,
+      "step": 13840
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.663023853605985,
+      "learning_rate": 3.8212023197334395e-06,
+      "loss": 0.4999,
+      "step": 13841
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 37.772670715186614,
+      "learning_rate": 3.819876438302332e-06,
+      "loss": 0.4836,
+      "step": 13842
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 47.185708096567204,
+      "learning_rate": 3.8185507326303236e-06,
+      "loss": 0.5125,
+      "step": 13843
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 42.95685183224071,
+      "learning_rate": 3.817225202755117e-06,
+      "loss": 0.4275,
+      "step": 13844
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 62.199950284033406,
+      "learning_rate": 3.815899848714412e-06,
+      "loss": 0.449,
+      "step": 13845
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.89014178557466,
+      "learning_rate": 3.8145746705458976e-06,
+      "loss": 0.4558,
+      "step": 13846
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 40.99776352039418,
+      "learning_rate": 3.8132496682872576e-06,
+      "loss": 0.5612,
+      "step": 13847
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 36.405339176424896,
+      "learning_rate": 3.811924841976178e-06,
+      "loss": 0.41,
+      "step": 13848
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 36.62738950765689,
+      "learning_rate": 3.81060019165034e-06,
+      "loss": 0.4293,
+      "step": 13849
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 46.919069384759105,
+      "learning_rate": 3.8092757173474114e-06,
+      "loss": 0.577,
+      "step": 13850
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 42.26648931553834,
+      "learning_rate": 3.807951419105056e-06,
+      "loss": 0.4582,
+      "step": 13851
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 32.822235196390885,
+      "learning_rate": 3.8066272969609407e-06,
+      "loss": 0.4058,
+      "step": 13852
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 58.742278999899646,
+      "learning_rate": 3.8053033509527214e-06,
+      "loss": 0.5315,
+      "step": 13853
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 44.24276283317147,
+      "learning_rate": 3.8039795811180547e-06,
+      "loss": 0.4431,
+      "step": 13854
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 45.45473516975172,
+      "learning_rate": 3.8026559874945844e-06,
+      "loss": 0.5561,
+      "step": 13855
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 40.4390959623732,
+      "learning_rate": 3.801332570119949e-06,
+      "loss": 0.4742,
+      "step": 13856
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 42.2907486570097,
+      "learning_rate": 3.8000093290317886e-06,
+      "loss": 0.5065,
+      "step": 13857
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 30.276772706433626,
+      "learning_rate": 3.7986862642677402e-06,
+      "loss": 0.4332,
+      "step": 13858
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 41.95099426769001,
+      "learning_rate": 3.7973633758654225e-06,
+      "loss": 0.4595,
+      "step": 13859
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 54.23083857351346,
+      "learning_rate": 3.796040663862467e-06,
+      "loss": 0.4366,
+      "step": 13860
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.79705175629791,
+      "learning_rate": 3.7947181282964806e-06,
+      "loss": 0.4972,
+      "step": 13861
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 43.20617316807709,
+      "learning_rate": 3.793395769205085e-06,
+      "loss": 0.448,
+      "step": 13862
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.49364645132413,
+      "learning_rate": 3.7920735866258796e-06,
+      "loss": 0.4436,
+      "step": 13863
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.68604457613621,
+      "learning_rate": 3.790751580596469e-06,
+      "loss": 0.4781,
+      "step": 13864
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 52.83274571526203,
+      "learning_rate": 3.7894297511544553e-06,
+      "loss": 0.5286,
+      "step": 13865
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.97055614223394,
+      "learning_rate": 3.7881080983374263e-06,
+      "loss": 0.6044,
+      "step": 13866
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 36.4347262561807,
+      "learning_rate": 3.786786622182966e-06,
+      "loss": 0.4205,
+      "step": 13867
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.748856625918904,
+      "learning_rate": 3.7854653227286586e-06,
+      "loss": 0.5139,
+      "step": 13868
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 49.3236100188051,
+      "learning_rate": 3.784144200012083e-06,
+      "loss": 0.5655,
+      "step": 13869
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 45.00192666697649,
+      "learning_rate": 3.782823254070813e-06,
+      "loss": 0.4824,
+      "step": 13870
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 44.960376089718906,
+      "learning_rate": 3.781502484942413e-06,
+      "loss": 0.5067,
+      "step": 13871
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 42.67837051742199,
+      "learning_rate": 3.7801818926644416e-06,
+      "loss": 0.4969,
+      "step": 13872
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 46.635307323244405,
+      "learning_rate": 3.778861477274458e-06,
+      "loss": 0.4544,
+      "step": 13873
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 38.7864308673667,
+      "learning_rate": 3.7775412388100186e-06,
+      "loss": 0.4709,
+      "step": 13874
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.91565159098829,
+      "learning_rate": 3.7762211773086656e-06,
+      "loss": 0.4603,
+      "step": 13875
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 42.478776448534035,
+      "learning_rate": 3.7749012928079387e-06,
+      "loss": 0.539,
+      "step": 13876
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 58.19687687254664,
+      "learning_rate": 3.773581585345377e-06,
+      "loss": 0.5896,
+      "step": 13877
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 46.84433217160859,
+      "learning_rate": 3.772262054958513e-06,
+      "loss": 0.4808,
+      "step": 13878
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.72091124694513,
+      "learning_rate": 3.77094270168488e-06,
+      "loss": 0.4756,
+      "step": 13879
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 35.57747212289663,
+      "learning_rate": 3.769623525561986e-06,
+      "loss": 0.5096,
+      "step": 13880
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.4243063246376,
+      "learning_rate": 3.768304526627354e-06,
+      "loss": 0.4504,
+      "step": 13881
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 34.84010716713059,
+      "learning_rate": 3.7669857049184953e-06,
+      "loss": 0.467,
+      "step": 13882
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.668340000248534,
+      "learning_rate": 3.765667060472922e-06,
+      "loss": 0.4865,
+      "step": 13883
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 39.581159765230126,
+      "learning_rate": 3.764348593328129e-06,
+      "loss": 0.4902,
+      "step": 13884
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.40255207850353,
+      "learning_rate": 3.763030303521611e-06,
+      "loss": 0.3943,
+      "step": 13885
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 59.035619365001644,
+      "learning_rate": 3.7617121910908627e-06,
+      "loss": 0.4982,
+      "step": 13886
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 39.67219409911952,
+      "learning_rate": 3.760394256073374e-06,
+      "loss": 0.4351,
+      "step": 13887
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 39.05422574110062,
+      "learning_rate": 3.7590764985066187e-06,
+      "loss": 0.4441,
+      "step": 13888
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.823032968374456,
+      "learning_rate": 3.7577589184280817e-06,
+      "loss": 0.4673,
+      "step": 13889
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 34.38508888619214,
+      "learning_rate": 3.7564415158752244e-06,
+      "loss": 0.4579,
+      "step": 13890
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 50.8521476427112,
+      "learning_rate": 3.755124290885522e-06,
+      "loss": 0.5467,
+      "step": 13891
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.998219242036186,
+      "learning_rate": 3.753807243496429e-06,
+      "loss": 0.5323,
+      "step": 13892
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 41.38924248396409,
+      "learning_rate": 3.752490373745403e-06,
+      "loss": 0.5231,
+      "step": 13893
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.713783942276194,
+      "learning_rate": 3.751173681669901e-06,
+      "loss": 0.4283,
+      "step": 13894
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 48.66260484125116,
+      "learning_rate": 3.749857167307359e-06,
+      "loss": 0.4628,
+      "step": 13895
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 45.640604843871756,
+      "learning_rate": 3.7485408306952263e-06,
+      "loss": 0.477,
+      "step": 13896
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 36.554382694715855,
+      "learning_rate": 3.7472246718709325e-06,
+      "loss": 0.4411,
+      "step": 13897
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.832402521669984,
+      "learning_rate": 3.7459086908719124e-06,
+      "loss": 0.4733,
+      "step": 13898
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 44.02612580263095,
+      "learning_rate": 3.744592887735592e-06,
+      "loss": 0.549,
+      "step": 13899
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.99290817264118,
+      "learning_rate": 3.7432772624993917e-06,
+      "loss": 0.4651,
+      "step": 13900
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.074032653720955,
+      "learning_rate": 3.741961815200723e-06,
+      "loss": 0.4258,
+      "step": 13901
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.53118020780936,
+      "learning_rate": 3.740646545876998e-06,
+      "loss": 0.4175,
+      "step": 13902
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 42.144467804843586,
+      "learning_rate": 3.7393314545656243e-06,
+      "loss": 0.4916,
+      "step": 13903
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.58114732629151,
+      "learning_rate": 3.7380165413040092e-06,
+      "loss": 0.5249,
+      "step": 13904
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.73432124009589,
+      "learning_rate": 3.7367018061295335e-06,
+      "loss": 0.4587,
+      "step": 13905
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 39.198831012140566,
+      "learning_rate": 3.735387249079594e-06,
+      "loss": 0.4915,
+      "step": 13906
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 32.767715944035686,
+      "learning_rate": 3.7340728701915764e-06,
+      "loss": 0.4016,
+      "step": 13907
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 34.2716585997185,
+      "learning_rate": 3.732758669502864e-06,
+      "loss": 0.446,
+      "step": 13908
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 51.687256080208726,
+      "learning_rate": 3.731444647050829e-06,
+      "loss": 0.5288,
+      "step": 13909
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 44.06723798885482,
+      "learning_rate": 3.730130802872839e-06,
+      "loss": 0.5104,
+      "step": 13910
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 44.112712289283685,
+      "learning_rate": 3.72881713700626e-06,
+      "loss": 0.4798,
+      "step": 13911
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 48.37830400149814,
+      "learning_rate": 3.7275036494884563e-06,
+      "loss": 0.5389,
+      "step": 13912
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 39.11468204057466,
+      "learning_rate": 3.7261903403567767e-06,
+      "loss": 0.4623,
+      "step": 13913
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 35.70297441513392,
+      "learning_rate": 3.7248772096485773e-06,
+      "loss": 0.4132,
+      "step": 13914
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 43.00359498011043,
+      "learning_rate": 3.7235642574011955e-06,
+      "loss": 0.463,
+      "step": 13915
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.76419605397395,
+      "learning_rate": 3.7222514836519784e-06,
+      "loss": 0.431,
+      "step": 13916
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 44.95786400173784,
+      "learning_rate": 3.7209388884382526e-06,
+      "loss": 0.5307,
+      "step": 13917
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 49.13502772828788,
+      "learning_rate": 3.719626471797352e-06,
+      "loss": 0.5139,
+      "step": 13918
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 44.05195190573651,
+      "learning_rate": 3.7183142337666045e-06,
+      "loss": 0.4725,
+      "step": 13919
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.7443228084338,
+      "learning_rate": 3.7170021743833216e-06,
+      "loss": 0.512,
+      "step": 13920
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 45.05369399820689,
+      "learning_rate": 3.715690293684825e-06,
+      "loss": 0.5425,
+      "step": 13921
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 37.355691569460674,
+      "learning_rate": 3.7143785917084163e-06,
+      "loss": 0.4532,
+      "step": 13922
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 46.04078135638964,
+      "learning_rate": 3.7130670684914083e-06,
+      "loss": 0.515,
+      "step": 13923
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 47.654611523087155,
+      "learning_rate": 3.71175572407109e-06,
+      "loss": 0.4597,
+      "step": 13924
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 45.36817510891994,
+      "learning_rate": 3.7104445584847647e-06,
+      "loss": 0.4869,
+      "step": 13925
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 36.56694987353143,
+      "learning_rate": 3.709133571769713e-06,
+      "loss": 0.4171,
+      "step": 13926
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 46.738154445656306,
+      "learning_rate": 3.7078227639632234e-06,
+      "loss": 0.4172,
+      "step": 13927
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 49.29966370543323,
+      "learning_rate": 3.7065121351025758e-06,
+      "loss": 0.5518,
+      "step": 13928
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 43.7734218218161,
+      "learning_rate": 3.705201685225043e-06,
+      "loss": 0.5148,
+      "step": 13929
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 38.06316270548863,
+      "learning_rate": 3.7038914143678874e-06,
+      "loss": 0.4364,
+      "step": 13930
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 52.11507524828404,
+      "learning_rate": 3.702581322568376e-06,
+      "loss": 0.5203,
+      "step": 13931
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 32.68282943384474,
+      "learning_rate": 3.701271409863769e-06,
+      "loss": 0.4094,
+      "step": 13932
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 38.49179771292568,
+      "learning_rate": 3.6999616762913226e-06,
+      "loss": 0.4495,
+      "step": 13933
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 36.61755782408715,
+      "learning_rate": 3.6986521218882798e-06,
+      "loss": 0.4642,
+      "step": 13934
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 40.971273791153244,
+      "learning_rate": 3.697342746691881e-06,
+      "loss": 0.4522,
+      "step": 13935
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 50.95620733189373,
+      "learning_rate": 3.6960335507393672e-06,
+      "loss": 0.5512,
+      "step": 13936
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 46.17893563905497,
+      "learning_rate": 3.694724534067976e-06,
+      "loss": 0.5036,
+      "step": 13937
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 44.2564635971037,
+      "learning_rate": 3.693415696714929e-06,
+      "loss": 0.4552,
+      "step": 13938
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 47.831837003787776,
+      "learning_rate": 3.6921070387174484e-06,
+      "loss": 0.5013,
+      "step": 13939
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 34.44776534693338,
+      "learning_rate": 3.690798560112754e-06,
+      "loss": 0.4338,
+      "step": 13940
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 42.612558322194054,
+      "learning_rate": 3.6894902609380566e-06,
+      "loss": 0.4961,
+      "step": 13941
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.587563818015,
+      "learning_rate": 3.6881821412305683e-06,
+      "loss": 0.4553,
+      "step": 13942
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 41.46480561091591,
+      "learning_rate": 3.686874201027487e-06,
+      "loss": 0.4079,
+      "step": 13943
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 54.344997196462415,
+      "learning_rate": 3.6855664403660073e-06,
+      "loss": 0.5058,
+      "step": 13944
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.55314401459009,
+      "learning_rate": 3.684258859283325e-06,
+      "loss": 0.4783,
+      "step": 13945
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 41.223441637102084,
+      "learning_rate": 3.6829514578166293e-06,
+      "loss": 0.4738,
+      "step": 13946
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 44.16209383749962,
+      "learning_rate": 3.6816442360030946e-06,
+      "loss": 0.5344,
+      "step": 13947
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.68851551348634,
+      "learning_rate": 3.680337193879906e-06,
+      "loss": 0.4593,
+      "step": 13948
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 38.03926949435905,
+      "learning_rate": 3.679030331484227e-06,
+      "loss": 0.5242,
+      "step": 13949
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 44.31266356089634,
+      "learning_rate": 3.677723648853231e-06,
+      "loss": 0.4428,
+      "step": 13950
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 40.45522886355072,
+      "learning_rate": 3.676417146024073e-06,
+      "loss": 0.476,
+      "step": 13951
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 46.75330158039917,
+      "learning_rate": 3.6751108230339116e-06,
+      "loss": 0.4756,
+      "step": 13952
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 48.26001892430946,
+      "learning_rate": 3.673804679919901e-06,
+      "loss": 0.4534,
+      "step": 13953
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 36.302531031179,
+      "learning_rate": 3.672498716719186e-06,
+      "loss": 0.3999,
+      "step": 13954
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 44.135241300684505,
+      "learning_rate": 3.6711929334689024e-06,
+      "loss": 0.5896,
+      "step": 13955
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 47.88685332189984,
+      "learning_rate": 3.6698873302061888e-06,
+      "loss": 0.547,
+      "step": 13956
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 43.93466550710766,
+      "learning_rate": 3.668581906968177e-06,
+      "loss": 0.5578,
+      "step": 13957
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 42.28210201941558,
+      "learning_rate": 3.6672766637919967e-06,
+      "loss": 0.4403,
+      "step": 13958
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 35.143433481955554,
+      "learning_rate": 3.6659716007147616e-06,
+      "loss": 0.4571,
+      "step": 13959
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 37.19791961038934,
+      "learning_rate": 3.664666717773586e-06,
+      "loss": 0.4495,
+      "step": 13960
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 44.01377168436289,
+      "learning_rate": 3.6633620150055835e-06,
+      "loss": 0.5309,
+      "step": 13961
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 43.702141492604014,
+      "learning_rate": 3.6620574924478615e-06,
+      "loss": 0.414,
+      "step": 13962
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 45.445842737859266,
+      "learning_rate": 3.660753150137517e-06,
+      "loss": 0.4924,
+      "step": 13963
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 42.646115557257446,
+      "learning_rate": 3.65944898811164e-06,
+      "loss": 0.5047,
+      "step": 13964
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 53.263442559039035,
+      "learning_rate": 3.6581450064073265e-06,
+      "loss": 0.439,
+      "step": 13965
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.41909839529996,
+      "learning_rate": 3.6568412050616574e-06,
+      "loss": 0.5236,
+      "step": 13966
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 38.896510367417136,
+      "learning_rate": 3.655537584111718e-06,
+      "loss": 0.4192,
+      "step": 13967
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 40.44861372233183,
+      "learning_rate": 3.654234143594578e-06,
+      "loss": 0.4477,
+      "step": 13968
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 49.873433631068124,
+      "learning_rate": 3.652930883547304e-06,
+      "loss": 0.4879,
+      "step": 13969
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 36.97754031012208,
+      "learning_rate": 3.651627804006963e-06,
+      "loss": 0.4394,
+      "step": 13970
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.894265579150314,
+      "learning_rate": 3.650324905010616e-06,
+      "loss": 0.4418,
+      "step": 13971
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 41.08136549306189,
+      "learning_rate": 3.6490221865953146e-06,
+      "loss": 0.4822,
+      "step": 13972
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.75676971045478,
+      "learning_rate": 3.6477196487981036e-06,
+      "loss": 0.4372,
+      "step": 13973
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 44.54697178429051,
+      "learning_rate": 3.6464172916560305e-06,
+      "loss": 0.4703,
+      "step": 13974
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 37.97703599977955,
+      "learning_rate": 3.645115115206136e-06,
+      "loss": 0.4094,
+      "step": 13975
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 46.331900421233165,
+      "learning_rate": 3.643813119485445e-06,
+      "loss": 0.4631,
+      "step": 13976
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.57152460770514,
+      "learning_rate": 3.642511304530996e-06,
+      "loss": 0.4738,
+      "step": 13977
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 46.86031326112716,
+      "learning_rate": 3.641209670379803e-06,
+      "loss": 0.5033,
+      "step": 13978
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 48.42080385956593,
+      "learning_rate": 3.6399082170688893e-06,
+      "loss": 0.5387,
+      "step": 13979
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 43.27403747010446,
+      "learning_rate": 3.6386069446352632e-06,
+      "loss": 0.4631,
+      "step": 13980
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 41.86653928544981,
+      "learning_rate": 3.6373058531159332e-06,
+      "loss": 0.5035,
+      "step": 13981
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 33.17708106926751,
+      "learning_rate": 3.636004942547907e-06,
+      "loss": 0.4348,
+      "step": 13982
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 46.60107817098388,
+      "learning_rate": 3.634704212968174e-06,
+      "loss": 0.5422,
+      "step": 13983
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 34.87935563373761,
+      "learning_rate": 3.6334036644137328e-06,
+      "loss": 0.4123,
+      "step": 13984
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 42.749134634974595,
+      "learning_rate": 3.632103296921563e-06,
+      "loss": 0.4643,
+      "step": 13985
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 52.7697591158986,
+      "learning_rate": 3.630803110528651e-06,
+      "loss": 0.6061,
+      "step": 13986
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 52.102598046453565,
+      "learning_rate": 3.6295031052719766e-06,
+      "loss": 0.5637,
+      "step": 13987
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 41.692562319205805,
+      "learning_rate": 3.628203281188506e-06,
+      "loss": 0.487,
+      "step": 13988
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.865869602788095,
+      "learning_rate": 3.6269036383152032e-06,
+      "loss": 0.5158,
+      "step": 13989
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 37.831418820112724,
+      "learning_rate": 3.6256041766890337e-06,
+      "loss": 0.5172,
+      "step": 13990
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 45.166917962375805,
+      "learning_rate": 3.624304896346953e-06,
+      "loss": 0.5032,
+      "step": 13991
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 48.057487449634436,
+      "learning_rate": 3.623005797325917e-06,
+      "loss": 0.5383,
+      "step": 13992
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.09837051349595,
+      "learning_rate": 3.6217068796628597e-06,
+      "loss": 0.4507,
+      "step": 13993
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 40.77869696097975,
+      "learning_rate": 3.6204081433947268e-06,
+      "loss": 0.4591,
+      "step": 13994
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 47.61677988455272,
+      "learning_rate": 3.619109588558455e-06,
+      "loss": 0.4336,
+      "step": 13995
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 54.214765502336704,
+      "learning_rate": 3.6178112151909763e-06,
+      "loss": 0.4462,
+      "step": 13996
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 43.29082539536192,
+      "learning_rate": 3.6165130233292133e-06,
+      "loss": 0.5097,
+      "step": 13997
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.354441732969974,
+      "learning_rate": 3.6152150130100825e-06,
+      "loss": 0.4752,
+      "step": 13998
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 36.578505274944035,
+      "learning_rate": 3.6139171842705024e-06,
+      "loss": 0.4012,
+      "step": 13999
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.67065104634864,
+      "learning_rate": 3.612619537147385e-06,
+      "loss": 0.4789,
+      "step": 14000
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.75502158706065,
+      "learning_rate": 3.611322071677628e-06,
+      "loss": 0.4538,
+      "step": 14001
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 36.06927794455306,
+      "learning_rate": 3.6100247878981365e-06,
+      "loss": 0.4487,
+      "step": 14002
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 43.959151542726396,
+      "learning_rate": 3.6087276858458e-06,
+      "loss": 0.4406,
+      "step": 14003
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 57.67504298651275,
+      "learning_rate": 3.607430765557508e-06,
+      "loss": 0.416,
+      "step": 14004
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 38.95929731240562,
+      "learning_rate": 3.606134027070151e-06,
+      "loss": 0.5664,
+      "step": 14005
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 39.32700952850792,
+      "learning_rate": 3.604837470420596e-06,
+      "loss": 0.4585,
+      "step": 14006
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 37.90816738880326,
+      "learning_rate": 3.603541095645727e-06,
+      "loss": 0.4606,
+      "step": 14007
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 34.53484705836306,
+      "learning_rate": 3.6022449027824035e-06,
+      "loss": 0.4505,
+      "step": 14008
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 45.985527034512955,
+      "learning_rate": 3.600948891867496e-06,
+      "loss": 0.5768,
+      "step": 14009
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 54.38973685758338,
+      "learning_rate": 3.599653062937856e-06,
+      "loss": 0.5436,
+      "step": 14010
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 39.46892124117735,
+      "learning_rate": 3.598357416030338e-06,
+      "loss": 0.4868,
+      "step": 14011
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 48.11224620808647,
+      "learning_rate": 3.597061951181794e-06,
+      "loss": 0.5191,
+      "step": 14012
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 40.76599342001331,
+      "learning_rate": 3.5957666684290626e-06,
+      "loss": 0.4697,
+      "step": 14013
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 33.38167537053217,
+      "learning_rate": 3.594471567808977e-06,
+      "loss": 0.4034,
+      "step": 14014
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.838554942995074,
+      "learning_rate": 3.593176649358373e-06,
+      "loss": 0.4911,
+      "step": 14015
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.439266782813526,
+      "learning_rate": 3.591881913114077e-06,
+      "loss": 0.4577,
+      "step": 14016
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.14150287808572,
+      "learning_rate": 3.590587359112917e-06,
+      "loss": 0.4807,
+      "step": 14017
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 49.40637200742806,
+      "learning_rate": 3.589292987391697e-06,
+      "loss": 0.4791,
+      "step": 14018
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 43.49881362941518,
+      "learning_rate": 3.5879987979872342e-06,
+      "loss": 0.4832,
+      "step": 14019
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 53.277762684285705,
+      "learning_rate": 3.586704790936335e-06,
+      "loss": 0.5878,
+      "step": 14020
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 47.725675508765136,
+      "learning_rate": 3.5854109662758043e-06,
+      "loss": 0.4969,
+      "step": 14021
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 45.59684324589569,
+      "learning_rate": 3.584117324042431e-06,
+      "loss": 0.4667,
+      "step": 14022
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 40.87392735896623,
+      "learning_rate": 3.5828238642730063e-06,
+      "loss": 0.525,
+      "step": 14023
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 36.10414503610135,
+      "learning_rate": 3.581530587004316e-06,
+      "loss": 0.426,
+      "step": 14024
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.04266873441816,
+      "learning_rate": 3.580237492273144e-06,
+      "loss": 0.4669,
+      "step": 14025
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 37.36981632812253,
+      "learning_rate": 3.578944580116264e-06,
+      "loss": 0.4439,
+      "step": 14026
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 38.373578290666686,
+      "learning_rate": 3.5776518505704383e-06,
+      "loss": 0.4122,
+      "step": 14027
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 31.413678132508874,
+      "learning_rate": 3.5763593036724387e-06,
+      "loss": 0.3862,
+      "step": 14028
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 38.41138439761875,
+      "learning_rate": 3.575066939459022e-06,
+      "loss": 0.4361,
+      "step": 14029
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 49.7365762437361,
+      "learning_rate": 3.5737747579669468e-06,
+      "loss": 0.4275,
+      "step": 14030
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.87770038836672,
+      "learning_rate": 3.572482759232958e-06,
+      "loss": 0.4973,
+      "step": 14031
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 34.31840723464353,
+      "learning_rate": 3.5711909432937964e-06,
+      "loss": 0.4201,
+      "step": 14032
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 42.19798585347261,
+      "learning_rate": 3.5698993101862034e-06,
+      "loss": 0.4883,
+      "step": 14033
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 45.567454629897355,
+      "learning_rate": 3.5686078599469166e-06,
+      "loss": 0.4984,
+      "step": 14034
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 45.76392781505689,
+      "learning_rate": 3.5673165926126553e-06,
+      "loss": 0.5268,
+      "step": 14035
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 72.88463846790467,
+      "learning_rate": 3.566025508220151e-06,
+      "loss": 0.4733,
+      "step": 14036
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 37.64209005593329,
+      "learning_rate": 3.5647346068061152e-06,
+      "loss": 0.4344,
+      "step": 14037
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 38.214223648244975,
+      "learning_rate": 3.5634438884072653e-06,
+      "loss": 0.4269,
+      "step": 14038
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 34.172469044542346,
+      "learning_rate": 3.5621533530603024e-06,
+      "loss": 0.4723,
+      "step": 14039
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 49.665192348031326,
+      "learning_rate": 3.560863000801932e-06,
+      "loss": 0.5261,
+      "step": 14040
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 63.94666459897571,
+      "learning_rate": 3.559572831668855e-06,
+      "loss": 0.553,
+      "step": 14041
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.23936346075278,
+      "learning_rate": 3.558282845697758e-06,
+      "loss": 0.5117,
+      "step": 14042
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.744079963898095,
+      "learning_rate": 3.5569930429253263e-06,
+      "loss": 0.4585,
+      "step": 14043
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 49.91062088315916,
+      "learning_rate": 3.5557034233882426e-06,
+      "loss": 0.5892,
+      "step": 14044
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 47.70847072692589,
+      "learning_rate": 3.5544139871231842e-06,
+      "loss": 0.5239,
+      "step": 14045
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.72421668593353,
+      "learning_rate": 3.5531247341668253e-06,
+      "loss": 0.4952,
+      "step": 14046
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 41.5183528051366,
+      "learning_rate": 3.551835664555827e-06,
+      "loss": 0.4934,
+      "step": 14047
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 48.23959656464454,
+      "learning_rate": 3.5505467783268465e-06,
+      "loss": 0.519,
+      "step": 14048
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 35.45946136380718,
+      "learning_rate": 3.549258075516544e-06,
+      "loss": 0.411,
+      "step": 14049
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 40.82584995218955,
+      "learning_rate": 3.54796955616157e-06,
+      "loss": 0.4774,
+      "step": 14050
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 34.55107817837627,
+      "learning_rate": 3.546681220298569e-06,
+      "loss": 0.4444,
+      "step": 14051
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 39.05901610135323,
+      "learning_rate": 3.545393067964176e-06,
+      "loss": 0.454,
+      "step": 14052
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.447057869615634,
+      "learning_rate": 3.5441050991950264e-06,
+      "loss": 0.5039,
+      "step": 14053
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 44.40282335691022,
+      "learning_rate": 3.5428173140277523e-06,
+      "loss": 0.4867,
+      "step": 14054
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 44.13093986731809,
+      "learning_rate": 3.54152971249898e-06,
+      "loss": 0.4773,
+      "step": 14055
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 46.118701659866225,
+      "learning_rate": 3.5402422946453242e-06,
+      "loss": 0.4972,
+      "step": 14056
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 37.006293528369675,
+      "learning_rate": 3.5389550605033963e-06,
+      "loss": 0.4361,
+      "step": 14057
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 41.992748542763685,
+      "learning_rate": 3.537668010109805e-06,
+      "loss": 0.5017,
+      "step": 14058
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.96010562707555,
+      "learning_rate": 3.5363811435011598e-06,
+      "loss": 0.4655,
+      "step": 14059
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 49.394772561518224,
+      "learning_rate": 3.53509446071405e-06,
+      "loss": 0.4914,
+      "step": 14060
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.91203520441182,
+      "learning_rate": 3.5338079617850762e-06,
+      "loss": 0.4458,
+      "step": 14061
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 47.5367274542322,
+      "learning_rate": 3.5325216467508172e-06,
+      "loss": 0.5151,
+      "step": 14062
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.96477686043312,
+      "learning_rate": 3.5312355156478616e-06,
+      "loss": 0.4928,
+      "step": 14063
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 39.758344638086996,
+      "learning_rate": 3.5299495685127816e-06,
+      "loss": 0.4261,
+      "step": 14064
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 41.73598280894185,
+      "learning_rate": 3.528663805382151e-06,
+      "loss": 0.4762,
+      "step": 14065
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.8650763069709,
+      "learning_rate": 3.5273782262925392e-06,
+      "loss": 0.4682,
+      "step": 14066
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 39.901409563583435,
+      "learning_rate": 3.5260928312805042e-06,
+      "loss": 0.4657,
+      "step": 14067
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 59.16739072043237,
+      "learning_rate": 3.5248076203825987e-06,
+      "loss": 0.4822,
+      "step": 14068
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 56.43328829033292,
+      "learning_rate": 3.5235225936353767e-06,
+      "loss": 0.5131,
+      "step": 14069
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.42721418545704,
+      "learning_rate": 3.522237751075387e-06,
+      "loss": 0.4522,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 36.487735013606795,
+      "learning_rate": 3.5209530927391632e-06,
+      "loss": 0.3937,
+      "step": 14071
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 47.52255598382914,
+      "learning_rate": 3.5196686186632457e-06,
+      "loss": 0.5071,
+      "step": 14072
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 38.95645150350242,
+      "learning_rate": 3.518384328884159e-06,
+      "loss": 0.4502,
+      "step": 14073
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.634858249938056,
+      "learning_rate": 3.517100223438431e-06,
+      "loss": 0.4822,
+      "step": 14074
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.97421283744422,
+      "learning_rate": 3.5158163023625825e-06,
+      "loss": 0.5185,
+      "step": 14075
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 38.871444564093444,
+      "learning_rate": 3.514532565693126e-06,
+      "loss": 0.4406,
+      "step": 14076
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.47849635103117,
+      "learning_rate": 3.513249013466565e-06,
+      "loss": 0.4698,
+      "step": 14077
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 39.69256671192329,
+      "learning_rate": 3.5119656457194086e-06,
+      "loss": 0.5122,
+      "step": 14078
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 45.58522361843621,
+      "learning_rate": 3.510682462488153e-06,
+      "loss": 0.4024,
+      "step": 14079
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 38.31205544517195,
+      "learning_rate": 3.5093994638093e-06,
+      "loss": 0.3844,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 38.934702752853575,
+      "learning_rate": 3.5081166497193208e-06,
+      "loss": 0.4221,
+      "step": 14081
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.32665524724459,
+      "learning_rate": 3.506834020254708e-06,
+      "loss": 0.4774,
+      "step": 14082
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 38.59723558631956,
+      "learning_rate": 3.5055515754519363e-06,
+      "loss": 0.3989,
+      "step": 14083
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 45.220192570535716,
+      "learning_rate": 3.504269315347483e-06,
+      "loss": 0.4407,
+      "step": 14084
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.28100185942871,
+      "learning_rate": 3.50298723997781e-06,
+      "loss": 0.495,
+      "step": 14085
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 45.454214812766786,
+      "learning_rate": 3.501705349379375e-06,
+      "loss": 0.5151,
+      "step": 14086
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 38.740512677080865,
+      "learning_rate": 3.5004236435886395e-06,
+      "loss": 0.4223,
+      "step": 14087
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 50.08895604069071,
+      "learning_rate": 3.4991421226420563e-06,
+      "loss": 0.5718,
+      "step": 14088
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.54268883991353,
+      "learning_rate": 3.497860786576065e-06,
+      "loss": 0.4177,
+      "step": 14089
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 48.86278236836496,
+      "learning_rate": 3.496579635427113e-06,
+      "loss": 0.5538,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.88940444191654,
+      "learning_rate": 3.4952986692316283e-06,
+      "loss": 0.4761,
+      "step": 14091
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 36.82380122624538,
+      "learning_rate": 3.4940178880260446e-06,
+      "loss": 0.4352,
+      "step": 14092
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.67526466877526,
+      "learning_rate": 3.4927372918467893e-06,
+      "loss": 0.5383,
+      "step": 14093
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.24245795249258,
+      "learning_rate": 3.491456880730275e-06,
+      "loss": 0.4856,
+      "step": 14094
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 47.082881784665034,
+      "learning_rate": 3.4901766547129234e-06,
+      "loss": 0.4902,
+      "step": 14095
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.70993146650552,
+      "learning_rate": 3.4888966138311363e-06,
+      "loss": 0.4831,
+      "step": 14096
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 41.66734941050334,
+      "learning_rate": 3.487616758121324e-06,
+      "loss": 0.4794,
+      "step": 14097
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 54.66659405701553,
+      "learning_rate": 3.4863370876198777e-06,
+      "loss": 0.5248,
+      "step": 14098
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.9469487928306,
+      "learning_rate": 3.4850576023631942e-06,
+      "loss": 0.5158,
+      "step": 14099
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 45.80373597462098,
+      "learning_rate": 3.483778302387665e-06,
+      "loss": 0.485,
+      "step": 14100
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 48.42361022656477,
+      "learning_rate": 3.4824991877296687e-06,
+      "loss": 0.5412,
+      "step": 14101
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.939155574459086,
+      "learning_rate": 3.4812202584255805e-06,
+      "loss": 0.5075,
+      "step": 14102
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 45.23770825585028,
+      "learning_rate": 3.4799415145117743e-06,
+      "loss": 0.4618,
+      "step": 14103
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.621757436004025,
+      "learning_rate": 3.4786629560246166e-06,
+      "loss": 0.485,
+      "step": 14104
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 41.8574722615118,
+      "learning_rate": 3.477384583000477e-06,
+      "loss": 0.4246,
+      "step": 14105
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 43.53617814560248,
+      "learning_rate": 3.4761063954756967e-06,
+      "loss": 0.4225,
+      "step": 14106
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.14019306525123,
+      "learning_rate": 3.4748283934866354e-06,
+      "loss": 0.433,
+      "step": 14107
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 37.632065293407685,
+      "learning_rate": 3.4735505770696365e-06,
+      "loss": 0.4336,
+      "step": 14108
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 44.61163838485198,
+      "learning_rate": 3.472272946261045e-06,
+      "loss": 0.5398,
+      "step": 14109
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 41.71557924601182,
+      "learning_rate": 3.4709955010971928e-06,
+      "loss": 0.509,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 32.799983730980856,
+      "learning_rate": 3.4697182416144047e-06,
+      "loss": 0.4453,
+      "step": 14111
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 46.98192585231527,
+      "learning_rate": 3.4684411678490104e-06,
+      "loss": 0.5334,
+      "step": 14112
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 47.287630042780386,
+      "learning_rate": 3.4671642798373316e-06,
+      "loss": 0.5253,
+      "step": 14113
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 47.25864302084661,
+      "learning_rate": 3.4658875776156763e-06,
+      "loss": 0.4701,
+      "step": 14114
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.30694645902795,
+      "learning_rate": 3.4646110612203586e-06,
+      "loss": 0.4511,
+      "step": 14115
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 42.020562315586815,
+      "learning_rate": 3.4633347306876763e-06,
+      "loss": 0.4336,
+      "step": 14116
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 36.662175645118545,
+      "learning_rate": 3.4620585860539303e-06,
+      "loss": 0.4161,
+      "step": 14117
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 51.77148519933266,
+      "learning_rate": 3.460782627355417e-06,
+      "loss": 0.5312,
+      "step": 14118
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 37.388702400536616,
+      "learning_rate": 3.4595068546284207e-06,
+      "loss": 0.4464,
+      "step": 14119
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 45.55475609776973,
+      "learning_rate": 3.458231267909219e-06,
+      "loss": 0.5506,
+      "step": 14120
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 49.88395648740831,
+      "learning_rate": 3.4569558672340943e-06,
+      "loss": 0.4389,
+      "step": 14121
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 48.16917629149828,
+      "learning_rate": 3.455680652639319e-06,
+      "loss": 0.5323,
+      "step": 14122
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 45.529634288640366,
+      "learning_rate": 3.4544056241611556e-06,
+      "loss": 0.477,
+      "step": 14123
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 50.497220350003246,
+      "learning_rate": 3.4531307818358705e-06,
+      "loss": 0.4739,
+      "step": 14124
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.10609343321973,
+      "learning_rate": 3.451856125699713e-06,
+      "loss": 0.4893,
+      "step": 14125
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 50.0330465781686,
+      "learning_rate": 3.4505816557889393e-06,
+      "loss": 0.5038,
+      "step": 14126
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 46.819115103857705,
+      "learning_rate": 3.4493073721397895e-06,
+      "loss": 0.4486,
+      "step": 14127
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 43.26930862566473,
+      "learning_rate": 3.4480332747885047e-06,
+      "loss": 0.4878,
+      "step": 14128
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.374048475848255,
+      "learning_rate": 3.446759363771326e-06,
+      "loss": 0.5235,
+      "step": 14129
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 49.801667475072854,
+      "learning_rate": 3.445485639124476e-06,
+      "loss": 0.556,
+      "step": 14130
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 45.118743779807794,
+      "learning_rate": 3.444212100884177e-06,
+      "loss": 0.4705,
+      "step": 14131
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.39929293298202,
+      "learning_rate": 3.4429387490866506e-06,
+      "loss": 0.504,
+      "step": 14132
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 48.973321220518656,
+      "learning_rate": 3.4416655837681113e-06,
+      "loss": 0.531,
+      "step": 14133
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 60.09467045579459,
+      "learning_rate": 3.4403926049647684e-06,
+      "loss": 0.4934,
+      "step": 14134
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.292862637217254,
+      "learning_rate": 3.4391198127128232e-06,
+      "loss": 0.4968,
+      "step": 14135
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.93874786973564,
+      "learning_rate": 3.4378472070484704e-06,
+      "loss": 0.5127,
+      "step": 14136
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 54.51597026563641,
+      "learning_rate": 3.4365747880079027e-06,
+      "loss": 0.5493,
+      "step": 14137
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 36.67674770416536,
+      "learning_rate": 3.435302555627312e-06,
+      "loss": 0.4972,
+      "step": 14138
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 40.699746575609254,
+      "learning_rate": 3.4340305099428774e-06,
+      "loss": 0.5433,
+      "step": 14139
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 41.85816410444893,
+      "learning_rate": 3.4327586509907695e-06,
+      "loss": 0.4263,
+      "step": 14140
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 42.715872103256636,
+      "learning_rate": 3.431486978807165e-06,
+      "loss": 0.4435,
+      "step": 14141
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.74752999093697,
+      "learning_rate": 3.4302154934282272e-06,
+      "loss": 0.5343,
+      "step": 14142
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 46.28550503351706,
+      "learning_rate": 3.4289441948901214e-06,
+      "loss": 0.4978,
+      "step": 14143
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.526962041684676,
+      "learning_rate": 3.4276730832289997e-06,
+      "loss": 0.5126,
+      "step": 14144
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 42.38086741400917,
+      "learning_rate": 3.4264021584810058e-06,
+      "loss": 0.4767,
+      "step": 14145
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 37.78724504208779,
+      "learning_rate": 3.4251314206822894e-06,
+      "loss": 0.4279,
+      "step": 14146
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 44.17711134195133,
+      "learning_rate": 3.4238608698689933e-06,
+      "loss": 0.4959,
+      "step": 14147
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 40.93855062985669,
+      "learning_rate": 3.4225905060772436e-06,
+      "loss": 0.4311,
+      "step": 14148
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 36.96324684283852,
+      "learning_rate": 3.421320329343175e-06,
+      "loss": 0.4699,
+      "step": 14149
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 47.42433376819582,
+      "learning_rate": 3.420050339702905e-06,
+      "loss": 0.5117,
+      "step": 14150
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 45.153588139309605,
+      "learning_rate": 3.4187805371925575e-06,
+      "loss": 0.5282,
+      "step": 14151
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 50.405263705457294,
+      "learning_rate": 3.4175109218482383e-06,
+      "loss": 0.4803,
+      "step": 14152
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 43.81887178438104,
+      "learning_rate": 3.4162414937060584e-06,
+      "loss": 0.4809,
+      "step": 14153
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 41.22825696281156,
+      "learning_rate": 3.414972252802121e-06,
+      "loss": 0.4391,
+      "step": 14154
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 47.50780688679277,
+      "learning_rate": 3.413703199172519e-06,
+      "loss": 0.5341,
+      "step": 14155
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 55.142054149324515,
+      "learning_rate": 3.4124343328533473e-06,
+      "loss": 0.6065,
+      "step": 14156
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 31.903494033076605,
+      "learning_rate": 3.411165653880686e-06,
+      "loss": 0.3662,
+      "step": 14157
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 46.19172903065076,
+      "learning_rate": 3.40989716229062e-06,
+      "loss": 0.514,
+      "step": 14158
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 33.63670207182577,
+      "learning_rate": 3.408628858119226e-06,
+      "loss": 0.432,
+      "step": 14159
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 54.10868356931039,
+      "learning_rate": 3.4073607414025723e-06,
+      "loss": 0.5291,
+      "step": 14160
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 43.037333632565044,
+      "learning_rate": 3.406092812176719e-06,
+      "loss": 0.4837,
+      "step": 14161
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 53.51585025043836,
+      "learning_rate": 3.404825070477729e-06,
+      "loss": 0.5991,
+      "step": 14162
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.804024798835485,
+      "learning_rate": 3.403557516341658e-06,
+      "loss": 0.4179,
+      "step": 14163
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 37.55796179753057,
+      "learning_rate": 3.4022901498045536e-06,
+      "loss": 0.4372,
+      "step": 14164
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 42.926849110343944,
+      "learning_rate": 3.4010229709024545e-06,
+      "loss": 0.4852,
+      "step": 14165
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 37.43468497874267,
+      "learning_rate": 3.399755979671401e-06,
+      "loss": 0.4395,
+      "step": 14166
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 44.513470978873755,
+      "learning_rate": 3.398489176147427e-06,
+      "loss": 0.5079,
+      "step": 14167
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 46.63902101554177,
+      "learning_rate": 3.3972225603665653e-06,
+      "loss": 0.5041,
+      "step": 14168
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 51.43707374868923,
+      "learning_rate": 3.395956132364825e-06,
+      "loss": 0.4371,
+      "step": 14169
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 45.52670426150749,
+      "learning_rate": 3.394689892178229e-06,
+      "loss": 0.4678,
+      "step": 14170
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 54.262141068929445,
+      "learning_rate": 3.3934238398427886e-06,
+      "loss": 0.6046,
+      "step": 14171
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.53163218558498,
+      "learning_rate": 3.3921579753945132e-06,
+      "loss": 0.4979,
+      "step": 14172
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 42.55548929741075,
+      "learning_rate": 3.3908922988693995e-06,
+      "loss": 0.44,
+      "step": 14173
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.010416739496826,
+      "learning_rate": 3.389626810303439e-06,
+      "loss": 0.4475,
+      "step": 14174
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 50.89356280131615,
+      "learning_rate": 3.388361509732625e-06,
+      "loss": 0.5485,
+      "step": 14175
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.65903633394253,
+      "learning_rate": 3.387096397192945e-06,
+      "loss": 0.4554,
+      "step": 14176
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 37.14099178686765,
+      "learning_rate": 3.3858314727203724e-06,
+      "loss": 0.5452,
+      "step": 14177
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.96028572059018,
+      "learning_rate": 3.384566736350886e-06,
+      "loss": 0.4377,
+      "step": 14178
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 42.37567037429098,
+      "learning_rate": 3.3833021881204487e-06,
+      "loss": 0.5277,
+      "step": 14179
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 41.39119866658714,
+      "learning_rate": 3.3820378280650267e-06,
+      "loss": 0.519,
+      "step": 14180
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.59224307557924,
+      "learning_rate": 3.3807736562205805e-06,
+      "loss": 0.4273,
+      "step": 14181
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 36.93100301208385,
+      "learning_rate": 3.3795096726230558e-06,
+      "loss": 0.4469,
+      "step": 14182
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.70147477133268,
+      "learning_rate": 3.3782458773084072e-06,
+      "loss": 0.5782,
+      "step": 14183
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 36.33603052239331,
+      "learning_rate": 3.3769822703125686e-06,
+      "loss": 0.4479,
+      "step": 14184
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 38.94055710573785,
+      "learning_rate": 3.3757188516714833e-06,
+      "loss": 0.5038,
+      "step": 14185
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 46.92893723252799,
+      "learning_rate": 3.3744556214210756e-06,
+      "loss": 0.5157,
+      "step": 14186
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 32.34970959661282,
+      "learning_rate": 3.3731925795972743e-06,
+      "loss": 0.4026,
+      "step": 14187
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 36.26210448316645,
+      "learning_rate": 3.371929726236002e-06,
+      "loss": 0.4655,
+      "step": 14188
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.964998729132695,
+      "learning_rate": 3.370667061373172e-06,
+      "loss": 0.5312,
+      "step": 14189
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 36.83776707884111,
+      "learning_rate": 3.369404585044689e-06,
+      "loss": 0.4449,
+      "step": 14190
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.661269235942825,
+      "learning_rate": 3.3681422972864598e-06,
+      "loss": 0.4614,
+      "step": 14191
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 36.64817300252586,
+      "learning_rate": 3.366880198134386e-06,
+      "loss": 0.4564,
+      "step": 14192
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 44.83770178437591,
+      "learning_rate": 3.3656182876243636e-06,
+      "loss": 0.581,
+      "step": 14193
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 42.35418228634485,
+      "learning_rate": 3.3643565657922704e-06,
+      "loss": 0.5402,
+      "step": 14194
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 44.2366399586018,
+      "learning_rate": 3.3630950326739954e-06,
+      "loss": 0.4933,
+      "step": 14195
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 42.76961316713966,
+      "learning_rate": 3.3618336883054146e-06,
+      "loss": 0.4788,
+      "step": 14196
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.124696834803366,
+      "learning_rate": 3.360572532722404e-06,
+      "loss": 0.4231,
+      "step": 14197
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 35.77885733687987,
+      "learning_rate": 3.359311565960827e-06,
+      "loss": 0.4088,
+      "step": 14198
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 46.96521763009066,
+      "learning_rate": 3.3580507880565406e-06,
+      "loss": 0.4927,
+      "step": 14199
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 56.304102746799295,
+      "learning_rate": 3.3567901990454043e-06,
+      "loss": 0.4981,
+      "step": 14200
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 40.43045226032623,
+      "learning_rate": 3.355529798963272e-06,
+      "loss": 0.4732,
+      "step": 14201
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 38.856226552336764,
+      "learning_rate": 3.354269587845982e-06,
+      "loss": 0.4275,
+      "step": 14202
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 47.74568416903838,
+      "learning_rate": 3.3530095657293803e-06,
+      "loss": 0.4641,
+      "step": 14203
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.35222690494635,
+      "learning_rate": 3.3517497326492952e-06,
+      "loss": 0.4853,
+      "step": 14204
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 51.78207236576818,
+      "learning_rate": 3.350490088641558e-06,
+      "loss": 0.5942,
+      "step": 14205
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 37.06546479536194,
+      "learning_rate": 3.3492306337419967e-06,
+      "loss": 0.4761,
+      "step": 14206
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.604947943140935,
+      "learning_rate": 3.3479713679864223e-06,
+      "loss": 0.4408,
+      "step": 14207
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.393966315194696,
+      "learning_rate": 3.346712291410654e-06,
+      "loss": 0.4537,
+      "step": 14208
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 32.36960343633271,
+      "learning_rate": 3.345453404050493e-06,
+      "loss": 0.4158,
+      "step": 14209
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 57.55033149872332,
+      "learning_rate": 3.344194705941748e-06,
+      "loss": 0.4899,
+      "step": 14210
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.07542544357961,
+      "learning_rate": 3.342936197120208e-06,
+      "loss": 0.526,
+      "step": 14211
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 46.16331538914246,
+      "learning_rate": 3.341677877621672e-06,
+      "loss": 0.515,
+      "step": 14212
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.539988799674155,
+      "learning_rate": 3.340419747481919e-06,
+      "loss": 0.4493,
+      "step": 14213
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.686341025099736,
+      "learning_rate": 3.3391618067367347e-06,
+      "loss": 0.4348,
+      "step": 14214
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 54.46748185555498,
+      "learning_rate": 3.33790405542189e-06,
+      "loss": 0.4984,
+      "step": 14215
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.4985442315813,
+      "learning_rate": 3.3366464935731566e-06,
+      "loss": 0.469,
+      "step": 14216
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.73125650136696,
+      "learning_rate": 3.3353891212263023e-06,
+      "loss": 0.4792,
+      "step": 14217
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 45.22997333411123,
+      "learning_rate": 3.3341319384170824e-06,
+      "loss": 0.4478,
+      "step": 14218
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.74962903046404,
+      "learning_rate": 3.332874945181248e-06,
+      "loss": 0.4124,
+      "step": 14219
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 45.89674657166173,
+      "learning_rate": 3.33161814155455e-06,
+      "loss": 0.509,
+      "step": 14220
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.76664745479293,
+      "learning_rate": 3.330361527572731e-06,
+      "loss": 0.476,
+      "step": 14221
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 35.497497483820325,
+      "learning_rate": 3.3291051032715317e-06,
+      "loss": 0.4745,
+      "step": 14222
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 34.323367564009025,
+      "learning_rate": 3.3278488686866807e-06,
+      "loss": 0.4261,
+      "step": 14223
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.556077258027756,
+      "learning_rate": 3.3265928238539015e-06,
+      "loss": 0.4985,
+      "step": 14224
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.90039600158949,
+      "learning_rate": 3.3253369688089186e-06,
+      "loss": 0.5168,
+      "step": 14225
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.822704860329864,
+      "learning_rate": 3.324081303587451e-06,
+      "loss": 0.4441,
+      "step": 14226
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 47.7820366744155,
+      "learning_rate": 3.3228258282252056e-06,
+      "loss": 0.4396,
+      "step": 14227
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.699537441612065,
+      "learning_rate": 3.321570542757885e-06,
+      "loss": 0.46,
+      "step": 14228
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.808223273082746,
+      "learning_rate": 3.320315447221191e-06,
+      "loss": 0.476,
+      "step": 14229
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 40.59274482216706,
+      "learning_rate": 3.3190605416508182e-06,
+      "loss": 0.4619,
+      "step": 14230
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.96844938044351,
+      "learning_rate": 3.3178058260824586e-06,
+      "loss": 0.4624,
+      "step": 14231
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.783244134624894,
+      "learning_rate": 3.3165513005517925e-06,
+      "loss": 0.4482,
+      "step": 14232
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 49.47626193973403,
+      "learning_rate": 3.3152969650944943e-06,
+      "loss": 0.4967,
+      "step": 14233
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 45.69010447163952,
+      "learning_rate": 3.3140428197462406e-06,
+      "loss": 0.4742,
+      "step": 14234
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.636716175539796,
+      "learning_rate": 3.312788864542701e-06,
+      "loss": 0.4467,
+      "step": 14235
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 43.33317537034416,
+      "learning_rate": 3.3115350995195293e-06,
+      "loss": 0.5211,
+      "step": 14236
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 49.320831136576416,
+      "learning_rate": 3.310281524712392e-06,
+      "loss": 0.5783,
+      "step": 14237
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 36.74847586708559,
+      "learning_rate": 3.3090281401569302e-06,
+      "loss": 0.5026,
+      "step": 14238
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 35.750533683501374,
+      "learning_rate": 3.307774945888799e-06,
+      "loss": 0.3678,
+      "step": 14239
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.19215690033729,
+      "learning_rate": 3.306521941943628e-06,
+      "loss": 0.4171,
+      "step": 14240
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 48.98183756557791,
+      "learning_rate": 3.3052691283570593e-06,
+      "loss": 0.5136,
+      "step": 14241
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 42.58444698330736,
+      "learning_rate": 3.304016505164722e-06,
+      "loss": 0.453,
+      "step": 14242
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 45.68367847371165,
+      "learning_rate": 3.3027640724022357e-06,
+      "loss": 0.5262,
+      "step": 14243
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 55.82999371556775,
+      "learning_rate": 3.301511830105225e-06,
+      "loss": 0.5046,
+      "step": 14244
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 37.59353793013857,
+      "learning_rate": 3.300259778309296e-06,
+      "loss": 0.3907,
+      "step": 14245
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 36.67752924897876,
+      "learning_rate": 3.2990079170500587e-06,
+      "loss": 0.4507,
+      "step": 14246
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 34.56830745766188,
+      "learning_rate": 3.297756246363121e-06,
+      "loss": 0.4322,
+      "step": 14247
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.38364735218129,
+      "learning_rate": 3.2965047662840744e-06,
+      "loss": 0.5538,
+      "step": 14248
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.5932928238076,
+      "learning_rate": 3.2952534768485066e-06,
+      "loss": 0.4577,
+      "step": 14249
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.72965015106548,
+      "learning_rate": 3.294002378092008e-06,
+      "loss": 0.4969,
+      "step": 14250
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 34.91304705703728,
+      "learning_rate": 3.292751470050164e-06,
+      "loss": 0.46,
+      "step": 14251
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 38.72320124029936,
+      "learning_rate": 3.291500752758543e-06,
+      "loss": 0.4845,
+      "step": 14252
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 56.26101529667481,
+      "learning_rate": 3.2902502262527124e-06,
+      "loss": 0.5846,
+      "step": 14253
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 47.01783077888522,
+      "learning_rate": 3.288999890568242e-06,
+      "loss": 0.4909,
+      "step": 14254
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 38.198243760386006,
+      "learning_rate": 3.287749745740687e-06,
+      "loss": 0.4406,
+      "step": 14255
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 36.42231648757287,
+      "learning_rate": 3.2864997918056075e-06,
+      "loss": 0.5068,
+      "step": 14256
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 44.97928072991631,
+      "learning_rate": 3.2852500287985455e-06,
+      "loss": 0.5027,
+      "step": 14257
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 35.294911262831896,
+      "learning_rate": 3.284000456755042e-06,
+      "loss": 0.4342,
+      "step": 14258
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 47.39585140297793,
+      "learning_rate": 3.282751075710637e-06,
+      "loss": 0.4097,
+      "step": 14259
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 40.92176400803542,
+      "learning_rate": 3.2815018857008653e-06,
+      "loss": 0.4485,
+      "step": 14260
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 49.532727529347525,
+      "learning_rate": 3.2802528867612494e-06,
+      "loss": 0.5,
+      "step": 14261
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 42.303594607169906,
+      "learning_rate": 3.279004078927307e-06,
+      "loss": 0.4952,
+      "step": 14262
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 42.6179458480762,
+      "learning_rate": 3.2777554622345563e-06,
+      "loss": 0.5555,
+      "step": 14263
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 42.91922998489088,
+      "learning_rate": 3.2765070367185126e-06,
+      "loss": 0.4086,
+      "step": 14264
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.987527732083834,
+      "learning_rate": 3.2752588024146714e-06,
+      "loss": 0.5105,
+      "step": 14265
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.51311074117479,
+      "learning_rate": 3.274010759358539e-06,
+      "loss": 0.4164,
+      "step": 14266
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.95966486391277,
+      "learning_rate": 3.272762907585604e-06,
+      "loss": 0.4628,
+      "step": 14267
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 46.02505759691968,
+      "learning_rate": 3.271515247131356e-06,
+      "loss": 0.4526,
+      "step": 14268
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 49.88165314356961,
+      "learning_rate": 3.270267778031281e-06,
+      "loss": 0.5298,
+      "step": 14269
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.99602878052103,
+      "learning_rate": 3.2690205003208508e-06,
+      "loss": 0.5259,
+      "step": 14270
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 42.81888005408538,
+      "learning_rate": 3.2677734140355434e-06,
+      "loss": 0.4059,
+      "step": 14271
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 43.402634208475654,
+      "learning_rate": 3.2665265192108186e-06,
+      "loss": 0.4669,
+      "step": 14272
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 52.609062400095354,
+      "learning_rate": 3.265279815882144e-06,
+      "loss": 0.5792,
+      "step": 14273
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 43.93925132354125,
+      "learning_rate": 3.264033304084968e-06,
+      "loss": 0.4264,
+      "step": 14274
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 40.41003524740032,
+      "learning_rate": 3.2627869838547454e-06,
+      "loss": 0.4841,
+      "step": 14275
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 40.354853519679104,
+      "learning_rate": 3.2615408552269223e-06,
+      "loss": 0.4457,
+      "step": 14276
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.68904231019405,
+      "learning_rate": 3.260294918236935e-06,
+      "loss": 0.4648,
+      "step": 14277
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 37.74929443230524,
+      "learning_rate": 3.2590491729202146e-06,
+      "loss": 0.5338,
+      "step": 14278
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 48.4177945520515,
+      "learning_rate": 3.257803619312193e-06,
+      "loss": 0.5033,
+      "step": 14279
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 40.20409761891324,
+      "learning_rate": 3.256558257448292e-06,
+      "loss": 0.4822,
+      "step": 14280
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 55.057680891195645,
+      "learning_rate": 3.255313087363936e-06,
+      "loss": 0.5435,
+      "step": 14281
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 35.7790207347483,
+      "learning_rate": 3.2540681090945235e-06,
+      "loss": 0.4476,
+      "step": 14282
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 52.71289146174859,
+      "learning_rate": 3.252823322675468e-06,
+      "loss": 0.4207,
+      "step": 14283
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 46.40834445015945,
+      "learning_rate": 3.2515787281421697e-06,
+      "loss": 0.5024,
+      "step": 14284
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 40.52758113047008,
+      "learning_rate": 3.250334325530029e-06,
+      "loss": 0.5092,
+      "step": 14285
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 37.66264532398607,
+      "learning_rate": 3.24909011487443e-06,
+      "loss": 0.4173,
+      "step": 14286
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 53.459510044973825,
+      "learning_rate": 3.247846096210757e-06,
+      "loss": 0.4845,
+      "step": 14287
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 35.04322609548079,
+      "learning_rate": 3.2466022695743905e-06,
+      "loss": 0.404,
+      "step": 14288
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.56097004523397,
+      "learning_rate": 3.2453586350007084e-06,
+      "loss": 0.5023,
+      "step": 14289
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 46.347018285476736,
+      "learning_rate": 3.2441151925250724e-06,
+      "loss": 0.4406,
+      "step": 14290
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 50.532447800744976,
+      "learning_rate": 3.2428719421828515e-06,
+      "loss": 0.5373,
+      "step": 14291
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.70043088200935,
+      "learning_rate": 3.2416288840093955e-06,
+      "loss": 0.527,
+      "step": 14292
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 54.06025040196308,
+      "learning_rate": 3.2403860180400615e-06,
+      "loss": 0.57,
+      "step": 14293
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.89558322294048,
+      "learning_rate": 3.2391433443101984e-06,
+      "loss": 0.4721,
+      "step": 14294
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 38.34889461199977,
+      "learning_rate": 3.237900862855139e-06,
+      "loss": 0.4078,
+      "step": 14295
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.27330127250033,
+      "learning_rate": 3.236658573710227e-06,
+      "loss": 0.3884,
+      "step": 14296
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.30664001983183,
+      "learning_rate": 3.2354164769107866e-06,
+      "loss": 0.4641,
+      "step": 14297
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 50.15924857139518,
+      "learning_rate": 3.234174572492147e-06,
+      "loss": 0.4377,
+      "step": 14298
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.561069993002484,
+      "learning_rate": 3.2329328604896217e-06,
+      "loss": 0.4807,
+      "step": 14299
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.355110503807204,
+      "learning_rate": 3.2316913409385263e-06,
+      "loss": 0.4221,
+      "step": 14300
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 35.98519392982415,
+      "learning_rate": 3.2304500138741736e-06,
+      "loss": 0.4271,
+      "step": 14301
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 36.065049724256355,
+      "learning_rate": 3.229208879331862e-06,
+      "loss": 0.3813,
+      "step": 14302
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 44.2982835656713,
+      "learning_rate": 3.2279679373468866e-06,
+      "loss": 0.4826,
+      "step": 14303
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 48.226487958545505,
+      "learning_rate": 3.2267271879545413e-06,
+      "loss": 0.5223,
+      "step": 14304
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 46.168308674133584,
+      "learning_rate": 3.2254866311901125e-06,
+      "loss": 0.4843,
+      "step": 14305
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.67706192114693,
+      "learning_rate": 3.224246267088884e-06,
+      "loss": 0.4179,
+      "step": 14306
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 48.92361523126074,
+      "learning_rate": 3.2230060956861285e-06,
+      "loss": 0.501,
+      "step": 14307
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 45.077035086164116,
+      "learning_rate": 3.221766117017111e-06,
+      "loss": 0.5086,
+      "step": 14308
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 48.53315710482898,
+      "learning_rate": 3.220526331117101e-06,
+      "loss": 0.4747,
+      "step": 14309
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 40.77002377681773,
+      "learning_rate": 3.219286738021359e-06,
+      "loss": 0.485,
+      "step": 14310
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 38.689151460591454,
+      "learning_rate": 3.2180473377651355e-06,
+      "loss": 0.4263,
+      "step": 14311
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 39.52007802879299,
+      "learning_rate": 3.216808130383675e-06,
+      "loss": 0.4347,
+      "step": 14312
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.61090918558338,
+      "learning_rate": 3.2155691159122236e-06,
+      "loss": 0.4653,
+      "step": 14313
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 38.09122200480113,
+      "learning_rate": 3.214330294386021e-06,
+      "loss": 0.4819,
+      "step": 14314
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 41.04701143146984,
+      "learning_rate": 3.2130916658402956e-06,
+      "loss": 0.4287,
+      "step": 14315
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.97492585230662,
+      "learning_rate": 3.211853230310269e-06,
+      "loss": 0.4866,
+      "step": 14316
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 43.646465474773,
+      "learning_rate": 3.2106149878311655e-06,
+      "loss": 0.4192,
+      "step": 14317
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 39.132538078444085,
+      "learning_rate": 3.209376938438201e-06,
+      "loss": 0.4422,
+      "step": 14318
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 36.44445885806425,
+      "learning_rate": 3.208139082166587e-06,
+      "loss": 0.4211,
+      "step": 14319
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 43.3387108522785,
+      "learning_rate": 3.2069014190515237e-06,
+      "loss": 0.4688,
+      "step": 14320
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 50.031879113725594,
+      "learning_rate": 3.205663949128207e-06,
+      "loss": 0.5519,
+      "step": 14321
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.99077943979554,
+      "learning_rate": 3.204426672431834e-06,
+      "loss": 0.488,
+      "step": 14322
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 40.103474675273134,
+      "learning_rate": 3.2031895889975926e-06,
+      "loss": 0.4619,
+      "step": 14323
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.77513884810302,
+      "learning_rate": 3.201952698860662e-06,
+      "loss": 0.4586,
+      "step": 14324
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 34.86492353230993,
+      "learning_rate": 3.200716002056222e-06,
+      "loss": 0.4269,
+      "step": 14325
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 43.72028880379927,
+      "learning_rate": 3.199479498619439e-06,
+      "loss": 0.4986,
+      "step": 14326
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 41.20284988953768,
+      "learning_rate": 3.1982431885854837e-06,
+      "loss": 0.4449,
+      "step": 14327
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 49.72540887421925,
+      "learning_rate": 3.19700707198951e-06,
+      "loss": 0.5108,
+      "step": 14328
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 38.02565652153876,
+      "learning_rate": 3.1957711488666756e-06,
+      "loss": 0.4565,
+      "step": 14329
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.027218894358484,
+      "learning_rate": 3.1945354192521316e-06,
+      "loss": 0.4992,
+      "step": 14330
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 47.07377464129892,
+      "learning_rate": 3.193299883181017e-06,
+      "loss": 0.4325,
+      "step": 14331
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 33.94779312674717,
+      "learning_rate": 3.192064540688475e-06,
+      "loss": 0.4097,
+      "step": 14332
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 38.59868007476055,
+      "learning_rate": 3.190829391809631e-06,
+      "loss": 0.4324,
+      "step": 14333
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 59.42107436758919,
+      "learning_rate": 3.189594436579617e-06,
+      "loss": 0.4576,
+      "step": 14334
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 38.251293247774754,
+      "learning_rate": 3.1883596750335554e-06,
+      "loss": 0.4658,
+      "step": 14335
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 42.65959590602889,
+      "learning_rate": 3.1871251072065613e-06,
+      "loss": 0.4562,
+      "step": 14336
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 36.47678546914988,
+      "learning_rate": 3.1858907331337397e-06,
+      "loss": 0.4444,
+      "step": 14337
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 49.075474655063516,
+      "learning_rate": 3.1846565528501993e-06,
+      "loss": 0.5052,
+      "step": 14338
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 46.92655206083901,
+      "learning_rate": 3.1834225663910436e-06,
+      "loss": 0.5062,
+      "step": 14339
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.61936802004472,
+      "learning_rate": 3.1821887737913613e-06,
+      "loss": 0.5522,
+      "step": 14340
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 42.24317836710525,
+      "learning_rate": 3.18095517508624e-06,
+      "loss": 0.4683,
+      "step": 14341
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 39.92172354652656,
+      "learning_rate": 3.1797217703107643e-06,
+      "loss": 0.4984,
+      "step": 14342
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 36.514108818155734,
+      "learning_rate": 3.1784885595000114e-06,
+      "loss": 0.4135,
+      "step": 14343
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 35.276456982041296,
+      "learning_rate": 3.177255542689056e-06,
+      "loss": 0.3797,
+      "step": 14344
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 37.59623609838169,
+      "learning_rate": 3.1760227199129623e-06,
+      "loss": 0.4179,
+      "step": 14345
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 39.65992341639655,
+      "learning_rate": 3.1747900912067877e-06,
+      "loss": 0.4608,
+      "step": 14346
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 42.79185103154172,
+      "learning_rate": 3.17355765660559e-06,
+      "loss": 0.5111,
+      "step": 14347
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 40.828734059395074,
+      "learning_rate": 3.172325416144423e-06,
+      "loss": 0.4405,
+      "step": 14348
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 40.65937515785359,
+      "learning_rate": 3.171093369858322e-06,
+      "loss": 0.464,
+      "step": 14349
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 41.441139510147835,
+      "learning_rate": 3.1698615177823357e-06,
+      "loss": 0.4585,
+      "step": 14350
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 57.498230716187635,
+      "learning_rate": 3.1686298599514887e-06,
+      "loss": 0.6029,
+      "step": 14351
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 50.94534950389551,
+      "learning_rate": 3.167398396400816e-06,
+      "loss": 0.5047,
+      "step": 14352
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 50.900194720062544,
+      "learning_rate": 3.1661671271653317e-06,
+      "loss": 0.4714,
+      "step": 14353
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 46.45544334911188,
+      "learning_rate": 3.1649360522800578e-06,
+      "loss": 0.5324,
+      "step": 14354
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 47.42624252292331,
+      "learning_rate": 3.163705171780006e-06,
+      "loss": 0.6386,
+      "step": 14355
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 38.9847151135609,
+      "learning_rate": 3.1624744857001776e-06,
+      "loss": 0.4791,
+      "step": 14356
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 37.6396722542361,
+      "learning_rate": 3.161243994075578e-06,
+      "loss": 0.44,
+      "step": 14357
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 47.89706084426344,
+      "learning_rate": 3.1600136969411954e-06,
+      "loss": 0.5086,
+      "step": 14358
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 35.812943487433856,
+      "learning_rate": 3.1587835943320256e-06,
+      "loss": 0.4506,
+      "step": 14359
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 41.96582105106331,
+      "learning_rate": 3.1575536862830445e-06,
+      "loss": 0.4543,
+      "step": 14360
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 45.599333104267934,
+      "learning_rate": 3.156323972829237e-06,
+      "loss": 0.4536,
+      "step": 14361
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 49.84282196998643,
+      "learning_rate": 3.1550944540055693e-06,
+      "loss": 0.5185,
+      "step": 14362
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 51.67498207959371,
+      "learning_rate": 3.1538651298470126e-06,
+      "loss": 0.3996,
+      "step": 14363
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 41.71497587927311,
+      "learning_rate": 3.1526360003885283e-06,
+      "loss": 0.4247,
+      "step": 14364
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 44.88401696935411,
+      "learning_rate": 3.151407065665071e-06,
+      "loss": 0.4764,
+      "step": 14365
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 36.279301646092385,
+      "learning_rate": 3.150178325711587e-06,
+      "loss": 0.4616,
+      "step": 14366
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 43.06205737934273,
+      "learning_rate": 3.1489497805630243e-06,
+      "loss": 0.4702,
+      "step": 14367
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 46.962328392529585,
+      "learning_rate": 3.1477214302543225e-06,
+      "loss": 0.5049,
+      "step": 14368
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 46.6495216434031,
+      "learning_rate": 3.1464932748204215e-06,
+      "loss": 0.42,
+      "step": 14369
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 47.750994637001405,
+      "learning_rate": 3.1452653142962353e-06,
+      "loss": 0.4442,
+      "step": 14370
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 44.023118702530475,
+      "learning_rate": 3.144037548716694e-06,
+      "loss": 0.5163,
+      "step": 14371
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 40.51864294426041,
+      "learning_rate": 3.142809978116714e-06,
+      "loss": 0.4672,
+      "step": 14372
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.136816376972206,
+      "learning_rate": 3.1415826025312103e-06,
+      "loss": 0.4176,
+      "step": 14373
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 40.28883669177634,
+      "learning_rate": 3.140355421995086e-06,
+      "loss": 0.5061,
+      "step": 14374
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 41.09600888559985,
+      "learning_rate": 3.139128436543236e-06,
+      "loss": 0.4369,
+      "step": 14375
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.07243941729458,
+      "learning_rate": 3.1379016462105617e-06,
+      "loss": 0.4468,
+      "step": 14376
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 35.05761758874507,
+      "learning_rate": 3.136675051031952e-06,
+      "loss": 0.3722,
+      "step": 14377
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 41.30847975406293,
+      "learning_rate": 3.1354486510422865e-06,
+      "loss": 0.4469,
+      "step": 14378
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.33259565732597,
+      "learning_rate": 3.134222446276448e-06,
+      "loss": 0.5087,
+      "step": 14379
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 63.05190211489092,
+      "learning_rate": 3.1329964367693042e-06,
+      "loss": 0.5176,
+      "step": 14380
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 40.29871710907726,
+      "learning_rate": 3.1317706225557254e-06,
+      "loss": 0.4526,
+      "step": 14381
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.88418242158144,
+      "learning_rate": 3.130545003670574e-06,
+      "loss": 0.4623,
+      "step": 14382
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 42.08911991497123,
+      "learning_rate": 3.129319580148702e-06,
+      "loss": 0.4976,
+      "step": 14383
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 52.69136011502146,
+      "learning_rate": 3.128094352024965e-06,
+      "loss": 0.5242,
+      "step": 14384
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.96467234115134,
+      "learning_rate": 3.1268693193342014e-06,
+      "loss": 0.5086,
+      "step": 14385
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.23371232095853,
+      "learning_rate": 3.1256444821112573e-06,
+      "loss": 0.4919,
+      "step": 14386
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 47.778230031531976,
+      "learning_rate": 3.1244198403909587e-06,
+      "loss": 0.4902,
+      "step": 14387
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.449693018364925,
+      "learning_rate": 3.1231953942081385e-06,
+      "loss": 0.5118,
+      "step": 14388
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 43.15730940659971,
+      "learning_rate": 3.1219711435976207e-06,
+      "loss": 0.4735,
+      "step": 14389
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 53.58797751165671,
+      "learning_rate": 3.1207470885942213e-06,
+      "loss": 0.5174,
+      "step": 14390
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 37.381154638059385,
+      "learning_rate": 3.119523229232746e-06,
+      "loss": 0.41,
+      "step": 14391
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 48.22335708869044,
+      "learning_rate": 3.1182995655480065e-06,
+      "loss": 0.526,
+      "step": 14392
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 51.943573519714796,
+      "learning_rate": 3.1170760975748006e-06,
+      "loss": 0.5743,
+      "step": 14393
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 52.65775992834951,
+      "learning_rate": 3.115852825347927e-06,
+      "loss": 0.5561,
+      "step": 14394
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 38.566026396548445,
+      "learning_rate": 3.114629748902173e-06,
+      "loss": 0.5041,
+      "step": 14395
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 49.93934315704373,
+      "learning_rate": 3.1134068682723173e-06,
+      "loss": 0.4644,
+      "step": 14396
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 48.75743522675339,
+      "learning_rate": 3.112184183493142e-06,
+      "loss": 0.4789,
+      "step": 14397
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 46.23134448145069,
+      "learning_rate": 3.1109616945994225e-06,
+      "loss": 0.4935,
+      "step": 14398
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 42.4978365656033,
+      "learning_rate": 3.109739401625922e-06,
+      "loss": 0.4653,
+      "step": 14399
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.1036929287013,
+      "learning_rate": 3.108517304607399e-06,
+      "loss": 0.5254,
+      "step": 14400
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.95800054446399,
+      "learning_rate": 3.107295403578613e-06,
+      "loss": 0.4846,
+      "step": 14401
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 47.316746172222885,
+      "learning_rate": 3.1060736985743166e-06,
+      "loss": 0.5353,
+      "step": 14402
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 36.617879325072394,
+      "learning_rate": 3.1048521896292483e-06,
+      "loss": 0.4145,
+      "step": 14403
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 43.141005931969936,
+      "learning_rate": 3.103630876778153e-06,
+      "loss": 0.4868,
+      "step": 14404
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 36.03815011011345,
+      "learning_rate": 3.102409760055758e-06,
+      "loss": 0.4334,
+      "step": 14405
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.66233420646053,
+      "learning_rate": 3.101188839496795e-06,
+      "loss": 0.5073,
+      "step": 14406
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.31548447505631,
+      "learning_rate": 3.099968115135988e-06,
+      "loss": 0.4626,
+      "step": 14407
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 39.580728902297444,
+      "learning_rate": 3.098747587008053e-06,
+      "loss": 0.444,
+      "step": 14408
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 50.31010453107359,
+      "learning_rate": 3.0975272551476953e-06,
+      "loss": 0.5009,
+      "step": 14409
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 48.478969379587,
+      "learning_rate": 3.0963071195896243e-06,
+      "loss": 0.484,
+      "step": 14410
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 40.33736186492976,
+      "learning_rate": 3.0950871803685435e-06,
+      "loss": 0.4426,
+      "step": 14411
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 35.18844653020232,
+      "learning_rate": 3.0938674375191415e-06,
+      "loss": 0.4325,
+      "step": 14412
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 46.81890088923895,
+      "learning_rate": 3.0926478910761127e-06,
+      "loss": 0.5071,
+      "step": 14413
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 36.21532847294358,
+      "learning_rate": 3.091428541074134e-06,
+      "loss": 0.4819,
+      "step": 14414
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 43.67410916804852,
+      "learning_rate": 3.090209387547889e-06,
+      "loss": 0.4846,
+      "step": 14415
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 34.34790480441457,
+      "learning_rate": 3.088990430532044e-06,
+      "loss": 0.4079,
+      "step": 14416
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.14320066786157,
+      "learning_rate": 3.0877716700612683e-06,
+      "loss": 0.5001,
+      "step": 14417
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 32.41939838556032,
+      "learning_rate": 3.086553106170226e-06,
+      "loss": 0.4351,
+      "step": 14418
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 40.769000108516906,
+      "learning_rate": 3.085334738893565e-06,
+      "loss": 0.4664,
+      "step": 14419
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 47.976749824191764,
+      "learning_rate": 3.0841165682659436e-06,
+      "loss": 0.4227,
+      "step": 14420
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 43.29865678626599,
+      "learning_rate": 3.082898594321998e-06,
+      "loss": 0.4233,
+      "step": 14421
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 40.91834082386491,
+      "learning_rate": 3.0816808170963697e-06,
+      "loss": 0.4886,
+      "step": 14422
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 46.43411893075061,
+      "learning_rate": 3.080463236623694e-06,
+      "loss": 0.5177,
+      "step": 14423
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 39.38190698746133,
+      "learning_rate": 3.079245852938597e-06,
+      "loss": 0.4664,
+      "step": 14424
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 41.24128140297645,
+      "learning_rate": 3.0780286660756964e-06,
+      "loss": 0.4459,
+      "step": 14425
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 41.83208590746251,
+      "learning_rate": 3.0768116760696108e-06,
+      "loss": 0.4001,
+      "step": 14426
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 36.723656077428934,
+      "learning_rate": 3.0755948829549552e-06,
+      "loss": 0.4154,
+      "step": 14427
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.69294356689618,
+      "learning_rate": 3.0743782867663297e-06,
+      "loss": 0.5387,
+      "step": 14428
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 35.621814655648066,
+      "learning_rate": 3.0731618875383318e-06,
+      "loss": 0.3975,
+      "step": 14429
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 48.45616184178873,
+      "learning_rate": 3.071945685305557e-06,
+      "loss": 0.5186,
+      "step": 14430
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 42.67550645962375,
+      "learning_rate": 3.0707296801025954e-06,
+      "loss": 0.4662,
+      "step": 14431
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 37.13616834132105,
+      "learning_rate": 3.0695138719640307e-06,
+      "loss": 0.4095,
+      "step": 14432
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 45.12776651117885,
+      "learning_rate": 3.068298260924436e-06,
+      "loss": 0.438,
+      "step": 14433
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 41.29411310146272,
+      "learning_rate": 3.0670828470183823e-06,
+      "loss": 0.4608,
+      "step": 14434
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 48.11819340514892,
+      "learning_rate": 3.065867630280436e-06,
+      "loss": 0.4728,
+      "step": 14435
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 55.29023294906542,
+      "learning_rate": 3.0646526107451615e-06,
+      "loss": 0.5498,
+      "step": 14436
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 37.17817030807163,
+      "learning_rate": 3.0634377884471055e-06,
+      "loss": 0.418,
+      "step": 14437
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 44.95753837615631,
+      "learning_rate": 3.0622231634208254e-06,
+      "loss": 0.4414,
+      "step": 14438
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 49.69919930682334,
+      "learning_rate": 3.0610087357008567e-06,
+      "loss": 0.474,
+      "step": 14439
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 40.43847912718808,
+      "learning_rate": 3.0597945053217435e-06,
+      "loss": 0.4665,
+      "step": 14440
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.38954406864473,
+      "learning_rate": 3.0585804723180115e-06,
+      "loss": 0.4727,
+      "step": 14441
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 49.992653001781115,
+      "learning_rate": 3.0573666367241893e-06,
+      "loss": 0.5387,
+      "step": 14442
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 40.71011457414932,
+      "learning_rate": 3.056152998574803e-06,
+      "loss": 0.4926,
+      "step": 14443
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.8543535707286,
+      "learning_rate": 3.0549395579043593e-06,
+      "loss": 0.4575,
+      "step": 14444
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 38.81242492533378,
+      "learning_rate": 3.053726314747376e-06,
+      "loss": 0.4259,
+      "step": 14445
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 54.067318550539504,
+      "learning_rate": 3.0525132691383486e-06,
+      "loss": 0.476,
+      "step": 14446
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 40.17406501275478,
+      "learning_rate": 3.0513004211117805e-06,
+      "loss": 0.4214,
+      "step": 14447
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 40.98011194498682,
+      "learning_rate": 3.050087770702167e-06,
+      "loss": 0.4544,
+      "step": 14448
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 36.94407856639745,
+      "learning_rate": 3.048875317943991e-06,
+      "loss": 0.456,
+      "step": 14449
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 48.903656429465194,
+      "learning_rate": 3.047663062871731e-06,
+      "loss": 0.4957,
+      "step": 14450
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 42.64536526598143,
+      "learning_rate": 3.046451005519867e-06,
+      "loss": 0.5109,
+      "step": 14451
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 53.50844448505607,
+      "learning_rate": 3.0452391459228726e-06,
+      "loss": 0.4814,
+      "step": 14452
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.70252332385722,
+      "learning_rate": 3.0440274841152072e-06,
+      "loss": 0.5025,
+      "step": 14453
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 35.27781530780415,
+      "learning_rate": 3.0428160201313295e-06,
+      "loss": 0.475,
+      "step": 14454
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 41.97689323926874,
+      "learning_rate": 3.0416047540056924e-06,
+      "loss": 0.4606,
+      "step": 14455
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.15512419937218,
+      "learning_rate": 3.040393685772747e-06,
+      "loss": 0.5021,
+      "step": 14456
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 46.86070614380802,
+      "learning_rate": 3.0391828154669367e-06,
+      "loss": 0.4834,
+      "step": 14457
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.80238098908242,
+      "learning_rate": 3.037972143122695e-06,
+      "loss": 0.497,
+      "step": 14458
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 39.93091791981829,
+      "learning_rate": 3.0367616687744504e-06,
+      "loss": 0.4531,
+      "step": 14459
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 40.66360679759555,
+      "learning_rate": 3.035551392456629e-06,
+      "loss": 0.4147,
+      "step": 14460
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 46.63458258819032,
+      "learning_rate": 3.0343413142036573e-06,
+      "loss": 0.5758,
+      "step": 14461
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 48.90079911052963,
+      "learning_rate": 3.0331314340499416e-06,
+      "loss": 0.5034,
+      "step": 14462
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 48.6805010748882,
+      "learning_rate": 3.0319217520298904e-06,
+      "loss": 0.4952,
+      "step": 14463
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 51.98907764973326,
+      "learning_rate": 3.0307122681779088e-06,
+      "loss": 0.5073,
+      "step": 14464
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 46.66321544941793,
+      "learning_rate": 3.0295029825283952e-06,
+      "loss": 0.4929,
+      "step": 14465
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 42.28140315713936,
+      "learning_rate": 3.028293895115737e-06,
+      "loss": 0.411,
+      "step": 14466
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 39.67845247122577,
+      "learning_rate": 3.027085005974325e-06,
+      "loss": 0.4632,
+      "step": 14467
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 49.471668665772924,
+      "learning_rate": 3.025876315138533e-06,
+      "loss": 0.515,
+      "step": 14468
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 49.024305716193666,
+      "learning_rate": 3.0246678226427396e-06,
+      "loss": 0.5641,
+      "step": 14469
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 42.630316662830374,
+      "learning_rate": 3.0234595285213154e-06,
+      "loss": 0.4845,
+      "step": 14470
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 45.77512451698706,
+      "learning_rate": 3.022251432808618e-06,
+      "loss": 0.4285,
+      "step": 14471
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 44.07518539714281,
+      "learning_rate": 3.021043535539011e-06,
+      "loss": 0.4974,
+      "step": 14472
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 39.39420027889329,
+      "learning_rate": 3.0198358367468407e-06,
+      "loss": 0.4487,
+      "step": 14473
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.82215870113414,
+      "learning_rate": 3.018628336466458e-06,
+      "loss": 0.5406,
+      "step": 14474
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 38.055460775370264,
+      "learning_rate": 3.0174210347322e-06,
+      "loss": 0.4368,
+      "step": 14475
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 49.19643685417709,
+      "learning_rate": 3.016213931578401e-06,
+      "loss": 0.454,
+      "step": 14476
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 44.642784645893634,
+      "learning_rate": 3.0150070270393973e-06,
+      "loss": 0.4503,
+      "step": 14477
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 38.516640742219074,
+      "learning_rate": 3.013800321149506e-06,
+      "loss": 0.4179,
+      "step": 14478
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 36.330568909835655,
+      "learning_rate": 3.012593813943044e-06,
+      "loss": 0.3998,
+      "step": 14479
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 39.51513117688685,
+      "learning_rate": 3.0113875054543273e-06,
+      "loss": 0.3998,
+      "step": 14480
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 31.759329934366434,
+      "learning_rate": 3.0101813957176617e-06,
+      "loss": 0.406,
+      "step": 14481
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 47.765719697238836,
+      "learning_rate": 3.008975484767351e-06,
+      "loss": 0.4484,
+      "step": 14482
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.87929165600509,
+      "learning_rate": 3.0077697726376873e-06,
+      "loss": 0.4501,
+      "step": 14483
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 33.40218245719732,
+      "learning_rate": 3.0065642593629597e-06,
+      "loss": 0.3754,
+      "step": 14484
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 36.4644993100196,
+      "learning_rate": 3.005358944977452e-06,
+      "loss": 0.4173,
+      "step": 14485
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 38.768986585043706,
+      "learning_rate": 3.004153829515447e-06,
+      "loss": 0.4453,
+      "step": 14486
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 48.82201006632043,
+      "learning_rate": 3.0029489130112155e-06,
+      "loss": 0.4254,
+      "step": 14487
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 40.64766670535881,
+      "learning_rate": 3.00174419549902e-06,
+      "loss": 0.4136,
+      "step": 14488
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 48.588594669761974,
+      "learning_rate": 3.0005396770131266e-06,
+      "loss": 0.4177,
+      "step": 14489
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 37.68521638217718,
+      "learning_rate": 2.9993353575877926e-06,
+      "loss": 0.3926,
+      "step": 14490
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 36.299019390974436,
+      "learning_rate": 2.9981312372572624e-06,
+      "loss": 0.461,
+      "step": 14491
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 47.13947991436674,
+      "learning_rate": 2.9969273160557877e-06,
+      "loss": 0.4528,
+      "step": 14492
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 46.468694901064936,
+      "learning_rate": 2.9957235940176e-06,
+      "loss": 0.506,
+      "step": 14493
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 51.39458885691233,
+      "learning_rate": 2.9945200711769362e-06,
+      "loss": 0.4917,
+      "step": 14494
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 39.27792222001547,
+      "learning_rate": 2.993316747568026e-06,
+      "loss": 0.4667,
+      "step": 14495
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 44.50680572458519,
+      "learning_rate": 2.992113623225086e-06,
+      "loss": 0.525,
+      "step": 14496
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 49.95997331656728,
+      "learning_rate": 2.990910698182338e-06,
+      "loss": 0.4751,
+      "step": 14497
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 42.379683531128194,
+      "learning_rate": 2.9897079724739866e-06,
+      "loss": 0.4598,
+      "step": 14498
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 39.13214188433511,
+      "learning_rate": 2.9885054461342423e-06,
+      "loss": 0.4578,
+      "step": 14499
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 42.081850569006164,
+      "learning_rate": 2.987303119197299e-06,
+      "loss": 0.4176,
+      "step": 14500
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 40.40137390312225,
+      "learning_rate": 2.986100991697355e-06,
+      "loss": 0.4388,
+      "step": 14501
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 39.98399481046658,
+      "learning_rate": 2.9848990636685937e-06,
+      "loss": 0.4513,
+      "step": 14502
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 36.279150410044615,
+      "learning_rate": 2.9836973351452014e-06,
+      "loss": 0.4701,
+      "step": 14503
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 39.38956959939853,
+      "learning_rate": 2.9824958061613497e-06,
+      "loss": 0.4608,
+      "step": 14504
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 46.268693007367474,
+      "learning_rate": 2.9812944767512132e-06,
+      "loss": 0.467,
+      "step": 14505
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 43.677928230719615,
+      "learning_rate": 2.980093346948958e-06,
+      "loss": 0.4825,
+      "step": 14506
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 53.78930662234113,
+      "learning_rate": 2.978892416788738e-06,
+      "loss": 0.4743,
+      "step": 14507
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 35.87525823546719,
+      "learning_rate": 2.9776916863047146e-06,
+      "loss": 0.4263,
+      "step": 14508
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 48.203892523344635,
+      "learning_rate": 2.976491155531027e-06,
+      "loss": 0.5006,
+      "step": 14509
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 45.742839669807765,
+      "learning_rate": 2.975290824501824e-06,
+      "loss": 0.454,
+      "step": 14510
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.19302056786239,
+      "learning_rate": 2.9740906932512427e-06,
+      "loss": 0.3841,
+      "step": 14511
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 36.085009222485354,
+      "learning_rate": 2.9728907618134127e-06,
+      "loss": 0.4318,
+      "step": 14512
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.839943394289875,
+      "learning_rate": 2.9716910302224556e-06,
+      "loss": 0.453,
+      "step": 14513
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 60.48784872460239,
+      "learning_rate": 2.970491498512493e-06,
+      "loss": 0.5204,
+      "step": 14514
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 49.02752858633821,
+      "learning_rate": 2.9692921667176443e-06,
+      "loss": 0.457,
+      "step": 14515
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 44.56052353637004,
+      "learning_rate": 2.968093034872014e-06,
+      "loss": 0.441,
+      "step": 14516
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 45.34608852526498,
+      "learning_rate": 2.9668941030097e-06,
+      "loss": 0.4688,
+      "step": 14517
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.67210043126375,
+      "learning_rate": 2.965695371164804e-06,
+      "loss": 0.4147,
+      "step": 14518
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 43.69556074312362,
+      "learning_rate": 2.964496839371417e-06,
+      "loss": 0.4381,
+      "step": 14519
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 37.661703283123295,
+      "learning_rate": 2.963298507663627e-06,
+      "loss": 0.4045,
+      "step": 14520
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 35.35322228344724,
+      "learning_rate": 2.9621003760755105e-06,
+      "loss": 0.3989,
+      "step": 14521
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 39.00873237440532,
+      "learning_rate": 2.9609024446411394e-06,
+      "loss": 0.4688,
+      "step": 14522
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 37.51834336504813,
+      "learning_rate": 2.9597047133945855e-06,
+      "loss": 0.4726,
+      "step": 14523
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 43.54799336125525,
+      "learning_rate": 2.9585071823699143e-06,
+      "loss": 0.4559,
+      "step": 14524
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 46.74296983092545,
+      "learning_rate": 2.9573098516011756e-06,
+      "loss": 0.4412,
+      "step": 14525
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 50.057361972521186,
+      "learning_rate": 2.9561127211224283e-06,
+      "loss": 0.5099,
+      "step": 14526
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 38.95747663334555,
+      "learning_rate": 2.9549157909677116e-06,
+      "loss": 0.3786,
+      "step": 14527
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 54.314102904236506,
+      "learning_rate": 2.953719061171072e-06,
+      "loss": 0.4977,
+      "step": 14528
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 51.01317215027765,
+      "learning_rate": 2.9525225317665375e-06,
+      "loss": 0.478,
+      "step": 14529
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.729202292711676,
+      "learning_rate": 2.9513262027881396e-06,
+      "loss": 0.4107,
+      "step": 14530
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.04966484146736,
+      "learning_rate": 2.9501300742699036e-06,
+      "loss": 0.4651,
+      "step": 14531
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 61.92389204199139,
+      "learning_rate": 2.948934146245841e-06,
+      "loss": 0.5506,
+      "step": 14532
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 51.923117644288595,
+      "learning_rate": 2.94773841874997e-06,
+      "loss": 0.5285,
+      "step": 14533
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 51.48931719802319,
+      "learning_rate": 2.94654289181629e-06,
+      "loss": 0.5828,
+      "step": 14534
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 51.680870853407605,
+      "learning_rate": 2.9453475654788032e-06,
+      "loss": 0.4794,
+      "step": 14535
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 47.712885578087636,
+      "learning_rate": 2.9441524397715094e-06,
+      "loss": 0.444,
+      "step": 14536
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 50.33120582623155,
+      "learning_rate": 2.942957514728392e-06,
+      "loss": 0.5184,
+      "step": 14537
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 37.46733781318384,
+      "learning_rate": 2.9417627903834313e-06,
+      "loss": 0.4141,
+      "step": 14538
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 38.456860084598375,
+      "learning_rate": 2.940568266770607e-06,
+      "loss": 0.4496,
+      "step": 14539
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 40.15047635678261,
+      "learning_rate": 2.9393739439238955e-06,
+      "loss": 0.4241,
+      "step": 14540
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 47.548812611111785,
+      "learning_rate": 2.9381798218772594e-06,
+      "loss": 0.5397,
+      "step": 14541
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 52.77882821471026,
+      "learning_rate": 2.936985900664654e-06,
+      "loss": 0.5088,
+      "step": 14542
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 42.83783462945827,
+      "learning_rate": 2.9357921803200386e-06,
+      "loss": 0.4547,
+      "step": 14543
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 52.30197332517897,
+      "learning_rate": 2.93459866087736e-06,
+      "loss": 0.5403,
+      "step": 14544
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 40.89745030418598,
+      "learning_rate": 2.9334053423705666e-06,
+      "loss": 0.4152,
+      "step": 14545
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 43.102634955224175,
+      "learning_rate": 2.9322122248335905e-06,
+      "loss": 0.4786,
+      "step": 14546
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 37.951810082955106,
+      "learning_rate": 2.9310193083003614e-06,
+      "loss": 0.4468,
+      "step": 14547
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 38.817251863623184,
+      "learning_rate": 2.9298265928048086e-06,
+      "loss": 0.4229,
+      "step": 14548
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 39.93143108981846,
+      "learning_rate": 2.9286340783808543e-06,
+      "loss": 0.4455,
+      "step": 14549
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 37.207502899784956,
+      "learning_rate": 2.9274417650624097e-06,
+      "loss": 0.3989,
+      "step": 14550
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 35.997536154383845,
+      "learning_rate": 2.926249652883382e-06,
+      "loss": 0.4419,
+      "step": 14551
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 40.07385807635459,
+      "learning_rate": 2.925057741877675e-06,
+      "loss": 0.4562,
+      "step": 14552
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 42.474398026177575,
+      "learning_rate": 2.92386603207919e-06,
+      "loss": 0.5084,
+      "step": 14553
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 49.14519098898902,
+      "learning_rate": 2.9226745235218134e-06,
+      "loss": 0.4777,
+      "step": 14554
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 42.27953515589433,
+      "learning_rate": 2.921483216239436e-06,
+      "loss": 0.4617,
+      "step": 14555
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 49.379832873851505,
+      "learning_rate": 2.920292110265932e-06,
+      "loss": 0.4656,
+      "step": 14556
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 38.8476005726411,
+      "learning_rate": 2.9191012056351787e-06,
+      "loss": 0.4206,
+      "step": 14557
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 50.19226071591348,
+      "learning_rate": 2.917910502381048e-06,
+      "loss": 0.4673,
+      "step": 14558
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 39.67597983459789,
+      "learning_rate": 2.9167200005373976e-06,
+      "loss": 0.4378,
+      "step": 14559
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 42.49807654684418,
+      "learning_rate": 2.915529700138089e-06,
+      "loss": 0.4259,
+      "step": 14560
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 38.64665180412366,
+      "learning_rate": 2.9143396012169677e-06,
+      "loss": 0.4317,
+      "step": 14561
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 46.61927977645904,
+      "learning_rate": 2.913149703807887e-06,
+      "loss": 0.4568,
+      "step": 14562
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.74184640198497,
+      "learning_rate": 2.9119600079446807e-06,
+      "loss": 0.4097,
+      "step": 14563
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.617021385965586,
+      "learning_rate": 2.910770513661184e-06,
+      "loss": 0.4607,
+      "step": 14564
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 45.14592659234573,
+      "learning_rate": 2.909581220991231e-06,
+      "loss": 0.4345,
+      "step": 14565
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 43.32393533264686,
+      "learning_rate": 2.9083921299686403e-06,
+      "loss": 0.431,
+      "step": 14566
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 43.277068906369784,
+      "learning_rate": 2.9072032406272243e-06,
+      "loss": 0.4476,
+      "step": 14567
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 46.6002868170773,
+      "learning_rate": 2.9060145530008013e-06,
+      "loss": 0.5484,
+      "step": 14568
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 45.510953140269194,
+      "learning_rate": 2.9048260671231733e-06,
+      "loss": 0.4118,
+      "step": 14569
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 43.69651941998973,
+      "learning_rate": 2.9036377830281435e-06,
+      "loss": 0.427,
+      "step": 14570
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 45.71856657083966,
+      "learning_rate": 2.902449700749506e-06,
+      "loss": 0.5006,
+      "step": 14571
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 39.25540468337992,
+      "learning_rate": 2.901261820321042e-06,
+      "loss": 0.4359,
+      "step": 14572
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 43.87861822701584,
+      "learning_rate": 2.90007414177654e-06,
+      "loss": 0.4809,
+      "step": 14573
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 49.85837299771043,
+      "learning_rate": 2.8988866651497793e-06,
+      "loss": 0.4417,
+      "step": 14574
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 38.388083585366914,
+      "learning_rate": 2.8976993904745277e-06,
+      "loss": 0.4602,
+      "step": 14575
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 49.445263586985995,
+      "learning_rate": 2.8965123177845476e-06,
+      "loss": 0.4622,
+      "step": 14576
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.344106648762526,
+      "learning_rate": 2.8953254471136017e-06,
+      "loss": 0.3916,
+      "step": 14577
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 39.58322900390261,
+      "learning_rate": 2.8941387784954476e-06,
+      "loss": 0.4242,
+      "step": 14578
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 44.06747184461228,
+      "learning_rate": 2.8929523119638257e-06,
+      "loss": 0.4819,
+      "step": 14579
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 39.87375522930755,
+      "learning_rate": 2.8917660475524867e-06,
+      "loss": 0.433,
+      "step": 14580
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 47.94077586254774,
+      "learning_rate": 2.89057998529516e-06,
+      "loss": 0.4367,
+      "step": 14581
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.516535504232394,
+      "learning_rate": 2.8893941252255784e-06,
+      "loss": 0.4684,
+      "step": 14582
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 46.291663758145894,
+      "learning_rate": 2.8882084673774722e-06,
+      "loss": 0.368,
+      "step": 14583
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.98535333422734,
+      "learning_rate": 2.8870230117845543e-06,
+      "loss": 0.4877,
+      "step": 14584
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.79388391655702,
+      "learning_rate": 2.885837758480543e-06,
+      "loss": 0.467,
+      "step": 14585
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 38.66736666207456,
+      "learning_rate": 2.8846527074991427e-06,
+      "loss": 0.4667,
+      "step": 14586
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 44.35699239353858,
+      "learning_rate": 2.8834678588740585e-06,
+      "loss": 0.4416,
+      "step": 14587
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 39.08986493267615,
+      "learning_rate": 2.8822832126389834e-06,
+      "loss": 0.423,
+      "step": 14588
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 39.13516698613283,
+      "learning_rate": 2.8810987688276092e-06,
+      "loss": 0.4308,
+      "step": 14589
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.00755449882341,
+      "learning_rate": 2.8799145274736253e-06,
+      "loss": 0.49,
+      "step": 14590
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 46.591581062579856,
+      "learning_rate": 2.878730488610706e-06,
+      "loss": 0.4277,
+      "step": 14591
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 51.3854501333488,
+      "learning_rate": 2.8775466522725238e-06,
+      "loss": 0.4619,
+      "step": 14592
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 56.178014240961005,
+      "learning_rate": 2.876363018492747e-06,
+      "loss": 0.5775,
+      "step": 14593
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 55.091169867032846,
+      "learning_rate": 2.8751795873050405e-06,
+      "loss": 0.5062,
+      "step": 14594
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.67728971611218,
+      "learning_rate": 2.873996358743061e-06,
+      "loss": 0.4934,
+      "step": 14595
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 48.25922334302972,
+      "learning_rate": 2.872813332840456e-06,
+      "loss": 0.4984,
+      "step": 14596
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.34733658694483,
+      "learning_rate": 2.8716305096308674e-06,
+      "loss": 0.4561,
+      "step": 14597
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 57.794611713020245,
+      "learning_rate": 2.870447889147938e-06,
+      "loss": 0.5447,
+      "step": 14598
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 40.71727412748291,
+      "learning_rate": 2.869265471425302e-06,
+      "loss": 0.4248,
+      "step": 14599
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.80947100355099,
+      "learning_rate": 2.8680832564965853e-06,
+      "loss": 0.4703,
+      "step": 14600
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 64.9433405785701,
+      "learning_rate": 2.866901244395406e-06,
+      "loss": 0.53,
+      "step": 14601
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.05252223379937,
+      "learning_rate": 2.8657194351553817e-06,
+      "loss": 0.5413,
+      "step": 14602
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.61151219524187,
+      "learning_rate": 2.8645378288101267e-06,
+      "loss": 0.507,
+      "step": 14603
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 40.35870419844568,
+      "learning_rate": 2.863356425393242e-06,
+      "loss": 0.4729,
+      "step": 14604
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 40.190488010105106,
+      "learning_rate": 2.862175224938323e-06,
+      "loss": 0.4478,
+      "step": 14605
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 40.58473214056735,
+      "learning_rate": 2.860994227478964e-06,
+      "loss": 0.4208,
+      "step": 14606
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.88859736747695,
+      "learning_rate": 2.8598134330487536e-06,
+      "loss": 0.4478,
+      "step": 14607
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 45.49683127006495,
+      "learning_rate": 2.8586328416812746e-06,
+      "loss": 0.474,
+      "step": 14608
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 48.0870453620919,
+      "learning_rate": 2.857452453410101e-06,
+      "loss": 0.4963,
+      "step": 14609
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 36.38924587153923,
+      "learning_rate": 2.856272268268798e-06,
+      "loss": 0.4224,
+      "step": 14610
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 48.82979606027258,
+      "learning_rate": 2.8550922862909327e-06,
+      "loss": 0.4387,
+      "step": 14611
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 33.49959636203086,
+      "learning_rate": 2.853912507510066e-06,
+      "loss": 0.3528,
+      "step": 14612
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 49.03683883295172,
+      "learning_rate": 2.852732931959744e-06,
+      "loss": 0.4463,
+      "step": 14613
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 49.718215052366034,
+      "learning_rate": 2.85155355967352e-06,
+      "loss": 0.5033,
+      "step": 14614
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 39.85088558254036,
+      "learning_rate": 2.850374390684928e-06,
+      "loss": 0.4386,
+      "step": 14615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.83875461730282,
+      "learning_rate": 2.849195425027509e-06,
+      "loss": 0.4641,
+      "step": 14616
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 51.53245070188101,
+      "learning_rate": 2.8480166627347847e-06,
+      "loss": 0.4739,
+      "step": 14617
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 49.99712349246762,
+      "learning_rate": 2.8468381038402836e-06,
+      "loss": 0.4835,
+      "step": 14618
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 51.7800591285253,
+      "learning_rate": 2.845659748377526e-06,
+      "loss": 0.4468,
+      "step": 14619
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 49.085359275189255,
+      "learning_rate": 2.8444815963800164e-06,
+      "loss": 0.5091,
+      "step": 14620
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 43.50094999174881,
+      "learning_rate": 2.843303647881267e-06,
+      "loss": 0.4533,
+      "step": 14621
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 38.788999983476884,
+      "learning_rate": 2.842125902914773e-06,
+      "loss": 0.4194,
+      "step": 14622
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.63984187608244,
+      "learning_rate": 2.840948361514031e-06,
+      "loss": 0.4609,
+      "step": 14623
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.11953172367257,
+      "learning_rate": 2.839771023712532e-06,
+      "loss": 0.4882,
+      "step": 14624
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 45.42821505416212,
+      "learning_rate": 2.838593889543757e-06,
+      "loss": 0.5272,
+      "step": 14625
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 42.953100521542765,
+      "learning_rate": 2.8374169590411793e-06,
+      "loss": 0.481,
+      "step": 14626
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 33.35712455645153,
+      "learning_rate": 2.836240232238274e-06,
+      "loss": 0.4463,
+      "step": 14627
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 45.46826040522156,
+      "learning_rate": 2.8350637091685085e-06,
+      "loss": 0.487,
+      "step": 14628
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.717122460065404,
+      "learning_rate": 2.8338873898653394e-06,
+      "loss": 0.4469,
+      "step": 14629
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 38.02316539283768,
+      "learning_rate": 2.8327112743622176e-06,
+      "loss": 0.4575,
+      "step": 14630
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.15851624081726,
+      "learning_rate": 2.8315353626925943e-06,
+      "loss": 0.4448,
+      "step": 14631
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 42.06739816586334,
+      "learning_rate": 2.8303596548899127e-06,
+      "loss": 0.461,
+      "step": 14632
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 38.09269731364227,
+      "learning_rate": 2.829184150987612e-06,
+      "loss": 0.4322,
+      "step": 14633
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 57.358743911966634,
+      "learning_rate": 2.8280088510191174e-06,
+      "loss": 0.4484,
+      "step": 14634
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 57.78767361789163,
+      "learning_rate": 2.8268337550178537e-06,
+      "loss": 0.4893,
+      "step": 14635
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.49414154365226,
+      "learning_rate": 2.8256588630172412e-06,
+      "loss": 0.4701,
+      "step": 14636
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 36.49959925862672,
+      "learning_rate": 2.8244841750506967e-06,
+      "loss": 0.4364,
+      "step": 14637
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 34.67724098879422,
+      "learning_rate": 2.823309691151622e-06,
+      "loss": 0.4159,
+      "step": 14638
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.99343048776231,
+      "learning_rate": 2.8221354113534248e-06,
+      "loss": 0.4655,
+      "step": 14639
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 54.882011973560076,
+      "learning_rate": 2.8209613356894948e-06,
+      "loss": 0.6291,
+      "step": 14640
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 37.97552611869919,
+      "learning_rate": 2.8197874641932276e-06,
+      "loss": 0.5012,
+      "step": 14641
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.62601431820267,
+      "learning_rate": 2.818613796898001e-06,
+      "loss": 0.4927,
+      "step": 14642
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.69788405273578,
+      "learning_rate": 2.8174403338371968e-06,
+      "loss": 0.429,
+      "step": 14643
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 48.23275718533793,
+      "learning_rate": 2.8162670750441924e-06,
+      "loss": 0.45,
+      "step": 14644
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 46.73278238075589,
+      "learning_rate": 2.815094020552346e-06,
+      "loss": 0.4786,
+      "step": 14645
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 44.188855092422564,
+      "learning_rate": 2.8139211703950255e-06,
+      "loss": 0.4569,
+      "step": 14646
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 41.180313050597995,
+      "learning_rate": 2.81274852460558e-06,
+      "loss": 0.4215,
+      "step": 14647
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 48.34095695924645,
+      "learning_rate": 2.8115760832173657e-06,
+      "loss": 0.4821,
+      "step": 14648
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.459834740154484,
+      "learning_rate": 2.8104038462637195e-06,
+      "loss": 0.4366,
+      "step": 14649
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 40.354447008195386,
+      "learning_rate": 2.809231813777984e-06,
+      "loss": 0.4657,
+      "step": 14650
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 36.19540542271142,
+      "learning_rate": 2.8080599857934876e-06,
+      "loss": 0.3836,
+      "step": 14651
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 49.775048961069274,
+      "learning_rate": 2.806888362343557e-06,
+      "loss": 0.4598,
+      "step": 14652
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 54.45973598898847,
+      "learning_rate": 2.805716943461517e-06,
+      "loss": 0.5744,
+      "step": 14653
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 46.12635798639119,
+      "learning_rate": 2.8045457291806787e-06,
+      "loss": 0.4508,
+      "step": 14654
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.87169359795832,
+      "learning_rate": 2.8033747195343474e-06,
+      "loss": 0.4853,
+      "step": 14655
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 45.04593054885872,
+      "learning_rate": 2.802203914555829e-06,
+      "loss": 0.4918,
+      "step": 14656
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 49.445988436349275,
+      "learning_rate": 2.8010333142784217e-06,
+      "loss": 0.4923,
+      "step": 14657
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 35.75438827978244,
+      "learning_rate": 2.7998629187354178e-06,
+      "loss": 0.3918,
+      "step": 14658
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 48.043960696279825,
+      "learning_rate": 2.798692727960101e-06,
+      "loss": 0.4398,
+      "step": 14659
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.609302138850886,
+      "learning_rate": 2.7975227419857466e-06,
+      "loss": 0.5108,
+      "step": 14660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 39.49033925667826,
+      "learning_rate": 2.796352960845632e-06,
+      "loss": 0.4286,
+      "step": 14661
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.05230573400904,
+      "learning_rate": 2.795183384573029e-06,
+      "loss": 0.437,
+      "step": 14662
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 34.523853645656686,
+      "learning_rate": 2.7940140132011963e-06,
+      "loss": 0.3633,
+      "step": 14663
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 39.91063225805693,
+      "learning_rate": 2.792844846763385e-06,
+      "loss": 0.4316,
+      "step": 14664
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 48.96263857586446,
+      "learning_rate": 2.7916758852928514e-06,
+      "loss": 0.4813,
+      "step": 14665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 46.85293060590116,
+      "learning_rate": 2.7905071288228415e-06,
+      "loss": 0.4871,
+      "step": 14666
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.9090624616373,
+      "learning_rate": 2.789338577386589e-06,
+      "loss": 0.4676,
+      "step": 14667
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 40.30538218151521,
+      "learning_rate": 2.788170231017332e-06,
+      "loss": 0.4487,
+      "step": 14668
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.24546115961472,
+      "learning_rate": 2.787002089748291e-06,
+      "loss": 0.4832,
+      "step": 14669
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 41.61136827840705,
+      "learning_rate": 2.785834153612692e-06,
+      "loss": 0.4507,
+      "step": 14670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 45.34962873254316,
+      "learning_rate": 2.7846664226437524e-06,
+      "loss": 0.5208,
+      "step": 14671
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.521012010781256,
+      "learning_rate": 2.783498896874676e-06,
+      "loss": 0.474,
+      "step": 14672
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 45.23339587059468,
+      "learning_rate": 2.7823315763386715e-06,
+      "loss": 0.455,
+      "step": 14673
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 45.53012005536903,
+      "learning_rate": 2.7811644610689327e-06,
+      "loss": 0.4682,
+      "step": 14674
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 44.98629991945114,
+      "learning_rate": 2.7799975510986566e-06,
+      "loss": 0.4754,
+      "step": 14675
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 40.98978791757381,
+      "learning_rate": 2.7788308464610225e-06,
+      "loss": 0.4617,
+      "step": 14676
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 54.91677247569301,
+      "learning_rate": 2.777664347189216e-06,
+      "loss": 0.4373,
+      "step": 14677
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 41.64642283315181,
+      "learning_rate": 2.776498053316413e-06,
+      "loss": 0.4618,
+      "step": 14678
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 40.95725967860157,
+      "learning_rate": 2.7753319648757802e-06,
+      "loss": 0.4738,
+      "step": 14679
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.622917781422764,
+      "learning_rate": 2.7741660819004766e-06,
+      "loss": 0.4216,
+      "step": 14680
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 41.39737412238264,
+      "learning_rate": 2.7730004044236636e-06,
+      "loss": 0.4654,
+      "step": 14681
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 37.56677702865254,
+      "learning_rate": 2.7718349324784897e-06,
+      "loss": 0.4233,
+      "step": 14682
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 47.56148247250715,
+      "learning_rate": 2.770669666098107e-06,
+      "loss": 0.4194,
+      "step": 14683
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 45.90501268146299,
+      "learning_rate": 2.769504605315648e-06,
+      "loss": 0.5148,
+      "step": 14684
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 38.90063461163608,
+      "learning_rate": 2.7683397501642463e-06,
+      "loss": 0.3971,
+      "step": 14685
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 46.46982901576818,
+      "learning_rate": 2.767175100677032e-06,
+      "loss": 0.4834,
+      "step": 14686
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 36.63669486695314,
+      "learning_rate": 2.7660106568871293e-06,
+      "loss": 0.374,
+      "step": 14687
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 36.811881171888075,
+      "learning_rate": 2.7648464188276514e-06,
+      "loss": 0.4125,
+      "step": 14688
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 41.379368162498416,
+      "learning_rate": 2.7636823865317065e-06,
+      "loss": 0.4664,
+      "step": 14689
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.363452819974135,
+      "learning_rate": 2.7625185600324013e-06,
+      "loss": 0.4338,
+      "step": 14690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 56.79709522986447,
+      "learning_rate": 2.761354939362838e-06,
+      "loss": 0.4864,
+      "step": 14691
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 42.86072081185357,
+      "learning_rate": 2.7601915245561017e-06,
+      "loss": 0.4563,
+      "step": 14692
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.48513335897628,
+      "learning_rate": 2.7590283156452867e-06,
+      "loss": 0.4933,
+      "step": 14693
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 49.20948339997703,
+      "learning_rate": 2.757865312663468e-06,
+      "loss": 0.5057,
+      "step": 14694
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 44.68468503288901,
+      "learning_rate": 2.7567025156437223e-06,
+      "loss": 0.4436,
+      "step": 14695
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 44.807086931548916,
+      "learning_rate": 2.755539924619124e-06,
+      "loss": 0.4487,
+      "step": 14696
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 37.88310569481371,
+      "learning_rate": 2.754377539622731e-06,
+      "loss": 0.4054,
+      "step": 14697
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 36.67428628725015,
+      "learning_rate": 2.7532153606876e-06,
+      "loss": 0.4223,
+      "step": 14698
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.07890726489176,
+      "learning_rate": 2.7520533878467846e-06,
+      "loss": 0.5457,
+      "step": 14699
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 51.10451779074501,
+      "learning_rate": 2.7508916211333346e-06,
+      "loss": 0.4563,
+      "step": 14700
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 59.018106401839255,
+      "learning_rate": 2.7497300605802823e-06,
+      "loss": 0.4875,
+      "step": 14701
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 41.52330447415193,
+      "learning_rate": 2.74856870622067e-06,
+      "loss": 0.4209,
+      "step": 14702
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 39.1224799670009,
+      "learning_rate": 2.7474075580875182e-06,
+      "loss": 0.4373,
+      "step": 14703
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 59.47474023634766,
+      "learning_rate": 2.7462466162138557e-06,
+      "loss": 0.5381,
+      "step": 14704
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 49.49463127383475,
+      "learning_rate": 2.745085880632693e-06,
+      "loss": 0.4675,
+      "step": 14705
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 55.2352684283267,
+      "learning_rate": 2.7439253513770434e-06,
+      "loss": 0.5019,
+      "step": 14706
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 52.4511377243928,
+      "learning_rate": 2.7427650284799157e-06,
+      "loss": 0.5198,
+      "step": 14707
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 43.90137018528935,
+      "learning_rate": 2.7416049119743025e-06,
+      "loss": 0.4532,
+      "step": 14708
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 54.52081280266863,
+      "learning_rate": 2.7404450018932027e-06,
+      "loss": 0.5483,
+      "step": 14709
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.36544933323095,
+      "learning_rate": 2.739285298269596e-06,
+      "loss": 0.4304,
+      "step": 14710
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 38.31225641244739,
+      "learning_rate": 2.7381258011364697e-06,
+      "loss": 0.3768,
+      "step": 14711
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 40.344191816940906,
+      "learning_rate": 2.7369665105267995e-06,
+      "loss": 0.4605,
+      "step": 14712
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 51.59691599530679,
+      "learning_rate": 2.7358074264735534e-06,
+      "loss": 0.4199,
+      "step": 14713
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 58.8566745374522,
+      "learning_rate": 2.7346485490096908e-06,
+      "loss": 0.6164,
+      "step": 14714
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.40513130910716,
+      "learning_rate": 2.733489878168175e-06,
+      "loss": 0.4129,
+      "step": 14715
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.64155306252189,
+      "learning_rate": 2.732331413981958e-06,
+      "loss": 0.487,
+      "step": 14716
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 41.91035956728895,
+      "learning_rate": 2.731173156483985e-06,
+      "loss": 0.4291,
+      "step": 14717
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.11884175260583,
+      "learning_rate": 2.730015105707191e-06,
+      "loss": 0.4989,
+      "step": 14718
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.12733408597003,
+      "learning_rate": 2.7288572616845154e-06,
+      "loss": 0.4423,
+      "step": 14719
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 36.98001218912657,
+      "learning_rate": 2.7276996244488864e-06,
+      "loss": 0.4099,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 37.341424156127374,
+      "learning_rate": 2.7265421940332283e-06,
+      "loss": 0.4422,
+      "step": 14721
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 45.78130029148295,
+      "learning_rate": 2.725384970470455e-06,
+      "loss": 0.447,
+      "step": 14722
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 48.53484447601699,
+      "learning_rate": 2.724227953793475e-06,
+      "loss": 0.3985,
+      "step": 14723
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 40.9323581058021,
+      "learning_rate": 2.7230711440351952e-06,
+      "loss": 0.4704,
+      "step": 14724
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.57681846363919,
+      "learning_rate": 2.7219145412285187e-06,
+      "loss": 0.4685,
+      "step": 14725
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 49.18780614667959,
+      "learning_rate": 2.7207581454063314e-06,
+      "loss": 0.44,
+      "step": 14726
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 60.82628716809216,
+      "learning_rate": 2.7196019566015286e-06,
+      "loss": 0.4771,
+      "step": 14727
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 38.98627082327821,
+      "learning_rate": 2.7184459748469836e-06,
+      "loss": 0.4509,
+      "step": 14728
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 45.47213934886488,
+      "learning_rate": 2.7172902001755787e-06,
+      "loss": 0.4424,
+      "step": 14729
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 40.2378446101458,
+      "learning_rate": 2.7161346326201763e-06,
+      "loss": 0.4493,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.03908052357229,
+      "learning_rate": 2.7149792722136446e-06,
+      "loss": 0.3841,
+      "step": 14731
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 51.314654682562,
+      "learning_rate": 2.713824118988844e-06,
+      "loss": 0.4467,
+      "step": 14732
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 47.82990308196281,
+      "learning_rate": 2.7126691729786202e-06,
+      "loss": 0.4798,
+      "step": 14733
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 45.780720949208224,
+      "learning_rate": 2.711514434215825e-06,
+      "loss": 0.4891,
+      "step": 14734
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 53.62771605791466,
+      "learning_rate": 2.710359902733293e-06,
+      "loss": 0.5108,
+      "step": 14735
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 55.181770191426,
+      "learning_rate": 2.709205578563859e-06,
+      "loss": 0.4782,
+      "step": 14736
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 37.67464539361644,
+      "learning_rate": 2.7080514617403584e-06,
+      "loss": 0.4164,
+      "step": 14737
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 63.78723843846264,
+      "learning_rate": 2.706897552295609e-06,
+      "loss": 0.4976,
+      "step": 14738
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 44.385916750577366,
+      "learning_rate": 2.705743850262422e-06,
+      "loss": 0.4396,
+      "step": 14739
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 43.12658951082894,
+      "learning_rate": 2.704590355673614e-06,
+      "loss": 0.4537,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.40400212556071,
+      "learning_rate": 2.7034370685619894e-06,
+      "loss": 0.4234,
+      "step": 14741
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 44.62681402798878,
+      "learning_rate": 2.7022839889603514e-06,
+      "loss": 0.4418,
+      "step": 14742
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.5750418598849,
+      "learning_rate": 2.7011311169014807e-06,
+      "loss": 0.4601,
+      "step": 14743
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 55.91640675688523,
+      "learning_rate": 2.6999784524181727e-06,
+      "loss": 0.6281,
+      "step": 14744
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 39.13697030370868,
+      "learning_rate": 2.698825995543207e-06,
+      "loss": 0.4157,
+      "step": 14745
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 36.82527247999477,
+      "learning_rate": 2.6976737463093617e-06,
+      "loss": 0.4035,
+      "step": 14746
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 39.61992901024002,
+      "learning_rate": 2.696521704749403e-06,
+      "loss": 0.4341,
+      "step": 14747
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 39.3139439031103,
+      "learning_rate": 2.6953698708960916e-06,
+      "loss": 0.3993,
+      "step": 14748
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 50.79306176480803,
+      "learning_rate": 2.6942182447821884e-06,
+      "loss": 0.5091,
+      "step": 14749
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.3686522171371,
+      "learning_rate": 2.693066826440447e-06,
+      "loss": 0.4255,
+      "step": 14750
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 42.90688776930031,
+      "learning_rate": 2.691915615903611e-06,
+      "loss": 0.4655,
+      "step": 14751
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 45.03532447886093,
+      "learning_rate": 2.6907646132044174e-06,
+      "loss": 0.5245,
+      "step": 14752
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 64.60414459201895,
+      "learning_rate": 2.6896138183756015e-06,
+      "loss": 0.4466,
+      "step": 14753
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 57.277554789809116,
+      "learning_rate": 2.6884632314498947e-06,
+      "loss": 0.5897,
+      "step": 14754
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 46.10695581608224,
+      "learning_rate": 2.6873128524600144e-06,
+      "loss": 0.4687,
+      "step": 14755
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 44.120593018411256,
+      "learning_rate": 2.686162681438682e-06,
+      "loss": 0.3844,
+      "step": 14756
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 63.43200513973016,
+      "learning_rate": 2.6850127184186015e-06,
+      "loss": 0.5906,
+      "step": 14757
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 50.14659614706537,
+      "learning_rate": 2.6838629634324796e-06,
+      "loss": 0.4752,
+      "step": 14758
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 39.91501953028952,
+      "learning_rate": 2.6827134165130175e-06,
+      "loss": 0.4578,
+      "step": 14759
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 51.04566651015184,
+      "learning_rate": 2.681564077692903e-06,
+      "loss": 0.5169,
+      "step": 14760
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 62.431936999482275,
+      "learning_rate": 2.6804149470048293e-06,
+      "loss": 0.5303,
+      "step": 14761
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 51.17104113649226,
+      "learning_rate": 2.6792660244814683e-06,
+      "loss": 0.499,
+      "step": 14762
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 45.337952414738275,
+      "learning_rate": 2.6781173101555014e-06,
+      "loss": 0.4643,
+      "step": 14763
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.851476757879624,
+      "learning_rate": 2.6769688040595933e-06,
+      "loss": 0.5061,
+      "step": 14764
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 57.31495374212967,
+      "learning_rate": 2.6758205062264076e-06,
+      "loss": 0.5553,
+      "step": 14765
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 40.31868799015895,
+      "learning_rate": 2.6746724166886063e-06,
+      "loss": 0.3796,
+      "step": 14766
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 44.92513781299798,
+      "learning_rate": 2.6735245354788354e-06,
+      "loss": 0.4959,
+      "step": 14767
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 48.44648587461156,
+      "learning_rate": 2.6723768626297387e-06,
+      "loss": 0.4925,
+      "step": 14768
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 54.920019938367375,
+      "learning_rate": 2.671229398173957e-06,
+      "loss": 0.4689,
+      "step": 14769
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 42.84229757461144,
+      "learning_rate": 2.670082142144124e-06,
+      "loss": 0.4191,
+      "step": 14770
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 47.359576218245856,
+      "learning_rate": 2.66893509457287e-06,
+      "loss": 0.4258,
+      "step": 14771
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 47.392591124367165,
+      "learning_rate": 2.6677882554928126e-06,
+      "loss": 0.4469,
+      "step": 14772
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.759120241961135,
+      "learning_rate": 2.666641624936566e-06,
+      "loss": 0.4564,
+      "step": 14773
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 45.2145051156663,
+      "learning_rate": 2.665495202936741e-06,
+      "loss": 0.4688,
+      "step": 14774
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 50.09260020420839,
+      "learning_rate": 2.6643489895259454e-06,
+      "loss": 0.4783,
+      "step": 14775
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 40.505884714338,
+      "learning_rate": 2.6632029847367734e-06,
+      "loss": 0.4461,
+      "step": 14776
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 51.01048526781374,
+      "learning_rate": 2.662057188601813e-06,
+      "loss": 0.4718,
+      "step": 14777
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 39.643888887212924,
+      "learning_rate": 2.660911601153654e-06,
+      "loss": 0.5612,
+      "step": 14778
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 51.58993257370661,
+      "learning_rate": 2.659766222424879e-06,
+      "loss": 0.513,
+      "step": 14779
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 42.26359001520888,
+      "learning_rate": 2.658621052448055e-06,
+      "loss": 0.4568,
+      "step": 14780
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 45.476944757141744,
+      "learning_rate": 2.6574760912557575e-06,
+      "loss": 0.5169,
+      "step": 14781
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 43.657342122971045,
+      "learning_rate": 2.656331338880541e-06,
+      "loss": 0.4609,
+      "step": 14782
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 36.36170351369776,
+      "learning_rate": 2.6551867953549657e-06,
+      "loss": 0.4153,
+      "step": 14783
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 45.07268965560889,
+      "learning_rate": 2.6540424607115845e-06,
+      "loss": 0.4956,
+      "step": 14784
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 39.03402291764348,
+      "learning_rate": 2.6528983349829353e-06,
+      "loss": 0.4327,
+      "step": 14785
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 43.17452255066758,
+      "learning_rate": 2.651754418201563e-06,
+      "loss": 0.4333,
+      "step": 14786
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 43.3169974982405,
+      "learning_rate": 2.6506107103999924e-06,
+      "loss": 0.4224,
+      "step": 14787
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 31.74184128921987,
+      "learning_rate": 2.6494672116107577e-06,
+      "loss": 0.4132,
+      "step": 14788
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 44.22979678140727,
+      "learning_rate": 2.6483239218663724e-06,
+      "loss": 0.5345,
+      "step": 14789
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 46.85672895765013,
+      "learning_rate": 2.647180841199354e-06,
+      "loss": 0.5364,
+      "step": 14790
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 48.64876536070008,
+      "learning_rate": 2.6460379696422146e-06,
+      "loss": 0.4082,
+      "step": 14791
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.01003422299203,
+      "learning_rate": 2.6448953072274532e-06,
+      "loss": 0.4399,
+      "step": 14792
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 47.308245690588066,
+      "learning_rate": 2.6437528539875633e-06,
+      "loss": 0.5091,
+      "step": 14793
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 47.02500826456828,
+      "learning_rate": 2.6426106099550396e-06,
+      "loss": 0.5751,
+      "step": 14794
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 40.51238370464974,
+      "learning_rate": 2.6414685751623682e-06,
+      "loss": 0.5062,
+      "step": 14795
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.45236231522493,
+      "learning_rate": 2.6403267496420238e-06,
+      "loss": 0.4091,
+      "step": 14796
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 50.295251507849365,
+      "learning_rate": 2.639185133426484e-06,
+      "loss": 0.4849,
+      "step": 14797
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 46.54864501148823,
+      "learning_rate": 2.6380437265482107e-06,
+      "loss": 0.4762,
+      "step": 14798
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 51.193331436348124,
+      "learning_rate": 2.6369025290396664e-06,
+      "loss": 0.4955,
+      "step": 14799
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 44.76684634687921,
+      "learning_rate": 2.6357615409333092e-06,
+      "loss": 0.4444,
+      "step": 14800
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 43.29960136406273,
+      "learning_rate": 2.634620762261587e-06,
+      "loss": 0.4408,
+      "step": 14801
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 58.81709244864841,
+      "learning_rate": 2.6334801930569387e-06,
+      "loss": 0.5114,
+      "step": 14802
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.680723443068,
+      "learning_rate": 2.6323398333518047e-06,
+      "loss": 0.4235,
+      "step": 14803
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 42.38676699986382,
+      "learning_rate": 2.631199683178615e-06,
+      "loss": 0.5019,
+      "step": 14804
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.08417393028829,
+      "learning_rate": 2.6300597425698015e-06,
+      "loss": 0.4594,
+      "step": 14805
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 36.43609970110981,
+      "learning_rate": 2.6289200115577727e-06,
+      "loss": 0.4221,
+      "step": 14806
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 61.34170859968136,
+      "learning_rate": 2.627780490174946e-06,
+      "loss": 0.5151,
+      "step": 14807
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 45.558166415553785,
+      "learning_rate": 2.6266411784537294e-06,
+      "loss": 0.543,
+      "step": 14808
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 55.944908239712305,
+      "learning_rate": 2.625502076426528e-06,
+      "loss": 0.4881,
+      "step": 14809
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 50.28340439244046,
+      "learning_rate": 2.6243631841257332e-06,
+      "loss": 0.444,
+      "step": 14810
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 36.65994544560545,
+      "learning_rate": 2.6232245015837323e-06,
+      "loss": 0.401,
+      "step": 14811
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 35.5745889333002,
+      "learning_rate": 2.6220860288329108e-06,
+      "loss": 0.4006,
+      "step": 14812
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 48.233907956209165,
+      "learning_rate": 2.6209477659056503e-06,
+      "loss": 0.5068,
+      "step": 14813
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 36.71002146764682,
+      "learning_rate": 2.6198097128343147e-06,
+      "loss": 0.467,
+      "step": 14814
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 43.28251557961479,
+      "learning_rate": 2.618671869651278e-06,
+      "loss": 0.4327,
+      "step": 14815
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 52.190141668103664,
+      "learning_rate": 2.617534236388891e-06,
+      "loss": 0.4962,
+      "step": 14816
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 43.0540081476569,
+      "learning_rate": 2.616396813079516e-06,
+      "loss": 0.4245,
+      "step": 14817
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 40.895153350281376,
+      "learning_rate": 2.615259599755492e-06,
+      "loss": 0.5434,
+      "step": 14818
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.07474530327389,
+      "learning_rate": 2.614122596449166e-06,
+      "loss": 0.447,
+      "step": 14819
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 37.923343138062435,
+      "learning_rate": 2.6129858031928746e-06,
+      "loss": 0.4488,
+      "step": 14820
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.07295106826233,
+      "learning_rate": 2.611849220018944e-06,
+      "loss": 0.3721,
+      "step": 14821
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 45.76118175209427,
+      "learning_rate": 2.610712846959702e-06,
+      "loss": 0.5029,
+      "step": 14822
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.79141751320005,
+      "learning_rate": 2.609576684047461e-06,
+      "loss": 0.4377,
+      "step": 14823
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 31.889969426042047,
+      "learning_rate": 2.6084407313145354e-06,
+      "loss": 0.3878,
+      "step": 14824
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 49.059943337856815,
+      "learning_rate": 2.6073049887932357e-06,
+      "loss": 0.4801,
+      "step": 14825
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 52.051562851410644,
+      "learning_rate": 2.606169456515857e-06,
+      "loss": 0.5504,
+      "step": 14826
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.479684393820015,
+      "learning_rate": 2.6050341345146902e-06,
+      "loss": 0.4335,
+      "step": 14827
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 34.52877284697294,
+      "learning_rate": 2.6038990228220275e-06,
+      "loss": 0.3372,
+      "step": 14828
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 34.69178732417983,
+      "learning_rate": 2.6027641214701495e-06,
+      "loss": 0.3949,
+      "step": 14829
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.85179219299485,
+      "learning_rate": 2.601629430491338e-06,
+      "loss": 0.4186,
+      "step": 14830
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 51.26889794065696,
+      "learning_rate": 2.6004949499178522e-06,
+      "loss": 0.5346,
+      "step": 14831
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.94342498312316,
+      "learning_rate": 2.5993606797819616e-06,
+      "loss": 0.4798,
+      "step": 14832
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 44.52908838480037,
+      "learning_rate": 2.5982266201159236e-06,
+      "loss": 0.4687,
+      "step": 14833
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 48.702837808868,
+      "learning_rate": 2.5970927709519942e-06,
+      "loss": 0.459,
+      "step": 14834
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 53.20337419493541,
+      "learning_rate": 2.5959591323224152e-06,
+      "loss": 0.5189,
+      "step": 14835
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 39.568679423398066,
+      "learning_rate": 2.5948257042594248e-06,
+      "loss": 0.4431,
+      "step": 14836
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 38.865906587129516,
+      "learning_rate": 2.5936924867952596e-06,
+      "loss": 0.4019,
+      "step": 14837
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 45.898097995232895,
+      "learning_rate": 2.5925594799621503e-06,
+      "loss": 0.4438,
+      "step": 14838
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 37.754252971037886,
+      "learning_rate": 2.5914266837923153e-06,
+      "loss": 0.4124,
+      "step": 14839
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.08398846720731,
+      "learning_rate": 2.5902940983179694e-06,
+      "loss": 0.4898,
+      "step": 14840
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.35883020230258,
+      "learning_rate": 2.589161723571324e-06,
+      "loss": 0.3698,
+      "step": 14841
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 46.789930405961435,
+      "learning_rate": 2.5880295595845885e-06,
+      "loss": 0.5155,
+      "step": 14842
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 42.763659319513174,
+      "learning_rate": 2.586897606389952e-06,
+      "loss": 0.4409,
+      "step": 14843
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 48.405998219474924,
+      "learning_rate": 2.585765864019615e-06,
+      "loss": 0.4954,
+      "step": 14844
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 36.55211056200682,
+      "learning_rate": 2.5846343325057556e-06,
+      "loss": 0.421,
+      "step": 14845
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 39.420610871006666,
+      "learning_rate": 2.583503011880558e-06,
+      "loss": 0.4425,
+      "step": 14846
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 44.35209521856828,
+      "learning_rate": 2.5823719021761993e-06,
+      "loss": 0.4593,
+      "step": 14847
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 44.66280079891659,
+      "learning_rate": 2.5812410034248426e-06,
+      "loss": 0.4465,
+      "step": 14848
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 60.779600537824166,
+      "learning_rate": 2.5801103156586536e-06,
+      "loss": 0.4153,
+      "step": 14849
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 36.39691781209919,
+      "learning_rate": 2.5789798389097855e-06,
+      "loss": 0.4136,
+      "step": 14850
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.57693481756312,
+      "learning_rate": 2.5778495732103914e-06,
+      "loss": 0.456,
+      "step": 14851
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.393745068693384,
+      "learning_rate": 2.5767195185926098e-06,
+      "loss": 0.4399,
+      "step": 14852
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 42.445395163629875,
+      "learning_rate": 2.5755896750885834e-06,
+      "loss": 0.446,
+      "step": 14853
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 49.86271622334284,
+      "learning_rate": 2.5744600427304467e-06,
+      "loss": 0.5274,
+      "step": 14854
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 35.6314947298214,
+      "learning_rate": 2.573330621550323e-06,
+      "loss": 0.4175,
+      "step": 14855
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.41158279338192,
+      "learning_rate": 2.5722014115803284e-06,
+      "loss": 0.4288,
+      "step": 14856
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.7119677172994,
+      "learning_rate": 2.57107241285258e-06,
+      "loss": 0.3975,
+      "step": 14857
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.84209823478973,
+      "learning_rate": 2.569943625399187e-06,
+      "loss": 0.4768,
+      "step": 14858
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 43.103585064914306,
+      "learning_rate": 2.5688150492522533e-06,
+      "loss": 0.4403,
+      "step": 14859
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 46.49122517864966,
+      "learning_rate": 2.567686684443873e-06,
+      "loss": 0.4698,
+      "step": 14860
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.45897047966588,
+      "learning_rate": 2.5665585310061327e-06,
+      "loss": 0.4869,
+      "step": 14861
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 37.890233290702405,
+      "learning_rate": 2.5654305889711182e-06,
+      "loss": 0.4116,
+      "step": 14862
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 45.76663973208527,
+      "learning_rate": 2.564302858370913e-06,
+      "loss": 0.4573,
+      "step": 14863
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.45578941642613,
+      "learning_rate": 2.5631753392375846e-06,
+      "loss": 0.5274,
+      "step": 14864
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.58361144855177,
+      "learning_rate": 2.562048031603196e-06,
+      "loss": 0.4341,
+      "step": 14865
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.54425150151213,
+      "learning_rate": 2.5609209354998097e-06,
+      "loss": 0.4108,
+      "step": 14866
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 42.261601993799,
+      "learning_rate": 2.5597940509594845e-06,
+      "loss": 0.3621,
+      "step": 14867
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 54.50052019185981,
+      "learning_rate": 2.55866737801426e-06,
+      "loss": 0.5004,
+      "step": 14868
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 42.37567023363328,
+      "learning_rate": 2.557540916696185e-06,
+      "loss": 0.4677,
+      "step": 14869
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 35.206741868491804,
+      "learning_rate": 2.5564146670372904e-06,
+      "loss": 0.3675,
+      "step": 14870
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 47.45905472148829,
+      "learning_rate": 2.5552886290696075e-06,
+      "loss": 0.4602,
+      "step": 14871
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 45.1342497509779,
+      "learning_rate": 2.554162802825165e-06,
+      "loss": 0.4427,
+      "step": 14872
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 63.4228916129026,
+      "learning_rate": 2.553037188335973e-06,
+      "loss": 0.4708,
+      "step": 14873
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 48.15416113935475,
+      "learning_rate": 2.5519117856340503e-06,
+      "loss": 0.5067,
+      "step": 14874
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 51.24795762156105,
+      "learning_rate": 2.5507865947513966e-06,
+      "loss": 0.4746,
+      "step": 14875
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 51.18784216648829,
+      "learning_rate": 2.549661615720017e-06,
+      "loss": 0.5426,
+      "step": 14876
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 40.58255033044485,
+      "learning_rate": 2.5485368485719e-06,
+      "loss": 0.4446,
+      "step": 14877
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 46.700856023188365,
+      "learning_rate": 2.5474122933390355e-06,
+      "loss": 0.4568,
+      "step": 14878
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 43.66909791582057,
+      "learning_rate": 2.54628795005341e-06,
+      "loss": 0.4581,
+      "step": 14879
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 39.319795086308275,
+      "learning_rate": 2.545163818746994e-06,
+      "loss": 0.4502,
+      "step": 14880
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.00438011719101,
+      "learning_rate": 2.5440398994517546e-06,
+      "loss": 0.4076,
+      "step": 14881
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 41.09524579244892,
+      "learning_rate": 2.54291619219966e-06,
+      "loss": 0.3948,
+      "step": 14882
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 42.53548232076272,
+      "learning_rate": 2.541792697022666e-06,
+      "loss": 0.458,
+      "step": 14883
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 38.21795092017927,
+      "learning_rate": 2.540669413952728e-06,
+      "loss": 0.5005,
+      "step": 14884
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.22539120869214,
+      "learning_rate": 2.5395463430217883e-06,
+      "loss": 0.4719,
+      "step": 14885
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 49.98015855019873,
+      "learning_rate": 2.538423484261783e-06,
+      "loss": 0.5176,
+      "step": 14886
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 41.33420981482814,
+      "learning_rate": 2.537300837704648e-06,
+      "loss": 0.4494,
+      "step": 14887
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.77993859505691,
+      "learning_rate": 2.536178403382317e-06,
+      "loss": 0.4426,
+      "step": 14888
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 34.10720328993996,
+      "learning_rate": 2.535056181326704e-06,
+      "loss": 0.4019,
+      "step": 14889
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 43.21475356264211,
+      "learning_rate": 2.533934171569724e-06,
+      "loss": 0.454,
+      "step": 14890
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 42.86447218659995,
+      "learning_rate": 2.5328123741432887e-06,
+      "loss": 0.4356,
+      "step": 14891
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 41.5316592699355,
+      "learning_rate": 2.5316907890793007e-06,
+      "loss": 0.4171,
+      "step": 14892
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 43.65577083917237,
+      "learning_rate": 2.530569416409665e-06,
+      "loss": 0.4549,
+      "step": 14893
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 54.0636351890674,
+      "learning_rate": 2.529448256166259e-06,
+      "loss": 0.5402,
+      "step": 14894
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.09157264618362,
+      "learning_rate": 2.5283273083809746e-06,
+      "loss": 0.441,
+      "step": 14895
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 42.89474938857129,
+      "learning_rate": 2.52720657308569e-06,
+      "loss": 0.5023,
+      "step": 14896
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.79551773777783,
+      "learning_rate": 2.526086050312283e-06,
+      "loss": 0.5325,
+      "step": 14897
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.15213079028047,
+      "learning_rate": 2.5249657400926165e-06,
+      "loss": 0.45,
+      "step": 14898
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 39.98434625634267,
+      "learning_rate": 2.5238456424585477e-06,
+      "loss": 0.4569,
+      "step": 14899
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 44.24778676290174,
+      "learning_rate": 2.522725757441936e-06,
+      "loss": 0.4752,
+      "step": 14900
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 44.242083555271556,
+      "learning_rate": 2.5216060850746326e-06,
+      "loss": 0.4385,
+      "step": 14901
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 43.20367309884865,
+      "learning_rate": 2.520486625388475e-06,
+      "loss": 0.4865,
+      "step": 14902
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 40.246800810867484,
+      "learning_rate": 2.519367378415305e-06,
+      "loss": 0.3743,
+      "step": 14903
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 68.33487565308182,
+      "learning_rate": 2.5182483441869477e-06,
+      "loss": 0.5516,
+      "step": 14904
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.97774054120818,
+      "learning_rate": 2.5171295227352332e-06,
+      "loss": 0.4797,
+      "step": 14905
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 44.799604930796136,
+      "learning_rate": 2.516010914091975e-06,
+      "loss": 0.5041,
+      "step": 14906
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 40.21825147199768,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.4514,
+      "step": 14907
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 39.91498027469874,
+      "learning_rate": 2.5137743353580833e-06,
+      "loss": 0.4571,
+      "step": 14908
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 40.79523798720144,
+      "learning_rate": 2.512656365331053e-06,
+      "loss": 0.4524,
+      "step": 14909
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 50.18581848367954,
+      "learning_rate": 2.5115386082396996e-06,
+      "loss": 0.5243,
+      "step": 14910
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.220965143424614,
+      "learning_rate": 2.5104210641158045e-06,
+      "loss": 0.4063,
+      "step": 14911
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 44.64256694977311,
+      "learning_rate": 2.509303732991153e-06,
+      "loss": 0.4287,
+      "step": 14912
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.52389669636816,
+      "learning_rate": 2.508186614897524e-06,
+      "loss": 0.4559,
+      "step": 14913
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 39.17931226891058,
+      "learning_rate": 2.507069709866685e-06,
+      "loss": 0.415,
+      "step": 14914
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 47.48698601380545,
+      "learning_rate": 2.5059530179303983e-06,
+      "loss": 0.5425,
+      "step": 14915
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.527608048744675,
+      "learning_rate": 2.5048365391204233e-06,
+      "loss": 0.4602,
+      "step": 14916
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 38.4595601301882,
+      "learning_rate": 2.5037202734685117e-06,
+      "loss": 0.4236,
+      "step": 14917
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 36.75965426266854,
+      "learning_rate": 2.502604221006417e-06,
+      "loss": 0.4099,
+      "step": 14918
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 43.127372095963096,
+      "learning_rate": 2.5014883817658662e-06,
+      "loss": 0.47,
+      "step": 14919
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 41.02863703801981,
+      "learning_rate": 2.5003727557785997e-06,
+      "loss": 0.4902,
+      "step": 14920
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 34.8347652671815,
+      "learning_rate": 2.4992573430763447e-06,
+      "loss": 0.4248,
+      "step": 14921
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 47.09214376483437,
+      "learning_rate": 2.498142143690826e-06,
+      "loss": 0.4451,
+      "step": 14922
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 48.90675346432242,
+      "learning_rate": 2.4970271576537563e-06,
+      "loss": 0.4693,
+      "step": 14923
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.601215329203974,
+      "learning_rate": 2.4959123849968427e-06,
+      "loss": 0.5039,
+      "step": 14924
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 47.30785203424248,
+      "learning_rate": 2.4947978257517903e-06,
+      "loss": 0.4686,
+      "step": 14925
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 51.75907973555523,
+      "learning_rate": 2.4936834799503005e-06,
+      "loss": 0.5109,
+      "step": 14926
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 35.15637469799391,
+      "learning_rate": 2.4925693476240587e-06,
+      "loss": 0.3965,
+      "step": 14927
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 42.961636401795715,
+      "learning_rate": 2.4914554288047553e-06,
+      "loss": 0.4795,
+      "step": 14928
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 51.91547509331766,
+      "learning_rate": 2.4903417235240647e-06,
+      "loss": 0.5148,
+      "step": 14929
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.426655142845824,
+      "learning_rate": 2.489228231813665e-06,
+      "loss": 0.5374,
+      "step": 14930
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.585166217257814,
+      "learning_rate": 2.4881149537052183e-06,
+      "loss": 0.4458,
+      "step": 14931
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 35.55060714077144,
+      "learning_rate": 2.4870018892303883e-06,
+      "loss": 0.3933,
+      "step": 14932
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.96311567652785,
+      "learning_rate": 2.485889038420831e-06,
+      "loss": 0.4858,
+      "step": 14933
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 49.51068474346353,
+      "learning_rate": 2.4847764013081923e-06,
+      "loss": 0.4561,
+      "step": 14934
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 39.41820336357945,
+      "learning_rate": 2.483663977924119e-06,
+      "loss": 0.3774,
+      "step": 14935
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 48.49668439858014,
+      "learning_rate": 2.482551768300242e-06,
+      "loss": 0.4574,
+      "step": 14936
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 39.84624240933451,
+      "learning_rate": 2.4814397724681995e-06,
+      "loss": 0.4231,
+      "step": 14937
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 36.65167028204173,
+      "learning_rate": 2.4803279904596067e-06,
+      "loss": 0.3468,
+      "step": 14938
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 53.65571913903689,
+      "learning_rate": 2.479216422306091e-06,
+      "loss": 0.5218,
+      "step": 14939
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 51.8772040392978,
+      "learning_rate": 2.4781050680392572e-06,
+      "loss": 0.5347,
+      "step": 14940
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.508242322696795,
+      "learning_rate": 2.4769939276907163e-06,
+      "loss": 0.4652,
+      "step": 14941
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 46.01040365391517,
+      "learning_rate": 2.4758830012920688e-06,
+      "loss": 0.4932,
+      "step": 14942
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 38.10517725048478,
+      "learning_rate": 2.4747722888749082e-06,
+      "loss": 0.3908,
+      "step": 14943
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.77066157150625,
+      "learning_rate": 2.4736617904708184e-06,
+      "loss": 0.4625,
+      "step": 14944
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 43.88499201054461,
+      "learning_rate": 2.4725515061113835e-06,
+      "loss": 0.4382,
+      "step": 14945
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 41.10772520239676,
+      "learning_rate": 2.471441435828181e-06,
+      "loss": 0.4272,
+      "step": 14946
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 43.774155800988886,
+      "learning_rate": 2.470331579652784e-06,
+      "loss": 0.4565,
+      "step": 14947
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 53.12446385281356,
+      "learning_rate": 2.4692219376167504e-06,
+      "loss": 0.4555,
+      "step": 14948
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 49.498914620030774,
+      "learning_rate": 2.4681125097516376e-06,
+      "loss": 0.5255,
+      "step": 14949
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 37.06469878289036,
+      "learning_rate": 2.467003296088999e-06,
+      "loss": 0.4379,
+      "step": 14950
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 41.486767773489454,
+      "learning_rate": 2.465894296660383e-06,
+      "loss": 0.4466,
+      "step": 14951
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 46.83227778271307,
+      "learning_rate": 2.464785511497325e-06,
+      "loss": 0.4854,
+      "step": 14952
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 34.47590288040803,
+      "learning_rate": 2.4636769406313576e-06,
+      "loss": 0.4068,
+      "step": 14953
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 58.145325845626594,
+      "learning_rate": 2.4625685840940083e-06,
+      "loss": 0.5289,
+      "step": 14954
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 42.86200753310369,
+      "learning_rate": 2.4614604419167997e-06,
+      "loss": 0.3974,
+      "step": 14955
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 46.6957903526891,
+      "learning_rate": 2.460352514131249e-06,
+      "loss": 0.5019,
+      "step": 14956
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 38.79817526916396,
+      "learning_rate": 2.459244800768862e-06,
+      "loss": 0.3893,
+      "step": 14957
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 37.982457830078886,
+      "learning_rate": 2.458137301861139e-06,
+      "loss": 0.404,
+      "step": 14958
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 34.644034417925795,
+      "learning_rate": 2.457030017439579e-06,
+      "loss": 0.4168,
+      "step": 14959
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.72342872234596,
+      "learning_rate": 2.455922947535675e-06,
+      "loss": 0.5246,
+      "step": 14960
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 39.299241142250644,
+      "learning_rate": 2.4548160921809073e-06,
+      "loss": 0.4443,
+      "step": 14961
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 50.27276559660813,
+      "learning_rate": 2.4537094514067596e-06,
+      "loss": 0.4701,
+      "step": 14962
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 46.67235191773299,
+      "learning_rate": 2.4526030252446963e-06,
+      "loss": 0.4299,
+      "step": 14963
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 45.55372102301316,
+      "learning_rate": 2.4514968137261918e-06,
+      "loss": 0.4946,
+      "step": 14964
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 38.966741110896734,
+      "learning_rate": 2.4503908168826985e-06,
+      "loss": 0.4341,
+      "step": 14965
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.150271347552334,
+      "learning_rate": 2.449285034745674e-06,
+      "loss": 0.4093,
+      "step": 14966
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 35.321804454725935,
+      "learning_rate": 2.4481794673465697e-06,
+      "loss": 0.357,
+      "step": 14967
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.46474718878966,
+      "learning_rate": 2.4470741147168242e-06,
+      "loss": 0.4431,
+      "step": 14968
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 44.55708625191702,
+      "learning_rate": 2.445968976887868e-06,
+      "loss": 0.4693,
+      "step": 14969
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 43.42316478228935,
+      "learning_rate": 2.4448640538911363e-06,
+      "loss": 0.4482,
+      "step": 14970
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 33.092024832940425,
+      "learning_rate": 2.443759345758051e-06,
+      "loss": 0.3905,
+      "step": 14971
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 42.73902275851478,
+      "learning_rate": 2.442654852520032e-06,
+      "loss": 0.4349,
+      "step": 14972
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 57.96609286159012,
+      "learning_rate": 2.441550574208489e-06,
+      "loss": 0.6311,
+      "step": 14973
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 41.96930039784136,
+      "learning_rate": 2.4404465108548226e-06,
+      "loss": 0.482,
+      "step": 14974
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.57961235949529,
+      "learning_rate": 2.4393426624904347e-06,
+      "loss": 0.3935,
+      "step": 14975
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 46.65839380550037,
+      "learning_rate": 2.438239029146723e-06,
+      "loss": 0.4975,
+      "step": 14976
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 39.98036760802863,
+      "learning_rate": 2.437135610855068e-06,
+      "loss": 0.437,
+      "step": 14977
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 42.793825902324144,
+      "learning_rate": 2.436032407646851e-06,
+      "loss": 0.5062,
+      "step": 14978
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 45.439273857322206,
+      "learning_rate": 2.434929419553447e-06,
+      "loss": 0.4542,
+      "step": 14979
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 43.19801610777759,
+      "learning_rate": 2.4338266466062242e-06,
+      "loss": 0.4231,
+      "step": 14980
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 44.92291195410681,
+      "learning_rate": 2.432724088836549e-06,
+      "loss": 0.4852,
+      "step": 14981
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 55.390781741330755,
+      "learning_rate": 2.431621746275773e-06,
+      "loss": 0.5301,
+      "step": 14982
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.16473835092859,
+      "learning_rate": 2.430519618955245e-06,
+      "loss": 0.456,
+      "step": 14983
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 44.187491992911944,
+      "learning_rate": 2.429417706906311e-06,
+      "loss": 0.4969,
+      "step": 14984
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 56.100281112564964,
+      "learning_rate": 2.428316010160312e-06,
+      "loss": 0.4762,
+      "step": 14985
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 37.69887526678502,
+      "learning_rate": 2.427214528748576e-06,
+      "loss": 0.3703,
+      "step": 14986
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 49.056708138579616,
+      "learning_rate": 2.4261132627024254e-06,
+      "loss": 0.4899,
+      "step": 14987
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 46.80601798076584,
+      "learning_rate": 2.425012212053184e-06,
+      "loss": 0.4567,
+      "step": 14988
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 45.20667767137171,
+      "learning_rate": 2.4239113768321666e-06,
+      "loss": 0.4499,
+      "step": 14989
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 51.12611488208593,
+      "learning_rate": 2.422810757070675e-06,
+      "loss": 0.5036,
+      "step": 14990
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 57.37722067597375,
+      "learning_rate": 2.4217103528000153e-06,
+      "loss": 0.5506,
+      "step": 14991
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 52.656537665874616,
+      "learning_rate": 2.4206101640514778e-06,
+      "loss": 0.3858,
+      "step": 14992
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 46.490676801493294,
+      "learning_rate": 2.4195101908563556e-06,
+      "loss": 0.5099,
+      "step": 14993
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.397730065538205,
+      "learning_rate": 2.4184104332459267e-06,
+      "loss": 0.3594,
+      "step": 14994
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 49.823814265434784,
+      "learning_rate": 2.4173108912514696e-06,
+      "loss": 0.4853,
+      "step": 14995
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 38.033682301085214,
+      "learning_rate": 2.4162115649042584e-06,
+      "loss": 0.4713,
+      "step": 14996
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.357590294262664,
+      "learning_rate": 2.415112454235551e-06,
+      "loss": 0.3935,
+      "step": 14997
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 49.56119753075048,
+      "learning_rate": 2.41401355927661e-06,
+      "loss": 0.558,
+      "step": 14998
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.16167816997694,
+      "learning_rate": 2.4129148800586846e-06,
+      "loss": 0.4434,
+      "step": 14999
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.923659639307985,
+      "learning_rate": 2.411816416613021e-06,
+      "loss": 0.4416,
+      "step": 15000
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 39.01744807410193,
+      "learning_rate": 2.4107181689708614e-06,
+      "loss": 0.4108,
+      "step": 15001
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 56.74004806105064,
+      "learning_rate": 2.4096201371634387e-06,
+      "loss": 0.6059,
+      "step": 15002
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 47.98136158769081,
+      "learning_rate": 2.4085223212219767e-06,
+      "loss": 0.4289,
+      "step": 15003
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.108862730588214,
+      "learning_rate": 2.407424721177698e-06,
+      "loss": 0.4361,
+      "step": 15004
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 35.850829212755414,
+      "learning_rate": 2.406327337061818e-06,
+      "loss": 0.3798,
+      "step": 15005
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 50.48652053802835,
+      "learning_rate": 2.4052301689055536e-06,
+      "loss": 0.5167,
+      "step": 15006
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 47.15352719322691,
+      "learning_rate": 2.4041332167400944e-06,
+      "loss": 0.5134,
+      "step": 15007
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 45.171233704064676,
+      "learning_rate": 2.4030364805966423e-06,
+      "loss": 0.4736,
+      "step": 15008
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 49.4356736473962,
+      "learning_rate": 2.401939960506389e-06,
+      "loss": 0.4654,
+      "step": 15009
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 54.03842246975493,
+      "learning_rate": 2.400843656500521e-06,
+      "loss": 0.5516,
+      "step": 15010
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 44.47346756789392,
+      "learning_rate": 2.3997475686102147e-06,
+      "loss": 0.452,
+      "step": 15011
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 48.756261731525214,
+      "learning_rate": 2.3986516968666385e-06,
+      "loss": 0.4872,
+      "step": 15012
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 47.24065382928127,
+      "learning_rate": 2.397556041300961e-06,
+      "loss": 0.4738,
+      "step": 15013
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 49.32405110725933,
+      "learning_rate": 2.396460601944346e-06,
+      "loss": 0.4834,
+      "step": 15014
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 47.536827999062616,
+      "learning_rate": 2.3953653788279395e-06,
+      "loss": 0.4165,
+      "step": 15015
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 38.478466142529946,
+      "learning_rate": 2.3942703719828965e-06,
+      "loss": 0.3965,
+      "step": 15016
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 66.4420048504327,
+      "learning_rate": 2.393175581440352e-06,
+      "loss": 0.4786,
+      "step": 15017
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.73618412137042,
+      "learning_rate": 2.3920810072314473e-06,
+      "loss": 0.4206,
+      "step": 15018
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 46.311805884633344,
+      "learning_rate": 2.3909866493873048e-06,
+      "loss": 0.5678,
+      "step": 15019
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 41.90130416506552,
+      "learning_rate": 2.389892507939051e-06,
+      "loss": 0.4234,
+      "step": 15020
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 34.505731037523304,
+      "learning_rate": 2.388798582917806e-06,
+      "loss": 0.4344,
+      "step": 15021
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 43.55370599436233,
+      "learning_rate": 2.387704874354674e-06,
+      "loss": 0.5111,
+      "step": 15022
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 48.12991345081359,
+      "learning_rate": 2.3866113822807645e-06,
+      "loss": 0.4155,
+      "step": 15023
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 46.778389944589854,
+      "learning_rate": 2.3855181067271703e-06,
+      "loss": 0.4964,
+      "step": 15024
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.34042173720371,
+      "learning_rate": 2.3844250477249877e-06,
+      "loss": 0.4696,
+      "step": 15025
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 47.55191502731081,
+      "learning_rate": 2.3833322053053045e-06,
+      "loss": 0.4604,
+      "step": 15026
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 41.4592081287557,
+      "learning_rate": 2.382239579499198e-06,
+      "loss": 0.4516,
+      "step": 15027
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 49.02905057954579,
+      "learning_rate": 2.3811471703377376e-06,
+      "loss": 0.5154,
+      "step": 15028
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 56.11918073022181,
+      "learning_rate": 2.3800549778519956e-06,
+      "loss": 0.3998,
+      "step": 15029
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 55.27082380066334,
+      "learning_rate": 2.3789630020730327e-06,
+      "loss": 0.5709,
+      "step": 15030
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 33.83873517896714,
+      "learning_rate": 2.3778712430319095e-06,
+      "loss": 0.3779,
+      "step": 15031
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 46.8504102576989,
+      "learning_rate": 2.376779700759664e-06,
+      "loss": 0.4736,
+      "step": 15032
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 40.58911887542114,
+      "learning_rate": 2.3756883752873426e-06,
+      "loss": 0.4242,
+      "step": 15033
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 44.85863285653278,
+      "learning_rate": 2.3745972666459847e-06,
+      "loss": 0.4588,
+      "step": 15034
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 43.47112833423471,
+      "learning_rate": 2.373506374866623e-06,
+      "loss": 0.4286,
+      "step": 15035
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 46.521132517614355,
+      "learning_rate": 2.3724156999802795e-06,
+      "loss": 0.4607,
+      "step": 15036
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 36.321348097181975,
+      "learning_rate": 2.3713252420179668e-06,
+      "loss": 0.4088,
+      "step": 15037
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 59.463799251005064,
+      "learning_rate": 2.3702350010107023e-06,
+      "loss": 0.5275,
+      "step": 15038
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 47.79534628831149,
+      "learning_rate": 2.3691449769894946e-06,
+      "loss": 0.4352,
+      "step": 15039
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 40.981386547666894,
+      "learning_rate": 2.368055169985339e-06,
+      "loss": 0.4467,
+      "step": 15040
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.12857813819915,
+      "learning_rate": 2.366965580029227e-06,
+      "loss": 0.4534,
+      "step": 15041
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 37.43685785166356,
+      "learning_rate": 2.365876207152149e-06,
+      "loss": 0.4508,
+      "step": 15042
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 47.55432519325049,
+      "learning_rate": 2.3647870513850857e-06,
+      "loss": 0.4897,
+      "step": 15043
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 44.28745787293113,
+      "learning_rate": 2.3636981127590155e-06,
+      "loss": 0.4536,
+      "step": 15044
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 41.411383944670774,
+      "learning_rate": 2.362609391304903e-06,
+      "loss": 0.3895,
+      "step": 15045
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.520045626359156,
+      "learning_rate": 2.361520887053709e-06,
+      "loss": 0.4514,
+      "step": 15046
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.55428718908316,
+      "learning_rate": 2.360432600036393e-06,
+      "loss": 0.4421,
+      "step": 15047
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 38.41459552673602,
+      "learning_rate": 2.359344530283908e-06,
+      "loss": 0.4261,
+      "step": 15048
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 42.37744071142998,
+      "learning_rate": 2.358256677827191e-06,
+      "loss": 0.4816,
+      "step": 15049
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 37.877258809406726,
+      "learning_rate": 2.357169042697187e-06,
+      "loss": 0.3949,
+      "step": 15050
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.273767362484,
+      "learning_rate": 2.356081624924822e-06,
+      "loss": 0.438,
+      "step": 15051
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 36.09237858532478,
+      "learning_rate": 2.354994424541027e-06,
+      "loss": 0.4027,
+      "step": 15052
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 48.07959771589418,
+      "learning_rate": 2.353907441576715e-06,
+      "loss": 0.4585,
+      "step": 15053
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.42937934226486,
+      "learning_rate": 2.3528206760628026e-06,
+      "loss": 0.4219,
+      "step": 15054
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 47.93119469645791,
+      "learning_rate": 2.3517341280301997e-06,
+      "loss": 0.4426,
+      "step": 15055
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 48.91390859338699,
+      "learning_rate": 2.350647797509804e-06,
+      "loss": 0.4845,
+      "step": 15056
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 38.3696737849009,
+      "learning_rate": 2.3495616845325074e-06,
+      "loss": 0.3892,
+      "step": 15057
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 34.21207522085714,
+      "learning_rate": 2.348475789129202e-06,
+      "loss": 0.3785,
+      "step": 15058
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 44.26870237480121,
+      "learning_rate": 2.3473901113307685e-06,
+      "loss": 0.4338,
+      "step": 15059
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.56461069008764,
+      "learning_rate": 2.3463046511680864e-06,
+      "loss": 0.4849,
+      "step": 15060
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 40.04659124386232,
+      "learning_rate": 2.3452194086720225e-06,
+      "loss": 0.4569,
+      "step": 15061
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 48.99528390707349,
+      "learning_rate": 2.3441343838734375e-06,
+      "loss": 0.5595,
+      "step": 15062
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.301657174505614,
+      "learning_rate": 2.3430495768031923e-06,
+      "loss": 0.454,
+      "step": 15063
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.131356515146905,
+      "learning_rate": 2.3419649874921413e-06,
+      "loss": 0.4084,
+      "step": 15064
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 40.4881886107162,
+      "learning_rate": 2.3408806159711262e-06,
+      "loss": 0.4113,
+      "step": 15065
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 49.03902118918742,
+      "learning_rate": 2.3397964622709825e-06,
+      "loss": 0.4612,
+      "step": 15066
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.795460047166195,
+      "learning_rate": 2.3387125264225475e-06,
+      "loss": 0.4015,
+      "step": 15067
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 50.00093635635994,
+      "learning_rate": 2.3376288084566445e-06,
+      "loss": 0.536,
+      "step": 15068
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 37.93867658019232,
+      "learning_rate": 2.3365453084041e-06,
+      "loss": 0.4135,
+      "step": 15069
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 48.33051132097419,
+      "learning_rate": 2.3354620262957238e-06,
+      "loss": 0.5123,
+      "step": 15070
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 38.07935101488506,
+      "learning_rate": 2.3343789621623213e-06,
+      "loss": 0.3997,
+      "step": 15071
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 38.360568939709495,
+      "learning_rate": 2.3332961160346966e-06,
+      "loss": 0.4417,
+      "step": 15072
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.88473632977837,
+      "learning_rate": 2.3322134879436487e-06,
+      "loss": 0.4701,
+      "step": 15073
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 42.86971567359071,
+      "learning_rate": 2.3311310779199603e-06,
+      "loss": 0.5099,
+      "step": 15074
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.0768876618291,
+      "learning_rate": 2.3300488859944217e-06,
+      "loss": 0.3626,
+      "step": 15075
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 63.50029102018456,
+      "learning_rate": 2.328966912197802e-06,
+      "loss": 0.4938,
+      "step": 15076
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 39.78411657429075,
+      "learning_rate": 2.3278851565608782e-06,
+      "loss": 0.4815,
+      "step": 15077
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 41.0231917894626,
+      "learning_rate": 2.3268036191144117e-06,
+      "loss": 0.5005,
+      "step": 15078
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 47.36093077185569,
+      "learning_rate": 2.3257222998891603e-06,
+      "loss": 0.4613,
+      "step": 15079
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 53.76870950167225,
+      "learning_rate": 2.32464119891588e-06,
+      "loss": 0.4248,
+      "step": 15080
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 44.15309735468128,
+      "learning_rate": 2.323560316225314e-06,
+      "loss": 0.3926,
+      "step": 15081
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 41.54428258628809,
+      "learning_rate": 2.3224796518481995e-06,
+      "loss": 0.5145,
+      "step": 15082
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 43.53593166614008,
+      "learning_rate": 2.3213992058152733e-06,
+      "loss": 0.3707,
+      "step": 15083
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 46.3738897761376,
+      "learning_rate": 2.320318978157263e-06,
+      "loss": 0.5208,
+      "step": 15084
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 47.95004093001352,
+      "learning_rate": 2.319238968904888e-06,
+      "loss": 0.4412,
+      "step": 15085
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 42.315389627481416,
+      "learning_rate": 2.318159178088865e-06,
+      "loss": 0.4713,
+      "step": 15086
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.611650405586126,
+      "learning_rate": 2.317079605739898e-06,
+      "loss": 0.4925,
+      "step": 15087
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 40.955346255007086,
+      "learning_rate": 2.3160002518886927e-06,
+      "loss": 0.4456,
+      "step": 15088
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 48.770124419137524,
+      "learning_rate": 2.314921116565948e-06,
+      "loss": 0.4451,
+      "step": 15089
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 65.03682933009208,
+      "learning_rate": 2.313842199802352e-06,
+      "loss": 0.3995,
+      "step": 15090
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 48.15768211656272,
+      "learning_rate": 2.312763501628584e-06,
+      "loss": 0.4639,
+      "step": 15091
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 46.16979931395079,
+      "learning_rate": 2.3116850220753253e-06,
+      "loss": 0.4363,
+      "step": 15092
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 42.66750544354382,
+      "learning_rate": 2.3106067611732473e-06,
+      "loss": 0.4519,
+      "step": 15093
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 41.863985389425174,
+      "learning_rate": 2.3095287189530203e-06,
+      "loss": 0.4665,
+      "step": 15094
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 36.177995372420376,
+      "learning_rate": 2.3084508954452923e-06,
+      "loss": 0.4113,
+      "step": 15095
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 42.47123419630706,
+      "learning_rate": 2.307373290680721e-06,
+      "loss": 0.4252,
+      "step": 15096
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 53.71546647720164,
+      "learning_rate": 2.3062959046899535e-06,
+      "loss": 0.4828,
+      "step": 15097
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 44.900650142474085,
+      "learning_rate": 2.3052187375036327e-06,
+      "loss": 0.5146,
+      "step": 15098
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 54.58577792781897,
+      "learning_rate": 2.3041417891523897e-06,
+      "loss": 0.5149,
+      "step": 15099
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 48.346209931374624,
+      "learning_rate": 2.303065059666849e-06,
+      "loss": 0.5239,
+      "step": 15100
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 54.765607142207436,
+      "learning_rate": 2.301988549077636e-06,
+      "loss": 0.4914,
+      "step": 15101
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 56.160272130497056,
+      "learning_rate": 2.3009122574153673e-06,
+      "loss": 0.5044,
+      "step": 15102
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 44.774407145474626,
+      "learning_rate": 2.2998361847106487e-06,
+      "loss": 0.4657,
+      "step": 15103
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 56.23047120921429,
+      "learning_rate": 2.2987603309940866e-06,
+      "loss": 0.5062,
+      "step": 15104
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 38.63655672356607,
+      "learning_rate": 2.297684696296273e-06,
+      "loss": 0.4139,
+      "step": 15105
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 39.38196894115521,
+      "learning_rate": 2.2966092806478e-06,
+      "loss": 0.4425,
+      "step": 15106
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 55.65977842101096,
+      "learning_rate": 2.2955340840792563e-06,
+      "loss": 0.5191,
+      "step": 15107
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 47.16391232314819,
+      "learning_rate": 2.294459106621214e-06,
+      "loss": 0.434,
+      "step": 15108
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 52.56146189767876,
+      "learning_rate": 2.2933843483042495e-06,
+      "loss": 0.4612,
+      "step": 15109
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 52.12094816290332,
+      "learning_rate": 2.2923098091589224e-06,
+      "loss": 0.4866,
+      "step": 15110
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 48.4132090295255,
+      "learning_rate": 2.2912354892157995e-06,
+      "loss": 0.478,
+      "step": 15111
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 44.70121888007268,
+      "learning_rate": 2.2901613885054252e-06,
+      "loss": 0.4305,
+      "step": 15112
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 47.3823290367023,
+      "learning_rate": 2.289087507058353e-06,
+      "loss": 0.4192,
+      "step": 15113
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 50.24698065231759,
+      "learning_rate": 2.2880138449051227e-06,
+      "loss": 0.4979,
+      "step": 15114
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 36.98681058225043,
+      "learning_rate": 2.2869404020762676e-06,
+      "loss": 0.3499,
+      "step": 15115
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 42.218708955983416,
+      "learning_rate": 2.285867178602312e-06,
+      "loss": 0.425,
+      "step": 15116
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 40.28057928540534,
+      "learning_rate": 2.2847941745137826e-06,
+      "loss": 0.4246,
+      "step": 15117
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 39.85103122975516,
+      "learning_rate": 2.2837213898411925e-06,
+      "loss": 0.4134,
+      "step": 15118
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 46.18441330605951,
+      "learning_rate": 2.282648824615059e-06,
+      "loss": 0.3999,
+      "step": 15119
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 53.39820022586103,
+      "learning_rate": 2.281576478865872e-06,
+      "loss": 0.4956,
+      "step": 15120
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 47.70079378691154,
+      "learning_rate": 2.2805043526241342e-06,
+      "loss": 0.4621,
+      "step": 15121
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.25259371431739,
+      "learning_rate": 2.2794324459203377e-06,
+      "loss": 0.5129,
+      "step": 15122
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.70654218891465,
+      "learning_rate": 2.278360758784969e-06,
+      "loss": 0.4797,
+      "step": 15123
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.682700934458,
+      "learning_rate": 2.277289291248502e-06,
+      "loss": 0.49,
+      "step": 15124
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 53.108601164111285,
+      "learning_rate": 2.2762180433414083e-06,
+      "loss": 0.5248,
+      "step": 15125
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 55.67815876972737,
+      "learning_rate": 2.2751470150941558e-06,
+      "loss": 0.5751,
+      "step": 15126
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 49.21473259893988,
+      "learning_rate": 2.2740762065372046e-06,
+      "loss": 0.4887,
+      "step": 15127
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 44.679009960871774,
+      "learning_rate": 2.2730056177010075e-06,
+      "loss": 0.4842,
+      "step": 15128
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 33.98668631926158,
+      "learning_rate": 2.2719352486160074e-06,
+      "loss": 0.3696,
+      "step": 15129
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 49.25975216004398,
+      "learning_rate": 2.270865099312648e-06,
+      "loss": 0.5104,
+      "step": 15130
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 44.24691369991037,
+      "learning_rate": 2.2697951698213628e-06,
+      "loss": 0.446,
+      "step": 15131
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 39.53616830137077,
+      "learning_rate": 2.268725460172585e-06,
+      "loss": 0.4486,
+      "step": 15132
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 55.48410932627147,
+      "learning_rate": 2.2676559703967317e-06,
+      "loss": 0.4883,
+      "step": 15133
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 49.645054754933355,
+      "learning_rate": 2.266586700524217e-06,
+      "loss": 0.4755,
+      "step": 15134
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 39.09321304184277,
+      "learning_rate": 2.2655176505854527e-06,
+      "loss": 0.4343,
+      "step": 15135
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.072071278508446,
+      "learning_rate": 2.264448820610844e-06,
+      "loss": 0.4541,
+      "step": 15136
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 36.786472374383514,
+      "learning_rate": 2.2633802106307835e-06,
+      "loss": 0.3976,
+      "step": 15137
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 50.8437330671492,
+      "learning_rate": 2.2623118206756668e-06,
+      "loss": 0.4727,
+      "step": 15138
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 50.38521939403923,
+      "learning_rate": 2.2612436507758726e-06,
+      "loss": 0.5396,
+      "step": 15139
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 46.08646912049086,
+      "learning_rate": 2.260175700961785e-06,
+      "loss": 0.49,
+      "step": 15140
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 47.51139919628622,
+      "learning_rate": 2.2591079712637698e-06,
+      "loss": 0.4818,
+      "step": 15141
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 38.68526481864174,
+      "learning_rate": 2.2580404617121954e-06,
+      "loss": 0.4049,
+      "step": 15142
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 49.75216859854632,
+      "learning_rate": 2.2569731723374243e-06,
+      "loss": 0.5129,
+      "step": 15143
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.58232839011932,
+      "learning_rate": 2.255906103169806e-06,
+      "loss": 0.449,
+      "step": 15144
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 41.80855058398544,
+      "learning_rate": 2.2548392542396856e-06,
+      "loss": 0.4369,
+      "step": 15145
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 46.43878467266598,
+      "learning_rate": 2.2537726255774063e-06,
+      "loss": 0.3906,
+      "step": 15146
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 49.10228051485326,
+      "learning_rate": 2.252706217213302e-06,
+      "loss": 0.5136,
+      "step": 15147
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 50.443092083460044,
+      "learning_rate": 2.251640029177704e-06,
+      "loss": 0.3874,
+      "step": 15148
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.26466666442225,
+      "learning_rate": 2.250574061500931e-06,
+      "loss": 0.4818,
+      "step": 15149
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 43.96461998820293,
+      "learning_rate": 2.2495083142132944e-06,
+      "loss": 0.459,
+      "step": 15150
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 41.72326861709743,
+      "learning_rate": 2.2484427873451086e-06,
+      "loss": 0.4235,
+      "step": 15151
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 38.837506864367256,
+      "learning_rate": 2.247377480926678e-06,
+      "loss": 0.4313,
+      "step": 15152
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 48.31871808936343,
+      "learning_rate": 2.246312394988296e-06,
+      "loss": 0.5326,
+      "step": 15153
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 43.01037412354772,
+      "learning_rate": 2.2452475295602518e-06,
+      "loss": 0.4005,
+      "step": 15154
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 46.361820204032895,
+      "learning_rate": 2.2441828846728307e-06,
+      "loss": 0.5069,
+      "step": 15155
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 46.24537954190575,
+      "learning_rate": 2.243118460356312e-06,
+      "loss": 0.4381,
+      "step": 15156
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 38.5384556080506,
+      "learning_rate": 2.2420542566409686e-06,
+      "loss": 0.4062,
+      "step": 15157
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 41.65429058460362,
+      "learning_rate": 2.2409902735570643e-06,
+      "loss": 0.4238,
+      "step": 15158
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 47.03785974434438,
+      "learning_rate": 2.2399265111348558e-06,
+      "loss": 0.5376,
+      "step": 15159
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 50.82911033859695,
+      "learning_rate": 2.238862969404596e-06,
+      "loss": 0.5406,
+      "step": 15160
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 45.26636170856655,
+      "learning_rate": 2.2377996483965368e-06,
+      "loss": 0.4995,
+      "step": 15161
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.24047811615425,
+      "learning_rate": 2.236736548140913e-06,
+      "loss": 0.4618,
+      "step": 15162
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 47.11888838784495,
+      "learning_rate": 2.2356736686679624e-06,
+      "loss": 0.533,
+      "step": 15163
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 62.16098666470946,
+      "learning_rate": 2.2346110100079076e-06,
+      "loss": 0.5442,
+      "step": 15164
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 45.005491472656004,
+      "learning_rate": 2.2335485721909766e-06,
+      "loss": 0.4401,
+      "step": 15165
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.93670701867764,
+      "learning_rate": 2.2324863552473776e-06,
+      "loss": 0.4053,
+      "step": 15166
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.402547169519686,
+      "learning_rate": 2.2314243592073226e-06,
+      "loss": 0.4611,
+      "step": 15167
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.365613885665205,
+      "learning_rate": 2.230362584101018e-06,
+      "loss": 0.4022,
+      "step": 15168
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 38.73753089251153,
+      "learning_rate": 2.229301029958656e-06,
+      "loss": 0.4403,
+      "step": 15169
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 43.023332281557224,
+      "learning_rate": 2.2282396968104235e-06,
+      "loss": 0.4707,
+      "step": 15170
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.532313833635605,
+      "learning_rate": 2.227178584686509e-06,
+      "loss": 0.4075,
+      "step": 15171
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 58.24113318284069,
+      "learning_rate": 2.226117693617088e-06,
+      "loss": 0.48,
+      "step": 15172
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.917135892852215,
+      "learning_rate": 2.2250570236323344e-06,
+      "loss": 0.4228,
+      "step": 15173
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.91146554305013,
+      "learning_rate": 2.2239965747624117e-06,
+      "loss": 0.4542,
+      "step": 15174
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 39.95559332678083,
+      "learning_rate": 2.222936347037474e-06,
+      "loss": 0.4559,
+      "step": 15175
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 55.28118380234778,
+      "learning_rate": 2.2218763404876775e-06,
+      "loss": 0.4421,
+      "step": 15176
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.63450471841122,
+      "learning_rate": 2.2208165551431706e-06,
+      "loss": 0.4809,
+      "step": 15177
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.368013644371835,
+      "learning_rate": 2.21975699103409e-06,
+      "loss": 0.4159,
+      "step": 15178
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 38.60190207373137,
+      "learning_rate": 2.2186976481905663e-06,
+      "loss": 0.3947,
+      "step": 15179
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 48.0648532286379,
+      "learning_rate": 2.2176385266427302e-06,
+      "loss": 0.4577,
+      "step": 15180
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 35.385916019779664,
+      "learning_rate": 2.216579626420702e-06,
+      "loss": 0.4485,
+      "step": 15181
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.77881515072169,
+      "learning_rate": 2.2155209475546013e-06,
+      "loss": 0.4398,
+      "step": 15182
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.60953165033648,
+      "learning_rate": 2.2144624900745247e-06,
+      "loss": 0.4182,
+      "step": 15183
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 48.07249619564228,
+      "learning_rate": 2.2134042540105814e-06,
+      "loss": 0.4607,
+      "step": 15184
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 40.8077841074366,
+      "learning_rate": 2.2123462393928663e-06,
+      "loss": 0.4546,
+      "step": 15185
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 43.766725765332666,
+      "learning_rate": 2.2112884462514707e-06,
+      "loss": 0.4349,
+      "step": 15186
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.3826454668439,
+      "learning_rate": 2.210230874616475e-06,
+      "loss": 0.4541,
+      "step": 15187
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.178827303363384,
+      "learning_rate": 2.2091735245179548e-06,
+      "loss": 0.4607,
+      "step": 15188
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.21879427135978,
+      "learning_rate": 2.208116395985981e-06,
+      "loss": 0.434,
+      "step": 15189
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 45.563700619591515,
+      "learning_rate": 2.2070594890506216e-06,
+      "loss": 0.4352,
+      "step": 15190
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 53.11581201423355,
+      "learning_rate": 2.206002803741929e-06,
+      "loss": 0.5222,
+      "step": 15191
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.813090992200124,
+      "learning_rate": 2.2049463400899606e-06,
+      "loss": 0.5211,
+      "step": 15192
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 37.3447007062234,
+      "learning_rate": 2.2038900981247545e-06,
+      "loss": 0.362,
+      "step": 15193
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 48.90309389308282,
+      "learning_rate": 2.2028340778763544e-06,
+      "loss": 0.5228,
+      "step": 15194
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 41.08433344760487,
+      "learning_rate": 2.201778279374794e-06,
+      "loss": 0.4073,
+      "step": 15195
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.59868899490562,
+      "learning_rate": 2.2007227026500956e-06,
+      "loss": 0.4172,
+      "step": 15196
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.72031993229478,
+      "learning_rate": 2.199667347732284e-06,
+      "loss": 0.4249,
+      "step": 15197
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 36.614771394350655,
+      "learning_rate": 2.1986122146513654e-06,
+      "loss": 0.4133,
+      "step": 15198
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 49.057419784815465,
+      "learning_rate": 2.1975573034373563e-06,
+      "loss": 0.4915,
+      "step": 15199
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 56.179954255507454,
+      "learning_rate": 2.1965026141202495e-06,
+      "loss": 0.5197,
+      "step": 15200
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 54.202578724681295,
+      "learning_rate": 2.195448146730044e-06,
+      "loss": 0.5349,
+      "step": 15201
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 46.298344424331425,
+      "learning_rate": 2.1943939012967295e-06,
+      "loss": 0.5147,
+      "step": 15202
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 43.01004824669586,
+      "learning_rate": 2.193339877850288e-06,
+      "loss": 0.4269,
+      "step": 15203
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 54.522032230281894,
+      "learning_rate": 2.19228607642069e-06,
+      "loss": 0.4969,
+      "step": 15204
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 45.095332704131586,
+      "learning_rate": 2.1912324970379084e-06,
+      "loss": 0.4485,
+      "step": 15205
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.42461024533858,
+      "learning_rate": 2.190179139731906e-06,
+      "loss": 0.4927,
+      "step": 15206
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.77632130958339,
+      "learning_rate": 2.1891260045326467e-06,
+      "loss": 0.4163,
+      "step": 15207
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 33.61857591475096,
+      "learning_rate": 2.1880730914700687e-06,
+      "loss": 0.3545,
+      "step": 15208
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.92104391195856,
+      "learning_rate": 2.1870204005741226e-06,
+      "loss": 0.4812,
+      "step": 15209
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.58854665602466,
+      "learning_rate": 2.185967931874745e-06,
+      "loss": 0.4381,
+      "step": 15210
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 53.20910744886836,
+      "learning_rate": 2.1849156854018715e-06,
+      "loss": 0.4897,
+      "step": 15211
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 36.875163475579434,
+      "learning_rate": 2.1838636611854246e-06,
+      "loss": 0.3656,
+      "step": 15212
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 48.958072211468775,
+      "learning_rate": 2.1828118592553195e-06,
+      "loss": 0.4691,
+      "step": 15213
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 42.34446615479201,
+      "learning_rate": 2.181760279641473e-06,
+      "loss": 0.4363,
+      "step": 15214
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.52486694103063,
+      "learning_rate": 2.180708922373792e-06,
+      "loss": 0.389,
+      "step": 15215
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 48.29710339751674,
+      "learning_rate": 2.1796577874821734e-06,
+      "loss": 0.5082,
+      "step": 15216
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 42.47365830824382,
+      "learning_rate": 2.178606874996515e-06,
+      "loss": 0.3939,
+      "step": 15217
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 45.64840508126396,
+      "learning_rate": 2.1775561849466987e-06,
+      "loss": 0.4652,
+      "step": 15218
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.453142821052985,
+      "learning_rate": 2.176505717362609e-06,
+      "loss": 0.4298,
+      "step": 15219
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.88364313040617,
+      "learning_rate": 2.1754554722741227e-06,
+      "loss": 0.4639,
+      "step": 15220
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 40.94555555102992,
+      "learning_rate": 2.1744054497111033e-06,
+      "loss": 0.4866,
+      "step": 15221
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 51.41634052344226,
+      "learning_rate": 2.173355649703417e-06,
+      "loss": 0.5339,
+      "step": 15222
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.00713965559196,
+      "learning_rate": 2.1723060722809155e-06,
+      "loss": 0.4532,
+      "step": 15223
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 32.87283334010525,
+      "learning_rate": 2.1712567174734523e-06,
+      "loss": 0.4186,
+      "step": 15224
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.58508910690027,
+      "learning_rate": 2.170207585310865e-06,
+      "loss": 0.5123,
+      "step": 15225
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 45.88199761159239,
+      "learning_rate": 2.1691586758229976e-06,
+      "loss": 0.4317,
+      "step": 15226
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 44.289285979192435,
+      "learning_rate": 2.168109989039674e-06,
+      "loss": 0.4237,
+      "step": 15227
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 46.527091701946766,
+      "learning_rate": 2.1670615249907234e-06,
+      "loss": 0.3751,
+      "step": 15228
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.46389618640053,
+      "learning_rate": 2.1660132837059576e-06,
+      "loss": 0.4165,
+      "step": 15229
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 53.86935721406036,
+      "learning_rate": 2.164965265215191e-06,
+      "loss": 0.4205,
+      "step": 15230
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 38.80935981812239,
+      "learning_rate": 2.1639174695482322e-06,
+      "loss": 0.4107,
+      "step": 15231
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 52.09367973354935,
+      "learning_rate": 2.162869896734876e-06,
+      "loss": 0.4934,
+      "step": 15232
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.81091537004295,
+      "learning_rate": 2.161822546804914e-06,
+      "loss": 0.4513,
+      "step": 15233
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 36.13715652334256,
+      "learning_rate": 2.1607754197881324e-06,
+      "loss": 0.3691,
+      "step": 15234
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 44.997763811286866,
+      "learning_rate": 2.1597285157143122e-06,
+      "loss": 0.4187,
+      "step": 15235
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 48.05925507374549,
+      "learning_rate": 2.15868183461323e-06,
+      "loss": 0.4447,
+      "step": 15236
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.13570637142369,
+      "learning_rate": 2.1576353765146486e-06,
+      "loss": 0.4213,
+      "step": 15237
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 54.73843400101285,
+      "learning_rate": 2.1565891414483266e-06,
+      "loss": 0.5658,
+      "step": 15238
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.648072135256605,
+      "learning_rate": 2.155543129444021e-06,
+      "loss": 0.3863,
+      "step": 15239
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 40.82100978712074,
+      "learning_rate": 2.154497340531484e-06,
+      "loss": 0.398,
+      "step": 15240
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 51.77093716719082,
+      "learning_rate": 2.1534517747404516e-06,
+      "loss": 0.4024,
+      "step": 15241
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 36.25510871667835,
+      "learning_rate": 2.152406432100659e-06,
+      "loss": 0.4469,
+      "step": 15242
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 38.44752314071708,
+      "learning_rate": 2.1513613126418364e-06,
+      "loss": 0.4273,
+      "step": 15243
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.58708292250874,
+      "learning_rate": 2.1503164163937062e-06,
+      "loss": 0.4035,
+      "step": 15244
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 50.6543649067375,
+      "learning_rate": 2.1492717433859887e-06,
+      "loss": 0.5119,
+      "step": 15245
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.83418669577933,
+      "learning_rate": 2.1482272936483896e-06,
+      "loss": 0.4311,
+      "step": 15246
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 47.60320324858467,
+      "learning_rate": 2.147183067210611e-06,
+      "loss": 0.4505,
+      "step": 15247
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 45.605006260619994,
+      "learning_rate": 2.1461390641023516e-06,
+      "loss": 0.4886,
+      "step": 15248
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.02210749878318,
+      "learning_rate": 2.145095284353307e-06,
+      "loss": 0.4341,
+      "step": 15249
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 36.40658966036983,
+      "learning_rate": 2.1440517279931528e-06,
+      "loss": 0.4037,
+      "step": 15250
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 43.749157401554,
+      "learning_rate": 2.1430083950515755e-06,
+      "loss": 0.4998,
+      "step": 15251
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 50.633963401273725,
+      "learning_rate": 2.1419652855582406e-06,
+      "loss": 0.4233,
+      "step": 15252
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 45.73025103118285,
+      "learning_rate": 2.1409223995428187e-06,
+      "loss": 0.4703,
+      "step": 15253
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 44.77932925474166,
+      "learning_rate": 2.1398797370349644e-06,
+      "loss": 0.4292,
+      "step": 15254
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 46.21601259637925,
+      "learning_rate": 2.1388372980643315e-06,
+      "loss": 0.5263,
+      "step": 15255
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 36.71914616147867,
+      "learning_rate": 2.13779508266057e-06,
+      "loss": 0.4203,
+      "step": 15256
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.826568273783124,
+      "learning_rate": 2.136753090853314e-06,
+      "loss": 0.4363,
+      "step": 15257
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 53.63408561690264,
+      "learning_rate": 2.1357113226722036e-06,
+      "loss": 0.5262,
+      "step": 15258
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 48.89454524951334,
+      "learning_rate": 2.1346697781468593e-06,
+      "loss": 0.4632,
+      "step": 15259
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 52.088625504244085,
+      "learning_rate": 2.1336284573069067e-06,
+      "loss": 0.5162,
+      "step": 15260
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 36.263471287156,
+      "learning_rate": 2.1325873601819613e-06,
+      "loss": 0.3401,
+      "step": 15261
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 31.66911548548901,
+      "learning_rate": 2.1315464868016287e-06,
+      "loss": 0.3438,
+      "step": 15262
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 41.27507831143229,
+      "learning_rate": 2.130505837195508e-06,
+      "loss": 0.4585,
+      "step": 15263
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 53.93151058390915,
+      "learning_rate": 2.129465411393198e-06,
+      "loss": 0.5009,
+      "step": 15264
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 41.73603584986604,
+      "learning_rate": 2.1284252094242908e-06,
+      "loss": 0.4082,
+      "step": 15265
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 47.433974224236856,
+      "learning_rate": 2.1273852313183663e-06,
+      "loss": 0.4715,
+      "step": 15266
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.64877762475611,
+      "learning_rate": 2.126345477104996e-06,
+      "loss": 0.3653,
+      "step": 15267
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 59.105909531822746,
+      "learning_rate": 2.125305946813756e-06,
+      "loss": 0.4913,
+      "step": 15268
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.871510459918184,
+      "learning_rate": 2.1242666404742074e-06,
+      "loss": 0.3944,
+      "step": 15269
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.06110757705221,
+      "learning_rate": 2.1232275581159123e-06,
+      "loss": 0.3589,
+      "step": 15270
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 40.65075506890892,
+      "learning_rate": 2.122188699768416e-06,
+      "loss": 0.4188,
+      "step": 15271
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.403900548047346,
+      "learning_rate": 2.1211500654612625e-06,
+      "loss": 0.4703,
+      "step": 15272
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 40.82457023062681,
+      "learning_rate": 2.120111655223993e-06,
+      "loss": 0.4135,
+      "step": 15273
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 44.4777894660785,
+      "learning_rate": 2.1190734690861403e-06,
+      "loss": 0.3934,
+      "step": 15274
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 60.08180600560541,
+      "learning_rate": 2.1180355070772287e-06,
+      "loss": 0.4341,
+      "step": 15275
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 40.40875415757257,
+      "learning_rate": 2.116997769226773e-06,
+      "loss": 0.439,
+      "step": 15276
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 56.44137873791823,
+      "learning_rate": 2.11596025556429e-06,
+      "loss": 0.5305,
+      "step": 15277
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 43.041617319916426,
+      "learning_rate": 2.114922966119287e-06,
+      "loss": 0.4966,
+      "step": 15278
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 40.107574909922306,
+      "learning_rate": 2.113885900921261e-06,
+      "loss": 0.438,
+      "step": 15279
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 49.702218444097355,
+      "learning_rate": 2.1128490599997078e-06,
+      "loss": 0.436,
+      "step": 15280
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 59.82396919679167,
+      "learning_rate": 2.1118124433841114e-06,
+      "loss": 0.5583,
+      "step": 15281
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.865684053761356,
+      "learning_rate": 2.1107760511039553e-06,
+      "loss": 0.4357,
+      "step": 15282
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 45.94506589385184,
+      "learning_rate": 2.109739883188715e-06,
+      "loss": 0.4208,
+      "step": 15283
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 46.91945640181587,
+      "learning_rate": 2.1087039396678544e-06,
+      "loss": 0.4594,
+      "step": 15284
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 51.94873184789404,
+      "learning_rate": 2.107668220570841e-06,
+      "loss": 0.4823,
+      "step": 15285
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 44.40305168129157,
+      "learning_rate": 2.1066327259271223e-06,
+      "loss": 0.4949,
+      "step": 15286
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 44.898835761389975,
+      "learning_rate": 2.1055974557661553e-06,
+      "loss": 0.474,
+      "step": 15287
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 45.61147416169469,
+      "learning_rate": 2.1045624101173754e-06,
+      "loss": 0.4318,
+      "step": 15288
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 39.225556422044,
+      "learning_rate": 2.1035275890102214e-06,
+      "loss": 0.4119,
+      "step": 15289
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 48.02301228752191,
+      "learning_rate": 2.1024929924741265e-06,
+      "loss": 0.4863,
+      "step": 15290
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 46.44520379672151,
+      "learning_rate": 2.1014586205385113e-06,
+      "loss": 0.4435,
+      "step": 15291
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 43.76151350293818,
+      "learning_rate": 2.1004244732327896e-06,
+      "loss": 0.4563,
+      "step": 15292
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.699136089188755,
+      "learning_rate": 2.0993905505863755e-06,
+      "loss": 0.4361,
+      "step": 15293
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 43.339168292674714,
+      "learning_rate": 2.098356852628671e-06,
+      "loss": 0.4489,
+      "step": 15294
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 43.37709584243369,
+      "learning_rate": 2.0973233793890812e-06,
+      "loss": 0.4301,
+      "step": 15295
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 35.034030575956635,
+      "learning_rate": 2.0962901308969864e-06,
+      "loss": 0.4143,
+      "step": 15296
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 38.74621708155457,
+      "learning_rate": 2.0952571071817763e-06,
+      "loss": 0.4279,
+      "step": 15297
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.978722913814686,
+      "learning_rate": 2.094224308272831e-06,
+      "loss": 0.4304,
+      "step": 15298
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 49.74330196626986,
+      "learning_rate": 2.0931917341995233e-06,
+      "loss": 0.4101,
+      "step": 15299
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 44.929304901017,
+      "learning_rate": 2.092159384991217e-06,
+      "loss": 0.4639,
+      "step": 15300
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 44.200419122339945,
+      "learning_rate": 2.09112726067727e-06,
+      "loss": 0.4676,
+      "step": 15301
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 58.0265662501687,
+      "learning_rate": 2.0900953612870367e-06,
+      "loss": 0.4375,
+      "step": 15302
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 48.068955304177706,
+      "learning_rate": 2.089063686849867e-06,
+      "loss": 0.4544,
+      "step": 15303
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 57.824184313346834,
+      "learning_rate": 2.0880322373950957e-06,
+      "loss": 0.55,
+      "step": 15304
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.933295395904224,
+      "learning_rate": 2.0870010129520624e-06,
+      "loss": 0.449,
+      "step": 15305
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.340428329755426,
+      "learning_rate": 2.085970013550088e-06,
+      "loss": 0.4255,
+      "step": 15306
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 54.32616268077651,
+      "learning_rate": 2.0849392392184963e-06,
+      "loss": 0.486,
+      "step": 15307
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 37.059190914392616,
+      "learning_rate": 2.0839086899866055e-06,
+      "loss": 0.367,
+      "step": 15308
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 46.77709213666998,
+      "learning_rate": 2.0828783658837194e-06,
+      "loss": 0.4835,
+      "step": 15309
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 46.36310107792088,
+      "learning_rate": 2.0818482669391428e-06,
+      "loss": 0.4301,
+      "step": 15310
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 38.46381411583997,
+      "learning_rate": 2.080818393182167e-06,
+      "loss": 0.3909,
+      "step": 15311
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 39.58597706381894,
+      "learning_rate": 2.079788744642085e-06,
+      "loss": 0.4373,
+      "step": 15312
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 63.4577385700663,
+      "learning_rate": 2.078759321348177e-06,
+      "loss": 0.5349,
+      "step": 15313
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 39.536328103411286,
+      "learning_rate": 2.0777301233297187e-06,
+      "loss": 0.4235,
+      "step": 15314
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.31323714860853,
+      "learning_rate": 2.076701150615985e-06,
+      "loss": 0.4783,
+      "step": 15315
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 60.37257483056154,
+      "learning_rate": 2.075672403236235e-06,
+      "loss": 0.4626,
+      "step": 15316
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 46.84121265117828,
+      "learning_rate": 2.0746438812197244e-06,
+      "loss": 0.4621,
+      "step": 15317
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.4491470677104,
+      "learning_rate": 2.0736155845957053e-06,
+      "loss": 0.4266,
+      "step": 15318
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 36.872489756472106,
+      "learning_rate": 2.0725875133934216e-06,
+      "loss": 0.3995,
+      "step": 15319
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.336021951217035,
+      "learning_rate": 2.0715596676421167e-06,
+      "loss": 0.4266,
+      "step": 15320
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.91113827931095,
+      "learning_rate": 2.070532047371012e-06,
+      "loss": 0.468,
+      "step": 15321
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 45.828939699201186,
+      "learning_rate": 2.0695046526093375e-06,
+      "loss": 0.4856,
+      "step": 15322
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 37.189457372233726,
+      "learning_rate": 2.0684774833863117e-06,
+      "loss": 0.4371,
+      "step": 15323
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.125781709842,
+      "learning_rate": 2.067450539731148e-06,
+      "loss": 0.4632,
+      "step": 15324
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 48.54637221857795,
+      "learning_rate": 2.0664238216730513e-06,
+      "loss": 0.4525,
+      "step": 15325
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.58801516026643,
+      "learning_rate": 2.0653973292412177e-06,
+      "loss": 0.3908,
+      "step": 15326
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 42.105954644570275,
+      "learning_rate": 2.0643710624648427e-06,
+      "loss": 0.4557,
+      "step": 15327
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.770463816593036,
+      "learning_rate": 2.063345021373114e-06,
+      "loss": 0.4192,
+      "step": 15328
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 34.40879795019154,
+      "learning_rate": 2.0623192059952114e-06,
+      "loss": 0.3745,
+      "step": 15329
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 45.55813407126368,
+      "learning_rate": 2.061293616360304e-06,
+      "loss": 0.4003,
+      "step": 15330
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 38.07498077384578,
+      "learning_rate": 2.0602682524975616e-06,
+      "loss": 0.4073,
+      "step": 15331
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 41.06086683291549,
+      "learning_rate": 2.0592431144361458e-06,
+      "loss": 0.4,
+      "step": 15332
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 44.356687200315086,
+      "learning_rate": 2.0582182022052134e-06,
+      "loss": 0.4387,
+      "step": 15333
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.70471195715595,
+      "learning_rate": 2.0571935158339084e-06,
+      "loss": 0.4723,
+      "step": 15334
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.41711420985996,
+      "learning_rate": 2.0561690553513725e-06,
+      "loss": 0.4234,
+      "step": 15335
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 48.037384490272046,
+      "learning_rate": 2.0551448207867407e-06,
+      "loss": 0.4302,
+      "step": 15336
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 45.0496293098195,
+      "learning_rate": 2.0541208121691458e-06,
+      "loss": 0.4175,
+      "step": 15337
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 46.38267589290036,
+      "learning_rate": 2.0530970295277032e-06,
+      "loss": 0.4194,
+      "step": 15338
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 43.6573436712024,
+      "learning_rate": 2.0520734728915358e-06,
+      "loss": 0.3945,
+      "step": 15339
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 46.682856776973665,
+      "learning_rate": 2.0510501422897466e-06,
+      "loss": 0.4618,
+      "step": 15340
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.92769886579277,
+      "learning_rate": 2.050027037751444e-06,
+      "loss": 0.4507,
+      "step": 15341
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 36.03258333940131,
+      "learning_rate": 2.0490041593057185e-06,
+      "loss": 0.4273,
+      "step": 15342
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 47.06508062514018,
+      "learning_rate": 2.0479815069816643e-06,
+      "loss": 0.456,
+      "step": 15343
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 36.666589929038004,
+      "learning_rate": 2.0469590808083674e-06,
+      "loss": 0.4285,
+      "step": 15344
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 41.644612228419305,
+      "learning_rate": 2.0459368808148983e-06,
+      "loss": 0.3867,
+      "step": 15345
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.33425052968345,
+      "learning_rate": 2.0449149070303344e-06,
+      "loss": 0.4034,
+      "step": 15346
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 47.26091162517895,
+      "learning_rate": 2.043893159483734e-06,
+      "loss": 0.5135,
+      "step": 15347
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 38.56431891233728,
+      "learning_rate": 2.042871638204158e-06,
+      "loss": 0.3402,
+      "step": 15348
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 45.521814461732816,
+      "learning_rate": 2.0418503432206604e-06,
+      "loss": 0.4493,
+      "step": 15349
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 46.100662851015926,
+      "learning_rate": 2.040829274562284e-06,
+      "loss": 0.4991,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 39.301940834130164,
+      "learning_rate": 2.0398084322580634e-06,
+      "loss": 0.4485,
+      "step": 15351
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.95481082165479,
+      "learning_rate": 2.0387878163370354e-06,
+      "loss": 0.452,
+      "step": 15352
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.12255277192169,
+      "learning_rate": 2.0377674268282275e-06,
+      "loss": 0.4996,
+      "step": 15353
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 43.31253954512647,
+      "learning_rate": 2.0367472637606554e-06,
+      "loss": 0.4285,
+      "step": 15354
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 43.71962390868776,
+      "learning_rate": 2.0357273271633304e-06,
+      "loss": 0.4464,
+      "step": 15355
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 43.915624819105226,
+      "learning_rate": 2.0347076170652624e-06,
+      "loss": 0.4782,
+      "step": 15356
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 53.64806227957634,
+      "learning_rate": 2.0336881334954484e-06,
+      "loss": 0.5066,
+      "step": 15357
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 44.627949673273726,
+      "learning_rate": 2.0326688764828873e-06,
+      "loss": 0.5571,
+      "step": 15358
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.800857250656215,
+      "learning_rate": 2.0316498460565627e-06,
+      "loss": 0.4634,
+      "step": 15359
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 40.79305910741363,
+      "learning_rate": 2.030631042245452e-06,
+      "loss": 0.4377,
+      "step": 15360
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 51.48009329836289,
+      "learning_rate": 2.0296124650785333e-06,
+      "loss": 0.4622,
+      "step": 15361
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 37.739428909551584,
+      "learning_rate": 2.0285941145847754e-06,
+      "loss": 0.3591,
+      "step": 15362
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 54.594267743892225,
+      "learning_rate": 2.0275759907931356e-06,
+      "loss": 0.476,
+      "step": 15363
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 40.07307858290195,
+      "learning_rate": 2.0265580937325735e-06,
+      "loss": 0.4107,
+      "step": 15364
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 36.08930617880812,
+      "learning_rate": 2.025540423432032e-06,
+      "loss": 0.3732,
+      "step": 15365
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 38.13321767950082,
+      "learning_rate": 2.024522979920458e-06,
+      "loss": 0.3985,
+      "step": 15366
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 40.24633421701553,
+      "learning_rate": 2.023505763226783e-06,
+      "loss": 0.399,
+      "step": 15367
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.70680092748856,
+      "learning_rate": 2.0224887733799383e-06,
+      "loss": 0.5121,
+      "step": 15368
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.17780770756714,
+      "learning_rate": 2.021472010408848e-06,
+      "loss": 0.4291,
+      "step": 15369
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 51.71344314500753,
+      "learning_rate": 2.020455474342424e-06,
+      "loss": 0.4455,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 44.83564766183466,
+      "learning_rate": 2.019439165209581e-06,
+      "loss": 0.4124,
+      "step": 15371
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 62.33574450346208,
+      "learning_rate": 2.018423083039218e-06,
+      "loss": 0.4386,
+      "step": 15372
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 38.457333120204574,
+      "learning_rate": 2.0174072278602352e-06,
+      "loss": 0.3888,
+      "step": 15373
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.597229634229635,
+      "learning_rate": 2.0163915997015195e-06,
+      "loss": 0.4173,
+      "step": 15374
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.49501284341363,
+      "learning_rate": 2.0153761985919575e-06,
+      "loss": 0.4511,
+      "step": 15375
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 48.16560290766812,
+      "learning_rate": 2.014361024560424e-06,
+      "loss": 0.5344,
+      "step": 15376
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 38.746672142858266,
+      "learning_rate": 2.0133460776357906e-06,
+      "loss": 0.3954,
+      "step": 15377
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 51.0458390452176,
+      "learning_rate": 2.012331357846926e-06,
+      "loss": 0.4711,
+      "step": 15378
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 37.6567768614068,
+      "learning_rate": 2.0113168652226843e-06,
+      "loss": 0.3943,
+      "step": 15379
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 55.05168542733176,
+      "learning_rate": 2.0103025997919155e-06,
+      "loss": 0.4704,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 41.50789845804324,
+      "learning_rate": 2.009288561583468e-06,
+      "loss": 0.565,
+      "step": 15381
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 50.98573487058493,
+      "learning_rate": 2.008274750626178e-06,
+      "loss": 0.5009,
+      "step": 15382
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 54.20682556328837,
+      "learning_rate": 2.007261166948885e-06,
+      "loss": 0.4406,
+      "step": 15383
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 47.11284534330811,
+      "learning_rate": 2.006247810580403e-06,
+      "loss": 0.4719,
+      "step": 15384
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 43.418481425811954,
+      "learning_rate": 2.0052346815495573e-06,
+      "loss": 0.4957,
+      "step": 15385
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 54.58554945322306,
+      "learning_rate": 2.0042217798851606e-06,
+      "loss": 0.4895,
+      "step": 15386
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 43.18862593655715,
+      "learning_rate": 2.003209105616021e-06,
+      "loss": 0.3951,
+      "step": 15387
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 46.29585256229666,
+      "learning_rate": 2.0021966587709372e-06,
+      "loss": 0.4416,
+      "step": 15388
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.090600581965326,
+      "learning_rate": 2.0011844393786983e-06,
+      "loss": 0.4164,
+      "step": 15389
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.32608144476559,
+      "learning_rate": 2.0001724474680963e-06,
+      "loss": 0.4615,
+      "step": 15390
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 48.62892543471518,
+      "learning_rate": 1.999160683067911e-06,
+      "loss": 0.4778,
+      "step": 15391
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 77.47990045168383,
+      "learning_rate": 1.9981491462069146e-06,
+      "loss": 0.4717,
+      "step": 15392
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 51.762072638465746,
+      "learning_rate": 1.9971378369138773e-06,
+      "loss": 0.4842,
+      "step": 15393
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 42.541742195239195,
+      "learning_rate": 1.9961267552175558e-06,
+      "loss": 0.4674,
+      "step": 15394
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.11793673553518,
+      "learning_rate": 1.995115901146707e-06,
+      "loss": 0.4655,
+      "step": 15395
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 47.03858084591239,
+      "learning_rate": 1.9941052747300826e-06,
+      "loss": 0.4536,
+      "step": 15396
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 41.37008907527062,
+      "learning_rate": 1.9930948759964175e-06,
+      "loss": 0.422,
+      "step": 15397
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 43.442261562497634,
+      "learning_rate": 1.992084704974453e-06,
+      "loss": 0.4766,
+      "step": 15398
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 39.797893066324704,
+      "learning_rate": 1.991074761692913e-06,
+      "loss": 0.4018,
+      "step": 15399
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 44.95228616716126,
+      "learning_rate": 1.9900650461805237e-06,
+      "loss": 0.4455,
+      "step": 15400
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 49.11936193003222,
+      "learning_rate": 1.9890555584659965e-06,
+      "loss": 0.4338,
+      "step": 15401
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 45.54433771986776,
+      "learning_rate": 1.9880462985780423e-06,
+      "loss": 0.4266,
+      "step": 15402
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 39.41518678904717,
+      "learning_rate": 1.9870372665453673e-06,
+      "loss": 0.4198,
+      "step": 15403
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 39.751796315948994,
+      "learning_rate": 1.986028462396666e-06,
+      "loss": 0.3949,
+      "step": 15404
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 46.28589447125595,
+      "learning_rate": 1.985019886160624e-06,
+      "loss": 0.399,
+      "step": 15405
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 35.48762750270145,
+      "learning_rate": 1.9840115378659275e-06,
+      "loss": 0.4188,
+      "step": 15406
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 46.62173335416115,
+      "learning_rate": 1.983003417541254e-06,
+      "loss": 0.4044,
+      "step": 15407
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 51.90187680851274,
+      "learning_rate": 1.9819955252152755e-06,
+      "loss": 0.4999,
+      "step": 15408
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 41.95850305734431,
+      "learning_rate": 1.9809878609166546e-06,
+      "loss": 0.428,
+      "step": 15409
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 49.435223195422985,
+      "learning_rate": 1.979980424674045e-06,
+      "loss": 0.4462,
+      "step": 15410
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 55.045235368237734,
+      "learning_rate": 1.9789732165161e-06,
+      "loss": 0.5272,
+      "step": 15411
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 36.81862496487716,
+      "learning_rate": 1.977966236471468e-06,
+      "loss": 0.3594,
+      "step": 15412
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 44.28095291027265,
+      "learning_rate": 1.9769594845687833e-06,
+      "loss": 0.4689,
+      "step": 15413
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 47.49278379637948,
+      "learning_rate": 1.9759529608366744e-06,
+      "loss": 0.4515,
+      "step": 15414
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 60.3151527216718,
+      "learning_rate": 1.9749466653037707e-06,
+      "loss": 0.5522,
+      "step": 15415
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 36.701659587839046,
+      "learning_rate": 1.9739405979986904e-06,
+      "loss": 0.3981,
+      "step": 15416
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 40.640157817896466,
+      "learning_rate": 1.9729347589500426e-06,
+      "loss": 0.4451,
+      "step": 15417
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 38.054329833055384,
+      "learning_rate": 1.9719291481864366e-06,
+      "loss": 0.4283,
+      "step": 15418
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 46.43676517464806,
+      "learning_rate": 1.970923765736468e-06,
+      "loss": 0.4001,
+      "step": 15419
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 38.86916712641226,
+      "learning_rate": 1.9699186116287295e-06,
+      "loss": 0.3902,
+      "step": 15420
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 41.66999661617938,
+      "learning_rate": 1.9689136858918112e-06,
+      "loss": 0.4175,
+      "step": 15421
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 42.768963639906026,
+      "learning_rate": 1.967908988554289e-06,
+      "loss": 0.4237,
+      "step": 15422
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.99454957465551,
+      "learning_rate": 1.9669045196447345e-06,
+      "loss": 0.4579,
+      "step": 15423
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 38.11759338350454,
+      "learning_rate": 1.9659002791917157e-06,
+      "loss": 0.4304,
+      "step": 15424
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 53.3822553246316,
+      "learning_rate": 1.964896267223797e-06,
+      "loss": 0.5001,
+      "step": 15425
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.365962887189006,
+      "learning_rate": 1.963892483769524e-06,
+      "loss": 0.4718,
+      "step": 15426
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 35.51560182314719,
+      "learning_rate": 1.9628889288574514e-06,
+      "loss": 0.3599,
+      "step": 15427
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 50.24344925385298,
+      "learning_rate": 1.9618856025161127e-06,
+      "loss": 0.4599,
+      "step": 15428
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 50.641808156731315,
+      "learning_rate": 1.9608825047740486e-06,
+      "loss": 0.4596,
+      "step": 15429
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 50.24927592823401,
+      "learning_rate": 1.9598796356597806e-06,
+      "loss": 0.4484,
+      "step": 15430
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.71049252377951,
+      "learning_rate": 1.958876995201833e-06,
+      "loss": 0.4719,
+      "step": 15431
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 40.56110821703478,
+      "learning_rate": 1.9578745834287204e-06,
+      "loss": 0.4202,
+      "step": 15432
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.36110217780091,
+      "learning_rate": 1.9568724003689486e-06,
+      "loss": 0.4693,
+      "step": 15433
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 46.15489289826836,
+      "learning_rate": 1.9558704460510235e-06,
+      "loss": 0.4819,
+      "step": 15434
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 45.997937189891175,
+      "learning_rate": 1.9548687205034346e-06,
+      "loss": 0.4415,
+      "step": 15435
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 39.782768154617735,
+      "learning_rate": 1.953867223754673e-06,
+      "loss": 0.4001,
+      "step": 15436
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.810372207315034,
+      "learning_rate": 1.952865955833223e-06,
+      "loss": 0.4354,
+      "step": 15437
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 45.96480527298984,
+      "learning_rate": 1.9518649167675585e-06,
+      "loss": 0.4821,
+      "step": 15438
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 54.77103683571236,
+      "learning_rate": 1.9508641065861445e-06,
+      "loss": 0.5035,
+      "step": 15439
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 47.95969357068127,
+      "learning_rate": 1.949863525317447e-06,
+      "loss": 0.4609,
+      "step": 15440
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.50095265229578,
+      "learning_rate": 1.9488631729899243e-06,
+      "loss": 0.5123,
+      "step": 15441
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 41.27041878059586,
+      "learning_rate": 1.9478630496320227e-06,
+      "loss": 0.4376,
+      "step": 15442
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 63.32533192116501,
+      "learning_rate": 1.946863155272183e-06,
+      "loss": 0.483,
+      "step": 15443
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 39.171419710876954,
+      "learning_rate": 1.9458634899388453e-06,
+      "loss": 0.3863,
+      "step": 15444
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.90386843504136,
+      "learning_rate": 1.944864053660437e-06,
+      "loss": 0.4679,
+      "step": 15445
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 54.802808685616704,
+      "learning_rate": 1.9438648464653865e-06,
+      "loss": 0.4679,
+      "step": 15446
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 44.964335963901675,
+      "learning_rate": 1.9428658683821066e-06,
+      "loss": 0.4318,
+      "step": 15447
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 46.07208156439769,
+      "learning_rate": 1.9418671194390047e-06,
+      "loss": 0.5022,
+      "step": 15448
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.7196760156896,
+      "learning_rate": 1.9408685996644893e-06,
+      "loss": 0.4181,
+      "step": 15449
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 43.306641015880444,
+      "learning_rate": 1.9398703090869585e-06,
+      "loss": 0.5157,
+      "step": 15450
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 43.95559020982567,
+      "learning_rate": 1.938872247734799e-06,
+      "loss": 0.4831,
+      "step": 15451
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 44.80847248828846,
+      "learning_rate": 1.9378744156363993e-06,
+      "loss": 0.5168,
+      "step": 15452
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 42.23416194657398,
+      "learning_rate": 1.936876812820133e-06,
+      "loss": 0.4523,
+      "step": 15453
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 52.177773059143085,
+      "learning_rate": 1.9358794393143755e-06,
+      "loss": 0.5258,
+      "step": 15454
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 43.49137724444913,
+      "learning_rate": 1.934882295147487e-06,
+      "loss": 0.3889,
+      "step": 15455
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 42.034735490437036,
+      "learning_rate": 1.9338853803478274e-06,
+      "loss": 0.4361,
+      "step": 15456
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 51.726232479525116,
+      "learning_rate": 1.932888694943752e-06,
+      "loss": 0.504,
+      "step": 15457
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 49.23128878291783,
+      "learning_rate": 1.931892238963601e-06,
+      "loss": 0.4816,
+      "step": 15458
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.99651161738877,
+      "learning_rate": 1.9308960124357167e-06,
+      "loss": 0.4339,
+      "step": 15459
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 40.64520014991992,
+      "learning_rate": 1.9299000153884274e-06,
+      "loss": 0.4541,
+      "step": 15460
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 45.907802490615516,
+      "learning_rate": 1.928904247850061e-06,
+      "loss": 0.4035,
+      "step": 15461
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 45.16256870293057,
+      "learning_rate": 1.927908709848938e-06,
+      "loss": 0.4972,
+      "step": 15462
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 43.961119398381726,
+      "learning_rate": 1.9269134014133706e-06,
+      "loss": 0.4117,
+      "step": 15463
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.96516505790402,
+      "learning_rate": 1.9259183225716594e-06,
+      "loss": 0.4146,
+      "step": 15464
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 50.43120264780826,
+      "learning_rate": 1.924923473352108e-06,
+      "loss": 0.4654,
+      "step": 15465
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 44.00807044943882,
+      "learning_rate": 1.9239288537830136e-06,
+      "loss": 0.4372,
+      "step": 15466
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 49.96543511613408,
+      "learning_rate": 1.922934463892657e-06,
+      "loss": 0.4703,
+      "step": 15467
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 44.687692100272066,
+      "learning_rate": 1.9219403037093164e-06,
+      "loss": 0.4747,
+      "step": 15468
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 39.56790585217727,
+      "learning_rate": 1.9209463732612687e-06,
+      "loss": 0.433,
+      "step": 15469
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 36.97040636700354,
+      "learning_rate": 1.919952672576779e-06,
+      "loss": 0.4273,
+      "step": 15470
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 47.369511223442245,
+      "learning_rate": 1.9189592016841154e-06,
+      "loss": 0.4381,
+      "step": 15471
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.43916827115002,
+      "learning_rate": 1.9179659606115187e-06,
+      "loss": 0.4907,
+      "step": 15472
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 45.88511449026602,
+      "learning_rate": 1.9169729493872423e-06,
+      "loss": 0.4215,
+      "step": 15473
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.752233717078326,
+      "learning_rate": 1.9159801680395264e-06,
+      "loss": 0.4827,
+      "step": 15474
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 38.14047452466412,
+      "learning_rate": 1.9149876165966088e-06,
+      "loss": 0.4021,
+      "step": 15475
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 37.00470038147483,
+      "learning_rate": 1.9139952950867135e-06,
+      "loss": 0.3789,
+      "step": 15476
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.2007959247743,
+      "learning_rate": 1.9130032035380595e-06,
+      "loss": 0.4421,
+      "step": 15477
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 47.34903004225839,
+      "learning_rate": 1.9120113419788635e-06,
+      "loss": 0.3969,
+      "step": 15478
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 50.00314775095326,
+      "learning_rate": 1.911019710437335e-06,
+      "loss": 0.488,
+      "step": 15479
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 48.147458783449046,
+      "learning_rate": 1.9100283089416725e-06,
+      "loss": 0.4774,
+      "step": 15480
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 43.65111737126565,
+      "learning_rate": 1.9090371375200755e-06,
+      "loss": 0.4702,
+      "step": 15481
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 53.00823958695019,
+      "learning_rate": 1.9080461962007257e-06,
+      "loss": 0.4591,
+      "step": 15482
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 52.779894050811265,
+      "learning_rate": 1.9070554850118095e-06,
+      "loss": 0.518,
+      "step": 15483
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 45.635208377857666,
+      "learning_rate": 1.906065003981503e-06,
+      "loss": 0.4035,
+      "step": 15484
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 48.33678604597027,
+      "learning_rate": 1.9050747531379698e-06,
+      "loss": 0.4579,
+      "step": 15485
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 41.035910144909074,
+      "learning_rate": 1.9040847325093791e-06,
+      "loss": 0.3653,
+      "step": 15486
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 38.3564130024075,
+      "learning_rate": 1.9030949421238787e-06,
+      "loss": 0.4081,
+      "step": 15487
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 58.88320327918623,
+      "learning_rate": 1.9021053820096246e-06,
+      "loss": 0.5714,
+      "step": 15488
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.74705561377239,
+      "learning_rate": 1.9011160521947548e-06,
+      "loss": 0.469,
+      "step": 15489
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 34.016225855947766,
+      "learning_rate": 1.900126952707405e-06,
+      "loss": 0.3943,
+      "step": 15490
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 41.40776213565898,
+      "learning_rate": 1.89913808357571e-06,
+      "loss": 0.4388,
+      "step": 15491
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 51.23839567043275,
+      "learning_rate": 1.8981494448277893e-06,
+      "loss": 0.4415,
+      "step": 15492
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 47.31990924000198,
+      "learning_rate": 1.897161036491756e-06,
+      "loss": 0.4236,
+      "step": 15493
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 47.02448467375852,
+      "learning_rate": 1.8961728585957229e-06,
+      "loss": 0.3987,
+      "step": 15494
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 45.475088284675756,
+      "learning_rate": 1.8951849111677922e-06,
+      "loss": 0.4143,
+      "step": 15495
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 37.773321605611095,
+      "learning_rate": 1.894197194236065e-06,
+      "loss": 0.3626,
+      "step": 15496
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 43.021327396292946,
+      "learning_rate": 1.8932097078286283e-06,
+      "loss": 0.4708,
+      "step": 15497
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 44.755344641778414,
+      "learning_rate": 1.8922224519735611e-06,
+      "loss": 0.4889,
+      "step": 15498
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 42.82115411653426,
+      "learning_rate": 1.8912354266989453e-06,
+      "loss": 0.421,
+      "step": 15499
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.82577217512219,
+      "learning_rate": 1.890248632032854e-06,
+      "loss": 0.4957,
+      "step": 15500
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 61.14371593216648,
+      "learning_rate": 1.8892620680033457e-06,
+      "loss": 0.514,
+      "step": 15501
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 65.94826451520498,
+      "learning_rate": 1.8882757346384785e-06,
+      "loss": 0.4586,
+      "step": 15502
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 34.94057272977189,
+      "learning_rate": 1.8872896319663038e-06,
+      "loss": 0.4257,
+      "step": 15503
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 49.94843732730222,
+      "learning_rate": 1.886303760014868e-06,
+      "loss": 0.5048,
+      "step": 15504
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 44.08243263612882,
+      "learning_rate": 1.885318118812205e-06,
+      "loss": 0.4109,
+      "step": 15505
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 41.78600545852761,
+      "learning_rate": 1.8843327083863495e-06,
+      "loss": 0.3744,
+      "step": 15506
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.544096900627615,
+      "learning_rate": 1.8833475287653225e-06,
+      "loss": 0.4859,
+      "step": 15507
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 54.31087813682937,
+      "learning_rate": 1.8823625799771428e-06,
+      "loss": 0.4246,
+      "step": 15508
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 60.595313230577325,
+      "learning_rate": 1.8813778620498268e-06,
+      "loss": 0.4131,
+      "step": 15509
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 42.409632222533475,
+      "learning_rate": 1.8803933750113712e-06,
+      "loss": 0.4696,
+      "step": 15510
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 43.99087281007931,
+      "learning_rate": 1.8794091188897812e-06,
+      "loss": 0.359,
+      "step": 15511
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 41.80547131135267,
+      "learning_rate": 1.8784250937130433e-06,
+      "loss": 0.4062,
+      "step": 15512
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 51.852027259894534,
+      "learning_rate": 1.8774412995091462e-06,
+      "loss": 0.5518,
+      "step": 15513
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 42.12255254591534,
+      "learning_rate": 1.8764577363060654e-06,
+      "loss": 0.4475,
+      "step": 15514
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 56.75617775922432,
+      "learning_rate": 1.8754744041317763e-06,
+      "loss": 0.4792,
+      "step": 15515
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 38.60204806366955,
+      "learning_rate": 1.8744913030142409e-06,
+      "loss": 0.4015,
+      "step": 15516
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 42.02038735339176,
+      "learning_rate": 1.8735084329814213e-06,
+      "loss": 0.3836,
+      "step": 15517
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 59.707932319368155,
+      "learning_rate": 1.872525794061265e-06,
+      "loss": 0.526,
+      "step": 15518
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 48.38587593323285,
+      "learning_rate": 1.8715433862817224e-06,
+      "loss": 0.429,
+      "step": 15519
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 44.67075689943972,
+      "learning_rate": 1.8705612096707314e-06,
+      "loss": 0.4832,
+      "step": 15520
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 43.31353628617044,
+      "learning_rate": 1.8695792642562226e-06,
+      "loss": 0.4358,
+      "step": 15521
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 39.96147014900576,
+      "learning_rate": 1.8685975500661257e-06,
+      "loss": 0.4117,
+      "step": 15522
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 54.99071780686487,
+      "learning_rate": 1.867616067128355e-06,
+      "loss": 0.6138,
+      "step": 15523
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 56.83141714885568,
+      "learning_rate": 1.8666348154708269e-06,
+      "loss": 0.4688,
+      "step": 15524
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 44.52679094175055,
+      "learning_rate": 1.8656537951214482e-06,
+      "loss": 0.4362,
+      "step": 15525
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 48.658103701757454,
+      "learning_rate": 1.8646730061081174e-06,
+      "loss": 0.4536,
+      "step": 15526
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 44.371141373408705,
+      "learning_rate": 1.8636924484587248e-06,
+      "loss": 0.4428,
+      "step": 15527
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 47.78496253171503,
+      "learning_rate": 1.86271212220116e-06,
+      "loss": 0.4859,
+      "step": 15528
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 45.91671927507343,
+      "learning_rate": 1.8617320273633044e-06,
+      "loss": 0.4107,
+      "step": 15529
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 55.536211314888185,
+      "learning_rate": 1.8607521639730285e-06,
+      "loss": 0.4159,
+      "step": 15530
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 45.66775644990608,
+      "learning_rate": 1.8597725320581982e-06,
+      "loss": 0.5105,
+      "step": 15531
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 45.23877572113104,
+      "learning_rate": 1.8587931316466746e-06,
+      "loss": 0.444,
+      "step": 15532
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.78769015119267,
+      "learning_rate": 1.8578139627663127e-06,
+      "loss": 0.4939,
+      "step": 15533
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 42.73606660552888,
+      "learning_rate": 1.8568350254449596e-06,
+      "loss": 0.44,
+      "step": 15534
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 47.62421169279389,
+      "learning_rate": 1.8558563197104562e-06,
+      "loss": 0.5168,
+      "step": 15535
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 39.750708585697346,
+      "learning_rate": 1.8548778455906325e-06,
+      "loss": 0.3848,
+      "step": 15536
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 49.28449010750294,
+      "learning_rate": 1.8538996031133172e-06,
+      "loss": 0.4886,
+      "step": 15537
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 44.465619854327244,
+      "learning_rate": 1.8529215923063348e-06,
+      "loss": 0.4937,
+      "step": 15538
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 50.792687995954395,
+      "learning_rate": 1.8519438131974932e-06,
+      "loss": 0.3984,
+      "step": 15539
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 52.39110967023399,
+      "learning_rate": 1.8509662658146067e-06,
+      "loss": 0.42,
+      "step": 15540
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 57.58315560092171,
+      "learning_rate": 1.8499889501854685e-06,
+      "loss": 0.5189,
+      "step": 15541
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 37.428231290813,
+      "learning_rate": 1.8490118663378809e-06,
+      "loss": 0.4235,
+      "step": 15542
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.11926047236814,
+      "learning_rate": 1.848035014299625e-06,
+      "loss": 0.4449,
+      "step": 15543
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.53736566111593,
+      "learning_rate": 1.8470583940984832e-06,
+      "loss": 0.3961,
+      "step": 15544
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 42.64536025669008,
+      "learning_rate": 1.8460820057622353e-06,
+      "loss": 0.4334,
+      "step": 15545
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 53.70644434331382,
+      "learning_rate": 1.8451058493186424e-06,
+      "loss": 0.4836,
+      "step": 15546
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 37.03080738945914,
+      "learning_rate": 1.8441299247954713e-06,
+      "loss": 0.376,
+      "step": 15547
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 39.80633764318711,
+      "learning_rate": 1.843154232220471e-06,
+      "loss": 0.3889,
+      "step": 15548
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 36.42026328993015,
+      "learning_rate": 1.8421787716213934e-06,
+      "loss": 0.4233,
+      "step": 15549
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 37.78346272465903,
+      "learning_rate": 1.8412035430259811e-06,
+      "loss": 0.4188,
+      "step": 15550
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.71971231837089,
+      "learning_rate": 1.8402285464619673e-06,
+      "loss": 0.4697,
+      "step": 15551
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 47.40863613493068,
+      "learning_rate": 1.839253781957079e-06,
+      "loss": 0.4421,
+      "step": 15552
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 38.09077392544004,
+      "learning_rate": 1.8382792495390378e-06,
+      "loss": 0.4134,
+      "step": 15553
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 41.122823535871646,
+      "learning_rate": 1.8373049492355633e-06,
+      "loss": 0.413,
+      "step": 15554
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 48.97877621816205,
+      "learning_rate": 1.8363308810743619e-06,
+      "loss": 0.4269,
+      "step": 15555
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 41.74603996798104,
+      "learning_rate": 1.8353570450831303e-06,
+      "loss": 0.4338,
+      "step": 15556
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 52.2607440123388,
+      "learning_rate": 1.83438344128957e-06,
+      "loss": 0.549,
+      "step": 15557
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 47.94319923173041,
+      "learning_rate": 1.8334100697213664e-06,
+      "loss": 0.3941,
+      "step": 15558
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 51.567650300791364,
+      "learning_rate": 1.8324369304062062e-06,
+      "loss": 0.4911,
+      "step": 15559
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.84954444405032,
+      "learning_rate": 1.8314640233717617e-06,
+      "loss": 0.4876,
+      "step": 15560
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 39.49169362228181,
+      "learning_rate": 1.8304913486456998e-06,
+      "loss": 0.354,
+      "step": 15561
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 47.08505183884335,
+      "learning_rate": 1.8295189062556851e-06,
+      "loss": 0.4222,
+      "step": 15562
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.76844183107772,
+      "learning_rate": 1.8285466962293752e-06,
+      "loss": 0.463,
+      "step": 15563
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 46.804051119356316,
+      "learning_rate": 1.827574718594417e-06,
+      "loss": 0.4637,
+      "step": 15564
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 57.36867958361257,
+      "learning_rate": 1.8266029733784506e-06,
+      "loss": 0.526,
+      "step": 15565
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 35.05493050055959,
+      "learning_rate": 1.8256314606091153e-06,
+      "loss": 0.3526,
+      "step": 15566
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 60.993378252455614,
+      "learning_rate": 1.8246601803140407e-06,
+      "loss": 0.5222,
+      "step": 15567
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 40.23476282617999,
+      "learning_rate": 1.823689132520846e-06,
+      "loss": 0.3773,
+      "step": 15568
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 59.390369622480094,
+      "learning_rate": 1.8227183172571516e-06,
+      "loss": 0.5425,
+      "step": 15569
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 46.073455636050696,
+      "learning_rate": 1.8217477345505619e-06,
+      "loss": 0.4337,
+      "step": 15570
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 58.109889143833286,
+      "learning_rate": 1.8207773844286835e-06,
+      "loss": 0.5102,
+      "step": 15571
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 44.352026816849396,
+      "learning_rate": 1.8198072669191136e-06,
+      "loss": 0.4092,
+      "step": 15572
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 56.87820655251441,
+      "learning_rate": 1.8188373820494376e-06,
+      "loss": 0.5167,
+      "step": 15573
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 62.59763063236849,
+      "learning_rate": 1.8178677298472425e-06,
+      "loss": 0.4742,
+      "step": 15574
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 40.92655653536438,
+      "learning_rate": 1.8168983103401006e-06,
+      "loss": 0.3872,
+      "step": 15575
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 40.53609856962932,
+      "learning_rate": 1.8159291235555864e-06,
+      "loss": 0.3886,
+      "step": 15576
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 41.73625732998545,
+      "learning_rate": 1.8149601695212571e-06,
+      "loss": 0.4437,
+      "step": 15577
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 51.285491688898375,
+      "learning_rate": 1.8139914482646737e-06,
+      "loss": 0.5078,
+      "step": 15578
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.5396911778316,
+      "learning_rate": 1.8130229598133865e-06,
+      "loss": 0.4565,
+      "step": 15579
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 48.899843755523136,
+      "learning_rate": 1.8120547041949366e-06,
+      "loss": 0.3767,
+      "step": 15580
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 53.52807757435487,
+      "learning_rate": 1.81108668143686e-06,
+      "loss": 0.5186,
+      "step": 15581
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 59.617161949371194,
+      "learning_rate": 1.8101188915666867e-06,
+      "loss": 0.4591,
+      "step": 15582
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.579766656667566,
+      "learning_rate": 1.8091513346119415e-06,
+      "loss": 0.3902,
+      "step": 15583
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 110.68560271572953,
+      "learning_rate": 1.8081840106001435e-06,
+      "loss": 0.4923,
+      "step": 15584
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 44.56758609226099,
+      "learning_rate": 1.8072169195587996e-06,
+      "loss": 0.4048,
+      "step": 15585
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 39.25184108279479,
+      "learning_rate": 1.8062500615154121e-06,
+      "loss": 0.4492,
+      "step": 15586
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 51.849502653607814,
+      "learning_rate": 1.8052834364974791e-06,
+      "loss": 0.4966,
+      "step": 15587
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.25976656743637,
+      "learning_rate": 1.8043170445324943e-06,
+      "loss": 0.4533,
+      "step": 15588
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 43.11504935582382,
+      "learning_rate": 1.8033508856479387e-06,
+      "loss": 0.4418,
+      "step": 15589
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 44.42818437518528,
+      "learning_rate": 1.8023849598712861e-06,
+      "loss": 0.4439,
+      "step": 15590
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.42788885982682,
+      "learning_rate": 1.801419267230009e-06,
+      "loss": 0.3947,
+      "step": 15591
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 55.985758128475986,
+      "learning_rate": 1.8004538077515755e-06,
+      "loss": 0.521,
+      "step": 15592
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 52.72718169707479,
+      "learning_rate": 1.7994885814634366e-06,
+      "loss": 0.4976,
+      "step": 15593
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 44.12336765220152,
+      "learning_rate": 1.7985235883930475e-06,
+      "loss": 0.4498,
+      "step": 15594
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.80602819880352,
+      "learning_rate": 1.797558828567847e-06,
+      "loss": 0.4037,
+      "step": 15595
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.595038245234726,
+      "learning_rate": 1.7965943020152753e-06,
+      "loss": 0.4343,
+      "step": 15596
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 42.997158017610886,
+      "learning_rate": 1.7956300087627655e-06,
+      "loss": 0.4275,
+      "step": 15597
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 43.632935384305824,
+      "learning_rate": 1.794665948837736e-06,
+      "loss": 0.5028,
+      "step": 15598
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.96308098266691,
+      "learning_rate": 1.7937021222676098e-06,
+      "loss": 0.4159,
+      "step": 15599
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 94.08158712284236,
+      "learning_rate": 1.7927385290797917e-06,
+      "loss": 0.4568,
+      "step": 15600
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.51723691428838,
+      "learning_rate": 1.7917751693016915e-06,
+      "loss": 0.443,
+      "step": 15601
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 37.262742196146625,
+      "learning_rate": 1.7908120429607024e-06,
+      "loss": 0.3757,
+      "step": 15602
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 50.020376754937296,
+      "learning_rate": 1.7898491500842152e-06,
+      "loss": 0.4696,
+      "step": 15603
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 51.2173171317189,
+      "learning_rate": 1.7888864906996183e-06,
+      "loss": 0.4426,
+      "step": 15604
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 33.475632363287225,
+      "learning_rate": 1.7879240648342854e-06,
+      "loss": 0.3742,
+      "step": 15605
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.81309612060203,
+      "learning_rate": 1.7869618725155868e-06,
+      "loss": 0.3956,
+      "step": 15606
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.96114665369761,
+      "learning_rate": 1.7859999137708872e-06,
+      "loss": 0.4281,
+      "step": 15607
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 48.37636649003139,
+      "learning_rate": 1.7850381886275458e-06,
+      "loss": 0.4475,
+      "step": 15608
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 43.97370220132173,
+      "learning_rate": 1.7840766971129153e-06,
+      "loss": 0.4995,
+      "step": 15609
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 47.24398176023095,
+      "learning_rate": 1.783115439254337e-06,
+      "loss": 0.4941,
+      "step": 15610
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 58.5210196846421,
+      "learning_rate": 1.7821544150791459e-06,
+      "loss": 0.4943,
+      "step": 15611
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.21653511028919,
+      "learning_rate": 1.7811936246146765e-06,
+      "loss": 0.4226,
+      "step": 15612
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.2177187291667,
+      "learning_rate": 1.7802330678882552e-06,
+      "loss": 0.3776,
+      "step": 15613
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.53341362265793,
+      "learning_rate": 1.7792727449271962e-06,
+      "loss": 0.4081,
+      "step": 15614
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 38.288368942682446,
+      "learning_rate": 1.7783126557588092e-06,
+      "loss": 0.417,
+      "step": 15615
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 47.58018846468903,
+      "learning_rate": 1.7773528004104012e-06,
+      "loss": 0.4045,
+      "step": 15616
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 48.966646389062234,
+      "learning_rate": 1.7763931789092715e-06,
+      "loss": 0.4088,
+      "step": 15617
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 48.34311716051414,
+      "learning_rate": 1.7754337912827092e-06,
+      "loss": 0.4019,
+      "step": 15618
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 45.06416355824193,
+      "learning_rate": 1.7744746375579968e-06,
+      "loss": 0.4373,
+      "step": 15619
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 37.29881854001508,
+      "learning_rate": 1.7735157177624129e-06,
+      "loss": 0.3982,
+      "step": 15620
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 42.1971184510567,
+      "learning_rate": 1.77255703192323e-06,
+      "loss": 0.3841,
+      "step": 15621
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.115083814960244,
+      "learning_rate": 1.7715985800677148e-06,
+      "loss": 0.4396,
+      "step": 15622
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 44.163016479132544,
+      "learning_rate": 1.7706403622231228e-06,
+      "loss": 0.4259,
+      "step": 15623
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 43.265446302624206,
+      "learning_rate": 1.7696823784167017e-06,
+      "loss": 0.3626,
+      "step": 15624
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 51.84767463917702,
+      "learning_rate": 1.7687246286756999e-06,
+      "loss": 0.4226,
+      "step": 15625
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 51.99118285895888,
+      "learning_rate": 1.7677671130273567e-06,
+      "loss": 0.4806,
+      "step": 15626
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.73712743584626,
+      "learning_rate": 1.766809831498898e-06,
+      "loss": 0.4301,
+      "step": 15627
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.85183172447682,
+      "learning_rate": 1.7658527841175543e-06,
+      "loss": 0.398,
+      "step": 15628
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.190168289894935,
+      "learning_rate": 1.7648959709105384e-06,
+      "loss": 0.4136,
+      "step": 15629
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 50.92996873937397,
+      "learning_rate": 1.7639393919050662e-06,
+      "loss": 0.4437,
+      "step": 15630
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 45.880741030404955,
+      "learning_rate": 1.762983047128337e-06,
+      "loss": 0.4152,
+      "step": 15631
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.09527364560361,
+      "learning_rate": 1.7620269366075514e-06,
+      "loss": 0.382,
+      "step": 15632
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 47.99726565341317,
+      "learning_rate": 1.7610710603699033e-06,
+      "loss": 0.4536,
+      "step": 15633
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.438025348628294,
+      "learning_rate": 1.7601154184425717e-06,
+      "loss": 0.3537,
+      "step": 15634
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 44.68570785663532,
+      "learning_rate": 1.7591600108527407e-06,
+      "loss": 0.4242,
+      "step": 15635
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 45.60584571829614,
+      "learning_rate": 1.758204837627575e-06,
+      "loss": 0.4815,
+      "step": 15636
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.31782460489875,
+      "learning_rate": 1.757249898794242e-06,
+      "loss": 0.4433,
+      "step": 15637
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.19528668339188,
+      "learning_rate": 1.7562951943799033e-06,
+      "loss": 0.4538,
+      "step": 15638
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 46.48054582483801,
+      "learning_rate": 1.755340724411707e-06,
+      "loss": 0.409,
+      "step": 15639
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.220311162500586,
+      "learning_rate": 1.754386488916795e-06,
+      "loss": 0.5064,
+      "step": 15640
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 43.04896735283102,
+      "learning_rate": 1.7534324879223065e-06,
+      "loss": 0.4264,
+      "step": 15641
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 47.76616286304687,
+      "learning_rate": 1.7524787214553774e-06,
+      "loss": 0.4694,
+      "step": 15642
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 49.469554771677636,
+      "learning_rate": 1.7515251895431295e-06,
+      "loss": 0.4602,
+      "step": 15643
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 55.07879478101453,
+      "learning_rate": 1.750571892212677e-06,
+      "loss": 0.483,
+      "step": 15644
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.82866385190039,
+      "learning_rate": 1.7496188294911332e-06,
+      "loss": 0.3852,
+      "step": 15645
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.81677175835169,
+      "learning_rate": 1.7486660014056046e-06,
+      "loss": 0.3919,
+      "step": 15646
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 46.28023187932849,
+      "learning_rate": 1.7477134079831893e-06,
+      "loss": 0.4305,
+      "step": 15647
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.52091884783798,
+      "learning_rate": 1.746761049250979e-06,
+      "loss": 0.4791,
+      "step": 15648
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.45753262268008,
+      "learning_rate": 1.7458089252360522e-06,
+      "loss": 0.3968,
+      "step": 15649
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.86544979015392,
+      "learning_rate": 1.7448570359654925e-06,
+      "loss": 0.4625,
+      "step": 15650
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.782611049579714,
+      "learning_rate": 1.743905381466371e-06,
+      "loss": 0.4324,
+      "step": 15651
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 41.49422724340622,
+      "learning_rate": 1.7429539617657487e-06,
+      "loss": 0.4266,
+      "step": 15652
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 48.36177545607664,
+      "learning_rate": 1.7420027768906877e-06,
+      "loss": 0.4459,
+      "step": 15653
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.46324979352011,
+      "learning_rate": 1.7410518268682353e-06,
+      "loss": 0.3694,
+      "step": 15654
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 63.190996799626895,
+      "learning_rate": 1.7401011117254397e-06,
+      "loss": 0.4534,
+      "step": 15655
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.76043163427849,
+      "learning_rate": 1.7391506314893337e-06,
+      "loss": 0.3963,
+      "step": 15656
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 43.9180641108815,
+      "learning_rate": 1.7382003861869512e-06,
+      "loss": 0.4583,
+      "step": 15657
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 44.84925444577607,
+      "learning_rate": 1.7372503758453196e-06,
+      "loss": 0.4314,
+      "step": 15658
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 52.64889279819784,
+      "learning_rate": 1.7363006004914507e-06,
+      "loss": 0.462,
+      "step": 15659
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 52.18252101377825,
+      "learning_rate": 1.7353510601523616e-06,
+      "loss": 0.4743,
+      "step": 15660
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 56.99948687071472,
+      "learning_rate": 1.7344017548550506e-06,
+      "loss": 0.4774,
+      "step": 15661
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 52.087380864863306,
+      "learning_rate": 1.733452684626522e-06,
+      "loss": 0.4352,
+      "step": 15662
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 43.94095981047936,
+      "learning_rate": 1.7325038494937597e-06,
+      "loss": 0.4266,
+      "step": 15663
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 44.2927715627828,
+      "learning_rate": 1.7315552494837552e-06,
+      "loss": 0.4128,
+      "step": 15664
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 49.19184313919145,
+      "learning_rate": 1.7306068846234791e-06,
+      "loss": 0.4118,
+      "step": 15665
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.56952307366969,
+      "learning_rate": 1.7296587549399057e-06,
+      "loss": 0.4383,
+      "step": 15666
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.82875905894923,
+      "learning_rate": 1.7287108604600023e-06,
+      "loss": 0.4344,
+      "step": 15667
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.21602942475497,
+      "learning_rate": 1.7277632012107236e-06,
+      "loss": 0.4609,
+      "step": 15668
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 41.058055068096984,
+      "learning_rate": 1.7268157772190174e-06,
+      "loss": 0.4633,
+      "step": 15669
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.89116760078597,
+      "learning_rate": 1.725868588511831e-06,
+      "loss": 0.416,
+      "step": 15670
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.428807215472865,
+      "learning_rate": 1.7249216351161013e-06,
+      "loss": 0.4845,
+      "step": 15671
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 55.27725331213673,
+      "learning_rate": 1.7239749170587628e-06,
+      "loss": 0.4938,
+      "step": 15672
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 53.85924558565878,
+      "learning_rate": 1.723028434366736e-06,
+      "loss": 0.5048,
+      "step": 15673
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 52.692215062552094,
+      "learning_rate": 1.7220821870669358e-06,
+      "loss": 0.4547,
+      "step": 15674
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 44.54856132384951,
+      "learning_rate": 1.7211361751862754e-06,
+      "loss": 0.4256,
+      "step": 15675
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 48.41357264040845,
+      "learning_rate": 1.720190398751662e-06,
+      "loss": 0.4501,
+      "step": 15676
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.96944267741699,
+      "learning_rate": 1.7192448577899912e-06,
+      "loss": 0.4183,
+      "step": 15677
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 39.294337954824,
+      "learning_rate": 1.7182995523281488e-06,
+      "loss": 0.3943,
+      "step": 15678
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.85435583105556,
+      "learning_rate": 1.7173544823930232e-06,
+      "loss": 0.4008,
+      "step": 15679
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 43.680340004176685,
+      "learning_rate": 1.7164096480114934e-06,
+      "loss": 0.4221,
+      "step": 15680
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.36262708056681,
+      "learning_rate": 1.715465049210424e-06,
+      "loss": 0.4475,
+      "step": 15681
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.960958802899476,
+      "learning_rate": 1.7145206860166863e-06,
+      "loss": 0.4636,
+      "step": 15682
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.61464994950826,
+      "learning_rate": 1.7135765584571296e-06,
+      "loss": 0.4542,
+      "step": 15683
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.85697680153807,
+      "learning_rate": 1.712632666558609e-06,
+      "loss": 0.4332,
+      "step": 15684
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 41.77208882329418,
+      "learning_rate": 1.7116890103479688e-06,
+      "loss": 0.423,
+      "step": 15685
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 49.39444173651655,
+      "learning_rate": 1.710745589852043e-06,
+      "loss": 0.442,
+      "step": 15686
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 48.25400294366236,
+      "learning_rate": 1.7098024050976657e-06,
+      "loss": 0.4925,
+      "step": 15687
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 55.52396310665113,
+      "learning_rate": 1.7088594561116567e-06,
+      "loss": 0.4387,
+      "step": 15688
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.51926720527152,
+      "learning_rate": 1.7079167429208366e-06,
+      "loss": 0.4489,
+      "step": 15689
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 63.10173033323233,
+      "learning_rate": 1.7069742655520106e-06,
+      "loss": 0.5725,
+      "step": 15690
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 37.99860463592102,
+      "learning_rate": 1.706032024031985e-06,
+      "loss": 0.4085,
+      "step": 15691
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 53.8035353128302,
+      "learning_rate": 1.7050900183875608e-06,
+      "loss": 0.4968,
+      "step": 15692
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.76842619781583,
+      "learning_rate": 1.704148248645523e-06,
+      "loss": 0.4298,
+      "step": 15693
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 51.44850820059846,
+      "learning_rate": 1.7032067148326536e-06,
+      "loss": 0.5125,
+      "step": 15694
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 43.67797854496472,
+      "learning_rate": 1.7022654169757312e-06,
+      "loss": 0.4086,
+      "step": 15695
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 56.20408756809446,
+      "learning_rate": 1.701324355101528e-06,
+      "loss": 0.4447,
+      "step": 15696
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 63.99509727802658,
+      "learning_rate": 1.7003835292368064e-06,
+      "loss": 0.4463,
+      "step": 15697
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 40.87222183480544,
+      "learning_rate": 1.6994429394083235e-06,
+      "loss": 0.4269,
+      "step": 15698
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 44.03341188396106,
+      "learning_rate": 1.698502585642824e-06,
+      "loss": 0.4363,
+      "step": 15699
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.22039936381978,
+      "learning_rate": 1.6975624679670556e-06,
+      "loss": 0.4811,
+      "step": 15700
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 44.14849227067516,
+      "learning_rate": 1.6966225864077557e-06,
+      "loss": 0.4874,
+      "step": 15701
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 34.572000539483284,
+      "learning_rate": 1.695682940991652e-06,
+      "loss": 0.3685,
+      "step": 15702
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 44.565097503076096,
+      "learning_rate": 1.6947435317454653e-06,
+      "loss": 0.467,
+      "step": 15703
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 50.299851214831556,
+      "learning_rate": 1.6938043586959141e-06,
+      "loss": 0.4988,
+      "step": 15704
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.90051732621847,
+      "learning_rate": 1.6928654218697105e-06,
+      "loss": 0.444,
+      "step": 15705
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 52.08642871305216,
+      "learning_rate": 1.6919267212935531e-06,
+      "loss": 0.4863,
+      "step": 15706
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 50.89994718665061,
+      "learning_rate": 1.6909882569941405e-06,
+      "loss": 0.4498,
+      "step": 15707
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 43.27768871377511,
+      "learning_rate": 1.6900500289981603e-06,
+      "loss": 0.384,
+      "step": 15708
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 50.12076108327234,
+      "learning_rate": 1.6891120373322955e-06,
+      "loss": 0.4189,
+      "step": 15709
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 37.85892098375392,
+      "learning_rate": 1.6881742820232249e-06,
+      "loss": 0.4087,
+      "step": 15710
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 47.061288850962754,
+      "learning_rate": 1.687236763097615e-06,
+      "loss": 0.4721,
+      "step": 15711
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 52.91297772032142,
+      "learning_rate": 1.6862994805821275e-06,
+      "loss": 0.4522,
+      "step": 15712
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 38.73075178834544,
+      "learning_rate": 1.685362434503418e-06,
+      "loss": 0.3764,
+      "step": 15713
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 50.78835866616744,
+      "learning_rate": 1.6844256248881408e-06,
+      "loss": 0.5166,
+      "step": 15714
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.29684281388296,
+      "learning_rate": 1.6834890517629309e-06,
+      "loss": 0.4722,
+      "step": 15715
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 47.062357903578835,
+      "learning_rate": 1.6825527151544307e-06,
+      "loss": 0.4546,
+      "step": 15716
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 49.19601886273053,
+      "learning_rate": 1.6816166150892622e-06,
+      "loss": 0.4191,
+      "step": 15717
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 54.59441183613783,
+      "learning_rate": 1.6806807515940537e-06,
+      "loss": 0.515,
+      "step": 15718
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.11119070130328,
+      "learning_rate": 1.6797451246954166e-06,
+      "loss": 0.4019,
+      "step": 15719
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 48.69224856214633,
+      "learning_rate": 1.6788097344199594e-06,
+      "loss": 0.4645,
+      "step": 15720
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 33.73933772042551,
+      "learning_rate": 1.6778745807942899e-06,
+      "loss": 0.3796,
+      "step": 15721
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 51.237474976547695,
+      "learning_rate": 1.6769396638449954e-06,
+      "loss": 0.5027,
+      "step": 15722
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.36806009358984,
+      "learning_rate": 1.6760049835986703e-06,
+      "loss": 0.4232,
+      "step": 15723
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 37.185100032394345,
+      "learning_rate": 1.6750705400818913e-06,
+      "loss": 0.3984,
+      "step": 15724
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 54.284185889067686,
+      "learning_rate": 1.6741363333212368e-06,
+      "loss": 0.4886,
+      "step": 15725
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 49.38899525141172,
+      "learning_rate": 1.673202363343277e-06,
+      "loss": 0.4105,
+      "step": 15726
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 39.63601227395093,
+      "learning_rate": 1.6722686301745716e-06,
+      "loss": 0.4015,
+      "step": 15727
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.05011608731198,
+      "learning_rate": 1.6713351338416707e-06,
+      "loss": 0.4482,
+      "step": 15728
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 37.03521884300946,
+      "learning_rate": 1.670401874371128e-06,
+      "loss": 0.3983,
+      "step": 15729
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.132960733054915,
+      "learning_rate": 1.6694688517894852e-06,
+      "loss": 0.4235,
+      "step": 15730
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 43.5066461199557,
+      "learning_rate": 1.6685360661232742e-06,
+      "loss": 0.4881,
+      "step": 15731
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.36407585049566,
+      "learning_rate": 1.6676035173990225e-06,
+      "loss": 0.4623,
+      "step": 15732
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 43.0869857508397,
+      "learning_rate": 1.6666712056432522e-06,
+      "loss": 0.4108,
+      "step": 15733
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.715759221239196,
+      "learning_rate": 1.6657391308824777e-06,
+      "loss": 0.4624,
+      "step": 15734
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 56.8683236364088,
+      "learning_rate": 1.6648072931432091e-06,
+      "loss": 0.5585,
+      "step": 15735
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.93228960698616,
+      "learning_rate": 1.663875692451945e-06,
+      "loss": 0.4568,
+      "step": 15736
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 47.02616912163181,
+      "learning_rate": 1.6629443288351777e-06,
+      "loss": 0.4182,
+      "step": 15737
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 114.13448495714653,
+      "learning_rate": 1.6620132023193968e-06,
+      "loss": 0.4182,
+      "step": 15738
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.791916166510376,
+      "learning_rate": 1.661082312931085e-06,
+      "loss": 0.4354,
+      "step": 15739
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 39.28035189406881,
+      "learning_rate": 1.6601516606967127e-06,
+      "loss": 0.3681,
+      "step": 15740
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.794879957356756,
+      "learning_rate": 1.6592212456427492e-06,
+      "loss": 0.4178,
+      "step": 15741
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 41.41883275426861,
+      "learning_rate": 1.6582910677956532e-06,
+      "loss": 0.4321,
+      "step": 15742
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 44.904589378748554,
+      "learning_rate": 1.6573611271818812e-06,
+      "loss": 0.494,
+      "step": 15743
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 39.92434592466652,
+      "learning_rate": 1.6564314238278755e-06,
+      "loss": 0.4026,
+      "step": 15744
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 47.328308397292275,
+      "learning_rate": 1.6555019577600806e-06,
+      "loss": 0.4091,
+      "step": 15745
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 43.07059939638514,
+      "learning_rate": 1.6545727290049307e-06,
+      "loss": 0.42,
+      "step": 15746
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 53.35155167651461,
+      "learning_rate": 1.653643737588847e-06,
+      "loss": 0.4622,
+      "step": 15747
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 39.18000993529403,
+      "learning_rate": 1.652714983538256e-06,
+      "loss": 0.4243,
+      "step": 15748
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 52.278602492020354,
+      "learning_rate": 1.6517864668795658e-06,
+      "loss": 0.51,
+      "step": 15749
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.96556120799219,
+      "learning_rate": 1.650858187639185e-06,
+      "loss": 0.406,
+      "step": 15750
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 43.45667045803531,
+      "learning_rate": 1.6499301458435146e-06,
+      "loss": 0.4216,
+      "step": 15751
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 50.11898467958829,
+      "learning_rate": 1.6490023415189472e-06,
+      "loss": 0.4532,
+      "step": 15752
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.47976800450232,
+      "learning_rate": 1.6480747746918657e-06,
+      "loss": 0.4389,
+      "step": 15753
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 51.48477582066388,
+      "learning_rate": 1.6471474453886504e-06,
+      "loss": 0.5233,
+      "step": 15754
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.435494809538596,
+      "learning_rate": 1.6462203536356769e-06,
+      "loss": 0.4087,
+      "step": 15755
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 38.53466405539699,
+      "learning_rate": 1.645293499459314e-06,
+      "loss": 0.3552,
+      "step": 15756
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 54.481448332038376,
+      "learning_rate": 1.6443668828859126e-06,
+      "loss": 0.4785,
+      "step": 15757
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 48.78250758757361,
+      "learning_rate": 1.6434405039418277e-06,
+      "loss": 0.4985,
+      "step": 15758
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 41.67978484724237,
+      "learning_rate": 1.642514362653408e-06,
+      "loss": 0.4627,
+      "step": 15759
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 46.15385560970943,
+      "learning_rate": 1.6415884590469922e-06,
+      "loss": 0.535,
+      "step": 15760
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 52.930625167367594,
+      "learning_rate": 1.6406627931489116e-06,
+      "loss": 0.4733,
+      "step": 15761
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.171407707596316,
+      "learning_rate": 1.6397373649854874e-06,
+      "loss": 0.4311,
+      "step": 15762
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.04746808205638,
+      "learning_rate": 1.6388121745830433e-06,
+      "loss": 0.3559,
+      "step": 15763
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 51.16868709971773,
+      "learning_rate": 1.637887221967892e-06,
+      "loss": 0.5137,
+      "step": 15764
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 41.0803896285916,
+      "learning_rate": 1.6369625071663365e-06,
+      "loss": 0.3843,
+      "step": 15765
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 44.6122755186633,
+      "learning_rate": 1.636038030204672e-06,
+      "loss": 0.4045,
+      "step": 15766
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.07007213309563,
+      "learning_rate": 1.6351137911091942e-06,
+      "loss": 0.4806,
+      "step": 15767
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.71507482564675,
+      "learning_rate": 1.6341897899061887e-06,
+      "loss": 0.4476,
+      "step": 15768
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 35.797433370084235,
+      "learning_rate": 1.6332660266219292e-06,
+      "loss": 0.405,
+      "step": 15769
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.804284325779975,
+      "learning_rate": 1.632342501282692e-06,
+      "loss": 0.4371,
+      "step": 15770
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 43.51762238169777,
+      "learning_rate": 1.6314192139147366e-06,
+      "loss": 0.4122,
+      "step": 15771
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 48.96944842786859,
+      "learning_rate": 1.630496164544323e-06,
+      "loss": 0.5134,
+      "step": 15772
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 48.273126169322225,
+      "learning_rate": 1.6295733531977054e-06,
+      "loss": 0.5197,
+      "step": 15773
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.27574784231589,
+      "learning_rate": 1.628650779901122e-06,
+      "loss": 0.425,
+      "step": 15774
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.46184037183226,
+      "learning_rate": 1.6277284446808162e-06,
+      "loss": 0.426,
+      "step": 15775
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 45.14127182752917,
+      "learning_rate": 1.6268063475630136e-06,
+      "loss": 0.4684,
+      "step": 15776
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 55.96669297349291,
+      "learning_rate": 1.6258844885739434e-06,
+      "loss": 0.4764,
+      "step": 15777
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 47.55260404682604,
+      "learning_rate": 1.6249628677398155e-06,
+      "loss": 0.4723,
+      "step": 15778
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 46.989123734622744,
+      "learning_rate": 1.6240414850868457e-06,
+      "loss": 0.4448,
+      "step": 15779
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 55.60460879294329,
+      "learning_rate": 1.6231203406412377e-06,
+      "loss": 0.4532,
+      "step": 15780
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 48.971046501017476,
+      "learning_rate": 1.622199434429188e-06,
+      "loss": 0.4721,
+      "step": 15781
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 44.77605106178686,
+      "learning_rate": 1.6212787664768826e-06,
+      "loss": 0.403,
+      "step": 15782
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 47.99856385923735,
+      "learning_rate": 1.620358336810507e-06,
+      "loss": 0.4176,
+      "step": 15783
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.093778394270096,
+      "learning_rate": 1.6194381454562392e-06,
+      "loss": 0.4249,
+      "step": 15784
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.75730688020209,
+      "learning_rate": 1.6185181924402504e-06,
+      "loss": 0.3986,
+      "step": 15785
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 47.003541320941366,
+      "learning_rate": 1.617598477788701e-06,
+      "loss": 0.4407,
+      "step": 15786
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.320937702926216,
+      "learning_rate": 1.6166790015277446e-06,
+      "loss": 0.4623,
+      "step": 15787
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 43.55694516202363,
+      "learning_rate": 1.6157597636835332e-06,
+      "loss": 0.4386,
+      "step": 15788
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 39.66794223230435,
+      "learning_rate": 1.6148407642822128e-06,
+      "loss": 0.3715,
+      "step": 15789
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 44.72004029273878,
+      "learning_rate": 1.6139220033499148e-06,
+      "loss": 0.4802,
+      "step": 15790
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.19657091179353,
+      "learning_rate": 1.6130034809127682e-06,
+      "loss": 0.4065,
+      "step": 15791
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.08435814179369,
+      "learning_rate": 1.6120851969968954e-06,
+      "loss": 0.4314,
+      "step": 15792
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 45.57574218390508,
+      "learning_rate": 1.6111671516284155e-06,
+      "loss": 0.4613,
+      "step": 15793
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.19344965409115,
+      "learning_rate": 1.6102493448334312e-06,
+      "loss": 0.4533,
+      "step": 15794
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 58.748855746062915,
+      "learning_rate": 1.6093317766380511e-06,
+      "loss": 0.4986,
+      "step": 15795
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 41.17303312869667,
+      "learning_rate": 1.608414447068365e-06,
+      "loss": 0.4583,
+      "step": 15796
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 51.21685200049528,
+      "learning_rate": 1.6074973561504614e-06,
+      "loss": 0.4608,
+      "step": 15797
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 45.253403256402954,
+      "learning_rate": 1.6065805039104275e-06,
+      "loss": 0.4358,
+      "step": 15798
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 37.2634561095476,
+      "learning_rate": 1.6056638903743306e-06,
+      "loss": 0.3919,
+      "step": 15799
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 49.82971163042146,
+      "learning_rate": 1.6047475155682445e-06,
+      "loss": 0.4562,
+      "step": 15800
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.36196704454297,
+      "learning_rate": 1.603831379518227e-06,
+      "loss": 0.3921,
+      "step": 15801
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 43.16784216717518,
+      "learning_rate": 1.6029154822503346e-06,
+      "loss": 0.4029,
+      "step": 15802
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 61.270680746982144,
+      "learning_rate": 1.601999823790611e-06,
+      "loss": 0.5495,
+      "step": 15803
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 56.61961372812361,
+      "learning_rate": 1.601084404165103e-06,
+      "loss": 0.5381,
+      "step": 15804
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 55.64456797819844,
+      "learning_rate": 1.600169223399839e-06,
+      "loss": 0.526,
+      "step": 15805
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.36180409238036,
+      "learning_rate": 1.59925428152085e-06,
+      "loss": 0.459,
+      "step": 15806
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.306424943041236,
+      "learning_rate": 1.5983395785541533e-06,
+      "loss": 0.4023,
+      "step": 15807
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 50.18136305478553,
+      "learning_rate": 1.5974251145257635e-06,
+      "loss": 0.4342,
+      "step": 15808
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 58.389248589556,
+      "learning_rate": 1.596510889461691e-06,
+      "loss": 0.4502,
+      "step": 15809
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 53.079848527929364,
+      "learning_rate": 1.5955969033879304e-06,
+      "loss": 0.5706,
+      "step": 15810
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.481814554885545,
+      "learning_rate": 1.59468315633048e-06,
+      "loss": 0.4266,
+      "step": 15811
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 41.31136083835797,
+      "learning_rate": 1.5937696483153208e-06,
+      "loss": 0.4493,
+      "step": 15812
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 45.70986686864914,
+      "learning_rate": 1.592856379368436e-06,
+      "loss": 0.4176,
+      "step": 15813
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 47.53194795155021,
+      "learning_rate": 1.5919433495157998e-06,
+      "loss": 0.4817,
+      "step": 15814
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 43.40955488779941,
+      "learning_rate": 1.5910305587833763e-06,
+      "loss": 0.4372,
+      "step": 15815
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.53007768187934,
+      "learning_rate": 1.5901180071971224e-06,
+      "loss": 0.4562,
+      "step": 15816
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 48.776171095459624,
+      "learning_rate": 1.5892056947829914e-06,
+      "loss": 0.4484,
+      "step": 15817
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 49.82998729346609,
+      "learning_rate": 1.588293621566932e-06,
+      "loss": 0.4654,
+      "step": 15818
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.934439193799896,
+      "learning_rate": 1.5873817875748854e-06,
+      "loss": 0.459,
+      "step": 15819
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.88815101308184,
+      "learning_rate": 1.586470192832774e-06,
+      "loss": 0.4264,
+      "step": 15820
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.28900241754862,
+      "learning_rate": 1.5855588373665298e-06,
+      "loss": 0.4012,
+      "step": 15821
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.087244125896106,
+      "learning_rate": 1.5846477212020695e-06,
+      "loss": 0.4947,
+      "step": 15822
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.8721683056445,
+      "learning_rate": 1.5837368443653067e-06,
+      "loss": 0.4472,
+      "step": 15823
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.5256751931404,
+      "learning_rate": 1.5828262068821453e-06,
+      "loss": 0.4239,
+      "step": 15824
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 44.19352052025066,
+      "learning_rate": 1.5819158087784802e-06,
+      "loss": 0.4307,
+      "step": 15825
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 52.77742588606302,
+      "learning_rate": 1.5810056500802063e-06,
+      "loss": 0.4312,
+      "step": 15826
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 47.8647864530098,
+      "learning_rate": 1.580095730813208e-06,
+      "loss": 0.4307,
+      "step": 15827
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.61624405898583,
+      "learning_rate": 1.5791860510033584e-06,
+      "loss": 0.4238,
+      "step": 15828
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 51.07515932694062,
+      "learning_rate": 1.5782766106765357e-06,
+      "loss": 0.4468,
+      "step": 15829
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 43.732882614267616,
+      "learning_rate": 1.5773674098585955e-06,
+      "loss": 0.4573,
+      "step": 15830
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 44.17553716806816,
+      "learning_rate": 1.5764584485754031e-06,
+      "loss": 0.4091,
+      "step": 15831
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 43.6539262495854,
+      "learning_rate": 1.5755497268528008e-06,
+      "loss": 0.3877,
+      "step": 15832
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 40.827943000723586,
+      "learning_rate": 1.5746412447166371e-06,
+      "loss": 0.4498,
+      "step": 15833
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 47.564616092322346,
+      "learning_rate": 1.57373300219275e-06,
+      "loss": 0.4531,
+      "step": 15834
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.4179887714582,
+      "learning_rate": 1.5728249993069645e-06,
+      "loss": 0.4635,
+      "step": 15835
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 44.085031516652386,
+      "learning_rate": 1.5719172360851088e-06,
+      "loss": 0.4119,
+      "step": 15836
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 58.88573756778054,
+      "learning_rate": 1.5710097125529943e-06,
+      "loss": 0.5011,
+      "step": 15837
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 44.1455714602202,
+      "learning_rate": 1.570102428736432e-06,
+      "loss": 0.4562,
+      "step": 15838
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 65.59024784358279,
+      "learning_rate": 1.5691953846612284e-06,
+      "loss": 0.5539,
+      "step": 15839
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 36.746759300980834,
+      "learning_rate": 1.5682885803531767e-06,
+      "loss": 0.347,
+      "step": 15840
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 55.41003781972605,
+      "learning_rate": 1.5673820158380615e-06,
+      "loss": 0.5584,
+      "step": 15841
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 46.643482538606776,
+      "learning_rate": 1.56647569114167e-06,
+      "loss": 0.4582,
+      "step": 15842
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 56.93695339438805,
+      "learning_rate": 1.5655696062897763e-06,
+      "loss": 0.5595,
+      "step": 15843
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.19082767219483,
+      "learning_rate": 1.5646637613081529e-06,
+      "loss": 0.4282,
+      "step": 15844
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 33.38707239069263,
+      "learning_rate": 1.5637581562225535e-06,
+      "loss": 0.3692,
+      "step": 15845
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 49.26162116915477,
+      "learning_rate": 1.5628527910587376e-06,
+      "loss": 0.4056,
+      "step": 15846
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.87799285940498,
+      "learning_rate": 1.5619476658424516e-06,
+      "loss": 0.3997,
+      "step": 15847
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 51.771860854187864,
+      "learning_rate": 1.5610427805994411e-06,
+      "loss": 0.4789,
+      "step": 15848
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 37.505120983144785,
+      "learning_rate": 1.5601381353554369e-06,
+      "loss": 0.4147,
+      "step": 15849
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 48.98013163656768,
+      "learning_rate": 1.5592337301361648e-06,
+      "loss": 0.489,
+      "step": 15850
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 36.991051764611406,
+      "learning_rate": 1.5583295649673468e-06,
+      "loss": 0.3908,
+      "step": 15851
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 41.8370864154762,
+      "learning_rate": 1.5574256398747011e-06,
+      "loss": 0.388,
+      "step": 15852
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 34.68621637185944,
+      "learning_rate": 1.5565219548839317e-06,
+      "loss": 0.3636,
+      "step": 15853
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 33.43441512236205,
+      "learning_rate": 1.5556185100207356e-06,
+      "loss": 0.4176,
+      "step": 15854
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 45.04894868151107,
+      "learning_rate": 1.5547153053108088e-06,
+      "loss": 0.4353,
+      "step": 15855
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 39.875034212677726,
+      "learning_rate": 1.5538123407798422e-06,
+      "loss": 0.3644,
+      "step": 15856
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 41.39994694941342,
+      "learning_rate": 1.5529096164535084e-06,
+      "loss": 0.4441,
+      "step": 15857
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 36.81308487458292,
+      "learning_rate": 1.552007132357486e-06,
+      "loss": 0.4536,
+      "step": 15858
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 41.412435865597246,
+      "learning_rate": 1.5511048885174361e-06,
+      "loss": 0.3954,
+      "step": 15859
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.834330236801385,
+      "learning_rate": 1.5502028849590212e-06,
+      "loss": 0.4254,
+      "step": 15860
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 48.06392569343421,
+      "learning_rate": 1.5493011217078958e-06,
+      "loss": 0.5273,
+      "step": 15861
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.20231260118756,
+      "learning_rate": 1.5483995987897006e-06,
+      "loss": 0.4203,
+      "step": 15862
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 50.20924524874668,
+      "learning_rate": 1.5474983162300794e-06,
+      "loss": 0.4093,
+      "step": 15863
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 49.244009389435035,
+      "learning_rate": 1.5465972740546586e-06,
+      "loss": 0.4888,
+      "step": 15864
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.2216461546268,
+      "learning_rate": 1.5456964722890688e-06,
+      "loss": 0.3993,
+      "step": 15865
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 51.761408690846096,
+      "learning_rate": 1.5447959109589228e-06,
+      "loss": 0.488,
+      "step": 15866
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 56.251682290246116,
+      "learning_rate": 1.5438955900898355e-06,
+      "loss": 0.4414,
+      "step": 15867
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 37.3169633218208,
+      "learning_rate": 1.5429955097074133e-06,
+      "loss": 0.3806,
+      "step": 15868
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 40.787397758736866,
+      "learning_rate": 1.5420956698372513e-06,
+      "loss": 0.4339,
+      "step": 15869
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 50.76691974034958,
+      "learning_rate": 1.5411960705049389e-06,
+      "loss": 0.4886,
+      "step": 15870
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 36.18533402715588,
+      "learning_rate": 1.5402967117360622e-06,
+      "loss": 0.3717,
+      "step": 15871
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 49.56965359970597,
+      "learning_rate": 1.5393975935561977e-06,
+      "loss": 0.4938,
+      "step": 15872
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 47.84810159179963,
+      "learning_rate": 1.538498715990918e-06,
+      "loss": 0.4381,
+      "step": 15873
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 46.195979293689454,
+      "learning_rate": 1.5376000790657864e-06,
+      "loss": 0.4319,
+      "step": 15874
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 44.03686525535451,
+      "learning_rate": 1.5367016828063553e-06,
+      "loss": 0.4603,
+      "step": 15875
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 45.82740297733945,
+      "learning_rate": 1.5358035272381765e-06,
+      "loss": 0.4463,
+      "step": 15876
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 45.75577992266974,
+      "learning_rate": 1.5349056123867977e-06,
+      "loss": 0.3676,
+      "step": 15877
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 54.29573370962181,
+      "learning_rate": 1.5340079382777506e-06,
+      "loss": 0.5713,
+      "step": 15878
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.31349398971041,
+      "learning_rate": 1.5331105049365624e-06,
+      "loss": 0.4039,
+      "step": 15879
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 45.5336302250966,
+      "learning_rate": 1.5322133123887595e-06,
+      "loss": 0.4067,
+      "step": 15880
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 51.82013082048282,
+      "learning_rate": 1.5313163606598591e-06,
+      "loss": 0.4804,
+      "step": 15881
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 46.769939872290344,
+      "learning_rate": 1.5304196497753642e-06,
+      "loss": 0.4187,
+      "step": 15882
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 44.402206134759076,
+      "learning_rate": 1.5295231797607824e-06,
+      "loss": 0.4562,
+      "step": 15883
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 47.765339978746276,
+      "learning_rate": 1.528626950641604e-06,
+      "loss": 0.4155,
+      "step": 15884
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 50.72760618291503,
+      "learning_rate": 1.527730962443319e-06,
+      "loss": 0.4568,
+      "step": 15885
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 39.472157455833845,
+      "learning_rate": 1.5268352151914124e-06,
+      "loss": 0.3936,
+      "step": 15886
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 42.00869125743949,
+      "learning_rate": 1.5259397089113525e-06,
+      "loss": 0.4089,
+      "step": 15887
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 52.323892763381814,
+      "learning_rate": 1.5250444436286138e-06,
+      "loss": 0.4297,
+      "step": 15888
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 42.03797861575995,
+      "learning_rate": 1.52414941936865e-06,
+      "loss": 0.3647,
+      "step": 15889
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 36.264606864209924,
+      "learning_rate": 1.5232546361569212e-06,
+      "loss": 0.3501,
+      "step": 15890
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 46.613665136177325,
+      "learning_rate": 1.5223600940188688e-06,
+      "loss": 0.4487,
+      "step": 15891
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 52.467323660988384,
+      "learning_rate": 1.521465792979937e-06,
+      "loss": 0.5068,
+      "step": 15892
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 38.666867055592924,
+      "learning_rate": 1.52057173306556e-06,
+      "loss": 0.3892,
+      "step": 15893
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 37.33823827186688,
+      "learning_rate": 1.5196779143011619e-06,
+      "loss": 0.3864,
+      "step": 15894
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 46.56328655321038,
+      "learning_rate": 1.5187843367121623e-06,
+      "loss": 0.4085,
+      "step": 15895
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 41.293662010827774,
+      "learning_rate": 1.5178910003239732e-06,
+      "loss": 0.4211,
+      "step": 15896
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 42.23059764853327,
+      "learning_rate": 1.5169979051620033e-06,
+      "loss": 0.3651,
+      "step": 15897
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 40.47487609919716,
+      "learning_rate": 1.5161050512516528e-06,
+      "loss": 0.4359,
+      "step": 15898
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 53.50744592688935,
+      "learning_rate": 1.515212438618312e-06,
+      "loss": 0.5006,
+      "step": 15899
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.97407729377395,
+      "learning_rate": 1.5143200672873626e-06,
+      "loss": 0.4253,
+      "step": 15900
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 44.401932551450834,
+      "learning_rate": 1.5134279372841876e-06,
+      "loss": 0.4423,
+      "step": 15901
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 42.4790636609849,
+      "learning_rate": 1.512536048634159e-06,
+      "loss": 0.3947,
+      "step": 15902
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 58.50885411447153,
+      "learning_rate": 1.5116444013626407e-06,
+      "loss": 0.4286,
+      "step": 15903
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 53.02529825273272,
+      "learning_rate": 1.5107529954949873e-06,
+      "loss": 0.466,
+      "step": 15904
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 36.944921958119174,
+      "learning_rate": 1.5098618310565527e-06,
+      "loss": 0.3603,
+      "step": 15905
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 46.80104108043718,
+      "learning_rate": 1.5089709080726811e-06,
+      "loss": 0.4564,
+      "step": 15906
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 49.47967873184522,
+      "learning_rate": 1.5080802265687145e-06,
+      "loss": 0.4433,
+      "step": 15907
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 39.2374270918194,
+      "learning_rate": 1.5071897865699748e-06,
+      "loss": 0.3519,
+      "step": 15908
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.90360922580791,
+      "learning_rate": 1.5062995881017882e-06,
+      "loss": 0.4052,
+      "step": 15909
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 53.306358842339435,
+      "learning_rate": 1.5054096311894727e-06,
+      "loss": 0.404,
+      "step": 15910
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 37.17197047153574,
+      "learning_rate": 1.5045199158583411e-06,
+      "loss": 0.3656,
+      "step": 15911
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 37.21766302922121,
+      "learning_rate": 1.5036304421336933e-06,
+      "loss": 0.3794,
+      "step": 15912
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 43.26789548570935,
+      "learning_rate": 1.5027412100408234e-06,
+      "loss": 0.4216,
+      "step": 15913
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 44.462303466405416,
+      "learning_rate": 1.5018522196050223e-06,
+      "loss": 0.4098,
+      "step": 15914
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 46.13257756314708,
+      "learning_rate": 1.5009634708515753e-06,
+      "loss": 0.3805,
+      "step": 15915
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 49.19964109983806,
+      "learning_rate": 1.5000749638057521e-06,
+      "loss": 0.4098,
+      "step": 15916
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 40.748415962979145,
+      "learning_rate": 1.4991866984928283e-06,
+      "loss": 0.3967,
+      "step": 15917
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 48.005810333789356,
+      "learning_rate": 1.498298674938059e-06,
+      "loss": 0.3755,
+      "step": 15918
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 37.70152316937404,
+      "learning_rate": 1.4974108931667042e-06,
+      "loss": 0.3487,
+      "step": 15919
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 54.51756342653063,
+      "learning_rate": 1.496523353204008e-06,
+      "loss": 0.4903,
+      "step": 15920
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 50.65811990137703,
+      "learning_rate": 1.4956360550752125e-06,
+      "loss": 0.4416,
+      "step": 15921
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 41.711959051843365,
+      "learning_rate": 1.4947489988055552e-06,
+      "loss": 0.4236,
+      "step": 15922
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 47.80462370679813,
+      "learning_rate": 1.4938621844202595e-06,
+      "loss": 0.4368,
+      "step": 15923
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 55.89255894660188,
+      "learning_rate": 1.4929756119445481e-06,
+      "loss": 0.4622,
+      "step": 15924
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.800456796530256,
+      "learning_rate": 1.4920892814036324e-06,
+      "loss": 0.3657,
+      "step": 15925
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 50.201361450981935,
+      "learning_rate": 1.4912031928227211e-06,
+      "loss": 0.42,
+      "step": 15926
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 41.591572372787205,
+      "learning_rate": 1.4903173462270148e-06,
+      "loss": 0.3516,
+      "step": 15927
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 52.656914554556685,
+      "learning_rate": 1.489431741641706e-06,
+      "loss": 0.4518,
+      "step": 15928
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.01161269738056,
+      "learning_rate": 1.4885463790919774e-06,
+      "loss": 0.3714,
+      "step": 15929
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 46.07787071201842,
+      "learning_rate": 1.4876612586030104e-06,
+      "loss": 0.4302,
+      "step": 15930
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 44.800788248101334,
+      "learning_rate": 1.4867763801999768e-06,
+      "loss": 0.4216,
+      "step": 15931
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 40.40709733775248,
+      "learning_rate": 1.485891743908049e-06,
+      "loss": 0.3699,
+      "step": 15932
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 44.99315595503598,
+      "learning_rate": 1.4850073497523743e-06,
+      "loss": 0.4349,
+      "step": 15933
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 52.838993582587655,
+      "learning_rate": 1.4841231977581094e-06,
+      "loss": 0.4352,
+      "step": 15934
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.44704656852037,
+      "learning_rate": 1.4832392879503986e-06,
+      "loss": 0.4118,
+      "step": 15935
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.75000721669276,
+      "learning_rate": 1.4823556203543833e-06,
+      "loss": 0.421,
+      "step": 15936
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 49.500365525639666,
+      "learning_rate": 1.4814721949951916e-06,
+      "loss": 0.4208,
+      "step": 15937
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 46.23192738728713,
+      "learning_rate": 1.4805890118979448e-06,
+      "loss": 0.4399,
+      "step": 15938
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 36.589758338474226,
+      "learning_rate": 1.479706071087762e-06,
+      "loss": 0.4052,
+      "step": 15939
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 48.2200277113469,
+      "learning_rate": 1.4788233725897583e-06,
+      "loss": 0.4149,
+      "step": 15940
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 49.01694272002607,
+      "learning_rate": 1.4779409164290292e-06,
+      "loss": 0.4349,
+      "step": 15941
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 41.561198637243166,
+      "learning_rate": 1.4770587026306782e-06,
+      "loss": 0.3473,
+      "step": 15942
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 39.78073618422933,
+      "learning_rate": 1.4761767312197906e-06,
+      "loss": 0.4216,
+      "step": 15943
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 50.894444772914916,
+      "learning_rate": 1.4752950022214519e-06,
+      "loss": 0.4338,
+      "step": 15944
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 39.66339140815137,
+      "learning_rate": 1.4744135156607332e-06,
+      "loss": 0.4017,
+      "step": 15945
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 46.95743350859556,
+      "learning_rate": 1.4735322715627075e-06,
+      "loss": 0.4429,
+      "step": 15946
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.54542192155525,
+      "learning_rate": 1.4726512699524399e-06,
+      "loss": 0.4025,
+      "step": 15947
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 38.42062526685568,
+      "learning_rate": 1.4717705108549773e-06,
+      "loss": 0.4443,
+      "step": 15948
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 43.0770333573448,
+      "learning_rate": 1.4708899942953757e-06,
+      "loss": 0.4218,
+      "step": 15949
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 47.60722258071266,
+      "learning_rate": 1.4700097202986719e-06,
+      "loss": 0.4679,
+      "step": 15950
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 50.75191927789995,
+      "learning_rate": 1.4691296888899031e-06,
+      "loss": 0.4351,
+      "step": 15951
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 43.37171725182534,
+      "learning_rate": 1.468249900094093e-06,
+      "loss": 0.3854,
+      "step": 15952
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 59.73390898845798,
+      "learning_rate": 1.4673703539362683e-06,
+      "loss": 0.4923,
+      "step": 15953
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 48.02019102393799,
+      "learning_rate": 1.4664910504414355e-06,
+      "loss": 0.4829,
+      "step": 15954
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 43.973231135386555,
+      "learning_rate": 1.4656119896346055e-06,
+      "loss": 0.4616,
+      "step": 15955
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 47.70066433809805,
+      "learning_rate": 1.4647331715407809e-06,
+      "loss": 0.4275,
+      "step": 15956
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 54.49282876789315,
+      "learning_rate": 1.4638545961849525e-06,
+      "loss": 0.4647,
+      "step": 15957
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 54.25181549743671,
+      "learning_rate": 1.462976263592103e-06,
+      "loss": 0.4576,
+      "step": 15958
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 40.16626329074356,
+      "learning_rate": 1.4620981737872142e-06,
+      "loss": 0.404,
+      "step": 15959
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 54.2764170575452,
+      "learning_rate": 1.4612203267952584e-06,
+      "loss": 0.4428,
+      "step": 15960
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 48.324201294465134,
+      "learning_rate": 1.4603427226412048e-06,
+      "loss": 0.4161,
+      "step": 15961
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 41.63274896222787,
+      "learning_rate": 1.4594653613500086e-06,
+      "loss": 0.4168,
+      "step": 15962
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 49.54021242326418,
+      "learning_rate": 1.458588242946618e-06,
+      "loss": 0.4954,
+      "step": 15963
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 64.84176990105505,
+      "learning_rate": 1.457711367455983e-06,
+      "loss": 0.419,
+      "step": 15964
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 45.26477763153474,
+      "learning_rate": 1.4568347349030398e-06,
+      "loss": 0.4079,
+      "step": 15965
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 44.7616684475007,
+      "learning_rate": 1.4559583453127213e-06,
+      "loss": 0.3881,
+      "step": 15966
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 46.56212675213826,
+      "learning_rate": 1.455082198709945e-06,
+      "loss": 0.4538,
+      "step": 15967
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 45.183906728881254,
+      "learning_rate": 1.4542062951196335e-06,
+      "loss": 0.3857,
+      "step": 15968
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 41.16643140155475,
+      "learning_rate": 1.4533306345666954e-06,
+      "loss": 0.3935,
+      "step": 15969
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 53.657698596735585,
+      "learning_rate": 1.4524552170760374e-06,
+      "loss": 0.434,
+      "step": 15970
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 42.180686372198146,
+      "learning_rate": 1.4515800426725524e-06,
+      "loss": 0.4301,
+      "step": 15971
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 47.83763504318262,
+      "learning_rate": 1.4507051113811277e-06,
+      "loss": 0.4052,
+      "step": 15972
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 36.653237458885165,
+      "learning_rate": 1.449830423226649e-06,
+      "loss": 0.3749,
+      "step": 15973
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 47.896112861276684,
+      "learning_rate": 1.4489559782339935e-06,
+      "loss": 0.4412,
+      "step": 15974
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.91697680214828,
+      "learning_rate": 1.4480817764280254e-06,
+      "loss": 0.3873,
+      "step": 15975
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.961184991906514,
+      "learning_rate": 1.4472078178336102e-06,
+      "loss": 0.4357,
+      "step": 15976
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 57.113913004831396,
+      "learning_rate": 1.4463341024756006e-06,
+      "loss": 0.4395,
+      "step": 15977
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 51.800486001198166,
+      "learning_rate": 1.4454606303788466e-06,
+      "loss": 0.3696,
+      "step": 15978
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 50.77020035000714,
+      "learning_rate": 1.4445874015681872e-06,
+      "loss": 0.4971,
+      "step": 15979
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.707125898301854,
+      "learning_rate": 1.4437144160684558e-06,
+      "loss": 0.4314,
+      "step": 15980
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.214132689884096,
+      "learning_rate": 1.4428416739044836e-06,
+      "loss": 0.4571,
+      "step": 15981
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.194812858021116,
+      "learning_rate": 1.441969175101089e-06,
+      "loss": 0.4187,
+      "step": 15982
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 49.152833779875515,
+      "learning_rate": 1.4410969196830827e-06,
+      "loss": 0.4015,
+      "step": 15983
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 46.28738198364443,
+      "learning_rate": 1.4402249076752716e-06,
+      "loss": 0.458,
+      "step": 15984
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.90549740928279,
+      "learning_rate": 1.4393531391024585e-06,
+      "loss": 0.4775,
+      "step": 15985
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.82188548772979,
+      "learning_rate": 1.4384816139894365e-06,
+      "loss": 0.3733,
+      "step": 15986
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 55.46518011076447,
+      "learning_rate": 1.437610332360988e-06,
+      "loss": 0.4382,
+      "step": 15987
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.786666683247745,
+      "learning_rate": 1.4367392942418912e-06,
+      "loss": 0.391,
+      "step": 15988
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 36.04983104264113,
+      "learning_rate": 1.4358684996569194e-06,
+      "loss": 0.3308,
+      "step": 15989
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.57371380495019,
+      "learning_rate": 1.4349979486308396e-06,
+      "loss": 0.39,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.37504700601743,
+      "learning_rate": 1.4341276411884076e-06,
+      "loss": 0.3903,
+      "step": 15991
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 47.61273241364431,
+      "learning_rate": 1.4332575773543732e-06,
+      "loss": 0.4198,
+      "step": 15992
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 44.82497188026836,
+      "learning_rate": 1.4323877571534817e-06,
+      "loss": 0.4526,
+      "step": 15993
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 50.79192125612855,
+      "learning_rate": 1.4315181806104694e-06,
+      "loss": 0.442,
+      "step": 15994
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 49.882948115440044,
+      "learning_rate": 1.4306488477500714e-06,
+      "loss": 0.4633,
+      "step": 15995
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 52.46760372888668,
+      "learning_rate": 1.4297797585970063e-06,
+      "loss": 0.4112,
+      "step": 15996
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 46.23521759307077,
+      "learning_rate": 1.4289109131759894e-06,
+      "loss": 0.4321,
+      "step": 15997
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 55.72769859115321,
+      "learning_rate": 1.4280423115117316e-06,
+      "loss": 0.4414,
+      "step": 15998
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 51.66389103273636,
+      "learning_rate": 1.4271739536289386e-06,
+      "loss": 0.3653,
+      "step": 15999
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 47.53495323707701,
+      "learning_rate": 1.4263058395523044e-06,
+      "loss": 0.4526,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 47.00963818867528,
+      "learning_rate": 1.4254379693065134e-06,
+      "loss": 0.4136,
+      "step": 16001
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 52.01743341237515,
+      "learning_rate": 1.4245703429162504e-06,
+      "loss": 0.4489,
+      "step": 16002
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 45.95301136053381,
+      "learning_rate": 1.4237029604061936e-06,
+      "loss": 0.4466,
+      "step": 16003
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.95083183715667,
+      "learning_rate": 1.422835821801004e-06,
+      "loss": 0.4589,
+      "step": 16004
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.99738328622506,
+      "learning_rate": 1.42196892712535e-06,
+      "loss": 0.4133,
+      "step": 16005
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 40.63532511448108,
+      "learning_rate": 1.4211022764038785e-06,
+      "loss": 0.4469,
+      "step": 16006
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 55.44151513527011,
+      "learning_rate": 1.4202358696612417e-06,
+      "loss": 0.5328,
+      "step": 16007
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 50.82597271935533,
+      "learning_rate": 1.4193697069220758e-06,
+      "loss": 0.4941,
+      "step": 16008
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 38.66167237992843,
+      "learning_rate": 1.4185037882110165e-06,
+      "loss": 0.367,
+      "step": 16009
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.25281196339298,
+      "learning_rate": 1.4176381135526906e-06,
+      "loss": 0.4358,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.454019506438584,
+      "learning_rate": 1.4167726829717155e-06,
+      "loss": 0.3992,
+      "step": 16011
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 40.390348876953745,
+      "learning_rate": 1.4159074964927055e-06,
+      "loss": 0.469,
+      "step": 16012
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 44.11651239001876,
+      "learning_rate": 1.4150425541402623e-06,
+      "loss": 0.419,
+      "step": 16013
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 50.65479637474829,
+      "learning_rate": 1.4141778559389875e-06,
+      "loss": 0.4634,
+      "step": 16014
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.39312364727703,
+      "learning_rate": 1.413313401913473e-06,
+      "loss": 0.4357,
+      "step": 16015
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 58.63756355082155,
+      "learning_rate": 1.4124491920883031e-06,
+      "loss": 0.5633,
+      "step": 16016
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 37.15448165163397,
+      "learning_rate": 1.4115852264880525e-06,
+      "loss": 0.3472,
+      "step": 16017
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 46.859411073615476,
+      "learning_rate": 1.4107215051372935e-06,
+      "loss": 0.3764,
+      "step": 16018
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 49.68599514391824,
+      "learning_rate": 1.409858028060591e-06,
+      "loss": 0.4428,
+      "step": 16019
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 37.946577164559386,
+      "learning_rate": 1.4089947952825057e-06,
+      "loss": 0.3672,
+      "step": 16020
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 39.89393179755105,
+      "learning_rate": 1.4081318068275773e-06,
+      "loss": 0.3565,
+      "step": 16021
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 46.761349478453695,
+      "learning_rate": 1.4072690627203544e-06,
+      "loss": 0.5175,
+      "step": 16022
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 54.044981637408625,
+      "learning_rate": 1.406406562985373e-06,
+      "loss": 0.4742,
+      "step": 16023
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 40.90458639097356,
+      "learning_rate": 1.4055443076471641e-06,
+      "loss": 0.3503,
+      "step": 16024
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 60.59914109405075,
+      "learning_rate": 1.4046822967302477e-06,
+      "loss": 0.535,
+      "step": 16025
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.098488552551075,
+      "learning_rate": 1.4038205302591368e-06,
+      "loss": 0.4323,
+      "step": 16026
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.815744843090194,
+      "learning_rate": 1.4029590082583411e-06,
+      "loss": 0.4552,
+      "step": 16027
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.25558474834746,
+      "learning_rate": 1.4020977307523642e-06,
+      "loss": 0.4174,
+      "step": 16028
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 47.96859088088494,
+      "learning_rate": 1.4012366977656954e-06,
+      "loss": 0.4316,
+      "step": 16029
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 44.27946771172549,
+      "learning_rate": 1.4003759093228275e-06,
+      "loss": 0.364,
+      "step": 16030
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 50.82258296647722,
+      "learning_rate": 1.3995153654482363e-06,
+      "loss": 0.4457,
+      "step": 16031
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 51.23648406652736,
+      "learning_rate": 1.3986550661663978e-06,
+      "loss": 0.4915,
+      "step": 16032
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 45.30758871888924,
+      "learning_rate": 1.3977950115017768e-06,
+      "loss": 0.4283,
+      "step": 16033
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 43.8511679227398,
+      "learning_rate": 1.3969352014788328e-06,
+      "loss": 0.4179,
+      "step": 16034
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.97462914214191,
+      "learning_rate": 1.3960756361220219e-06,
+      "loss": 0.3846,
+      "step": 16035
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 40.99015069041364,
+      "learning_rate": 1.395216315455784e-06,
+      "loss": 0.457,
+      "step": 16036
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 37.85647482160493,
+      "learning_rate": 1.3943572395045624e-06,
+      "loss": 0.4031,
+      "step": 16037
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 40.90134746761524,
+      "learning_rate": 1.393498408292785e-06,
+      "loss": 0.397,
+      "step": 16038
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 50.52488060172203,
+      "learning_rate": 1.3926398218448778e-06,
+      "loss": 0.4333,
+      "step": 16039
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 40.55078785856469,
+      "learning_rate": 1.3917814801852615e-06,
+      "loss": 0.388,
+      "step": 16040
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.791054313339394,
+      "learning_rate": 1.390923383338345e-06,
+      "loss": 0.4391,
+      "step": 16041
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 49.1848909735511,
+      "learning_rate": 1.3900655313285283e-06,
+      "loss": 0.4778,
+      "step": 16042
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 44.772672322833216,
+      "learning_rate": 1.3892079241802104e-06,
+      "loss": 0.4365,
+      "step": 16043
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 36.63053018070659,
+      "learning_rate": 1.3883505619177829e-06,
+      "loss": 0.3829,
+      "step": 16044
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.87600067823362,
+      "learning_rate": 1.3874934445656319e-06,
+      "loss": 0.4272,
+      "step": 16045
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 41.82845681607144,
+      "learning_rate": 1.3866365721481256e-06,
+      "loss": 0.4227,
+      "step": 16046
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 49.56140006096746,
+      "learning_rate": 1.3857799446896358e-06,
+      "loss": 0.4126,
+      "step": 16047
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 41.36629885546634,
+      "learning_rate": 1.3849235622145253e-06,
+      "loss": 0.4334,
+      "step": 16048
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 33.456521679170486,
+      "learning_rate": 1.3840674247471508e-06,
+      "loss": 0.3764,
+      "step": 16049
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 37.78384357697373,
+      "learning_rate": 1.383211532311859e-06,
+      "loss": 0.3852,
+      "step": 16050
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.67098804317023,
+      "learning_rate": 1.3823558849329877e-06,
+      "loss": 0.4651,
+      "step": 16051
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 39.19895816216479,
+      "learning_rate": 1.3815004826348744e-06,
+      "loss": 0.4487,
+      "step": 16052
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 47.4385464368152,
+      "learning_rate": 1.3806453254418483e-06,
+      "loss": 0.453,
+      "step": 16053
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 51.11593639924698,
+      "learning_rate": 1.3797904133782269e-06,
+      "loss": 0.502,
+      "step": 16054
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 47.60173762640018,
+      "learning_rate": 1.3789357464683206e-06,
+      "loss": 0.4212,
+      "step": 16055
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 59.36306210217819,
+      "learning_rate": 1.3780813247364388e-06,
+      "loss": 0.4682,
+      "step": 16056
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.69471535852136,
+      "learning_rate": 1.3772271482068799e-06,
+      "loss": 0.482,
+      "step": 16057
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 52.736003166777905,
+      "learning_rate": 1.3763732169039401e-06,
+      "loss": 0.4228,
+      "step": 16058
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 44.98403283400154,
+      "learning_rate": 1.375519530851901e-06,
+      "loss": 0.3924,
+      "step": 16059
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.93164064942167,
+      "learning_rate": 1.3746660900750386e-06,
+      "loss": 0.4189,
+      "step": 16060
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 48.3965579944908,
+      "learning_rate": 1.373812894597627e-06,
+      "loss": 0.4444,
+      "step": 16061
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 49.23956643363781,
+      "learning_rate": 1.3729599444439323e-06,
+      "loss": 0.4538,
+      "step": 16062
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 56.912759499698694,
+      "learning_rate": 1.3721072396382085e-06,
+      "loss": 0.4653,
+      "step": 16063
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 38.81601257216717,
+      "learning_rate": 1.3712547802047094e-06,
+      "loss": 0.4921,
+      "step": 16064
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 60.79778054415184,
+      "learning_rate": 1.3704025661676757e-06,
+      "loss": 0.4864,
+      "step": 16065
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 40.54492774653515,
+      "learning_rate": 1.3695505975513456e-06,
+      "loss": 0.4285,
+      "step": 16066
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 59.25834321782057,
+      "learning_rate": 1.3686988743799466e-06,
+      "loss": 0.5255,
+      "step": 16067
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 48.50412840821894,
+      "learning_rate": 1.3678473966777018e-06,
+      "loss": 0.4848,
+      "step": 16068
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.69418534259871,
+      "learning_rate": 1.36699616446883e-06,
+      "loss": 0.4422,
+      "step": 16069
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 47.14103113390821,
+      "learning_rate": 1.3661451777775369e-06,
+      "loss": 0.3752,
+      "step": 16070
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 50.984177032843924,
+      "learning_rate": 1.3652944366280218e-06,
+      "loss": 0.4503,
+      "step": 16071
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 49.763207743100374,
+      "learning_rate": 1.364443941044482e-06,
+      "loss": 0.4337,
+      "step": 16072
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 42.46292920955482,
+      "learning_rate": 1.363593691051105e-06,
+      "loss": 0.3886,
+      "step": 16073
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 44.91481625502906,
+      "learning_rate": 1.3627436866720734e-06,
+      "loss": 0.4361,
+      "step": 16074
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 41.68119359964275,
+      "learning_rate": 1.3618939279315591e-06,
+      "loss": 0.4682,
+      "step": 16075
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.25220497133338,
+      "learning_rate": 1.3610444148537261e-06,
+      "loss": 0.4893,
+      "step": 16076
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 41.16928787354959,
+      "learning_rate": 1.3601951474627362e-06,
+      "loss": 0.3984,
+      "step": 16077
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 52.99045299518536,
+      "learning_rate": 1.3593461257827433e-06,
+      "loss": 0.4254,
+      "step": 16078
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 36.19582922023432,
+      "learning_rate": 1.3584973498378928e-06,
+      "loss": 0.3488,
+      "step": 16079
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 45.01465305601858,
+      "learning_rate": 1.3576488196523207e-06,
+      "loss": 0.4171,
+      "step": 16080
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 55.181027038206736,
+      "learning_rate": 1.3568005352501602e-06,
+      "loss": 0.4761,
+      "step": 16081
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.53942012345283,
+      "learning_rate": 1.355952496655536e-06,
+      "loss": 0.4467,
+      "step": 16082
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 38.86567021001002,
+      "learning_rate": 1.3551047038925692e-06,
+      "loss": 0.397,
+      "step": 16083
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 39.47235737290594,
+      "learning_rate": 1.3542571569853669e-06,
+      "loss": 0.3907,
+      "step": 16084
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 47.70160299522657,
+      "learning_rate": 1.353409855958031e-06,
+      "loss": 0.5034,
+      "step": 16085
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.59024675847689,
+      "learning_rate": 1.3525628008346613e-06,
+      "loss": 0.3548,
+      "step": 16086
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 45.005167775650335,
+      "learning_rate": 1.3517159916393485e-06,
+      "loss": 0.4211,
+      "step": 16087
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 37.68893235048401,
+      "learning_rate": 1.350869428396172e-06,
+      "loss": 0.4127,
+      "step": 16088
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 40.68312006989159,
+      "learning_rate": 1.3500231111292118e-06,
+      "loss": 0.3514,
+      "step": 16089
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 48.589389107352275,
+      "learning_rate": 1.3491770398625315e-06,
+      "loss": 0.4065,
+      "step": 16090
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 53.17716627671825,
+      "learning_rate": 1.3483312146201988e-06,
+      "loss": 0.504,
+      "step": 16091
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 52.9523706233279,
+      "learning_rate": 1.3474856354262622e-06,
+      "loss": 0.4478,
+      "step": 16092
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.44146043670186,
+      "learning_rate": 1.3466403023047724e-06,
+      "loss": 0.4317,
+      "step": 16093
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 45.005034414116544,
+      "learning_rate": 1.3457952152797727e-06,
+      "loss": 0.4707,
+      "step": 16094
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 49.259348984062704,
+      "learning_rate": 1.3449503743752945e-06,
+      "loss": 0.4272,
+      "step": 16095
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 50.92786758244504,
+      "learning_rate": 1.3441057796153634e-06,
+      "loss": 0.4644,
+      "step": 16096
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 51.429873924375286,
+      "learning_rate": 1.3432614310239989e-06,
+      "loss": 0.4273,
+      "step": 16097
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 41.99741403248121,
+      "learning_rate": 1.3424173286252185e-06,
+      "loss": 0.463,
+      "step": 16098
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 60.6498827021912,
+      "learning_rate": 1.3415734724430219e-06,
+      "loss": 0.4798,
+      "step": 16099
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.80339913209941,
+      "learning_rate": 1.340729862501413e-06,
+      "loss": 0.4194,
+      "step": 16100
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 45.997994773115735,
+      "learning_rate": 1.3398864988243786e-06,
+      "loss": 0.4839,
+      "step": 16101
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 48.31314326323192,
+      "learning_rate": 1.3390433814359072e-06,
+      "loss": 0.5313,
+      "step": 16102
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.934078377931826,
+      "learning_rate": 1.338200510359977e-06,
+      "loss": 0.3746,
+      "step": 16103
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 46.39942550570723,
+      "learning_rate": 1.3373578856205583e-06,
+      "loss": 0.3951,
+      "step": 16104
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 51.262527530367905,
+      "learning_rate": 1.3365155072416103e-06,
+      "loss": 0.4752,
+      "step": 16105
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 50.00003011702584,
+      "learning_rate": 1.335673375247094e-06,
+      "loss": 0.4495,
+      "step": 16106
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 41.329969910832894,
+      "learning_rate": 1.3348314896609581e-06,
+      "loss": 0.4125,
+      "step": 16107
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 48.05427220412696,
+      "learning_rate": 1.3339898505071503e-06,
+      "loss": 0.3513,
+      "step": 16108
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 39.28443529784257,
+      "learning_rate": 1.3331484578095976e-06,
+      "loss": 0.3511,
+      "step": 16109
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 48.58473278906906,
+      "learning_rate": 1.3323073115922324e-06,
+      "loss": 0.4728,
+      "step": 16110
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 58.54192731082041,
+      "learning_rate": 1.3314664118789766e-06,
+      "loss": 0.4755,
+      "step": 16111
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 44.06120578677831,
+      "learning_rate": 1.3306257586937476e-06,
+      "loss": 0.4798,
+      "step": 16112
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.490216371722084,
+      "learning_rate": 1.3297853520604497e-06,
+      "loss": 0.4394,
+      "step": 16113
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 55.895404264737145,
+      "learning_rate": 1.3289451920029838e-06,
+      "loss": 0.5096,
+      "step": 16114
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 41.76042563074437,
+      "learning_rate": 1.328105278545243e-06,
+      "loss": 0.383,
+      "step": 16115
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 45.66106944460469,
+      "learning_rate": 1.3272656117111183e-06,
+      "loss": 0.4068,
+      "step": 16116
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.193114923046856,
+      "learning_rate": 1.3264261915244836e-06,
+      "loss": 0.4131,
+      "step": 16117
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.74746288333421,
+      "learning_rate": 1.3255870180092157e-06,
+      "loss": 0.441,
+      "step": 16118
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 50.68379854546775,
+      "learning_rate": 1.3247480911891775e-06,
+      "loss": 0.5355,
+      "step": 16119
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.735783566888884,
+      "learning_rate": 1.3239094110882279e-06,
+      "loss": 0.4533,
+      "step": 16120
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 56.48692102880501,
+      "learning_rate": 1.323070977730222e-06,
+      "loss": 0.4751,
+      "step": 16121
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.36612220759968,
+      "learning_rate": 1.3222327911389988e-06,
+      "loss": 0.4059,
+      "step": 16122
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 40.15986234692619,
+      "learning_rate": 1.3213948513384011e-06,
+      "loss": 0.4318,
+      "step": 16123
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.45698697548657,
+      "learning_rate": 1.320557158352256e-06,
+      "loss": 0.4636,
+      "step": 16124
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 52.244308645394725,
+      "learning_rate": 1.3197197122043892e-06,
+      "loss": 0.4993,
+      "step": 16125
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.030854674037045,
+      "learning_rate": 1.3188825129186134e-06,
+      "loss": 0.4403,
+      "step": 16126
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 46.77603983740156,
+      "learning_rate": 1.3180455605187404e-06,
+      "loss": 0.4607,
+      "step": 16127
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.290598222863174,
+      "learning_rate": 1.3172088550285756e-06,
+      "loss": 0.4218,
+      "step": 16128
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 40.38878075762871,
+      "learning_rate": 1.3163723964719122e-06,
+      "loss": 0.4356,
+      "step": 16129
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 53.96504803072481,
+      "learning_rate": 1.3155361848725357e-06,
+      "loss": 0.4696,
+      "step": 16130
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.70173820431508,
+      "learning_rate": 1.3147002202542291e-06,
+      "loss": 0.5022,
+      "step": 16131
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 44.700206726671475,
+      "learning_rate": 1.3138645026407682e-06,
+      "loss": 0.4246,
+      "step": 16132
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.85486398448194,
+      "learning_rate": 1.3130290320559235e-06,
+      "loss": 0.4689,
+      "step": 16133
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.20860647294844,
+      "learning_rate": 1.3121938085234466e-06,
+      "loss": 0.4136,
+      "step": 16134
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 36.98055199871322,
+      "learning_rate": 1.3113588320670967e-06,
+      "loss": 0.3608,
+      "step": 16135
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 46.119993956467034,
+      "learning_rate": 1.3105241027106187e-06,
+      "loss": 0.4693,
+      "step": 16136
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 45.43959798783626,
+      "learning_rate": 1.3096896204777532e-06,
+      "loss": 0.372,
+      "step": 16137
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.26396436564129,
+      "learning_rate": 1.3088553853922325e-06,
+      "loss": 0.4022,
+      "step": 16138
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 39.98578366730779,
+      "learning_rate": 1.3080213974777766e-06,
+      "loss": 0.4278,
+      "step": 16139
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.33062549188411,
+      "learning_rate": 1.3071876567581087e-06,
+      "loss": 0.4728,
+      "step": 16140
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 38.33224908636028,
+      "learning_rate": 1.306354163256941e-06,
+      "loss": 0.3925,
+      "step": 16141
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 36.856780600840175,
+      "learning_rate": 1.3055209169979754e-06,
+      "loss": 0.342,
+      "step": 16142
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 56.66720480195877,
+      "learning_rate": 1.3046879180049054e-06,
+      "loss": 0.5541,
+      "step": 16143
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.23859998292929,
+      "learning_rate": 1.3038551663014255e-06,
+      "loss": 0.386,
+      "step": 16144
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 45.31778513636597,
+      "learning_rate": 1.3030226619112175e-06,
+      "loss": 0.4829,
+      "step": 16145
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.76676782320913,
+      "learning_rate": 1.3021904048579592e-06,
+      "loss": 0.4815,
+      "step": 16146
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 50.21332024570733,
+      "learning_rate": 1.3013583951653185e-06,
+      "loss": 0.4348,
+      "step": 16147
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 51.92531231073191,
+      "learning_rate": 1.3005266328569533e-06,
+      "loss": 0.4633,
+      "step": 16148
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 44.157122080318,
+      "learning_rate": 1.2996951179565221e-06,
+      "loss": 0.4313,
+      "step": 16149
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 50.43527555701096,
+      "learning_rate": 1.2988638504876738e-06,
+      "loss": 0.477,
+      "step": 16150
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 49.00260746596647,
+      "learning_rate": 1.2980328304740464e-06,
+      "loss": 0.4168,
+      "step": 16151
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.62138649586141,
+      "learning_rate": 1.2972020579392763e-06,
+      "loss": 0.432,
+      "step": 16152
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 40.06161196910239,
+      "learning_rate": 1.296371532906987e-06,
+      "loss": 0.3801,
+      "step": 16153
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 46.85083822988512,
+      "learning_rate": 1.2955412554008018e-06,
+      "loss": 0.4002,
+      "step": 16154
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 44.5753821415976,
+      "learning_rate": 1.2947112254443283e-06,
+      "loss": 0.4074,
+      "step": 16155
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 47.3771872002974,
+      "learning_rate": 1.2938814430611756e-06,
+      "loss": 0.4527,
+      "step": 16156
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 44.705371692231395,
+      "learning_rate": 1.2930519082749448e-06,
+      "loss": 0.4563,
+      "step": 16157
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.08283223178007,
+      "learning_rate": 1.2922226211092237e-06,
+      "loss": 0.3323,
+      "step": 16158
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 46.60175629005524,
+      "learning_rate": 1.2913935815875945e-06,
+      "loss": 0.4506,
+      "step": 16159
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 42.11252162974689,
+      "learning_rate": 1.2905647897336382e-06,
+      "loss": 0.3719,
+      "step": 16160
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.0111172133151,
+      "learning_rate": 1.289736245570925e-06,
+      "loss": 0.4029,
+      "step": 16161
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 44.49558208801988,
+      "learning_rate": 1.2889079491230184e-06,
+      "loss": 0.4157,
+      "step": 16162
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 52.55803421499501,
+      "learning_rate": 1.2880799004134748e-06,
+      "loss": 0.4638,
+      "step": 16163
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 48.23173831100799,
+      "learning_rate": 1.2872520994658399e-06,
+      "loss": 0.4166,
+      "step": 16164
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 56.79598212504473,
+      "learning_rate": 1.2864245463036585e-06,
+      "loss": 0.4751,
+      "step": 16165
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 52.45336841280187,
+      "learning_rate": 1.285597240950468e-06,
+      "loss": 0.4356,
+      "step": 16166
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 42.81274138890734,
+      "learning_rate": 1.2847701834297944e-06,
+      "loss": 0.4192,
+      "step": 16167
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 60.283928795319454,
+      "learning_rate": 1.2839433737651564e-06,
+      "loss": 0.4728,
+      "step": 16168
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 40.93377270801856,
+      "learning_rate": 1.2831168119800696e-06,
+      "loss": 0.4124,
+      "step": 16169
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.23863750539048,
+      "learning_rate": 1.2822904980980422e-06,
+      "loss": 0.4971,
+      "step": 16170
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 43.47699995480588,
+      "learning_rate": 1.2814644321425752e-06,
+      "loss": 0.4341,
+      "step": 16171
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 50.62997610808503,
+      "learning_rate": 1.2806386141371596e-06,
+      "loss": 0.4462,
+      "step": 16172
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.84441181425274,
+      "learning_rate": 1.279813044105278e-06,
+      "loss": 0.4302,
+      "step": 16173
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 55.086176792610225,
+      "learning_rate": 1.2789877220704127e-06,
+      "loss": 0.472,
+      "step": 16174
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 54.99137363408754,
+      "learning_rate": 1.2781626480560384e-06,
+      "loss": 0.4548,
+      "step": 16175
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 46.52616043898413,
+      "learning_rate": 1.2773378220856126e-06,
+      "loss": 0.5273,
+      "step": 16176
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.5098915434824,
+      "learning_rate": 1.2765132441825989e-06,
+      "loss": 0.4246,
+      "step": 16177
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.03174172584773,
+      "learning_rate": 1.2756889143704443e-06,
+      "loss": 0.4603,
+      "step": 16178
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 46.47828203548078,
+      "learning_rate": 1.2748648326725943e-06,
+      "loss": 0.4031,
+      "step": 16179
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 46.55077980568396,
+      "learning_rate": 1.2740409991124826e-06,
+      "loss": 0.4635,
+      "step": 16180
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 45.13566128438302,
+      "learning_rate": 1.273217413713541e-06,
+      "loss": 0.4333,
+      "step": 16181
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 40.867391018077235,
+      "learning_rate": 1.2723940764991927e-06,
+      "loss": 0.4587,
+      "step": 16182
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 55.681036551585905,
+      "learning_rate": 1.2715709874928518e-06,
+      "loss": 0.4716,
+      "step": 16183
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 50.356185229741435,
+      "learning_rate": 1.2707481467179228e-06,
+      "loss": 0.4925,
+      "step": 16184
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.52740028034986,
+      "learning_rate": 1.2699255541978117e-06,
+      "loss": 0.4364,
+      "step": 16185
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.8946942409509,
+      "learning_rate": 1.2691032099559097e-06,
+      "loss": 0.453,
+      "step": 16186
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 52.107662006915696,
+      "learning_rate": 1.268281114015607e-06,
+      "loss": 0.4448,
+      "step": 16187
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 53.109264605980634,
+      "learning_rate": 1.2674592664002828e-06,
+      "loss": 0.4242,
+      "step": 16188
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 45.97457042906198,
+      "learning_rate": 1.2666376671333048e-06,
+      "loss": 0.473,
+      "step": 16189
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.514574235480715,
+      "learning_rate": 1.2658163162380443e-06,
+      "loss": 0.4494,
+      "step": 16190
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 39.725190865722524,
+      "learning_rate": 1.2649952137378596e-06,
+      "loss": 0.4163,
+      "step": 16191
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 43.77713944213933,
+      "learning_rate": 1.2641743596561007e-06,
+      "loss": 0.4205,
+      "step": 16192
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 58.964234713441314,
+      "learning_rate": 1.2633537540161123e-06,
+      "loss": 0.5327,
+      "step": 16193
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 62.2461550455836,
+      "learning_rate": 1.2625333968412312e-06,
+      "loss": 0.5105,
+      "step": 16194
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 36.7559780137381,
+      "learning_rate": 1.261713288154789e-06,
+      "loss": 0.3482,
+      "step": 16195
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 40.60260394817599,
+      "learning_rate": 1.2608934279801133e-06,
+      "loss": 0.3608,
+      "step": 16196
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 52.54262518960087,
+      "learning_rate": 1.2600738163405124e-06,
+      "loss": 0.515,
+      "step": 16197
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 39.736657734476765,
+      "learning_rate": 1.2592544532592988e-06,
+      "loss": 0.3796,
+      "step": 16198
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 41.793987429688734,
+      "learning_rate": 1.2584353387597758e-06,
+      "loss": 0.4104,
+      "step": 16199
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 52.383079146944674,
+      "learning_rate": 1.2576164728652406e-06,
+      "loss": 0.51,
+      "step": 16200
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 38.69331688358931,
+      "learning_rate": 1.256797855598978e-06,
+      "loss": 0.3432,
+      "step": 16201
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 45.951986361667615,
+      "learning_rate": 1.255979486984269e-06,
+      "loss": 0.4675,
+      "step": 16202
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 43.89714579558094,
+      "learning_rate": 1.2551613670443862e-06,
+      "loss": 0.3644,
+      "step": 16203
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 40.495625830723576,
+      "learning_rate": 1.2543434958026024e-06,
+      "loss": 0.4511,
+      "step": 16204
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.948717665898705,
+      "learning_rate": 1.253525873282171e-06,
+      "loss": 0.4391,
+      "step": 16205
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 44.77462699080877,
+      "learning_rate": 1.2527084995063488e-06,
+      "loss": 0.3803,
+      "step": 16206
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 49.010359642111844,
+      "learning_rate": 1.2518913744983786e-06,
+      "loss": 0.3931,
+      "step": 16207
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 40.07921225576957,
+      "learning_rate": 1.2510744982814993e-06,
+      "loss": 0.4237,
+      "step": 16208
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 38.72978564577105,
+      "learning_rate": 1.250257870878946e-06,
+      "loss": 0.4273,
+      "step": 16209
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.139973222160904,
+      "learning_rate": 1.2494414923139375e-06,
+      "loss": 0.4065,
+      "step": 16210
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 59.9294312210183,
+      "learning_rate": 1.2486253626096978e-06,
+      "loss": 0.5137,
+      "step": 16211
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.75367506973694,
+      "learning_rate": 1.2478094817894293e-06,
+      "loss": 0.4857,
+      "step": 16212
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 46.26664509110859,
+      "learning_rate": 1.2469938498763423e-06,
+      "loss": 0.3891,
+      "step": 16213
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 39.69808869769618,
+      "learning_rate": 1.2461784668936283e-06,
+      "loss": 0.3891,
+      "step": 16214
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.25823528789471,
+      "learning_rate": 1.2453633328644765e-06,
+      "loss": 0.4655,
+      "step": 16215
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 33.707529493790226,
+      "learning_rate": 1.244548447812074e-06,
+      "loss": 0.372,
+      "step": 16216
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 50.28279351804576,
+      "learning_rate": 1.243733811759591e-06,
+      "loss": 0.4766,
+      "step": 16217
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 40.37501462103517,
+      "learning_rate": 1.2429194247301934e-06,
+      "loss": 0.4168,
+      "step": 16218
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.34570925847762,
+      "learning_rate": 1.2421052867470452e-06,
+      "loss": 0.4421,
+      "step": 16219
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 59.39367708484502,
+      "learning_rate": 1.2412913978332997e-06,
+      "loss": 0.5203,
+      "step": 16220
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 43.01522192885683,
+      "learning_rate": 1.2404777580121075e-06,
+      "loss": 0.4986,
+      "step": 16221
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 42.28851109400147,
+      "learning_rate": 1.2396643673065988e-06,
+      "loss": 0.508,
+      "step": 16222
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 47.51717076394888,
+      "learning_rate": 1.238851225739911e-06,
+      "loss": 0.4657,
+      "step": 16223
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 44.17938176191803,
+      "learning_rate": 1.2380383333351687e-06,
+      "loss": 0.4013,
+      "step": 16224
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 45.54200813896036,
+      "learning_rate": 1.2372256901154934e-06,
+      "loss": 0.4721,
+      "step": 16225
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 38.926245108298765,
+      "learning_rate": 1.2364132961039933e-06,
+      "loss": 0.407,
+      "step": 16226
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 49.92985244554216,
+      "learning_rate": 1.2356011513237708e-06,
+      "loss": 0.4379,
+      "step": 16227
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 48.45399621440431,
+      "learning_rate": 1.2347892557979236e-06,
+      "loss": 0.4356,
+      "step": 16228
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 63.068272041987335,
+      "learning_rate": 1.233977609549546e-06,
+      "loss": 0.5217,
+      "step": 16229
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 52.481758104676544,
+      "learning_rate": 1.2331662126017141e-06,
+      "loss": 0.4647,
+      "step": 16230
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 42.72671186233057,
+      "learning_rate": 1.2323550649775085e-06,
+      "loss": 0.3956,
+      "step": 16231
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 67.90627511063948,
+      "learning_rate": 1.2315441666999939e-06,
+      "loss": 0.4388,
+      "step": 16232
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 57.021451155307744,
+      "learning_rate": 1.2307335177922342e-06,
+      "loss": 0.4635,
+      "step": 16233
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 46.49392415929544,
+      "learning_rate": 1.2299231182772852e-06,
+      "loss": 0.45,
+      "step": 16234
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 39.841623328804445,
+      "learning_rate": 1.229112968178191e-06,
+      "loss": 0.4408,
+      "step": 16235
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 42.09125890915072,
+      "learning_rate": 1.2283030675179951e-06,
+      "loss": 0.3919,
+      "step": 16236
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 59.04911649739111,
+      "learning_rate": 1.227493416319726e-06,
+      "loss": 0.4911,
+      "step": 16237
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 51.82057977311572,
+      "learning_rate": 1.226684014606414e-06,
+      "loss": 0.5021,
+      "step": 16238
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 49.056447937529974,
+      "learning_rate": 1.2258748624010752e-06,
+      "loss": 0.462,
+      "step": 16239
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 55.31925938140837,
+      "learning_rate": 1.2250659597267244e-06,
+      "loss": 0.4733,
+      "step": 16240
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 36.65370579806674,
+      "learning_rate": 1.2242573066063623e-06,
+      "loss": 0.3032,
+      "step": 16241
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 53.41074940325517,
+      "learning_rate": 1.2234489030629916e-06,
+      "loss": 0.5069,
+      "step": 16242
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 49.90158021943706,
+      "learning_rate": 1.2226407491195969e-06,
+      "loss": 0.3779,
+      "step": 16243
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 50.909513435646204,
+      "learning_rate": 1.2218328447991657e-06,
+      "loss": 0.4058,
+      "step": 16244
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 44.33319327444645,
+      "learning_rate": 1.2210251901246739e-06,
+      "loss": 0.4444,
+      "step": 16245
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 57.72005966344698,
+      "learning_rate": 1.2202177851190912e-06,
+      "loss": 0.5052,
+      "step": 16246
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 47.25971183674474,
+      "learning_rate": 1.2194106298053765e-06,
+      "loss": 0.4369,
+      "step": 16247
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 45.50255695431462,
+      "learning_rate": 1.2186037242064863e-06,
+      "loss": 0.4205,
+      "step": 16248
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 44.5859024424705,
+      "learning_rate": 1.21779706834537e-06,
+      "loss": 0.4231,
+      "step": 16249
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 48.146659589230985,
+      "learning_rate": 1.2169906622449701e-06,
+      "loss": 0.4734,
+      "step": 16250
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 45.18051072831263,
+      "learning_rate": 1.2161845059282174e-06,
+      "loss": 0.4235,
+      "step": 16251
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 56.329641897911024,
+      "learning_rate": 1.2153785994180366e-06,
+      "loss": 0.4209,
+      "step": 16252
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 40.37053084366289,
+      "learning_rate": 1.2145729427373499e-06,
+      "loss": 0.3704,
+      "step": 16253
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 44.72608600212712,
+      "learning_rate": 1.2137675359090705e-06,
+      "loss": 0.3942,
+      "step": 16254
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 48.782481635586834,
+      "learning_rate": 1.212962378956104e-06,
+      "loss": 0.4713,
+      "step": 16255
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 37.71959520355363,
+      "learning_rate": 1.212157471901344e-06,
+      "loss": 0.3954,
+      "step": 16256
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 51.64042755412389,
+      "learning_rate": 1.2113528147676855e-06,
+      "loss": 0.4553,
+      "step": 16257
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 41.22602314439898,
+      "learning_rate": 1.2105484075780117e-06,
+      "loss": 0.3836,
+      "step": 16258
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 49.10427128884919,
+      "learning_rate": 1.2097442503552003e-06,
+      "loss": 0.4544,
+      "step": 16259
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 50.304767824817695,
+      "learning_rate": 1.2089403431221213e-06,
+      "loss": 0.4624,
+      "step": 16260
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 47.6840516289738,
+      "learning_rate": 1.2081366859016353e-06,
+      "loss": 0.4197,
+      "step": 16261
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 50.21617853860321,
+      "learning_rate": 1.207333278716597e-06,
+      "loss": 0.4565,
+      "step": 16262
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 44.541912202482465,
+      "learning_rate": 1.2065301215898595e-06,
+      "loss": 0.3812,
+      "step": 16263
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 52.773999580648436,
+      "learning_rate": 1.20572721454426e-06,
+      "loss": 0.4165,
+      "step": 16264
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 47.42855288700933,
+      "learning_rate": 1.2049245576026346e-06,
+      "loss": 0.3988,
+      "step": 16265
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 41.28400976758534,
+      "learning_rate": 1.2041221507878087e-06,
+      "loss": 0.4304,
+      "step": 16266
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 46.60738927925331,
+      "learning_rate": 1.203319994122606e-06,
+      "loss": 0.4013,
+      "step": 16267
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 43.440891720750365,
+      "learning_rate": 1.2025180876298338e-06,
+      "loss": 0.4155,
+      "step": 16268
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 51.95650113027988,
+      "learning_rate": 1.2017164313323004e-06,
+      "loss": 0.5236,
+      "step": 16269
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 42.61857086090287,
+      "learning_rate": 1.200915025252808e-06,
+      "loss": 0.4069,
+      "step": 16270
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 57.98170898032143,
+      "learning_rate": 1.2001138694141423e-06,
+      "loss": 0.4435,
+      "step": 16271
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 42.07669673497233,
+      "learning_rate": 1.199312963839092e-06,
+      "loss": 0.444,
+      "step": 16272
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 37.7826730414402,
+      "learning_rate": 1.1985123085504312e-06,
+      "loss": 0.3646,
+      "step": 16273
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 45.79700057898503,
+      "learning_rate": 1.1977119035709329e-06,
+      "loss": 0.4346,
+      "step": 16274
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 48.8049628804295,
+      "learning_rate": 1.1969117489233595e-06,
+      "loss": 0.4028,
+      "step": 16275
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 50.86765150943322,
+      "learning_rate": 1.1961118446304666e-06,
+      "loss": 0.4204,
+      "step": 16276
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 45.158996023349026,
+      "learning_rate": 1.195312190715e-06,
+      "loss": 0.3971,
+      "step": 16277
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 42.49441860799034,
+      "learning_rate": 1.1945127871997042e-06,
+      "loss": 0.3903,
+      "step": 16278
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 38.34029750001629,
+      "learning_rate": 1.1937136341073163e-06,
+      "loss": 0.4276,
+      "step": 16279
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 46.19738975433716,
+      "learning_rate": 1.1929147314605617e-06,
+      "loss": 0.4137,
+      "step": 16280
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 46.82415677282359,
+      "learning_rate": 1.1921160792821572e-06,
+      "loss": 0.4552,
+      "step": 16281
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 44.94541343652434,
+      "learning_rate": 1.1913176775948187e-06,
+      "loss": 0.4996,
+      "step": 16282
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 39.309572975102306,
+      "learning_rate": 1.1905195264212532e-06,
+      "loss": 0.3777,
+      "step": 16283
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 42.9494321976545,
+      "learning_rate": 1.1897216257841605e-06,
+      "loss": 0.4234,
+      "step": 16284
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 54.91044631928067,
+      "learning_rate": 1.1889239757062309e-06,
+      "loss": 0.4764,
+      "step": 16285
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 49.24202010089806,
+      "learning_rate": 1.188126576210148e-06,
+      "loss": 0.4243,
+      "step": 16286
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 33.5466333175361,
+      "learning_rate": 1.1873294273185898e-06,
+      "loss": 0.3544,
+      "step": 16287
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 50.558684451471244,
+      "learning_rate": 1.1865325290542295e-06,
+      "loss": 0.4327,
+      "step": 16288
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 49.38374915529951,
+      "learning_rate": 1.185735881439728e-06,
+      "loss": 0.4321,
+      "step": 16289
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 47.699984349941985,
+      "learning_rate": 1.1849394844977402e-06,
+      "loss": 0.3838,
+      "step": 16290
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 57.03031802735408,
+      "learning_rate": 1.184143338250917e-06,
+      "loss": 0.5628,
+      "step": 16291
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 61.55830888762745,
+      "learning_rate": 1.1833474427219015e-06,
+      "loss": 0.4537,
+      "step": 16292
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 48.33715314679012,
+      "learning_rate": 1.1825517979333256e-06,
+      "loss": 0.4446,
+      "step": 16293
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 52.7662385018706,
+      "learning_rate": 1.18175640390782e-06,
+      "loss": 0.493,
+      "step": 16294
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 47.30103368153884,
+      "learning_rate": 1.180961260668002e-06,
+      "loss": 0.4025,
+      "step": 16295
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 45.366779625768594,
+      "learning_rate": 1.1801663682364873e-06,
+      "loss": 0.4045,
+      "step": 16296
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 45.45806039692696,
+      "learning_rate": 1.179371726635883e-06,
+      "loss": 0.4113,
+      "step": 16297
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.952509459702156,
+      "learning_rate": 1.1785773358887854e-06,
+      "loss": 0.4183,
+      "step": 16298
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 46.290030543510305,
+      "learning_rate": 1.1777831960177898e-06,
+      "loss": 0.406,
+      "step": 16299
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 46.40898551199535,
+      "learning_rate": 1.1769893070454774e-06,
+      "loss": 0.4258,
+      "step": 16300
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 64.83728943042823,
+      "learning_rate": 1.1761956689944288e-06,
+      "loss": 0.4523,
+      "step": 16301
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 54.33249571901537,
+      "learning_rate": 1.1754022818872123e-06,
+      "loss": 0.4359,
+      "step": 16302
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 49.408252103655876,
+      "learning_rate": 1.1746091457463927e-06,
+      "loss": 0.429,
+      "step": 16303
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 47.89268345850605,
+      "learning_rate": 1.173816260594529e-06,
+      "loss": 0.4421,
+      "step": 16304
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.048435212003746,
+      "learning_rate": 1.1730236264541661e-06,
+      "loss": 0.4507,
+      "step": 16305
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 59.34001296953857,
+      "learning_rate": 1.1722312433478467e-06,
+      "loss": 0.4761,
+      "step": 16306
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 48.33529817051389,
+      "learning_rate": 1.1714391112981071e-06,
+      "loss": 0.4178,
+      "step": 16307
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 38.57640876212506,
+      "learning_rate": 1.170647230327473e-06,
+      "loss": 0.3522,
+      "step": 16308
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 53.013816728454806,
+      "learning_rate": 1.1698556004584728e-06,
+      "loss": 0.4654,
+      "step": 16309
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 35.20482720104427,
+      "learning_rate": 1.1690642217136084e-06,
+      "loss": 0.3623,
+      "step": 16310
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 49.56927146690236,
+      "learning_rate": 1.1682730941153918e-06,
+      "loss": 0.4723,
+      "step": 16311
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 47.06868522901971,
+      "learning_rate": 1.167482217686322e-06,
+      "loss": 0.4444,
+      "step": 16312
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 63.07157533218449,
+      "learning_rate": 1.1666915924488931e-06,
+      "loss": 0.483,
+      "step": 16313
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 55.799908164847764,
+      "learning_rate": 1.165901218425588e-06,
+      "loss": 0.4673,
+      "step": 16314
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 34.44193877191042,
+      "learning_rate": 1.1651110956388822e-06,
+      "loss": 0.3007,
+      "step": 16315
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 46.40462079511706,
+      "learning_rate": 1.164321224111249e-06,
+      "loss": 0.4288,
+      "step": 16316
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 43.23612175967308,
+      "learning_rate": 1.1635316038651524e-06,
+      "loss": 0.3998,
+      "step": 16317
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 39.00970608412523,
+      "learning_rate": 1.1627422349230465e-06,
+      "loss": 0.3765,
+      "step": 16318
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 46.15038780171387,
+      "learning_rate": 1.161953117307385e-06,
+      "loss": 0.4006,
+      "step": 16319
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 44.21954759613079,
+      "learning_rate": 1.161164251040603e-06,
+      "loss": 0.4586,
+      "step": 16320
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 45.289629290582425,
+      "learning_rate": 1.1603756361451402e-06,
+      "loss": 0.4469,
+      "step": 16321
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 45.23259283320345,
+      "learning_rate": 1.1595872726434243e-06,
+      "loss": 0.4225,
+      "step": 16322
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 38.36623720216766,
+      "learning_rate": 1.158799160557874e-06,
+      "loss": 0.3583,
+      "step": 16323
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 43.171741536472936,
+      "learning_rate": 1.158011299910905e-06,
+      "loss": 0.3715,
+      "step": 16324
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 52.16146125609348,
+      "learning_rate": 1.15722369072492e-06,
+      "loss": 0.4711,
+      "step": 16325
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 32.75386816214779,
+      "learning_rate": 1.1564363330223227e-06,
+      "loss": 0.3537,
+      "step": 16326
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.15423377258162,
+      "learning_rate": 1.1556492268255004e-06,
+      "loss": 0.4252,
+      "step": 16327
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 55.87395570193013,
+      "learning_rate": 1.1548623721568409e-06,
+      "loss": 0.482,
+      "step": 16328
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 52.65630425243808,
+      "learning_rate": 1.1540757690387227e-06,
+      "loss": 0.4688,
+      "step": 16329
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 48.238299853634615,
+      "learning_rate": 1.153289417493515e-06,
+      "loss": 0.3986,
+      "step": 16330
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 49.68021742035256,
+      "learning_rate": 1.1525033175435796e-06,
+      "loss": 0.4398,
+      "step": 16331
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 39.49600691168475,
+      "learning_rate": 1.1517174692112742e-06,
+      "loss": 0.3976,
+      "step": 16332
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 47.151356002628994,
+      "learning_rate": 1.1509318725189477e-06,
+      "loss": 0.4951,
+      "step": 16333
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 43.446055194150084,
+      "learning_rate": 1.1501465274889457e-06,
+      "loss": 0.456,
+      "step": 16334
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 49.90565164788982,
+      "learning_rate": 1.1493614341435954e-06,
+      "loss": 0.3794,
+      "step": 16335
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 61.52533479217373,
+      "learning_rate": 1.1485765925052294e-06,
+      "loss": 0.4573,
+      "step": 16336
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 44.09702067204803,
+      "learning_rate": 1.1477920025961664e-06,
+      "loss": 0.3802,
+      "step": 16337
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 52.245433183433626,
+      "learning_rate": 1.1470076644387229e-06,
+      "loss": 0.4414,
+      "step": 16338
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 58.19663589248981,
+      "learning_rate": 1.1462235780552023e-06,
+      "loss": 0.4306,
+      "step": 16339
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.26829676054436,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.3951,
+      "step": 16340
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 47.133643336650245,
+      "learning_rate": 1.1446561606991158e-06,
+      "loss": 0.4404,
+      "step": 16341
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 53.43834124343544,
+      "learning_rate": 1.1438728297711288e-06,
+      "loss": 0.4698,
+      "step": 16342
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 39.02165882536146,
+      "learning_rate": 1.14308975070622e-06,
+      "loss": 0.3966,
+      "step": 16343
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 40.1635600965532,
+      "learning_rate": 1.1423069235266538e-06,
+      "loss": 0.3392,
+      "step": 16344
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.37578653207646,
+      "learning_rate": 1.1415243482546977e-06,
+      "loss": 0.4531,
+      "step": 16345
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 45.46836624770101,
+      "learning_rate": 1.1407420249126068e-06,
+      "loss": 0.4521,
+      "step": 16346
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 39.73091052181713,
+      "learning_rate": 1.1399599535226324e-06,
+      "loss": 0.405,
+      "step": 16347
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 39.449684953942494,
+      "learning_rate": 1.139178134107014e-06,
+      "loss": 0.3899,
+      "step": 16348
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 50.74292539510807,
+      "learning_rate": 1.1383965666879847e-06,
+      "loss": 0.436,
+      "step": 16349
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 51.406835781110786,
+      "learning_rate": 1.1376152512877725e-06,
+      "loss": 0.3917,
+      "step": 16350
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 41.0202351912866,
+      "learning_rate": 1.1368341879286004e-06,
+      "loss": 0.4058,
+      "step": 16351
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 39.17468484144388,
+      "learning_rate": 1.1360533766326765e-06,
+      "loss": 0.4022,
+      "step": 16352
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 54.36987880399301,
+      "learning_rate": 1.1352728174222128e-06,
+      "loss": 0.4289,
+      "step": 16353
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 42.989904715678,
+      "learning_rate": 1.1344925103194005e-06,
+      "loss": 0.4431,
+      "step": 16354
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 50.397127474184785,
+      "learning_rate": 1.1337124553464384e-06,
+      "loss": 0.4616,
+      "step": 16355
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 50.27811995413852,
+      "learning_rate": 1.1329326525255046e-06,
+      "loss": 0.4074,
+      "step": 16356
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 54.96784839682283,
+      "learning_rate": 1.1321531018787801e-06,
+      "loss": 0.484,
+      "step": 16357
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 58.58002071411506,
+      "learning_rate": 1.131373803428435e-06,
+      "loss": 0.5268,
+      "step": 16358
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 34.899200868338326,
+      "learning_rate": 1.1305947571966291e-06,
+      "loss": 0.3401,
+      "step": 16359
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 52.69438407425055,
+      "learning_rate": 1.1298159632055228e-06,
+      "loss": 0.3915,
+      "step": 16360
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 52.75831978620541,
+      "learning_rate": 1.1290374214772582e-06,
+      "loss": 0.4193,
+      "step": 16361
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 46.46250089533007,
+      "learning_rate": 1.1282591320339809e-06,
+      "loss": 0.4222,
+      "step": 16362
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 43.70668929032149,
+      "learning_rate": 1.1274810948978255e-06,
+      "loss": 0.4261,
+      "step": 16363
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 46.42276335501492,
+      "learning_rate": 1.1267033100909174e-06,
+      "loss": 0.4126,
+      "step": 16364
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.34432341980602,
+      "learning_rate": 1.125925777635375e-06,
+      "loss": 0.4341,
+      "step": 16365
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 42.20946228849774,
+      "learning_rate": 1.1251484975533123e-06,
+      "loss": 0.3878,
+      "step": 16366
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 46.79343685199747,
+      "learning_rate": 1.1243714698668363e-06,
+      "loss": 0.4096,
+      "step": 16367
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 33.96298952223726,
+      "learning_rate": 1.1235946945980435e-06,
+      "loss": 0.303,
+      "step": 16368
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 48.81008389851636,
+      "learning_rate": 1.1228181717690234e-06,
+      "loss": 0.4245,
+      "step": 16369
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 51.5692594039551,
+      "learning_rate": 1.1220419014018613e-06,
+      "loss": 0.4896,
+      "step": 16370
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 41.378054094106176,
+      "learning_rate": 1.121265883518634e-06,
+      "loss": 0.406,
+      "step": 16371
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 44.65345827509954,
+      "learning_rate": 1.1204901181414141e-06,
+      "loss": 0.3843,
+      "step": 16372
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.92075789635578,
+      "learning_rate": 1.1197146052922592e-06,
+      "loss": 0.4454,
+      "step": 16373
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 48.46627623235934,
+      "learning_rate": 1.118939344993225e-06,
+      "loss": 0.4111,
+      "step": 16374
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 54.80223279636664,
+      "learning_rate": 1.1181643372663608e-06,
+      "loss": 0.4927,
+      "step": 16375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 46.57717010020856,
+      "learning_rate": 1.1173895821337088e-06,
+      "loss": 0.428,
+      "step": 16376
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 59.22017906296522,
+      "learning_rate": 1.1166150796172981e-06,
+      "loss": 0.4588,
+      "step": 16377
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.3889767202617,
+      "learning_rate": 1.115840829739161e-06,
+      "loss": 0.4667,
+      "step": 16378
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 42.727044009873005,
+      "learning_rate": 1.115066832521311e-06,
+      "loss": 0.4168,
+      "step": 16379
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.598816870563255,
+      "learning_rate": 1.114293087985766e-06,
+      "loss": 0.4032,
+      "step": 16380
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 48.04471525967173,
+      "learning_rate": 1.1135195961545242e-06,
+      "loss": 0.4745,
+      "step": 16381
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 53.83873458413932,
+      "learning_rate": 1.1127463570495867e-06,
+      "loss": 0.4885,
+      "step": 16382
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 48.06456195093986,
+      "learning_rate": 1.111973370692947e-06,
+      "loss": 0.4759,
+      "step": 16383
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 44.47422260379093,
+      "learning_rate": 1.111200637106582e-06,
+      "loss": 0.4188,
+      "step": 16384
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 42.1599150125626,
+      "learning_rate": 1.1104281563124741e-06,
+      "loss": 0.3335,
+      "step": 16385
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 54.06134379479074,
+      "learning_rate": 1.109655928332587e-06,
+      "loss": 0.4861,
+      "step": 16386
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 40.38497216458322,
+      "learning_rate": 1.1088839531888862e-06,
+      "loss": 0.3957,
+      "step": 16387
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 50.57948676141924,
+      "learning_rate": 1.108112230903322e-06,
+      "loss": 0.4526,
+      "step": 16388
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 43.42883596942485,
+      "learning_rate": 1.1073407614978471e-06,
+      "loss": 0.3975,
+      "step": 16389
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 65.24261238033797,
+      "learning_rate": 1.106569544994397e-06,
+      "loss": 0.453,
+      "step": 16390
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 43.72010308681072,
+      "learning_rate": 1.1057985814149063e-06,
+      "loss": 0.4487,
+      "step": 16391
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 43.55366506975915,
+      "learning_rate": 1.1050278707813033e-06,
+      "loss": 0.4008,
+      "step": 16392
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 55.901612159260615,
+      "learning_rate": 1.1042574131155048e-06,
+      "loss": 0.4661,
+      "step": 16393
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.35235199288218,
+      "learning_rate": 1.1034872084394187e-06,
+      "loss": 0.4514,
+      "step": 16394
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 51.612674724295125,
+      "learning_rate": 1.1027172567749523e-06,
+      "loss": 0.4855,
+      "step": 16395
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 54.15291218011107,
+      "learning_rate": 1.101947558144002e-06,
+      "loss": 0.5321,
+      "step": 16396
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 39.374576905414585,
+      "learning_rate": 1.1011781125684618e-06,
+      "loss": 0.4062,
+      "step": 16397
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.484203241658356,
+      "learning_rate": 1.1004089200702072e-06,
+      "loss": 0.405,
+      "step": 16398
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 36.85502320683884,
+      "learning_rate": 1.0996399806711167e-06,
+      "loss": 0.3772,
+      "step": 16399
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 44.43177759638309,
+      "learning_rate": 1.098871294393058e-06,
+      "loss": 0.4052,
+      "step": 16400
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 45.085708852904375,
+      "learning_rate": 1.0981028612578949e-06,
+      "loss": 0.3978,
+      "step": 16401
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 44.333502108866924,
+      "learning_rate": 1.0973346812874796e-06,
+      "loss": 0.444,
+      "step": 16402
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 42.96911224212543,
+      "learning_rate": 1.0965667545036552e-06,
+      "loss": 0.3837,
+      "step": 16403
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 44.1597065527727,
+      "learning_rate": 1.0957990809282649e-06,
+      "loss": 0.4244,
+      "step": 16404
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 38.67225256533251,
+      "learning_rate": 1.0950316605831413e-06,
+      "loss": 0.3457,
+      "step": 16405
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 51.06334200460174,
+      "learning_rate": 1.0942644934901059e-06,
+      "loss": 0.5821,
+      "step": 16406
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 44.57652795433207,
+      "learning_rate": 1.0934975796709801e-06,
+      "loss": 0.4336,
+      "step": 16407
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.15602200492817,
+      "learning_rate": 1.0927309191475722e-06,
+      "loss": 0.4389,
+      "step": 16408
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 51.92867245348539,
+      "learning_rate": 1.0919645119416855e-06,
+      "loss": 0.4059,
+      "step": 16409
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 48.04911114459041,
+      "learning_rate": 1.0911983580751195e-06,
+      "loss": 0.4484,
+      "step": 16410
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 58.41359262862897,
+      "learning_rate": 1.090432457569659e-06,
+      "loss": 0.458,
+      "step": 16411
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 50.50863179001865,
+      "learning_rate": 1.0896668104470886e-06,
+      "loss": 0.4582,
+      "step": 16412
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 53.81515176067806,
+      "learning_rate": 1.08890141672918e-06,
+      "loss": 0.4227,
+      "step": 16413
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 45.58647401521756,
+      "learning_rate": 1.0881362764377046e-06,
+      "loss": 0.4463,
+      "step": 16414
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 34.130241421193,
+      "learning_rate": 1.087371389594417e-06,
+      "loss": 0.3695,
+      "step": 16415
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 45.52892912825441,
+      "learning_rate": 1.0866067562210748e-06,
+      "loss": 0.4104,
+      "step": 16416
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 48.74772051348111,
+      "learning_rate": 1.0858423763394243e-06,
+      "loss": 0.459,
+      "step": 16417
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 40.67263836607744,
+      "learning_rate": 1.085078249971201e-06,
+      "loss": 0.4286,
+      "step": 16418
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.143207058289136,
+      "learning_rate": 1.084314377138136e-06,
+      "loss": 0.4396,
+      "step": 16419
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.05695050919333,
+      "learning_rate": 1.0835507578619542e-06,
+      "loss": 0.4116,
+      "step": 16420
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 41.824469673456484,
+      "learning_rate": 1.0827873921643727e-06,
+      "loss": 0.4167,
+      "step": 16421
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 47.11284183885256,
+      "learning_rate": 1.0820242800671032e-06,
+      "loss": 0.3695,
+      "step": 16422
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 64.03568628524971,
+      "learning_rate": 1.0812614215918472e-06,
+      "loss": 0.5064,
+      "step": 16423
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 49.69325744572873,
+      "learning_rate": 1.080498816760296e-06,
+      "loss": 0.477,
+      "step": 16424
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 42.329474792700786,
+      "learning_rate": 1.0797364655941411e-06,
+      "loss": 0.4145,
+      "step": 16425
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 53.35032750916706,
+      "learning_rate": 1.0789743681150656e-06,
+      "loss": 0.4302,
+      "step": 16426
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 47.879055140294064,
+      "learning_rate": 1.0782125243447395e-06,
+      "loss": 0.4622,
+      "step": 16427
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 40.63393583470396,
+      "learning_rate": 1.077450934304829e-06,
+      "loss": 0.3527,
+      "step": 16428
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 48.80744880105766,
+      "learning_rate": 1.0766895980169933e-06,
+      "loss": 0.4703,
+      "step": 16429
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 54.247310488258336,
+      "learning_rate": 1.0759285155028887e-06,
+      "loss": 0.4711,
+      "step": 16430
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 43.17264463143883,
+      "learning_rate": 1.0751676867841553e-06,
+      "loss": 0.4127,
+      "step": 16431
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 58.364931132650874,
+      "learning_rate": 1.0744071118824306e-06,
+      "loss": 0.4242,
+      "step": 16432
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 41.944628580094815,
+      "learning_rate": 1.0736467908193471e-06,
+      "loss": 0.3388,
+      "step": 16433
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 40.054020863642684,
+      "learning_rate": 1.0728867236165264e-06,
+      "loss": 0.3886,
+      "step": 16434
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 42.40650232587168,
+      "learning_rate": 1.0721269102955866e-06,
+      "loss": 0.3852,
+      "step": 16435
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 48.05606693823839,
+      "learning_rate": 1.0713673508781353e-06,
+      "loss": 0.4738,
+      "step": 16436
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 43.11832611300283,
+      "learning_rate": 1.0706080453857714e-06,
+      "loss": 0.4367,
+      "step": 16437
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 53.353494830099024,
+      "learning_rate": 1.0698489938400914e-06,
+      "loss": 0.4143,
+      "step": 16438
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 54.24961449007099,
+      "learning_rate": 1.0690901962626843e-06,
+      "loss": 0.3947,
+      "step": 16439
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 50.55886513242912,
+      "learning_rate": 1.0683316526751253e-06,
+      "loss": 0.4578,
+      "step": 16440
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 43.36022476050938,
+      "learning_rate": 1.0675733630989904e-06,
+      "loss": 0.404,
+      "step": 16441
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 64.17340302302603,
+      "learning_rate": 1.0668153275558424e-06,
+      "loss": 0.4566,
+      "step": 16442
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 42.644398288531754,
+      "learning_rate": 1.066057546067243e-06,
+      "loss": 0.3839,
+      "step": 16443
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 49.978078118200145,
+      "learning_rate": 1.0653000186547379e-06,
+      "loss": 0.426,
+      "step": 16444
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 45.474571227683654,
+      "learning_rate": 1.0645427453398748e-06,
+      "loss": 0.4609,
+      "step": 16445
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 50.385722342050194,
+      "learning_rate": 1.0637857261441898e-06,
+      "loss": 0.4836,
+      "step": 16446
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 58.63554575812092,
+      "learning_rate": 1.0630289610892097e-06,
+      "loss": 0.4889,
+      "step": 16447
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 44.218198927825476,
+      "learning_rate": 1.06227245019646e-06,
+      "loss": 0.4129,
+      "step": 16448
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 41.19634801271043,
+      "learning_rate": 1.061516193487452e-06,
+      "loss": 0.4055,
+      "step": 16449
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 41.2309789325153,
+      "learning_rate": 1.060760190983694e-06,
+      "loss": 0.4024,
+      "step": 16450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 36.45424079086669,
+      "learning_rate": 1.060004442706688e-06,
+      "loss": 0.3704,
+      "step": 16451
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 40.92044632942344,
+      "learning_rate": 1.0592489486779267e-06,
+      "loss": 0.4059,
+      "step": 16452
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 54.10394659965843,
+      "learning_rate": 1.0584937089188941e-06,
+      "loss": 0.5179,
+      "step": 16453
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 40.45249553911196,
+      "learning_rate": 1.0577387234510684e-06,
+      "loss": 0.3925,
+      "step": 16454
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.95101814300337,
+      "learning_rate": 1.0569839922959247e-06,
+      "loss": 0.4149,
+      "step": 16455
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.19537110414822,
+      "learning_rate": 1.0562295154749248e-06,
+      "loss": 0.4316,
+      "step": 16456
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 44.77013043005244,
+      "learning_rate": 1.0554752930095236e-06,
+      "loss": 0.4769,
+      "step": 16457
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 44.94303500768903,
+      "learning_rate": 1.054721324921173e-06,
+      "loss": 0.4405,
+      "step": 16458
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 60.31363286352823,
+      "learning_rate": 1.0539676112313147e-06,
+      "loss": 0.5745,
+      "step": 16459
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 49.64943571970538,
+      "learning_rate": 1.053214151961386e-06,
+      "loss": 0.4385,
+      "step": 16460
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 49.67873124775366,
+      "learning_rate": 1.052460947132814e-06,
+      "loss": 0.3824,
+      "step": 16461
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 62.486329853435926,
+      "learning_rate": 1.0517079967670152e-06,
+      "loss": 0.4562,
+      "step": 16462
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 44.65546921086834,
+      "learning_rate": 1.0509553008854067e-06,
+      "loss": 0.4804,
+      "step": 16463
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.78787832248549,
+      "learning_rate": 1.0502028595093972e-06,
+      "loss": 0.477,
+      "step": 16464
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 49.08580123685738,
+      "learning_rate": 1.0494506726603804e-06,
+      "loss": 0.4793,
+      "step": 16465
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 38.78149598417877,
+      "learning_rate": 1.0486987403597526e-06,
+      "loss": 0.4519,
+      "step": 16466
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 48.56135488998655,
+      "learning_rate": 1.0479470626288946e-06,
+      "loss": 0.4561,
+      "step": 16467
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 58.130505627175914,
+      "learning_rate": 1.0471956394891868e-06,
+      "loss": 0.4576,
+      "step": 16468
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 46.78957735199725,
+      "learning_rate": 1.0464444709619959e-06,
+      "loss": 0.4522,
+      "step": 16469
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 49.587414635521554,
+      "learning_rate": 1.0456935570686866e-06,
+      "loss": 0.4129,
+      "step": 16470
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 53.83476362109714,
+      "learning_rate": 1.0449428978306164e-06,
+      "loss": 0.5102,
+      "step": 16471
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 38.4549888335844,
+      "learning_rate": 1.0441924932691293e-06,
+      "loss": 0.4116,
+      "step": 16472
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 50.906565707956545,
+      "learning_rate": 1.0434423434055719e-06,
+      "loss": 0.4794,
+      "step": 16473
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 45.439353280661784,
+      "learning_rate": 1.042692448261272e-06,
+      "loss": 0.3922,
+      "step": 16474
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 41.613840397141914,
+      "learning_rate": 1.041942807857559e-06,
+      "loss": 0.4064,
+      "step": 16475
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 53.097636425806286,
+      "learning_rate": 1.0411934222157538e-06,
+      "loss": 0.3843,
+      "step": 16476
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 50.46634573391385,
+      "learning_rate": 1.0404442913571678e-06,
+      "loss": 0.453,
+      "step": 16477
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 47.8163819152107,
+      "learning_rate": 1.0396954153031024e-06,
+      "loss": 0.4575,
+      "step": 16478
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 50.92899159709272,
+      "learning_rate": 1.0389467940748576e-06,
+      "loss": 0.4221,
+      "step": 16479
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 42.57798347099651,
+      "learning_rate": 1.0381984276937263e-06,
+      "loss": 0.3563,
+      "step": 16480
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 36.6928110843661,
+      "learning_rate": 1.037450316180989e-06,
+      "loss": 0.3419,
+      "step": 16481
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.018392537453714,
+      "learning_rate": 1.03670245955792e-06,
+      "loss": 0.4501,
+      "step": 16482
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 63.06290801627996,
+      "learning_rate": 1.03595485784579e-06,
+      "loss": 0.4992,
+      "step": 16483
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 46.73093307508055,
+      "learning_rate": 1.0352075110658588e-06,
+      "loss": 0.4725,
+      "step": 16484
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 44.924418029454614,
+      "learning_rate": 1.034460419239387e-06,
+      "loss": 0.4039,
+      "step": 16485
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 51.156947504600325,
+      "learning_rate": 1.033713582387611e-06,
+      "loss": 0.4594,
+      "step": 16486
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 40.31118298826261,
+      "learning_rate": 1.032967000531777e-06,
+      "loss": 0.403,
+      "step": 16487
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.61298976150244,
+      "learning_rate": 1.0322206736931152e-06,
+      "loss": 0.398,
+      "step": 16488
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 35.70148918027126,
+      "learning_rate": 1.0314746018928535e-06,
+      "loss": 0.3495,
+      "step": 16489
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 49.41781206342091,
+      "learning_rate": 1.0307287851522074e-06,
+      "loss": 0.4437,
+      "step": 16490
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 41.40949670118672,
+      "learning_rate": 1.0299832234923857e-06,
+      "loss": 0.3873,
+      "step": 16491
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 52.10689046675917,
+      "learning_rate": 1.0292379169345945e-06,
+      "loss": 0.4916,
+      "step": 16492
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 53.628648495990745,
+      "learning_rate": 1.0284928655000303e-06,
+      "loss": 0.4345,
+      "step": 16493
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 55.76289897571595,
+      "learning_rate": 1.0277480692098796e-06,
+      "loss": 0.431,
+      "step": 16494
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 47.178739959858845,
+      "learning_rate": 1.0270035280853275e-06,
+      "loss": 0.3976,
+      "step": 16495
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.17313721304217,
+      "learning_rate": 1.0262592421475436e-06,
+      "loss": 0.4207,
+      "step": 16496
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 51.67421017870448,
+      "learning_rate": 1.025515211417697e-06,
+      "loss": 0.364,
+      "step": 16497
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 44.49567160290178,
+      "learning_rate": 1.0247714359169502e-06,
+      "loss": 0.4572,
+      "step": 16498
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 39.98454890748128,
+      "learning_rate": 1.0240279156664512e-06,
+      "loss": 0.3183,
+      "step": 16499
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 53.30519909971715,
+      "learning_rate": 1.0232846506873495e-06,
+      "loss": 0.4619,
+      "step": 16500
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.03361514342157,
+      "learning_rate": 1.0225416410007794e-06,
+      "loss": 0.4176,
+      "step": 16501
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 56.68548970721844,
+      "learning_rate": 1.021798886627875e-06,
+      "loss": 0.5346,
+      "step": 16502
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 54.22275289660607,
+      "learning_rate": 1.0210563875897561e-06,
+      "loss": 0.3955,
+      "step": 16503
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 52.60309530042164,
+      "learning_rate": 1.0203141439075415e-06,
+      "loss": 0.4388,
+      "step": 16504
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.712576141572,
+      "learning_rate": 1.0195721556023409e-06,
+      "loss": 0.4848,
+      "step": 16505
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 46.697132465087016,
+      "learning_rate": 1.0188304226952562e-06,
+      "loss": 0.4431,
+      "step": 16506
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.42932636251928,
+      "learning_rate": 1.0180889452073772e-06,
+      "loss": 0.4349,
+      "step": 16507
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 41.1922470043146,
+      "learning_rate": 1.017347723159795e-06,
+      "loss": 0.4107,
+      "step": 16508
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.59878491385823,
+      "learning_rate": 1.0166067565735881e-06,
+      "loss": 0.4414,
+      "step": 16509
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 56.928191515951795,
+      "learning_rate": 1.015866045469832e-06,
+      "loss": 0.4044,
+      "step": 16510
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.308503025956334,
+      "learning_rate": 1.0151255898695911e-06,
+      "loss": 0.3542,
+      "step": 16511
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 51.64898714584666,
+      "learning_rate": 1.0143853897939193e-06,
+      "loss": 0.4382,
+      "step": 16512
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 47.139230466654226,
+      "learning_rate": 1.013645445263871e-06,
+      "loss": 0.3494,
+      "step": 16513
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.610969838557764,
+      "learning_rate": 1.012905756300492e-06,
+      "loss": 0.4696,
+      "step": 16514
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 58.249813951350724,
+      "learning_rate": 1.0121663229248145e-06,
+      "loss": 0.5718,
+      "step": 16515
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 59.917848104077805,
+      "learning_rate": 1.0114271451578684e-06,
+      "loss": 0.5049,
+      "step": 16516
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 49.78785616234669,
+      "learning_rate": 1.0106882230206749e-06,
+      "loss": 0.503,
+      "step": 16517
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 73.9097117349203,
+      "learning_rate": 1.0099495565342532e-06,
+      "loss": 0.4443,
+      "step": 16518
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 42.63809063495919,
+      "learning_rate": 1.0092111457196041e-06,
+      "loss": 0.3799,
+      "step": 16519
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 49.514888431731926,
+      "learning_rate": 1.0084729905977332e-06,
+      "loss": 0.475,
+      "step": 16520
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 43.77770613078458,
+      "learning_rate": 1.0077350911896278e-06,
+      "loss": 0.34,
+      "step": 16521
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 46.54975129787333,
+      "learning_rate": 1.006997447516276e-06,
+      "loss": 0.3794,
+      "step": 16522
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 54.10119073512838,
+      "learning_rate": 1.0062600595986582e-06,
+      "loss": 0.4197,
+      "step": 16523
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 57.87351595662846,
+      "learning_rate": 1.0055229274577417e-06,
+      "loss": 0.4468,
+      "step": 16524
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 43.66925545958577,
+      "learning_rate": 1.0047860511144937e-06,
+      "loss": 0.4425,
+      "step": 16525
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 55.815189232707496,
+      "learning_rate": 1.004049430589865e-06,
+      "loss": 0.4544,
+      "step": 16526
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 55.34746812148641,
+      "learning_rate": 1.0033130659048119e-06,
+      "loss": 0.534,
+      "step": 16527
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 42.05335308737937,
+      "learning_rate": 1.002576957080269e-06,
+      "loss": 0.3688,
+      "step": 16528
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 44.751134778131735,
+      "learning_rate": 1.0018411041371756e-06,
+      "loss": 0.4424,
+      "step": 16529
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 61.45521417321168,
+      "learning_rate": 1.001105507096457e-06,
+      "loss": 0.4658,
+      "step": 16530
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 42.99430713007,
+      "learning_rate": 1.0003701659790344e-06,
+      "loss": 0.4244,
+      "step": 16531
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 41.81604180780751,
+      "learning_rate": 9.996350808058175e-07,
+      "loss": 0.4188,
+      "step": 16532
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.68064561768432,
+      "learning_rate": 9.989002515977154e-07,
+      "loss": 0.4219,
+      "step": 16533
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 56.48791268272442,
+      "learning_rate": 9.981656783756255e-07,
+      "loss": 0.5097,
+      "step": 16534
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 36.31786926546068,
+      "learning_rate": 9.974313611604358e-07,
+      "loss": 0.3831,
+      "step": 16535
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 44.46338767601868,
+      "learning_rate": 9.96697299973034e-07,
+      "loss": 0.4001,
+      "step": 16536
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.34445552158748,
+      "learning_rate": 9.959634948342923e-07,
+      "loss": 0.4124,
+      "step": 16537
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 51.57356351589176,
+      "learning_rate": 9.952299457650805e-07,
+      "loss": 0.3527,
+      "step": 16538
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 43.1261398220084,
+      "learning_rate": 9.944966527862644e-07,
+      "loss": 0.3591,
+      "step": 16539
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 109.63461265045645,
+      "learning_rate": 9.93763615918696e-07,
+      "loss": 0.4326,
+      "step": 16540
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.78995440655122,
+      "learning_rate": 9.930308351832185e-07,
+      "loss": 0.4167,
+      "step": 16541
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.96631964401946,
+      "learning_rate": 9.922983106006766e-07,
+      "loss": 0.4409,
+      "step": 16542
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.781151775996854,
+      "learning_rate": 9.915660421919027e-07,
+      "loss": 0.4157,
+      "step": 16543
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 48.7544398022249,
+      "learning_rate": 9.908340299777208e-07,
+      "loss": 0.4611,
+      "step": 16544
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 42.209006116174024,
+      "learning_rate": 9.901022739789468e-07,
+      "loss": 0.3763,
+      "step": 16545
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 45.30654265606157,
+      "learning_rate": 9.893707742163926e-07,
+      "loss": 0.4025,
+      "step": 16546
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 50.24885585298433,
+      "learning_rate": 9.886395307108643e-07,
+      "loss": 0.4643,
+      "step": 16547
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 55.14793926013405,
+      "learning_rate": 9.87908543483157e-07,
+      "loss": 0.4565,
+      "step": 16548
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 44.40406354746813,
+      "learning_rate": 9.871778125540587e-07,
+      "loss": 0.4088,
+      "step": 16549
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 40.65967214667956,
+      "learning_rate": 9.864473379443495e-07,
+      "loss": 0.3927,
+      "step": 16550
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 53.08440820373262,
+      "learning_rate": 9.85717119674806e-07,
+      "loss": 0.3983,
+      "step": 16551
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 50.40370116438145,
+      "learning_rate": 9.84987157766195e-07,
+      "loss": 0.4294,
+      "step": 16552
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 50.92936067202745,
+      "learning_rate": 9.842574522392744e-07,
+      "loss": 0.4518,
+      "step": 16553
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 48.01479496534816,
+      "learning_rate": 9.835280031147999e-07,
+      "loss": 0.3547,
+      "step": 16554
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.41252019951579,
+      "learning_rate": 9.827988104135122e-07,
+      "loss": 0.3899,
+      "step": 16555
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 44.13398706267956,
+      "learning_rate": 9.82069874156154e-07,
+      "loss": 0.3682,
+      "step": 16556
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 50.05028698569305,
+      "learning_rate": 9.81341194363451e-07,
+      "loss": 0.4369,
+      "step": 16557
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 41.49973139618503,
+      "learning_rate": 9.806127710561274e-07,
+      "loss": 0.3902,
+      "step": 16558
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 52.715786871542,
+      "learning_rate": 9.798846042549037e-07,
+      "loss": 0.4891,
+      "step": 16559
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.75941332557387,
+      "learning_rate": 9.79156693980483e-07,
+      "loss": 0.4416,
+      "step": 16560
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 44.02215735699067,
+      "learning_rate": 9.784290402535713e-07,
+      "loss": 0.3932,
+      "step": 16561
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 40.73506367067976,
+      "learning_rate": 9.777016430948572e-07,
+      "loss": 0.4015,
+      "step": 16562
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 46.14499308536571,
+      "learning_rate": 9.76974502525031e-07,
+      "loss": 0.3803,
+      "step": 16563
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 55.27403131064919,
+      "learning_rate": 9.76247618564774e-07,
+      "loss": 0.5198,
+      "step": 16564
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 60.825251631210385,
+      "learning_rate": 9.755209912347552e-07,
+      "loss": 0.4516,
+      "step": 16565
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 43.221868938958345,
+      "learning_rate": 9.747946205556391e-07,
+      "loss": 0.4662,
+      "step": 16566
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 55.16403113098074,
+      "learning_rate": 9.740685065480837e-07,
+      "loss": 0.5011,
+      "step": 16567
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.167354620143584,
+      "learning_rate": 9.733426492327425e-07,
+      "loss": 0.3821,
+      "step": 16568
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.686573807164486,
+      "learning_rate": 9.726170486302554e-07,
+      "loss": 0.4077,
+      "step": 16569
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 52.39591494554518,
+      "learning_rate": 9.718917047612575e-07,
+      "loss": 0.4204,
+      "step": 16570
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 50.879177423439124,
+      "learning_rate": 9.711666176463775e-07,
+      "loss": 0.4537,
+      "step": 16571
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 51.793936469726454,
+      "learning_rate": 9.704417873062366e-07,
+      "loss": 0.4906,
+      "step": 16572
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.89054242478326,
+      "learning_rate": 9.697172137614518e-07,
+      "loss": 0.3887,
+      "step": 16573
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 58.64091308669353,
+      "learning_rate": 9.689928970326268e-07,
+      "loss": 0.474,
+      "step": 16574
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 41.81131350397635,
+      "learning_rate": 9.682688371403593e-07,
+      "loss": 0.3554,
+      "step": 16575
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 38.01995019104779,
+      "learning_rate": 9.675450341052427e-07,
+      "loss": 0.3915,
+      "step": 16576
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 53.06372656600876,
+      "learning_rate": 9.668214879478622e-07,
+      "loss": 0.4787,
+      "step": 16577
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 48.53603018697275,
+      "learning_rate": 9.66098198688795e-07,
+      "loss": 0.4676,
+      "step": 16578
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.48102137731635,
+      "learning_rate": 9.65375166348609e-07,
+      "loss": 0.4478,
+      "step": 16579
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 61.57943722639041,
+      "learning_rate": 9.646523909478677e-07,
+      "loss": 0.5143,
+      "step": 16580
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 39.70779493566665,
+      "learning_rate": 9.63929872507129e-07,
+      "loss": 0.4585,
+      "step": 16581
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 54.57150901187198,
+      "learning_rate": 9.632076110469368e-07,
+      "loss": 0.4817,
+      "step": 16582
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 54.07711078582211,
+      "learning_rate": 9.624856065878351e-07,
+      "loss": 0.4701,
+      "step": 16583
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.303773662330485,
+      "learning_rate": 9.617638591503542e-07,
+      "loss": 0.4233,
+      "step": 16584
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 52.30225438789801,
+      "learning_rate": 9.61042368755023e-07,
+      "loss": 0.4395,
+      "step": 16585
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.19671453986542,
+      "learning_rate": 9.603211354223597e-07,
+      "loss": 0.4488,
+      "step": 16586
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 51.857517440183145,
+      "learning_rate": 9.596001591728743e-07,
+      "loss": 0.5125,
+      "step": 16587
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 50.66594903002687,
+      "learning_rate": 9.58879440027074e-07,
+      "loss": 0.4621,
+      "step": 16588
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 44.80565666607235,
+      "learning_rate": 9.5815897800545e-07,
+      "loss": 0.4116,
+      "step": 16589
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.001568585456894,
+      "learning_rate": 9.574387731284984e-07,
+      "loss": 0.4147,
+      "step": 16590
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 43.113460804803076,
+      "learning_rate": 9.56718825416697e-07,
+      "loss": 0.5101,
+      "step": 16591
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 51.37213070600034,
+      "learning_rate": 9.559991348905218e-07,
+      "loss": 0.4846,
+      "step": 16592
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 46.68103696013764,
+      "learning_rate": 9.55279701570442e-07,
+      "loss": 0.4225,
+      "step": 16593
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 49.985887223764834,
+      "learning_rate": 9.545605254769164e-07,
+      "loss": 0.5112,
+      "step": 16594
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.19886845448672,
+      "learning_rate": 9.538416066303956e-07,
+      "loss": 0.4615,
+      "step": 16595
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 40.36123979892476,
+      "learning_rate": 9.531229450513291e-07,
+      "loss": 0.3679,
+      "step": 16596
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 42.76129150455964,
+      "learning_rate": 9.524045407601534e-07,
+      "loss": 0.4373,
+      "step": 16597
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 37.28922872867933,
+      "learning_rate": 9.516863937773013e-07,
+      "loss": 0.3735,
+      "step": 16598
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 55.32273554522396,
+      "learning_rate": 9.509685041231953e-07,
+      "loss": 0.5128,
+      "step": 16599
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 49.82430826176851,
+      "learning_rate": 9.502508718182491e-07,
+      "loss": 0.4391,
+      "step": 16600
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 42.43918394475366,
+      "learning_rate": 9.49533496882874e-07,
+      "loss": 0.4431,
+      "step": 16601
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 46.50370313398983,
+      "learning_rate": 9.488163793374749e-07,
+      "loss": 0.4866,
+      "step": 16602
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 41.715834026368796,
+      "learning_rate": 9.480995192024423e-07,
+      "loss": 0.3723,
+      "step": 16603
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 48.48848701272975,
+      "learning_rate": 9.473829164981629e-07,
+      "loss": 0.4421,
+      "step": 16604
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 51.68684801938217,
+      "learning_rate": 9.466665712450174e-07,
+      "loss": 0.4764,
+      "step": 16605
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 46.424752206916544,
+      "learning_rate": 9.459504834633804e-07,
+      "loss": 0.4064,
+      "step": 16606
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 38.308280357742994,
+      "learning_rate": 9.452346531736134e-07,
+      "loss": 0.3943,
+      "step": 16607
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 44.15968171193301,
+      "learning_rate": 9.44519080396078e-07,
+      "loss": 0.4279,
+      "step": 16608
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.0521561209324,
+      "learning_rate": 9.438037651511201e-07,
+      "loss": 0.403,
+      "step": 16609
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 52.18479196980499,
+      "learning_rate": 9.430887074590855e-07,
+      "loss": 0.4579,
+      "step": 16610
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.723536899824786,
+      "learning_rate": 9.423739073403126e-07,
+      "loss": 0.4003,
+      "step": 16611
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 49.50529910150669,
+      "learning_rate": 9.416593648151251e-07,
+      "loss": 0.4261,
+      "step": 16612
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 51.70396972777779,
+      "learning_rate": 9.409450799038478e-07,
+      "loss": 0.4499,
+      "step": 16613
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 56.95218323398579,
+      "learning_rate": 9.40231052626791e-07,
+      "loss": 0.4011,
+      "step": 16614
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.6477245622104,
+      "learning_rate": 9.395172830042653e-07,
+      "loss": 0.4094,
+      "step": 16615
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.946733673798896,
+      "learning_rate": 9.388037710565667e-07,
+      "loss": 0.5003,
+      "step": 16616
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 38.96865657804375,
+      "learning_rate": 9.380905168039878e-07,
+      "loss": 0.3636,
+      "step": 16617
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.02717326292011,
+      "learning_rate": 9.373775202668156e-07,
+      "loss": 0.3842,
+      "step": 16618
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.424455317851894,
+      "learning_rate": 9.36664781465324e-07,
+      "loss": 0.4588,
+      "step": 16619
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.966710896543475,
+      "learning_rate": 9.359523004197835e-07,
+      "loss": 0.4423,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 48.7262766824951,
+      "learning_rate": 9.352400771504566e-07,
+      "loss": 0.439,
+      "step": 16621
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 48.945136204042036,
+      "learning_rate": 9.345281116775995e-07,
+      "loss": 0.5105,
+      "step": 16622
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 51.96133328179389,
+      "learning_rate": 9.338164040214614e-07,
+      "loss": 0.4355,
+      "step": 16623
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.88285602096524,
+      "learning_rate": 9.331049542022818e-07,
+      "loss": 0.3772,
+      "step": 16624
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 52.847232679025446,
+      "learning_rate": 9.323937622402912e-07,
+      "loss": 0.4298,
+      "step": 16625
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 59.636261922073956,
+      "learning_rate": 9.316828281557178e-07,
+      "loss": 0.4459,
+      "step": 16626
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.128545759255424,
+      "learning_rate": 9.309721519687809e-07,
+      "loss": 0.4219,
+      "step": 16627
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 57.22366906881083,
+      "learning_rate": 9.30261733699691e-07,
+      "loss": 0.5717,
+      "step": 16628
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 41.3715008958836,
+      "learning_rate": 9.295515733686511e-07,
+      "loss": 0.3473,
+      "step": 16629
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 61.17150561536632,
+      "learning_rate": 9.28841670995857e-07,
+      "loss": 0.5618,
+      "step": 16630
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 72.95305135645185,
+      "learning_rate": 9.281320266015015e-07,
+      "loss": 0.5204,
+      "step": 16631
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 54.380730570346834,
+      "learning_rate": 9.274226402057651e-07,
+      "loss": 0.4937,
+      "step": 16632
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 44.78560850411675,
+      "learning_rate": 9.267135118288184e-07,
+      "loss": 0.3727,
+      "step": 16633
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 45.69999288667419,
+      "learning_rate": 9.26004641490833e-07,
+      "loss": 0.4121,
+      "step": 16634
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 43.31660594992194,
+      "learning_rate": 9.252960292119661e-07,
+      "loss": 0.427,
+      "step": 16635
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 53.009330827842696,
+      "learning_rate": 9.24587675012375e-07,
+      "loss": 0.4133,
+      "step": 16636
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.814709291963574,
+      "learning_rate": 9.238795789122002e-07,
+      "loss": 0.4139,
+      "step": 16637
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.61685749267665,
+      "learning_rate": 9.231717409315788e-07,
+      "loss": 0.5098,
+      "step": 16638
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.36666178224503,
+      "learning_rate": 9.224641610906437e-07,
+      "loss": 0.3864,
+      "step": 16639
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 40.979538748906755,
+      "learning_rate": 9.2175683940952e-07,
+      "loss": 0.3769,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.80730022922458,
+      "learning_rate": 9.210497759083193e-07,
+      "loss": 0.4398,
+      "step": 16641
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 42.90513350443424,
+      "learning_rate": 9.203429706071531e-07,
+      "loss": 0.3905,
+      "step": 16642
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 59.22223392595402,
+      "learning_rate": 9.196364235261202e-07,
+      "loss": 0.4672,
+      "step": 16643
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 44.57946329818877,
+      "learning_rate": 9.189301346853163e-07,
+      "loss": 0.4605,
+      "step": 16644
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 52.14163482300868,
+      "learning_rate": 9.182241041048267e-07,
+      "loss": 0.4506,
+      "step": 16645
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.60791339389697,
+      "learning_rate": 9.175183318047298e-07,
+      "loss": 0.492,
+      "step": 16646
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.67344130093044,
+      "learning_rate": 9.168128178051005e-07,
+      "loss": 0.3854,
+      "step": 16647
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 52.61661187687558,
+      "learning_rate": 9.161075621259997e-07,
+      "loss": 0.4141,
+      "step": 16648
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.70210079589254,
+      "learning_rate": 9.154025647874875e-07,
+      "loss": 0.4789,
+      "step": 16649
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 53.47055155908908,
+      "learning_rate": 9.146978258096107e-07,
+      "loss": 0.4464,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 47.801411408771116,
+      "learning_rate": 9.139933452124117e-07,
+      "loss": 0.4654,
+      "step": 16651
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 40.08780219306438,
+      "learning_rate": 9.132891230159302e-07,
+      "loss": 0.4106,
+      "step": 16652
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.567805647232596,
+      "learning_rate": 9.125851592401891e-07,
+      "loss": 0.448,
+      "step": 16653
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 44.648879781086016,
+      "learning_rate": 9.11881453905209e-07,
+      "loss": 0.3694,
+      "step": 16654
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.382144126237435,
+      "learning_rate": 9.111780070310038e-07,
+      "loss": 0.4304,
+      "step": 16655
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.63937021992537,
+      "learning_rate": 9.104748186375811e-07,
+      "loss": 0.4379,
+      "step": 16656
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 52.38264342642743,
+      "learning_rate": 9.097718887449381e-07,
+      "loss": 0.4782,
+      "step": 16657
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.383829715748774,
+      "learning_rate": 9.090692173730619e-07,
+      "loss": 0.4906,
+      "step": 16658
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 48.86554427521116,
+      "learning_rate": 9.083668045419391e-07,
+      "loss": 0.4067,
+      "step": 16659
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.9971933439948,
+      "learning_rate": 9.076646502715469e-07,
+      "loss": 0.4272,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 52.67517245644182,
+      "learning_rate": 9.069627545818549e-07,
+      "loss": 0.4983,
+      "step": 16661
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.42463200248367,
+      "learning_rate": 9.062611174928226e-07,
+      "loss": 0.4477,
+      "step": 16662
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 54.28200627548459,
+      "learning_rate": 9.05559739024402e-07,
+      "loss": 0.4419,
+      "step": 16663
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.58474545678584,
+      "learning_rate": 9.048586191965437e-07,
+      "loss": 0.4519,
+      "step": 16664
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 50.62467259430684,
+      "learning_rate": 9.04157758029186e-07,
+      "loss": 0.4817,
+      "step": 16665
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 45.280219919981526,
+      "learning_rate": 9.0345715554226e-07,
+      "loss": 0.4281,
+      "step": 16666
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 51.08492371270341,
+      "learning_rate": 9.027568117556928e-07,
+      "loss": 0.4364,
+      "step": 16667
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 46.75791034481102,
+      "learning_rate": 9.020567266893976e-07,
+      "loss": 0.4445,
+      "step": 16668
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 50.86997397244213,
+      "learning_rate": 9.013569003632894e-07,
+      "loss": 0.4769,
+      "step": 16669
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 39.28515953328708,
+      "learning_rate": 9.006573327972667e-07,
+      "loss": 0.3997,
+      "step": 16670
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 51.14342358021279,
+      "learning_rate": 8.99958024011226e-07,
+      "loss": 0.4543,
+      "step": 16671
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 41.5991548713705,
+      "learning_rate": 8.992589740250579e-07,
+      "loss": 0.4453,
+      "step": 16672
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 42.54042658865012,
+      "learning_rate": 8.985601828586399e-07,
+      "loss": 0.3912,
+      "step": 16673
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.30432246760174,
+      "learning_rate": 8.978616505318461e-07,
+      "loss": 0.4521,
+      "step": 16674
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 43.464071563665996,
+      "learning_rate": 8.971633770645416e-07,
+      "loss": 0.4888,
+      "step": 16675
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 51.88194635601968,
+      "learning_rate": 8.964653624765874e-07,
+      "loss": 0.5178,
+      "step": 16676
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 42.36056851204545,
+      "learning_rate": 8.957676067878307e-07,
+      "loss": 0.4078,
+      "step": 16677
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 59.48351340725784,
+      "learning_rate": 8.950701100181203e-07,
+      "loss": 0.4251,
+      "step": 16678
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 41.698440511619545,
+      "learning_rate": 8.943728721872868e-07,
+      "loss": 0.3977,
+      "step": 16679
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 51.69085056095898,
+      "learning_rate": 8.936758933151623e-07,
+      "loss": 0.419,
+      "step": 16680
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.705327398227986,
+      "learning_rate": 8.92979173421571e-07,
+      "loss": 0.3755,
+      "step": 16681
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 52.88259501728789,
+      "learning_rate": 8.922827125263234e-07,
+      "loss": 0.4572,
+      "step": 16682
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 55.32567900212113,
+      "learning_rate": 8.91586510649226e-07,
+      "loss": 0.512,
+      "step": 16683
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 46.24635620197378,
+      "learning_rate": 8.90890567810081e-07,
+      "loss": 0.4106,
+      "step": 16684
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 46.58846834310949,
+      "learning_rate": 8.901948840286789e-07,
+      "loss": 0.3841,
+      "step": 16685
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 45.01183213339922,
+      "learning_rate": 8.894994593248063e-07,
+      "loss": 0.4032,
+      "step": 16686
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 39.049940217151864,
+      "learning_rate": 8.888042937182406e-07,
+      "loss": 0.406,
+      "step": 16687
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 39.52429873298601,
+      "learning_rate": 8.881093872287483e-07,
+      "loss": 0.3769,
+      "step": 16688
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 46.19012343062934,
+      "learning_rate": 8.874147398760957e-07,
+      "loss": 0.4214,
+      "step": 16689
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 67.82987844805798,
+      "learning_rate": 8.867203516800382e-07,
+      "loss": 0.5264,
+      "step": 16690
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 44.25373385996261,
+      "learning_rate": 8.860262226603234e-07,
+      "loss": 0.4615,
+      "step": 16691
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 47.403867688101954,
+      "learning_rate": 8.853323528366886e-07,
+      "loss": 0.4616,
+      "step": 16692
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 40.36214044705532,
+      "learning_rate": 8.846387422288704e-07,
+      "loss": 0.3413,
+      "step": 16693
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 48.98524593084,
+      "learning_rate": 8.839453908565965e-07,
+      "loss": 0.389,
+      "step": 16694
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 53.641779957445685,
+      "learning_rate": 8.832522987395798e-07,
+      "loss": 0.4198,
+      "step": 16695
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 37.94935098213722,
+      "learning_rate": 8.825594658975367e-07,
+      "loss": 0.3406,
+      "step": 16696
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 46.45016258589065,
+      "learning_rate": 8.818668923501683e-07,
+      "loss": 0.4479,
+      "step": 16697
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 57.41933192843388,
+      "learning_rate": 8.811745781171699e-07,
+      "loss": 0.4926,
+      "step": 16698
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 52.94592747268519,
+      "learning_rate": 8.804825232182345e-07,
+      "loss": 0.4825,
+      "step": 16699
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.161629896013615,
+      "learning_rate": 8.7979072767304e-07,
+      "loss": 0.4148,
+      "step": 16700
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 54.41318010763536,
+      "learning_rate": 8.790991915012636e-07,
+      "loss": 0.4679,
+      "step": 16701
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 46.03798960793852,
+      "learning_rate": 8.784079147225688e-07,
+      "loss": 0.3927,
+      "step": 16702
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 55.3881150638743,
+      "learning_rate": 8.777168973566186e-07,
+      "loss": 0.44,
+      "step": 16703
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 48.105178397228194,
+      "learning_rate": 8.770261394230617e-07,
+      "loss": 0.4361,
+      "step": 16704
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 47.11906722589814,
+      "learning_rate": 8.763356409415447e-07,
+      "loss": 0.4405,
+      "step": 16705
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 44.510980987125855,
+      "learning_rate": 8.756454019317063e-07,
+      "loss": 0.3864,
+      "step": 16706
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 47.732076816208355,
+      "learning_rate": 8.749554224131751e-07,
+      "loss": 0.4223,
+      "step": 16707
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 50.59850824705296,
+      "learning_rate": 8.742657024055723e-07,
+      "loss": 0.4294,
+      "step": 16708
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.42156600704939,
+      "learning_rate": 8.735762419285143e-07,
+      "loss": 0.3435,
+      "step": 16709
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 40.987868462267826,
+      "learning_rate": 8.728870410016099e-07,
+      "loss": 0.3727,
+      "step": 16710
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 47.26353591715899,
+      "learning_rate": 8.721980996444601e-07,
+      "loss": 0.4044,
+      "step": 16711
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.375892016494156,
+      "learning_rate": 8.715094178766559e-07,
+      "loss": 0.4155,
+      "step": 16712
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 40.199154363819176,
+      "learning_rate": 8.708209957177826e-07,
+      "loss": 0.3576,
+      "step": 16713
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 49.323942720669976,
+      "learning_rate": 8.701328331874204e-07,
+      "loss": 0.3879,
+      "step": 16714
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.80466083044795,
+      "learning_rate": 8.694449303051411e-07,
+      "loss": 0.33,
+      "step": 16715
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.23936588554628,
+      "learning_rate": 8.68757287090507e-07,
+      "loss": 0.4137,
+      "step": 16716
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 52.247725806868814,
+      "learning_rate": 8.680699035630713e-07,
+      "loss": 0.4436,
+      "step": 16717
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 61.24564476877773,
+      "learning_rate": 8.673827797423862e-07,
+      "loss": 0.6015,
+      "step": 16718
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 52.46586891793351,
+      "learning_rate": 8.666959156479938e-07,
+      "loss": 0.4339,
+      "step": 16719
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 38.61651855543795,
+      "learning_rate": 8.660093112994261e-07,
+      "loss": 0.3906,
+      "step": 16720
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.52602789700112,
+      "learning_rate": 8.653229667162111e-07,
+      "loss": 0.4181,
+      "step": 16721
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.59161956394174,
+      "learning_rate": 8.646368819178652e-07,
+      "loss": 0.4148,
+      "step": 16722
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.3512330513015,
+      "learning_rate": 8.639510569239029e-07,
+      "loss": 0.388,
+      "step": 16723
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 43.20076821712625,
+      "learning_rate": 8.632654917538297e-07,
+      "loss": 0.4633,
+      "step": 16724
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 50.726431950586665,
+      "learning_rate": 8.625801864271411e-07,
+      "loss": 0.4578,
+      "step": 16725
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 48.24972686294963,
+      "learning_rate": 8.618951409633258e-07,
+      "loss": 0.4289,
+      "step": 16726
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 58.23746235214362,
+      "learning_rate": 8.612103553818663e-07,
+      "loss": 0.4877,
+      "step": 16727
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 47.41632988826351,
+      "learning_rate": 8.605258297022401e-07,
+      "loss": 0.4656,
+      "step": 16728
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 55.81680857910966,
+      "learning_rate": 8.598415639439107e-07,
+      "loss": 0.4938,
+      "step": 16729
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 41.43954404619679,
+      "learning_rate": 8.591575581263422e-07,
+      "loss": 0.3973,
+      "step": 16730
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.07444168554503,
+      "learning_rate": 8.58473812268984e-07,
+      "loss": 0.3953,
+      "step": 16731
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.728158306844556,
+      "learning_rate": 8.577903263912846e-07,
+      "loss": 0.3544,
+      "step": 16732
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 52.98174892477272,
+      "learning_rate": 8.571071005126785e-07,
+      "loss": 0.4408,
+      "step": 16733
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 37.17173506015703,
+      "learning_rate": 8.564241346525992e-07,
+      "loss": 0.3525,
+      "step": 16734
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 45.78943731118458,
+      "learning_rate": 8.557414288304689e-07,
+      "loss": 0.387,
+      "step": 16735
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 53.97222979293529,
+      "learning_rate": 8.550589830657019e-07,
+      "loss": 0.4555,
+      "step": 16736
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 50.86123276934452,
+      "learning_rate": 8.543767973777095e-07,
+      "loss": 0.459,
+      "step": 16737
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 47.763934498909016,
+      "learning_rate": 8.536948717858895e-07,
+      "loss": 0.429,
+      "step": 16738
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.07184124704183,
+      "learning_rate": 8.530132063096364e-07,
+      "loss": 0.3671,
+      "step": 16739
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 53.82206370826663,
+      "learning_rate": 8.52331800968339e-07,
+      "loss": 0.4252,
+      "step": 16740
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 49.779781729387096,
+      "learning_rate": 8.516506557813742e-07,
+      "loss": 0.4196,
+      "step": 16741
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.35495151117413,
+      "learning_rate": 8.509697707681108e-07,
+      "loss": 0.3983,
+      "step": 16742
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 47.100651711816646,
+      "learning_rate": 8.502891459479145e-07,
+      "loss": 0.3966,
+      "step": 16743
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 44.67969595516663,
+      "learning_rate": 8.496087813401454e-07,
+      "loss": 0.4172,
+      "step": 16744
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 50.701437660201606,
+      "learning_rate": 8.489286769641492e-07,
+      "loss": 0.4226,
+      "step": 16745
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.24150548784897,
+      "learning_rate": 8.482488328392668e-07,
+      "loss": 0.4748,
+      "step": 16746
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 41.70437065085043,
+      "learning_rate": 8.475692489848342e-07,
+      "loss": 0.491,
+      "step": 16747
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 50.18800589851612,
+      "learning_rate": 8.468899254201768e-07,
+      "loss": 0.4448,
+      "step": 16748
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 57.88561347551327,
+      "learning_rate": 8.462108621646182e-07,
+      "loss": 0.5343,
+      "step": 16749
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.26598585529793,
+      "learning_rate": 8.455320592374683e-07,
+      "loss": 0.4276,
+      "step": 16750
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 43.81357691226284,
+      "learning_rate": 8.448535166580286e-07,
+      "loss": 0.3821,
+      "step": 16751
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.679585088953644,
+      "learning_rate": 8.441752344456001e-07,
+      "loss": 0.3958,
+      "step": 16752
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 41.028794955320876,
+      "learning_rate": 8.43497212619474e-07,
+      "loss": 0.3925,
+      "step": 16753
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 45.013737007014875,
+      "learning_rate": 8.428194511989285e-07,
+      "loss": 0.4025,
+      "step": 16754
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 55.07330084702136,
+      "learning_rate": 8.421419502032425e-07,
+      "loss": 0.4443,
+      "step": 16755
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 43.416832481819355,
+      "learning_rate": 8.414647096516804e-07,
+      "loss": 0.3987,
+      "step": 16756
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.72110700025565,
+      "learning_rate": 8.40787729563507e-07,
+      "loss": 0.4372,
+      "step": 16757
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 47.436036397698366,
+      "learning_rate": 8.401110099579702e-07,
+      "loss": 0.408,
+      "step": 16758
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.604788803413854,
+      "learning_rate": 8.394345508543178e-07,
+      "loss": 0.4324,
+      "step": 16759
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.65943486968197,
+      "learning_rate": 8.387583522717901e-07,
+      "loss": 0.3496,
+      "step": 16760
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 39.929373754658094,
+      "learning_rate": 8.38082414229614e-07,
+      "loss": 0.397,
+      "step": 16761
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 42.609695590036836,
+      "learning_rate": 8.374067367470151e-07,
+      "loss": 0.3911,
+      "step": 16762
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 56.15555241342021,
+      "learning_rate": 8.367313198432081e-07,
+      "loss": 0.5037,
+      "step": 16763
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 50.425802982747335,
+      "learning_rate": 8.36056163537402e-07,
+      "loss": 0.4387,
+      "step": 16764
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 55.47489088854862,
+      "learning_rate": 8.353812678487993e-07,
+      "loss": 0.4545,
+      "step": 16765
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 40.039016574740245,
+      "learning_rate": 8.347066327965925e-07,
+      "loss": 0.3561,
+      "step": 16766
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.77788617634489,
+      "learning_rate": 8.340322583999649e-07,
+      "loss": 0.4693,
+      "step": 16767
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 42.34184412872438,
+      "learning_rate": 8.333581446780981e-07,
+      "loss": 0.4146,
+      "step": 16768
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 45.69770925924193,
+      "learning_rate": 8.326842916501654e-07,
+      "loss": 0.3951,
+      "step": 16769
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 64.55783977842657,
+      "learning_rate": 8.320106993353294e-07,
+      "loss": 0.5562,
+      "step": 16770
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 42.120358004755744,
+      "learning_rate": 8.313373677527437e-07,
+      "loss": 0.4278,
+      "step": 16771
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.05913274147307,
+      "learning_rate": 8.306642969215595e-07,
+      "loss": 0.4529,
+      "step": 16772
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.778262408002035,
+      "learning_rate": 8.299914868609193e-07,
+      "loss": 0.3984,
+      "step": 16773
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 50.70251369489864,
+      "learning_rate": 8.293189375899579e-07,
+      "loss": 0.3794,
+      "step": 16774
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 41.84373270138418,
+      "learning_rate": 8.286466491278023e-07,
+      "loss": 0.4184,
+      "step": 16775
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 45.85704632489188,
+      "learning_rate": 8.27974621493568e-07,
+      "loss": 0.4222,
+      "step": 16776
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 35.58660283058185,
+      "learning_rate": 8.27302854706371e-07,
+      "loss": 0.382,
+      "step": 16777
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 41.6914853960471,
+      "learning_rate": 8.266313487853162e-07,
+      "loss": 0.3935,
+      "step": 16778
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 54.44245890686108,
+      "learning_rate": 8.259601037494991e-07,
+      "loss": 0.4156,
+      "step": 16779
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.79641374386955,
+      "learning_rate": 8.252891196180091e-07,
+      "loss": 0.4789,
+      "step": 16780
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 51.74830199879022,
+      "learning_rate": 8.246183964099286e-07,
+      "loss": 0.4053,
+      "step": 16781
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 52.593222027835445,
+      "learning_rate": 8.239479341443357e-07,
+      "loss": 0.4626,
+      "step": 16782
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.79105320237247,
+      "learning_rate": 8.232777328402941e-07,
+      "loss": 0.4581,
+      "step": 16783
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 38.62180149698875,
+      "learning_rate": 8.226077925168674e-07,
+      "loss": 0.4089,
+      "step": 16784
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.92351004679515,
+      "learning_rate": 8.219381131931048e-07,
+      "loss": 0.4166,
+      "step": 16785
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 44.90643857094733,
+      "learning_rate": 8.212686948880521e-07,
+      "loss": 0.4143,
+      "step": 16786
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 51.810164565771316,
+      "learning_rate": 8.20599537620751e-07,
+      "loss": 0.451,
+      "step": 16787
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.1402452473085,
+      "learning_rate": 8.199306414102282e-07,
+      "loss": 0.4494,
+      "step": 16788
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 52.08149158985279,
+      "learning_rate": 8.192620062755085e-07,
+      "loss": 0.4666,
+      "step": 16789
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.791121299768285,
+      "learning_rate": 8.185936322356047e-07,
+      "loss": 0.3784,
+      "step": 16790
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 39.46193445336334,
+      "learning_rate": 8.179255193095292e-07,
+      "loss": 0.4383,
+      "step": 16791
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 54.80863814630478,
+      "learning_rate": 8.172576675162791e-07,
+      "loss": 0.4043,
+      "step": 16792
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.367755988219436,
+      "learning_rate": 8.165900768748491e-07,
+      "loss": 0.4532,
+      "step": 16793
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 52.15748768429879,
+      "learning_rate": 8.159227474042276e-07,
+      "loss": 0.5253,
+      "step": 16794
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 45.31866744589872,
+      "learning_rate": 8.15255679123389e-07,
+      "loss": 0.4297,
+      "step": 16795
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 43.494790987564365,
+      "learning_rate": 8.145888720513051e-07,
+      "loss": 0.3979,
+      "step": 16796
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 66.58001015891963,
+      "learning_rate": 8.139223262069407e-07,
+      "loss": 0.5363,
+      "step": 16797
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.15731552020215,
+      "learning_rate": 8.132560416092516e-07,
+      "loss": 0.3856,
+      "step": 16798
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 43.00441172612237,
+      "learning_rate": 8.125900182771874e-07,
+      "loss": 0.4341,
+      "step": 16799
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 56.52960494533169,
+      "learning_rate": 8.119242562296892e-07,
+      "loss": 0.4394,
+      "step": 16800
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 43.6716375081765,
+      "learning_rate": 8.112587554856888e-07,
+      "loss": 0.3985,
+      "step": 16801
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 47.717421642468366,
+      "learning_rate": 8.105935160641143e-07,
+      "loss": 0.443,
+      "step": 16802
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 49.85321838947875,
+      "learning_rate": 8.099285379838862e-07,
+      "loss": 0.3995,
+      "step": 16803
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.38563522069217,
+      "learning_rate": 8.092638212639136e-07,
+      "loss": 0.3838,
+      "step": 16804
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 41.681138990274135,
+      "learning_rate": 8.085993659231006e-07,
+      "loss": 0.4038,
+      "step": 16805
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 60.78741939977832,
+      "learning_rate": 8.079351719803441e-07,
+      "loss": 0.5433,
+      "step": 16806
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.31817552292847,
+      "learning_rate": 8.072712394545368e-07,
+      "loss": 0.4199,
+      "step": 16807
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 43.548559724968065,
+      "learning_rate": 8.066075683645557e-07,
+      "loss": 0.4647,
+      "step": 16808
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 47.30343621234694,
+      "learning_rate": 8.059441587292782e-07,
+      "loss": 0.3932,
+      "step": 16809
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 47.79316593280705,
+      "learning_rate": 8.052810105675702e-07,
+      "loss": 0.4441,
+      "step": 16810
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 43.40704213704206,
+      "learning_rate": 8.046181238982908e-07,
+      "loss": 0.4222,
+      "step": 16811
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 52.528987025637626,
+      "learning_rate": 8.039554987402942e-07,
+      "loss": 0.4463,
+      "step": 16812
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.94953431774515,
+      "learning_rate": 8.03293135112423e-07,
+      "loss": 0.4192,
+      "step": 16813
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.75068915180077,
+      "learning_rate": 8.026310330335163e-07,
+      "loss": 0.4623,
+      "step": 16814
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.45604928651389,
+      "learning_rate": 8.019691925224004e-07,
+      "loss": 0.4625,
+      "step": 16815
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 48.959431967346354,
+      "learning_rate": 8.013076135979025e-07,
+      "loss": 0.4305,
+      "step": 16816
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 65.90139953794282,
+      "learning_rate": 8.006462962788331e-07,
+      "loss": 0.4173,
+      "step": 16817
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.10822935535778,
+      "learning_rate": 7.999852405840025e-07,
+      "loss": 0.4089,
+      "step": 16818
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 56.694356543629446,
+      "learning_rate": 7.993244465322092e-07,
+      "loss": 0.4496,
+      "step": 16819
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.498978493568615,
+      "learning_rate": 7.986639141422469e-07,
+      "loss": 0.427,
+      "step": 16820
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 44.27757792837899,
+      "learning_rate": 7.980036434328997e-07,
+      "loss": 0.3815,
+      "step": 16821
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.376318684938525,
+      "learning_rate": 7.973436344229458e-07,
+      "loss": 0.3939,
+      "step": 16822
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.67716335984124,
+      "learning_rate": 7.966838871311566e-07,
+      "loss": 0.4348,
+      "step": 16823
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 53.086350260428816,
+      "learning_rate": 7.960244015762919e-07,
+      "loss": 0.4407,
+      "step": 16824
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.36080275159282,
+      "learning_rate": 7.953651777771121e-07,
+      "loss": 0.4754,
+      "step": 16825
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 46.07621828974086,
+      "learning_rate": 7.947062157523589e-07,
+      "loss": 0.4377,
+      "step": 16826
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 46.055200637663624,
+      "learning_rate": 7.940475155207772e-07,
+      "loss": 0.3904,
+      "step": 16827
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 38.29301303050395,
+      "learning_rate": 7.933890771010999e-07,
+      "loss": 0.3753,
+      "step": 16828
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 47.68657091468091,
+      "learning_rate": 7.927309005120521e-07,
+      "loss": 0.5133,
+      "step": 16829
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 48.6854028556182,
+      "learning_rate": 7.920729857723486e-07,
+      "loss": 0.3941,
+      "step": 16830
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 60.98155566732498,
+      "learning_rate": 7.914153329007035e-07,
+      "loss": 0.5122,
+      "step": 16831
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.98014138706846,
+      "learning_rate": 7.907579419158196e-07,
+      "loss": 0.4968,
+      "step": 16832
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.5459106604954,
+      "learning_rate": 7.901008128363963e-07,
+      "loss": 0.3934,
+      "step": 16833
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 46.78987471709464,
+      "learning_rate": 7.894439456811143e-07,
+      "loss": 0.4246,
+      "step": 16834
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 43.82186985904022,
+      "learning_rate": 7.887873404686586e-07,
+      "loss": 0.3621,
+      "step": 16835
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 44.98019183028496,
+      "learning_rate": 7.88130997217702e-07,
+      "loss": 0.4252,
+      "step": 16836
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 56.79052656918019,
+      "learning_rate": 7.874749159469131e-07,
+      "loss": 0.4441,
+      "step": 16837
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 47.52227010175745,
+      "learning_rate": 7.868190966749489e-07,
+      "loss": 0.3939,
+      "step": 16838
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 56.37492545264617,
+      "learning_rate": 7.861635394204581e-07,
+      "loss": 0.4121,
+      "step": 16839
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 49.35937263317289,
+      "learning_rate": 7.855082442020867e-07,
+      "loss": 0.4064,
+      "step": 16840
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 43.79392250716285,
+      "learning_rate": 7.84853211038471e-07,
+      "loss": 0.3788,
+      "step": 16841
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.20457779584281,
+      "learning_rate": 7.841984399482383e-07,
+      "loss": 0.4241,
+      "step": 16842
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.25194192902926,
+      "learning_rate": 7.835439309500126e-07,
+      "loss": 0.3867,
+      "step": 16843
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 56.51272586962158,
+      "learning_rate": 7.828896840624045e-07,
+      "loss": 0.3767,
+      "step": 16844
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.40637940547852,
+      "learning_rate": 7.822356993040236e-07,
+      "loss": 0.4057,
+      "step": 16845
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 48.48218221866061,
+      "learning_rate": 7.815819766934651e-07,
+      "loss": 0.4524,
+      "step": 16846
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 54.30854141154906,
+      "learning_rate": 7.80928516249323e-07,
+      "loss": 0.4541,
+      "step": 16847
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 48.8883294175869,
+      "learning_rate": 7.802753179901823e-07,
+      "loss": 0.3867,
+      "step": 16848
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.533273636326385,
+      "learning_rate": 7.796223819346171e-07,
+      "loss": 0.3943,
+      "step": 16849
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 41.09242042450251,
+      "learning_rate": 7.789697081011982e-07,
+      "loss": 0.3511,
+      "step": 16850
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.255142595887094,
+      "learning_rate": 7.783172965084851e-07,
+      "loss": 0.446,
+      "step": 16851
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 53.60980696297054,
+      "learning_rate": 7.77665147175034e-07,
+      "loss": 0.4941,
+      "step": 16852
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 46.48515258651593,
+      "learning_rate": 7.770132601193936e-07,
+      "loss": 0.4698,
+      "step": 16853
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 51.10322746080433,
+      "learning_rate": 7.763616353601e-07,
+      "loss": 0.4446,
+      "step": 16854
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.56281158487243,
+      "learning_rate": 7.75710272915684e-07,
+      "loss": 0.454,
+      "step": 16855
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 44.19254550428034,
+      "learning_rate": 7.750591728046719e-07,
+      "loss": 0.3991,
+      "step": 16856
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.14584928680576,
+      "learning_rate": 7.744083350455811e-07,
+      "loss": 0.437,
+      "step": 16857
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 42.974718185708,
+      "learning_rate": 7.737577596569223e-07,
+      "loss": 0.3798,
+      "step": 16858
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 43.148736828717716,
+      "learning_rate": 7.731074466571942e-07,
+      "loss": 0.3638,
+      "step": 16859
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 66.24693077463303,
+      "learning_rate": 7.724573960648907e-07,
+      "loss": 0.5226,
+      "step": 16860
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 48.473976355216266,
+      "learning_rate": 7.718076078985026e-07,
+      "loss": 0.4358,
+      "step": 16861
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 43.25894756043962,
+      "learning_rate": 7.711580821765085e-07,
+      "loss": 0.4056,
+      "step": 16862
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.3881027418429,
+      "learning_rate": 7.705088189173804e-07,
+      "loss": 0.4577,
+      "step": 16863
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 48.20976621607918,
+      "learning_rate": 7.6985981813958e-07,
+      "loss": 0.4111,
+      "step": 16864
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 47.96618794113303,
+      "learning_rate": 7.692110798615682e-07,
+      "loss": 0.4221,
+      "step": 16865
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.62527392277961,
+      "learning_rate": 7.685626041017935e-07,
+      "loss": 0.5031,
+      "step": 16866
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 57.92453393217694,
+      "learning_rate": 7.67914390878699e-07,
+      "loss": 0.5178,
+      "step": 16867
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 50.623099981374544,
+      "learning_rate": 7.672664402107166e-07,
+      "loss": 0.4244,
+      "step": 16868
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 44.64416008699116,
+      "learning_rate": 7.66618752116276e-07,
+      "loss": 0.4052,
+      "step": 16869
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 38.10406253161557,
+      "learning_rate": 7.659713266137992e-07,
+      "loss": 0.3494,
+      "step": 16870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 41.818383679590674,
+      "learning_rate": 7.653241637216924e-07,
+      "loss": 0.43,
+      "step": 16871
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 47.17536310906248,
+      "learning_rate": 7.646772634583677e-07,
+      "loss": 0.4656,
+      "step": 16872
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 54.98017231564875,
+      "learning_rate": 7.64030625842217e-07,
+      "loss": 0.4752,
+      "step": 16873
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 57.09858206276125,
+      "learning_rate": 7.633842508916323e-07,
+      "loss": 0.4507,
+      "step": 16874
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 46.9864183088792,
+      "learning_rate": 7.627381386249976e-07,
+      "loss": 0.4696,
+      "step": 16875
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 57.346197706988164,
+      "learning_rate": 7.620922890606852e-07,
+      "loss": 0.4712,
+      "step": 16876
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 58.23652326440695,
+      "learning_rate": 7.614467022170658e-07,
+      "loss": 0.3878,
+      "step": 16877
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 43.44030063025893,
+      "learning_rate": 7.608013781124956e-07,
+      "loss": 0.3655,
+      "step": 16878
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.186036032024234,
+      "learning_rate": 7.601563167653314e-07,
+      "loss": 0.375,
+      "step": 16879
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 49.78622211020212,
+      "learning_rate": 7.59511518193915e-07,
+      "loss": 0.419,
+      "step": 16880
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 43.02710229612837,
+      "learning_rate": 7.588669824165851e-07,
+      "loss": 0.4031,
+      "step": 16881
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 44.28774499741979,
+      "learning_rate": 7.582227094516747e-07,
+      "loss": 0.4007,
+      "step": 16882
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 51.87687659488656,
+      "learning_rate": 7.575786993175028e-07,
+      "loss": 0.4388,
+      "step": 16883
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 46.654735492058286,
+      "learning_rate": 7.569349520323854e-07,
+      "loss": 0.4027,
+      "step": 16884
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 63.75571484849249,
+      "learning_rate": 7.562914676146304e-07,
+      "loss": 0.5621,
+      "step": 16885
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 44.95210498849897,
+      "learning_rate": 7.556482460825399e-07,
+      "loss": 0.4008,
+      "step": 16886
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 51.04629490116694,
+      "learning_rate": 7.550052874544056e-07,
+      "loss": 0.455,
+      "step": 16887
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 47.49480270763221,
+      "learning_rate": 7.543625917485142e-07,
+      "loss": 0.465,
+      "step": 16888
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 50.919659163989216,
+      "learning_rate": 7.537201589831389e-07,
+      "loss": 0.4367,
+      "step": 16889
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 63.81319296719327,
+      "learning_rate": 7.530779891765549e-07,
+      "loss": 0.5502,
+      "step": 16890
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 54.52262632720668,
+      "learning_rate": 7.524360823470244e-07,
+      "loss": 0.4955,
+      "step": 16891
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 53.711155184217965,
+      "learning_rate": 7.517944385128018e-07,
+      "loss": 0.4541,
+      "step": 16892
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 46.87406931652581,
+      "learning_rate": 7.511530576921344e-07,
+      "loss": 0.4216,
+      "step": 16893
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 41.69334878741334,
+      "learning_rate": 7.505119399032623e-07,
+      "loss": 0.4003,
+      "step": 16894
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 42.06772721616715,
+      "learning_rate": 7.498710851644231e-07,
+      "loss": 0.3675,
+      "step": 16895
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.208239652805865,
+      "learning_rate": 7.492304934938365e-07,
+      "loss": 0.4369,
+      "step": 16896
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 45.852818216927524,
+      "learning_rate": 7.48590164909725e-07,
+      "loss": 0.3797,
+      "step": 16897
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 45.63821354613564,
+      "learning_rate": 7.479500994302957e-07,
+      "loss": 0.3803,
+      "step": 16898
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 45.539500330117505,
+      "learning_rate": 7.473102970737534e-07,
+      "loss": 0.4527,
+      "step": 16899
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 43.78082390631649,
+      "learning_rate": 7.466707578582954e-07,
+      "loss": 0.3702,
+      "step": 16900
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 51.71810661906865,
+      "learning_rate": 7.460314818021053e-07,
+      "loss": 0.4608,
+      "step": 16901
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 46.73739388273506,
+      "learning_rate": 7.453924689233693e-07,
+      "loss": 0.4816,
+      "step": 16902
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.52532662464021,
+      "learning_rate": 7.447537192402554e-07,
+      "loss": 0.4393,
+      "step": 16903
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 60.6077016475045,
+      "learning_rate": 7.441152327709334e-07,
+      "loss": 0.5187,
+      "step": 16904
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 60.28010538219816,
+      "learning_rate": 7.434770095335575e-07,
+      "loss": 0.4523,
+      "step": 16905
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.43158345173525,
+      "learning_rate": 7.428390495462812e-07,
+      "loss": 0.4162,
+      "step": 16906
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 55.766346369657875,
+      "learning_rate": 7.422013528272487e-07,
+      "loss": 0.4073,
+      "step": 16907
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 44.67782314070759,
+      "learning_rate": 7.415639193945945e-07,
+      "loss": 0.4516,
+      "step": 16908
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 52.25545689501433,
+      "learning_rate": 7.409267492664441e-07,
+      "loss": 0.4449,
+      "step": 16909
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 53.40306921114506,
+      "learning_rate": 7.402898424609206e-07,
+      "loss": 0.4879,
+      "step": 16910
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 44.74351677203042,
+      "learning_rate": 7.396531989961364e-07,
+      "loss": 0.4348,
+      "step": 16911
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 45.538089968901275,
+      "learning_rate": 7.390168188902014e-07,
+      "loss": 0.3885,
+      "step": 16912
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 46.859047736245905,
+      "learning_rate": 7.383807021612089e-07,
+      "loss": 0.4059,
+      "step": 16913
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 55.46260963946673,
+      "learning_rate": 7.377448488272509e-07,
+      "loss": 0.4841,
+      "step": 16914
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 60.509366721182396,
+      "learning_rate": 7.3710925890641e-07,
+      "loss": 0.4241,
+      "step": 16915
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 42.732498685662364,
+      "learning_rate": 7.364739324167658e-07,
+      "loss": 0.3619,
+      "step": 16916
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 50.93809124188057,
+      "learning_rate": 7.35838869376384e-07,
+      "loss": 0.3956,
+      "step": 16917
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 55.28210511093412,
+      "learning_rate": 7.352040698033236e-07,
+      "loss": 0.4388,
+      "step": 16918
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 42.84933943845803,
+      "learning_rate": 7.345695337156389e-07,
+      "loss": 0.4343,
+      "step": 16919
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 50.58870783580582,
+      "learning_rate": 7.339352611313777e-07,
+      "loss": 0.3973,
+      "step": 16920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.31715461446847,
+      "learning_rate": 7.33301252068579e-07,
+      "loss": 0.4069,
+      "step": 16921
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 51.02099327829737,
+      "learning_rate": 7.326675065452693e-07,
+      "loss": 0.4512,
+      "step": 16922
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 49.155787185071375,
+      "learning_rate": 7.320340245794755e-07,
+      "loss": 0.4548,
+      "step": 16923
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 41.60314620321268,
+      "learning_rate": 7.314008061892108e-07,
+      "loss": 0.4417,
+      "step": 16924
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 45.67314466446706,
+      "learning_rate": 7.307678513924877e-07,
+      "loss": 0.4391,
+      "step": 16925
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 35.61354094805293,
+      "learning_rate": 7.301351602073048e-07,
+      "loss": 0.337,
+      "step": 16926
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 43.646769128325396,
+      "learning_rate": 7.295027326516535e-07,
+      "loss": 0.3814,
+      "step": 16927
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 41.76664379859556,
+      "learning_rate": 7.288705687435204e-07,
+      "loss": 0.3647,
+      "step": 16928
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 50.31899281657467,
+      "learning_rate": 7.28238668500888e-07,
+      "loss": 0.4175,
+      "step": 16929
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.17427699422979,
+      "learning_rate": 7.276070319417216e-07,
+      "loss": 0.4619,
+      "step": 16930
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 48.40921456436779,
+      "learning_rate": 7.269756590839883e-07,
+      "loss": 0.3802,
+      "step": 16931
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 51.1073843353445,
+      "learning_rate": 7.263445499456412e-07,
+      "loss": 0.5244,
+      "step": 16932
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 40.791064365422685,
+      "learning_rate": 7.257137045446327e-07,
+      "loss": 0.3778,
+      "step": 16933
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 47.21852785837065,
+      "learning_rate": 7.250831228988986e-07,
+      "loss": 0.3863,
+      "step": 16934
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.704773016703996,
+      "learning_rate": 7.244528050263744e-07,
+      "loss": 0.3722,
+      "step": 16935
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 44.67947421529109,
+      "learning_rate": 7.238227509449891e-07,
+      "loss": 0.4262,
+      "step": 16936
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 38.80155529746747,
+      "learning_rate": 7.231929606726562e-07,
+      "loss": 0.3517,
+      "step": 16937
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 51.453427320531084,
+      "learning_rate": 7.225634342272903e-07,
+      "loss": 0.4061,
+      "step": 16938
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 52.88822387474015,
+      "learning_rate": 7.219341716267925e-07,
+      "loss": 0.4403,
+      "step": 16939
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.75473451323736,
+      "learning_rate": 7.213051728890586e-07,
+      "loss": 0.3976,
+      "step": 16940
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 46.592147349292944,
+      "learning_rate": 7.206764380319786e-07,
+      "loss": 0.4103,
+      "step": 16941
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 52.11326168727029,
+      "learning_rate": 7.20047967073434e-07,
+      "loss": 0.4711,
+      "step": 16942
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 42.35721027917534,
+      "learning_rate": 7.194197600312936e-07,
+      "loss": 0.4344,
+      "step": 16943
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 51.76759540044485,
+      "learning_rate": 7.187918169234265e-07,
+      "loss": 0.4019,
+      "step": 16944
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 58.66452495592322,
+      "learning_rate": 7.181641377676918e-07,
+      "loss": 0.4046,
+      "step": 16945
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 52.61764924909477,
+      "learning_rate": 7.175367225819418e-07,
+      "loss": 0.4116,
+      "step": 16946
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 43.68616127724452,
+      "learning_rate": 7.169095713840135e-07,
+      "loss": 0.4213,
+      "step": 16947
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 49.85907625487226,
+      "learning_rate": 7.16282684191747e-07,
+      "loss": 0.4338,
+      "step": 16948
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 44.82051390917977,
+      "learning_rate": 7.156560610229701e-07,
+      "loss": 0.4064,
+      "step": 16949
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 44.726859817516164,
+      "learning_rate": 7.150297018955055e-07,
+      "loss": 0.355,
+      "step": 16950
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 43.005634764816705,
+      "learning_rate": 7.144036068271654e-07,
+      "loss": 0.3944,
+      "step": 16951
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 42.34018492946142,
+      "learning_rate": 7.137777758357523e-07,
+      "loss": 0.4057,
+      "step": 16952
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 43.3509705572366,
+      "learning_rate": 7.131522089390663e-07,
+      "loss": 0.401,
+      "step": 16953
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 49.86409617384533,
+      "learning_rate": 7.125269061549012e-07,
+      "loss": 0.4839,
+      "step": 16954
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 46.070352053289724,
+      "learning_rate": 7.119018675010369e-07,
+      "loss": 0.3956,
+      "step": 16955
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 61.64623035622347,
+      "learning_rate": 7.112770929952506e-07,
+      "loss": 0.5058,
+      "step": 16956
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 53.89542671682324,
+      "learning_rate": 7.10652582655309e-07,
+      "loss": 0.4674,
+      "step": 16957
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 50.991960247100906,
+      "learning_rate": 7.100283364989757e-07,
+      "loss": 0.4363,
+      "step": 16958
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 42.37549156702794,
+      "learning_rate": 7.094043545439999e-07,
+      "loss": 0.4244,
+      "step": 16959
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 46.901723288737564,
+      "learning_rate": 7.087806368081296e-07,
+      "loss": 0.4108,
+      "step": 16960
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 58.43840929435705,
+      "learning_rate": 7.081571833091039e-07,
+      "loss": 0.4784,
+      "step": 16961
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.20042018236316,
+      "learning_rate": 7.075339940646508e-07,
+      "loss": 0.4203,
+      "step": 16962
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.322677022464894,
+      "learning_rate": 7.069110690924974e-07,
+      "loss": 0.3582,
+      "step": 16963
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 56.760154886867774,
+      "learning_rate": 7.06288408410355e-07,
+      "loss": 0.5073,
+      "step": 16964
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 46.00405688640545,
+      "learning_rate": 7.056660120359349e-07,
+      "loss": 0.4387,
+      "step": 16965
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 44.745361151623676,
+      "learning_rate": 7.050438799869352e-07,
+      "loss": 0.4493,
+      "step": 16966
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 49.67067851221206,
+      "learning_rate": 7.044220122810508e-07,
+      "loss": 0.4585,
+      "step": 16967
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 42.28234575970529,
+      "learning_rate": 7.038004089359663e-07,
+      "loss": 0.3994,
+      "step": 16968
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 49.39164273771132,
+      "learning_rate": 7.031790699693586e-07,
+      "loss": 0.3976,
+      "step": 16969
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 38.443769607328086,
+      "learning_rate": 7.025579953989026e-07,
+      "loss": 0.3516,
+      "step": 16970
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 47.83999679365275,
+      "learning_rate": 7.019371852422574e-07,
+      "loss": 0.4601,
+      "step": 16971
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 44.780176537179806,
+      "learning_rate": 7.01316639517079e-07,
+      "loss": 0.4184,
+      "step": 16972
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 50.91129610495969,
+      "learning_rate": 7.006963582410153e-07,
+      "loss": 0.4516,
+      "step": 16973
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 52.039969745190305,
+      "learning_rate": 7.000763414317069e-07,
+      "loss": 0.4273,
+      "step": 16974
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 48.0677710322723,
+      "learning_rate": 6.994565891067884e-07,
+      "loss": 0.3839,
+      "step": 16975
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 38.581934311688876,
+      "learning_rate": 6.988371012838846e-07,
+      "loss": 0.4133,
+      "step": 16976
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 56.51565218063431,
+      "learning_rate": 6.982178779806104e-07,
+      "loss": 0.4258,
+      "step": 16977
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 56.17258533776728,
+      "learning_rate": 6.975989192145782e-07,
+      "loss": 0.4557,
+      "step": 16978
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 52.81471608043811,
+      "learning_rate": 6.96980225003393e-07,
+      "loss": 0.4307,
+      "step": 16979
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 42.96812412985239,
+      "learning_rate": 6.963617953646484e-07,
+      "loss": 0.3501,
+      "step": 16980
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 44.46418370696429,
+      "learning_rate": 6.957436303159304e-07,
+      "loss": 0.418,
+      "step": 16981
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 50.56657562993673,
+      "learning_rate": 6.951257298748204e-07,
+      "loss": 0.4915,
+      "step": 16982
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 65.23849890185929,
+      "learning_rate": 6.94508094058891e-07,
+      "loss": 0.5434,
+      "step": 16983
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 50.13786746746139,
+      "learning_rate": 6.938907228857095e-07,
+      "loss": 0.4232,
+      "step": 16984
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 47.53114785328259,
+      "learning_rate": 6.932736163728327e-07,
+      "loss": 0.4424,
+      "step": 16985
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 62.6074353672811,
+      "learning_rate": 6.926567745378076e-07,
+      "loss": 0.5204,
+      "step": 16986
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 45.03653245554157,
+      "learning_rate": 6.920401973981794e-07,
+      "loss": 0.4087,
+      "step": 16987
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.01819884734827,
+      "learning_rate": 6.914238849714849e-07,
+      "loss": 0.354,
+      "step": 16988
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 46.86331689834352,
+      "learning_rate": 6.90807837275248e-07,
+      "loss": 0.4056,
+      "step": 16989
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 48.189472445085705,
+      "learning_rate": 6.901920543269925e-07,
+      "loss": 0.4401,
+      "step": 16990
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 43.565108293687466,
+      "learning_rate": 6.895765361442264e-07,
+      "loss": 0.4387,
+      "step": 16991
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 47.45006451677734,
+      "learning_rate": 6.889612827444592e-07,
+      "loss": 0.3602,
+      "step": 16992
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 50.54432423212233,
+      "learning_rate": 6.883462941451846e-07,
+      "loss": 0.4461,
+      "step": 16993
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 45.06257991016514,
+      "learning_rate": 6.877315703638943e-07,
+      "loss": 0.3836,
+      "step": 16994
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.47070411384254,
+      "learning_rate": 6.871171114180719e-07,
+      "loss": 0.4407,
+      "step": 16995
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 54.217670273817085,
+      "learning_rate": 6.865029173251914e-07,
+      "loss": 0.4522,
+      "step": 16996
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 56.88287275971547,
+      "learning_rate": 6.858889881027187e-07,
+      "loss": 0.4863,
+      "step": 16997
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 46.860487784477556,
+      "learning_rate": 6.852753237681131e-07,
+      "loss": 0.4427,
+      "step": 16998
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 51.6215900929866,
+      "learning_rate": 6.846619243388297e-07,
+      "loss": 0.4245,
+      "step": 16999
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 42.983952300067436,
+      "learning_rate": 6.840487898323122e-07,
+      "loss": 0.4246,
+      "step": 17000
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.31206535528889,
+      "learning_rate": 6.83435920265999e-07,
+      "loss": 0.3799,
+      "step": 17001
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 49.16591033784857,
+      "learning_rate": 6.82823315657315e-07,
+      "loss": 0.4542,
+      "step": 17002
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.60650266758496,
+      "learning_rate": 6.822109760236861e-07,
+      "loss": 0.4348,
+      "step": 17003
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 49.648314459503446,
+      "learning_rate": 6.815989013825285e-07,
+      "loss": 0.5032,
+      "step": 17004
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 60.30183934289603,
+      "learning_rate": 6.809870917512462e-07,
+      "loss": 0.4462,
+      "step": 17005
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 51.24136199728251,
+      "learning_rate": 6.803755471472384e-07,
+      "loss": 0.4094,
+      "step": 17006
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 50.44676847514577,
+      "learning_rate": 6.79764267587898e-07,
+      "loss": 0.4124,
+      "step": 17007
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 61.17994987033063,
+      "learning_rate": 6.791532530906098e-07,
+      "loss": 0.425,
+      "step": 17008
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 51.245199053055146,
+      "learning_rate": 6.785425036727522e-07,
+      "loss": 0.4091,
+      "step": 17009
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 42.85071907152311,
+      "learning_rate": 6.779320193516925e-07,
+      "loss": 0.4063,
+      "step": 17010
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.63429268658016,
+      "learning_rate": 6.773218001447923e-07,
+      "loss": 0.4333,
+      "step": 17011
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.57054081434094,
+      "learning_rate": 6.767118460694056e-07,
+      "loss": 0.4382,
+      "step": 17012
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 62.49365328798346,
+      "learning_rate": 6.761021571428816e-07,
+      "loss": 0.4328,
+      "step": 17013
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.35410628944363,
+      "learning_rate": 6.75492733382559e-07,
+      "loss": 0.4623,
+      "step": 17014
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 44.84134390088087,
+      "learning_rate": 6.748835748057658e-07,
+      "loss": 0.449,
+      "step": 17015
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 41.578146314131686,
+      "learning_rate": 6.742746814298285e-07,
+      "loss": 0.4257,
+      "step": 17016
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 37.51978006354881,
+      "learning_rate": 6.736660532720662e-07,
+      "loss": 0.3977,
+      "step": 17017
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 43.72396592061937,
+      "learning_rate": 6.730576903497832e-07,
+      "loss": 0.4933,
+      "step": 17018
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 43.39751273336487,
+      "learning_rate": 6.724495926802854e-07,
+      "loss": 0.4195,
+      "step": 17019
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.41258265631845,
+      "learning_rate": 6.718417602808636e-07,
+      "loss": 0.4448,
+      "step": 17020
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 47.50010009052372,
+      "learning_rate": 6.71234193168806e-07,
+      "loss": 0.4949,
+      "step": 17021
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 55.38019734139789,
+      "learning_rate": 6.7062689136139e-07,
+      "loss": 0.4322,
+      "step": 17022
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 51.59885313326818,
+      "learning_rate": 6.700198548758874e-07,
+      "loss": 0.4411,
+      "step": 17023
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.76720240175,
+      "learning_rate": 6.694130837295632e-07,
+      "loss": 0.437,
+      "step": 17024
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 55.19003299589807,
+      "learning_rate": 6.688065779396713e-07,
+      "loss": 0.5295,
+      "step": 17025
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.853506526450744,
+      "learning_rate": 6.682003375234636e-07,
+      "loss": 0.4627,
+      "step": 17026
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 49.66010716451696,
+      "learning_rate": 6.675943624981785e-07,
+      "loss": 0.414,
+      "step": 17027
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 40.91981664304215,
+      "learning_rate": 6.669886528810498e-07,
+      "loss": 0.3749,
+      "step": 17028
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.49958470746812,
+      "learning_rate": 6.663832086893052e-07,
+      "loss": 0.3608,
+      "step": 17029
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 47.15996406444607,
+      "learning_rate": 6.657780299401628e-07,
+      "loss": 0.4159,
+      "step": 17030
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 49.62142234800952,
+      "learning_rate": 6.651731166508313e-07,
+      "loss": 0.4203,
+      "step": 17031
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.740295280490336,
+      "learning_rate": 6.645684688385156e-07,
+      "loss": 0.4412,
+      "step": 17032
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.274478840002665,
+      "learning_rate": 6.639640865204111e-07,
+      "loss": 0.4632,
+      "step": 17033
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 60.47419047955173,
+      "learning_rate": 6.633599697137105e-07,
+      "loss": 0.4963,
+      "step": 17034
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.9337414551411,
+      "learning_rate": 6.627561184355869e-07,
+      "loss": 0.4367,
+      "step": 17035
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 55.03264673768613,
+      "learning_rate": 6.621525327032174e-07,
+      "loss": 0.4525,
+      "step": 17036
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 39.149243585725934,
+      "learning_rate": 6.615492125337663e-07,
+      "loss": 0.3656,
+      "step": 17037
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 54.33827471886143,
+      "learning_rate": 6.609461579443954e-07,
+      "loss": 0.4396,
+      "step": 17038
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 48.02346135503928,
+      "learning_rate": 6.603433689522531e-07,
+      "loss": 0.468,
+      "step": 17039
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 55.68080658936818,
+      "learning_rate": 6.597408455744792e-07,
+      "loss": 0.4225,
+      "step": 17040
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 36.25516051687624,
+      "learning_rate": 6.59138587828212e-07,
+      "loss": 0.343,
+      "step": 17041
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.35964880408081,
+      "learning_rate": 6.585365957305812e-07,
+      "loss": 0.4531,
+      "step": 17042
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 64.77640837386423,
+      "learning_rate": 6.57934869298703e-07,
+      "loss": 0.5068,
+      "step": 17043
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 43.75898483659257,
+      "learning_rate": 6.573334085496941e-07,
+      "loss": 0.3466,
+      "step": 17044
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 49.65510951965718,
+      "learning_rate": 6.567322135006559e-07,
+      "loss": 0.4765,
+      "step": 17045
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 42.548854912098705,
+      "learning_rate": 6.561312841686895e-07,
+      "loss": 0.3785,
+      "step": 17046
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 40.53390353655563,
+      "learning_rate": 6.555306205708812e-07,
+      "loss": 0.4147,
+      "step": 17047
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 54.80643873292068,
+      "learning_rate": 6.54930222724317e-07,
+      "loss": 0.4169,
+      "step": 17048
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 42.48219815647304,
+      "learning_rate": 6.543300906460714e-07,
+      "loss": 0.401,
+      "step": 17049
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 49.669589330566446,
+      "learning_rate": 6.537302243532095e-07,
+      "loss": 0.4221,
+      "step": 17050
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 40.86874358682776,
+      "learning_rate": 6.531306238627932e-07,
+      "loss": 0.3785,
+      "step": 17051
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 46.34295353169687,
+      "learning_rate": 6.525312891918745e-07,
+      "loss": 0.4366,
+      "step": 17052
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.62632762270306,
+      "learning_rate": 6.519322203574962e-07,
+      "loss": 0.4436,
+      "step": 17053
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 44.4587950672077,
+      "learning_rate": 6.513334173766994e-07,
+      "loss": 0.3567,
+      "step": 17054
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 51.76678705911746,
+      "learning_rate": 6.507348802665115e-07,
+      "loss": 0.4514,
+      "step": 17055
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 45.76962321676632,
+      "learning_rate": 6.501366090439521e-07,
+      "loss": 0.4286,
+      "step": 17056
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 47.43430155750854,
+      "learning_rate": 6.495386037260387e-07,
+      "loss": 0.4291,
+      "step": 17057
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.9509733308487,
+      "learning_rate": 6.489408643297778e-07,
+      "loss": 0.5641,
+      "step": 17058
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.260556774421744,
+      "learning_rate": 6.483433908721715e-07,
+      "loss": 0.4488,
+      "step": 17059
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 51.84735361167696,
+      "learning_rate": 6.477461833702047e-07,
+      "loss": 0.376,
+      "step": 17060
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 43.13570510293711,
+      "learning_rate": 6.471492418408665e-07,
+      "loss": 0.3924,
+      "step": 17061
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 44.490583977342425,
+      "learning_rate": 6.465525663011318e-07,
+      "loss": 0.3606,
+      "step": 17062
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 46.025737346894815,
+      "learning_rate": 6.459561567679728e-07,
+      "loss": 0.4008,
+      "step": 17063
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 51.549805991753985,
+      "learning_rate": 6.453600132583482e-07,
+      "loss": 0.4646,
+      "step": 17064
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 50.161283436558996,
+      "learning_rate": 6.44764135789211e-07,
+      "loss": 0.3812,
+      "step": 17065
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 36.1452360662682,
+      "learning_rate": 6.441685243775097e-07,
+      "loss": 0.3507,
+      "step": 17066
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 48.50121645532545,
+      "learning_rate": 6.435731790401833e-07,
+      "loss": 0.4168,
+      "step": 17067
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 42.005461496661184,
+      "learning_rate": 6.429780997941626e-07,
+      "loss": 0.4247,
+      "step": 17068
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 49.3046158022564,
+      "learning_rate": 6.423832866563696e-07,
+      "loss": 0.4423,
+      "step": 17069
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 46.92975443337857,
+      "learning_rate": 6.417887396437217e-07,
+      "loss": 0.4163,
+      "step": 17070
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 48.199012172833484,
+      "learning_rate": 6.411944587731279e-07,
+      "loss": 0.4221,
+      "step": 17071
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 52.716282391642565,
+      "learning_rate": 6.406004440614911e-07,
+      "loss": 0.3846,
+      "step": 17072
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 69.58886524832606,
+      "learning_rate": 6.400066955257023e-07,
+      "loss": 0.4554,
+      "step": 17073
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 47.17180735383265,
+      "learning_rate": 6.394132131826447e-07,
+      "loss": 0.4867,
+      "step": 17074
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 46.08571958773269,
+      "learning_rate": 6.388199970492015e-07,
+      "loss": 0.4371,
+      "step": 17075
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 43.500280943873136,
+      "learning_rate": 6.382270471422413e-07,
+      "loss": 0.4258,
+      "step": 17076
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 40.31053142620319,
+      "learning_rate": 6.376343634786264e-07,
+      "loss": 0.3324,
+      "step": 17077
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 34.75491151816814,
+      "learning_rate": 6.370419460752153e-07,
+      "loss": 0.372,
+      "step": 17078
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 48.15104933869827,
+      "learning_rate": 6.364497949488524e-07,
+      "loss": 0.4575,
+      "step": 17079
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 52.07899798952858,
+      "learning_rate": 6.35857910116382e-07,
+      "loss": 0.5258,
+      "step": 17080
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 48.952336217014896,
+      "learning_rate": 6.352662915946317e-07,
+      "loss": 0.4272,
+      "step": 17081
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 42.464571064415736,
+      "learning_rate": 6.346749394004314e-07,
+      "loss": 0.385,
+      "step": 17082
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 57.440558782744496,
+      "learning_rate": 6.340838535505978e-07,
+      "loss": 0.4708,
+      "step": 17083
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 58.467662105908914,
+      "learning_rate": 6.334930340619405e-07,
+      "loss": 0.4426,
+      "step": 17084
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 53.92242059069608,
+      "learning_rate": 6.329024809512607e-07,
+      "loss": 0.4442,
+      "step": 17085
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 47.827369611695076,
+      "learning_rate": 6.323121942353549e-07,
+      "loss": 0.4112,
+      "step": 17086
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 67.15075141332937,
+      "learning_rate": 6.317221739310109e-07,
+      "loss": 0.4257,
+      "step": 17087
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 41.97006023574444,
+      "learning_rate": 6.311324200550084e-07,
+      "loss": 0.3464,
+      "step": 17088
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 46.0790975008194,
+      "learning_rate": 6.305429326241208e-07,
+      "loss": 0.354,
+      "step": 17089
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 66.98158072179696,
+      "learning_rate": 6.29953711655108e-07,
+      "loss": 0.5085,
+      "step": 17090
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 52.407464078739764,
+      "learning_rate": 6.29364757164731e-07,
+      "loss": 0.4208,
+      "step": 17091
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 52.504937988751756,
+      "learning_rate": 6.287760691697409e-07,
+      "loss": 0.5012,
+      "step": 17092
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 48.713684175902074,
+      "learning_rate": 6.281876476868764e-07,
+      "loss": 0.396,
+      "step": 17093
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 64.48336929856978,
+      "learning_rate": 6.27599492732871e-07,
+      "loss": 0.4885,
+      "step": 17094
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 44.0957971228793,
+      "learning_rate": 6.270116043244545e-07,
+      "loss": 0.3421,
+      "step": 17095
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 51.34414711704022,
+      "learning_rate": 6.264239824783447e-07,
+      "loss": 0.4345,
+      "step": 17096
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 57.530195451361664,
+      "learning_rate": 6.258366272112537e-07,
+      "loss": 0.433,
+      "step": 17097
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 49.652559771055486,
+      "learning_rate": 6.252495385398871e-07,
+      "loss": 0.4225,
+      "step": 17098
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 41.34654659463444,
+      "learning_rate": 6.246627164809371e-07,
+      "loss": 0.3696,
+      "step": 17099
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 58.90494125895611,
+      "learning_rate": 6.240761610510948e-07,
+      "loss": 0.3965,
+      "step": 17100
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 40.96608593947896,
+      "learning_rate": 6.234898722670435e-07,
+      "loss": 0.402,
+      "step": 17101
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 58.39837825300088,
+      "learning_rate": 6.229038501454532e-07,
+      "loss": 0.4363,
+      "step": 17102
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 50.42065532238785,
+      "learning_rate": 6.223180947029939e-07,
+      "loss": 0.4246,
+      "step": 17103
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 49.39005326275329,
+      "learning_rate": 6.217326059563211e-07,
+      "loss": 0.4406,
+      "step": 17104
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 49.335791687134524,
+      "learning_rate": 6.211473839220883e-07,
+      "loss": 0.4168,
+      "step": 17105
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 45.08797051396025,
+      "learning_rate": 6.205624286169354e-07,
+      "loss": 0.4096,
+      "step": 17106
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 44.03161334503117,
+      "learning_rate": 6.199777400575013e-07,
+      "loss": 0.3691,
+      "step": 17107
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 45.540315551586,
+      "learning_rate": 6.193933182604128e-07,
+      "loss": 0.4532,
+      "step": 17108
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 51.20350201688833,
+      "learning_rate": 6.188091632422921e-07,
+      "loss": 0.4908,
+      "step": 17109
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 78.8719967060245,
+      "learning_rate": 6.182252750197493e-07,
+      "loss": 0.4625,
+      "step": 17110
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 46.130225172495955,
+      "learning_rate": 6.17641653609391e-07,
+      "loss": 0.4307,
+      "step": 17111
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 45.795172564367476,
+      "learning_rate": 6.170582990278173e-07,
+      "loss": 0.4209,
+      "step": 17112
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 57.44597220382494,
+      "learning_rate": 6.16475211291615e-07,
+      "loss": 0.426,
+      "step": 17113
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 39.16051334513191,
+      "learning_rate": 6.158923904173709e-07,
+      "loss": 0.3763,
+      "step": 17114
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 42.121533361583815,
+      "learning_rate": 6.153098364216548e-07,
+      "loss": 0.478,
+      "step": 17115
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 43.99047433629934,
+      "learning_rate": 6.147275493210381e-07,
+      "loss": 0.4483,
+      "step": 17116
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 45.22859376585219,
+      "learning_rate": 6.141455291320808e-07,
+      "loss": 0.4396,
+      "step": 17117
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 40.60920754046256,
+      "learning_rate": 6.135637758713342e-07,
+      "loss": 0.405,
+      "step": 17118
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 46.916503827482664,
+      "learning_rate": 6.129822895553417e-07,
+      "loss": 0.3606,
+      "step": 17119
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 51.69165355410182,
+      "learning_rate": 6.124010702006411e-07,
+      "loss": 0.4147,
+      "step": 17120
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 45.30411247760769,
+      "learning_rate": 6.118201178237626e-07,
+      "loss": 0.3943,
+      "step": 17121
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 32.841503095489706,
+      "learning_rate": 6.112394324412308e-07,
+      "loss": 0.3706,
+      "step": 17122
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 50.75428379000331,
+      "learning_rate": 6.106590140695545e-07,
+      "loss": 0.4094,
+      "step": 17123
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 43.1881197872488,
+      "learning_rate": 6.100788627252441e-07,
+      "loss": 0.369,
+      "step": 17124
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 55.51736298880919,
+      "learning_rate": 6.094989784247973e-07,
+      "loss": 0.3742,
+      "step": 17125
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 39.97801077051183,
+      "learning_rate": 6.089193611847066e-07,
+      "loss": 0.3816,
+      "step": 17126
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 46.70688379961058,
+      "learning_rate": 6.083400110214577e-07,
+      "loss": 0.409,
+      "step": 17127
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 38.976124320412524,
+      "learning_rate": 6.077609279515217e-07,
+      "loss": 0.3957,
+      "step": 17128
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.66182725328818,
+      "learning_rate": 6.071821119913713e-07,
+      "loss": 0.3687,
+      "step": 17129
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.177391124495145,
+      "learning_rate": 6.066035631574685e-07,
+      "loss": 0.405,
+      "step": 17130
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 54.39328558380678,
+      "learning_rate": 6.060252814662637e-07,
+      "loss": 0.4877,
+      "step": 17131
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 57.187450237982105,
+      "learning_rate": 6.05447266934206e-07,
+      "loss": 0.4537,
+      "step": 17132
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 53.79120837261878,
+      "learning_rate": 6.04869519577731e-07,
+      "loss": 0.4291,
+      "step": 17133
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 48.485477957441404,
+      "learning_rate": 6.042920394132712e-07,
+      "loss": 0.4383,
+      "step": 17134
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.0229156320919,
+      "learning_rate": 6.037148264572512e-07,
+      "loss": 0.3719,
+      "step": 17135
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 44.03091085071865,
+      "learning_rate": 6.031378807260823e-07,
+      "loss": 0.406,
+      "step": 17136
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 40.5786403612997,
+      "learning_rate": 6.025612022361771e-07,
+      "loss": 0.3686,
+      "step": 17137
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 50.94465408168819,
+      "learning_rate": 6.019847910039334e-07,
+      "loss": 0.4504,
+      "step": 17138
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.43189159104647,
+      "learning_rate": 6.014086470457448e-07,
+      "loss": 0.4477,
+      "step": 17139
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 52.68664743767671,
+      "learning_rate": 6.008327703779948e-07,
+      "loss": 0.5006,
+      "step": 17140
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.337595468719826,
+      "learning_rate": 6.002571610170627e-07,
+      "loss": 0.3742,
+      "step": 17141
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 49.892626078611165,
+      "learning_rate": 5.996818189793207e-07,
+      "loss": 0.4253,
+      "step": 17142
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 51.352027588013065,
+      "learning_rate": 5.991067442811272e-07,
+      "loss": 0.457,
+      "step": 17143
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 46.13512479072943,
+      "learning_rate": 5.985319369388376e-07,
+      "loss": 0.4185,
+      "step": 17144
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.9677300045039,
+      "learning_rate": 5.979573969688001e-07,
+      "loss": 0.4433,
+      "step": 17145
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 53.074866643190965,
+      "learning_rate": 5.973831243873551e-07,
+      "loss": 0.4238,
+      "step": 17146
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 48.70803802352854,
+      "learning_rate": 5.96809119210835e-07,
+      "loss": 0.4051,
+      "step": 17147
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 43.38438776756978,
+      "learning_rate": 5.96235381455561e-07,
+      "loss": 0.3429,
+      "step": 17148
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 54.88987287076105,
+      "learning_rate": 5.956619111378514e-07,
+      "loss": 0.455,
+      "step": 17149
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 48.74162981607278,
+      "learning_rate": 5.950887082740153e-07,
+      "loss": 0.4613,
+      "step": 17150
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 43.2184081377809,
+      "learning_rate": 5.945157728803563e-07,
+      "loss": 0.4197,
+      "step": 17151
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 43.967841813414914,
+      "learning_rate": 5.93943104973167e-07,
+      "loss": 0.3989,
+      "step": 17152
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.054882724945365,
+      "learning_rate": 5.93370704568732e-07,
+      "loss": 0.4504,
+      "step": 17153
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 62.02658213389211,
+      "learning_rate": 5.927985716833317e-07,
+      "loss": 0.3808,
+      "step": 17154
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 46.010963066483384,
+      "learning_rate": 5.922267063332376e-07,
+      "loss": 0.405,
+      "step": 17155
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 44.98572530571538,
+      "learning_rate": 5.916551085347134e-07,
+      "loss": 0.3891,
+      "step": 17156
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 49.65374679449797,
+      "learning_rate": 5.910837783040113e-07,
+      "loss": 0.4621,
+      "step": 17157
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 43.51413913018595,
+      "learning_rate": 5.905127156573842e-07,
+      "loss": 0.369,
+      "step": 17158
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.811168328637905,
+      "learning_rate": 5.899419206110702e-07,
+      "loss": 0.4212,
+      "step": 17159
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 41.533523284948195,
+      "learning_rate": 5.89371393181305e-07,
+      "loss": 0.3821,
+      "step": 17160
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 51.23788279594901,
+      "learning_rate": 5.888011333843113e-07,
+      "loss": 0.4717,
+      "step": 17161
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 41.81858968366213,
+      "learning_rate": 5.882311412363073e-07,
+      "loss": 0.42,
+      "step": 17162
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 44.85554898912694,
+      "learning_rate": 5.876614167535044e-07,
+      "loss": 0.4189,
+      "step": 17163
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 54.96086611441482,
+      "learning_rate": 5.870919599521052e-07,
+      "loss": 0.4364,
+      "step": 17164
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 53.87449890045292,
+      "learning_rate": 5.865227708483034e-07,
+      "loss": 0.4294,
+      "step": 17165
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 49.693603035248074,
+      "learning_rate": 5.859538494582895e-07,
+      "loss": 0.4023,
+      "step": 17166
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 50.48929816337923,
+      "learning_rate": 5.853851957982381e-07,
+      "loss": 0.4039,
+      "step": 17167
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 48.10355744678932,
+      "learning_rate": 5.848168098843265e-07,
+      "loss": 0.4482,
+      "step": 17168
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 54.695585278780435,
+      "learning_rate": 5.842486917327162e-07,
+      "loss": 0.4589,
+      "step": 17169
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 54.44342896050766,
+      "learning_rate": 5.836808413595641e-07,
+      "loss": 0.3972,
+      "step": 17170
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 51.68105618256802,
+      "learning_rate": 5.831132587810228e-07,
+      "loss": 0.476,
+      "step": 17171
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 49.336550215709195,
+      "learning_rate": 5.825459440132308e-07,
+      "loss": 0.4196,
+      "step": 17172
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 53.43717021728837,
+      "learning_rate": 5.819788970723217e-07,
+      "loss": 0.4881,
+      "step": 17173
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 61.23402367695729,
+      "learning_rate": 5.814121179744248e-07,
+      "loss": 0.5277,
+      "step": 17174
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.05340254717257,
+      "learning_rate": 5.80845606735656e-07,
+      "loss": 0.4055,
+      "step": 17175
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 46.47784503879894,
+      "learning_rate": 5.802793633721304e-07,
+      "loss": 0.4138,
+      "step": 17176
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 44.968845242736776,
+      "learning_rate": 5.797133878999484e-07,
+      "loss": 0.3864,
+      "step": 17177
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.11291216829142,
+      "learning_rate": 5.791476803352058e-07,
+      "loss": 0.4065,
+      "step": 17178
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 49.232082290558786,
+      "learning_rate": 5.785822406939934e-07,
+      "loss": 0.4351,
+      "step": 17179
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 52.115035666678914,
+      "learning_rate": 5.780170689923902e-07,
+      "loss": 0.4302,
+      "step": 17180
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 49.11796244717876,
+      "learning_rate": 5.774521652464715e-07,
+      "loss": 0.4828,
+      "step": 17181
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.919372438523084,
+      "learning_rate": 5.768875294722987e-07,
+      "loss": 0.4056,
+      "step": 17182
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 46.347392182817615,
+      "learning_rate": 5.763231616859333e-07,
+      "loss": 0.4209,
+      "step": 17183
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 69.14338301335643,
+      "learning_rate": 5.757590619034236e-07,
+      "loss": 0.5001,
+      "step": 17184
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.64374747948739,
+      "learning_rate": 5.751952301408148e-07,
+      "loss": 0.3924,
+      "step": 17185
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 48.92514199697778,
+      "learning_rate": 5.746316664141394e-07,
+      "loss": 0.4177,
+      "step": 17186
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 50.46021909626112,
+      "learning_rate": 5.740683707394256e-07,
+      "loss": 0.5281,
+      "step": 17187
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.24306586388976,
+      "learning_rate": 5.735053431326931e-07,
+      "loss": 0.3937,
+      "step": 17188
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.419443576577784,
+      "learning_rate": 5.729425836099556e-07,
+      "loss": 0.4567,
+      "step": 17189
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 57.090042720504066,
+      "learning_rate": 5.723800921872147e-07,
+      "loss": 0.5099,
+      "step": 17190
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 42.74619757043487,
+      "learning_rate": 5.718178688804699e-07,
+      "loss": 0.3956,
+      "step": 17191
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 46.7659101952276,
+      "learning_rate": 5.712559137057094e-07,
+      "loss": 0.3987,
+      "step": 17192
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 51.618417278621145,
+      "learning_rate": 5.706942266789162e-07,
+      "loss": 0.4618,
+      "step": 17193
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 56.685789198121384,
+      "learning_rate": 5.701328078160606e-07,
+      "loss": 0.5275,
+      "step": 17194
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 48.90228295857606,
+      "learning_rate": 5.695716571331134e-07,
+      "loss": 0.446,
+      "step": 17195
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.1904695614556,
+      "learning_rate": 5.690107746460316e-07,
+      "loss": 0.4492,
+      "step": 17196
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 50.61570735462374,
+      "learning_rate": 5.684501603707671e-07,
+      "loss": 0.4146,
+      "step": 17197
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.39951547202007,
+      "learning_rate": 5.678898143232614e-07,
+      "loss": 0.4175,
+      "step": 17198
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 37.514063671181475,
+      "learning_rate": 5.673297365194508e-07,
+      "loss": 0.3609,
+      "step": 17199
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 39.940660372534026,
+      "learning_rate": 5.667699269752658e-07,
+      "loss": 0.3866,
+      "step": 17200
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 43.81606754900866,
+      "learning_rate": 5.66210385706627e-07,
+      "loss": 0.4586,
+      "step": 17201
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.33724717317658,
+      "learning_rate": 5.65651112729445e-07,
+      "loss": 0.4448,
+      "step": 17202
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 35.97218038604804,
+      "learning_rate": 5.650921080596261e-07,
+      "loss": 0.3745,
+      "step": 17203
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 45.341089603084725,
+      "learning_rate": 5.645333717130685e-07,
+      "loss": 0.3772,
+      "step": 17204
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 49.83975973868897,
+      "learning_rate": 5.63974903705663e-07,
+      "loss": 0.4395,
+      "step": 17205
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 57.187482698364974,
+      "learning_rate": 5.634167040532922e-07,
+      "loss": 0.5189,
+      "step": 17206
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 49.23513993347278,
+      "learning_rate": 5.628587727718282e-07,
+      "loss": 0.4453,
+      "step": 17207
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 43.784107101412594,
+      "learning_rate": 5.623011098771391e-07,
+      "loss": 0.3799,
+      "step": 17208
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 62.97529554399935,
+      "learning_rate": 5.617437153850868e-07,
+      "loss": 0.4417,
+      "step": 17209
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 53.374553924233126,
+      "learning_rate": 5.611865893115243e-07,
+      "loss": 0.3567,
+      "step": 17210
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.67815489350929,
+      "learning_rate": 5.60629731672292e-07,
+      "loss": 0.4232,
+      "step": 17211
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 45.003157886328616,
+      "learning_rate": 5.600731424832273e-07,
+      "loss": 0.4669,
+      "step": 17212
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.31492218509432,
+      "learning_rate": 5.595168217601599e-07,
+      "loss": 0.4094,
+      "step": 17213
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 48.49225468582507,
+      "learning_rate": 5.589607695189136e-07,
+      "loss": 0.4141,
+      "step": 17214
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 46.529180266781786,
+      "learning_rate": 5.584049857752993e-07,
+      "loss": 0.3579,
+      "step": 17215
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 43.97088197095563,
+      "learning_rate": 5.578494705451232e-07,
+      "loss": 0.387,
+      "step": 17216
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 52.92544074670115,
+      "learning_rate": 5.572942238441847e-07,
+      "loss": 0.4037,
+      "step": 17217
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 52.546375046551375,
+      "learning_rate": 5.567392456882758e-07,
+      "loss": 0.4151,
+      "step": 17218
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 44.13819240652482,
+      "learning_rate": 5.561845360931784e-07,
+      "loss": 0.3605,
+      "step": 17219
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.17146375052759,
+      "learning_rate": 5.556300950746684e-07,
+      "loss": 0.4655,
+      "step": 17220
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 40.95183891014762,
+      "learning_rate": 5.550759226485126e-07,
+      "loss": 0.3936,
+      "step": 17221
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 45.42691867620375,
+      "learning_rate": 5.545220188304723e-07,
+      "loss": 0.3814,
+      "step": 17222
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.077378452888176,
+      "learning_rate": 5.539683836363019e-07,
+      "loss": 0.442,
+      "step": 17223
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 43.018267759928584,
+      "learning_rate": 5.534150170817431e-07,
+      "loss": 0.3641,
+      "step": 17224
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 44.62611028214348,
+      "learning_rate": 5.52861919182538e-07,
+      "loss": 0.4245,
+      "step": 17225
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 55.793035276272214,
+      "learning_rate": 5.523090899544104e-07,
+      "loss": 0.4517,
+      "step": 17226
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.657301659823744,
+      "learning_rate": 5.517565294130877e-07,
+      "loss": 0.4964,
+      "step": 17227
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.534787118534766,
+      "learning_rate": 5.51204237574281e-07,
+      "loss": 0.4481,
+      "step": 17228
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 49.2790982073552,
+      "learning_rate": 5.506522144536975e-07,
+      "loss": 0.4438,
+      "step": 17229
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.33338679838966,
+      "learning_rate": 5.501004600670401e-07,
+      "loss": 0.3773,
+      "step": 17230
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.54036371973597,
+      "learning_rate": 5.495489744299965e-07,
+      "loss": 0.4287,
+      "step": 17231
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 51.438682459128984,
+      "learning_rate": 5.489977575582506e-07,
+      "loss": 0.4471,
+      "step": 17232
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 52.92286118071982,
+      "learning_rate": 5.4844680946748e-07,
+      "loss": 0.4747,
+      "step": 17233
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 49.543137110113,
+      "learning_rate": 5.478961301733531e-07,
+      "loss": 0.4447,
+      "step": 17234
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 46.619670783238384,
+      "learning_rate": 5.473457196915332e-07,
+      "loss": 0.3722,
+      "step": 17235
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 41.44288800452988,
+      "learning_rate": 5.467955780376688e-07,
+      "loss": 0.4087,
+      "step": 17236
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 56.57182687397966,
+      "learning_rate": 5.462457052274084e-07,
+      "loss": 0.3891,
+      "step": 17237
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 45.15914144267513,
+      "learning_rate": 5.456961012763906e-07,
+      "loss": 0.4972,
+      "step": 17238
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 52.68404616939951,
+      "learning_rate": 5.451467662002452e-07,
+      "loss": 0.4842,
+      "step": 17239
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 49.1257672140766,
+      "learning_rate": 5.445977000145952e-07,
+      "loss": 0.3873,
+      "step": 17240
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.527969307324355,
+      "learning_rate": 5.440489027350548e-07,
+      "loss": 0.3568,
+      "step": 17241
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 40.231881710685485,
+      "learning_rate": 5.435003743772305e-07,
+      "loss": 0.4007,
+      "step": 17242
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 40.253000615504284,
+      "learning_rate": 5.429521149567263e-07,
+      "loss": 0.3801,
+      "step": 17243
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.5463343134286,
+      "learning_rate": 5.424041244891298e-07,
+      "loss": 0.4888,
+      "step": 17244
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 51.097602992937354,
+      "learning_rate": 5.418564029900286e-07,
+      "loss": 0.4483,
+      "step": 17245
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 43.4464119914578,
+      "learning_rate": 5.413089504749979e-07,
+      "loss": 0.4453,
+      "step": 17246
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 46.23763882692755,
+      "learning_rate": 5.407617669596066e-07,
+      "loss": 0.4164,
+      "step": 17247
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 63.331664864831964,
+      "learning_rate": 5.402148524594198e-07,
+      "loss": 0.4454,
+      "step": 17248
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 52.38505063548858,
+      "learning_rate": 5.396682069899861e-07,
+      "loss": 0.4815,
+      "step": 17249
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.52647131097713,
+      "learning_rate": 5.391218305668566e-07,
+      "loss": 0.38,
+      "step": 17250
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 52.127562892122675,
+      "learning_rate": 5.385757232055655e-07,
+      "loss": 0.4143,
+      "step": 17251
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 54.60754426826283,
+      "learning_rate": 5.380298849216481e-07,
+      "loss": 0.4622,
+      "step": 17252
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 49.56599107066827,
+      "learning_rate": 5.374843157306253e-07,
+      "loss": 0.4102,
+      "step": 17253
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 50.77651804312743,
+      "learning_rate": 5.369390156480126e-07,
+      "loss": 0.3958,
+      "step": 17254
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 48.446641608173515,
+      "learning_rate": 5.363939846893184e-07,
+      "loss": 0.4114,
+      "step": 17255
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 42.06506134102186,
+      "learning_rate": 5.35849222870044e-07,
+      "loss": 0.3697,
+      "step": 17256
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 48.65388772628669,
+      "learning_rate": 5.353047302056802e-07,
+      "loss": 0.3543,
+      "step": 17257
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 47.32621077667825,
+      "learning_rate": 5.347605067117134e-07,
+      "loss": 0.3949,
+      "step": 17258
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.013676252895046,
+      "learning_rate": 5.342165524036224e-07,
+      "loss": 0.4472,
+      "step": 17259
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 45.12786859286672,
+      "learning_rate": 5.33672867296875e-07,
+      "loss": 0.4123,
+      "step": 17260
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 44.409708794393026,
+      "learning_rate": 5.331294514069318e-07,
+      "loss": 0.4154,
+      "step": 17261
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 50.12690052986172,
+      "learning_rate": 5.325863047492496e-07,
+      "loss": 0.4139,
+      "step": 17262
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 55.07625517139976,
+      "learning_rate": 5.320434273392738e-07,
+      "loss": 0.4832,
+      "step": 17263
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 39.21404587699373,
+      "learning_rate": 5.315008191924464e-07,
+      "loss": 0.3898,
+      "step": 17264
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.86673392195048,
+      "learning_rate": 5.309584803241973e-07,
+      "loss": 0.4592,
+      "step": 17265
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 46.51499261961199,
+      "learning_rate": 5.304164107499477e-07,
+      "loss": 0.3776,
+      "step": 17266
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 52.94194635411127,
+      "learning_rate": 5.29874610485116e-07,
+      "loss": 0.4164,
+      "step": 17267
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 41.85361271034803,
+      "learning_rate": 5.293330795451112e-07,
+      "loss": 0.4022,
+      "step": 17268
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 42.44771126664281,
+      "learning_rate": 5.287918179453344e-07,
+      "loss": 0.4266,
+      "step": 17269
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 44.15139413362699,
+      "learning_rate": 5.282508257011764e-07,
+      "loss": 0.4094,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 48.49870738047149,
+      "learning_rate": 5.27710102828024e-07,
+      "loss": 0.3954,
+      "step": 17271
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 40.10166616926202,
+      "learning_rate": 5.271696493412548e-07,
+      "loss": 0.3844,
+      "step": 17272
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 45.469560617254,
+      "learning_rate": 5.266294652562409e-07,
+      "loss": 0.4067,
+      "step": 17273
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 38.164921583582114,
+      "learning_rate": 5.260895505883423e-07,
+      "loss": 0.3667,
+      "step": 17274
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 42.59026038021972,
+      "learning_rate": 5.255499053529145e-07,
+      "loss": 0.372,
+      "step": 17275
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 46.63191912244746,
+      "learning_rate": 5.250105295653052e-07,
+      "loss": 0.4583,
+      "step": 17276
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 62.42759038663128,
+      "learning_rate": 5.244714232408544e-07,
+      "loss": 0.5304,
+      "step": 17277
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 56.34318869718982,
+      "learning_rate": 5.23932586394893e-07,
+      "loss": 0.4825,
+      "step": 17278
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 44.101605765200695,
+      "learning_rate": 5.233940190427456e-07,
+      "loss": 0.4068,
+      "step": 17279
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 63.70786858516587,
+      "learning_rate": 5.228557211997276e-07,
+      "loss": 0.4224,
+      "step": 17280
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.957735111226086,
+      "learning_rate": 5.223176928811502e-07,
+      "loss": 0.406,
+      "step": 17281
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 57.72844571237351,
+      "learning_rate": 5.217799341023122e-07,
+      "loss": 0.4581,
+      "step": 17282
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.10488597593285,
+      "learning_rate": 5.212424448785092e-07,
+      "loss": 0.4448,
+      "step": 17283
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 42.05131518090978,
+      "learning_rate": 5.207052252250266e-07,
+      "loss": 0.3818,
+      "step": 17284
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.8080709747835,
+      "learning_rate": 5.201682751571402e-07,
+      "loss": 0.4398,
+      "step": 17285
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.85405540002229,
+      "learning_rate": 5.196315946901254e-07,
+      "loss": 0.4247,
+      "step": 17286
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 55.03496442365116,
+      "learning_rate": 5.1909518383924e-07,
+      "loss": 0.4655,
+      "step": 17287
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 47.7643180720782,
+      "learning_rate": 5.185590426197406e-07,
+      "loss": 0.4465,
+      "step": 17288
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.99305299658608,
+      "learning_rate": 5.180231710468775e-07,
+      "loss": 0.4196,
+      "step": 17289
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 54.594540621399325,
+      "learning_rate": 5.174875691358894e-07,
+      "loss": 0.4938,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 55.231187543394476,
+      "learning_rate": 5.169522369020052e-07,
+      "loss": 0.5155,
+      "step": 17291
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.850035736851716,
+      "learning_rate": 5.164171743604529e-07,
+      "loss": 0.4077,
+      "step": 17292
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 50.91763530544468,
+      "learning_rate": 5.158823815264491e-07,
+      "loss": 0.4326,
+      "step": 17293
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 56.407364895018084,
+      "learning_rate": 5.153478584152028e-07,
+      "loss": 0.5864,
+      "step": 17294
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 41.897721921268285,
+      "learning_rate": 5.148136050419139e-07,
+      "loss": 0.3986,
+      "step": 17295
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 37.156677455384354,
+      "learning_rate": 5.14279621421776e-07,
+      "loss": 0.3541,
+      "step": 17296
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 45.61632417207164,
+      "learning_rate": 5.13745907569978e-07,
+      "loss": 0.4158,
+      "step": 17297
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 36.899504456126046,
+      "learning_rate": 5.132124635016977e-07,
+      "loss": 0.3492,
+      "step": 17298
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.90695405868331,
+      "learning_rate": 5.126792892321064e-07,
+      "loss": 0.3917,
+      "step": 17299
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 57.771352573248166,
+      "learning_rate": 5.121463847763641e-07,
+      "loss": 0.4912,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.31025987816175,
+      "learning_rate": 5.116137501496288e-07,
+      "loss": 0.4674,
+      "step": 17301
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.78751658085714,
+      "learning_rate": 5.110813853670482e-07,
+      "loss": 0.3955,
+      "step": 17302
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.183032603587876,
+      "learning_rate": 5.105492904437636e-07,
+      "loss": 0.4393,
+      "step": 17303
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 56.15406485558682,
+      "learning_rate": 5.100174653949031e-07,
+      "loss": 0.4399,
+      "step": 17304
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 50.67723026496102,
+      "learning_rate": 5.094859102355953e-07,
+      "loss": 0.4325,
+      "step": 17305
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 52.682270795133206,
+      "learning_rate": 5.089546249809584e-07,
+      "loss": 0.4636,
+      "step": 17306
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 42.47556562378272,
+      "learning_rate": 5.08423609646097e-07,
+      "loss": 0.3986,
+      "step": 17307
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 36.571141716630905,
+      "learning_rate": 5.078928642461178e-07,
+      "loss": 0.3207,
+      "step": 17308
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 51.51549933351387,
+      "learning_rate": 5.073623887961121e-07,
+      "loss": 0.4083,
+      "step": 17309
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 49.436994345064726,
+      "learning_rate": 5.068321833111667e-07,
+      "loss": 0.4383,
+      "step": 17310
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 46.94089095036069,
+      "learning_rate": 5.063022478063617e-07,
+      "loss": 0.4178,
+      "step": 17311
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 56.1272387748363,
+      "learning_rate": 5.057725822967663e-07,
+      "loss": 0.4597,
+      "step": 17312
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 57.48772927255318,
+      "learning_rate": 5.05243186797445e-07,
+      "loss": 0.4857,
+      "step": 17313
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 43.659499681636184,
+      "learning_rate": 5.047140613234524e-07,
+      "loss": 0.3555,
+      "step": 17314
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 58.318297115387104,
+      "learning_rate": 5.041852058898389e-07,
+      "loss": 0.4883,
+      "step": 17315
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 45.45169068266972,
+      "learning_rate": 5.036566205116422e-07,
+      "loss": 0.4344,
+      "step": 17316
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 62.28926875454562,
+      "learning_rate": 5.031283052038949e-07,
+      "loss": 0.4867,
+      "step": 17317
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.08026382658296,
+      "learning_rate": 5.026002599816248e-07,
+      "loss": 0.3797,
+      "step": 17318
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 55.928497736606865,
+      "learning_rate": 5.02072484859848e-07,
+      "loss": 0.4338,
+      "step": 17319
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 50.9298419414624,
+      "learning_rate": 5.015449798535721e-07,
+      "loss": 0.4245,
+      "step": 17320
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 52.11387513273164,
+      "learning_rate": 5.010177449778009e-07,
+      "loss": 0.4468,
+      "step": 17321
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 56.2247535718824,
+      "learning_rate": 5.004907802475278e-07,
+      "loss": 0.4102,
+      "step": 17322
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 47.372033459123884,
+      "learning_rate": 4.999640856777421e-07,
+      "loss": 0.4157,
+      "step": 17323
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 50.55541521944869,
+      "learning_rate": 4.994376612834185e-07,
+      "loss": 0.5061,
+      "step": 17324
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 42.47239492015491,
+      "learning_rate": 4.989115070795303e-07,
+      "loss": 0.4029,
+      "step": 17325
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 43.01609583332684,
+      "learning_rate": 4.983856230810402e-07,
+      "loss": 0.3939,
+      "step": 17326
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 39.380855579032655,
+      "learning_rate": 4.978600093029062e-07,
+      "loss": 0.3556,
+      "step": 17327
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 47.356759346796345,
+      "learning_rate": 4.973346657600752e-07,
+      "loss": 0.4114,
+      "step": 17328
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.47624808072564,
+      "learning_rate": 4.968095924674854e-07,
+      "loss": 0.5179,
+      "step": 17329
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 48.74116867475402,
+      "learning_rate": 4.962847894400724e-07,
+      "loss": 0.4021,
+      "step": 17330
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.57760354389079,
+      "learning_rate": 4.95760256692761e-07,
+      "loss": 0.4799,
+      "step": 17331
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 42.703204878262326,
+      "learning_rate": 4.952359942404672e-07,
+      "loss": 0.3858,
+      "step": 17332
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 43.72663083598495,
+      "learning_rate": 4.947120020981034e-07,
+      "loss": 0.3793,
+      "step": 17333
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 53.058120708579764,
+      "learning_rate": 4.941882802805675e-07,
+      "loss": 0.4133,
+      "step": 17334
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 45.724068166123836,
+      "learning_rate": 4.936648288027568e-07,
+      "loss": 0.4152,
+      "step": 17335
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 41.23513538834692,
+      "learning_rate": 4.931416476795592e-07,
+      "loss": 0.3671,
+      "step": 17336
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 48.42850303032197,
+      "learning_rate": 4.926187369258495e-07,
+      "loss": 0.4138,
+      "step": 17337
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 44.55792598068491,
+      "learning_rate": 4.920960965565036e-07,
+      "loss": 0.4097,
+      "step": 17338
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 63.101435309921044,
+      "learning_rate": 4.915737265863807e-07,
+      "loss": 0.5115,
+      "step": 17339
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 53.40774619750486,
+      "learning_rate": 4.9105162703034e-07,
+      "loss": 0.4386,
+      "step": 17340
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 38.10219372431193,
+      "learning_rate": 4.905297979032264e-07,
+      "loss": 0.3325,
+      "step": 17341
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 39.722125374439564,
+      "learning_rate": 4.900082392198835e-07,
+      "loss": 0.3683,
+      "step": 17342
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 50.68788246076345,
+      "learning_rate": 4.894869509951449e-07,
+      "loss": 0.3963,
+      "step": 17343
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 47.01254719795469,
+      "learning_rate": 4.889659332438334e-07,
+      "loss": 0.4178,
+      "step": 17344
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 51.15904560501529,
+      "learning_rate": 4.884451859807648e-07,
+      "loss": 0.395,
+      "step": 17345
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 40.77138280067812,
+      "learning_rate": 4.879247092207518e-07,
+      "loss": 0.4351,
+      "step": 17346
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 62.274877311190586,
+      "learning_rate": 4.874045029785957e-07,
+      "loss": 0.4849,
+      "step": 17347
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.77765020129317,
+      "learning_rate": 4.868845672690937e-07,
+      "loss": 0.406,
+      "step": 17348
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 46.30397068406596,
+      "learning_rate": 4.863649021070261e-07,
+      "loss": 0.4025,
+      "step": 17349
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 43.49006108491756,
+      "learning_rate": 4.858455075071766e-07,
+      "loss": 0.4397,
+      "step": 17350
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 42.00663931345104,
+      "learning_rate": 4.853263834843136e-07,
+      "loss": 0.4237,
+      "step": 17351
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 46.48886215991968,
+      "learning_rate": 4.848075300532051e-07,
+      "loss": 0.4332,
+      "step": 17352
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 43.64626638824315,
+      "learning_rate": 4.842889472286039e-07,
+      "loss": 0.3541,
+      "step": 17353
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 51.02974704320916,
+      "learning_rate": 4.837706350252569e-07,
+      "loss": 0.4325,
+      "step": 17354
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.99319624386088,
+      "learning_rate": 4.832525934579058e-07,
+      "loss": 0.4366,
+      "step": 17355
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 51.019305009123435,
+      "learning_rate": 4.827348225412864e-07,
+      "loss": 0.4463,
+      "step": 17356
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 40.4115424824159,
+      "learning_rate": 4.822173222901194e-07,
+      "loss": 0.376,
+      "step": 17357
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 45.66869527193024,
+      "learning_rate": 4.81700092719124e-07,
+      "loss": 0.4739,
+      "step": 17358
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 48.08783706644084,
+      "learning_rate": 4.811831338430095e-07,
+      "loss": 0.3859,
+      "step": 17359
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 45.53769059666201,
+      "learning_rate": 4.806664456764787e-07,
+      "loss": 0.329,
+      "step": 17360
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 42.415585675857564,
+      "learning_rate": 4.801500282342264e-07,
+      "loss": 0.4069,
+      "step": 17361
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 45.32159956492167,
+      "learning_rate": 4.796338815309387e-07,
+      "loss": 0.4148,
+      "step": 17362
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 51.24358701357818,
+      "learning_rate": 4.791180055812928e-07,
+      "loss": 0.4512,
+      "step": 17363
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.64918812110065,
+      "learning_rate": 4.786024003999612e-07,
+      "loss": 0.3616,
+      "step": 17364
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 58.55605878795683,
+      "learning_rate": 4.780870660016091e-07,
+      "loss": 0.4362,
+      "step": 17365
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 58.36725911067246,
+      "learning_rate": 4.77572002400889e-07,
+      "loss": 0.5372,
+      "step": 17366
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 38.60041675824422,
+      "learning_rate": 4.770572096124515e-07,
+      "loss": 0.3411,
+      "step": 17367
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 61.27719817123294,
+      "learning_rate": 4.7654268765093604e-07,
+      "loss": 0.4351,
+      "step": 17368
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 46.71221018663782,
+      "learning_rate": 4.760284365309753e-07,
+      "loss": 0.406,
+      "step": 17369
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 46.764621661403574,
+      "learning_rate": 4.755144562671943e-07,
+      "loss": 0.3976,
+      "step": 17370
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 49.11148815360311,
+      "learning_rate": 4.7500074687421017e-07,
+      "loss": 0.3866,
+      "step": 17371
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 42.56089412219667,
+      "learning_rate": 4.744873083666346e-07,
+      "loss": 0.4227,
+      "step": 17372
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 67.26180236091648,
+      "learning_rate": 4.739741407590659e-07,
+      "loss": 0.4719,
+      "step": 17373
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 38.53467707309464,
+      "learning_rate": 4.7346124406610125e-07,
+      "loss": 0.3604,
+      "step": 17374
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 47.00173149064558,
+      "learning_rate": 4.729486183023246e-07,
+      "loss": 0.4465,
+      "step": 17375
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 46.02247405448123,
+      "learning_rate": 4.724362634823165e-07,
+      "loss": 0.3708,
+      "step": 17376
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 48.39233466858213,
+      "learning_rate": 4.7192417962064865e-07,
+      "loss": 0.3904,
+      "step": 17377
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 45.31096643417815,
+      "learning_rate": 4.714123667318837e-07,
+      "loss": 0.4496,
+      "step": 17378
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 44.089507227999306,
+      "learning_rate": 4.7090082483057577e-07,
+      "loss": 0.3804,
+      "step": 17379
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 54.30878121448564,
+      "learning_rate": 4.7038955393127306e-07,
+      "loss": 0.5906,
+      "step": 17380
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 48.2472706749173,
+      "learning_rate": 4.6987855404851955e-07,
+      "loss": 0.4172,
+      "step": 17381
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.613099479145056,
+      "learning_rate": 4.6936782519684365e-07,
+      "loss": 0.4512,
+      "step": 17382
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 56.26231353345786,
+      "learning_rate": 4.6885736739077146e-07,
+      "loss": 0.4421,
+      "step": 17383
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 45.71912134088823,
+      "learning_rate": 4.683471806448192e-07,
+      "loss": 0.3457,
+      "step": 17384
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 44.25079624623438,
+      "learning_rate": 4.6783726497349747e-07,
+      "loss": 0.4136,
+      "step": 17385
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 55.771250963933305,
+      "learning_rate": 4.673276203913091e-07,
+      "loss": 0.4233,
+      "step": 17386
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 59.8342505263406,
+      "learning_rate": 4.668182469127469e-07,
+      "loss": 0.5731,
+      "step": 17387
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.751973141120395,
+      "learning_rate": 4.6630914455229493e-07,
+      "loss": 0.4238,
+      "step": 17388
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 56.14172776262175,
+      "learning_rate": 4.6580031332443487e-07,
+      "loss": 0.4483,
+      "step": 17389
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.82394272700635,
+      "learning_rate": 4.652917532436374e-07,
+      "loss": 0.4235,
+      "step": 17390
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 53.36242045121598,
+      "learning_rate": 4.6478346432436426e-07,
+      "loss": 0.3685,
+      "step": 17391
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 38.99456208171118,
+      "learning_rate": 4.642754465810717e-07,
+      "loss": 0.3384,
+      "step": 17392
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 43.91145658743777,
+      "learning_rate": 4.6376770002820593e-07,
+      "loss": 0.3886,
+      "step": 17393
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 42.29971704704033,
+      "learning_rate": 4.632602246802109e-07,
+      "loss": 0.4343,
+      "step": 17394
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 44.055101669050465,
+      "learning_rate": 4.6275302055151293e-07,
+      "loss": 0.3777,
+      "step": 17395
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 53.0500352909715,
+      "learning_rate": 4.6224608765654153e-07,
+      "loss": 0.4944,
+      "step": 17396
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 53.0991391415688,
+      "learning_rate": 4.61739426009713e-07,
+      "loss": 0.4343,
+      "step": 17397
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 44.33417793176887,
+      "learning_rate": 4.612330356254335e-07,
+      "loss": 0.3833,
+      "step": 17398
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 54.95474317140621,
+      "learning_rate": 4.607269165181083e-07,
+      "loss": 0.4625,
+      "step": 17399
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 49.740753574951704,
+      "learning_rate": 4.6022106870212804e-07,
+      "loss": 0.386,
+      "step": 17400
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.44473848108491,
+      "learning_rate": 4.597154921918812e-07,
+      "loss": 0.4008,
+      "step": 17401
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 44.11485179232446,
+      "learning_rate": 4.592101870017429e-07,
+      "loss": 0.4082,
+      "step": 17402
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 45.2042952912494,
+      "learning_rate": 4.587051531460873e-07,
+      "loss": 0.4099,
+      "step": 17403
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.854383563452345,
+      "learning_rate": 4.5820039063927514e-07,
+      "loss": 0.4134,
+      "step": 17404
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 42.94995124889622,
+      "learning_rate": 4.5769589949566153e-07,
+      "loss": 0.3754,
+      "step": 17405
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.06526108519022,
+      "learning_rate": 4.57191679729595e-07,
+      "loss": 0.3694,
+      "step": 17406
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.82415986785569,
+      "learning_rate": 4.566877313554152e-07,
+      "loss": 0.4198,
+      "step": 17407
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 65.89091732981592,
+      "learning_rate": 4.561840543874529e-07,
+      "loss": 0.4838,
+      "step": 17408
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 56.11813900478264,
+      "learning_rate": 4.5568064884003337e-07,
+      "loss": 0.4,
+      "step": 17409
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.92890505569473,
+      "learning_rate": 4.551775147274717e-07,
+      "loss": 0.4037,
+      "step": 17410
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.86236336322604,
+      "learning_rate": 4.5467465206408103e-07,
+      "loss": 0.389,
+      "step": 17411
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 53.61672067195807,
+      "learning_rate": 4.541720608641575e-07,
+      "loss": 0.4921,
+      "step": 17412
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.67398142666636,
+      "learning_rate": 4.5366974114199546e-07,
+      "loss": 0.4439,
+      "step": 17413
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 53.86476820505384,
+      "learning_rate": 4.5316769291188223e-07,
+      "loss": 0.4849,
+      "step": 17414
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 47.16296344997489,
+      "learning_rate": 4.526659161880964e-07,
+      "loss": 0.4172,
+      "step": 17415
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 45.52304930134366,
+      "learning_rate": 4.521644109849066e-07,
+      "loss": 0.3909,
+      "step": 17416
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.57580444237415,
+      "learning_rate": 4.5166317731657363e-07,
+      "loss": 0.4428,
+      "step": 17417
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 52.44609582260043,
+      "learning_rate": 4.5116221519735493e-07,
+      "loss": 0.41,
+      "step": 17418
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.29948915483761,
+      "learning_rate": 4.5066152464149914e-07,
+      "loss": 0.4677,
+      "step": 17419
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 43.53380212470551,
+      "learning_rate": 4.5016110566324044e-07,
+      "loss": 0.3951,
+      "step": 17420
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 43.484002312212134,
+      "learning_rate": 4.4966095827681524e-07,
+      "loss": 0.4266,
+      "step": 17421
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 49.91166982440418,
+      "learning_rate": 4.491610824964454e-07,
+      "loss": 0.4384,
+      "step": 17422
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 45.01504450022875,
+      "learning_rate": 4.486614783363463e-07,
+      "loss": 0.3552,
+      "step": 17423
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 48.32816900528176,
+      "learning_rate": 4.481621458107288e-07,
+      "loss": 0.4445,
+      "step": 17424
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 44.73260981694284,
+      "learning_rate": 4.476630849337904e-07,
+      "loss": 0.3992,
+      "step": 17425
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 48.40502971000063,
+      "learning_rate": 4.471642957197275e-07,
+      "loss": 0.3801,
+      "step": 17426
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 53.454667288512205,
+      "learning_rate": 4.4666577818272327e-07,
+      "loss": 0.4083,
+      "step": 17427
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 43.72932077241047,
+      "learning_rate": 4.461675323369563e-07,
+      "loss": 0.3721,
+      "step": 17428
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 47.95822720995092,
+      "learning_rate": 4.456695581965942e-07,
+      "loss": 0.4895,
+      "step": 17429
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 41.52267048253839,
+      "learning_rate": 4.4517185577580226e-07,
+      "loss": 0.422,
+      "step": 17430
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 61.383557292182005,
+      "learning_rate": 4.446744250887336e-07,
+      "loss": 0.5132,
+      "step": 17431
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 43.22131090720729,
+      "learning_rate": 4.441772661495347e-07,
+      "loss": 0.4214,
+      "step": 17432
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 57.9331517485936,
+      "learning_rate": 4.4368037897234317e-07,
+      "loss": 0.4064,
+      "step": 17433
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 47.72490586980427,
+      "learning_rate": 4.4318376357129103e-07,
+      "loss": 0.3816,
+      "step": 17434
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 59.92702948770459,
+      "learning_rate": 4.4268741996050244e-07,
+      "loss": 0.4527,
+      "step": 17435
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 37.65734553346127,
+      "learning_rate": 4.421913481540929e-07,
+      "loss": 0.3399,
+      "step": 17436
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 47.37031992540321,
+      "learning_rate": 4.4169554816617224e-07,
+      "loss": 0.4156,
+      "step": 17437
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 46.10747494946801,
+      "learning_rate": 4.4120002001083575e-07,
+      "loss": 0.3883,
+      "step": 17438
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 71.80289440802845,
+      "learning_rate": 4.4070476370218e-07,
+      "loss": 0.5051,
+      "step": 17439
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 39.553869960412406,
+      "learning_rate": 4.402097792542892e-07,
+      "loss": 0.3424,
+      "step": 17440
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 50.61070779692525,
+      "learning_rate": 4.397150666812411e-07,
+      "loss": 0.3976,
+      "step": 17441
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 47.35386536560111,
+      "learning_rate": 4.392206259971021e-07,
+      "loss": 0.4072,
+      "step": 17442
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 57.95492019402969,
+      "learning_rate": 4.3872645721593556e-07,
+      "loss": 0.4186,
+      "step": 17443
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 48.99811351317088,
+      "learning_rate": 4.382325603517956e-07,
+      "loss": 0.3869,
+      "step": 17444
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 51.22825524118055,
+      "learning_rate": 4.3773893541873005e-07,
+      "loss": 0.4809,
+      "step": 17445
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 40.218785251106986,
+      "learning_rate": 4.372455824307731e-07,
+      "loss": 0.4233,
+      "step": 17446
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 58.50753411252141,
+      "learning_rate": 4.36752501401958e-07,
+      "loss": 0.4539,
+      "step": 17447
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 53.661141654149596,
+      "learning_rate": 4.3625969234630694e-07,
+      "loss": 0.4893,
+      "step": 17448
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 53.71515686551983,
+      "learning_rate": 4.3576715527783755e-07,
+      "loss": 0.4479,
+      "step": 17449
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 50.2832064677682,
+      "learning_rate": 4.3527489021055414e-07,
+      "loss": 0.4092,
+      "step": 17450
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 57.56783927650185,
+      "learning_rate": 4.347828971584578e-07,
+      "loss": 0.4697,
+      "step": 17451
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 42.0287574000605,
+      "learning_rate": 4.342911761355395e-07,
+      "loss": 0.3642,
+      "step": 17452
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 38.882117122834394,
+      "learning_rate": 4.3379972715578587e-07,
+      "loss": 0.3496,
+      "step": 17453
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.76076805825139,
+      "learning_rate": 4.333085502331713e-07,
+      "loss": 0.4014,
+      "step": 17454
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 52.18746601491953,
+      "learning_rate": 4.328176453816657e-07,
+      "loss": 0.438,
+      "step": 17455
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 43.911078891405026,
+      "learning_rate": 4.32327012615229e-07,
+      "loss": 0.4096,
+      "step": 17456
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 45.912166190861775,
+      "learning_rate": 4.3183665194781564e-07,
+      "loss": 0.4221,
+      "step": 17457
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 47.348729824121385,
+      "learning_rate": 4.3134656339337e-07,
+      "loss": 0.474,
+      "step": 17458
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 63.49360997045355,
+      "learning_rate": 4.308567469658298e-07,
+      "loss": 0.5135,
+      "step": 17459
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 55.64045216962822,
+      "learning_rate": 4.3036720267912833e-07,
+      "loss": 0.4157,
+      "step": 17460
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.79256075557773,
+      "learning_rate": 4.2987793054718343e-07,
+      "loss": 0.4917,
+      "step": 17461
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 54.870163754392124,
+      "learning_rate": 4.2938893058391385e-07,
+      "loss": 0.4633,
+      "step": 17462
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 39.687055158750326,
+      "learning_rate": 4.289002028032219e-07,
+      "loss": 0.3412,
+      "step": 17463
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 50.341954090897815,
+      "learning_rate": 4.284117472190108e-07,
+      "loss": 0.4041,
+      "step": 17464
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 35.03247812383,
+      "learning_rate": 4.2792356384517063e-07,
+      "loss": 0.3775,
+      "step": 17465
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.043969587266126,
+      "learning_rate": 4.2743565269558475e-07,
+      "loss": 0.3616,
+      "step": 17466
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.712155492796015,
+      "learning_rate": 4.2694801378412867e-07,
+      "loss": 0.4097,
+      "step": 17467
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 55.709049268281575,
+      "learning_rate": 4.2646064712467016e-07,
+      "loss": 0.4568,
+      "step": 17468
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 47.744422991349325,
+      "learning_rate": 4.2597355273107267e-07,
+      "loss": 0.4504,
+      "step": 17469
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 51.486798899990056,
+      "learning_rate": 4.2548673061718614e-07,
+      "loss": 0.4509,
+      "step": 17470
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 44.818287649923064,
+      "learning_rate": 4.25000180796854e-07,
+      "loss": 0.4002,
+      "step": 17471
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 43.513269653132475,
+      "learning_rate": 4.245139032839163e-07,
+      "loss": 0.402,
+      "step": 17472
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 45.82109232285273,
+      "learning_rate": 4.240278980922019e-07,
+      "loss": 0.4089,
+      "step": 17473
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 48.37453619663166,
+      "learning_rate": 4.2354216523553314e-07,
+      "loss": 0.4038,
+      "step": 17474
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 51.98383196405802,
+      "learning_rate": 4.230567047277234e-07,
+      "loss": 0.4169,
+      "step": 17475
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 40.94884194074537,
+      "learning_rate": 4.2257151658257724e-07,
+      "loss": 0.3763,
+      "step": 17476
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 45.89801866186059,
+      "learning_rate": 4.2208660081389463e-07,
+      "loss": 0.3978,
+      "step": 17477
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 48.234072838452526,
+      "learning_rate": 4.2160195743546797e-07,
+      "loss": 0.4139,
+      "step": 17478
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 52.71963571687435,
+      "learning_rate": 4.2111758646107617e-07,
+      "loss": 0.4234,
+      "step": 17479
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 44.53717891375023,
+      "learning_rate": 4.2063348790449823e-07,
+      "loss": 0.3847,
+      "step": 17480
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 42.849322072856936,
+      "learning_rate": 4.201496617794998e-07,
+      "loss": 0.4163,
+      "step": 17481
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 51.66976731082031,
+      "learning_rate": 4.19666108099841e-07,
+      "loss": 0.4281,
+      "step": 17482
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.357080148868185,
+      "learning_rate": 4.1918282687927304e-07,
+      "loss": 0.3693,
+      "step": 17483
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 43.64910131099572,
+      "learning_rate": 4.186998181315405e-07,
+      "loss": 0.3742,
+      "step": 17484
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.61212730146188,
+      "learning_rate": 4.1821708187038236e-07,
+      "loss": 0.426,
+      "step": 17485
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 37.45383438056032,
+      "learning_rate": 4.177346181095232e-07,
+      "loss": 0.3549,
+      "step": 17486
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.25188279830182,
+      "learning_rate": 4.172524268626876e-07,
+      "loss": 0.356,
+      "step": 17487
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 39.303587165463426,
+      "learning_rate": 4.167705081435858e-07,
+      "loss": 0.3717,
+      "step": 17488
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 56.248067975243124,
+      "learning_rate": 4.162888619659244e-07,
+      "loss": 0.435,
+      "step": 17489
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 50.85723615564327,
+      "learning_rate": 4.158074883434038e-07,
+      "loss": 0.4829,
+      "step": 17490
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 54.792544413257616,
+      "learning_rate": 4.1532638728971065e-07,
+      "loss": 0.4744,
+      "step": 17491
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.506292476257855,
+      "learning_rate": 4.148455588185274e-07,
+      "loss": 0.4062,
+      "step": 17492
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 43.79464688248989,
+      "learning_rate": 4.143650029435287e-07,
+      "loss": 0.3653,
+      "step": 17493
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 54.633730943515765,
+      "learning_rate": 4.138847196783835e-07,
+      "loss": 0.3958,
+      "step": 17494
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 36.97225760735178,
+      "learning_rate": 4.134047090367488e-07,
+      "loss": 0.3488,
+      "step": 17495
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 47.51793395343556,
+      "learning_rate": 4.129249710322758e-07,
+      "loss": 0.3954,
+      "step": 17496
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.98242913024332,
+      "learning_rate": 4.124455056786081e-07,
+      "loss": 0.4034,
+      "step": 17497
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.10349206645824,
+      "learning_rate": 4.119663129893814e-07,
+      "loss": 0.3915,
+      "step": 17498
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 37.084904278830564,
+      "learning_rate": 4.114873929782259e-07,
+      "loss": 0.3484,
+      "step": 17499
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 52.93876852702237,
+      "learning_rate": 4.1100874565875745e-07,
+      "loss": 0.3999,
+      "step": 17500
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.509265448265595,
+      "learning_rate": 4.105303710445918e-07,
+      "loss": 0.4283,
+      "step": 17501
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 48.26717501417847,
+      "learning_rate": 4.100522691493325e-07,
+      "loss": 0.4456,
+      "step": 17502
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.53984641198016,
+      "learning_rate": 4.0957443998657754e-07,
+      "loss": 0.3946,
+      "step": 17503
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 47.343442365698124,
+      "learning_rate": 4.0909688356991495e-07,
+      "loss": 0.3888,
+      "step": 17504
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 39.831736266602114,
+      "learning_rate": 4.086195999129261e-07,
+      "loss": 0.3629,
+      "step": 17505
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.33662337478045,
+      "learning_rate": 4.081425890291846e-07,
+      "loss": 0.4111,
+      "step": 17506
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 43.68159453681478,
+      "learning_rate": 4.076658509322573e-07,
+      "loss": 0.3557,
+      "step": 17507
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 51.94460910384565,
+      "learning_rate": 4.0718938563570233e-07,
+      "loss": 0.4215,
+      "step": 17508
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 46.064402738943514,
+      "learning_rate": 4.067131931530699e-07,
+      "loss": 0.4183,
+      "step": 17509
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.12237697977604,
+      "learning_rate": 4.0623727349790034e-07,
+      "loss": 0.4093,
+      "step": 17510
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.75014715690136,
+      "learning_rate": 4.0576162668373164e-07,
+      "loss": 0.3955,
+      "step": 17511
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 57.087325629896625,
+      "learning_rate": 4.0528625272409083e-07,
+      "loss": 0.4792,
+      "step": 17512
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 55.26732396692431,
+      "learning_rate": 4.0481115163249483e-07,
+      "loss": 0.4291,
+      "step": 17513
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.62797481527453,
+      "learning_rate": 4.0433632342245735e-07,
+      "loss": 0.4069,
+      "step": 17514
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 46.098788688823625,
+      "learning_rate": 4.038617681074808e-07,
+      "loss": 0.3687,
+      "step": 17515
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 46.223171226426466,
+      "learning_rate": 4.0338748570106333e-07,
+      "loss": 0.3883,
+      "step": 17516
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 59.09346249095385,
+      "learning_rate": 4.0291347621669084e-07,
+      "loss": 0.5027,
+      "step": 17517
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 41.964208298933514,
+      "learning_rate": 4.0243973966784477e-07,
+      "loss": 0.4364,
+      "step": 17518
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 53.968607948173705,
+      "learning_rate": 4.019662760679988e-07,
+      "loss": 0.4395,
+      "step": 17519
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.11163192248361,
+      "learning_rate": 4.014930854306176e-07,
+      "loss": 0.4147,
+      "step": 17520
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.606438296129596,
+      "learning_rate": 4.010201677691572e-07,
+      "loss": 0.4481,
+      "step": 17521
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.527172671965886,
+      "learning_rate": 4.005475230970679e-07,
+      "loss": 0.452,
+      "step": 17522
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 48.483887153947045,
+      "learning_rate": 4.000751514277912e-07,
+      "loss": 0.4827,
+      "step": 17523
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.81278959199058,
+      "learning_rate": 3.9960305277476294e-07,
+      "loss": 0.4456,
+      "step": 17524
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 52.546429348909584,
+      "learning_rate": 3.9913122715140804e-07,
+      "loss": 0.4388,
+      "step": 17525
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.18776619185592,
+      "learning_rate": 3.986596745711424e-07,
+      "loss": 0.4048,
+      "step": 17526
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 58.81181728278924,
+      "learning_rate": 3.9818839504737974e-07,
+      "loss": 0.5097,
+      "step": 17527
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 44.971006992899824,
+      "learning_rate": 3.977173885935237e-07,
+      "loss": 0.4164,
+      "step": 17528
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 44.64834367770585,
+      "learning_rate": 3.972466552229681e-07,
+      "loss": 0.4372,
+      "step": 17529
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 46.41508417782049,
+      "learning_rate": 3.967761949490978e-07,
+      "loss": 0.4143,
+      "step": 17530
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 44.59229507692625,
+      "learning_rate": 3.963060077852965e-07,
+      "loss": 0.5038,
+      "step": 17531
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 46.89517450681886,
+      "learning_rate": 3.958360937449335e-07,
+      "loss": 0.4051,
+      "step": 17532
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 41.294054129155946,
+      "learning_rate": 3.953664528413737e-07,
+      "loss": 0.4283,
+      "step": 17533
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 54.70857402964528,
+      "learning_rate": 3.948970850879752e-07,
+      "loss": 0.474,
+      "step": 17534
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 48.1562294387526,
+      "learning_rate": 3.94427990498083e-07,
+      "loss": 0.4419,
+      "step": 17535
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 53.01074910923235,
+      "learning_rate": 3.9395916908503863e-07,
+      "loss": 0.415,
+      "step": 17536
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 57.8313977836486,
+      "learning_rate": 3.9349062086217915e-07,
+      "loss": 0.4601,
+      "step": 17537
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 44.85785712076106,
+      "learning_rate": 3.930223458428239e-07,
+      "loss": 0.3585,
+      "step": 17538
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 49.739992294893796,
+      "learning_rate": 3.9255434404029567e-07,
+      "loss": 0.381,
+      "step": 17539
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 44.15537137441795,
+      "learning_rate": 3.9208661546789927e-07,
+      "loss": 0.3961,
+      "step": 17540
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 43.29903656744472,
+      "learning_rate": 3.9161916013894186e-07,
+      "loss": 0.445,
+      "step": 17541
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 51.05892523739018,
+      "learning_rate": 3.9115197806671277e-07,
+      "loss": 0.421,
+      "step": 17542
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 57.26495964631751,
+      "learning_rate": 3.9068506926450146e-07,
+      "loss": 0.4877,
+      "step": 17543
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 48.84908832218624,
+      "learning_rate": 3.9021843374558385e-07,
+      "loss": 0.4199,
+      "step": 17544
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 54.02176877835157,
+      "learning_rate": 3.897520715232339e-07,
+      "loss": 0.443,
+      "step": 17545
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 59.20703964198104,
+      "learning_rate": 3.8928598261071313e-07,
+      "loss": 0.46,
+      "step": 17546
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 51.692817801896915,
+      "learning_rate": 3.8882016702127544e-07,
+      "loss": 0.438,
+      "step": 17547
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.32410009025279,
+      "learning_rate": 3.883546247681713e-07,
+      "loss": 0.4208,
+      "step": 17548
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 48.255973525207004,
+      "learning_rate": 3.8788935586463906e-07,
+      "loss": 0.4441,
+      "step": 17549
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 51.213748098802,
+      "learning_rate": 3.8742436032391027e-07,
+      "loss": 0.3841,
+      "step": 17550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 49.092494635311894,
+      "learning_rate": 3.8695963815920887e-07,
+      "loss": 0.388,
+      "step": 17551
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.53632648152007,
+      "learning_rate": 3.864951893837521e-07,
+      "loss": 0.4212,
+      "step": 17552
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 32.64707587936089,
+      "learning_rate": 3.8603101401074927e-07,
+      "loss": 0.3733,
+      "step": 17553
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.65144189305463,
+      "learning_rate": 3.855671120533999e-07,
+      "loss": 0.4072,
+      "step": 17554
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 43.53027827375709,
+      "learning_rate": 3.851034835248979e-07,
+      "loss": 0.3901,
+      "step": 17555
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 39.543719759272804,
+      "learning_rate": 3.8464012843842714e-07,
+      "loss": 0.4235,
+      "step": 17556
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.56574913119526,
+      "learning_rate": 3.8417704680716704e-07,
+      "loss": 0.4131,
+      "step": 17557
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 40.94496392791516,
+      "learning_rate": 3.8371423864428826e-07,
+      "loss": 0.3783,
+      "step": 17558
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 51.38206754460238,
+      "learning_rate": 3.8325170396294907e-07,
+      "loss": 0.4774,
+      "step": 17559
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 47.011294823808534,
+      "learning_rate": 3.827894427763057e-07,
+      "loss": 0.4087,
+      "step": 17560
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 54.31056747246818,
+      "learning_rate": 3.8232745509750423e-07,
+      "loss": 0.4112,
+      "step": 17561
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 49.03524048748695,
+      "learning_rate": 3.818657409396853e-07,
+      "loss": 0.397,
+      "step": 17562
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 46.624743345847946,
+      "learning_rate": 3.8140430031597844e-07,
+      "loss": 0.3919,
+      "step": 17563
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 54.22397047519105,
+      "learning_rate": 3.809431332395053e-07,
+      "loss": 0.4359,
+      "step": 17564
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 53.56799614927433,
+      "learning_rate": 3.804822397233832e-07,
+      "loss": 0.4646,
+      "step": 17565
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 45.14235807256342,
+      "learning_rate": 3.800216197807194e-07,
+      "loss": 0.4461,
+      "step": 17566
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 42.17371385317096,
+      "learning_rate": 3.7956127342461127e-07,
+      "loss": 0.4079,
+      "step": 17567
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 50.35875010065718,
+      "learning_rate": 3.791012006681549e-07,
+      "loss": 0.5008,
+      "step": 17568
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 52.49814476186568,
+      "learning_rate": 3.786414015244311e-07,
+      "loss": 0.4256,
+      "step": 17569
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 42.645164273294824,
+      "learning_rate": 3.781818760065181e-07,
+      "loss": 0.4098,
+      "step": 17570
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 51.77450458191963,
+      "learning_rate": 3.777226241274834e-07,
+      "loss": 0.4172,
+      "step": 17571
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 51.93324420261855,
+      "learning_rate": 3.772636459003887e-07,
+      "loss": 0.4177,
+      "step": 17572
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 56.734869616261086,
+      "learning_rate": 3.768049413382868e-07,
+      "loss": 0.4678,
+      "step": 17573
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.76276653308669,
+      "learning_rate": 3.76346510454223e-07,
+      "loss": 0.3724,
+      "step": 17574
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 56.80078117864331,
+      "learning_rate": 3.758883532612356e-07,
+      "loss": 0.4707,
+      "step": 17575
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 42.98277653031153,
+      "learning_rate": 3.754304697723521e-07,
+      "loss": 0.3555,
+      "step": 17576
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.56299501123658,
+      "learning_rate": 3.749728600005953e-07,
+      "loss": 0.4269,
+      "step": 17577
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 50.35553222349228,
+      "learning_rate": 3.745155239589815e-07,
+      "loss": 0.4695,
+      "step": 17578
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.53570153108967,
+      "learning_rate": 3.7405846166051585e-07,
+      "loss": 0.4478,
+      "step": 17579
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 57.64952596517511,
+      "learning_rate": 3.736016731181946e-07,
+      "loss": 0.4431,
+      "step": 17580
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 43.008651581079334,
+      "learning_rate": 3.7314515834501075e-07,
+      "loss": 0.3551,
+      "step": 17581
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 49.62208448632837,
+      "learning_rate": 3.726889173539483e-07,
+      "loss": 0.4194,
+      "step": 17582
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.04047538056191,
+      "learning_rate": 3.7223295015798133e-07,
+      "loss": 0.3783,
+      "step": 17583
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 43.68187591195994,
+      "learning_rate": 3.7177725677007615e-07,
+      "loss": 0.3965,
+      "step": 17584
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 45.08247148262323,
+      "learning_rate": 3.713218372031935e-07,
+      "loss": 0.3748,
+      "step": 17585
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 38.07895626934272,
+      "learning_rate": 3.708666914702852e-07,
+      "loss": 0.3927,
+      "step": 17586
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 45.79974685769733,
+      "learning_rate": 3.704118195842965e-07,
+      "loss": 0.4511,
+      "step": 17587
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 41.88079292880227,
+      "learning_rate": 3.699572215581615e-07,
+      "loss": 0.4089,
+      "step": 17588
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 47.86518744363149,
+      "learning_rate": 3.695028974048098e-07,
+      "loss": 0.4122,
+      "step": 17589
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.81124007359213,
+      "learning_rate": 3.6904884713716114e-07,
+      "loss": 0.3896,
+      "step": 17590
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 49.83094566689588,
+      "learning_rate": 3.6859507076813073e-07,
+      "loss": 0.4289,
+      "step": 17591
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 52.64075764123879,
+      "learning_rate": 3.6814156831062264e-07,
+      "loss": 0.3741,
+      "step": 17592
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 88.24140041045884,
+      "learning_rate": 3.6768833977753214e-07,
+      "loss": 0.5267,
+      "step": 17593
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.78526125319811,
+      "learning_rate": 3.672353851817512e-07,
+      "loss": 0.4602,
+      "step": 17594
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 62.118205345583995,
+      "learning_rate": 3.667827045361616e-07,
+      "loss": 0.5092,
+      "step": 17595
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.681387501823295,
+      "learning_rate": 3.663302978536354e-07,
+      "loss": 0.3909,
+      "step": 17596
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 47.208779277642435,
+      "learning_rate": 3.658781651470422e-07,
+      "loss": 0.4034,
+      "step": 17597
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 47.81551581451872,
+      "learning_rate": 3.654263064292363e-07,
+      "loss": 0.3775,
+      "step": 17598
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 36.32973242372152,
+      "learning_rate": 3.649747217130695e-07,
+      "loss": 0.3945,
+      "step": 17599
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 37.5742040099339,
+      "learning_rate": 3.645234110113871e-07,
+      "loss": 0.3767,
+      "step": 17600
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 56.27550162486233,
+      "learning_rate": 3.6407237433702116e-07,
+      "loss": 0.4577,
+      "step": 17601
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.96241908448483,
+      "learning_rate": 3.636216117028013e-07,
+      "loss": 0.3851,
+      "step": 17602
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 42.495597392536624,
+      "learning_rate": 3.63171123121544e-07,
+      "loss": 0.3982,
+      "step": 17603
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 47.88247544613,
+      "learning_rate": 3.627209086060635e-07,
+      "loss": 0.38,
+      "step": 17604
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.7490300810953,
+      "learning_rate": 3.6227096816916274e-07,
+      "loss": 0.4389,
+      "step": 17605
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 59.602500957141636,
+      "learning_rate": 3.6182130182363716e-07,
+      "loss": 0.4505,
+      "step": 17606
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 56.60649425432414,
+      "learning_rate": 3.6137190958227655e-07,
+      "loss": 0.4803,
+      "step": 17607
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 42.82773317410396,
+      "learning_rate": 3.609227914578606e-07,
+      "loss": 0.3441,
+      "step": 17608
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 55.94836913911732,
+      "learning_rate": 3.6047394746316023e-07,
+      "loss": 0.4767,
+      "step": 17609
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 54.138441588396674,
+      "learning_rate": 3.6002537761094193e-07,
+      "loss": 0.4319,
+      "step": 17610
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.41339465798781,
+      "learning_rate": 3.595770819139632e-07,
+      "loss": 0.3564,
+      "step": 17611
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 53.84042616474294,
+      "learning_rate": 3.5912906038497287e-07,
+      "loss": 0.4775,
+      "step": 17612
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 45.5007398880547,
+      "learning_rate": 3.5868131303671393e-07,
+      "loss": 0.3845,
+      "step": 17613
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 53.39475157266196,
+      "learning_rate": 3.582338398819163e-07,
+      "loss": 0.478,
+      "step": 17614
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 45.88424929787572,
+      "learning_rate": 3.577866409333075e-07,
+      "loss": 0.3693,
+      "step": 17615
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 49.399509305486056,
+      "learning_rate": 3.5733971620360853e-07,
+      "loss": 0.4002,
+      "step": 17616
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.90794732650527,
+      "learning_rate": 3.5689306570552584e-07,
+      "loss": 0.479,
+      "step": 17617
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.53918512065804,
+      "learning_rate": 3.5644668945176264e-07,
+      "loss": 0.3609,
+      "step": 17618
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 47.15963339628329,
+      "learning_rate": 3.5600058745501543e-07,
+      "loss": 0.3701,
+      "step": 17619
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.89522995426382,
+      "learning_rate": 3.555547597279696e-07,
+      "loss": 0.4757,
+      "step": 17620
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 43.589201910751335,
+      "learning_rate": 3.551092062833039e-07,
+      "loss": 0.4042,
+      "step": 17621
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.9813670327806,
+      "learning_rate": 3.546639271336916e-07,
+      "loss": 0.4292,
+      "step": 17622
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 42.84376288816563,
+      "learning_rate": 3.5421892229179354e-07,
+      "loss": 0.384,
+      "step": 17623
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 36.89885868282008,
+      "learning_rate": 3.537741917702664e-07,
+      "loss": 0.3723,
+      "step": 17624
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 50.49895851789253,
+      "learning_rate": 3.5332973558175996e-07,
+      "loss": 0.4502,
+      "step": 17625
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 61.397716878634306,
+      "learning_rate": 3.52885553738912e-07,
+      "loss": 0.47,
+      "step": 17626
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.42641618075404,
+      "learning_rate": 3.524416462543556e-07,
+      "loss": 0.4303,
+      "step": 17627
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 51.76282531186511,
+      "learning_rate": 3.519980131407152e-07,
+      "loss": 0.4471,
+      "step": 17628
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 49.81062602941924,
+      "learning_rate": 3.515546544106074e-07,
+      "loss": 0.433,
+      "step": 17629
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 47.68369237015048,
+      "learning_rate": 3.5111157007663986e-07,
+      "loss": 0.4509,
+      "step": 17630
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 46.031971705023935,
+      "learning_rate": 3.506687601514158e-07,
+      "loss": 0.3924,
+      "step": 17631
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 50.42622266430375,
+      "learning_rate": 3.502262246475285e-07,
+      "loss": 0.4626,
+      "step": 17632
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 49.31892946824822,
+      "learning_rate": 3.497839635775613e-07,
+      "loss": 0.4423,
+      "step": 17633
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 45.00186801318086,
+      "learning_rate": 3.493419769540929e-07,
+      "loss": 0.3571,
+      "step": 17634
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 48.5347564865013,
+      "learning_rate": 3.489002647896933e-07,
+      "loss": 0.3998,
+      "step": 17635
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 45.83932798312448,
+      "learning_rate": 3.4845882709692356e-07,
+      "loss": 0.4135,
+      "step": 17636
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 58.56365613905552,
+      "learning_rate": 3.4801766388834037e-07,
+      "loss": 0.3851,
+      "step": 17637
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 45.66874106190292,
+      "learning_rate": 3.4757677517648916e-07,
+      "loss": 0.3851,
+      "step": 17638
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 58.62022931141613,
+      "learning_rate": 3.471361609739055e-07,
+      "loss": 0.4905,
+      "step": 17639
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 46.90876158433363,
+      "learning_rate": 3.4669582129312373e-07,
+      "loss": 0.3988,
+      "step": 17640
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 43.651287236410674,
+      "learning_rate": 3.462557561466662e-07,
+      "loss": 0.3771,
+      "step": 17641
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 47.57981669608328,
+      "learning_rate": 3.458159655470483e-07,
+      "loss": 0.4107,
+      "step": 17642
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 45.06945799971215,
+      "learning_rate": 3.4537644950677576e-07,
+      "loss": 0.3742,
+      "step": 17643
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 52.14601207564848,
+      "learning_rate": 3.449372080383484e-07,
+      "loss": 0.3915,
+      "step": 17644
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 45.25021652043423,
+      "learning_rate": 3.4449824115426076e-07,
+      "loss": 0.4297,
+      "step": 17645
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 48.318212849941354,
+      "learning_rate": 3.440595488669951e-07,
+      "loss": 0.4443,
+      "step": 17646
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 48.000062569935075,
+      "learning_rate": 3.4362113118902475e-07,
+      "loss": 0.3955,
+      "step": 17647
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 41.42002946274214,
+      "learning_rate": 3.4318298813282193e-07,
+      "loss": 0.3633,
+      "step": 17648
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 58.44834462278099,
+      "learning_rate": 3.4274511971084556e-07,
+      "loss": 0.4353,
+      "step": 17649
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 48.58414513949532,
+      "learning_rate": 3.423075259355491e-07,
+      "loss": 0.4063,
+      "step": 17650
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 41.52990149841091,
+      "learning_rate": 3.41870206819378e-07,
+      "loss": 0.375,
+      "step": 17651
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 57.25882286068307,
+      "learning_rate": 3.4143316237476686e-07,
+      "loss": 0.429,
+      "step": 17652
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 48.43422800149214,
+      "learning_rate": 3.409963926141457e-07,
+      "loss": 0.3925,
+      "step": 17653
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 49.773894911894494,
+      "learning_rate": 3.4055989754993913e-07,
+      "loss": 0.4132,
+      "step": 17654
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 46.772236350014424,
+      "learning_rate": 3.4012367719455595e-07,
+      "loss": 0.3648,
+      "step": 17655
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 52.17508140415043,
+      "learning_rate": 3.3968773156040745e-07,
+      "loss": 0.4812,
+      "step": 17656
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 44.496055803273556,
+      "learning_rate": 3.392520606598859e-07,
+      "loss": 0.4438,
+      "step": 17657
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 58.18000360010896,
+      "learning_rate": 3.3881666450538695e-07,
+      "loss": 0.4019,
+      "step": 17658
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 63.08280608823179,
+      "learning_rate": 3.383815431092885e-07,
+      "loss": 0.4442,
+      "step": 17659
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 47.95263447888208,
+      "learning_rate": 3.3794669648396727e-07,
+      "loss": 0.3697,
+      "step": 17660
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 49.05329761094429,
+      "learning_rate": 3.3751212464179007e-07,
+      "loss": 0.4166,
+      "step": 17661
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 46.48779145836914,
+      "learning_rate": 3.370778275951159e-07,
+      "loss": 0.421,
+      "step": 17662
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 47.75671397436717,
+      "learning_rate": 3.3664380535629595e-07,
+      "loss": 0.4077,
+      "step": 17663
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 41.8379813584582,
+      "learning_rate": 3.362100579376726e-07,
+      "loss": 0.3695,
+      "step": 17664
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 53.165062085914286,
+      "learning_rate": 3.357765853515815e-07,
+      "loss": 0.4134,
+      "step": 17665
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 43.542210427985,
+      "learning_rate": 3.353433876103529e-07,
+      "loss": 0.3563,
+      "step": 17666
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 54.028508780349384,
+      "learning_rate": 3.349104647263046e-07,
+      "loss": 0.456,
+      "step": 17667
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 48.32386405308327,
+      "learning_rate": 3.3447781671174684e-07,
+      "loss": 0.4266,
+      "step": 17668
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 54.00418015036356,
+      "learning_rate": 3.340454435789864e-07,
+      "loss": 0.4352,
+      "step": 17669
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 58.21764290412744,
+      "learning_rate": 3.336133453403201e-07,
+      "loss": 0.4741,
+      "step": 17670
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 46.78149184659469,
+      "learning_rate": 3.331815220080359e-07,
+      "loss": 0.3863,
+      "step": 17671
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 57.91535933228883,
+      "learning_rate": 3.3274997359441174e-07,
+      "loss": 0.4956,
+      "step": 17672
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 42.147467855731655,
+      "learning_rate": 3.323187001117245e-07,
+      "loss": 0.4008,
+      "step": 17673
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 44.4942399701151,
+      "learning_rate": 3.318877015722377e-07,
+      "loss": 0.4407,
+      "step": 17674
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 43.058039313768695,
+      "learning_rate": 3.3145697798821044e-07,
+      "loss": 0.3554,
+      "step": 17675
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 54.78490140315057,
+      "learning_rate": 3.3102652937189064e-07,
+      "loss": 0.405,
+      "step": 17676
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 44.34352808030497,
+      "learning_rate": 3.3059635573551854e-07,
+      "loss": 0.3881,
+      "step": 17677
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 50.15358628659795,
+      "learning_rate": 3.301664570913299e-07,
+      "loss": 0.3933,
+      "step": 17678
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 46.82137734920696,
+      "learning_rate": 3.297368334515516e-07,
+      "loss": 0.3704,
+      "step": 17679
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 50.52725713521156,
+      "learning_rate": 3.2930748482840056e-07,
+      "loss": 0.4426,
+      "step": 17680
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 44.82708367261452,
+      "learning_rate": 3.2887841123408817e-07,
+      "loss": 0.3819,
+      "step": 17681
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 49.224141395407926,
+      "learning_rate": 3.2844961268081457e-07,
+      "loss": 0.3897,
+      "step": 17682
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 52.753579881513694,
+      "learning_rate": 3.2802108918077777e-07,
+      "loss": 0.3761,
+      "step": 17683
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 64.87563453760903,
+      "learning_rate": 3.275928407461615e-07,
+      "loss": 0.4656,
+      "step": 17684
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 41.66547601842428,
+      "learning_rate": 3.271648673891481e-07,
+      "loss": 0.3337,
+      "step": 17685
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 41.59062759178131,
+      "learning_rate": 3.267371691219068e-07,
+      "loss": 0.3992,
+      "step": 17686
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 52.690371614985715,
+      "learning_rate": 3.2630974595660226e-07,
+      "loss": 0.471,
+      "step": 17687
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 40.602888472108965,
+      "learning_rate": 3.2588259790538923e-07,
+      "loss": 0.3795,
+      "step": 17688
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 47.57656982586188,
+      "learning_rate": 3.2545572498041576e-07,
+      "loss": 0.3896,
+      "step": 17689
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 53.931127695600765,
+      "learning_rate": 3.2502912719382327e-07,
+      "loss": 0.4397,
+      "step": 17690
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 45.097620046542424,
+      "learning_rate": 3.2460280455774096e-07,
+      "loss": 0.4149,
+      "step": 17691
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 44.1948868163069,
+      "learning_rate": 3.2417675708429574e-07,
+      "loss": 0.4171,
+      "step": 17692
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 48.435744857983366,
+      "learning_rate": 3.237509847856035e-07,
+      "loss": 0.4267,
+      "step": 17693
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 47.52420655882021,
+      "learning_rate": 3.2332548767377237e-07,
+      "loss": 0.4466,
+      "step": 17694
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 56.12432805701989,
+      "learning_rate": 3.229002657609037e-07,
+      "loss": 0.4785,
+      "step": 17695
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 42.12832993496518,
+      "learning_rate": 3.224753190590912e-07,
+      "loss": 0.409,
+      "step": 17696
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 42.05370943511357,
+      "learning_rate": 3.220506475804186e-07,
+      "loss": 0.4135,
+      "step": 17697
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 46.22995267901514,
+      "learning_rate": 3.21626251336965e-07,
+      "loss": 0.4136,
+      "step": 17698
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 42.02251396008305,
+      "learning_rate": 3.212021303407975e-07,
+      "loss": 0.3326,
+      "step": 17699
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 49.84567449316346,
+      "learning_rate": 3.20778284603982e-07,
+      "loss": 0.4264,
+      "step": 17700
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 41.62039885994774,
+      "learning_rate": 3.203547141385688e-07,
+      "loss": 0.4066,
+      "step": 17701
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 61.48804460793444,
+      "learning_rate": 3.1993141895660385e-07,
+      "loss": 0.3885,
+      "step": 17702
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.78076588158327,
+      "learning_rate": 3.1950839907012755e-07,
+      "loss": 0.3494,
+      "step": 17703
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.69303900405267,
+      "learning_rate": 3.1908565449117026e-07,
+      "loss": 0.3637,
+      "step": 17704
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 47.07344018192444,
+      "learning_rate": 3.1866318523175344e-07,
+      "loss": 0.3943,
+      "step": 17705
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.28131115079341,
+      "learning_rate": 3.18240991303892e-07,
+      "loss": 0.3888,
+      "step": 17706
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 45.80576056313787,
+      "learning_rate": 3.178190727195918e-07,
+      "loss": 0.3966,
+      "step": 17707
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 49.958186547342365,
+      "learning_rate": 3.173974294908555e-07,
+      "loss": 0.3897,
+      "step": 17708
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 48.61120718749709,
+      "learning_rate": 3.169760616296702e-07,
+      "loss": 0.4205,
+      "step": 17709
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 48.46774628397443,
+      "learning_rate": 3.1655496914802296e-07,
+      "loss": 0.4398,
+      "step": 17710
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 50.8297747799546,
+      "learning_rate": 3.1613415205788643e-07,
+      "loss": 0.3938,
+      "step": 17711
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 42.16525569949972,
+      "learning_rate": 3.1571361037122996e-07,
+      "loss": 0.3834,
+      "step": 17712
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 40.13318059246383,
+      "learning_rate": 3.152933441000139e-07,
+      "loss": 0.348,
+      "step": 17713
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 50.237326451951205,
+      "learning_rate": 3.148733532561887e-07,
+      "loss": 0.39,
+      "step": 17714
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 40.63239088758388,
+      "learning_rate": 3.1445363785170155e-07,
+      "loss": 0.3379,
+      "step": 17715
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 49.831739665971995,
+      "learning_rate": 3.140341978984851e-07,
+      "loss": 0.418,
+      "step": 17716
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 64.83301998477843,
+      "learning_rate": 3.13615033408472e-07,
+      "loss": 0.4852,
+      "step": 17717
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.34310334176385,
+      "learning_rate": 3.1319614439357827e-07,
+      "loss": 0.3841,
+      "step": 17718
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 38.715362714703744,
+      "learning_rate": 3.1277753086572105e-07,
+      "loss": 0.3722,
+      "step": 17719
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 39.40993579927997,
+      "learning_rate": 3.123591928368042e-07,
+      "loss": 0.3822,
+      "step": 17720
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 60.65620978942199,
+      "learning_rate": 3.1194113031872584e-07,
+      "loss": 0.434,
+      "step": 17721
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 40.526829333614,
+      "learning_rate": 3.115233433233722e-07,
+      "loss": 0.4499,
+      "step": 17722
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 49.032479800184795,
+      "learning_rate": 3.111058318626281e-07,
+      "loss": 0.3866,
+      "step": 17723
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.128583757728684,
+      "learning_rate": 3.106885959483652e-07,
+      "loss": 0.3529,
+      "step": 17724
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 52.97890659957979,
+      "learning_rate": 3.102716355924518e-07,
+      "loss": 0.4432,
+      "step": 17725
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 45.76476533185904,
+      "learning_rate": 3.09854950806745e-07,
+      "loss": 0.4384,
+      "step": 17726
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 44.67687781474839,
+      "learning_rate": 3.094385416030943e-07,
+      "loss": 0.4215,
+      "step": 17727
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 53.5381559158535,
+      "learning_rate": 3.0902240799334237e-07,
+      "loss": 0.4739,
+      "step": 17728
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.287081532847395,
+      "learning_rate": 3.0860654998932536e-07,
+      "loss": 0.3817,
+      "step": 17729
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 48.32779890580025,
+      "learning_rate": 3.0819096760286826e-07,
+      "loss": 0.4226,
+      "step": 17730
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 52.09116707698897,
+      "learning_rate": 3.0777566084578934e-07,
+      "loss": 0.3487,
+      "step": 17731
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 52.98768111780902,
+      "learning_rate": 3.0736062972990145e-07,
+      "loss": 0.4278,
+      "step": 17732
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.304058130251164,
+      "learning_rate": 3.069458742670073e-07,
+      "loss": 0.3739,
+      "step": 17733
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 56.53719866628254,
+      "learning_rate": 3.065313944689019e-07,
+      "loss": 0.4365,
+      "step": 17734
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 55.773646775213855,
+      "learning_rate": 3.0611719034737363e-07,
+      "loss": 0.408,
+      "step": 17735
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 48.506997006836194,
+      "learning_rate": 3.05703261914202e-07,
+      "loss": 0.4134,
+      "step": 17736
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.594289944494086,
+      "learning_rate": 3.0528960918115744e-07,
+      "loss": 0.4089,
+      "step": 17737
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.49066950494313,
+      "learning_rate": 3.048762321600063e-07,
+      "loss": 0.395,
+      "step": 17738
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.606381659727944,
+      "learning_rate": 3.0446313086250346e-07,
+      "loss": 0.3452,
+      "step": 17739
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.98133961218055,
+      "learning_rate": 3.0405030530039734e-07,
+      "loss": 0.362,
+      "step": 17740
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 48.70879058634143,
+      "learning_rate": 3.0363775548542753e-07,
+      "loss": 0.3671,
+      "step": 17741
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 54.445262662043966,
+      "learning_rate": 3.0322548142932893e-07,
+      "loss": 0.4051,
+      "step": 17742
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.65502419522298,
+      "learning_rate": 3.0281348314382453e-07,
+      "loss": 0.4086,
+      "step": 17743
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 47.29608339643449,
+      "learning_rate": 3.0240176064063266e-07,
+      "loss": 0.411,
+      "step": 17744
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 38.97067151995363,
+      "learning_rate": 3.0199031393146174e-07,
+      "loss": 0.3524,
+      "step": 17745
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 39.66622950142028,
+      "learning_rate": 3.015791430280135e-07,
+      "loss": 0.3244,
+      "step": 17746
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 51.53764666291953,
+      "learning_rate": 3.0116824794197977e-07,
+      "loss": 0.3913,
+      "step": 17747
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 58.57436641459536,
+      "learning_rate": 3.0075762868504777e-07,
+      "loss": 0.4661,
+      "step": 17748
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 57.19851930275692,
+      "learning_rate": 3.003472852688949e-07,
+      "loss": 0.4393,
+      "step": 17749
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 48.7428643701938,
+      "learning_rate": 2.999372177051918e-07,
+      "loss": 0.4876,
+      "step": 17750
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 47.23186865594869,
+      "learning_rate": 2.9952742600559917e-07,
+      "loss": 0.4251,
+      "step": 17751
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 57.34761313748667,
+      "learning_rate": 2.9911791018177204e-07,
+      "loss": 0.5667,
+      "step": 17752
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 45.894730822942336,
+      "learning_rate": 2.9870867024535675e-07,
+      "loss": 0.3983,
+      "step": 17753
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 46.884878292635655,
+      "learning_rate": 2.9829970620799176e-07,
+      "loss": 0.4263,
+      "step": 17754
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 42.59172732839566,
+      "learning_rate": 2.978910180813088e-07,
+      "loss": 0.4129,
+      "step": 17755
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 52.20349826125535,
+      "learning_rate": 2.9748260587692867e-07,
+      "loss": 0.4675,
+      "step": 17756
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 43.30424132883744,
+      "learning_rate": 2.9707446960646644e-07,
+      "loss": 0.3688,
+      "step": 17757
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 52.93390823934065,
+      "learning_rate": 2.966666092815318e-07,
+      "loss": 0.3794,
+      "step": 17758
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 40.850213506227036,
+      "learning_rate": 2.962590249137232e-07,
+      "loss": 0.3364,
+      "step": 17759
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 47.14585585965192,
+      "learning_rate": 2.958517165146302e-07,
+      "loss": 0.4427,
+      "step": 17760
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 42.933640810517524,
+      "learning_rate": 2.954446840958369e-07,
+      "loss": 0.352,
+      "step": 17761
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 62.19743846282926,
+      "learning_rate": 2.950379276689197e-07,
+      "loss": 0.5563,
+      "step": 17762
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.789064040427355,
+      "learning_rate": 2.946314472454481e-07,
+      "loss": 0.3782,
+      "step": 17763
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 51.80851638260099,
+      "learning_rate": 2.9422524283698074e-07,
+      "loss": 0.3793,
+      "step": 17764
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 42.36323872516164,
+      "learning_rate": 2.9381931445506826e-07,
+      "loss": 0.4277,
+      "step": 17765
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 57.70388441001864,
+      "learning_rate": 2.934136621112571e-07,
+      "loss": 0.363,
+      "step": 17766
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 48.7411499304318,
+      "learning_rate": 2.9300828581708464e-07,
+      "loss": 0.3824,
+      "step": 17767
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 69.1648155868588,
+      "learning_rate": 2.926031855840761e-07,
+      "loss": 0.4881,
+      "step": 17768
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 56.242763765914304,
+      "learning_rate": 2.9219836142375557e-07,
+      "loss": 0.438,
+      "step": 17769
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 36.627096565310296,
+      "learning_rate": 2.917938133476339e-07,
+      "loss": 0.3975,
+      "step": 17770
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 51.644392910847,
+      "learning_rate": 2.913895413672174e-07,
+      "loss": 0.4972,
+      "step": 17771
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 49.64968929078303,
+      "learning_rate": 2.909855454940025e-07,
+      "loss": 0.3746,
+      "step": 17772
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 48.279069588323445,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.3318,
+      "step": 17773
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 61.411721515283,
+      "learning_rate": 2.901783821151305e-07,
+      "loss": 0.4012,
+      "step": 17774
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 45.59958043511933,
+      "learning_rate": 2.897752146324262e-07,
+      "loss": 0.4539,
+      "step": 17775
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 40.515638605211876,
+      "learning_rate": 2.893723233028367e-07,
+      "loss": 0.391,
+      "step": 17776
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 54.86793017628218,
+      "learning_rate": 2.889697081378162e-07,
+      "loss": 0.468,
+      "step": 17777
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 46.895765994529334,
+      "learning_rate": 2.885673691488167e-07,
+      "loss": 0.4223,
+      "step": 17778
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 64.81455814547908,
+      "learning_rate": 2.8816530634728e-07,
+      "loss": 0.3877,
+      "step": 17779
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 46.32491186105791,
+      "learning_rate": 2.877635197446427e-07,
+      "loss": 0.4179,
+      "step": 17780
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 50.54356944671152,
+      "learning_rate": 2.873620093523266e-07,
+      "loss": 0.473,
+      "step": 17781
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 40.89343537762446,
+      "learning_rate": 2.869607751817538e-07,
+      "loss": 0.4161,
+      "step": 17782
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 47.54190845857072,
+      "learning_rate": 2.8655981724433625e-07,
+      "loss": 0.3648,
+      "step": 17783
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 53.50523218969883,
+      "learning_rate": 2.861591355514748e-07,
+      "loss": 0.4097,
+      "step": 17784
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 57.115142005063284,
+      "learning_rate": 2.857587301145637e-07,
+      "loss": 0.4107,
+      "step": 17785
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 45.88415464065015,
+      "learning_rate": 2.853586009449927e-07,
+      "loss": 0.4012,
+      "step": 17786
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 109.31565200684501,
+      "learning_rate": 2.8495874805413937e-07,
+      "loss": 0.4887,
+      "step": 17787
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 50.215235826348206,
+      "learning_rate": 2.845591714533769e-07,
+      "loss": 0.3569,
+      "step": 17788
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 41.001818383994475,
+      "learning_rate": 2.841598711540694e-07,
+      "loss": 0.3854,
+      "step": 17789
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 47.008935748503276,
+      "learning_rate": 2.83760847167569e-07,
+      "loss": 0.4201,
+      "step": 17790
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 44.91870116590375,
+      "learning_rate": 2.8336209950522774e-07,
+      "loss": 0.4296,
+      "step": 17791
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 41.6746957551575,
+      "learning_rate": 2.829636281783843e-07,
+      "loss": 0.4141,
+      "step": 17792
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 34.12592884518323,
+      "learning_rate": 2.825654331983707e-07,
+      "loss": 0.362,
+      "step": 17793
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 44.06446738213494,
+      "learning_rate": 2.8216751457651124e-07,
+      "loss": 0.4092,
+      "step": 17794
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 43.126055596742354,
+      "learning_rate": 2.8176987232412354e-07,
+      "loss": 0.4195,
+      "step": 17795
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 56.01591925337582,
+      "learning_rate": 2.8137250645251636e-07,
+      "loss": 0.4717,
+      "step": 17796
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 43.48091325540388,
+      "learning_rate": 2.8097541697298835e-07,
+      "loss": 0.3683,
+      "step": 17797
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 39.21187764781412,
+      "learning_rate": 2.80578603896835e-07,
+      "loss": 0.3975,
+      "step": 17798
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 48.2062808594221,
+      "learning_rate": 2.8018206723533947e-07,
+      "loss": 0.4077,
+      "step": 17799
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 41.85178933893041,
+      "learning_rate": 2.7978580699978054e-07,
+      "loss": 0.3773,
+      "step": 17800
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 64.96839323399998,
+      "learning_rate": 2.7938982320142693e-07,
+      "loss": 0.4518,
+      "step": 17801
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 38.909443852115004,
+      "learning_rate": 2.789941158515408e-07,
+      "loss": 0.3913,
+      "step": 17802
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 44.95693679893782,
+      "learning_rate": 2.785986849613753e-07,
+      "loss": 0.3943,
+      "step": 17803
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 61.41506992509093,
+      "learning_rate": 2.78203530542176e-07,
+      "loss": 0.4609,
+      "step": 17804
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 45.87185883746865,
+      "learning_rate": 2.7780865260518265e-07,
+      "loss": 0.4017,
+      "step": 17805
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 48.08498893688479,
+      "learning_rate": 2.774140511616219e-07,
+      "loss": 0.3867,
+      "step": 17806
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 49.92589391252888,
+      "learning_rate": 2.770197262227181e-07,
+      "loss": 0.3819,
+      "step": 17807
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 46.71835917944091,
+      "learning_rate": 2.7662567779968677e-07,
+      "loss": 0.3594,
+      "step": 17808
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 48.5606311453224,
+      "learning_rate": 2.762319059037333e-07,
+      "loss": 0.395,
+      "step": 17809
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 62.64576600374241,
+      "learning_rate": 2.758384105460554e-07,
+      "loss": 0.5237,
+      "step": 17810
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 50.17921671556697,
+      "learning_rate": 2.7544519173784423e-07,
+      "loss": 0.4194,
+      "step": 17811
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 57.20506847240457,
+      "learning_rate": 2.7505224949028297e-07,
+      "loss": 0.4401,
+      "step": 17812
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 62.15725965720301,
+      "learning_rate": 2.7465958381454714e-07,
+      "loss": 0.4996,
+      "step": 17813
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 60.692029436291236,
+      "learning_rate": 2.7426719472180454e-07,
+      "loss": 0.4693,
+      "step": 17814
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 43.977547584895255,
+      "learning_rate": 2.738750822232128e-07,
+      "loss": 0.414,
+      "step": 17815
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 52.97851918922799,
+      "learning_rate": 2.7348324632992416e-07,
+      "loss": 0.4535,
+      "step": 17816
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 51.03529425046521,
+      "learning_rate": 2.7309168705308196e-07,
+      "loss": 0.4417,
+      "step": 17817
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 44.515156340081894,
+      "learning_rate": 2.7270040440382283e-07,
+      "loss": 0.3787,
+      "step": 17818
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 49.793805739078415,
+      "learning_rate": 2.723093983932734e-07,
+      "loss": 0.3915,
+      "step": 17819
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 47.15796478268131,
+      "learning_rate": 2.719186690325537e-07,
+      "loss": 0.4205,
+      "step": 17820
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 37.38225157186351,
+      "learning_rate": 2.7152821633277705e-07,
+      "loss": 0.3656,
+      "step": 17821
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 43.29052111211786,
+      "learning_rate": 2.7113804030504564e-07,
+      "loss": 0.3469,
+      "step": 17822
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 51.883303214453605,
+      "learning_rate": 2.7074814096045954e-07,
+      "loss": 0.4617,
+      "step": 17823
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 56.0796376617143,
+      "learning_rate": 2.7035851831010206e-07,
+      "loss": 0.4491,
+      "step": 17824
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 43.73363150523079,
+      "learning_rate": 2.699691723650577e-07,
+      "loss": 0.4057,
+      "step": 17825
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 63.21624989263664,
+      "learning_rate": 2.6958010313639873e-07,
+      "loss": 0.3943,
+      "step": 17826
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 64.69457826818471,
+      "learning_rate": 2.6919131063518956e-07,
+      "loss": 0.4366,
+      "step": 17827
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 62.943043181528665,
+      "learning_rate": 2.6880279487248697e-07,
+      "loss": 0.4403,
+      "step": 17828
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 53.12953807521589,
+      "learning_rate": 2.684145558593398e-07,
+      "loss": 0.414,
+      "step": 17829
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 43.30386102896977,
+      "learning_rate": 2.680265936067905e-07,
+      "loss": 0.3893,
+      "step": 17830
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 59.96970700230409,
+      "learning_rate": 2.6763890812587126e-07,
+      "loss": 0.4545,
+      "step": 17831
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 45.428889260204166,
+      "learning_rate": 2.6725149942760875e-07,
+      "loss": 0.3933,
+      "step": 17832
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 45.892606204747935,
+      "learning_rate": 2.668643675230198e-07,
+      "loss": 0.3878,
+      "step": 17833
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 56.26588747937379,
+      "learning_rate": 2.664775124231156e-07,
+      "loss": 0.4542,
+      "step": 17834
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 41.784082954361274,
+      "learning_rate": 2.660909341388951e-07,
+      "loss": 0.368,
+      "step": 17835
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 44.02321626189591,
+      "learning_rate": 2.65704632681355e-07,
+      "loss": 0.3646,
+      "step": 17836
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 55.46237368617656,
+      "learning_rate": 2.653186080614811e-07,
+      "loss": 0.4572,
+      "step": 17837
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 65.57167006548173,
+      "learning_rate": 2.649328602902501e-07,
+      "loss": 0.4313,
+      "step": 17838
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 46.06308788664571,
+      "learning_rate": 2.6454738937863545e-07,
+      "loss": 0.378,
+      "step": 17839
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 47.111230138687304,
+      "learning_rate": 2.6416219533759724e-07,
+      "loss": 0.4374,
+      "step": 17840
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 61.90116162789329,
+      "learning_rate": 2.637772781780901e-07,
+      "loss": 0.5306,
+      "step": 17841
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 51.08373733545367,
+      "learning_rate": 2.6339263791106294e-07,
+      "loss": 0.4287,
+      "step": 17842
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 49.53099890323375,
+      "learning_rate": 2.630082745474538e-07,
+      "loss": 0.3912,
+      "step": 17843
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 56.01556948947684,
+      "learning_rate": 2.6262418809819166e-07,
+      "loss": 0.4358,
+      "step": 17844
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 36.89358003923887,
+      "learning_rate": 2.6224037857420225e-07,
+      "loss": 0.3403,
+      "step": 17845
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 44.64414371859992,
+      "learning_rate": 2.6185684598640013e-07,
+      "loss": 0.3783,
+      "step": 17846
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 44.28620983551891,
+      "learning_rate": 2.614735903456933e-07,
+      "loss": 0.4055,
+      "step": 17847
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 48.42972804567966,
+      "learning_rate": 2.610906116629797e-07,
+      "loss": 0.3558,
+      "step": 17848
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 56.317850106445206,
+      "learning_rate": 2.607079099491516e-07,
+      "loss": 0.4076,
+      "step": 17849
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 85.2204098177879,
+      "learning_rate": 2.6032548521509384e-07,
+      "loss": 0.4918,
+      "step": 17850
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.45740381960505,
+      "learning_rate": 2.59943337471682e-07,
+      "loss": 0.4365,
+      "step": 17851
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 46.92212497315589,
+      "learning_rate": 2.595614667297841e-07,
+      "loss": 0.4157,
+      "step": 17852
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 42.151616341652336,
+      "learning_rate": 2.591798730002593e-07,
+      "loss": 0.3319,
+      "step": 17853
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 50.22229986931463,
+      "learning_rate": 2.587985562939599e-07,
+      "loss": 0.4262,
+      "step": 17854
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 44.88711578217451,
+      "learning_rate": 2.5841751662173287e-07,
+      "loss": 0.413,
+      "step": 17855
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 46.92424353821549,
+      "learning_rate": 2.5803675399441173e-07,
+      "loss": 0.3919,
+      "step": 17856
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 54.50564198644198,
+      "learning_rate": 2.576562684228279e-07,
+      "loss": 0.4072,
+      "step": 17857
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 48.67481975923699,
+      "learning_rate": 2.5727605991779925e-07,
+      "loss": 0.4075,
+      "step": 17858
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 54.97608177990765,
+      "learning_rate": 2.5689612849014166e-07,
+      "loss": 0.4521,
+      "step": 17859
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 60.82799081030605,
+      "learning_rate": 2.565164741506576e-07,
+      "loss": 0.4639,
+      "step": 17860
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 40.2862741904642,
+      "learning_rate": 2.56137096910144e-07,
+      "loss": 0.3968,
+      "step": 17861
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.954080316353355,
+      "learning_rate": 2.557579967793933e-07,
+      "loss": 0.4485,
+      "step": 17862
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 43.29951293354213,
+      "learning_rate": 2.5537917376918466e-07,
+      "loss": 0.4087,
+      "step": 17863
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 47.75845256110047,
+      "learning_rate": 2.5500062789029165e-07,
+      "loss": 0.3825,
+      "step": 17864
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 54.48594938497241,
+      "learning_rate": 2.546223591534802e-07,
+      "loss": 0.4232,
+      "step": 17865
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 62.65510969765335,
+      "learning_rate": 2.542443675695072e-07,
+      "loss": 0.5332,
+      "step": 17866
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 42.651610178700174,
+      "learning_rate": 2.538666531491241e-07,
+      "loss": 0.4113,
+      "step": 17867
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 50.117679303938324,
+      "learning_rate": 2.5348921590307216e-07,
+      "loss": 0.3559,
+      "step": 17868
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 73.07767808935355,
+      "learning_rate": 2.5311205584208523e-07,
+      "loss": 0.4688,
+      "step": 17869
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 49.08831334014597,
+      "learning_rate": 2.52735172976889e-07,
+      "loss": 0.4189,
+      "step": 17870
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 40.54516022277366,
+      "learning_rate": 2.5235856731820276e-07,
+      "loss": 0.4021,
+      "step": 17871
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.930745556114864,
+      "learning_rate": 2.519822388767379e-07,
+      "loss": 0.3669,
+      "step": 17872
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.44835478531778,
+      "learning_rate": 2.5160618766319477e-07,
+      "loss": 0.4422,
+      "step": 17873
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 44.547810876837225,
+      "learning_rate": 2.512304136882682e-07,
+      "loss": 0.359,
+      "step": 17874
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 42.27451884956347,
+      "learning_rate": 2.508549169626462e-07,
+      "loss": 0.3463,
+      "step": 17875
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 46.44357982384094,
+      "learning_rate": 2.504796974970081e-07,
+      "loss": 0.3732,
+      "step": 17876
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 43.62788096317928,
+      "learning_rate": 2.5010475530202414e-07,
+      "loss": 0.3809,
+      "step": 17877
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 57.696231342455384,
+      "learning_rate": 2.4973009038835593e-07,
+      "loss": 0.4912,
+      "step": 17878
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 48.03701290982513,
+      "learning_rate": 2.4935570276666157e-07,
+      "loss": 0.4083,
+      "step": 17879
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 55.16903528831413,
+      "learning_rate": 2.4898159244758693e-07,
+      "loss": 0.4491,
+      "step": 17880
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.059342738601806,
+      "learning_rate": 2.486077594417724e-07,
+      "loss": 0.4683,
+      "step": 17881
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 56.05805952503259,
+      "learning_rate": 2.482342037598473e-07,
+      "loss": 0.4388,
+      "step": 17882
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 46.64967701034515,
+      "learning_rate": 2.4786092541243645e-07,
+      "loss": 0.3876,
+      "step": 17883
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 47.458252891939004,
+      "learning_rate": 2.474879244101569e-07,
+      "loss": 0.425,
+      "step": 17884
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 57.076802987933945,
+      "learning_rate": 2.471152007636157e-07,
+      "loss": 0.4542,
+      "step": 17885
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.0007755834584,
+      "learning_rate": 2.4674275448341333e-07,
+      "loss": 0.4546,
+      "step": 17886
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 45.14882970119099,
+      "learning_rate": 2.463705855801413e-07,
+      "loss": 0.4526,
+      "step": 17887
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 45.67862068497648,
+      "learning_rate": 2.4599869406438327e-07,
+      "loss": 0.3484,
+      "step": 17888
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 48.77639497084869,
+      "learning_rate": 2.456270799467175e-07,
+      "loss": 0.4162,
+      "step": 17889
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 42.186976541167994,
+      "learning_rate": 2.452557432377101e-07,
+      "loss": 0.4164,
+      "step": 17890
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 53.57460213752437,
+      "learning_rate": 2.4488468394792463e-07,
+      "loss": 0.4442,
+      "step": 17891
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 47.353571652949434,
+      "learning_rate": 2.4451390208791173e-07,
+      "loss": 0.4375,
+      "step": 17892
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 47.659109181168134,
+      "learning_rate": 2.4414339766821614e-07,
+      "loss": 0.3699,
+      "step": 17893
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 43.98842130380754,
+      "learning_rate": 2.437731706993751e-07,
+      "loss": 0.3756,
+      "step": 17894
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 53.96235131090156,
+      "learning_rate": 2.434032211919179e-07,
+      "loss": 0.4516,
+      "step": 17895
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 58.642808807961146,
+      "learning_rate": 2.4303354915636735e-07,
+      "loss": 0.4987,
+      "step": 17896
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 61.07187227695062,
+      "learning_rate": 2.426641546032338e-07,
+      "loss": 0.5228,
+      "step": 17897
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 52.57152979667764,
+      "learning_rate": 2.422950375430233e-07,
+      "loss": 0.4352,
+      "step": 17898
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 66.75458645154575,
+      "learning_rate": 2.419261979862342e-07,
+      "loss": 0.4444,
+      "step": 17899
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 50.24164803686455,
+      "learning_rate": 2.415576359433558e-07,
+      "loss": 0.4311,
+      "step": 17900
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 51.22136287297043,
+      "learning_rate": 2.4118935142486975e-07,
+      "loss": 0.4297,
+      "step": 17901
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 43.089041167015765,
+      "learning_rate": 2.408213444412499e-07,
+      "loss": 0.4216,
+      "step": 17902
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 56.730252073310595,
+      "learning_rate": 2.404536150029624e-07,
+      "loss": 0.4362,
+      "step": 17903
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 43.90744698952908,
+      "learning_rate": 2.4008616312046315e-07,
+      "loss": 0.4125,
+      "step": 17904
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 44.262239910461496,
+      "learning_rate": 2.397189888042062e-07,
+      "loss": 0.4564,
+      "step": 17905
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 41.98280937475637,
+      "learning_rate": 2.3935209206463195e-07,
+      "loss": 0.3689,
+      "step": 17906
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 44.44804859640947,
+      "learning_rate": 2.389854729121721e-07,
+      "loss": 0.3969,
+      "step": 17907
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 61.88589808502549,
+      "learning_rate": 2.3861913135725613e-07,
+      "loss": 0.3944,
+      "step": 17908
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 49.89183571502623,
+      "learning_rate": 2.382530674103023e-07,
+      "loss": 0.4719,
+      "step": 17909
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 40.30232334468098,
+      "learning_rate": 2.3788728108172121e-07,
+      "loss": 0.4244,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 40.30278043455464,
+      "learning_rate": 2.3752177238191455e-07,
+      "loss": 0.3541,
+      "step": 17911
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 48.0234061831063,
+      "learning_rate": 2.371565413212773e-07,
+      "loss": 0.4695,
+      "step": 17912
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 56.856232449622254,
+      "learning_rate": 2.3679158791019673e-07,
+      "loss": 0.4789,
+      "step": 17913
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 62.25515195676509,
+      "learning_rate": 2.3642691215905344e-07,
+      "loss": 0.4573,
+      "step": 17914
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 52.66708402025669,
+      "learning_rate": 2.360625140782169e-07,
+      "loss": 0.501,
+      "step": 17915
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 46.01535308705739,
+      "learning_rate": 2.3569839367805103e-07,
+      "loss": 0.3796,
+      "step": 17916
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 42.07460527965873,
+      "learning_rate": 2.3533455096890978e-07,
+      "loss": 0.4144,
+      "step": 17917
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 49.21958102054519,
+      "learning_rate": 2.3497098596114265e-07,
+      "loss": 0.4295,
+      "step": 17918
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 52.609824885367125,
+      "learning_rate": 2.3460769866508693e-07,
+      "loss": 0.4473,
+      "step": 17919
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 45.3859854659547,
+      "learning_rate": 2.3424468909107656e-07,
+      "loss": 0.3911,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 51.42281116524796,
+      "learning_rate": 2.338819572494355e-07,
+      "loss": 0.4086,
+      "step": 17921
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 55.44485571056178,
+      "learning_rate": 2.3351950315047777e-07,
+      "loss": 0.4417,
+      "step": 17922
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 53.437525815009415,
+      "learning_rate": 2.3315732680451175e-07,
+      "loss": 0.4245,
+      "step": 17923
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 50.28026346076944,
+      "learning_rate": 2.3279542822183698e-07,
+      "loss": 0.3404,
+      "step": 17924
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 55.61345468783317,
+      "learning_rate": 2.3243380741274745e-07,
+      "loss": 0.3891,
+      "step": 17925
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 42.79248514045786,
+      "learning_rate": 2.3207246438752719e-07,
+      "loss": 0.3738,
+      "step": 17926
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 42.40901664192888,
+      "learning_rate": 2.3171139915645235e-07,
+      "loss": 0.3334,
+      "step": 17927
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 52.60331484517253,
+      "learning_rate": 2.3135061172979034e-07,
+      "loss": 0.4615,
+      "step": 17928
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 51.45343025364874,
+      "learning_rate": 2.3099010211780183e-07,
+      "loss": 0.3555,
+      "step": 17929
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 51.75347221170049,
+      "learning_rate": 2.306298703307408e-07,
+      "loss": 0.4176,
+      "step": 17930
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 44.745614012698645,
+      "learning_rate": 2.3026991637885242e-07,
+      "loss": 0.4216,
+      "step": 17931
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 47.597751827402305,
+      "learning_rate": 2.2991024027237075e-07,
+      "loss": 0.3882,
+      "step": 17932
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 43.1567213404258,
+      "learning_rate": 2.2955084202152644e-07,
+      "loss": 0.3824,
+      "step": 17933
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 35.540049577994004,
+      "learning_rate": 2.2919172163654134e-07,
+      "loss": 0.3264,
+      "step": 17934
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 63.13385876729309,
+      "learning_rate": 2.2883287912762842e-07,
+      "loss": 0.4898,
+      "step": 17935
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 54.916153828199775,
+      "learning_rate": 2.2847431450499169e-07,
+      "loss": 0.469,
+      "step": 17936
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 45.16248857638977,
+      "learning_rate": 2.281160277788297e-07,
+      "loss": 0.3788,
+      "step": 17937
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 50.1585884380978,
+      "learning_rate": 2.277580189593298e-07,
+      "loss": 0.4819,
+      "step": 17938
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 47.02951357858222,
+      "learning_rate": 2.2740028805667725e-07,
+      "loss": 0.4851,
+      "step": 17939
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 40.652084664994284,
+      "learning_rate": 2.2704283508104386e-07,
+      "loss": 0.3814,
+      "step": 17940
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 50.31370519770218,
+      "learning_rate": 2.2668566004259486e-07,
+      "loss": 0.4493,
+      "step": 17941
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 46.22437076295063,
+      "learning_rate": 2.2632876295148765e-07,
+      "loss": 0.4111,
+      "step": 17942
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 47.323850512360934,
+      "learning_rate": 2.2597214381787301e-07,
+      "loss": 0.4577,
+      "step": 17943
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 66.22477480172172,
+      "learning_rate": 2.2561580265189397e-07,
+      "loss": 0.4876,
+      "step": 17944
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 62.416061974801984,
+      "learning_rate": 2.2525973946368352e-07,
+      "loss": 0.4642,
+      "step": 17945
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 42.25114492975561,
+      "learning_rate": 2.2490395426336687e-07,
+      "loss": 0.3662,
+      "step": 17946
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 49.16141820901173,
+      "learning_rate": 2.2454844706106482e-07,
+      "loss": 0.3914,
+      "step": 17947
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 44.31473418873009,
+      "learning_rate": 2.2419321786688485e-07,
+      "loss": 0.3887,
+      "step": 17948
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 44.01707652186557,
+      "learning_rate": 2.2383826669093222e-07,
+      "loss": 0.4371,
+      "step": 17949
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 48.26856007146557,
+      "learning_rate": 2.2348359354330106e-07,
+      "loss": 0.4556,
+      "step": 17950
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 49.48477998019147,
+      "learning_rate": 2.2312919843407555e-07,
+      "loss": 0.3481,
+      "step": 17951
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 53.63048438279349,
+      "learning_rate": 2.227750813733376e-07,
+      "loss": 0.4232,
+      "step": 17952
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 55.26287375933998,
+      "learning_rate": 2.2242124237115693e-07,
+      "loss": 0.4338,
+      "step": 17953
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 46.671690816295545,
+      "learning_rate": 2.220676814375955e-07,
+      "loss": 0.4191,
+      "step": 17954
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 48.22445616269618,
+      "learning_rate": 2.217143985827097e-07,
+      "loss": 0.4317,
+      "step": 17955
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 53.032652261477935,
+      "learning_rate": 2.2136139381654708e-07,
+      "loss": 0.3657,
+      "step": 17956
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 58.789312293450315,
+      "learning_rate": 2.2100866714914514e-07,
+      "loss": 0.4966,
+      "step": 17957
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 43.262378591495526,
+      "learning_rate": 2.2065621859053587e-07,
+      "loss": 0.4377,
+      "step": 17958
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 48.28802943856566,
+      "learning_rate": 2.2030404815074346e-07,
+      "loss": 0.4572,
+      "step": 17959
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 50.88909698056945,
+      "learning_rate": 2.1995215583978435e-07,
+      "loss": 0.3411,
+      "step": 17960
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 51.56611938720419,
+      "learning_rate": 2.1960054166766386e-07,
+      "loss": 0.4539,
+      "step": 17961
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 49.83179360608374,
+      "learning_rate": 2.1924920564438178e-07,
+      "loss": 0.407,
+      "step": 17962
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 47.29640448052345,
+      "learning_rate": 2.1889814777993013e-07,
+      "loss": 0.3956,
+      "step": 17963
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 43.09837819918121,
+      "learning_rate": 2.1854736808429533e-07,
+      "loss": 0.3641,
+      "step": 17964
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 51.52570879067882,
+      "learning_rate": 2.1819686656745053e-07,
+      "loss": 0.3917,
+      "step": 17965
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 55.59207065869885,
+      "learning_rate": 2.1784664323936445e-07,
+      "loss": 0.4222,
+      "step": 17966
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 41.98097596856418,
+      "learning_rate": 2.1749669810999686e-07,
+      "loss": 0.39,
+      "step": 17967
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 46.7441842138691,
+      "learning_rate": 2.1714703118930203e-07,
+      "loss": 0.3747,
+      "step": 17968
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 54.72608861942684,
+      "learning_rate": 2.16797642487222e-07,
+      "loss": 0.457,
+      "step": 17969
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 41.543383497168435,
+      "learning_rate": 2.1644853201369444e-07,
+      "loss": 0.3953,
+      "step": 17970
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 50.01349586500238,
+      "learning_rate": 2.1609969977864687e-07,
+      "loss": 0.4078,
+      "step": 17971
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 37.48514828951679,
+      "learning_rate": 2.157511457920014e-07,
+      "loss": 0.3329,
+      "step": 17972
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 47.43576036060365,
+      "learning_rate": 2.1540287006366788e-07,
+      "loss": 0.3495,
+      "step": 17973
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 57.73691885458121,
+      "learning_rate": 2.1505487260355396e-07,
+      "loss": 0.4723,
+      "step": 17974
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 50.78912246951706,
+      "learning_rate": 2.147071534215561e-07,
+      "loss": 0.4679,
+      "step": 17975
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 65.97708906976536,
+      "learning_rate": 2.1435971252756093e-07,
+      "loss": 0.4031,
+      "step": 17976
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 52.22227323689686,
+      "learning_rate": 2.140125499314527e-07,
+      "loss": 0.4344,
+      "step": 17977
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 48.52038641276513,
+      "learning_rate": 2.1366566564310244e-07,
+      "loss": 0.3388,
+      "step": 17978
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 46.73810342387426,
+      "learning_rate": 2.1331905967237553e-07,
+      "loss": 0.4326,
+      "step": 17979
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 46.16544229327592,
+      "learning_rate": 2.1297273202912972e-07,
+      "loss": 0.3961,
+      "step": 17980
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 48.367745459260014,
+      "learning_rate": 2.1262668272321486e-07,
+      "loss": 0.4318,
+      "step": 17981
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 44.430798772414036,
+      "learning_rate": 2.1228091176447086e-07,
+      "loss": 0.3786,
+      "step": 17982
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 45.835639220522566,
+      "learning_rate": 2.1193541916273097e-07,
+      "loss": 0.4141,
+      "step": 17983
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 66.08546073282514,
+      "learning_rate": 2.11590204927824e-07,
+      "loss": 0.4268,
+      "step": 17984
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 42.04324952325704,
+      "learning_rate": 2.1124526906956545e-07,
+      "loss": 0.4046,
+      "step": 17985
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 49.390365054427924,
+      "learning_rate": 2.109006115977641e-07,
+      "loss": 0.4634,
+      "step": 17986
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 53.58606232871769,
+      "learning_rate": 2.1055623252222211e-07,
+      "loss": 0.4059,
+      "step": 17987
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 56.85522805165927,
+      "learning_rate": 2.1021213185273504e-07,
+      "loss": 0.5018,
+      "step": 17988
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 49.93550252440602,
+      "learning_rate": 2.0986830959908943e-07,
+      "loss": 0.3963,
+      "step": 17989
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 51.22642893829782,
+      "learning_rate": 2.0952476577106085e-07,
+      "loss": 0.3707,
+      "step": 17990
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 51.80886782073444,
+      "learning_rate": 2.0918150037842032e-07,
+      "loss": 0.4199,
+      "step": 17991
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 34.09265450344335,
+      "learning_rate": 2.088385134309312e-07,
+      "loss": 0.3862,
+      "step": 17992
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 48.726925318744705,
+      "learning_rate": 2.0849580493834675e-07,
+      "loss": 0.4001,
+      "step": 17993
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 53.49597406837026,
+      "learning_rate": 2.0815337491041476e-07,
+      "loss": 0.4528,
+      "step": 17994
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 52.54686602530855,
+      "learning_rate": 2.0781122335687075e-07,
+      "loss": 0.4267,
+      "step": 17995
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 70.64556889065929,
+      "learning_rate": 2.0746935028744808e-07,
+      "loss": 0.4514,
+      "step": 17996
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 40.471589119197304,
+      "learning_rate": 2.0712775571186895e-07,
+      "loss": 0.3601,
+      "step": 17997
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 53.64252658565815,
+      "learning_rate": 2.067864396398478e-07,
+      "loss": 0.4756,
+      "step": 17998
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 49.02073860672686,
+      "learning_rate": 2.0644540208109133e-07,
+      "loss": 0.3882,
+      "step": 17999
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 53.04603938707249,
+      "learning_rate": 2.0610464304529843e-07,
+      "loss": 0.4019,
+      "step": 18000
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 48.582809291104965,
+      "learning_rate": 2.0576416254216024e-07,
+      "loss": 0.4329,
+      "step": 18001
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 48.59327414423767,
+      "learning_rate": 2.0542396058136015e-07,
+      "loss": 0.3436,
+      "step": 18002
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 39.33260436648723,
+      "learning_rate": 2.0508403717257263e-07,
+      "loss": 0.3264,
+      "step": 18003
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 44.01027218164808,
+      "learning_rate": 2.047443923254655e-07,
+      "loss": 0.3601,
+      "step": 18004
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 52.69865304865097,
+      "learning_rate": 2.0440502604969659e-07,
+      "loss": 0.4553,
+      "step": 18005
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 46.42293875293058,
+      "learning_rate": 2.040659383549204e-07,
+      "loss": 0.3553,
+      "step": 18006
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 40.01797085238422,
+      "learning_rate": 2.0372712925077697e-07,
+      "loss": 0.3767,
+      "step": 18007
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 56.51338376132791,
+      "learning_rate": 2.0338859874690308e-07,
+      "loss": 0.4413,
+      "step": 18008
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 55.773441979958065,
+      "learning_rate": 2.030503468529288e-07,
+      "loss": 0.5535,
+      "step": 18009
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 54.545677458206484,
+      "learning_rate": 2.0271237357847085e-07,
+      "loss": 0.4046,
+      "step": 18010
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 42.123495556146004,
+      "learning_rate": 2.0237467893314044e-07,
+      "loss": 0.3725,
+      "step": 18011
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 40.30819617074518,
+      "learning_rate": 2.0203726292654324e-07,
+      "loss": 0.3158,
+      "step": 18012
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 46.44735151199299,
+      "learning_rate": 2.0170012556827379e-07,
+      "loss": 0.4561,
+      "step": 18013
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 56.59370250057458,
+      "learning_rate": 2.0136326686792217e-07,
+      "loss": 0.4461,
+      "step": 18014
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 49.79290262973388,
+      "learning_rate": 2.0102668683506743e-07,
+      "loss": 0.4414,
+      "step": 18015
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 49.80782533103304,
+      "learning_rate": 2.0069038547928078e-07,
+      "loss": 0.4663,
+      "step": 18016
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 55.23337411796493,
+      "learning_rate": 2.0035436281012678e-07,
+      "loss": 0.4522,
+      "step": 18017
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 58.3935134005885,
+      "learning_rate": 2.0001861883716335e-07,
+      "loss": 0.4425,
+      "step": 18018
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 46.68561634926247,
+      "learning_rate": 1.996831535699373e-07,
+      "loss": 0.4089,
+      "step": 18019
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 61.518480579081185,
+      "learning_rate": 1.9934796701798876e-07,
+      "loss": 0.4455,
+      "step": 18020
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 42.22327381352591,
+      "learning_rate": 1.9901305919085123e-07,
+      "loss": 0.3568,
+      "step": 18021
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 47.902819785022295,
+      "learning_rate": 1.9867843009804822e-07,
+      "loss": 0.4169,
+      "step": 18022
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 46.40775987628496,
+      "learning_rate": 1.9834407974909764e-07,
+      "loss": 0.3334,
+      "step": 18023
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 52.96838096107501,
+      "learning_rate": 1.9801000815350745e-07,
+      "loss": 0.4075,
+      "step": 18024
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 50.43291952707612,
+      "learning_rate": 1.9767621532077896e-07,
+      "loss": 0.4452,
+      "step": 18025
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 48.59725531786835,
+      "learning_rate": 1.9734270126040344e-07,
+      "loss": 0.3901,
+      "step": 18026
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 47.716176071647354,
+      "learning_rate": 1.9700946598186887e-07,
+      "loss": 0.4179,
+      "step": 18027
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 44.799689580701916,
+      "learning_rate": 1.966765094946499e-07,
+      "loss": 0.4267,
+      "step": 18028
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 44.38909845570596,
+      "learning_rate": 1.963438318082156e-07,
+      "loss": 0.4103,
+      "step": 18029
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 49.652438479007465,
+      "learning_rate": 1.960114329320284e-07,
+      "loss": 0.4648,
+      "step": 18030
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 49.77781185482955,
+      "learning_rate": 1.9567931287554075e-07,
+      "loss": 0.3707,
+      "step": 18031
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 45.3965355601566,
+      "learning_rate": 1.9534747164819733e-07,
+      "loss": 0.4276,
+      "step": 18032
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 45.212407635506636,
+      "learning_rate": 1.9501590925943726e-07,
+      "loss": 0.4297,
+      "step": 18033
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 41.71676849345695,
+      "learning_rate": 1.9468462571868852e-07,
+      "loss": 0.4106,
+      "step": 18034
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 54.919513027577175,
+      "learning_rate": 1.9435362103537357e-07,
+      "loss": 0.4117,
+      "step": 18035
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 51.69882745944198,
+      "learning_rate": 1.9402289521890495e-07,
+      "loss": 0.4138,
+      "step": 18036
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 50.006692981986824,
+      "learning_rate": 1.936924482786884e-07,
+      "loss": 0.4289,
+      "step": 18037
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 62.10251526000342,
+      "learning_rate": 1.9336228022412306e-07,
+      "loss": 0.3987,
+      "step": 18038
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 48.91153360899322,
+      "learning_rate": 1.9303239106459703e-07,
+      "loss": 0.3679,
+      "step": 18039
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 56.16134191660959,
+      "learning_rate": 1.927027808094939e-07,
+      "loss": 0.3984,
+      "step": 18040
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 45.89729683485373,
+      "learning_rate": 1.9237344946818614e-07,
+      "loss": 0.4253,
+      "step": 18041
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 57.28408409387321,
+      "learning_rate": 1.9204439705003964e-07,
+      "loss": 0.3966,
+      "step": 18042
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 32.88477564466777,
+      "learning_rate": 1.9171562356441353e-07,
+      "loss": 0.3234,
+      "step": 18043
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 54.34816984132606,
+      "learning_rate": 1.9138712902065703e-07,
+      "loss": 0.4013,
+      "step": 18044
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 52.02695169193012,
+      "learning_rate": 1.9105891342811268e-07,
+      "loss": 0.3909,
+      "step": 18045
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 76.16113494746651,
+      "learning_rate": 1.9073097679611518e-07,
+      "loss": 0.5069,
+      "step": 18046
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 43.58186832242639,
+      "learning_rate": 1.9040331913398934e-07,
+      "loss": 0.4209,
+      "step": 18047
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 42.28823626218684,
+      "learning_rate": 1.900759404510577e-07,
+      "loss": 0.4201,
+      "step": 18048
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 49.43403248255001,
+      "learning_rate": 1.8974884075662503e-07,
+      "loss": 0.3558,
+      "step": 18049
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 45.33251977580874,
+      "learning_rate": 1.894220200599961e-07,
+      "loss": 0.4346,
+      "step": 18050
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 38.2981068363617,
+      "learning_rate": 1.8909547837046683e-07,
+      "loss": 0.3387,
+      "step": 18051
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 54.50216476270838,
+      "learning_rate": 1.887692156973231e-07,
+      "loss": 0.4965,
+      "step": 18052
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 43.97605896521042,
+      "learning_rate": 1.8844323204984417e-07,
+      "loss": 0.3566,
+      "step": 18053
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 42.957425320954506,
+      "learning_rate": 1.8811752743729817e-07,
+      "loss": 0.37,
+      "step": 18054
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 40.08278011778275,
+      "learning_rate": 1.8779210186895102e-07,
+      "loss": 0.4047,
+      "step": 18055
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 48.370623351148545,
+      "learning_rate": 1.8746695535405645e-07,
+      "loss": 0.4422,
+      "step": 18056
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 64.29377935658627,
+      "learning_rate": 1.871420879018615e-07,
+      "loss": 0.4489,
+      "step": 18057
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 43.22550461421054,
+      "learning_rate": 1.8681749952160544e-07,
+      "loss": 0.3515,
+      "step": 18058
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 47.13506776740278,
+      "learning_rate": 1.8649319022251865e-07,
+      "loss": 0.4016,
+      "step": 18059
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 45.75946119587309,
+      "learning_rate": 1.8616916001382602e-07,
+      "loss": 0.4202,
+      "step": 18060
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 43.522117935743985,
+      "learning_rate": 1.8584540890474012e-07,
+      "loss": 0.4253,
+      "step": 18061
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 54.49063207012974,
+      "learning_rate": 1.855219369044703e-07,
+      "loss": 0.4484,
+      "step": 18062
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 49.03195262378757,
+      "learning_rate": 1.8519874402221472e-07,
+      "loss": 0.4243,
+      "step": 18063
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 53.49136188327345,
+      "learning_rate": 1.8487583026716604e-07,
+      "loss": 0.4905,
+      "step": 18064
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 42.53514431369882,
+      "learning_rate": 1.8455319564850694e-07,
+      "loss": 0.4027,
+      "step": 18065
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 43.09766962650548,
+      "learning_rate": 1.8423084017541227e-07,
+      "loss": 0.4148,
+      "step": 18066
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 63.66841162126105,
+      "learning_rate": 1.8390876385705136e-07,
+      "loss": 0.4813,
+      "step": 18067
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 39.352023066146145,
+      "learning_rate": 1.8358696670258357e-07,
+      "loss": 0.3895,
+      "step": 18068
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 53.686252410745794,
+      "learning_rate": 1.8326544872115936e-07,
+      "loss": 0.4811,
+      "step": 18069
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 51.569796971803,
+      "learning_rate": 1.829442099219225e-07,
+      "loss": 0.4366,
+      "step": 18070
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 43.60027701897941,
+      "learning_rate": 1.826232503140102e-07,
+      "loss": 0.4145,
+      "step": 18071
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 52.36591935020128,
+      "learning_rate": 1.823025699065495e-07,
+      "loss": 0.4233,
+      "step": 18072
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 47.32934823632587,
+      "learning_rate": 1.8198216870866093e-07,
+      "loss": 0.3991,
+      "step": 18073
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 52.69377691634501,
+      "learning_rate": 1.8166204672945496e-07,
+      "loss": 0.4368,
+      "step": 18074
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 60.29029754168533,
+      "learning_rate": 1.813422039780377e-07,
+      "loss": 0.4341,
+      "step": 18075
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 45.575545783653936,
+      "learning_rate": 1.8102264046350404e-07,
+      "loss": 0.3733,
+      "step": 18076
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 49.92341140508082,
+      "learning_rate": 1.8070335619494227e-07,
+      "loss": 0.3576,
+      "step": 18077
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 52.00280738551561,
+      "learning_rate": 1.8038435118143405e-07,
+      "loss": 0.4135,
+      "step": 18078
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 47.28399428366305,
+      "learning_rate": 1.8006562543204986e-07,
+      "loss": 0.3987,
+      "step": 18079
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 51.73914184545796,
+      "learning_rate": 1.797471789558547e-07,
+      "loss": 0.4028,
+      "step": 18080
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 51.42054627196227,
+      "learning_rate": 1.7942901176190574e-07,
+      "loss": 0.3968,
+      "step": 18081
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 51.80719441857636,
+      "learning_rate": 1.7911112385925022e-07,
+      "loss": 0.4476,
+      "step": 18082
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 53.7885910423522,
+      "learning_rate": 1.7879351525692978e-07,
+      "loss": 0.4672,
+      "step": 18083
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 64.26117558638451,
+      "learning_rate": 1.7847618596397498e-07,
+      "loss": 0.5121,
+      "step": 18084
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 47.62455640170742,
+      "learning_rate": 1.7815913598941303e-07,
+      "loss": 0.4031,
+      "step": 18085
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 45.803038251042125,
+      "learning_rate": 1.7784236534226006e-07,
+      "loss": 0.3955,
+      "step": 18086
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 47.107337596354036,
+      "learning_rate": 1.7752587403152444e-07,
+      "loss": 0.4152,
+      "step": 18087
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 53.98193161288849,
+      "learning_rate": 1.7720966206620672e-07,
+      "loss": 0.4726,
+      "step": 18088
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 44.642417099991036,
+      "learning_rate": 1.768937294552997e-07,
+      "loss": 0.3917,
+      "step": 18089
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 58.170013164324004,
+      "learning_rate": 1.7657807620778956e-07,
+      "loss": 0.4379,
+      "step": 18090
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 51.334996773450406,
+      "learning_rate": 1.762627023326513e-07,
+      "loss": 0.4357,
+      "step": 18091
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 43.26961361459757,
+      "learning_rate": 1.7594760783885557e-07,
+      "loss": 0.3972,
+      "step": 18092
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 41.98578608490793,
+      "learning_rate": 1.7563279273536294e-07,
+      "loss": 0.3883,
+      "step": 18093
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 43.49272757061786,
+      "learning_rate": 1.7531825703112736e-07,
+      "loss": 0.4094,
+      "step": 18094
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 46.39593291503926,
+      "learning_rate": 1.7500400073509172e-07,
+      "loss": 0.4126,
+      "step": 18095
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 54.82554118128076,
+      "learning_rate": 1.7469002385619548e-07,
+      "loss": 0.415,
+      "step": 18096
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 48.90632707388658,
+      "learning_rate": 1.7437632640336822e-07,
+      "loss": 0.3803,
+      "step": 18097
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 47.09735651008653,
+      "learning_rate": 1.7406290838552943e-07,
+      "loss": 0.4098,
+      "step": 18098
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 56.77237178946084,
+      "learning_rate": 1.737497698115942e-07,
+      "loss": 0.4531,
+      "step": 18099
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 65.02512958840755,
+      "learning_rate": 1.7343691069046654e-07,
+      "loss": 0.4985,
+      "step": 18100
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 62.14605933913559,
+      "learning_rate": 1.7312433103104486e-07,
+      "loss": 0.4839,
+      "step": 18101
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 49.96255876119182,
+      "learning_rate": 1.7281203084221876e-07,
+      "loss": 0.3619,
+      "step": 18102
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 49.63311016678623,
+      "learning_rate": 1.725000101328711e-07,
+      "loss": 0.3816,
+      "step": 18103
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 52.35700308112529,
+      "learning_rate": 1.7218826891187258e-07,
+      "loss": 0.4624,
+      "step": 18104
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 42.34973797565096,
+      "learning_rate": 1.7187680718809053e-07,
+      "loss": 0.4025,
+      "step": 18105
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 45.606942781401465,
+      "learning_rate": 1.7156562497038342e-07,
+      "loss": 0.3917,
+      "step": 18106
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 44.29705974129935,
+      "learning_rate": 1.7125472226760087e-07,
+      "loss": 0.3389,
+      "step": 18107
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 66.47735290380236,
+      "learning_rate": 1.7094409908858357e-07,
+      "loss": 0.4763,
+      "step": 18108
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 44.8129392812525,
+      "learning_rate": 1.7063375544216666e-07,
+      "loss": 0.4316,
+      "step": 18109
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 45.26331135134031,
+      "learning_rate": 1.703236913371753e-07,
+      "loss": 0.4084,
+      "step": 18110
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 37.37727266757017,
+      "learning_rate": 1.7001390678242803e-07,
+      "loss": 0.3489,
+      "step": 18111
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 51.937273761793605,
+      "learning_rate": 1.6970440178673552e-07,
+      "loss": 0.4258,
+      "step": 18112
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 51.740500757716006,
+      "learning_rate": 1.6939517635889857e-07,
+      "loss": 0.3841,
+      "step": 18113
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 48.36304527738457,
+      "learning_rate": 1.6908623050771234e-07,
+      "loss": 0.3855,
+      "step": 18114
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 55.05268488338478,
+      "learning_rate": 1.6877756424196312e-07,
+      "loss": 0.4285,
+      "step": 18115
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 55.73766596740118,
+      "learning_rate": 1.6846917757042836e-07,
+      "loss": 0.3937,
+      "step": 18116
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 51.12163080711767,
+      "learning_rate": 1.6816107050187992e-07,
+      "loss": 0.4001,
+      "step": 18117
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 50.2160348116657,
+      "learning_rate": 1.6785324304507745e-07,
+      "loss": 0.4363,
+      "step": 18118
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 47.40003004113384,
+      "learning_rate": 1.675456952087795e-07,
+      "loss": 0.3902,
+      "step": 18119
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 53.17293387677803,
+      "learning_rate": 1.67238427001728e-07,
+      "loss": 0.4481,
+      "step": 18120
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 49.96265098848641,
+      "learning_rate": 1.669314384326648e-07,
+      "loss": 0.4112,
+      "step": 18121
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 56.06353962932632,
+      "learning_rate": 1.666247295103185e-07,
+      "loss": 0.4511,
+      "step": 18122
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 48.41673857683016,
+      "learning_rate": 1.6631830024341323e-07,
+      "loss": 0.4653,
+      "step": 18123
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 51.87192421952236,
+      "learning_rate": 1.66012150640662e-07,
+      "loss": 0.3907,
+      "step": 18124
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 49.51305740262267,
+      "learning_rate": 1.6570628071077343e-07,
+      "loss": 0.4217,
+      "step": 18125
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 52.63818241384529,
+      "learning_rate": 1.65400690462445e-07,
+      "loss": 0.4521,
+      "step": 18126
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 52.717458156276564,
+      "learning_rate": 1.6509537990436753e-07,
+      "loss": 0.4861,
+      "step": 18127
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 56.75855446925827,
+      "learning_rate": 1.6479034904522518e-07,
+      "loss": 0.5005,
+      "step": 18128
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 50.004433525934516,
+      "learning_rate": 1.6448559789369102e-07,
+      "loss": 0.4347,
+      "step": 18129
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 42.52554080042907,
+      "learning_rate": 1.6418112645843253e-07,
+      "loss": 0.388,
+      "step": 18130
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 49.190851448787846,
+      "learning_rate": 1.638769347481095e-07,
+      "loss": 0.4425,
+      "step": 18131
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 48.753835077312516,
+      "learning_rate": 1.6357302277137388e-07,
+      "loss": 0.4627,
+      "step": 18132
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 54.933306502517766,
+      "learning_rate": 1.632693905368654e-07,
+      "loss": 0.4352,
+      "step": 18133
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 52.9083482356653,
+      "learning_rate": 1.6296603805322163e-07,
+      "loss": 0.5279,
+      "step": 18134
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 39.58491172387507,
+      "learning_rate": 1.62662965329069e-07,
+      "loss": 0.389,
+      "step": 18135
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 52.86261043631762,
+      "learning_rate": 1.6236017237302838e-07,
+      "loss": 0.4067,
+      "step": 18136
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 44.61394033503791,
+      "learning_rate": 1.6205765919370843e-07,
+      "loss": 0.4443,
+      "step": 18137
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 50.70656594942415,
+      "learning_rate": 1.6175542579971336e-07,
+      "loss": 0.3844,
+      "step": 18138
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 47.3094703833109,
+      "learning_rate": 1.6145347219963857e-07,
+      "loss": 0.456,
+      "step": 18139
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 55.103717328591,
+      "learning_rate": 1.6115179840207274e-07,
+      "loss": 0.4376,
+      "step": 18140
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 49.86404157059625,
+      "learning_rate": 1.608504044155934e-07,
+      "loss": 0.4797,
+      "step": 18141
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 38.58911036661305,
+      "learning_rate": 1.6054929024877375e-07,
+      "loss": 0.4256,
+      "step": 18142
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 47.34967786483995,
+      "learning_rate": 1.6024845591017468e-07,
+      "loss": 0.4098,
+      "step": 18143
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 58.42759695749447,
+      "learning_rate": 1.599479014083549e-07,
+      "loss": 0.4443,
+      "step": 18144
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 60.61345627986798,
+      "learning_rate": 1.5964762675185984e-07,
+      "loss": 0.4596,
+      "step": 18145
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 57.67141620925813,
+      "learning_rate": 1.5934763194923043e-07,
+      "loss": 0.4303,
+      "step": 18146
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 42.138691717432685,
+      "learning_rate": 1.5904791700899758e-07,
+      "loss": 0.401,
+      "step": 18147
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 42.934842900625924,
+      "learning_rate": 1.587484819396856e-07,
+      "loss": 0.3985,
+      "step": 18148
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 53.20104812847997,
+      "learning_rate": 1.5844932674980885e-07,
+      "loss": 0.4256,
+      "step": 18149
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 43.964599977458896,
+      "learning_rate": 1.58150451447876e-07,
+      "loss": 0.361,
+      "step": 18150
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 45.00749150542294,
+      "learning_rate": 1.5785185604238805e-07,
+      "loss": 0.4268,
+      "step": 18151
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 41.917424167945086,
+      "learning_rate": 1.5755354054183491e-07,
+      "loss": 0.4149,
+      "step": 18152
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 53.13112223634166,
+      "learning_rate": 1.57255504954702e-07,
+      "loss": 0.4825,
+      "step": 18153
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 48.346693944280716,
+      "learning_rate": 1.5695774928946473e-07,
+      "loss": 0.4138,
+      "step": 18154
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 54.825683639411,
+      "learning_rate": 1.5666027355458967e-07,
+      "loss": 0.4475,
+      "step": 18155
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 54.14310821519204,
+      "learning_rate": 1.563630777585401e-07,
+      "loss": 0.3893,
+      "step": 18156
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 43.995275254600784,
+      "learning_rate": 1.5606616190976587e-07,
+      "loss": 0.3775,
+      "step": 18157
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 44.33477009450461,
+      "learning_rate": 1.5576952601671135e-07,
+      "loss": 0.3746,
+      "step": 18158
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 49.145436978921566,
+      "learning_rate": 1.5547317008781205e-07,
+      "loss": 0.3639,
+      "step": 18159
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.3091843727001,
+      "learning_rate": 1.5517709413149674e-07,
+      "loss": 0.4608,
+      "step": 18160
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 44.40888292747191,
+      "learning_rate": 1.5488129815618757e-07,
+      "loss": 0.4296,
+      "step": 18161
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 42.7584179738308,
+      "learning_rate": 1.5458578217029342e-07,
+      "loss": 0.4161,
+      "step": 18162
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.017671372569914,
+      "learning_rate": 1.5429054618222083e-07,
+      "loss": 0.3792,
+      "step": 18163
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.25784700820932,
+      "learning_rate": 1.5399559020036537e-07,
+      "loss": 0.4881,
+      "step": 18164
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 46.871191943301284,
+      "learning_rate": 1.5370091423311584e-07,
+      "loss": 0.4009,
+      "step": 18165
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 52.15412776919464,
+      "learning_rate": 1.5340651828885223e-07,
+      "loss": 0.4465,
+      "step": 18166
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 49.22983225231118,
+      "learning_rate": 1.531124023759467e-07,
+      "loss": 0.4098,
+      "step": 18167
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 52.077602018745615,
+      "learning_rate": 1.5281856650276482e-07,
+      "loss": 0.4244,
+      "step": 18168
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 50.738559753560516,
+      "learning_rate": 1.5252501067766324e-07,
+      "loss": 0.4391,
+      "step": 18169
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 45.0428773956656,
+      "learning_rate": 1.5223173490898856e-07,
+      "loss": 0.3893,
+      "step": 18170
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 40.3537396710515,
+      "learning_rate": 1.5193873920508306e-07,
+      "loss": 0.4217,
+      "step": 18171
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 38.94559242201503,
+      "learning_rate": 1.5164602357427781e-07,
+      "loss": 0.3985,
+      "step": 18172
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 48.492476582760204,
+      "learning_rate": 1.5135358802489952e-07,
+      "loss": 0.3717,
+      "step": 18173
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 46.09553001975489,
+      "learning_rate": 1.5106143256526374e-07,
+      "loss": 0.3959,
+      "step": 18174
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 57.45968704989514,
+      "learning_rate": 1.5076955720367937e-07,
+      "loss": 0.497,
+      "step": 18175
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 46.95500745506581,
+      "learning_rate": 1.5047796194844754e-07,
+      "loss": 0.4393,
+      "step": 18176
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 41.440099453410916,
+      "learning_rate": 1.5018664680785943e-07,
+      "loss": 0.3722,
+      "step": 18177
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 51.650188249612384,
+      "learning_rate": 1.4989561179020283e-07,
+      "loss": 0.4737,
+      "step": 18178
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 52.21630128589978,
+      "learning_rate": 1.4960485690375115e-07,
+      "loss": 0.4391,
+      "step": 18179
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.88040677408524,
+      "learning_rate": 1.4931438215677663e-07,
+      "loss": 0.3956,
+      "step": 18180
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 52.639029257035574,
+      "learning_rate": 1.490241875575371e-07,
+      "loss": 0.4497,
+      "step": 18181
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 56.914703135134694,
+      "learning_rate": 1.4873427311428822e-07,
+      "loss": 0.4617,
+      "step": 18182
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 43.027632495467124,
+      "learning_rate": 1.484446388352734e-07,
+      "loss": 0.4034,
+      "step": 18183
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 46.18652672574554,
+      "learning_rate": 1.481552847287293e-07,
+      "loss": 0.3756,
+      "step": 18184
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 51.7634007891182,
+      "learning_rate": 1.478662108028872e-07,
+      "loss": 0.4133,
+      "step": 18185
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 45.923736534535124,
+      "learning_rate": 1.4757741706596605e-07,
+      "loss": 0.4321,
+      "step": 18186
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 46.053045195335464,
+      "learning_rate": 1.472889035261793e-07,
+      "loss": 0.4307,
+      "step": 18187
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 42.256249264707364,
+      "learning_rate": 1.4700067019173258e-07,
+      "loss": 0.3929,
+      "step": 18188
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 39.33261565924946,
+      "learning_rate": 1.4671271707082268e-07,
+      "loss": 0.3354,
+      "step": 18189
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 37.1737031972126,
+      "learning_rate": 1.4642504417163973e-07,
+      "loss": 0.3169,
+      "step": 18190
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 55.39867320055643,
+      "learning_rate": 1.4613765150236382e-07,
+      "loss": 0.447,
+      "step": 18191
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 42.78603680666732,
+      "learning_rate": 1.4585053907116953e-07,
+      "loss": 0.3714,
+      "step": 18192
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.5523262973134,
+      "learning_rate": 1.4556370688622036e-07,
+      "loss": 0.4124,
+      "step": 18193
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 56.82535424346534,
+      "learning_rate": 1.4527715495567529e-07,
+      "loss": 0.4255,
+      "step": 18194
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 41.62772755091995,
+      "learning_rate": 1.4499088328768341e-07,
+      "loss": 0.3674,
+      "step": 18195
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 45.568946884623585,
+      "learning_rate": 1.4470489189038595e-07,
+      "loss": 0.3823,
+      "step": 18196
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 48.47784831810874,
+      "learning_rate": 1.444191807719153e-07,
+      "loss": 0.3834,
+      "step": 18197
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.998668920419234,
+      "learning_rate": 1.4413374994039718e-07,
+      "loss": 0.3609,
+      "step": 18198
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 54.80553539052546,
+      "learning_rate": 1.438485994039518e-07,
+      "loss": 0.4064,
+      "step": 18199
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 42.43404628593476,
+      "learning_rate": 1.4356372917068595e-07,
+      "loss": 0.3659,
+      "step": 18200
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 58.33560684496613,
+      "learning_rate": 1.4327913924870097e-07,
+      "loss": 0.393,
+      "step": 18201
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 55.41710727296597,
+      "learning_rate": 1.4299482964609035e-07,
+      "loss": 0.4481,
+      "step": 18202
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 50.19965147034225,
+      "learning_rate": 1.427108003709421e-07,
+      "loss": 0.423,
+      "step": 18203
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 52.581941526747464,
+      "learning_rate": 1.4242705143133195e-07,
+      "loss": 0.3786,
+      "step": 18204
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.29105966769416,
+      "learning_rate": 1.4214358283533124e-07,
+      "loss": 0.3641,
+      "step": 18205
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 41.78434166208314,
+      "learning_rate": 1.41860394590998e-07,
+      "loss": 0.4212,
+      "step": 18206
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 51.273440485263535,
+      "learning_rate": 1.4157748670639015e-07,
+      "loss": 0.479,
+      "step": 18207
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 56.54672899778121,
+      "learning_rate": 1.412948591895502e-07,
+      "loss": 0.4426,
+      "step": 18208
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 43.9222901717207,
+      "learning_rate": 1.4101251204851841e-07,
+      "loss": 0.4497,
+      "step": 18209
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 48.37466722319729,
+      "learning_rate": 1.4073044529132273e-07,
+      "loss": 0.4366,
+      "step": 18210
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 48.86212703809392,
+      "learning_rate": 1.4044865892598681e-07,
+      "loss": 0.4514,
+      "step": 18211
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.73202719151208,
+      "learning_rate": 1.4016715296052198e-07,
+      "loss": 0.4067,
+      "step": 18212
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 51.18214870143198,
+      "learning_rate": 1.3988592740293629e-07,
+      "loss": 0.4226,
+      "step": 18213
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 52.81128956688587,
+      "learning_rate": 1.3960498226122554e-07,
+      "loss": 0.4255,
+      "step": 18214
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 42.562213627864345,
+      "learning_rate": 1.3932431754338227e-07,
+      "loss": 0.4022,
+      "step": 18215
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 50.638981435552495,
+      "learning_rate": 1.3904393325738785e-07,
+      "loss": 0.4876,
+      "step": 18216
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 55.8366388063648,
+      "learning_rate": 1.387638294112137e-07,
+      "loss": 0.5011,
+      "step": 18217
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 46.062647049357295,
+      "learning_rate": 1.3848400601282784e-07,
+      "loss": 0.3724,
+      "step": 18218
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 57.04688184158756,
+      "learning_rate": 1.3820446307018843e-07,
+      "loss": 0.4988,
+      "step": 18219
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 52.34753752712692,
+      "learning_rate": 1.379252005912457e-07,
+      "loss": 0.4275,
+      "step": 18220
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 58.982694738846966,
+      "learning_rate": 1.3764621858394e-07,
+      "loss": 0.4248,
+      "step": 18221
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 60.64596177531664,
+      "learning_rate": 1.373675170562072e-07,
+      "loss": 0.4451,
+      "step": 18222
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 59.505769409828865,
+      "learning_rate": 1.370890960159721e-07,
+      "loss": 0.5309,
+      "step": 18223
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 36.16459024455882,
+      "learning_rate": 1.3681095547115497e-07,
+      "loss": 0.3709,
+      "step": 18224
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 53.96173263014793,
+      "learning_rate": 1.3653309542966286e-07,
+      "loss": 0.4494,
+      "step": 18225
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 43.67144477645974,
+      "learning_rate": 1.3625551589940056e-07,
+      "loss": 0.338,
+      "step": 18226
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 39.60837238555167,
+      "learning_rate": 1.3597821688826063e-07,
+      "loss": 0.3534,
+      "step": 18227
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 60.15692066561896,
+      "learning_rate": 1.357011984041301e-07,
+      "loss": 0.4547,
+      "step": 18228
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 53.73702983947785,
+      "learning_rate": 1.3542446045488712e-07,
+      "loss": 0.4233,
+      "step": 18229
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 54.637456708797984,
+      "learning_rate": 1.3514800304840203e-07,
+      "loss": 0.4251,
+      "step": 18230
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 46.356476006864064,
+      "learning_rate": 1.3487182619253636e-07,
+      "loss": 0.4321,
+      "step": 18231
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 48.587843380817006,
+      "learning_rate": 1.3459592989514604e-07,
+      "loss": 0.4735,
+      "step": 18232
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 52.91519432823027,
+      "learning_rate": 1.3432031416407476e-07,
+      "loss": 0.3883,
+      "step": 18233
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 60.23270424431811,
+      "learning_rate": 1.3404497900716406e-07,
+      "loss": 0.4755,
+      "step": 18234
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 61.59452697790277,
+      "learning_rate": 1.3376992443224213e-07,
+      "loss": 0.4668,
+      "step": 18235
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 51.858820724628494,
+      "learning_rate": 1.3349515044713157e-07,
+      "loss": 0.4615,
+      "step": 18236
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 62.88216881049932,
+      "learning_rate": 1.3322065705964838e-07,
+      "loss": 0.51,
+      "step": 18237
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 43.829238931844294,
+      "learning_rate": 1.3294644427759628e-07,
+      "loss": 0.4455,
+      "step": 18238
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 60.62431384612662,
+      "learning_rate": 1.3267251210877686e-07,
+      "loss": 0.4935,
+      "step": 18239
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 45.838730280469,
+      "learning_rate": 1.3239886056097718e-07,
+      "loss": 0.3986,
+      "step": 18240
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 47.773783178318695,
+      "learning_rate": 1.3212548964198324e-07,
+      "loss": 0.4029,
+      "step": 18241
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 46.403666938793926,
+      "learning_rate": 1.3185239935956662e-07,
+      "loss": 0.41,
+      "step": 18242
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 56.51502189704655,
+      "learning_rate": 1.3157958972149553e-07,
+      "loss": 0.4059,
+      "step": 18243
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 51.2081451422904,
+      "learning_rate": 1.3130706073552824e-07,
+      "loss": 0.4122,
+      "step": 18244
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 46.91877882450835,
+      "learning_rate": 1.310348124094152e-07,
+      "loss": 0.396,
+      "step": 18245
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 51.58569508734915,
+      "learning_rate": 1.3076284475089906e-07,
+      "loss": 0.4421,
+      "step": 18246
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 51.18322502235738,
+      "learning_rate": 1.304911577677137e-07,
+      "loss": 0.4215,
+      "step": 18247
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 59.883113550274224,
+      "learning_rate": 1.3021975146758514e-07,
+      "loss": 0.3849,
+      "step": 18248
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 52.941152975787766,
+      "learning_rate": 1.2994862585823608e-07,
+      "loss": 0.4892,
+      "step": 18249
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 46.29550279585087,
+      "learning_rate": 1.296777809473726e-07,
+      "loss": 0.3347,
+      "step": 18250
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 49.24057371680723,
+      "learning_rate": 1.294072167426985e-07,
+      "loss": 0.3977,
+      "step": 18251
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 53.16046788313993,
+      "learning_rate": 1.291369332519099e-07,
+      "loss": 0.4052,
+      "step": 18252
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 47.43060485324684,
+      "learning_rate": 1.2886693048269284e-07,
+      "loss": 0.4408,
+      "step": 18253
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 39.97999693084789,
+      "learning_rate": 1.285972084427256e-07,
+      "loss": 0.3474,
+      "step": 18254
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 48.09673047760625,
+      "learning_rate": 1.2832776713967876e-07,
+      "loss": 0.5026,
+      "step": 18255
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 47.596925941346356,
+      "learning_rate": 1.2805860658121505e-07,
+      "loss": 0.4521,
+      "step": 18256
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 43.03320577328247,
+      "learning_rate": 1.2778972677499057e-07,
+      "loss": 0.38,
+      "step": 18257
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 56.47903458502323,
+      "learning_rate": 1.2752112772865144e-07,
+      "loss": 0.3687,
+      "step": 18258
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 52.07440716153748,
+      "learning_rate": 1.27252809449836e-07,
+      "loss": 0.4336,
+      "step": 18259
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 49.687640884487585,
+      "learning_rate": 1.2698477194617475e-07,
+      "loss": 0.4281,
+      "step": 18260
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 53.937020804706606,
+      "learning_rate": 1.2671701522529055e-07,
+      "loss": 0.4101,
+      "step": 18261
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 50.687374886241926,
+      "learning_rate": 1.2644953929479954e-07,
+      "loss": 0.4473,
+      "step": 18262
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 48.9107552887684,
+      "learning_rate": 1.2618234416230778e-07,
+      "loss": 0.4139,
+      "step": 18263
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 51.14386237154385,
+      "learning_rate": 1.2591542983541371e-07,
+      "loss": 0.4504,
+      "step": 18264
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 53.88788302450696,
+      "learning_rate": 1.25648796321709e-07,
+      "loss": 0.4421,
+      "step": 18265
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 73.97314313728417,
+      "learning_rate": 1.253824436287765e-07,
+      "loss": 0.5269,
+      "step": 18266
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 48.27242495179757,
+      "learning_rate": 1.2511637176419012e-07,
+      "loss": 0.4113,
+      "step": 18267
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 45.02420970513351,
+      "learning_rate": 1.2485058073551825e-07,
+      "loss": 0.3748,
+      "step": 18268
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 55.857881740004395,
+      "learning_rate": 1.2458507055031822e-07,
+      "loss": 0.4586,
+      "step": 18269
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 56.09210794749724,
+      "learning_rate": 1.2431984121614282e-07,
+      "loss": 0.423,
+      "step": 18270
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 40.64769939373467,
+      "learning_rate": 1.2405489274053273e-07,
+      "loss": 0.3851,
+      "step": 18271
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 52.070062900869324,
+      "learning_rate": 1.2379022513102524e-07,
+      "loss": 0.4157,
+      "step": 18272
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 53.491048768153306,
+      "learning_rate": 1.2352583839514655e-07,
+      "loss": 0.4512,
+      "step": 18273
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 41.535277034340304,
+      "learning_rate": 1.2326173254041506e-07,
+      "loss": 0.4006,
+      "step": 18274
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 43.465971895649645,
+      "learning_rate": 1.229979075743415e-07,
+      "loss": 0.3704,
+      "step": 18275
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 49.973923201264526,
+      "learning_rate": 1.227343635044298e-07,
+      "loss": 0.3973,
+      "step": 18276
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 60.524920926836145,
+      "learning_rate": 1.224711003381751e-07,
+      "loss": 0.4509,
+      "step": 18277
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 57.424084264074295,
+      "learning_rate": 1.2220811808306477e-07,
+      "loss": 0.4841,
+      "step": 18278
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 52.55173945646817,
+      "learning_rate": 1.219454167465761e-07,
+      "loss": 0.4484,
+      "step": 18279
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 60.18287526046761,
+      "learning_rate": 1.2168299633618209e-07,
+      "loss": 0.499,
+      "step": 18280
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 57.986232665106506,
+      "learning_rate": 1.2142085685934446e-07,
+      "loss": 0.4796,
+      "step": 18281
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 50.67556761944565,
+      "learning_rate": 1.2115899832351952e-07,
+      "loss": 0.4089,
+      "step": 18282
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.58515670754572,
+      "learning_rate": 1.2089742073615352e-07,
+      "loss": 0.4262,
+      "step": 18283
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.2716423101064,
+      "learning_rate": 1.2063612410468496e-07,
+      "loss": 0.4089,
+      "step": 18284
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 51.515141793490656,
+      "learning_rate": 1.2037510843654565e-07,
+      "loss": 0.4379,
+      "step": 18285
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 50.50390322869122,
+      "learning_rate": 1.201143737391597e-07,
+      "loss": 0.4458,
+      "step": 18286
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 48.04178675559068,
+      "learning_rate": 1.1985392001994112e-07,
+      "loss": 0.4281,
+      "step": 18287
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 53.243097937028196,
+      "learning_rate": 1.1959374728629847e-07,
+      "loss": 0.4423,
+      "step": 18288
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 46.618382932928824,
+      "learning_rate": 1.1933385554562803e-07,
+      "loss": 0.4037,
+      "step": 18289
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 41.86993786404748,
+      "learning_rate": 1.1907424480532282e-07,
+      "loss": 0.3573,
+      "step": 18290
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 57.83081991723621,
+      "learning_rate": 1.1881491507276688e-07,
+      "loss": 0.4867,
+      "step": 18291
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 51.754075700711525,
+      "learning_rate": 1.1855586635533434e-07,
+      "loss": 0.4445,
+      "step": 18292
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 48.49467401166241,
+      "learning_rate": 1.1829709866039263e-07,
+      "loss": 0.4108,
+      "step": 18293
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 50.574417067905216,
+      "learning_rate": 1.180386119952992e-07,
+      "loss": 0.3834,
+      "step": 18294
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 43.519983157626214,
+      "learning_rate": 1.1778040636740928e-07,
+      "loss": 0.3112,
+      "step": 18295
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 51.43248834680222,
+      "learning_rate": 1.1752248178406144e-07,
+      "loss": 0.3851,
+      "step": 18296
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 44.98923829989985,
+      "learning_rate": 1.1726483825259427e-07,
+      "loss": 0.426,
+      "step": 18297
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 66.07753782446686,
+      "learning_rate": 1.1700747578033412e-07,
+      "loss": 0.387,
+      "step": 18298
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 44.19304406378171,
+      "learning_rate": 1.1675039437459956e-07,
+      "loss": 0.43,
+      "step": 18299
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 40.978248688048524,
+      "learning_rate": 1.1649359404270144e-07,
+      "loss": 0.3391,
+      "step": 18300
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 41.968197555484956,
+      "learning_rate": 1.1623707479194501e-07,
+      "loss": 0.3685,
+      "step": 18301
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 45.78157726600511,
+      "learning_rate": 1.1598083662962335e-07,
+      "loss": 0.4165,
+      "step": 18302
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 35.65426420253655,
+      "learning_rate": 1.1572487956302613e-07,
+      "loss": 0.3401,
+      "step": 18303
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 36.944131781913,
+      "learning_rate": 1.1546920359943092e-07,
+      "loss": 0.3842,
+      "step": 18304
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 41.86242026215799,
+      "learning_rate": 1.1521380874610854e-07,
+      "loss": 0.3993,
+      "step": 18305
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 53.17326074723123,
+      "learning_rate": 1.1495869501032209e-07,
+      "loss": 0.414,
+      "step": 18306
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.95890561957428,
+      "learning_rate": 1.1470386239932907e-07,
+      "loss": 0.3771,
+      "step": 18307
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 51.432966146046375,
+      "learning_rate": 1.1444931092037592e-07,
+      "loss": 0.4133,
+      "step": 18308
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 42.24487727896543,
+      "learning_rate": 1.141950405807013e-07,
+      "loss": 0.3719,
+      "step": 18309
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 49.055254358672855,
+      "learning_rate": 1.1394105138753498e-07,
+      "loss": 0.4238,
+      "step": 18310
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 61.07612299376612,
+      "learning_rate": 1.1368734334810339e-07,
+      "loss": 0.4586,
+      "step": 18311
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.249943697860516,
+      "learning_rate": 1.1343391646962076e-07,
+      "loss": 0.4334,
+      "step": 18312
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 40.72547897691811,
+      "learning_rate": 1.1318077075929357e-07,
+      "loss": 0.3881,
+      "step": 18313
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.5465642973457,
+      "learning_rate": 1.129279062243216e-07,
+      "loss": 0.4422,
+      "step": 18314
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 41.49679538694177,
+      "learning_rate": 1.1267532287189576e-07,
+      "loss": 0.3645,
+      "step": 18315
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 45.81369806729336,
+      "learning_rate": 1.1242302070920141e-07,
+      "loss": 0.3552,
+      "step": 18316
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 41.63383713125479,
+      "learning_rate": 1.1217099974341173e-07,
+      "loss": 0.3983,
+      "step": 18317
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 45.00492344519498,
+      "learning_rate": 1.1191925998169317e-07,
+      "loss": 0.4387,
+      "step": 18318
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.48647217608476,
+      "learning_rate": 1.1166780143120781e-07,
+      "loss": 0.4391,
+      "step": 18319
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 46.47481779952345,
+      "learning_rate": 1.1141662409910547e-07,
+      "loss": 0.3822,
+      "step": 18320
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 45.304860669426894,
+      "learning_rate": 1.1116572799252934e-07,
+      "loss": 0.3617,
+      "step": 18321
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 53.6767132972929,
+      "learning_rate": 1.109151131186148e-07,
+      "loss": 0.4578,
+      "step": 18322
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 59.88315725805846,
+      "learning_rate": 1.106647794844895e-07,
+      "loss": 0.4534,
+      "step": 18323
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 54.305260240489524,
+      "learning_rate": 1.104147270972733e-07,
+      "loss": 0.4165,
+      "step": 18324
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 40.99841716083572,
+      "learning_rate": 1.1016495596407717e-07,
+      "loss": 0.3577,
+      "step": 18325
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 54.90432563062583,
+      "learning_rate": 1.0991546609200432e-07,
+      "loss": 0.42,
+      "step": 18326
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 41.6457323846298,
+      "learning_rate": 1.0966625748815019e-07,
+      "loss": 0.3908,
+      "step": 18327
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 49.4344690891207,
+      "learning_rate": 1.0941733015960132e-07,
+      "loss": 0.3885,
+      "step": 18328
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 46.58331035017471,
+      "learning_rate": 1.0916868411343872e-07,
+      "loss": 0.4466,
+      "step": 18329
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 62.38346233566906,
+      "learning_rate": 1.089203193567312e-07,
+      "loss": 0.503,
+      "step": 18330
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 60.199823276602565,
+      "learning_rate": 1.0867223589654418e-07,
+      "loss": 0.4283,
+      "step": 18331
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 53.822551289238554,
+      "learning_rate": 1.0842443373993427e-07,
+      "loss": 0.3946,
+      "step": 18332
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 64.20237727413328,
+      "learning_rate": 1.0817691289394583e-07,
+      "loss": 0.4313,
+      "step": 18333
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 57.210856982378544,
+      "learning_rate": 1.0792967336561877e-07,
+      "loss": 0.4066,
+      "step": 18334
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 54.26410528153452,
+      "learning_rate": 1.0768271516198525e-07,
+      "loss": 0.3949,
+      "step": 18335
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 46.200388401946114,
+      "learning_rate": 1.0743603829006854e-07,
+      "loss": 0.3611,
+      "step": 18336
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 53.99360277995031,
+      "learning_rate": 1.0718964275688526e-07,
+      "loss": 0.484,
+      "step": 18337
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 47.412272936365014,
+      "learning_rate": 1.069435285694409e-07,
+      "loss": 0.4176,
+      "step": 18338
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 48.846985263784205,
+      "learning_rate": 1.0669769573473432e-07,
+      "loss": 0.4904,
+      "step": 18339
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 48.176092957360716,
+      "learning_rate": 1.064521442597577e-07,
+      "loss": 0.3472,
+      "step": 18340
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 51.548664331383705,
+      "learning_rate": 1.0620687415149655e-07,
+      "loss": 0.4138,
+      "step": 18341
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.921047038117095,
+      "learning_rate": 1.0596188541692309e-07,
+      "loss": 0.4065,
+      "step": 18342
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 35.69345820004915,
+      "learning_rate": 1.0571717806300508e-07,
+      "loss": 0.3084,
+      "step": 18343
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.677669483002546,
+      "learning_rate": 1.0547275209670249e-07,
+      "loss": 0.3763,
+      "step": 18344
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 43.86336629908398,
+      "learning_rate": 1.0522860752496755e-07,
+      "loss": 0.3641,
+      "step": 18345
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.47434749501344,
+      "learning_rate": 1.0498474435474249e-07,
+      "loss": 0.4182,
+      "step": 18346
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 43.97682516507852,
+      "learning_rate": 1.0474116259296285e-07,
+      "loss": 0.3615,
+      "step": 18347
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 44.16309281530846,
+      "learning_rate": 1.0449786224655644e-07,
+      "loss": 0.4061,
+      "step": 18348
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.83049599055581,
+      "learning_rate": 1.0425484332244107e-07,
+      "loss": 0.395,
+      "step": 18349
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 45.47439623511065,
+      "learning_rate": 1.0401210582753007e-07,
+      "loss": 0.4051,
+      "step": 18350
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 61.677369813576405,
+      "learning_rate": 1.0376964976872461e-07,
+      "loss": 0.3697,
+      "step": 18351
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 53.835284328636334,
+      "learning_rate": 1.035274751529225e-07,
+      "loss": 0.5079,
+      "step": 18352
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 51.62378177637134,
+      "learning_rate": 1.0328558198700933e-07,
+      "loss": 0.4068,
+      "step": 18353
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.68079058660944,
+      "learning_rate": 1.0304397027786516e-07,
+      "loss": 0.4816,
+      "step": 18354
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 53.86340220030619,
+      "learning_rate": 1.0280264003236006e-07,
+      "loss": 0.4243,
+      "step": 18355
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.78466562620938,
+      "learning_rate": 1.0256159125735742e-07,
+      "loss": 0.4127,
+      "step": 18356
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.6297101013156,
+      "learning_rate": 1.023208239597151e-07,
+      "loss": 0.457,
+      "step": 18357
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 59.947950229806615,
+      "learning_rate": 1.0208033814627872e-07,
+      "loss": 0.4088,
+      "step": 18358
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 50.380463315734616,
+      "learning_rate": 1.0184013382388613e-07,
+      "loss": 0.4457,
+      "step": 18359
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 43.99108239860368,
+      "learning_rate": 1.0160021099936967e-07,
+      "loss": 0.4161,
+      "step": 18360
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 56.55422581804222,
+      "learning_rate": 1.0136056967955388e-07,
+      "loss": 0.4147,
+      "step": 18361
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 49.73401552310695,
+      "learning_rate": 1.0112120987125329e-07,
+      "loss": 0.4414,
+      "step": 18362
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.00784289077099,
+      "learning_rate": 1.0088213158127358e-07,
+      "loss": 0.4186,
+      "step": 18363
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 50.123522420399965,
+      "learning_rate": 1.0064333481641597e-07,
+      "loss": 0.3669,
+      "step": 18364
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 38.95980505313899,
+      "learning_rate": 1.0040481958346948e-07,
+      "loss": 0.3769,
+      "step": 18365
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 40.5654811380808,
+      "learning_rate": 1.0016658588922091e-07,
+      "loss": 0.3826,
+      "step": 18366
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 48.579767606147186,
+      "learning_rate": 9.99286337404426e-08,
+      "loss": 0.436,
+      "step": 18367
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.53356300848152,
+      "learning_rate": 9.969096314390248e-08,
+      "loss": 0.4223,
+      "step": 18368
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.645948693731135,
+      "learning_rate": 9.945357410635959e-08,
+      "loss": 0.3915,
+      "step": 18369
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 49.38793676362315,
+      "learning_rate": 9.92164666345663e-08,
+      "loss": 0.3846,
+      "step": 18370
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 49.23256519493653,
+      "learning_rate": 9.8979640735265e-08,
+      "loss": 0.4159,
+      "step": 18371
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 37.107754070016405,
+      "learning_rate": 9.87430964151892e-08,
+      "loss": 0.3424,
+      "step": 18372
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.305110001508204,
+      "learning_rate": 9.850683368106794e-08,
+      "loss": 0.3749,
+      "step": 18373
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 39.85753064167142,
+      "learning_rate": 9.827085253962144e-08,
+      "loss": 0.3608,
+      "step": 18374
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 51.76034993872202,
+      "learning_rate": 9.803515299755872e-08,
+      "loss": 0.449,
+      "step": 18375
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.981292855030425,
+      "learning_rate": 9.779973506158446e-08,
+      "loss": 0.4718,
+      "step": 18376
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 40.28605232584893,
+      "learning_rate": 9.75645987383922e-08,
+      "loss": 0.3655,
+      "step": 18377
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 56.51592535730627,
+      "learning_rate": 9.732974403466988e-08,
+      "loss": 0.423,
+      "step": 18378
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 62.38614360058624,
+      "learning_rate": 9.709517095709775e-08,
+      "loss": 0.503,
+      "step": 18379
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 46.78372541566143,
+      "learning_rate": 9.68608795123449e-08,
+      "loss": 0.4599,
+      "step": 18380
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 46.772098443721035,
+      "learning_rate": 9.662686970707602e-08,
+      "loss": 0.4158,
+      "step": 18381
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 48.53707108034258,
+      "learning_rate": 9.639314154794576e-08,
+      "loss": 0.3979,
+      "step": 18382
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 45.809528389763756,
+      "learning_rate": 9.615969504160106e-08,
+      "loss": 0.4519,
+      "step": 18383
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 49.78315263789236,
+      "learning_rate": 9.5926530194681e-08,
+      "loss": 0.388,
+      "step": 18384
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 63.18320207477373,
+      "learning_rate": 9.569364701381589e-08,
+      "loss": 0.423,
+      "step": 18385
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 72.53377505225195,
+      "learning_rate": 9.546104550563152e-08,
+      "loss": 0.5783,
+      "step": 18386
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 51.541724788085105,
+      "learning_rate": 9.522872567673924e-08,
+      "loss": 0.4121,
+      "step": 18387
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 45.58143285675173,
+      "learning_rate": 9.499668753374936e-08,
+      "loss": 0.378,
+      "step": 18388
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 57.89713414878991,
+      "learning_rate": 9.476493108325768e-08,
+      "loss": 0.4015,
+      "step": 18389
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 43.34168828173886,
+      "learning_rate": 9.453345633185784e-08,
+      "loss": 0.4348,
+      "step": 18390
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.61300850133744,
+      "learning_rate": 9.430226328613234e-08,
+      "loss": 0.3754,
+      "step": 18391
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 42.69149653770498,
+      "learning_rate": 9.407135195265593e-08,
+      "loss": 0.3929,
+      "step": 18392
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 49.22652321120575,
+      "learning_rate": 9.384072233799447e-08,
+      "loss": 0.4082,
+      "step": 18393
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 56.35480995398664,
+      "learning_rate": 9.361037444870824e-08,
+      "loss": 0.4792,
+      "step": 18394
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 46.05982509491211,
+      "learning_rate": 9.33803082913487e-08,
+      "loss": 0.392,
+      "step": 18395
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 48.533567055180484,
+      "learning_rate": 9.315052387245726e-08,
+      "loss": 0.4858,
+      "step": 18396
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 42.46686203656902,
+      "learning_rate": 9.292102119856983e-08,
+      "loss": 0.3677,
+      "step": 18397
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 55.62931089519934,
+      "learning_rate": 9.269180027621228e-08,
+      "loss": 0.4657,
+      "step": 18398
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 58.11336993286152,
+      "learning_rate": 9.246286111190495e-08,
+      "loss": 0.4763,
+      "step": 18399
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 46.42849379700679,
+      "learning_rate": 9.223420371215818e-08,
+      "loss": 0.4018,
+      "step": 18400
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 47.43997736004374,
+      "learning_rate": 9.200582808347569e-08,
+      "loss": 0.4027,
+      "step": 18401
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 59.522498813712296,
+      "learning_rate": 9.177773423235004e-08,
+      "loss": 0.4978,
+      "step": 18402
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 74.22975036694233,
+      "learning_rate": 9.154992216526937e-08,
+      "loss": 0.4433,
+      "step": 18403
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 49.61629015688179,
+      "learning_rate": 9.132239188871405e-08,
+      "loss": 0.3991,
+      "step": 18404
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 54.35030485592352,
+      "learning_rate": 9.109514340915226e-08,
+      "loss": 0.4449,
+      "step": 18405
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 45.631259546303724,
+      "learning_rate": 9.086817673304882e-08,
+      "loss": 0.3755,
+      "step": 18406
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 39.55834552230845,
+      "learning_rate": 9.064149186685744e-08,
+      "loss": 0.3313,
+      "step": 18407
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.91706280696599,
+      "learning_rate": 9.041508881702521e-08,
+      "loss": 0.455,
+      "step": 18408
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 45.881371915712705,
+      "learning_rate": 9.018896758999029e-08,
+      "loss": 0.3649,
+      "step": 18409
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 50.4134112318539,
+      "learning_rate": 8.996312819218533e-08,
+      "loss": 0.4469,
+      "step": 18410
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 54.80489837758175,
+      "learning_rate": 8.973757063003075e-08,
+      "loss": 0.4261,
+      "step": 18411
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 44.29748812458037,
+      "learning_rate": 8.95122949099425e-08,
+      "loss": 0.4165,
+      "step": 18412
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 41.5034267435536,
+      "learning_rate": 8.928730103832773e-08,
+      "loss": 0.378,
+      "step": 18413
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 44.28117303296428,
+      "learning_rate": 8.90625890215846e-08,
+      "loss": 0.4324,
+      "step": 18414
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 58.519087799247856,
+      "learning_rate": 8.883815886610358e-08,
+      "loss": 0.4569,
+      "step": 18415
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 44.97946831356295,
+      "learning_rate": 8.861401057826624e-08,
+      "loss": 0.3637,
+      "step": 18416
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 43.879156844976244,
+      "learning_rate": 8.839014416444969e-08,
+      "loss": 0.4262,
+      "step": 18417
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 44.679509180723336,
+      "learning_rate": 8.816655963101883e-08,
+      "loss": 0.3625,
+      "step": 18418
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.27435164165862,
+      "learning_rate": 8.794325698433193e-08,
+      "loss": 0.4258,
+      "step": 18419
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.895463190671435,
+      "learning_rate": 8.772023623074166e-08,
+      "loss": 0.464,
+      "step": 18420
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 51.63561580676762,
+      "learning_rate": 8.749749737658964e-08,
+      "loss": 0.4414,
+      "step": 18421
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 51.70892223552541,
+      "learning_rate": 8.727504042820966e-08,
+      "loss": 0.4657,
+      "step": 18422
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 50.125167932015025,
+      "learning_rate": 8.705286539192781e-08,
+      "loss": 0.5326,
+      "step": 18423
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 46.72203728583425,
+      "learning_rate": 8.683097227406456e-08,
+      "loss": 0.4527,
+      "step": 18424
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 51.56483912855384,
+      "learning_rate": 8.660936108093043e-08,
+      "loss": 0.4684,
+      "step": 18425
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 46.438763007012035,
+      "learning_rate": 8.638803181882483e-08,
+      "loss": 0.4916,
+      "step": 18426
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 52.43905690988429,
+      "learning_rate": 8.616698449404492e-08,
+      "loss": 0.4757,
+      "step": 18427
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 44.47710290389875,
+      "learning_rate": 8.594621911287682e-08,
+      "loss": 0.3886,
+      "step": 18428
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 43.66366034756138,
+      "learning_rate": 8.572573568159881e-08,
+      "loss": 0.3649,
+      "step": 18429
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 43.20687593315426,
+      "learning_rate": 8.550553420648145e-08,
+      "loss": 0.3661,
+      "step": 18430
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 45.37638517635212,
+      "learning_rate": 8.528561469378637e-08,
+      "loss": 0.3873,
+      "step": 18431
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 49.77225699077867,
+      "learning_rate": 8.50659771497686e-08,
+      "loss": 0.4036,
+      "step": 18432
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 42.97820447077991,
+      "learning_rate": 8.484662158067425e-08,
+      "loss": 0.4357,
+      "step": 18433
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 56.39494913021282,
+      "learning_rate": 8.462754799274275e-08,
+      "loss": 0.4185,
+      "step": 18434
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 58.36346650801104,
+      "learning_rate": 8.440875639220247e-08,
+      "loss": 0.4844,
+      "step": 18435
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 48.038052893064226,
+      "learning_rate": 8.41902467852762e-08,
+      "loss": 0.5229,
+      "step": 18436
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 59.38967990298897,
+      "learning_rate": 8.397201917817899e-08,
+      "loss": 0.4345,
+      "step": 18437
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.53573417704102,
+      "learning_rate": 8.375407357711806e-08,
+      "loss": 0.3297,
+      "step": 18438
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 54.608897670684414,
+      "learning_rate": 8.35364099882896e-08,
+      "loss": 0.4628,
+      "step": 18439
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.54632558263881,
+      "learning_rate": 8.331902841788531e-08,
+      "loss": 0.3748,
+      "step": 18440
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 61.73891514733593,
+      "learning_rate": 8.310192887208579e-08,
+      "loss": 0.5071,
+      "step": 18441
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 48.87879827623583,
+      "learning_rate": 8.288511135706611e-08,
+      "loss": 0.4,
+      "step": 18442
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 55.36233562118647,
+      "learning_rate": 8.266857587899357e-08,
+      "loss": 0.4305,
+      "step": 18443
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 49.85028108959476,
+      "learning_rate": 8.245232244402435e-08,
+      "loss": 0.402,
+      "step": 18444
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 38.7559695844713,
+      "learning_rate": 8.223635105831018e-08,
+      "loss": 0.3572,
+      "step": 18445
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 53.693095882580025,
+      "learning_rate": 8.202066172799172e-08,
+      "loss": 0.4536,
+      "step": 18446
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.73051928042915,
+      "learning_rate": 8.180525445920407e-08,
+      "loss": 0.3891,
+      "step": 18447
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 44.07901089875707,
+      "learning_rate": 8.159012925807341e-08,
+      "loss": 0.3873,
+      "step": 18448
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 53.85554438094779,
+      "learning_rate": 8.1375286130716e-08,
+      "loss": 0.491,
+      "step": 18449
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 43.61366359681471,
+      "learning_rate": 8.116072508324579e-08,
+      "loss": 0.4068,
+      "step": 18450
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 61.772223925149014,
+      "learning_rate": 8.094644612176017e-08,
+      "loss": 0.4409,
+      "step": 18451
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 39.3299599038741,
+      "learning_rate": 8.073244925235535e-08,
+      "loss": 0.3253,
+      "step": 18452
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 52.20168147393391,
+      "learning_rate": 8.051873448111758e-08,
+      "loss": 0.4098,
+      "step": 18453
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 46.224298692193415,
+      "learning_rate": 8.030530181412422e-08,
+      "loss": 0.3639,
+      "step": 18454
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 62.658837738252686,
+      "learning_rate": 8.009215125744486e-08,
+      "loss": 0.4133,
+      "step": 18455
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 59.5696761220491,
+      "learning_rate": 7.987928281714241e-08,
+      "loss": 0.4166,
+      "step": 18456
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 57.584451878012395,
+      "learning_rate": 7.966669649926872e-08,
+      "loss": 0.4287,
+      "step": 18457
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 64.3475665960994,
+      "learning_rate": 7.945439230987228e-08,
+      "loss": 0.5152,
+      "step": 18458
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 48.87356393075077,
+      "learning_rate": 7.924237025498938e-08,
+      "loss": 0.4034,
+      "step": 18459
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 40.831702229301776,
+      "learning_rate": 7.903063034064962e-08,
+      "loss": 0.3407,
+      "step": 18460
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 53.76362315683018,
+      "learning_rate": 7.881917257287485e-08,
+      "loss": 0.4825,
+      "step": 18461
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 60.82704325080812,
+      "learning_rate": 7.860799695767918e-08,
+      "loss": 0.4204,
+      "step": 18462
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 65.6014848776564,
+      "learning_rate": 7.839710350106888e-08,
+      "loss": 0.4993,
+      "step": 18463
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 45.217912456609426,
+      "learning_rate": 7.81864922090414e-08,
+      "loss": 0.4303,
+      "step": 18464
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 54.47205579993692,
+      "learning_rate": 7.797616308758526e-08,
+      "loss": 0.3905,
+      "step": 18465
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 39.95041277902549,
+      "learning_rate": 7.776611614268236e-08,
+      "loss": 0.3737,
+      "step": 18466
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 47.90657952102632,
+      "learning_rate": 7.755635138030682e-08,
+      "loss": 0.4213,
+      "step": 18467
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 45.91553623079503,
+      "learning_rate": 7.734686880642495e-08,
+      "loss": 0.4014,
+      "step": 18468
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 46.051117327841666,
+      "learning_rate": 7.71376684269931e-08,
+      "loss": 0.3748,
+      "step": 18469
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 53.63400703316842,
+      "learning_rate": 7.692875024796099e-08,
+      "loss": 0.4728,
+      "step": 18470
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 43.07088045644949,
+      "learning_rate": 7.672011427527159e-08,
+      "loss": 0.4105,
+      "step": 18471
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 44.271891647554384,
+      "learning_rate": 7.651176051485576e-08,
+      "loss": 0.3774,
+      "step": 18472
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 44.77257873425139,
+      "learning_rate": 7.630368897264096e-08,
+      "loss": 0.3731,
+      "step": 18473
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 50.76846410268617,
+      "learning_rate": 7.609589965454356e-08,
+      "loss": 0.4766,
+      "step": 18474
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 43.27708688813473,
+      "learning_rate": 7.58883925664744e-08,
+      "loss": 0.3709,
+      "step": 18475
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 51.23588232654178,
+      "learning_rate": 7.56811677143332e-08,
+      "loss": 0.4543,
+      "step": 18476
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 63.496710068789675,
+      "learning_rate": 7.547422510401414e-08,
+      "loss": 0.4272,
+      "step": 18477
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 45.81825129693269,
+      "learning_rate": 7.526756474140252e-08,
+      "loss": 0.4225,
+      "step": 18478
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 55.27015414241847,
+      "learning_rate": 7.506118663237583e-08,
+      "loss": 0.4027,
+      "step": 18479
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 54.56339288966316,
+      "learning_rate": 7.485509078280384e-08,
+      "loss": 0.4505,
+      "step": 18480
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 55.29624824351974,
+      "learning_rate": 7.46492771985452e-08,
+      "loss": 0.3747,
+      "step": 18481
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 49.43716502407706,
+      "learning_rate": 7.444374588545522e-08,
+      "loss": 0.3734,
+      "step": 18482
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 53.27349633323955,
+      "learning_rate": 7.423849684938033e-08,
+      "loss": 0.4546,
+      "step": 18483
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 48.859885076160474,
+      "learning_rate": 7.403353009615588e-08,
+      "loss": 0.398,
+      "step": 18484
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 54.50644205674334,
+      "learning_rate": 7.382884563161052e-08,
+      "loss": 0.443,
+      "step": 18485
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 54.765480369997256,
+      "learning_rate": 7.362444346156517e-08,
+      "loss": 0.4407,
+      "step": 18486
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 55.44365897903823,
+      "learning_rate": 7.342032359183515e-08,
+      "loss": 0.446,
+      "step": 18487
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 52.43467288756513,
+      "learning_rate": 7.321648602822473e-08,
+      "loss": 0.4671,
+      "step": 18488
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 46.5672388519945,
+      "learning_rate": 7.301293077653038e-08,
+      "loss": 0.4092,
+      "step": 18489
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 41.47891276978148,
+      "learning_rate": 7.280965784253968e-08,
+      "loss": 0.4232,
+      "step": 18490
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 46.99086719411711,
+      "learning_rate": 7.26066672320358e-08,
+      "loss": 0.3848,
+      "step": 18491
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 47.606064984669594,
+      "learning_rate": 7.240395895079189e-08,
+      "loss": 0.4284,
+      "step": 18492
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 46.52813080528787,
+      "learning_rate": 7.22015330045711e-08,
+      "loss": 0.471,
+      "step": 18493
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 54.137706903640506,
+      "learning_rate": 7.199938939913109e-08,
+      "loss": 0.412,
+      "step": 18494
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 68.7128639909674,
+      "learning_rate": 7.179752814022056e-08,
+      "loss": 0.5156,
+      "step": 18495
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 43.51198374445535,
+      "learning_rate": 7.159594923358159e-08,
+      "loss": 0.3822,
+      "step": 18496
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 52.16244990650133,
+      "learning_rate": 7.139465268494405e-08,
+      "loss": 0.4431,
+      "step": 18497
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 48.1863978515928,
+      "learning_rate": 7.119363850003558e-08,
+      "loss": 0.3631,
+      "step": 18498
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 45.19070233279787,
+      "learning_rate": 7.09929066845727e-08,
+      "loss": 0.4082,
+      "step": 18499
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 47.59585622734439,
+      "learning_rate": 7.079245724426198e-08,
+      "loss": 0.4311,
+      "step": 18500
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 48.741287029916464,
+      "learning_rate": 7.059229018480551e-08,
+      "loss": 0.4207,
+      "step": 18501
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 62.26500848826295,
+      "learning_rate": 7.039240551189653e-08,
+      "loss": 0.3797,
+      "step": 18502
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 51.180197300800835,
+      "learning_rate": 7.019280323121714e-08,
+      "loss": 0.3854,
+      "step": 18503
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 49.69247306072612,
+      "learning_rate": 6.999348334844724e-08,
+      "loss": 0.3934,
+      "step": 18504
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 47.221781746299726,
+      "learning_rate": 6.979444586925455e-08,
+      "loss": 0.4005,
+      "step": 18505
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 52.84430619843248,
+      "learning_rate": 6.959569079929674e-08,
+      "loss": 0.5047,
+      "step": 18506
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 51.84497375147682,
+      "learning_rate": 6.939721814422928e-08,
+      "loss": 0.3873,
+      "step": 18507
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 82.198755527738,
+      "learning_rate": 6.919902790969657e-08,
+      "loss": 0.4894,
+      "step": 18508
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 77.07153692418424,
+      "learning_rate": 6.900112010133297e-08,
+      "loss": 0.4026,
+      "step": 18509
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 44.163588748903216,
+      "learning_rate": 6.880349472476844e-08,
+      "loss": 0.3749,
+      "step": 18510
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 45.53734548868694,
+      "learning_rate": 6.86061517856229e-08,
+      "loss": 0.3652,
+      "step": 18511
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 50.97536667864498,
+      "learning_rate": 6.840909128950857e-08,
+      "loss": 0.4143,
+      "step": 18512
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 47.07831714235149,
+      "learning_rate": 6.821231324203093e-08,
+      "loss": 0.4108,
+      "step": 18513
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 46.01047376934732,
+      "learning_rate": 6.801581764878329e-08,
+      "loss": 0.3923,
+      "step": 18514
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 45.42342904460089,
+      "learning_rate": 6.781960451535674e-08,
+      "loss": 0.3626,
+      "step": 18515
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 54.80167798828541,
+      "learning_rate": 6.762367384733015e-08,
+      "loss": 0.4056,
+      "step": 18516
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 66.48429623160979,
+      "learning_rate": 6.742802565027684e-08,
+      "loss": 0.4721,
+      "step": 18517
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 54.5864286355592,
+      "learning_rate": 6.723265992975903e-08,
+      "loss": 0.4332,
+      "step": 18518
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 50.697272730914754,
+      "learning_rate": 6.703757669133448e-08,
+      "loss": 0.4286,
+      "step": 18519
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 46.112877454528714,
+      "learning_rate": 6.6842775940551e-08,
+      "loss": 0.4388,
+      "step": 18520
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 37.89523673081636,
+      "learning_rate": 6.664825768294747e-08,
+      "loss": 0.3355,
+      "step": 18521
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 42.73743785084894,
+      "learning_rate": 6.645402192405726e-08,
+      "loss": 0.4098,
+      "step": 18522
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 52.73369675037644,
+      "learning_rate": 6.626006866940371e-08,
+      "loss": 0.4336,
+      "step": 18523
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 64.02477042839602,
+      "learning_rate": 6.606639792450131e-08,
+      "loss": 0.4629,
+      "step": 18524
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 53.48735946544096,
+      "learning_rate": 6.587300969486122e-08,
+      "loss": 0.4121,
+      "step": 18525
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 49.459685169225246,
+      "learning_rate": 6.567990398598012e-08,
+      "loss": 0.4214,
+      "step": 18526
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 48.46584755565222,
+      "learning_rate": 6.548708080335143e-08,
+      "loss": 0.4005,
+      "step": 18527
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 50.81816934205282,
+      "learning_rate": 6.529454015245962e-08,
+      "loss": 0.4511,
+      "step": 18528
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 47.27688496289912,
+      "learning_rate": 6.510228203877811e-08,
+      "loss": 0.4174,
+      "step": 18529
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 55.19431777349326,
+      "learning_rate": 6.491030646777696e-08,
+      "loss": 0.4777,
+      "step": 18530
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 44.60980147176271,
+      "learning_rate": 6.471861344491514e-08,
+      "loss": 0.3752,
+      "step": 18531
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 62.284918264112015,
+      "learning_rate": 6.452720297564275e-08,
+      "loss": 0.476,
+      "step": 18532
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 44.956312361586065,
+      "learning_rate": 6.43360750654054e-08,
+      "loss": 0.3897,
+      "step": 18533
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 53.316248056513984,
+      "learning_rate": 6.414522971963766e-08,
+      "loss": 0.4528,
+      "step": 18534
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 52.919099052974744,
+      "learning_rate": 6.395466694376739e-08,
+      "loss": 0.4923,
+      "step": 18535
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 44.31397972272713,
+      "learning_rate": 6.376438674321361e-08,
+      "loss": 0.3624,
+      "step": 18536
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 62.31173586974716,
+      "learning_rate": 6.357438912338864e-08,
+      "loss": 0.4578,
+      "step": 18537
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 44.11795616122026,
+      "learning_rate": 6.338467408969484e-08,
+      "loss": 0.3876,
+      "step": 18538
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 50.18479309763021,
+      "learning_rate": 6.319524164752899e-08,
+      "loss": 0.4378,
+      "step": 18539
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 50.177531894956694,
+      "learning_rate": 6.300609180227679e-08,
+      "loss": 0.4455,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 51.874820148202204,
+      "learning_rate": 6.281722455931839e-08,
+      "loss": 0.4567,
+      "step": 18541
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 58.91917293870637,
+      "learning_rate": 6.262863992402612e-08,
+      "loss": 0.4199,
+      "step": 18542
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 46.54817317242103,
+      "learning_rate": 6.24403379017613e-08,
+      "loss": 0.4283,
+      "step": 18543
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 46.94347072371468,
+      "learning_rate": 6.225231849787961e-08,
+      "loss": 0.4671,
+      "step": 18544
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 47.232148678599216,
+      "learning_rate": 6.2064581717729e-08,
+      "loss": 0.4299,
+      "step": 18545
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 48.81976761973989,
+      "learning_rate": 6.187712756664854e-08,
+      "loss": 0.467,
+      "step": 18546
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 48.97114550312691,
+      "learning_rate": 6.168995604996841e-08,
+      "loss": 0.3928,
+      "step": 18547
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 52.1451700782837,
+      "learning_rate": 6.150306717301213e-08,
+      "loss": 0.4098,
+      "step": 18548
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 43.891194156203476,
+      "learning_rate": 6.131646094109545e-08,
+      "loss": 0.3943,
+      "step": 18549
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 51.106384115193585,
+      "learning_rate": 6.1130137359523e-08,
+      "loss": 0.4023,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 53.03818224329443,
+      "learning_rate": 6.094409643359722e-08,
+      "loss": 0.4438,
+      "step": 18551
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 51.2960331708589,
+      "learning_rate": 6.07583381686061e-08,
+      "loss": 0.4409,
+      "step": 18552
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 51.87314315138326,
+      "learning_rate": 6.05728625698343e-08,
+      "loss": 0.432,
+      "step": 18553
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 47.085358381468076,
+      "learning_rate": 6.038766964255538e-08,
+      "loss": 0.4035,
+      "step": 18554
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 50.496545125791656,
+      "learning_rate": 6.020275939203734e-08,
+      "loss": 0.3972,
+      "step": 18555
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 47.748261494968915,
+      "learning_rate": 6.00181318235371e-08,
+      "loss": 0.4234,
+      "step": 18556
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 39.79493478000174,
+      "learning_rate": 5.983378694230713e-08,
+      "loss": 0.4134,
+      "step": 18557
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 44.13310636779436,
+      "learning_rate": 5.964972475358876e-08,
+      "loss": 0.355,
+      "step": 18558
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 43.091976469528696,
+      "learning_rate": 5.946594526261895e-08,
+      "loss": 0.3911,
+      "step": 18559
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 44.924014374583344,
+      "learning_rate": 5.928244847462128e-08,
+      "loss": 0.3818,
+      "step": 18560
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 52.48410443857723,
+      "learning_rate": 5.909923439481491e-08,
+      "loss": 0.3921,
+      "step": 18561
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 47.40717971584132,
+      "learning_rate": 5.8916303028412334e-08,
+      "loss": 0.3477,
+      "step": 18562
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 58.862778409541285,
+      "learning_rate": 5.873365438061385e-08,
+      "loss": 0.3686,
+      "step": 18563
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 51.94997005663716,
+      "learning_rate": 5.8551288456616395e-08,
+      "loss": 0.3798,
+      "step": 18564
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 38.62806578694047,
+      "learning_rate": 5.83692052616025e-08,
+      "loss": 0.4048,
+      "step": 18565
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 65.1128384113214,
+      "learning_rate": 5.8187404800753575e-08,
+      "loss": 0.4326,
+      "step": 18566
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 48.2364234007428,
+      "learning_rate": 5.800588707923993e-08,
+      "loss": 0.4309,
+      "step": 18567
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 54.8835531446772,
+      "learning_rate": 5.782465210222299e-08,
+      "loss": 0.4724,
+      "step": 18568
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 63.43882051042489,
+      "learning_rate": 5.7643699874855296e-08,
+      "loss": 0.4203,
+      "step": 18569
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 69.58388596351448,
+      "learning_rate": 5.746303040228607e-08,
+      "loss": 0.4343,
+      "step": 18570
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 46.640892332195875,
+      "learning_rate": 5.7282643689652306e-08,
+      "loss": 0.4074,
+      "step": 18571
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 55.70015284245065,
+      "learning_rate": 5.710253974208324e-08,
+      "loss": 0.4204,
+      "step": 18572
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 66.63106965120299,
+      "learning_rate": 5.692271856470144e-08,
+      "loss": 0.427,
+      "step": 18573
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 46.15753971640504,
+      "learning_rate": 5.674318016262059e-08,
+      "loss": 0.419,
+      "step": 18574
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 58.024358674022025,
+      "learning_rate": 5.656392454094661e-08,
+      "loss": 0.388,
+      "step": 18575
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 53.93695468118592,
+      "learning_rate": 5.638495170477876e-08,
+      "loss": 0.4714,
+      "step": 18576
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 52.401996646944326,
+      "learning_rate": 5.6206261659206284e-08,
+      "loss": 0.4162,
+      "step": 18577
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 58.596230294183826,
+      "learning_rate": 5.602785440931069e-08,
+      "loss": 0.4167,
+      "step": 18578
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 52.42304900399687,
+      "learning_rate": 5.584972996016569e-08,
+      "loss": 0.4045,
+      "step": 18579
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 45.55984147613618,
+      "learning_rate": 5.567188831683723e-08,
+      "loss": 0.418,
+      "step": 18580
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 58.02254834565389,
+      "learning_rate": 5.549432948438238e-08,
+      "loss": 0.3244,
+      "step": 18581
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 58.501792652018416,
+      "learning_rate": 5.531705346785265e-08,
+      "loss": 0.468,
+      "step": 18582
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 51.35102824228169,
+      "learning_rate": 5.514006027228735e-08,
+      "loss": 0.3831,
+      "step": 18583
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 48.30792792080329,
+      "learning_rate": 5.4963349902722454e-08,
+      "loss": 0.4271,
+      "step": 18584
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 43.45888701379095,
+      "learning_rate": 5.478692236418059e-08,
+      "loss": 0.4246,
+      "step": 18585
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 44.95822671495204,
+      "learning_rate": 5.4610777661681105e-08,
+      "loss": 0.3866,
+      "step": 18586
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 44.61139970717477,
+      "learning_rate": 5.443491580023441e-08,
+      "loss": 0.4325,
+      "step": 18587
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 43.247231852417194,
+      "learning_rate": 5.4259336784839855e-08,
+      "loss": 0.4193,
+      "step": 18588
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 47.085526581827736,
+      "learning_rate": 5.408404062049233e-08,
+      "loss": 0.4084,
+      "step": 18589
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 42.6046679659251,
+      "learning_rate": 5.390902731217562e-08,
+      "loss": 0.3732,
+      "step": 18590
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 45.53491855155804,
+      "learning_rate": 5.373429686486797e-08,
+      "loss": 0.4051,
+      "step": 18591
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 50.456367249094484,
+      "learning_rate": 5.355984928353986e-08,
+      "loss": 0.4216,
+      "step": 18592
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 41.58521755296905,
+      "learning_rate": 5.338568457314952e-08,
+      "loss": 0.4247,
+      "step": 18593
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 47.891843208388956,
+      "learning_rate": 5.321180273865301e-08,
+      "loss": 0.3928,
+      "step": 18594
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 36.047223642132394,
+      "learning_rate": 5.303820378499303e-08,
+      "loss": 0.3372,
+      "step": 18595
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 50.64943827179209,
+      "learning_rate": 5.286488771710785e-08,
+      "loss": 0.4389,
+      "step": 18596
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 50.15029768755597,
+      "learning_rate": 5.2691854539926865e-08,
+      "loss": 0.389,
+      "step": 18597
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 62.29010247740658,
+      "learning_rate": 5.251910425837059e-08,
+      "loss": 0.4341,
+      "step": 18598
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 57.10677447575533,
+      "learning_rate": 5.234663687735064e-08,
+      "loss": 0.5385,
+      "step": 18599
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 52.04018131159278,
+      "learning_rate": 5.217445240177421e-08,
+      "loss": 0.4069,
+      "step": 18600
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 56.305048188078054,
+      "learning_rate": 5.200255083653738e-08,
+      "loss": 0.4273,
+      "step": 18601
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 42.55487552417097,
+      "learning_rate": 5.183093218652846e-08,
+      "loss": 0.3585,
+      "step": 18602
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 49.97557777482987,
+      "learning_rate": 5.1659596456627995e-08,
+      "loss": 0.378,
+      "step": 18603
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 44.41280766495167,
+      "learning_rate": 5.1488543651708746e-08,
+      "loss": 0.4062,
+      "step": 18604
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 47.327445497921225,
+      "learning_rate": 5.1317773776635715e-08,
+      "loss": 0.3617,
+      "step": 18605
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 49.39605173325199,
+      "learning_rate": 5.114728683626613e-08,
+      "loss": 0.4263,
+      "step": 18606
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 54.17883440046023,
+      "learning_rate": 5.0977082835447224e-08,
+      "loss": 0.4694,
+      "step": 18607
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 67.88861529116,
+      "learning_rate": 5.0807161779019565e-08,
+      "loss": 0.4828,
+      "step": 18608
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 45.38999431913886,
+      "learning_rate": 5.0637523671817065e-08,
+      "loss": 0.3964,
+      "step": 18609
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 46.66485135203431,
+      "learning_rate": 5.046816851866254e-08,
+      "loss": 0.433,
+      "step": 18610
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 41.61009313097605,
+      "learning_rate": 5.029909632437324e-08,
+      "loss": 0.3594,
+      "step": 18611
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 49.523331817307174,
+      "learning_rate": 5.013030709375644e-08,
+      "loss": 0.4335,
+      "step": 18612
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 53.02210294731057,
+      "learning_rate": 4.996180083161384e-08,
+      "loss": 0.4757,
+      "step": 18613
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 57.01587752006396,
+      "learning_rate": 4.979357754273717e-08,
+      "loss": 0.4255,
+      "step": 18614
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 60.180796177593315,
+      "learning_rate": 4.9625637231910385e-08,
+      "loss": 0.4913,
+      "step": 18615
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 52.057864638803174,
+      "learning_rate": 4.9457979903909656e-08,
+      "loss": 0.4682,
+      "step": 18616
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 61.05454683781856,
+      "learning_rate": 4.92906055635034e-08,
+      "loss": 0.4848,
+      "step": 18617
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 41.682637052556075,
+      "learning_rate": 4.912351421545114e-08,
+      "loss": 0.3546,
+      "step": 18618
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 43.956540308297,
+      "learning_rate": 4.8956705864504625e-08,
+      "loss": 0.3715,
+      "step": 18619
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 72.41374231659309,
+      "learning_rate": 4.879018051540785e-08,
+      "loss": 0.4456,
+      "step": 18620
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 46.66350353908246,
+      "learning_rate": 4.8623938172898125e-08,
+      "loss": 0.3706,
+      "step": 18621
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 49.349257323098286,
+      "learning_rate": 4.8457978841702777e-08,
+      "loss": 0.4114,
+      "step": 18622
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 67.7899460162893,
+      "learning_rate": 4.8292302526539156e-08,
+      "loss": 0.4656,
+      "step": 18623
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 42.78205419415689,
+      "learning_rate": 4.8126909232121265e-08,
+      "loss": 0.3598,
+      "step": 18624
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 51.9695061303492,
+      "learning_rate": 4.7961798963153116e-08,
+      "loss": 0.4043,
+      "step": 18625
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 57.61269368896238,
+      "learning_rate": 4.7796971724329844e-08,
+      "loss": 0.5224,
+      "step": 18626
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 42.01546604499526,
+      "learning_rate": 4.763242752033881e-08,
+      "loss": 0.409,
+      "step": 18627
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 66.03527108481171,
+      "learning_rate": 4.746816635585849e-08,
+      "loss": 0.4198,
+      "step": 18628
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 47.26385418561843,
+      "learning_rate": 4.730418823556182e-08,
+      "loss": 0.4295,
+      "step": 18629
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 45.502995575817124,
+      "learning_rate": 4.7140493164112844e-08,
+      "loss": 0.3796,
+      "step": 18630
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 38.285579115866916,
+      "learning_rate": 4.6977081146165614e-08,
+      "loss": 0.3551,
+      "step": 18631
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 56.60176764346397,
+      "learning_rate": 4.6813952186366415e-08,
+      "loss": 0.4388,
+      "step": 18632
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 38.76095832187212,
+      "learning_rate": 4.6651106289357095e-08,
+      "loss": 0.4215,
+      "step": 18633
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 48.968088555752786,
+      "learning_rate": 4.648854345976839e-08,
+      "loss": 0.4681,
+      "step": 18634
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 59.012811441696584,
+      "learning_rate": 4.632626370222215e-08,
+      "loss": 0.4826,
+      "step": 18635
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 48.355440867787536,
+      "learning_rate": 4.6164267021334693e-08,
+      "loss": 0.4081,
+      "step": 18636
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 56.31248592668668,
+      "learning_rate": 4.6002553421711226e-08,
+      "loss": 0.4145,
+      "step": 18637
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 53.40419832756321,
+      "learning_rate": 4.5841122907953614e-08,
+      "loss": 0.4396,
+      "step": 18638
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 53.94579043718745,
+      "learning_rate": 4.567997548465153e-08,
+      "loss": 0.4327,
+      "step": 18639
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 43.66355441677115,
+      "learning_rate": 4.551911115638685e-08,
+      "loss": 0.3996,
+      "step": 18640
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 44.003459176722885,
+      "learning_rate": 4.5358529927735925e-08,
+      "loss": 0.3604,
+      "step": 18641
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 44.489651762632015,
+      "learning_rate": 4.5198231803265103e-08,
+      "loss": 0.3708,
+      "step": 18642
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 55.507635897221924,
+      "learning_rate": 4.503821678753406e-08,
+      "loss": 0.4586,
+      "step": 18643
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 47.46366625064521,
+      "learning_rate": 4.487848488509139e-08,
+      "loss": 0.3739,
+      "step": 18644
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 47.098442334944714,
+      "learning_rate": 4.4719036100481225e-08,
+      "loss": 0.43,
+      "step": 18645
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 39.4038928943143,
+      "learning_rate": 4.455987043823884e-08,
+      "loss": 0.3603,
+      "step": 18646
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 58.40117624042175,
+      "learning_rate": 4.440098790288949e-08,
+      "loss": 0.4655,
+      "step": 18647
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 46.01069155089806,
+      "learning_rate": 4.4242388498951797e-08,
+      "loss": 0.3365,
+      "step": 18648
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 50.65517583481221,
+      "learning_rate": 4.408407223093658e-08,
+      "loss": 0.4312,
+      "step": 18649
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 51.50395394859376,
+      "learning_rate": 4.3926039103346915e-08,
+      "loss": 0.4151,
+      "step": 18650
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 43.816537330923126,
+      "learning_rate": 4.376828912067699e-08,
+      "loss": 0.4398,
+      "step": 18651
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 47.348234793882426,
+      "learning_rate": 4.361082228741209e-08,
+      "loss": 0.3381,
+      "step": 18652
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 46.04602453859308,
+      "learning_rate": 4.3453638608030865e-08,
+      "loss": 0.3834,
+      "step": 18653
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 51.70918649722212,
+      "learning_rate": 4.329673808700308e-08,
+      "loss": 0.4002,
+      "step": 18654
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 44.073720555768034,
+      "learning_rate": 4.3140120728792925e-08,
+      "loss": 0.3559,
+      "step": 18655
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 42.90876156429093,
+      "learning_rate": 4.29837865378524e-08,
+      "loss": 0.3864,
+      "step": 18656
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 52.35604762675407,
+      "learning_rate": 4.2827735518629065e-08,
+      "loss": 0.4122,
+      "step": 18657
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 67.66452014584665,
+      "learning_rate": 4.2671967675559365e-08,
+      "loss": 0.475,
+      "step": 18658
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 47.41300037682189,
+      "learning_rate": 4.2516483013074205e-08,
+      "loss": 0.4059,
+      "step": 18659
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 50.549790967175156,
+      "learning_rate": 4.23612815355956e-08,
+      "loss": 0.4355,
+      "step": 18660
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 46.32775048141528,
+      "learning_rate": 4.2206363247536684e-08,
+      "loss": 0.3792,
+      "step": 18661
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.861174127912676,
+      "learning_rate": 4.205172815330394e-08,
+      "loss": 0.4026,
+      "step": 18662
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 44.091621996426106,
+      "learning_rate": 4.189737625729384e-08,
+      "loss": 0.3819,
+      "step": 18663
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 50.36336468626292,
+      "learning_rate": 4.174330756389844e-08,
+      "loss": 0.3852,
+      "step": 18664
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.94175764054081,
+      "learning_rate": 4.1589522077497556e-08,
+      "loss": 0.4151,
+      "step": 18665
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.65187610231484,
+      "learning_rate": 4.143601980246437e-08,
+      "loss": 0.3832,
+      "step": 18666
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 56.488323097960034,
+      "learning_rate": 4.128280074316649e-08,
+      "loss": 0.4203,
+      "step": 18667
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 58.560694114012165,
+      "learning_rate": 4.112986490395931e-08,
+      "loss": 0.3782,
+      "step": 18668
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 55.453826598896214,
+      "learning_rate": 4.0977212289192717e-08,
+      "loss": 0.4754,
+      "step": 18669
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 52.89474859799406,
+      "learning_rate": 4.082484290320876e-08,
+      "loss": 0.4242,
+      "step": 18670
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.10777311764848,
+      "learning_rate": 4.067275675034066e-08,
+      "loss": 0.4281,
+      "step": 18671
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 63.22387178258337,
+      "learning_rate": 4.0520953834912724e-08,
+      "loss": 0.5094,
+      "step": 18672
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 40.443660596979505,
+      "learning_rate": 4.0369434161242616e-08,
+      "loss": 0.3879,
+      "step": 18673
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 49.97982360017993,
+      "learning_rate": 4.021819773364022e-08,
+      "loss": 0.4077,
+      "step": 18674
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 45.98512960651496,
+      "learning_rate": 4.0067244556405424e-08,
+      "loss": 0.3989,
+      "step": 18675
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 41.74421285734485,
+      "learning_rate": 3.9916574633832586e-08,
+      "loss": 0.3965,
+      "step": 18676
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 36.227893942263364,
+      "learning_rate": 3.976618797020493e-08,
+      "loss": 0.3837,
+      "step": 18677
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 42.168250127068255,
+      "learning_rate": 3.961608456980126e-08,
+      "loss": 0.3585,
+      "step": 18678
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 43.58481157882449,
+      "learning_rate": 3.946626443688817e-08,
+      "loss": 0.4073,
+      "step": 18679
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 44.437438084898574,
+      "learning_rate": 3.9316727575728906e-08,
+      "loss": 0.4183,
+      "step": 18680
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 55.762631187122174,
+      "learning_rate": 3.9167473990575635e-08,
+      "loss": 0.4731,
+      "step": 18681
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 44.403524074872614,
+      "learning_rate": 3.901850368567161e-08,
+      "loss": 0.4483,
+      "step": 18682
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.5136166953934,
+      "learning_rate": 3.886981666525347e-08,
+      "loss": 0.4217,
+      "step": 18683
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 52.01638654046542,
+      "learning_rate": 3.8721412933552246e-08,
+      "loss": 0.3843,
+      "step": 18684
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 45.09494464267303,
+      "learning_rate": 3.85732924947857e-08,
+      "loss": 0.4082,
+      "step": 18685
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 44.61277495085988,
+      "learning_rate": 3.8425455353168215e-08,
+      "loss": 0.3845,
+      "step": 18686
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 44.028262879713985,
+      "learning_rate": 3.827790151290312e-08,
+      "loss": 0.3869,
+      "step": 18687
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 61.35627936550688,
+      "learning_rate": 3.813063097818703e-08,
+      "loss": 0.4898,
+      "step": 18688
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 45.7873519840915,
+      "learning_rate": 3.798364375320773e-08,
+      "loss": 0.3946,
+      "step": 18689
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 54.36359398136476,
+      "learning_rate": 3.783693984214743e-08,
+      "loss": 0.4185,
+      "step": 18690
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 49.96834066866274,
+      "learning_rate": 3.7690519249174996e-08,
+      "loss": 0.3988,
+      "step": 18691
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 64.54813571973982,
+      "learning_rate": 3.754438197845822e-08,
+      "loss": 0.3955,
+      "step": 18692
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 41.18151241189358,
+      "learning_rate": 3.739852803415045e-08,
+      "loss": 0.3808,
+      "step": 18693
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 53.65080569978586,
+      "learning_rate": 3.725295742040058e-08,
+      "loss": 0.4058,
+      "step": 18694
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 66.41266313888585,
+      "learning_rate": 3.710767014134864e-08,
+      "loss": 0.4551,
+      "step": 18695
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 39.865505188243624,
+      "learning_rate": 3.696266620112576e-08,
+      "loss": 0.379,
+      "step": 18696
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 42.6682082800466,
+      "learning_rate": 3.681794560385754e-08,
+      "loss": 0.3739,
+      "step": 18697
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 55.29729033586614,
+      "learning_rate": 3.667350835365846e-08,
+      "loss": 0.4562,
+      "step": 18698
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 56.347182725209215,
+      "learning_rate": 3.6529354454635236e-08,
+      "loss": 0.364,
+      "step": 18699
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 40.183067249354316,
+      "learning_rate": 3.638548391089014e-08,
+      "loss": 0.402,
+      "step": 18700
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.79206317748098,
+      "learning_rate": 3.6241896726513236e-08,
+      "loss": 0.4292,
+      "step": 18701
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.728484369823185,
+      "learning_rate": 3.6098592905587925e-08,
+      "loss": 0.3765,
+      "step": 18702
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 52.56781078819854,
+      "learning_rate": 3.595557245218983e-08,
+      "loss": 0.4695,
+      "step": 18703
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 50.66459030311736,
+      "learning_rate": 3.581283537038571e-08,
+      "loss": 0.369,
+      "step": 18704
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 47.256260648861684,
+      "learning_rate": 3.567038166423675e-08,
+      "loss": 0.4816,
+      "step": 18705
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 48.535479005780466,
+      "learning_rate": 3.552821133779305e-08,
+      "loss": 0.444,
+      "step": 18706
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 58.5154264520554,
+      "learning_rate": 3.5386324395096924e-08,
+      "loss": 0.5657,
+      "step": 18707
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 49.74573713278879,
+      "learning_rate": 3.5244720840186266e-08,
+      "loss": 0.4097,
+      "step": 18708
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 41.85839028106184,
+      "learning_rate": 3.510340067708562e-08,
+      "loss": 0.4313,
+      "step": 18709
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 53.55385697915179,
+      "learning_rate": 3.4962363909815114e-08,
+      "loss": 0.4719,
+      "step": 18710
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 53.00366496963793,
+      "learning_rate": 3.482161054238486e-08,
+      "loss": 0.4734,
+      "step": 18711
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 43.55528139265431,
+      "learning_rate": 3.4681140578799453e-08,
+      "loss": 0.4325,
+      "step": 18712
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 50.11876179697013,
+      "learning_rate": 3.4540954023052356e-08,
+      "loss": 0.4416,
+      "step": 18713
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 51.67155692686378,
+      "learning_rate": 3.440105087913148e-08,
+      "loss": 0.3987,
+      "step": 18714
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 53.638226134044125,
+      "learning_rate": 3.426143115101477e-08,
+      "loss": 0.4527,
+      "step": 18715
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 52.30942758358929,
+      "learning_rate": 3.412209484267237e-08,
+      "loss": 0.4447,
+      "step": 18716
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 64.19790325900195,
+      "learning_rate": 3.39830419580689e-08,
+      "loss": 0.4618,
+      "step": 18717
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 49.73788608018657,
+      "learning_rate": 3.384427250115674e-08,
+      "loss": 0.3756,
+      "step": 18718
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 46.13276786317531,
+      "learning_rate": 3.370578647588496e-08,
+      "loss": 0.4191,
+      "step": 18719
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 55.33406114769843,
+      "learning_rate": 3.3567583886189304e-08,
+      "loss": 0.4468,
+      "step": 18720
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 45.89216582608277,
+      "learning_rate": 3.3429664736001064e-08,
+      "loss": 0.4027,
+      "step": 18721
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 53.92655617206534,
+      "learning_rate": 3.3292029029243777e-08,
+      "loss": 0.4496,
+      "step": 18722
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 55.995425940085035,
+      "learning_rate": 3.315467676982986e-08,
+      "loss": 0.5343,
+      "step": 18723
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 46.77033682760007,
+      "learning_rate": 3.301760796166731e-08,
+      "loss": 0.4463,
+      "step": 18724
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 47.52308817453975,
+      "learning_rate": 3.2880822608653e-08,
+      "loss": 0.407,
+      "step": 18725
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 61.49200741074268,
+      "learning_rate": 3.274432071467826e-08,
+      "loss": 0.4668,
+      "step": 18726
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 52.15812934794562,
+      "learning_rate": 3.260810228362332e-08,
+      "loss": 0.4935,
+      "step": 18727
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 35.99714681789493,
+      "learning_rate": 3.247216731936398e-08,
+      "loss": 0.3388,
+      "step": 18728
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 54.184166001620234,
+      "learning_rate": 3.233651582576491e-08,
+      "loss": 0.4372,
+      "step": 18729
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 50.06831960391926,
+      "learning_rate": 3.220114780668415e-08,
+      "loss": 0.4807,
+      "step": 18730
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 46.08747465703571,
+      "learning_rate": 3.206606326597306e-08,
+      "loss": 0.4354,
+      "step": 18731
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 47.68748673105118,
+      "learning_rate": 3.193126220747078e-08,
+      "loss": 0.4301,
+      "step": 18732
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 48.65738830998893,
+      "learning_rate": 3.1796744635013144e-08,
+      "loss": 0.4107,
+      "step": 18733
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 40.74056553012247,
+      "learning_rate": 3.166251055242375e-08,
+      "loss": 0.3876,
+      "step": 18734
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 52.715892757283235,
+      "learning_rate": 3.1528559963522886e-08,
+      "loss": 0.4458,
+      "step": 18735
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.80761163920397,
+      "learning_rate": 3.139489287211639e-08,
+      "loss": 0.3628,
+      "step": 18736
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 56.0372361465631,
+      "learning_rate": 3.1261509282009e-08,
+      "loss": 0.372,
+      "step": 18737
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 42.38579608593977,
+      "learning_rate": 3.112840919699212e-08,
+      "loss": 0.3565,
+      "step": 18738
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 58.4874867236454,
+      "learning_rate": 3.099559262085272e-08,
+      "loss": 0.4065,
+      "step": 18739
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 49.86595157775254,
+      "learning_rate": 3.086305955736557e-08,
+      "loss": 0.3853,
+      "step": 18740
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 48.96003815083794,
+      "learning_rate": 3.073081001030209e-08,
+      "loss": 0.4073,
+      "step": 18741
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 51.19328047237026,
+      "learning_rate": 3.05988439834215e-08,
+      "loss": 0.4118,
+      "step": 18742
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 55.07331225856379,
+      "learning_rate": 3.046716148047968e-08,
+      "loss": 0.3823,
+      "step": 18743
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.6558107776651,
+      "learning_rate": 3.033576250521919e-08,
+      "loss": 0.3813,
+      "step": 18744
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 50.52065846468626,
+      "learning_rate": 3.0204647061375936e-08,
+      "loss": 0.4504,
+      "step": 18745
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 59.54813592076736,
+      "learning_rate": 3.0073815152681374e-08,
+      "loss": 0.4949,
+      "step": 18746
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 42.313265217407746,
+      "learning_rate": 2.994326678285586e-08,
+      "loss": 0.4035,
+      "step": 18747
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 52.406270690804696,
+      "learning_rate": 2.981300195561088e-08,
+      "loss": 0.4343,
+      "step": 18748
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 57.488514117913454,
+      "learning_rate": 2.968302067465234e-08,
+      "loss": 0.4171,
+      "step": 18749
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 62.12756929515063,
+      "learning_rate": 2.955332294367508e-08,
+      "loss": 0.4569,
+      "step": 18750
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 49.80710086734772,
+      "learning_rate": 2.942390876636947e-08,
+      "loss": 0.458,
+      "step": 18751
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 55.959689290293205,
+      "learning_rate": 2.9294778146415902e-08,
+      "loss": 0.4231,
+      "step": 18752
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 59.49220109364585,
+      "learning_rate": 2.916593108748589e-08,
+      "loss": 0.4321,
+      "step": 18753
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 49.09496901739122,
+      "learning_rate": 2.903736759324316e-08,
+      "loss": 0.4513,
+      "step": 18754
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.03081502710969,
+      "learning_rate": 2.8909087667345905e-08,
+      "loss": 0.4259,
+      "step": 18755
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 51.23031130092799,
+      "learning_rate": 2.878109131344009e-08,
+      "loss": 0.3758,
+      "step": 18756
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 47.1562533166883,
+      "learning_rate": 2.8653378535168363e-08,
+      "loss": 0.3903,
+      "step": 18757
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 52.745676618281415,
+      "learning_rate": 2.852594933616004e-08,
+      "loss": 0.4158,
+      "step": 18758
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 41.53114262574885,
+      "learning_rate": 2.8398803720042223e-08,
+      "loss": 0.4155,
+      "step": 18759
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.01029188295827,
+      "learning_rate": 2.8271941690427574e-08,
+      "loss": 0.3694,
+      "step": 18760
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 38.38017475390543,
+      "learning_rate": 2.8145363250926537e-08,
+      "loss": 0.3785,
+      "step": 18761
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.96162530753025,
+      "learning_rate": 2.801906840513735e-08,
+      "loss": 0.34,
+      "step": 18762
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 55.87592044814637,
+      "learning_rate": 2.7893057156653802e-08,
+      "loss": 0.4041,
+      "step": 18763
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 68.83807750021184,
+      "learning_rate": 2.776732950905636e-08,
+      "loss": 0.3644,
+      "step": 18764
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 42.91128397151198,
+      "learning_rate": 2.764188546592439e-08,
+      "loss": 0.3688,
+      "step": 18765
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 52.05826544055243,
+      "learning_rate": 2.7516725030821702e-08,
+      "loss": 0.4484,
+      "step": 18766
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 51.97249103048241,
+      "learning_rate": 2.7391848207311e-08,
+      "loss": 0.3856,
+      "step": 18767
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 55.97481391555502,
+      "learning_rate": 2.7267254998941673e-08,
+      "loss": 0.454,
+      "step": 18768
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 51.756850194658256,
+      "learning_rate": 2.714294540925866e-08,
+      "loss": 0.4475,
+      "step": 18769
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.998978935187615,
+      "learning_rate": 2.701891944179469e-08,
+      "loss": 0.4133,
+      "step": 18770
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 43.38051481669582,
+      "learning_rate": 2.6895177100079163e-08,
+      "loss": 0.3929,
+      "step": 18771
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 48.057619459683124,
+      "learning_rate": 2.6771718387631485e-08,
+      "loss": 0.4639,
+      "step": 18772
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 56.28641869228343,
+      "learning_rate": 2.6648543307962183e-08,
+      "loss": 0.4304,
+      "step": 18773
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 51.58303783262264,
+      "learning_rate": 2.65256518645729e-08,
+      "loss": 0.4211,
+      "step": 18774
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 50.40284409105041,
+      "learning_rate": 2.6403044060959726e-08,
+      "loss": 0.3744,
+      "step": 18775
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 55.79243012699579,
+      "learning_rate": 2.6280719900609874e-08,
+      "loss": 0.4758,
+      "step": 18776
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 54.46564453341769,
+      "learning_rate": 2.6158679387002782e-08,
+      "loss": 0.3526,
+      "step": 18777
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 44.57455425221435,
+      "learning_rate": 2.60369225236079e-08,
+      "loss": 0.3968,
+      "step": 18778
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 60.317673325902476,
+      "learning_rate": 2.591544931388801e-08,
+      "loss": 0.4271,
+      "step": 18779
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 38.98874295825729,
+      "learning_rate": 2.5794259761298124e-08,
+      "loss": 0.3737,
+      "step": 18780
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 60.23482337319694,
+      "learning_rate": 2.5673353869284378e-08,
+      "loss": 0.4603,
+      "step": 18781
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 50.827860294735146,
+      "learning_rate": 2.5552731641286243e-08,
+      "loss": 0.4686,
+      "step": 18782
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 61.921856106512905,
+      "learning_rate": 2.5432393080733197e-08,
+      "loss": 0.4306,
+      "step": 18783
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 51.809700226460436,
+      "learning_rate": 2.5312338191048057e-08,
+      "loss": 0.4251,
+      "step": 18784
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 70.83263694720054,
+      "learning_rate": 2.5192566975644762e-08,
+      "loss": 0.484,
+      "step": 18785
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 45.18238129386111,
+      "learning_rate": 2.507307943793058e-08,
+      "loss": 0.4185,
+      "step": 18786
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 57.58800703984306,
+      "learning_rate": 2.495387558130169e-08,
+      "loss": 0.3985,
+      "step": 18787
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 40.68507249683879,
+      "learning_rate": 2.4834955409149818e-08,
+      "loss": 0.3939,
+      "step": 18788
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 52.467360114023215,
+      "learning_rate": 2.471631892485671e-08,
+      "loss": 0.4331,
+      "step": 18789
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 48.32750769836671,
+      "learning_rate": 2.4597966131796324e-08,
+      "loss": 0.4393,
+      "step": 18790
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 42.292168791222046,
+      "learning_rate": 2.4479897033333756e-08,
+      "loss": 0.4447,
+      "step": 18791
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.17569407148274,
+      "learning_rate": 2.4362111632827423e-08,
+      "loss": 0.4015,
+      "step": 18792
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 49.54124344093403,
+      "learning_rate": 2.4244609933627982e-08,
+      "loss": 0.3807,
+      "step": 18793
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 50.62670572021096,
+      "learning_rate": 2.4127391939076094e-08,
+      "loss": 0.4338,
+      "step": 18794
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 47.763041967069036,
+      "learning_rate": 2.401045765250465e-08,
+      "loss": 0.3799,
+      "step": 18795
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.06640255983904,
+      "learning_rate": 2.3893807077239872e-08,
+      "loss": 0.3781,
+      "step": 18796
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 52.68058177104664,
+      "learning_rate": 2.3777440216600224e-08,
+      "loss": 0.4688,
+      "step": 18797
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 62.42163883398496,
+      "learning_rate": 2.3661357073894166e-08,
+      "loss": 0.457,
+      "step": 18798
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 54.32181298465741,
+      "learning_rate": 2.3545557652422392e-08,
+      "loss": 0.3971,
+      "step": 18799
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.16322220795933,
+      "learning_rate": 2.343004195547893e-08,
+      "loss": 0.3672,
+      "step": 18800
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.468882152007744,
+      "learning_rate": 2.3314809986348942e-08,
+      "loss": 0.4058,
+      "step": 18801
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 60.558605714511486,
+      "learning_rate": 2.3199861748309793e-08,
+      "loss": 0.4111,
+      "step": 18802
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 41.6183129687091,
+      "learning_rate": 2.3085197244631097e-08,
+      "loss": 0.4352,
+      "step": 18803
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 49.65304800936679,
+      "learning_rate": 2.2970816478572466e-08,
+      "loss": 0.4565,
+      "step": 18804
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 52.82391884957077,
+      "learning_rate": 2.2856719453386856e-08,
+      "loss": 0.4231,
+      "step": 18805
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 62.52647895284062,
+      "learning_rate": 2.2742906172320555e-08,
+      "loss": 0.4649,
+      "step": 18806
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 60.919675917950016,
+      "learning_rate": 2.262937663860876e-08,
+      "loss": 0.4614,
+      "step": 18807
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 54.90425970370661,
+      "learning_rate": 2.25161308554811e-08,
+      "loss": 0.4682,
+      "step": 18808
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 57.641501821506786,
+      "learning_rate": 2.240316882615834e-08,
+      "loss": 0.401,
+      "step": 18809
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 56.09120327579544,
+      "learning_rate": 2.2290490553852352e-08,
+      "loss": 0.4147,
+      "step": 18810
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.658514710105756,
+      "learning_rate": 2.217809604176835e-08,
+      "loss": 0.4119,
+      "step": 18811
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 45.58916908976279,
+      "learning_rate": 2.2065985293102664e-08,
+      "loss": 0.3612,
+      "step": 18812
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 41.528217991473554,
+      "learning_rate": 2.195415831104275e-08,
+      "loss": 0.3753,
+      "step": 18813
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 45.9447278285727,
+      "learning_rate": 2.1842615098769394e-08,
+      "loss": 0.3613,
+      "step": 18814
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 45.06971826821198,
+      "learning_rate": 2.1731355659456722e-08,
+      "loss": 0.3978,
+      "step": 18815
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 63.48926192075274,
+      "learning_rate": 2.162037999626554e-08,
+      "loss": 0.4207,
+      "step": 18816
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.90270104981675,
+      "learning_rate": 2.1509688112354432e-08,
+      "loss": 0.4947,
+      "step": 18817
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.64318624904864,
+      "learning_rate": 2.139928001086977e-08,
+      "loss": 0.3951,
+      "step": 18818
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 48.18458261329481,
+      "learning_rate": 2.128915569495238e-08,
+      "loss": 0.5005,
+      "step": 18819
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.94322559773298,
+      "learning_rate": 2.1179315167734194e-08,
+      "loss": 0.3715,
+      "step": 18820
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 50.056399760723004,
+      "learning_rate": 2.1069758432339382e-08,
+      "loss": 0.4597,
+      "step": 18821
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 47.2449207452034,
+      "learning_rate": 2.096048549188212e-08,
+      "loss": 0.4183,
+      "step": 18822
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 56.073362180624336,
+      "learning_rate": 2.0851496349472144e-08,
+      "loss": 0.4537,
+      "step": 18823
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 59.15132655448278,
+      "learning_rate": 2.0742791008206974e-08,
+      "loss": 0.4461,
+      "step": 18824
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 51.6032661705417,
+      "learning_rate": 2.0634369471179692e-08,
+      "loss": 0.3865,
+      "step": 18825
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 43.03272406574968,
+      "learning_rate": 2.052623174147339e-08,
+      "loss": 0.3808,
+      "step": 18826
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.67284134749924,
+      "learning_rate": 2.0418377822162272e-08,
+      "loss": 0.3693,
+      "step": 18827
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 43.39732956866945,
+      "learning_rate": 2.0310807716316107e-08,
+      "loss": 0.3802,
+      "step": 18828
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 42.66881011102627,
+      "learning_rate": 2.0203521426991333e-08,
+      "loss": 0.3906,
+      "step": 18829
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.467606987713296,
+      "learning_rate": 2.0096518957241072e-08,
+      "loss": 0.4591,
+      "step": 18830
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 48.14845923566103,
+      "learning_rate": 1.9989800310107333e-08,
+      "loss": 0.4457,
+      "step": 18831
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 43.09029406659008,
+      "learning_rate": 1.9883365488625462e-08,
+      "loss": 0.423,
+      "step": 18832
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.72704750564341,
+      "learning_rate": 1.977721449582304e-08,
+      "loss": 0.3545,
+      "step": 18833
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 58.27854499573506,
+      "learning_rate": 1.9671347334717648e-08,
+      "loss": 0.4514,
+      "step": 18834
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 42.47709540642718,
+      "learning_rate": 1.956576400832133e-08,
+      "loss": 0.4073,
+      "step": 18835
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 62.63605050758534,
+      "learning_rate": 1.9460464519636124e-08,
+      "loss": 0.4707,
+      "step": 18836
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.38623647134903,
+      "learning_rate": 1.9355448871657413e-08,
+      "loss": 0.4021,
+      "step": 18837
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 53.92145586583217,
+      "learning_rate": 1.9250717067370583e-08,
+      "loss": 0.3772,
+      "step": 18838
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 47.32186343463655,
+      "learning_rate": 1.9146269109755477e-08,
+      "loss": 0.3955,
+      "step": 18839
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 51.66917380536015,
+      "learning_rate": 1.904210500178083e-08,
+      "loss": 0.3524,
+      "step": 18840
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 41.45906759055963,
+      "learning_rate": 1.8938224746410938e-08,
+      "loss": 0.3837,
+      "step": 18841
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 49.18843187760564,
+      "learning_rate": 1.8834628346598993e-08,
+      "loss": 0.3608,
+      "step": 18842
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 71.87260733229299,
+      "learning_rate": 1.8731315805290418e-08,
+      "loss": 0.4437,
+      "step": 18843
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 49.64195754940578,
+      "learning_rate": 1.862828712542508e-08,
+      "loss": 0.3472,
+      "step": 18844
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.90357906150488,
+      "learning_rate": 1.8525542309932865e-08,
+      "loss": 0.4537,
+      "step": 18845
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.59575010758403,
+      "learning_rate": 1.8423081361734762e-08,
+      "loss": 0.3469,
+      "step": 18846
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 61.26333982043404,
+      "learning_rate": 1.832090428374511e-08,
+      "loss": 0.4513,
+      "step": 18847
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 44.050592634347446,
+      "learning_rate": 1.8219011078869365e-08,
+      "loss": 0.4389,
+      "step": 18848
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 58.71110185611585,
+      "learning_rate": 1.8117401750006314e-08,
+      "loss": 0.4307,
+      "step": 18849
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 43.17272206710084,
+      "learning_rate": 1.801607630004476e-08,
+      "loss": 0.3703,
+      "step": 18850
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 48.8610808852705,
+      "learning_rate": 1.7915034731867952e-08,
+      "loss": 0.4007,
+      "step": 18851
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 46.35181677301856,
+      "learning_rate": 1.781427704834693e-08,
+      "loss": 0.4249,
+      "step": 18852
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 53.427884130140775,
+      "learning_rate": 1.7713803252348283e-08,
+      "loss": 0.5342,
+      "step": 18853
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 47.45104114477564,
+      "learning_rate": 1.761361334673084e-08,
+      "loss": 0.4209,
+      "step": 18854
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 53.767286776620374,
+      "learning_rate": 1.7513707334341212e-08,
+      "loss": 0.4455,
+      "step": 18855
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 46.612502379045246,
+      "learning_rate": 1.741408521802379e-08,
+      "loss": 0.4145,
+      "step": 18856
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 64.04364309278724,
+      "learning_rate": 1.7314747000608532e-08,
+      "loss": 0.4279,
+      "step": 18857
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 55.50971161011675,
+      "learning_rate": 1.721569268492318e-08,
+      "loss": 0.4085,
+      "step": 18858
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 49.38454180880418,
+      "learning_rate": 1.7116922273783255e-08,
+      "loss": 0.4391,
+      "step": 18859
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 52.537833750370595,
+      "learning_rate": 1.7018435769998732e-08,
+      "loss": 0.408,
+      "step": 18860
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 44.55380419919606,
+      "learning_rate": 1.6920233176369594e-08,
+      "loss": 0.4023,
+      "step": 18861
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 44.27554988853475,
+      "learning_rate": 1.6822314495689164e-08,
+      "loss": 0.4016,
+      "step": 18862
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 47.859858511370085,
+      "learning_rate": 1.6724679730742986e-08,
+      "loss": 0.4305,
+      "step": 18863
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 57.10875010846509,
+      "learning_rate": 1.6627328884305516e-08,
+      "loss": 0.4144,
+      "step": 18864
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 57.45513691723927,
+      "learning_rate": 1.6530261959147863e-08,
+      "loss": 0.4275,
+      "step": 18865
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 41.53688263711288,
+      "learning_rate": 1.6433478958028938e-08,
+      "loss": 0.3905,
+      "step": 18866
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 47.983153883843706,
+      "learning_rate": 1.6336979883700976e-08,
+      "loss": 0.4436,
+      "step": 18867
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 51.89518550048335,
+      "learning_rate": 1.6240764738909566e-08,
+      "loss": 0.4817,
+      "step": 18868
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 40.970749748768284,
+      "learning_rate": 1.6144833526390292e-08,
+      "loss": 0.3118,
+      "step": 18869
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 52.087943565287105,
+      "learning_rate": 1.6049186248872084e-08,
+      "loss": 0.4471,
+      "step": 18870
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 61.675347666539444,
+      "learning_rate": 1.595382290907388e-08,
+      "loss": 0.4457,
+      "step": 18871
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 56.124935744828655,
+      "learning_rate": 1.585874350970906e-08,
+      "loss": 0.4721,
+      "step": 18872
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 49.34359785639939,
+      "learning_rate": 1.5763948053481026e-08,
+      "loss": 0.4425,
+      "step": 18873
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 40.82601803563241,
+      "learning_rate": 1.566943654308539e-08,
+      "loss": 0.3452,
+      "step": 18874
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.4376297389108,
+      "learning_rate": 1.557520898121001e-08,
+      "loss": 0.3998,
+      "step": 18875
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 52.47225931778963,
+      "learning_rate": 1.5481265370536068e-08,
+      "loss": 0.4751,
+      "step": 18876
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 50.13421110532892,
+      "learning_rate": 1.5387605713732545e-08,
+      "loss": 0.4772,
+      "step": 18877
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 68.71342728675958,
+      "learning_rate": 1.5294230013466194e-08,
+      "loss": 0.539,
+      "step": 18878
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 42.1989914622176,
+      "learning_rate": 1.520113827239045e-08,
+      "loss": 0.3868,
+      "step": 18879
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 60.905529865452515,
+      "learning_rate": 1.510833049315319e-08,
+      "loss": 0.3575,
+      "step": 18880
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 80.49851711755507,
+      "learning_rate": 1.501580667839453e-08,
+      "loss": 0.3649,
+      "step": 18881
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.11686069967807,
+      "learning_rate": 1.4923566830744586e-08,
+      "loss": 0.4189,
+      "step": 18882
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.08366069098272,
+      "learning_rate": 1.4831610952827925e-08,
+      "loss": 0.4527,
+      "step": 18883
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 58.2035480908883,
+      "learning_rate": 1.4739939047259122e-08,
+      "loss": 0.5035,
+      "step": 18884
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 40.60597604615432,
+      "learning_rate": 1.4648551116644982e-08,
+      "loss": 0.3963,
+      "step": 18885
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 54.185308449502536,
+      "learning_rate": 1.4557447163584538e-08,
+      "loss": 0.4618,
+      "step": 18886
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 54.70625910297477,
+      "learning_rate": 1.4466627190669047e-08,
+      "loss": 0.3959,
+      "step": 18887
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 53.102590425528476,
+      "learning_rate": 1.4376091200482002e-08,
+      "loss": 0.4893,
+      "step": 18888
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 46.11647407694644,
+      "learning_rate": 1.4285839195596896e-08,
+      "loss": 0.4311,
+      "step": 18889
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.85661099249171,
+      "learning_rate": 1.4195871178580567e-08,
+      "loss": 0.4117,
+      "step": 18890
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 58.44537527292341,
+      "learning_rate": 1.410618715199319e-08,
+      "loss": 0.4707,
+      "step": 18891
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 46.08472323204288,
+      "learning_rate": 1.4016787118383835e-08,
+      "loss": 0.3607,
+      "step": 18892
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 49.2325293250392,
+      "learning_rate": 1.3927671080294913e-08,
+      "loss": 0.396,
+      "step": 18893
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.004478801945,
+      "learning_rate": 1.3838839040262175e-08,
+      "loss": 0.4488,
+      "step": 18894
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 44.087551822204674,
+      "learning_rate": 1.3750291000811377e-08,
+      "loss": 0.3396,
+      "step": 18895
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.71918841171469,
+      "learning_rate": 1.3662026964459396e-08,
+      "loss": 0.4389,
+      "step": 18896
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 42.433653976031074,
+      "learning_rate": 1.3574046933717556e-08,
+      "loss": 0.3774,
+      "step": 18897
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 38.98394732957386,
+      "learning_rate": 1.3486350911089407e-08,
+      "loss": 0.4356,
+      "step": 18898
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 46.269885669927284,
+      "learning_rate": 1.3398938899066294e-08,
+      "loss": 0.4373,
+      "step": 18899
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 48.758574846710864,
+      "learning_rate": 1.3311810900135113e-08,
+      "loss": 0.3977,
+      "step": 18900
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 40.19676203213251,
+      "learning_rate": 1.3224966916774995e-08,
+      "loss": 0.3472,
+      "step": 18901
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 49.38331806278326,
+      "learning_rate": 1.3138406951453963e-08,
+      "loss": 0.4219,
+      "step": 18902
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 53.02905223026109,
+      "learning_rate": 1.3052131006634494e-08,
+      "loss": 0.4268,
+      "step": 18903
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 50.840073010004296,
+      "learning_rate": 1.296613908477018e-08,
+      "loss": 0.4491,
+      "step": 18904
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 52.12273468382643,
+      "learning_rate": 1.2880431188306841e-08,
+      "loss": 0.3742,
+      "step": 18905
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 43.32935716959057,
+      "learning_rate": 1.2795007319681418e-08,
+      "loss": 0.4178,
+      "step": 18906
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 55.03273594857627,
+      "learning_rate": 1.270986748132419e-08,
+      "loss": 0.4031,
+      "step": 18907
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 44.29593446732162,
+      "learning_rate": 1.262501167565655e-08,
+      "loss": 0.388,
+      "step": 18908
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 58.592916327077056,
+      "learning_rate": 1.2540439905089907e-08,
+      "loss": 0.4655,
+      "step": 18909
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 47.73634848253396,
+      "learning_rate": 1.2456152172031222e-08,
+      "loss": 0.3862,
+      "step": 18910
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 65.65370282221159,
+      "learning_rate": 1.2372148478876356e-08,
+      "loss": 0.4967,
+      "step": 18911
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 45.650501090325626,
+      "learning_rate": 1.228842882801562e-08,
+      "loss": 0.3793,
+      "step": 18912
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 53.06512286530083,
+      "learning_rate": 1.2204993221829331e-08,
+      "loss": 0.3671,
+      "step": 18913
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 39.83553982985252,
+      "learning_rate": 1.2121841662690037e-08,
+      "loss": 0.3508,
+      "step": 18914
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 50.83336744880574,
+      "learning_rate": 1.2038974152963623e-08,
+      "loss": 0.428,
+      "step": 18915
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 57.09032408609778,
+      "learning_rate": 1.1956390695004871e-08,
+      "loss": 0.3915,
+      "step": 18916
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 50.128833552809965,
+      "learning_rate": 1.1874091291164125e-08,
+      "loss": 0.3997,
+      "step": 18917
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 46.73832453316431,
+      "learning_rate": 1.1792075943781733e-08,
+      "loss": 0.38,
+      "step": 18918
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 53.541115443537706,
+      "learning_rate": 1.1710344655189166e-08,
+      "loss": 0.4016,
+      "step": 18919
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 42.688528224463596,
+      "learning_rate": 1.1628897427711227e-08,
+      "loss": 0.3393,
+      "step": 18920
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 49.05759786425294,
+      "learning_rate": 1.1547734263664957e-08,
+      "loss": 0.3901,
+      "step": 18921
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 44.20578133027076,
+      "learning_rate": 1.1466855165357393e-08,
+      "loss": 0.3918,
+      "step": 18922
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 48.98413788709184,
+      "learning_rate": 1.1386260135090033e-08,
+      "loss": 0.4184,
+      "step": 18923
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.38172549316224,
+      "learning_rate": 1.1305949175154374e-08,
+      "loss": 0.4349,
+      "step": 18924
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 43.359356733438574,
+      "learning_rate": 1.1225922287834146e-08,
+      "loss": 0.3778,
+      "step": 18925
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 60.54019123422736,
+      "learning_rate": 1.1146179475404195e-08,
+      "loss": 0.5019,
+      "step": 18926
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 52.11356088277201,
+      "learning_rate": 1.1066720740134928e-08,
+      "loss": 0.4201,
+      "step": 18927
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 58.19500379599565,
+      "learning_rate": 1.098754608428454e-08,
+      "loss": 0.4511,
+      "step": 18928
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 44.08834393707238,
+      "learning_rate": 1.090865551010456e-08,
+      "loss": 0.3886,
+      "step": 18929
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 48.044848302259695,
+      "learning_rate": 1.0830049019839862e-08,
+      "loss": 0.3791,
+      "step": 18930
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 56.852281044564016,
+      "learning_rate": 1.0751726615724212e-08,
+      "loss": 0.462,
+      "step": 18931
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 51.45689203146051,
+      "learning_rate": 1.0673688299985829e-08,
+      "loss": 0.3957,
+      "step": 18932
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 41.96182293793124,
+      "learning_rate": 1.059593407484516e-08,
+      "loss": 0.4045,
+      "step": 18933
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 52.67898911215872,
+      "learning_rate": 1.0518463942511547e-08,
+      "loss": 0.4132,
+      "step": 18934
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.15989862057415,
+      "learning_rate": 1.0441277905188785e-08,
+      "loss": 0.4179,
+      "step": 18935
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 52.451269622069596,
+      "learning_rate": 1.0364375965074003e-08,
+      "loss": 0.4628,
+      "step": 18936
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.11621122087254,
+      "learning_rate": 1.0287758124351011e-08,
+      "loss": 0.3776,
+      "step": 18937
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 44.39754624674582,
+      "learning_rate": 1.0211424385201396e-08,
+      "loss": 0.3581,
+      "step": 18938
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 46.85576293395006,
+      "learning_rate": 1.0135374749794536e-08,
+      "loss": 0.3873,
+      "step": 18939
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 45.65346345746652,
+      "learning_rate": 1.0059609220293142e-08,
+      "loss": 0.3884,
+      "step": 18940
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 48.36787656830204,
+      "learning_rate": 9.984127798853271e-09,
+      "loss": 0.3943,
+      "step": 18941
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 49.244187094293885,
+      "learning_rate": 9.908930487620983e-09,
+      "loss": 0.4104,
+      "step": 18942
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 51.5478874262065,
+      "learning_rate": 9.834017288734565e-09,
+      "loss": 0.4481,
+      "step": 18943
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 51.430397637239516,
+      "learning_rate": 9.759388204323428e-09,
+      "loss": 0.4191,
+      "step": 18944
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 65.16468729328591,
+      "learning_rate": 9.685043236512537e-09,
+      "loss": 0.5702,
+      "step": 18945
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 52.48780863762688,
+      "learning_rate": 9.610982387414647e-09,
+      "loss": 0.4647,
+      "step": 18946
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 43.948957216479606,
+      "learning_rate": 9.53720565913585e-09,
+      "loss": 0.3853,
+      "step": 18947
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 51.730231494470395,
+      "learning_rate": 9.46371305377447e-09,
+      "loss": 0.3839,
+      "step": 18948
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 60.16115184282377,
+      "learning_rate": 9.390504573422166e-09,
+      "loss": 0.4263,
+      "step": 18949
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 43.41537267394284,
+      "learning_rate": 9.317580220158385e-09,
+      "loss": 0.3373,
+      "step": 18950
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 42.130027476649694,
+      "learning_rate": 9.244939996058134e-09,
+      "loss": 0.3459,
+      "step": 18951
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 42.0891051699808,
+      "learning_rate": 9.172583903187537e-09,
+      "loss": 0.3883,
+      "step": 18952
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 63.248377700023866,
+      "learning_rate": 9.10051194360495e-09,
+      "loss": 0.4875,
+      "step": 18953
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.490466772409405,
+      "learning_rate": 9.028724119358734e-09,
+      "loss": 0.3896,
+      "step": 18954
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 49.53205218222099,
+      "learning_rate": 8.957220432490587e-09,
+      "loss": 0.4272,
+      "step": 18955
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.48198159434902,
+      "learning_rate": 8.886000885034441e-09,
+      "loss": 0.4125,
+      "step": 18956
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 51.505337531144114,
+      "learning_rate": 8.815065479016449e-09,
+      "loss": 0.4118,
+      "step": 18957
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 62.88545074217638,
+      "learning_rate": 8.74441421645278e-09,
+      "loss": 0.4625,
+      "step": 18958
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 53.55014594901433,
+      "learning_rate": 8.674047099354044e-09,
+      "loss": 0.4917,
+      "step": 18959
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 59.24096536765458,
+      "learning_rate": 8.603964129719755e-09,
+      "loss": 0.4763,
+      "step": 18960
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 41.63131006772264,
+      "learning_rate": 8.534165309544985e-09,
+      "loss": 0.3279,
+      "step": 18961
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 39.152077600982146,
+      "learning_rate": 8.464650640812588e-09,
+      "loss": 0.353,
+      "step": 18962
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.81819992496927,
+      "learning_rate": 8.395420125500986e-09,
+      "loss": 0.3754,
+      "step": 18963
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 57.43963430629609,
+      "learning_rate": 8.326473765579713e-09,
+      "loss": 0.3939,
+      "step": 18964
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 48.474126747532566,
+      "learning_rate": 8.2578115630072e-09,
+      "loss": 0.4275,
+      "step": 18965
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 74.50221150591337,
+      "learning_rate": 8.189433519738555e-09,
+      "loss": 0.4848,
+      "step": 18966
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 45.98772822606244,
+      "learning_rate": 8.121339637716663e-09,
+      "loss": 0.3902,
+      "step": 18967
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 45.00840618243246,
+      "learning_rate": 8.053529918878867e-09,
+      "loss": 0.4039,
+      "step": 18968
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 43.679852112303145,
+      "learning_rate": 7.986004365153621e-09,
+      "loss": 0.3821,
+      "step": 18969
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 45.94710164834416,
+      "learning_rate": 7.918762978461613e-09,
+      "loss": 0.4359,
+      "step": 18970
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 53.25940140934653,
+      "learning_rate": 7.851805760714647e-09,
+      "loss": 0.4831,
+      "step": 18971
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 54.495975568904484,
+      "learning_rate": 7.785132713816756e-09,
+      "loss": 0.4476,
+      "step": 18972
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.516385193714406,
+      "learning_rate": 7.7187438396642e-09,
+      "loss": 0.3601,
+      "step": 18973
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 50.64963358382216,
+      "learning_rate": 7.652639140146579e-09,
+      "loss": 0.4085,
+      "step": 18974
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 44.7326013318038,
+      "learning_rate": 7.586818617141279e-09,
+      "loss": 0.3874,
+      "step": 18975
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 45.04461751594462,
+      "learning_rate": 7.521282272521246e-09,
+      "loss": 0.3581,
+      "step": 18976
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 47.53737131964505,
+      "learning_rate": 7.456030108151657e-09,
+      "loss": 0.415,
+      "step": 18977
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 55.02773439614357,
+      "learning_rate": 7.391062125886583e-09,
+      "loss": 0.3986,
+      "step": 18978
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 54.477399785970796,
+      "learning_rate": 7.326378327574546e-09,
+      "loss": 0.4889,
+      "step": 18979
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 42.78095308790143,
+      "learning_rate": 7.261978715054074e-09,
+      "loss": 0.3892,
+      "step": 18980
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 49.13131189443108,
+      "learning_rate": 7.197863290157037e-09,
+      "loss": 0.4571,
+      "step": 18981
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 60.60199653496614,
+      "learning_rate": 7.134032054707529e-09,
+      "loss": 0.4754,
+      "step": 18982
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 49.43141105954561,
+      "learning_rate": 7.070485010520767e-09,
+      "loss": 0.3678,
+      "step": 18983
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 58.4523561575464,
+      "learning_rate": 7.007222159404192e-09,
+      "loss": 0.4534,
+      "step": 18984
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 66.16670042160415,
+      "learning_rate": 6.944243503155257e-09,
+      "loss": 0.4952,
+      "step": 18985
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 44.802108187544036,
+      "learning_rate": 6.881549043568081e-09,
+      "loss": 0.4331,
+      "step": 18986
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 47.74485681206935,
+      "learning_rate": 6.819138782422352e-09,
+      "loss": 0.3662,
+      "step": 18987
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 49.948849368279014,
+      "learning_rate": 6.757012721494427e-09,
+      "loss": 0.442,
+      "step": 18988
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 53.52611356322527,
+      "learning_rate": 6.6951708625517805e-09,
+      "loss": 0.4257,
+      "step": 18989
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 43.97129745532457,
+      "learning_rate": 6.633613207351897e-09,
+      "loss": 0.402,
+      "step": 18990
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.865973995816304,
+      "learning_rate": 6.5723397576467066e-09,
+      "loss": 0.4098,
+      "step": 18991
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 54.872546017287846,
+      "learning_rate": 6.51135051517815e-09,
+      "loss": 0.4312,
+      "step": 18992
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 40.64916821434664,
+      "learning_rate": 6.450645481681506e-09,
+      "loss": 0.3276,
+      "step": 18993
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 52.41999936337054,
+      "learning_rate": 6.39022465888095e-09,
+      "loss": 0.4021,
+      "step": 18994
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.88049368756028,
+      "learning_rate": 6.330088048497329e-09,
+      "loss": 0.4343,
+      "step": 18995
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 46.24413937733281,
+      "learning_rate": 6.270235652239276e-09,
+      "loss": 0.4714,
+      "step": 18996
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 45.0611504129115,
+      "learning_rate": 6.2106674718098725e-09,
+      "loss": 0.393,
+      "step": 18997
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 47.13002774563755,
+      "learning_rate": 6.1513835089033194e-09,
+      "loss": 0.4155,
+      "step": 18998
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 43.25287376349033,
+      "learning_rate": 6.0923837652038245e-09,
+      "loss": 0.3961,
+      "step": 18999
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 42.963684643508174,
+      "learning_rate": 6.033668242392266e-09,
+      "loss": 0.3508,
+      "step": 19000
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 45.57093733436501,
+      "learning_rate": 5.975236942136197e-09,
+      "loss": 0.4156,
+      "step": 19001
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 46.95278571541017,
+      "learning_rate": 5.917089866097625e-09,
+      "loss": 0.4257,
+      "step": 19002
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 48.943341380354674,
+      "learning_rate": 5.85922701593078e-09,
+      "loss": 0.3772,
+      "step": 19003
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 51.14126013091298,
+      "learning_rate": 5.801648393282122e-09,
+      "loss": 0.4889,
+      "step": 19004
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 61.20581718018338,
+      "learning_rate": 5.744353999787011e-09,
+      "loss": 0.4086,
+      "step": 19005
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 53.80907309206157,
+      "learning_rate": 5.6873438370763645e-09,
+      "loss": 0.4437,
+      "step": 19006
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 40.822877157584294,
+      "learning_rate": 5.630617906771108e-09,
+      "loss": 0.3533,
+      "step": 19007
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 51.751505571845826,
+      "learning_rate": 5.574176210484394e-09,
+      "loss": 0.4071,
+      "step": 19008
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 47.61357886868973,
+      "learning_rate": 5.518018749821607e-09,
+      "loss": 0.4229,
+      "step": 19009
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 51.14923393872843,
+      "learning_rate": 5.462145526380358e-09,
+      "loss": 0.4344,
+      "step": 19010
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 48.02077372858261,
+      "learning_rate": 5.406556541748265e-09,
+      "loss": 0.4105,
+      "step": 19011
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 55.58316375824751,
+      "learning_rate": 5.351251797507395e-09,
+      "loss": 0.4083,
+      "step": 19012
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 49.11013560323572,
+      "learning_rate": 5.296231295229826e-09,
+      "loss": 0.4064,
+      "step": 19013
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 55.430622624687764,
+      "learning_rate": 5.2414950364809706e-09,
+      "loss": 0.4168,
+      "step": 19014
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 46.144316466511455,
+      "learning_rate": 5.187043022817362e-09,
+      "loss": 0.415,
+      "step": 19015
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 54.42187588932936,
+      "learning_rate": 5.1328752557877615e-09,
+      "loss": 0.4119,
+      "step": 19016
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 53.81782828345653,
+      "learning_rate": 5.078991736932049e-09,
+      "loss": 0.4151,
+      "step": 19017
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 51.2429384809606,
+      "learning_rate": 5.025392467783441e-09,
+      "loss": 0.3671,
+      "step": 19018
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 54.44704975843887,
+      "learning_rate": 4.972077449865165e-09,
+      "loss": 0.4642,
+      "step": 19019
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 53.59261143311877,
+      "learning_rate": 4.919046684693785e-09,
+      "loss": 0.4338,
+      "step": 19020
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 69.49953917480939,
+      "learning_rate": 4.8663001737780934e-09,
+      "loss": 0.4767,
+      "step": 19021
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 48.27019491200187,
+      "learning_rate": 4.813837918618003e-09,
+      "loss": 0.4223,
+      "step": 19022
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.69667727907944,
+      "learning_rate": 4.761659920705652e-09,
+      "loss": 0.394,
+      "step": 19023
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 53.834491369863805,
+      "learning_rate": 4.7097661815243e-09,
+      "loss": 0.4958,
+      "step": 19024
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 48.99198581396728,
+      "learning_rate": 4.6581567025494324e-09,
+      "loss": 0.4143,
+      "step": 19025
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 59.47972124401259,
+      "learning_rate": 4.6068314852498745e-09,
+      "loss": 0.4478,
+      "step": 19026
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 51.60326670648753,
+      "learning_rate": 4.555790531085569e-09,
+      "loss": 0.3633,
+      "step": 19027
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 41.71881188663078,
+      "learning_rate": 4.505033841506468e-09,
+      "loss": 0.3906,
+      "step": 19028
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 48.85049118607184,
+      "learning_rate": 4.454561417958081e-09,
+      "loss": 0.4011,
+      "step": 19029
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 55.91181670738706,
+      "learning_rate": 4.404373261873707e-09,
+      "loss": 0.5029,
+      "step": 19030
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.141253231174865,
+      "learning_rate": 4.354469374682202e-09,
+      "loss": 0.3789,
+      "step": 19031
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 70.5525440933394,
+      "learning_rate": 4.304849757802432e-09,
+      "loss": 0.4349,
+      "step": 19032
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 54.600033433959844,
+      "learning_rate": 4.25551441264549e-09,
+      "loss": 0.4468,
+      "step": 19033
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 47.117064594607555,
+      "learning_rate": 4.206463340614697e-09,
+      "loss": 0.4209,
+      "step": 19034
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 54.351584884816596,
+      "learning_rate": 4.157696543103384e-09,
+      "loss": 0.4151,
+      "step": 19035
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 52.53985995343337,
+      "learning_rate": 4.109214021500441e-09,
+      "loss": 0.4594,
+      "step": 19036
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 46.82991081450053,
+      "learning_rate": 4.061015777184762e-09,
+      "loss": 0.4235,
+      "step": 19037
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 53.43792354125887,
+      "learning_rate": 4.013101811525255e-09,
+      "loss": 0.4501,
+      "step": 19038
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.42333598306211,
+      "learning_rate": 3.9654721258863826e-09,
+      "loss": 0.4387,
+      "step": 19039
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 45.763915363754975,
+      "learning_rate": 3.918126721621507e-09,
+      "loss": 0.4015,
+      "step": 19040
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 42.6963055829341,
+      "learning_rate": 3.871065600078439e-09,
+      "loss": 0.4012,
+      "step": 19041
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 56.50858601172961,
+      "learning_rate": 3.824288762593886e-09,
+      "loss": 0.3926,
+      "step": 19042
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 44.811746311633165,
+      "learning_rate": 3.777796210499008e-09,
+      "loss": 0.4274,
+      "step": 19043
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 47.07473610506728,
+      "learning_rate": 3.731587945116078e-09,
+      "loss": 0.4067,
+      "step": 19044
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 51.293539184531255,
+      "learning_rate": 3.6856639677584906e-09,
+      "loss": 0.396,
+      "step": 19045
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 45.375135910208954,
+      "learning_rate": 3.640024279734089e-09,
+      "loss": 0.3962,
+      "step": 19046
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 45.735604470698775,
+      "learning_rate": 3.5946688823396137e-09,
+      "loss": 0.3664,
+      "step": 19047
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 38.85254460069023,
+      "learning_rate": 3.5495977768640332e-09,
+      "loss": 0.3566,
+      "step": 19048
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 58.657468406123435,
+      "learning_rate": 3.5048109645907657e-09,
+      "loss": 0.4684,
+      "step": 19049
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 50.124660062707434,
+      "learning_rate": 3.460308446793237e-09,
+      "loss": 0.4916,
+      "step": 19050
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 51.081407559962926,
+      "learning_rate": 3.4160902247359904e-09,
+      "loss": 0.4474,
+      "step": 19051
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 45.94050937578615,
+      "learning_rate": 3.37215629967802e-09,
+      "loss": 0.4032,
+      "step": 19052
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 58.75361167682205,
+      "learning_rate": 3.3285066728672156e-09,
+      "loss": 0.4025,
+      "step": 19053
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 62.80346283310008,
+      "learning_rate": 3.285141345547027e-09,
+      "loss": 0.4886,
+      "step": 19054
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 46.582969211435945,
+      "learning_rate": 3.242060318948692e-09,
+      "loss": 0.372,
+      "step": 19055
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 45.92964012739214,
+      "learning_rate": 3.1992635942978965e-09,
+      "loss": 0.4502,
+      "step": 19056
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 42.58674505486666,
+      "learning_rate": 3.1567511728125555e-09,
+      "loss": 0.3975,
+      "step": 19057
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 36.97194780012555,
+      "learning_rate": 3.114523055700591e-09,
+      "loss": 0.3547,
+      "step": 19058
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 55.69555443978598,
+      "learning_rate": 3.0725792441643755e-09,
+      "loss": 0.4382,
+      "step": 19059
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 52.93585346638283,
+      "learning_rate": 3.0309197393962874e-09,
+      "loss": 0.4748,
+      "step": 19060
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 44.76961971121445,
+      "learning_rate": 2.9895445425798253e-09,
+      "loss": 0.4215,
+      "step": 19061
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 55.22353642373361,
+      "learning_rate": 2.9484536548929355e-09,
+      "loss": 0.4579,
+      "step": 19062
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 48.58290020922471,
+      "learning_rate": 2.907647077504683e-09,
+      "loss": 0.3449,
+      "step": 19063
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 53.80604839307121,
+      "learning_rate": 2.86712481157414e-09,
+      "loss": 0.4535,
+      "step": 19064
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 42.26685152935993,
+      "learning_rate": 2.826886858253719e-09,
+      "loss": 0.4118,
+      "step": 19065
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 56.48597277311259,
+      "learning_rate": 2.7869332186891697e-09,
+      "loss": 0.4131,
+      "step": 19066
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 46.92918829337257,
+      "learning_rate": 2.7472638940162498e-09,
+      "loss": 0.4306,
+      "step": 19067
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 48.96828485303272,
+      "learning_rate": 2.707878885362947e-09,
+      "loss": 0.4126,
+      "step": 19068
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 42.24924125215809,
+      "learning_rate": 2.6687781938483647e-09,
+      "loss": 0.4298,
+      "step": 19069
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 51.043622098042945,
+      "learning_rate": 2.629961820587168e-09,
+      "loss": 0.42,
+      "step": 19070
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 53.08912643499275,
+      "learning_rate": 2.5914297666806975e-09,
+      "loss": 0.4417,
+      "step": 19071
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 46.25871373132813,
+      "learning_rate": 2.5531820332247436e-09,
+      "loss": 0.4566,
+      "step": 19072
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 44.969064908916536,
+      "learning_rate": 2.515218621309545e-09,
+      "loss": 0.3856,
+      "step": 19073
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 47.73551901587406,
+      "learning_rate": 2.4775395320120187e-09,
+      "loss": 0.4233,
+      "step": 19074
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 59.1220906128534,
+      "learning_rate": 2.4401447664046398e-09,
+      "loss": 0.4856,
+      "step": 19075
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 51.13083515117847,
+      "learning_rate": 2.4030343255521115e-09,
+      "loss": 0.3919,
+      "step": 19076
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 65.62045325377613,
+      "learning_rate": 2.3662082105080365e-09,
+      "loss": 0.4788,
+      "step": 19077
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 47.817766644995906,
+      "learning_rate": 2.3296664223215747e-09,
+      "loss": 0.4133,
+      "step": 19078
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 59.06294802137528,
+      "learning_rate": 2.2934089620307852e-09,
+      "loss": 0.4671,
+      "step": 19079
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 41.432953741928564,
+      "learning_rate": 2.257435830665955e-09,
+      "loss": 0.3414,
+      "step": 19080
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 57.64913961094372,
+      "learning_rate": 2.2217470292518196e-09,
+      "loss": 0.4815,
+      "step": 19081
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 48.970183453152956,
+      "learning_rate": 2.1863425588020125e-09,
+      "loss": 0.3895,
+      "step": 19082
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 43.22311903907907,
+      "learning_rate": 2.1512224203246167e-09,
+      "loss": 0.3519,
+      "step": 19083
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 45.228318125483064,
+      "learning_rate": 2.116386614818833e-09,
+      "loss": 0.3725,
+      "step": 19084
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 54.903366499486985,
+      "learning_rate": 2.081835143272759e-09,
+      "loss": 0.4544,
+      "step": 19085
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 53.4447014401091,
+      "learning_rate": 2.0475680066711632e-09,
+      "loss": 0.4298,
+      "step": 19086
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 51.54339493043747,
+      "learning_rate": 2.0135852059888215e-09,
+      "loss": 0.3893,
+      "step": 19087
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 46.33285758948619,
+      "learning_rate": 1.9798867421905176e-09,
+      "loss": 0.408,
+      "step": 19088
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 44.82519142441903,
+      "learning_rate": 1.9464726162365944e-09,
+      "loss": 0.4172,
+      "step": 19089
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 51.35000118155325,
+      "learning_rate": 1.9133428290751822e-09,
+      "loss": 0.4521,
+      "step": 19090
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 48.1146695332465,
+      "learning_rate": 1.880497381651081e-09,
+      "loss": 0.4157,
+      "step": 19091
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 51.39835925081057,
+      "learning_rate": 1.8479362748957674e-09,
+      "loss": 0.4672,
+      "step": 19092
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 47.491295845207006,
+      "learning_rate": 1.8156595097362783e-09,
+      "loss": 0.4469,
+      "step": 19093
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 42.39898885407447,
+      "learning_rate": 1.783667087090768e-09,
+      "loss": 0.3837,
+      "step": 19094
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 48.057837529229715,
+      "learning_rate": 1.7519590078696192e-09,
+      "loss": 0.3715,
+      "step": 19095
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 45.30064750086527,
+      "learning_rate": 1.7205352729732227e-09,
+      "loss": 0.4506,
+      "step": 19096
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 52.691899958708504,
+      "learning_rate": 1.6893958832964186e-09,
+      "loss": 0.4904,
+      "step": 19097
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 64.4619231462512,
+      "learning_rate": 1.6585408397240544e-09,
+      "loss": 0.4439,
+      "step": 19098
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 53.25754974018383,
+      "learning_rate": 1.6279701431343165e-09,
+      "loss": 0.4688,
+      "step": 19099
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 49.12985502523303,
+      "learning_rate": 1.5976837943953993e-09,
+      "loss": 0.4136,
+      "step": 19100
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 50.93640818917719,
+      "learning_rate": 1.5676817943699462e-09,
+      "loss": 0.4442,
+      "step": 19101
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 53.66897458342344,
+      "learning_rate": 1.5379641439106086e-09,
+      "loss": 0.4154,
+      "step": 19102
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 52.1218758980171,
+      "learning_rate": 1.5085308438633762e-09,
+      "loss": 0.397,
+      "step": 19103
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 46.130814488206326,
+      "learning_rate": 1.4793818950642469e-09,
+      "loss": 0.4465,
+      "step": 19104
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 46.58123208949898,
+      "learning_rate": 1.4505172983425576e-09,
+      "loss": 0.353,
+      "step": 19105
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 42.19336379943409,
+      "learning_rate": 1.4219370545187628e-09,
+      "loss": 0.3859,
+      "step": 19106
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 53.5119555341596,
+      "learning_rate": 1.393641164405546e-09,
+      "loss": 0.4494,
+      "step": 19107
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 47.20225201847246,
+      "learning_rate": 1.3656296288089288e-09,
+      "loss": 0.4587,
+      "step": 19108
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 52.73937995274198,
+      "learning_rate": 1.3379024485249414e-09,
+      "loss": 0.4071,
+      "step": 19109
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 43.70510377519622,
+      "learning_rate": 1.310459624341842e-09,
+      "loss": 0.3782,
+      "step": 19110
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 59.16797739735167,
+      "learning_rate": 1.2833011570401177e-09,
+      "loss": 0.4519,
+      "step": 19111
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 44.241966035488325,
+      "learning_rate": 1.2564270473924834e-09,
+      "loss": 0.4346,
+      "step": 19112
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.677425411809125,
+      "learning_rate": 1.2298372961627726e-09,
+      "loss": 0.4579,
+      "step": 19113
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 45.183954688109935,
+      "learning_rate": 1.203531904107047e-09,
+      "loss": 0.3793,
+      "step": 19114
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 54.74084729720446,
+      "learning_rate": 1.1775108719735973e-09,
+      "loss": 0.4723,
+      "step": 19115
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 41.089808883528896,
+      "learning_rate": 1.1517742005029419e-09,
+      "loss": 0.357,
+      "step": 19116
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 59.53965424815599,
+      "learning_rate": 1.1263218904267181e-09,
+      "loss": 0.4253,
+      "step": 19117
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 42.84265501485907,
+      "learning_rate": 1.101153942468791e-09,
+      "loss": 0.3951,
+      "step": 19118
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 50.46087778303172,
+      "learning_rate": 1.0762703573452548e-09,
+      "loss": 0.4771,
+      "step": 19119
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 54.341331665035256,
+      "learning_rate": 1.051671135762211e-09,
+      "loss": 0.4805,
+      "step": 19120
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.81743277560386,
+      "learning_rate": 1.0273562784213209e-09,
+      "loss": 0.4453,
+      "step": 19121
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 49.809748383342004,
+      "learning_rate": 1.0033257860131428e-09,
+      "loss": 0.4428,
+      "step": 19122
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 50.531932196320476,
+      "learning_rate": 9.795796592204643e-10,
+      "loss": 0.4053,
+      "step": 19123
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 55.42475613896154,
+      "learning_rate": 9.561178987205211e-10,
+      "loss": 0.4048,
+      "step": 19124
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 39.379528318111504,
+      "learning_rate": 9.32940505178337e-10,
+      "loss": 0.3651,
+      "step": 19125
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 48.781670250264575,
+      "learning_rate": 9.100474792544944e-10,
+      "loss": 0.4098,
+      "step": 19126
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 43.80077282384902,
+      "learning_rate": 8.874388215995844e-10,
+      "loss": 0.4859,
+      "step": 19127
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 37.369149812332275,
+      "learning_rate": 8.651145328564259e-10,
+      "loss": 0.3555,
+      "step": 19128
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 49.4181827561338,
+      "learning_rate": 8.430746136600665e-10,
+      "loss": 0.4229,
+      "step": 19129
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 48.28163643099015,
+      "learning_rate": 8.21319064636672e-10,
+      "loss": 0.3526,
+      "step": 19130
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 94.32383952090443,
+      "learning_rate": 7.998478864068571e-10,
+      "loss": 0.4421,
+      "step": 19131
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 43.153822303678034,
+      "learning_rate": 7.786610795790239e-10,
+      "loss": 0.418,
+      "step": 19132
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.89798947054217,
+      "learning_rate": 7.577586447571339e-10,
+      "loss": 0.4166,
+      "step": 19133
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 58.7570705460807,
+      "learning_rate": 7.371405825351563e-10,
+      "loss": 0.4755,
+      "step": 19134
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 51.837066316135775,
+      "learning_rate": 7.168068934992889e-10,
+      "loss": 0.3652,
+      "step": 19135
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 46.74675585460442,
+      "learning_rate": 6.967575782279579e-10,
+      "loss": 0.4209,
+      "step": 19136
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 40.54275529398943,
+      "learning_rate": 6.769926372918178e-10,
+      "loss": 0.3553,
+      "step": 19137
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 47.544540804287166,
+      "learning_rate": 6.575120712526417e-10,
+      "loss": 0.4048,
+      "step": 19138
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 46.3340848077352,
+      "learning_rate": 6.383158806644307e-10,
+      "loss": 0.3914,
+      "step": 19139
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 46.14281067546393,
+      "learning_rate": 6.194040660723044e-10,
+      "loss": 0.399,
+      "step": 19140
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 45.35615016131999,
+      "learning_rate": 6.007766280158312e-10,
+      "loss": 0.4122,
+      "step": 19141
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 44.32011551513147,
+      "learning_rate": 5.824335670234771e-10,
+      "loss": 0.439,
+      "step": 19142
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 47.44899713463383,
+      "learning_rate": 5.643748836170471e-10,
+      "loss": 0.4343,
+      "step": 19143
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 54.06145093484341,
+      "learning_rate": 5.466005783105743e-10,
+      "loss": 0.4754,
+      "step": 19144
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 59.226349882955155,
+      "learning_rate": 5.291106516092103e-10,
+      "loss": 0.4488,
+      "step": 19145
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 40.57939257017206,
+      "learning_rate": 5.119051040092249e-10,
+      "loss": 0.4547,
+      "step": 19146
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 42.09304274295492,
+      "learning_rate": 4.949839360024467e-10,
+      "loss": 0.3747,
+      "step": 19147
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 40.6488802752115,
+      "learning_rate": 4.78347148068492e-10,
+      "loss": 0.393,
+      "step": 19148
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 50.47245741415969,
+      "learning_rate": 4.6199474068031603e-10,
+      "loss": 0.4575,
+      "step": 19149
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 55.97494065356974,
+      "learning_rate": 4.459267143042123e-10,
+      "loss": 0.4193,
+      "step": 19150
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.576127783528534,
+      "learning_rate": 4.301430693964825e-10,
+      "loss": 0.3749,
+      "step": 19151
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 62.14742018791926,
+      "learning_rate": 4.1464380640565683e-10,
+      "loss": 0.4639,
+      "step": 19152
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 52.80853280058918,
+      "learning_rate": 3.994289257724937e-10,
+      "loss": 0.4424,
+      "step": 19153
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 49.533366719453284,
+      "learning_rate": 3.8449842792998016e-10,
+      "loss": 0.448,
+      "step": 19154
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 55.803886757095974,
+      "learning_rate": 3.698523133022214e-10,
+      "loss": 0.4804,
+      "step": 19155
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 52.0366044322281,
+      "learning_rate": 3.5549058230777143e-10,
+      "loss": 0.4766,
+      "step": 19156
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 50.767543518127404,
+      "learning_rate": 3.4141323535186175e-10,
+      "loss": 0.4782,
+      "step": 19157
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 59.45279706947226,
+      "learning_rate": 3.276202728375033e-10,
+      "loss": 0.5087,
+      "step": 19158
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.64615076606931,
+      "learning_rate": 3.1411169515549457e-10,
+      "loss": 0.426,
+      "step": 19159
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 56.26929042649432,
+      "learning_rate": 3.0088750269108293e-10,
+      "loss": 0.4252,
+      "step": 19160
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 55.945346780592544,
+      "learning_rate": 2.8794769581952376e-10,
+      "loss": 0.4959,
+      "step": 19161
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.411694172735764,
+      "learning_rate": 2.7529227490941115e-10,
+      "loss": 0.3779,
+      "step": 19162
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 45.54271915903007,
+      "learning_rate": 2.629212403193471e-10,
+      "loss": 0.4085,
+      "step": 19163
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 55.49259322806382,
+      "learning_rate": 2.508345924023825e-10,
+      "loss": 0.4264,
+      "step": 19164
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 46.22987523795854,
+      "learning_rate": 2.390323315026866e-10,
+      "loss": 0.4073,
+      "step": 19165
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 40.21989976355432,
+      "learning_rate": 2.2751445795443638e-10,
+      "loss": 0.4159,
+      "step": 19166
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 53.06366112003673,
+      "learning_rate": 2.162809720873682e-10,
+      "loss": 0.4918,
+      "step": 19167
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 43.87554235204676,
+      "learning_rate": 2.0533187421789557e-10,
+      "loss": 0.397,
+      "step": 19168
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 58.54433343408665,
+      "learning_rate": 1.9466716466021162e-10,
+      "loss": 0.4243,
+      "step": 19169
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 46.08897341014682,
+      "learning_rate": 1.84286843716297e-10,
+      "loss": 0.3473,
+      "step": 19170
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 44.13826377832251,
+      "learning_rate": 1.7419091168036085e-10,
+      "loss": 0.3547,
+      "step": 19171
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 58.35351036253826,
+      "learning_rate": 1.6437936884217132e-10,
+      "loss": 0.4651,
+      "step": 19172
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 44.47212005412618,
+      "learning_rate": 1.5485221547817396e-10,
+      "loss": 0.3672,
+      "step": 19173
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 39.526634103628965,
+      "learning_rate": 1.4560945186148366e-10,
+      "loss": 0.4008,
+      "step": 19174
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 68.00852189239076,
+      "learning_rate": 1.3665107825300284e-10,
+      "loss": 0.4729,
+      "step": 19175
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 45.47957659944511,
+      "learning_rate": 1.2797709490919296e-10,
+      "loss": 0.4253,
+      "step": 19176
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 60.68638658533451,
+      "learning_rate": 1.1958750207541337e-10,
+      "loss": 0.4342,
+      "step": 19177
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 55.43611292947708,
+      "learning_rate": 1.114822999914722e-10,
+      "loss": 0.45,
+      "step": 19178
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 38.85235978732733,
+      "learning_rate": 1.0366148888607541e-10,
+      "loss": 0.3516,
+      "step": 19179
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 52.982509426075964,
+      "learning_rate": 9.612506898348805e-11,
+      "loss": 0.4734,
+      "step": 19180
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 47.51713461768703,
+      "learning_rate": 8.887304049798317e-11,
+      "loss": 0.4408,
+      "step": 19181
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 47.862681395136406,
+      "learning_rate": 8.190540363384181e-11,
+      "loss": 0.3964,
+      "step": 19182
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 45.95962818750659,
+      "learning_rate": 7.522215859090409e-11,
+      "loss": 0.3889,
+      "step": 19183
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 52.26575384879659,
+      "learning_rate": 6.882330556012839e-11,
+      "loss": 0.4394,
+      "step": 19184
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 53.14962701729415,
+      "learning_rate": 6.270884472026062e-11,
+      "loss": 0.4688,
+      "step": 19185
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 50.5263695043798,
+      "learning_rate": 5.687877624893645e-11,
+      "loss": 0.4183,
+      "step": 19186
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 47.997874689734395,
+      "learning_rate": 5.133310030935867e-11,
+      "loss": 0.4563,
+      "step": 19187
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 57.25991836987059,
+      "learning_rate": 4.607181705917896e-11,
+      "loss": 0.4165,
+      "step": 19188
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 49.10203174497339,
+      "learning_rate": 4.1094926650497855e-11,
+      "loss": 0.3606,
+      "step": 19189
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 44.595126670112606,
+      "learning_rate": 3.640242922209325e-11,
+      "loss": 0.3545,
+      "step": 19190
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 44.315975886844754,
+      "learning_rate": 3.199432490941234e-11,
+      "loss": 0.3364,
+      "step": 19191
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 57.789346708429406,
+      "learning_rate": 2.787061383680012e-11,
+      "loss": 0.4315,
+      "step": 19192
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 50.34506151893245,
+      "learning_rate": 2.4031296120829993e-11,
+      "loss": 0.4576,
+      "step": 19193
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 42.301974876557225,
+      "learning_rate": 2.047637187252427e-11,
+      "loss": 0.4154,
+      "step": 19194
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 58.305336879204845,
+      "learning_rate": 1.720584119180302e-11,
+      "loss": 0.3924,
+      "step": 19195
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 57.481470273645115,
+      "learning_rate": 1.4219704171924976e-11,
+      "loss": 0.4926,
+      "step": 19196
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 36.78134527331826,
+      "learning_rate": 1.1517960898377312e-11,
+      "loss": 0.3319,
+      "step": 19197
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 76.83274897217314,
+      "learning_rate": 9.100611446655194e-12,
+      "loss": 0.3956,
+      "step": 19198
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 53.39265281302791,
+      "learning_rate": 6.967655886702674e-12,
+      "loss": 0.4155,
+      "step": 19199
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 46.40695416979095,
+      "learning_rate": 5.119094278471792e-12,
+      "loss": 0.4526,
+      "step": 19200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 57.298556550470686,
+      "learning_rate": 3.5549266752532564e-12,
+      "loss": 0.4651,
+      "step": 19201
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 50.870943465410335,
+      "learning_rate": 2.275153120345763e-12,
+      "loss": 0.4442,
+      "step": 19202
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 59.43246324700678,
+      "learning_rate": 1.2797736514968962e-12,
+      "loss": 0.3961,
+      "step": 19203
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 51.09261099102838,
+      "learning_rate": 5.687882964622305e-13,
+      "loss": 0.4488,
+      "step": 19204
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 47.8347989809253,
+      "learning_rate": 1.4219707522578064e-13,
+      "loss": 0.443,
+      "step": 19205
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 61.58493904098984,
+      "learning_rate": 0.0,
+      "loss": 0.4655,
+      "step": 19206
+    },
+    {
+      "epoch": 3.0,
+      "step": 19206,
+      "total_flos": 2.4590557150021943e+19,
+      "train_loss": 0.6681029205302156,
+      "train_runtime": 34223.736,
+      "train_samples_per_second": 71.832,
+      "train_steps_per_second": 0.561
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 19206,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50000,
+  "total_flos": 2.4590557150021943e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}