diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,7034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.25130363762015456,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00025130363762015457,
+      "grad_norm": 2.0144803524017334,
+      "learning_rate": 0.0,
+      "loss": 2.7742,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005026072752403091,
+      "grad_norm": 1.5290026664733887,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.9485,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007539109128604636,
+      "grad_norm": 1.7576098442077637,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 3.2699,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010052145504806183,
+      "grad_norm": 1.4803149700164795,
+      "learning_rate": 6e-06,
+      "loss": 2.6531,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012565181881007727,
+      "grad_norm": 1.2919175624847412,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.7198,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015078218257209273,
+      "grad_norm": 1.0887187719345093,
+      "learning_rate": 1e-05,
+      "loss": 2.9058,
+      "step": 6
+    },
+    {
+      "epoch": 0.001759125463341082,
+      "grad_norm": 1.176196575164795,
+      "learning_rate": 1.2e-05,
+      "loss": 2.769,
+      "step": 7
+    },
+    {
+      "epoch": 0.0020104291009612365,
+      "grad_norm": 1.4506360292434692,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 2.8493,
+      "step": 8
+    },
+    {
+      "epoch": 0.002261732738581391,
+      "grad_norm": 0.8232998251914978,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.7691,
+      "step": 9
+    },
+    {
+      "epoch": 0.0025130363762015454,
+      "grad_norm": 0.8385952711105347,
+      "learning_rate": 1.8e-05,
+      "loss": 2.7073,
+      "step": 10
+    },
+    {
+      "epoch": 0.0027643400138217,
+      "grad_norm": 0.594434916973114,
+      "learning_rate": 2e-05,
+      "loss": 2.1646,
+      "step": 11
+    },
+    {
+      "epoch": 0.0030156436514418546,
+      "grad_norm": 0.8067689538002014,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 2.5741,
+      "step": 12
+    },
+    {
+      "epoch": 0.0032669472890620092,
+      "grad_norm": 0.8038071393966675,
+      "learning_rate": 2.4e-05,
+      "loss": 2.6733,
+      "step": 13
+    },
+    {
+      "epoch": 0.003518250926682164,
+      "grad_norm": 0.48843103647232056,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.598,
+      "step": 14
+    },
+    {
+      "epoch": 0.0037695545643023185,
+      "grad_norm": 0.8283182978630066,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 2.8211,
+      "step": 15
+    },
+    {
+      "epoch": 0.004020858201922473,
+      "grad_norm": 0.4953489601612091,
+      "learning_rate": 3e-05,
+      "loss": 2.3855,
+      "step": 16
+    },
+    {
+      "epoch": 0.004272161839542627,
+      "grad_norm": 0.4421069025993347,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 2.6665,
+      "step": 17
+    },
+    {
+      "epoch": 0.004523465477162782,
+      "grad_norm": 0.6795067191123962,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 2.5962,
+      "step": 18
+    },
+    {
+      "epoch": 0.0047747691147829365,
+      "grad_norm": 0.5411877036094666,
+      "learning_rate": 3.6e-05,
+      "loss": 2.4962,
+      "step": 19
+    },
+    {
+      "epoch": 0.005026072752403091,
+      "grad_norm": 0.5864161252975464,
+      "learning_rate": 3.8e-05,
+      "loss": 2.4219,
+      "step": 20
+    },
+    {
+      "epoch": 0.005277376390023246,
+      "grad_norm": 0.5454627871513367,
+      "learning_rate": 4e-05,
+      "loss": 2.8872,
+      "step": 21
+    },
+    {
+      "epoch": 0.0055286800276434,
+      "grad_norm": 0.4071284532546997,
+      "learning_rate": 4.2e-05,
+      "loss": 2.4573,
+      "step": 22
+    },
+    {
+      "epoch": 0.005779983665263555,
+      "grad_norm": 0.2068735957145691,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.8764,
+      "step": 23
+    },
+    {
+      "epoch": 0.006031287302883709,
+      "grad_norm": 0.5132932662963867,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 2.4306,
+      "step": 24
+    },
+    {
+      "epoch": 0.006282590940503863,
+      "grad_norm": 0.3241199851036072,
+      "learning_rate": 4.8e-05,
+      "loss": 2.1521,
+      "step": 25
+    },
+    {
+      "epoch": 0.0065338945781240184,
+      "grad_norm": 0.4157683551311493,
+      "learning_rate": 5e-05,
+      "loss": 2.7139,
+      "step": 26
+    },
+    {
+      "epoch": 0.006785198215744173,
+      "grad_norm": 0.5854943990707397,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 2.2383,
+      "step": 27
+    },
+    {
+      "epoch": 0.007036501853364328,
+      "grad_norm": 0.5483741760253906,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 2.4569,
+      "step": 28
+    },
+    {
+      "epoch": 0.007287805490984482,
+      "grad_norm": 0.41493722796440125,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 2.9473,
+      "step": 29
+    },
+    {
+      "epoch": 0.007539109128604637,
+      "grad_norm": 0.3696433901786804,
+      "learning_rate": 5.8e-05,
+      "loss": 2.2744,
+      "step": 30
+    },
+    {
+      "epoch": 0.007790412766224791,
+      "grad_norm": 0.18437762558460236,
+      "learning_rate": 6e-05,
+      "loss": 1.0493,
+      "step": 31
+    },
+    {
+      "epoch": 0.008041716403844946,
+      "grad_norm": 0.46951478719711304,
+      "learning_rate": 6.2e-05,
+      "loss": 2.4904,
+      "step": 32
+    },
+    {
+      "epoch": 0.0082930200414651,
+      "grad_norm": 0.32901376485824585,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 2.0627,
+      "step": 33
+    },
+    {
+      "epoch": 0.008544323679085255,
+      "grad_norm": 0.668229877948761,
+      "learning_rate": 6.6e-05,
+      "loss": 2.3694,
+      "step": 34
+    },
+    {
+      "epoch": 0.008795627316705409,
+      "grad_norm": 0.9680635333061218,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 2.3719,
+      "step": 35
+    },
+    {
+      "epoch": 0.009046930954325565,
+      "grad_norm": 0.6750502586364746,
+      "learning_rate": 7e-05,
+      "loss": 2.3538,
+      "step": 36
+    },
+    {
+      "epoch": 0.009298234591945719,
+      "grad_norm": 0.1759599894285202,
+      "learning_rate": 7.2e-05,
+      "loss": 1.0368,
+      "step": 37
+    },
+    {
+      "epoch": 0.009549538229565873,
+      "grad_norm": 0.5437096953392029,
+      "learning_rate": 7.4e-05,
+      "loss": 2.7952,
+      "step": 38
+    },
+    {
+      "epoch": 0.009800841867186027,
+      "grad_norm": 0.31723931431770325,
+      "learning_rate": 7.6e-05,
+      "loss": 2.4566,
+      "step": 39
+    },
+    {
+      "epoch": 0.010052145504806181,
+      "grad_norm": 0.34227266907691956,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 2.281,
+      "step": 40
+    },
+    {
+      "epoch": 0.010303449142426337,
+      "grad_norm": 0.3893303871154785,
+      "learning_rate": 8e-05,
+      "loss": 2.7128,
+      "step": 41
+    },
+    {
+      "epoch": 0.010554752780046492,
+      "grad_norm": 0.32380804419517517,
+      "learning_rate": 8.2e-05,
+      "loss": 2.7928,
+      "step": 42
+    },
+    {
+      "epoch": 0.010806056417666646,
+      "grad_norm": 0.29380202293395996,
+      "learning_rate": 8.4e-05,
+      "loss": 2.305,
+      "step": 43
+    },
+    {
+      "epoch": 0.0110573600552868,
+      "grad_norm": 0.4324714243412018,
+      "learning_rate": 8.6e-05,
+      "loss": 2.4824,
+      "step": 44
+    },
+    {
+      "epoch": 0.011308663692906954,
+      "grad_norm": 0.4421226978302002,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 2.1697,
+      "step": 45
+    },
+    {
+      "epoch": 0.01155996733052711,
+      "grad_norm": 0.5161215662956238,
+      "learning_rate": 9e-05,
+      "loss": 2.4065,
+      "step": 46
+    },
+    {
+      "epoch": 0.011811270968147264,
+      "grad_norm": 0.3017684817314148,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 2.4247,
+      "step": 47
+    },
+    {
+      "epoch": 0.012062574605767418,
+      "grad_norm": 0.5015223622322083,
+      "learning_rate": 9.4e-05,
+      "loss": 2.3724,
+      "step": 48
+    },
+    {
+      "epoch": 0.012313878243387573,
+      "grad_norm": 0.3141859471797943,
+      "learning_rate": 9.6e-05,
+      "loss": 1.8871,
+      "step": 49
+    },
+    {
+      "epoch": 0.012565181881007727,
+      "grad_norm": 0.1826818883419037,
+      "learning_rate": 9.8e-05,
+      "loss": 1.9633,
+      "step": 50
+    },
+    {
+      "epoch": 0.012816485518627883,
+      "grad_norm": 0.37197422981262207,
+      "learning_rate": 0.0001,
+      "loss": 2.4161,
+      "step": 51
+    },
+    {
+      "epoch": 0.013067789156248037,
+      "grad_norm": 0.3732236623764038,
+      "learning_rate": 9.99999840163606e-05,
+      "loss": 2.2943,
+      "step": 52
+    },
+    {
+      "epoch": 0.013319092793868191,
+      "grad_norm": 0.5340951681137085,
+      "learning_rate": 9.99999360654526e-05,
+      "loss": 2.2039,
+      "step": 53
+    },
+    {
+      "epoch": 0.013570396431488345,
+      "grad_norm": 0.394661545753479,
+      "learning_rate": 9.999985614730664e-05,
+      "loss": 2.3037,
+      "step": 54
+    },
+    {
+      "epoch": 0.013821700069108501,
+      "grad_norm": 0.3042058050632477,
+      "learning_rate": 9.999974426197384e-05,
+      "loss": 2.5117,
+      "step": 55
+    },
+    {
+      "epoch": 0.014073003706728655,
+      "grad_norm": 0.362106591463089,
+      "learning_rate": 9.999960040952574e-05,
+      "loss": 2.6846,
+      "step": 56
+    },
+    {
+      "epoch": 0.01432430734434881,
+      "grad_norm": 0.728489875793457,
+      "learning_rate": 9.99994245900543e-05,
+      "loss": 2.4936,
+      "step": 57
+    },
+    {
+      "epoch": 0.014575610981968964,
+      "grad_norm": 0.2623450756072998,
+      "learning_rate": 9.999921680367191e-05,
+      "loss": 2.6129,
+      "step": 58
+    },
+    {
+      "epoch": 0.014826914619589118,
+      "grad_norm": 0.24758170545101166,
+      "learning_rate": 9.999897705051145e-05,
+      "loss": 2.1818,
+      "step": 59
+    },
+    {
+      "epoch": 0.015078218257209274,
+      "grad_norm": 0.3045499324798584,
+      "learning_rate": 9.99987053307262e-05,
+      "loss": 2.3918,
+      "step": 60
+    },
+    {
+      "epoch": 0.015329521894829428,
+      "grad_norm": 0.3922071158885956,
+      "learning_rate": 9.999840164448984e-05,
+      "loss": 2.2996,
+      "step": 61
+    },
+    {
+      "epoch": 0.015580825532449582,
+      "grad_norm": 0.20261235535144806,
+      "learning_rate": 9.999806599199659e-05,
+      "loss": 1.9872,
+      "step": 62
+    },
+    {
+      "epoch": 0.015832129170069736,
+      "grad_norm": 0.3178957402706146,
+      "learning_rate": 9.999769837346103e-05,
+      "loss": 2.4412,
+      "step": 63
+    },
+    {
+      "epoch": 0.016083432807689892,
+      "grad_norm": 0.304047167301178,
+      "learning_rate": 9.999729878911816e-05,
+      "loss": 2.705,
+      "step": 64
+    },
+    {
+      "epoch": 0.016334736445310045,
+      "grad_norm": 0.5113909840583801,
+      "learning_rate": 9.99968672392235e-05,
+      "loss": 2.533,
+      "step": 65
+    },
+    {
+      "epoch": 0.0165860400829302,
+      "grad_norm": 0.29634010791778564,
+      "learning_rate": 9.999640372405295e-05,
+      "loss": 2.4283,
+      "step": 66
+    },
+    {
+      "epoch": 0.016837343720550357,
+      "grad_norm": 0.3125896453857422,
+      "learning_rate": 9.999590824390281e-05,
+      "loss": 2.4421,
+      "step": 67
+    },
+    {
+      "epoch": 0.01708864735817051,
+      "grad_norm": 0.38444283604621887,
+      "learning_rate": 9.999538079908993e-05,
+      "loss": 2.1928,
+      "step": 68
+    },
+    {
+      "epoch": 0.017339950995790665,
+      "grad_norm": 0.3269999325275421,
+      "learning_rate": 9.999482138995149e-05,
+      "loss": 2.6259,
+      "step": 69
+    },
+    {
+      "epoch": 0.017591254633410817,
+      "grad_norm": 0.30345460772514343,
+      "learning_rate": 9.999423001684513e-05,
+      "loss": 2.6393,
+      "step": 70
+    },
+    {
+      "epoch": 0.017842558271030973,
+      "grad_norm": 1.1956677436828613,
+      "learning_rate": 9.9993606680149e-05,
+      "loss": 2.9455,
+      "step": 71
+    },
+    {
+      "epoch": 0.01809386190865113,
+      "grad_norm": 0.4432305097579956,
+      "learning_rate": 9.999295138026157e-05,
+      "loss": 2.3158,
+      "step": 72
+    },
+    {
+      "epoch": 0.018345165546271282,
+      "grad_norm": 0.24386799335479736,
+      "learning_rate": 9.999226411760185e-05,
+      "loss": 2.6777,
+      "step": 73
+    },
+    {
+      "epoch": 0.018596469183891438,
+      "grad_norm": 0.34685081243515015,
+      "learning_rate": 9.99915448926092e-05,
+      "loss": 2.4423,
+      "step": 74
+    },
+    {
+      "epoch": 0.01884777282151159,
+      "grad_norm": 0.5708550214767456,
+      "learning_rate": 9.999079370574347e-05,
+      "loss": 2.2399,
+      "step": 75
+    },
+    {
+      "epoch": 0.019099076459131746,
+      "grad_norm": 0.2775181829929352,
+      "learning_rate": 9.999001055748492e-05,
+      "loss": 1.4073,
+      "step": 76
+    },
+    {
+      "epoch": 0.019350380096751902,
+      "grad_norm": 0.44185924530029297,
+      "learning_rate": 9.998919544833427e-05,
+      "loss": 2.2447,
+      "step": 77
+    },
+    {
+      "epoch": 0.019601683734372054,
+      "grad_norm": 0.35135698318481445,
+      "learning_rate": 9.998834837881263e-05,
+      "loss": 2.4701,
+      "step": 78
+    },
+    {
+      "epoch": 0.01985298737199221,
+      "grad_norm": 0.2450733631849289,
+      "learning_rate": 9.998746934946159e-05,
+      "loss": 2.1066,
+      "step": 79
+    },
+    {
+      "epoch": 0.020104291009612363,
+      "grad_norm": 0.3851965367794037,
+      "learning_rate": 9.998655836084316e-05,
+      "loss": 2.2371,
+      "step": 80
+    },
+    {
+      "epoch": 0.02035559464723252,
+      "grad_norm": 0.2562990188598633,
+      "learning_rate": 9.998561541353976e-05,
+      "loss": 2.5131,
+      "step": 81
+    },
+    {
+      "epoch": 0.020606898284852675,
+      "grad_norm": 0.35332947969436646,
+      "learning_rate": 9.998464050815426e-05,
+      "loss": 2.4474,
+      "step": 82
+    },
+    {
+      "epoch": 0.020858201922472827,
+      "grad_norm": 0.3566109836101532,
+      "learning_rate": 9.998363364530998e-05,
+      "loss": 2.5007,
+      "step": 83
+    },
+    {
+      "epoch": 0.021109505560092983,
+      "grad_norm": 0.27204859256744385,
+      "learning_rate": 9.998259482565063e-05,
+      "loss": 2.221,
+      "step": 84
+    },
+    {
+      "epoch": 0.021360809197713135,
+      "grad_norm": 0.1774623841047287,
+      "learning_rate": 9.998152404984036e-05,
+      "loss": 1.5693,
+      "step": 85
+    },
+    {
+      "epoch": 0.02161211283533329,
+      "grad_norm": 0.31386587023735046,
+      "learning_rate": 9.998042131856382e-05,
+      "loss": 2.1838,
+      "step": 86
+    },
+    {
+      "epoch": 0.021863416472953447,
+      "grad_norm": 0.4387069344520569,
+      "learning_rate": 9.997928663252601e-05,
+      "loss": 2.2176,
+      "step": 87
+    },
+    {
+      "epoch": 0.0221147201105736,
+      "grad_norm": 0.5427992343902588,
+      "learning_rate": 9.997811999245236e-05,
+      "loss": 2.2632,
+      "step": 88
+    },
+    {
+      "epoch": 0.022366023748193756,
+      "grad_norm": 0.14437389373779297,
+      "learning_rate": 9.997692139908879e-05,
+      "loss": 1.0033,
+      "step": 89
+    },
+    {
+      "epoch": 0.022617327385813908,
+      "grad_norm": 0.18704228103160858,
+      "learning_rate": 9.99756908532016e-05,
+      "loss": 1.1516,
+      "step": 90
+    },
+    {
+      "epoch": 0.022868631023434064,
+      "grad_norm": 0.23126636445522308,
+      "learning_rate": 9.997442835557753e-05,
+      "loss": 1.9892,
+      "step": 91
+    },
+    {
+      "epoch": 0.02311993466105422,
+      "grad_norm": 0.2967139780521393,
+      "learning_rate": 9.997313390702377e-05,
+      "loss": 2.0742,
+      "step": 92
+    },
+    {
+      "epoch": 0.023371238298674372,
+      "grad_norm": 0.25964412093162537,
+      "learning_rate": 9.997180750836792e-05,
+      "loss": 2.4486,
+      "step": 93
+    },
+    {
+      "epoch": 0.02362254193629453,
+      "grad_norm": 0.21999791264533997,
+      "learning_rate": 9.997044916045799e-05,
+      "loss": 1.0311,
+      "step": 94
+    },
+    {
+      "epoch": 0.02387384557391468,
+      "grad_norm": 0.22438450157642365,
+      "learning_rate": 9.996905886416244e-05,
+      "loss": 1.2788,
+      "step": 95
+    },
+    {
+      "epoch": 0.024125149211534837,
+      "grad_norm": 0.3627947270870209,
+      "learning_rate": 9.996763662037014e-05,
+      "loss": 2.2382,
+      "step": 96
+    },
+    {
+      "epoch": 0.024376452849154993,
+      "grad_norm": 0.2582647502422333,
+      "learning_rate": 9.996618242999042e-05,
+      "loss": 2.2999,
+      "step": 97
+    },
+    {
+      "epoch": 0.024627756486775145,
+      "grad_norm": 0.34755995869636536,
+      "learning_rate": 9.9964696293953e-05,
+      "loss": 2.2329,
+      "step": 98
+    },
+    {
+      "epoch": 0.0248790601243953,
+      "grad_norm": 0.15312296152114868,
+      "learning_rate": 9.996317821320802e-05,
+      "loss": 0.6043,
+      "step": 99
+    },
+    {
+      "epoch": 0.025130363762015454,
+      "grad_norm": 0.22593067586421967,
+      "learning_rate": 9.996162818872607e-05,
+      "loss": 2.3993,
+      "step": 100
+    },
+    {
+      "epoch": 0.02538166739963561,
+      "grad_norm": 0.3924887776374817,
+      "learning_rate": 9.996004622149814e-05,
+      "loss": 2.3133,
+      "step": 101
+    },
+    {
+      "epoch": 0.025632971037255765,
+      "grad_norm": 0.5138020515441895,
+      "learning_rate": 9.995843231253569e-05,
+      "loss": 2.4782,
+      "step": 102
+    },
+    {
+      "epoch": 0.025884274674875918,
+      "grad_norm": 0.42053958773612976,
+      "learning_rate": 9.995678646287053e-05,
+      "loss": 2.0569,
+      "step": 103
+    },
+    {
+      "epoch": 0.026135578312496074,
+      "grad_norm": 0.2818872034549713,
+      "learning_rate": 9.995510867355494e-05,
+      "loss": 2.7051,
+      "step": 104
+    },
+    {
+      "epoch": 0.02638688195011623,
+      "grad_norm": 0.4185803532600403,
+      "learning_rate": 9.995339894566158e-05,
+      "loss": 2.2182,
+      "step": 105
+    },
+    {
+      "epoch": 0.026638185587736382,
+      "grad_norm": 0.3512636721134186,
+      "learning_rate": 9.995165728028359e-05,
+      "loss": 2.5701,
+      "step": 106
+    },
+    {
+      "epoch": 0.026889489225356538,
+      "grad_norm": 0.3916986584663391,
+      "learning_rate": 9.994988367853451e-05,
+      "loss": 2.034,
+      "step": 107
+    },
+    {
+      "epoch": 0.02714079286297669,
+      "grad_norm": 0.3517094850540161,
+      "learning_rate": 9.994807814154824e-05,
+      "loss": 2.0577,
+      "step": 108
+    },
+    {
+      "epoch": 0.027392096500596846,
+      "grad_norm": 0.3244116008281708,
+      "learning_rate": 9.994624067047917e-05,
+      "loss": 2.1661,
+      "step": 109
+    },
+    {
+      "epoch": 0.027643400138217002,
+      "grad_norm": 0.42575743794441223,
+      "learning_rate": 9.994437126650207e-05,
+      "loss": 2.5129,
+      "step": 110
+    },
+    {
+      "epoch": 0.027894703775837155,
+      "grad_norm": 0.1691725254058838,
+      "learning_rate": 9.994246993081213e-05,
+      "loss": 1.5939,
+      "step": 111
+    },
+    {
+      "epoch": 0.02814600741345731,
+      "grad_norm": 0.33062514662742615,
+      "learning_rate": 9.994053666462498e-05,
+      "loss": 2.222,
+      "step": 112
+    },
+    {
+      "epoch": 0.028397311051077463,
+      "grad_norm": 0.3275945484638214,
+      "learning_rate": 9.993857146917662e-05,
+      "loss": 2.3336,
+      "step": 113
+    },
+    {
+      "epoch": 0.02864861468869762,
+      "grad_norm": 0.3942461907863617,
+      "learning_rate": 9.993657434572353e-05,
+      "loss": 2.3289,
+      "step": 114
+    },
+    {
+      "epoch": 0.028899918326317775,
+      "grad_norm": 0.3180122971534729,
+      "learning_rate": 9.993454529554251e-05,
+      "loss": 2.4353,
+      "step": 115
+    },
+    {
+      "epoch": 0.029151221963937927,
+      "grad_norm": 0.7072325944900513,
+      "learning_rate": 9.993248431993086e-05,
+      "loss": 1.9157,
+      "step": 116
+    },
+    {
+      "epoch": 0.029402525601558083,
+      "grad_norm": 0.32835853099823,
+      "learning_rate": 9.993039142020622e-05,
+      "loss": 2.8819,
+      "step": 117
+    },
+    {
+      "epoch": 0.029653829239178236,
+      "grad_norm": 0.2762772738933563,
+      "learning_rate": 9.992826659770672e-05,
+      "loss": 2.504,
+      "step": 118
+    },
+    {
+      "epoch": 0.029905132876798392,
+      "grad_norm": 0.2669197618961334,
+      "learning_rate": 9.992610985379082e-05,
+      "loss": 2.3047,
+      "step": 119
+    },
+    {
+      "epoch": 0.030156436514418548,
+      "grad_norm": 0.47167858481407166,
+      "learning_rate": 9.992392118983746e-05,
+      "loss": 2.2432,
+      "step": 120
+    },
+    {
+      "epoch": 0.0304077401520387,
+      "grad_norm": 0.18089920282363892,
+      "learning_rate": 9.992170060724593e-05,
+      "loss": 1.5848,
+      "step": 121
+    },
+    {
+      "epoch": 0.030659043789658856,
+      "grad_norm": 0.33109933137893677,
+      "learning_rate": 9.991944810743597e-05,
+      "loss": 2.3503,
+      "step": 122
+    },
+    {
+      "epoch": 0.03091034742727901,
+      "grad_norm": 0.24957267940044403,
+      "learning_rate": 9.991716369184765e-05,
+      "loss": 2.2894,
+      "step": 123
+    },
+    {
+      "epoch": 0.031161651064899164,
+      "grad_norm": 0.3461008667945862,
+      "learning_rate": 9.991484736194157e-05,
+      "loss": 2.2987,
+      "step": 124
+    },
+    {
+      "epoch": 0.03141295470251932,
+      "grad_norm": 0.2855774760246277,
+      "learning_rate": 9.991249911919862e-05,
+      "loss": 2.5193,
+      "step": 125
+    },
+    {
+      "epoch": 0.03166425834013947,
+      "grad_norm": 0.30880868434906006,
+      "learning_rate": 9.991011896512016e-05,
+      "loss": 2.5596,
+      "step": 126
+    },
+    {
+      "epoch": 0.031915561977759625,
+      "grad_norm": 0.4237726032733917,
+      "learning_rate": 9.990770690122793e-05,
+      "loss": 2.1401,
+      "step": 127
+    },
+    {
+      "epoch": 0.032166865615379785,
+      "grad_norm": 0.2743227183818817,
+      "learning_rate": 9.990526292906405e-05,
+      "loss": 1.858,
+      "step": 128
+    },
+    {
+      "epoch": 0.03241816925299994,
+      "grad_norm": 0.28934118151664734,
+      "learning_rate": 9.99027870501911e-05,
+      "loss": 1.9751,
+      "step": 129
+    },
+    {
+      "epoch": 0.03266947289062009,
+      "grad_norm": 0.4869619905948639,
+      "learning_rate": 9.990027926619197e-05,
+      "loss": 2.7332,
+      "step": 130
+    },
+    {
+      "epoch": 0.03292077652824025,
+      "grad_norm": 0.2246621996164322,
+      "learning_rate": 9.989773957867006e-05,
+      "loss": 2.2951,
+      "step": 131
+    },
+    {
+      "epoch": 0.0331720801658604,
+      "grad_norm": 0.46311891078948975,
+      "learning_rate": 9.989516798924908e-05,
+      "loss": 2.0357,
+      "step": 132
+    },
+    {
+      "epoch": 0.033423383803480554,
+      "grad_norm": 0.45459482073783875,
+      "learning_rate": 9.989256449957316e-05,
+      "loss": 2.2661,
+      "step": 133
+    },
+    {
+      "epoch": 0.03367468744110071,
+      "grad_norm": 0.4114730656147003,
+      "learning_rate": 9.988992911130683e-05,
+      "loss": 2.096,
+      "step": 134
+    },
+    {
+      "epoch": 0.033925991078720866,
+      "grad_norm": 0.468787282705307,
+      "learning_rate": 9.988726182613502e-05,
+      "loss": 2.2083,
+      "step": 135
+    },
+    {
+      "epoch": 0.03417729471634102,
+      "grad_norm": 0.30824795365333557,
+      "learning_rate": 9.988456264576305e-05,
+      "loss": 2.4525,
+      "step": 136
+    },
+    {
+      "epoch": 0.03442859835396117,
+      "grad_norm": 0.22959741950035095,
+      "learning_rate": 9.988183157191662e-05,
+      "loss": 2.4861,
+      "step": 137
+    },
+    {
+      "epoch": 0.03467990199158133,
+      "grad_norm": 0.37763601541519165,
+      "learning_rate": 9.987906860634184e-05,
+      "loss": 2.4736,
+      "step": 138
+    },
+    {
+      "epoch": 0.03493120562920148,
+      "grad_norm": 0.3474363386631012,
+      "learning_rate": 9.987627375080519e-05,
+      "loss": 2.4363,
+      "step": 139
+    },
+    {
+      "epoch": 0.035182509266821635,
+      "grad_norm": 0.3558574616909027,
+      "learning_rate": 9.987344700709356e-05,
+      "loss": 2.0973,
+      "step": 140
+    },
+    {
+      "epoch": 0.035433812904441794,
+      "grad_norm": 0.19336606562137604,
+      "learning_rate": 9.98705883770142e-05,
+      "loss": 1.3915,
+      "step": 141
+    },
+    {
+      "epoch": 0.03568511654206195,
+      "grad_norm": 0.4760534465312958,
+      "learning_rate": 9.986769786239477e-05,
+      "loss": 2.0163,
+      "step": 142
+    },
+    {
+      "epoch": 0.0359364201796821,
+      "grad_norm": 0.2839144170284271,
+      "learning_rate": 9.98647754650833e-05,
+      "loss": 1.7691,
+      "step": 143
+    },
+    {
+      "epoch": 0.03618772381730226,
+      "grad_norm": 0.3936741352081299,
+      "learning_rate": 9.986182118694825e-05,
+      "loss": 2.1423,
+      "step": 144
+    },
+    {
+      "epoch": 0.03643902745492241,
+      "grad_norm": 0.35313600301742554,
+      "learning_rate": 9.985883502987838e-05,
+      "loss": 2.3156,
+      "step": 145
+    },
+    {
+      "epoch": 0.036690331092542564,
+      "grad_norm": 0.3352813720703125,
+      "learning_rate": 9.985581699578287e-05,
+      "loss": 2.2992,
+      "step": 146
+    },
+    {
+      "epoch": 0.036941634730162716,
+      "grad_norm": 0.43075451254844666,
+      "learning_rate": 9.985276708659134e-05,
+      "loss": 2.7181,
+      "step": 147
+    },
+    {
+      "epoch": 0.037192938367782875,
+      "grad_norm": 0.4140123128890991,
+      "learning_rate": 9.984968530425369e-05,
+      "loss": 2.2454,
+      "step": 148
+    },
+    {
+      "epoch": 0.03744424200540303,
+      "grad_norm": 0.31063738465309143,
+      "learning_rate": 9.984657165074027e-05,
+      "loss": 2.4994,
+      "step": 149
+    },
+    {
+      "epoch": 0.03769554564302318,
+      "grad_norm": 0.22530898451805115,
+      "learning_rate": 9.984342612804176e-05,
+      "loss": 2.0645,
+      "step": 150
+    },
+    {
+      "epoch": 0.03794684928064334,
+      "grad_norm": 0.2811500132083893,
+      "learning_rate": 9.984024873816924e-05,
+      "loss": 2.3546,
+      "step": 151
+    },
+    {
+      "epoch": 0.03819815291826349,
+      "grad_norm": 0.4520877003669739,
+      "learning_rate": 9.983703948315417e-05,
+      "loss": 2.1208,
+      "step": 152
+    },
+    {
+      "epoch": 0.038449456555883645,
+      "grad_norm": 0.37516894936561584,
+      "learning_rate": 9.983379836504838e-05,
+      "loss": 2.3911,
+      "step": 153
+    },
+    {
+      "epoch": 0.038700760193503804,
+      "grad_norm": 0.4740954637527466,
+      "learning_rate": 9.983052538592404e-05,
+      "loss": 2.4538,
+      "step": 154
+    },
+    {
+      "epoch": 0.038952063831123956,
+      "grad_norm": 0.34146809577941895,
+      "learning_rate": 9.982722054787372e-05,
+      "loss": 2.5654,
+      "step": 155
+    },
+    {
+      "epoch": 0.03920336746874411,
+      "grad_norm": 0.3912610709667206,
+      "learning_rate": 9.982388385301038e-05,
+      "loss": 2.4403,
+      "step": 156
+    },
+    {
+      "epoch": 0.03945467110636426,
+      "grad_norm": 0.5822238922119141,
+      "learning_rate": 9.98205153034673e-05,
+      "loss": 2.3008,
+      "step": 157
+    },
+    {
+      "epoch": 0.03970597474398442,
+      "grad_norm": 0.1491788774728775,
+      "learning_rate": 9.981711490139814e-05,
+      "loss": 1.1215,
+      "step": 158
+    },
+    {
+      "epoch": 0.03995727838160457,
+      "grad_norm": 0.3438681960105896,
+      "learning_rate": 9.981368264897694e-05,
+      "loss": 2.7011,
+      "step": 159
+    },
+    {
+      "epoch": 0.040208582019224726,
+      "grad_norm": 0.3378736972808838,
+      "learning_rate": 9.98102185483981e-05,
+      "loss": 2.4961,
+      "step": 160
+    },
+    {
+      "epoch": 0.040459885656844885,
+      "grad_norm": 0.2592676281929016,
+      "learning_rate": 9.980672260187638e-05,
+      "loss": 1.8838,
+      "step": 161
+    },
+    {
+      "epoch": 0.04071118929446504,
+      "grad_norm": 0.31087106466293335,
+      "learning_rate": 9.980319481164688e-05,
+      "loss": 2.2719,
+      "step": 162
+    },
+    {
+      "epoch": 0.04096249293208519,
+      "grad_norm": 0.33119067549705505,
+      "learning_rate": 9.979963517996509e-05,
+      "loss": 2.0298,
+      "step": 163
+    },
+    {
+      "epoch": 0.04121379656970535,
+      "grad_norm": 0.4013742506504059,
+      "learning_rate": 9.979604370910685e-05,
+      "loss": 2.5852,
+      "step": 164
+    },
+    {
+      "epoch": 0.0414651002073255,
+      "grad_norm": 0.21013818681240082,
+      "learning_rate": 9.979242040136835e-05,
+      "loss": 1.1489,
+      "step": 165
+    },
+    {
+      "epoch": 0.041716403844945654,
+      "grad_norm": 0.5467535257339478,
+      "learning_rate": 9.978876525906613e-05,
+      "loss": 2.2094,
+      "step": 166
+    },
+    {
+      "epoch": 0.04196770748256581,
+      "grad_norm": 0.3550753593444824,
+      "learning_rate": 9.978507828453708e-05,
+      "loss": 2.2849,
+      "step": 167
+    },
+    {
+      "epoch": 0.042219011120185966,
+      "grad_norm": 0.2620997726917267,
+      "learning_rate": 9.978135948013847e-05,
+      "loss": 2.3229,
+      "step": 168
+    },
+    {
+      "epoch": 0.04247031475780612,
+      "grad_norm": 0.2706509530544281,
+      "learning_rate": 9.977760884824788e-05,
+      "loss": 2.2731,
+      "step": 169
+    },
+    {
+      "epoch": 0.04272161839542627,
+      "grad_norm": 0.41886723041534424,
+      "learning_rate": 9.977382639126328e-05,
+      "loss": 2.42,
+      "step": 170
+    },
+    {
+      "epoch": 0.04297292203304643,
+      "grad_norm": 0.33128440380096436,
+      "learning_rate": 9.977001211160296e-05,
+      "loss": 2.3203,
+      "step": 171
+    },
+    {
+      "epoch": 0.04322422567066658,
+      "grad_norm": 0.24908378720283508,
+      "learning_rate": 9.976616601170557e-05,
+      "loss": 1.8275,
+      "step": 172
+    },
+    {
+      "epoch": 0.043475529308286735,
+      "grad_norm": 0.3278106451034546,
+      "learning_rate": 9.976228809403008e-05,
+      "loss": 2.6484,
+      "step": 173
+    },
+    {
+      "epoch": 0.043726832945906895,
+      "grad_norm": 0.5737075209617615,
+      "learning_rate": 9.975837836105581e-05,
+      "loss": 2.2196,
+      "step": 174
+    },
+    {
+      "epoch": 0.04397813658352705,
+      "grad_norm": 0.3396255373954773,
+      "learning_rate": 9.975443681528247e-05,
+      "loss": 2.1756,
+      "step": 175
+    },
+    {
+      "epoch": 0.0442294402211472,
+      "grad_norm": 0.4173082411289215,
+      "learning_rate": 9.975046345923004e-05,
+      "loss": 1.9394,
+      "step": 176
+    },
+    {
+      "epoch": 0.04448074385876736,
+      "grad_norm": 0.2630142867565155,
+      "learning_rate": 9.974645829543889e-05,
+      "loss": 1.9798,
+      "step": 177
+    },
+    {
+      "epoch": 0.04473204749638751,
+      "grad_norm": 0.23479297757148743,
+      "learning_rate": 9.974242132646967e-05,
+      "loss": 1.8764,
+      "step": 178
+    },
+    {
+      "epoch": 0.044983351134007664,
+      "grad_norm": 0.2806346118450165,
+      "learning_rate": 9.973835255490343e-05,
+      "loss": 2.2942,
+      "step": 179
+    },
+    {
+      "epoch": 0.045234654771627816,
+      "grad_norm": 0.2745998203754425,
+      "learning_rate": 9.97342519833415e-05,
+      "loss": 2.1606,
+      "step": 180
+    },
+    {
+      "epoch": 0.045485958409247976,
+      "grad_norm": 0.46454527974128723,
+      "learning_rate": 9.973011961440559e-05,
+      "loss": 2.2312,
+      "step": 181
+    },
+    {
+      "epoch": 0.04573726204686813,
+      "grad_norm": 0.34017616510391235,
+      "learning_rate": 9.972595545073769e-05,
+      "loss": 2.3171,
+      "step": 182
+    },
+    {
+      "epoch": 0.04598856568448828,
+      "grad_norm": 0.34439125657081604,
+      "learning_rate": 9.972175949500012e-05,
+      "loss": 2.3098,
+      "step": 183
+    },
+    {
+      "epoch": 0.04623986932210844,
+      "grad_norm": 0.47229647636413574,
+      "learning_rate": 9.97175317498756e-05,
+      "loss": 2.1515,
+      "step": 184
+    },
+    {
+      "epoch": 0.04649117295972859,
+      "grad_norm": 0.3148500621318817,
+      "learning_rate": 9.971327221806706e-05,
+      "loss": 2.5827,
+      "step": 185
+    },
+    {
+      "epoch": 0.046742476597348745,
+      "grad_norm": 0.20196139812469482,
+      "learning_rate": 9.970898090229785e-05,
+      "loss": 1.5597,
+      "step": 186
+    },
+    {
+      "epoch": 0.046993780234968904,
+      "grad_norm": 0.26498332619667053,
+      "learning_rate": 9.97046578053116e-05,
+      "loss": 2.4454,
+      "step": 187
+    },
+    {
+      "epoch": 0.04724508387258906,
+      "grad_norm": 0.4051288068294525,
+      "learning_rate": 9.970030292987225e-05,
+      "loss": 2.3758,
+      "step": 188
+    },
+    {
+      "epoch": 0.04749638751020921,
+      "grad_norm": 0.3003399074077606,
+      "learning_rate": 9.969591627876409e-05,
+      "loss": 2.691,
+      "step": 189
+    },
+    {
+      "epoch": 0.04774769114782936,
+      "grad_norm": 0.22101780772209167,
+      "learning_rate": 9.96914978547917e-05,
+      "loss": 2.361,
+      "step": 190
+    },
+    {
+      "epoch": 0.04799899478544952,
+      "grad_norm": 0.342926025390625,
+      "learning_rate": 9.968704766077997e-05,
+      "loss": 2.4845,
+      "step": 191
+    },
+    {
+      "epoch": 0.048250298423069674,
+      "grad_norm": 0.3619941174983978,
+      "learning_rate": 9.968256569957411e-05,
+      "loss": 2.6731,
+      "step": 192
+    },
+    {
+      "epoch": 0.048501602060689826,
+      "grad_norm": 0.4262664020061493,
+      "learning_rate": 9.967805197403965e-05,
+      "loss": 2.6423,
+      "step": 193
+    },
+    {
+      "epoch": 0.048752905698309985,
+      "grad_norm": 0.21453841030597687,
+      "learning_rate": 9.96735064870624e-05,
+      "loss": 1.0993,
+      "step": 194
+    },
+    {
+      "epoch": 0.04900420933593014,
+      "grad_norm": 0.5756468176841736,
+      "learning_rate": 9.966892924154853e-05,
+      "loss": 2.1122,
+      "step": 195
+    },
+    {
+      "epoch": 0.04925551297355029,
+      "grad_norm": 0.5691533088684082,
+      "learning_rate": 9.96643202404245e-05,
+      "loss": 2.3865,
+      "step": 196
+    },
+    {
+      "epoch": 0.04950681661117045,
+      "grad_norm": 0.4420826733112335,
+      "learning_rate": 9.965967948663698e-05,
+      "loss": 2.1786,
+      "step": 197
+    },
+    {
+      "epoch": 0.0497581202487906,
+      "grad_norm": 0.4547330141067505,
+      "learning_rate": 9.965500698315306e-05,
+      "loss": 2.5167,
+      "step": 198
+    },
+    {
+      "epoch": 0.050009423886410755,
+      "grad_norm": 0.18379782140254974,
+      "learning_rate": 9.96503027329601e-05,
+      "loss": 1.0926,
+      "step": 199
+    },
+    {
+      "epoch": 0.05026072752403091,
+      "grad_norm": 0.20280440151691437,
+      "learning_rate": 9.964556673906572e-05,
+      "loss": 2.0984,
+      "step": 200
+    },
+    {
+      "epoch": 0.050512031161651066,
+      "grad_norm": 0.5554583072662354,
+      "learning_rate": 9.964079900449785e-05,
+      "loss": 2.8414,
+      "step": 201
+    },
+    {
+      "epoch": 0.05076333479927122,
+      "grad_norm": 0.3430071473121643,
+      "learning_rate": 9.963599953230473e-05,
+      "loss": 2.4415,
+      "step": 202
+    },
+    {
+      "epoch": 0.05101463843689137,
+      "grad_norm": 0.34130024909973145,
+      "learning_rate": 9.96311683255549e-05,
+      "loss": 2.1872,
+      "step": 203
+    },
+    {
+      "epoch": 0.05126594207451153,
+      "grad_norm": 0.3914225995540619,
+      "learning_rate": 9.962630538733715e-05,
+      "loss": 2.3896,
+      "step": 204
+    },
+    {
+      "epoch": 0.05151724571213168,
+      "grad_norm": 0.6636744141578674,
+      "learning_rate": 9.962141072076057e-05,
+      "loss": 2.5256,
+      "step": 205
+    },
+    {
+      "epoch": 0.051768549349751836,
+      "grad_norm": 0.419360876083374,
+      "learning_rate": 9.961648432895454e-05,
+      "loss": 2.1897,
+      "step": 206
+    },
+    {
+      "epoch": 0.052019852987371995,
+      "grad_norm": 0.3426320552825928,
+      "learning_rate": 9.961152621506876e-05,
+      "loss": 2.433,
+      "step": 207
+    },
+    {
+      "epoch": 0.05227115662499215,
+      "grad_norm": 0.24097320437431335,
+      "learning_rate": 9.960653638227315e-05,
+      "loss": 1.899,
+      "step": 208
+    },
+    {
+      "epoch": 0.0525224602626123,
+      "grad_norm": 0.40818917751312256,
+      "learning_rate": 9.960151483375795e-05,
+      "loss": 2.4313,
+      "step": 209
+    },
+    {
+      "epoch": 0.05277376390023246,
+      "grad_norm": 0.34848183393478394,
+      "learning_rate": 9.959646157273366e-05,
+      "loss": 1.9527,
+      "step": 210
+    },
+    {
+      "epoch": 0.05302506753785261,
+      "grad_norm": 0.18201400339603424,
+      "learning_rate": 9.959137660243105e-05,
+      "loss": 1.8784,
+      "step": 211
+    },
+    {
+      "epoch": 0.053276371175472764,
+      "grad_norm": 0.3388751745223999,
+      "learning_rate": 9.95862599261012e-05,
+      "loss": 2.4083,
+      "step": 212
+    },
+    {
+      "epoch": 0.05352767481309292,
+      "grad_norm": 0.2567930519580841,
+      "learning_rate": 9.958111154701542e-05,
+      "loss": 1.9071,
+      "step": 213
+    },
+    {
+      "epoch": 0.053778978450713076,
+      "grad_norm": 0.39157772064208984,
+      "learning_rate": 9.957593146846529e-05,
+      "loss": 2.072,
+      "step": 214
+    },
+    {
+      "epoch": 0.05403028208833323,
+      "grad_norm": 0.40348386764526367,
+      "learning_rate": 9.95707196937627e-05,
+      "loss": 2.427,
+      "step": 215
+    },
+    {
+      "epoch": 0.05428158572595338,
+      "grad_norm": 0.47843194007873535,
+      "learning_rate": 9.956547622623973e-05,
+      "loss": 2.0144,
+      "step": 216
+    },
+    {
+      "epoch": 0.05453288936357354,
+      "grad_norm": 0.337272971868515,
+      "learning_rate": 9.956020106924882e-05,
+      "loss": 1.7783,
+      "step": 217
+    },
+    {
+      "epoch": 0.05478419300119369,
+      "grad_norm": 0.40009111166000366,
+      "learning_rate": 9.955489422616258e-05,
+      "loss": 2.2961,
+      "step": 218
+    },
+    {
+      "epoch": 0.055035496638813845,
+      "grad_norm": 0.28611692786216736,
+      "learning_rate": 9.954955570037395e-05,
+      "loss": 2.44,
+      "step": 219
+    },
+    {
+      "epoch": 0.055286800276434005,
+      "grad_norm": 0.3352760374546051,
+      "learning_rate": 9.954418549529605e-05,
+      "loss": 2.2449,
+      "step": 220
+    },
+    {
+      "epoch": 0.05553810391405416,
+      "grad_norm": 0.295691579580307,
+      "learning_rate": 9.953878361436232e-05,
+      "loss": 1.9398,
+      "step": 221
+    },
+    {
+      "epoch": 0.05578940755167431,
+      "grad_norm": 0.41174155473709106,
+      "learning_rate": 9.953335006102643e-05,
+      "loss": 2.241,
+      "step": 222
+    },
+    {
+      "epoch": 0.05604071118929446,
+      "grad_norm": 0.32655176520347595,
+      "learning_rate": 9.95278848387623e-05,
+      "loss": 2.078,
+      "step": 223
+    },
+    {
+      "epoch": 0.05629201482691462,
+      "grad_norm": 0.26872923970222473,
+      "learning_rate": 9.95223879510641e-05,
+      "loss": 1.9153,
+      "step": 224
+    },
+    {
+      "epoch": 0.056543318464534774,
+      "grad_norm": 0.37920647859573364,
+      "learning_rate": 9.951685940144622e-05,
+      "loss": 2.5823,
+      "step": 225
+    },
+    {
+      "epoch": 0.056794622102154926,
+      "grad_norm": 0.18780489265918732,
+      "learning_rate": 9.951129919344334e-05,
+      "loss": 0.9544,
+      "step": 226
+    },
+    {
+      "epoch": 0.057045925739775086,
+      "grad_norm": 0.3492078185081482,
+      "learning_rate": 9.950570733061033e-05,
+      "loss": 2.3073,
+      "step": 227
+    },
+    {
+      "epoch": 0.05729722937739524,
+      "grad_norm": 0.29748043417930603,
+      "learning_rate": 9.950008381652235e-05,
+      "loss": 1.6764,
+      "step": 228
+    },
+    {
+      "epoch": 0.05754853301501539,
+      "grad_norm": 0.30301064252853394,
+      "learning_rate": 9.949442865477474e-05,
+      "loss": 2.4839,
+      "step": 229
+    },
+    {
+      "epoch": 0.05779983665263555,
+      "grad_norm": 1.056869387626648,
+      "learning_rate": 9.948874184898313e-05,
+      "loss": 1.9098,
+      "step": 230
+    },
+    {
+      "epoch": 0.0580511402902557,
+      "grad_norm": 0.33054810762405396,
+      "learning_rate": 9.948302340278333e-05,
+      "loss": 2.6708,
+      "step": 231
+    },
+    {
+      "epoch": 0.058302443927875855,
+      "grad_norm": 0.37433305382728577,
+      "learning_rate": 9.94772733198314e-05,
+      "loss": 2.4231,
+      "step": 232
+    },
+    {
+      "epoch": 0.05855374756549601,
+      "grad_norm": 0.35789650678634644,
+      "learning_rate": 9.947149160380366e-05,
+      "loss": 2.1926,
+      "step": 233
+    },
+    {
+      "epoch": 0.05880505120311617,
+      "grad_norm": 0.3637252151966095,
+      "learning_rate": 9.94656782583966e-05,
+      "loss": 2.0955,
+      "step": 234
+    },
+    {
+      "epoch": 0.05905635484073632,
+      "grad_norm": 0.37392285466194153,
+      "learning_rate": 9.945983328732698e-05,
+      "loss": 2.2662,
+      "step": 235
+    },
+    {
+      "epoch": 0.05930765847835647,
+      "grad_norm": 0.36526796221733093,
+      "learning_rate": 9.945395669433172e-05,
+      "loss": 2.4013,
+      "step": 236
+    },
+    {
+      "epoch": 0.05955896211597663,
+      "grad_norm": 0.3178480267524719,
+      "learning_rate": 9.944804848316802e-05,
+      "loss": 2.4751,
+      "step": 237
+    },
+    {
+      "epoch": 0.059810265753596784,
+      "grad_norm": 0.5924585461616516,
+      "learning_rate": 9.944210865761328e-05,
+      "loss": 2.4306,
+      "step": 238
+    },
+    {
+      "epoch": 0.060061569391216936,
+      "grad_norm": 0.19464784860610962,
+      "learning_rate": 9.943613722146505e-05,
+      "loss": 1.7291,
+      "step": 239
+    },
+    {
+      "epoch": 0.060312873028837095,
+      "grad_norm": 0.43970218300819397,
+      "learning_rate": 9.943013417854122e-05,
+      "loss": 2.6384,
+      "step": 240
+    },
+    {
+      "epoch": 0.06056417666645725,
+      "grad_norm": 0.24550621211528778,
+      "learning_rate": 9.942409953267972e-05,
+      "loss": 1.5974,
+      "step": 241
+    },
+    {
+      "epoch": 0.0608154803040774,
+      "grad_norm": 0.33134451508522034,
+      "learning_rate": 9.941803328773885e-05,
+      "loss": 2.4358,
+      "step": 242
+    },
+    {
+      "epoch": 0.06106678394169755,
+      "grad_norm": 0.5241190791130066,
+      "learning_rate": 9.941193544759699e-05,
+      "loss": 2.3483,
+      "step": 243
+    },
+    {
+      "epoch": 0.06131808757931771,
+      "grad_norm": 0.5123705863952637,
+      "learning_rate": 9.940580601615279e-05,
+      "loss": 1.9421,
+      "step": 244
+    },
+    {
+      "epoch": 0.061569391216937865,
+      "grad_norm": 0.3747979998588562,
+      "learning_rate": 9.939964499732507e-05,
+      "loss": 2.207,
+      "step": 245
+    },
+    {
+      "epoch": 0.06182069485455802,
+      "grad_norm": 0.3442586362361908,
+      "learning_rate": 9.939345239505284e-05,
+      "loss": 2.2754,
+      "step": 246
+    },
+    {
+      "epoch": 0.062071998492178176,
+      "grad_norm": 0.3288307785987854,
+      "learning_rate": 9.938722821329532e-05,
+      "loss": 2.3217,
+      "step": 247
+    },
+    {
+      "epoch": 0.06232330212979833,
+      "grad_norm": 0.33456501364707947,
+      "learning_rate": 9.938097245603193e-05,
+      "loss": 2.1507,
+      "step": 248
+    },
+    {
+      "epoch": 0.06257460576741848,
+      "grad_norm": 0.4627265930175781,
+      "learning_rate": 9.937468512726223e-05,
+      "loss": 2.4072,
+      "step": 249
+    },
+    {
+      "epoch": 0.06282590940503864,
+      "grad_norm": 0.2955438196659088,
+      "learning_rate": 9.9368366231006e-05,
+      "loss": 2.452,
+      "step": 250
+    },
+    {
+      "epoch": 0.06307721304265879,
+      "grad_norm": 0.7906093001365662,
+      "learning_rate": 9.936201577130324e-05,
+      "loss": 2.1179,
+      "step": 251
+    },
+    {
+      "epoch": 0.06332851668027895,
+      "grad_norm": 0.41081616282463074,
+      "learning_rate": 9.935563375221404e-05,
+      "loss": 2.0524,
+      "step": 252
+    },
+    {
+      "epoch": 0.0635798203178991,
+      "grad_norm": 0.18861036002635956,
+      "learning_rate": 9.934922017781873e-05,
+      "loss": 1.4761,
+      "step": 253
+    },
+    {
+      "epoch": 0.06383112395551925,
+      "grad_norm": 0.27071478962898254,
+      "learning_rate": 9.93427750522178e-05,
+      "loss": 2.2497,
+      "step": 254
+    },
+    {
+      "epoch": 0.06408242759313941,
+      "grad_norm": 0.335267573595047,
+      "learning_rate": 9.933629837953191e-05,
+      "loss": 2.3577,
+      "step": 255
+    },
+    {
+      "epoch": 0.06433373123075957,
+      "grad_norm": 0.3902337849140167,
+      "learning_rate": 9.932979016390192e-05,
+      "loss": 2.2602,
+      "step": 256
+    },
+    {
+      "epoch": 0.06458503486837971,
+      "grad_norm": 0.550452470779419,
+      "learning_rate": 9.932325040948878e-05,
+      "loss": 2.1322,
+      "step": 257
+    },
+    {
+      "epoch": 0.06483633850599987,
+      "grad_norm": 0.42229825258255005,
+      "learning_rate": 9.93166791204737e-05,
+      "loss": 2.449,
+      "step": 258
+    },
+    {
+      "epoch": 0.06508764214362003,
+      "grad_norm": 0.2767508924007416,
+      "learning_rate": 9.931007630105798e-05,
+      "loss": 2.552,
+      "step": 259
+    },
+    {
+      "epoch": 0.06533894578124018,
+      "grad_norm": 0.3925112783908844,
+      "learning_rate": 9.93034419554631e-05,
+      "loss": 2.5078,
+      "step": 260
+    },
+    {
+      "epoch": 0.06559024941886034,
+      "grad_norm": 0.38622385263442993,
+      "learning_rate": 9.929677608793072e-05,
+      "loss": 2.5889,
+      "step": 261
+    },
+    {
+      "epoch": 0.0658415530564805,
+      "grad_norm": 0.3186779022216797,
+      "learning_rate": 9.929007870272262e-05,
+      "loss": 2.3722,
+      "step": 262
+    },
+    {
+      "epoch": 0.06609285669410064,
+      "grad_norm": 0.308403342962265,
+      "learning_rate": 9.928334980412073e-05,
+      "loss": 2.8672,
+      "step": 263
+    },
+    {
+      "epoch": 0.0663441603317208,
+      "grad_norm": 0.37506890296936035,
+      "learning_rate": 9.927658939642716e-05,
+      "loss": 2.3802,
+      "step": 264
+    },
+    {
+      "epoch": 0.06659546396934096,
+      "grad_norm": 0.43554747104644775,
+      "learning_rate": 9.926979748396415e-05,
+      "loss": 2.1515,
+      "step": 265
+    },
+    {
+      "epoch": 0.06684676760696111,
+      "grad_norm": 0.4737273156642914,
+      "learning_rate": 9.926297407107406e-05,
+      "loss": 2.4804,
+      "step": 266
+    },
+    {
+      "epoch": 0.06709807124458127,
+      "grad_norm": 0.30481746792793274,
+      "learning_rate": 9.925611916211943e-05,
+      "loss": 2.5455,
+      "step": 267
+    },
+    {
+      "epoch": 0.06734937488220143,
+      "grad_norm": 0.29122671484947205,
+      "learning_rate": 9.92492327614829e-05,
+      "loss": 1.9563,
+      "step": 268
+    },
+    {
+      "epoch": 0.06760067851982157,
+      "grad_norm": 0.2516386806964874,
+      "learning_rate": 9.924231487356725e-05,
+      "loss": 2.1242,
+      "step": 269
+    },
+    {
+      "epoch": 0.06785198215744173,
+      "grad_norm": 0.1990005224943161,
+      "learning_rate": 9.923536550279544e-05,
+      "loss": 1.6236,
+      "step": 270
+    },
+    {
+      "epoch": 0.06810328579506188,
+      "grad_norm": 0.3531213402748108,
+      "learning_rate": 9.92283846536105e-05,
+      "loss": 2.3706,
+      "step": 271
+    },
+    {
+      "epoch": 0.06835458943268204,
+      "grad_norm": 0.358190655708313,
+      "learning_rate": 9.922137233047558e-05,
+      "loss": 2.3467,
+      "step": 272
+    },
+    {
+      "epoch": 0.0686058930703022,
+      "grad_norm": 0.3931328058242798,
+      "learning_rate": 9.9214328537874e-05,
+      "loss": 2.359,
+      "step": 273
+    },
+    {
+      "epoch": 0.06885719670792234,
+      "grad_norm": 0.30483195185661316,
+      "learning_rate": 9.92072532803092e-05,
+      "loss": 2.8789,
+      "step": 274
+    },
+    {
+      "epoch": 0.0691085003455425,
+      "grad_norm": 0.155193030834198,
+      "learning_rate": 9.920014656230468e-05,
+      "loss": 1.4218,
+      "step": 275
+    },
+    {
+      "epoch": 0.06935980398316266,
+      "grad_norm": 0.2718715965747833,
+      "learning_rate": 9.919300838840409e-05,
+      "loss": 2.6687,
+      "step": 276
+    },
+    {
+      "epoch": 0.0696111076207828,
+      "grad_norm": 0.45890912413597107,
+      "learning_rate": 9.91858387631712e-05,
+      "loss": 2.4566,
+      "step": 277
+    },
+    {
+      "epoch": 0.06986241125840296,
+      "grad_norm": 0.39152559638023376,
+      "learning_rate": 9.917863769118988e-05,
+      "loss": 2.7171,
+      "step": 278
+    },
+    {
+      "epoch": 0.07011371489602312,
+      "grad_norm": 0.423405259847641,
+      "learning_rate": 9.91714051770641e-05,
+      "loss": 2.4389,
+      "step": 279
+    },
+    {
+      "epoch": 0.07036501853364327,
+      "grad_norm": 0.2914890944957733,
+      "learning_rate": 9.916414122541794e-05,
+      "loss": 2.8702,
+      "step": 280
+    },
+    {
+      "epoch": 0.07061632217126343,
+      "grad_norm": 0.45607659220695496,
+      "learning_rate": 9.915684584089557e-05,
+      "loss": 2.2563,
+      "step": 281
+    },
+    {
+      "epoch": 0.07086762580888359,
+      "grad_norm": 0.8603537678718567,
+      "learning_rate": 9.914951902816128e-05,
+      "loss": 2.0017,
+      "step": 282
+    },
+    {
+      "epoch": 0.07111892944650373,
+      "grad_norm": 0.34362930059432983,
+      "learning_rate": 9.91421607918994e-05,
+      "loss": 2.5167,
+      "step": 283
+    },
+    {
+      "epoch": 0.0713702330841239,
+      "grad_norm": 0.3992220163345337,
+      "learning_rate": 9.913477113681441e-05,
+      "loss": 2.1986,
+      "step": 284
+    },
+    {
+      "epoch": 0.07162153672174405,
+      "grad_norm": 0.409201443195343,
+      "learning_rate": 9.912735006763085e-05,
+      "loss": 1.8954,
+      "step": 285
+    },
+    {
+      "epoch": 0.0718728403593642,
+      "grad_norm": 0.293445348739624,
+      "learning_rate": 9.911989758909335e-05,
+      "loss": 2.3246,
+      "step": 286
+    },
+    {
+      "epoch": 0.07212414399698436,
+      "grad_norm": 0.3958424925804138,
+      "learning_rate": 9.911241370596663e-05,
+      "loss": 2.35,
+      "step": 287
+    },
+    {
+      "epoch": 0.07237544763460452,
+      "grad_norm": 0.5355744957923889,
+      "learning_rate": 9.910489842303544e-05,
+      "loss": 2.3103,
+      "step": 288
+    },
+    {
+      "epoch": 0.07262675127222466,
+      "grad_norm": 0.4253835678100586,
+      "learning_rate": 9.909735174510467e-05,
+      "loss": 2.5327,
+      "step": 289
+    },
+    {
+      "epoch": 0.07287805490984482,
+      "grad_norm": 0.2699333727359772,
+      "learning_rate": 9.908977367699926e-05,
+      "loss": 1.4629,
+      "step": 290
+    },
+    {
+      "epoch": 0.07312935854746497,
+      "grad_norm": 0.44158613681793213,
+      "learning_rate": 9.90821642235642e-05,
+      "loss": 2.1242,
+      "step": 291
+    },
+    {
+      "epoch": 0.07338066218508513,
+      "grad_norm": 0.33713415265083313,
+      "learning_rate": 9.907452338966457e-05,
+      "loss": 2.5233,
+      "step": 292
+    },
+    {
+      "epoch": 0.07363196582270529,
+      "grad_norm": 0.21191152930259705,
+      "learning_rate": 9.906685118018549e-05,
+      "loss": 1.9315,
+      "step": 293
+    },
+    {
+      "epoch": 0.07388326946032543,
+      "grad_norm": 0.3546162247657776,
+      "learning_rate": 9.905914760003216e-05,
+      "loss": 2.3661,
+      "step": 294
+    },
+    {
+      "epoch": 0.07413457309794559,
+      "grad_norm": 0.33006587624549866,
+      "learning_rate": 9.905141265412984e-05,
+      "loss": 1.7989,
+      "step": 295
+    },
+    {
+      "epoch": 0.07438587673556575,
+      "grad_norm": 0.4529229700565338,
+      "learning_rate": 9.904364634742385e-05,
+      "loss": 2.4695,
+      "step": 296
+    },
+    {
+      "epoch": 0.0746371803731859,
+      "grad_norm": 0.29525983333587646,
+      "learning_rate": 9.90358486848795e-05,
+      "loss": 2.1218,
+      "step": 297
+    },
+    {
+      "epoch": 0.07488848401080606,
+      "grad_norm": 0.273483544588089,
+      "learning_rate": 9.902801967148219e-05,
+      "loss": 2.4696,
+      "step": 298
+    },
+    {
+      "epoch": 0.07513978764842622,
+      "grad_norm": 0.3772584795951843,
+      "learning_rate": 9.902015931223742e-05,
+      "loss": 2.3749,
+      "step": 299
+    },
+    {
+      "epoch": 0.07539109128604636,
+      "grad_norm": 0.15759634971618652,
+      "learning_rate": 9.901226761217062e-05,
+      "loss": 0.7498,
+      "step": 300
+    },
+    {
+      "epoch": 0.07564239492366652,
+      "grad_norm": 0.2802472710609436,
+      "learning_rate": 9.900434457632734e-05,
+      "loss": 2.3353,
+      "step": 301
+    },
+    {
+      "epoch": 0.07589369856128668,
+      "grad_norm": 0.7529959678649902,
+      "learning_rate": 9.899639020977314e-05,
+      "loss": 1.7885,
+      "step": 302
+    },
+    {
+      "epoch": 0.07614500219890682,
+      "grad_norm": 0.3906523585319519,
+      "learning_rate": 9.89884045175936e-05,
+      "loss": 2.4677,
+      "step": 303
+    },
+    {
+      "epoch": 0.07639630583652698,
+      "grad_norm": 0.31307291984558105,
+      "learning_rate": 9.898038750489433e-05,
+      "loss": 2.408,
+      "step": 304
+    },
+    {
+      "epoch": 0.07664760947414714,
+      "grad_norm": 0.3155834674835205,
+      "learning_rate": 9.897233917680098e-05,
+      "loss": 2.0308,
+      "step": 305
+    },
+    {
+      "epoch": 0.07689891311176729,
+      "grad_norm": 0.2838704586029053,
+      "learning_rate": 9.896425953845923e-05,
+      "loss": 2.1338,
+      "step": 306
+    },
+    {
+      "epoch": 0.07715021674938745,
+      "grad_norm": 0.3904295563697815,
+      "learning_rate": 9.895614859503472e-05,
+      "loss": 2.4403,
+      "step": 307
+    },
+    {
+      "epoch": 0.07740152038700761,
+      "grad_norm": 0.27329105138778687,
+      "learning_rate": 9.89480063517132e-05,
+      "loss": 1.7174,
+      "step": 308
+    },
+    {
+      "epoch": 0.07765282402462775,
+      "grad_norm": 0.20863182842731476,
+      "learning_rate": 9.893983281370034e-05,
+      "loss": 2.1524,
+      "step": 309
+    },
+    {
+      "epoch": 0.07790412766224791,
+      "grad_norm": 0.3970401883125305,
+      "learning_rate": 9.893162798622185e-05,
+      "loss": 3.0095,
+      "step": 310
+    },
+    {
+      "epoch": 0.07815543129986807,
+      "grad_norm": 0.3557833731174469,
+      "learning_rate": 9.892339187452347e-05,
+      "loss": 2.2407,
+      "step": 311
+    },
+    {
+      "epoch": 0.07840673493748822,
+      "grad_norm": 0.3668101131916046,
+      "learning_rate": 9.891512448387092e-05,
+      "loss": 1.8899,
+      "step": 312
+    },
+    {
+      "epoch": 0.07865803857510838,
+      "grad_norm": 0.2897646129131317,
+      "learning_rate": 9.890682581954991e-05,
+      "loss": 2.4487,
+      "step": 313
+    },
+    {
+      "epoch": 0.07890934221272852,
+      "grad_norm": 0.31541863083839417,
+      "learning_rate": 9.889849588686617e-05,
+      "loss": 1.9136,
+      "step": 314
+    },
+    {
+      "epoch": 0.07916064585034868,
+      "grad_norm": 0.4142214357852936,
+      "learning_rate": 9.889013469114539e-05,
+      "loss": 2.4315,
+      "step": 315
+    },
+    {
+      "epoch": 0.07941194948796884,
+      "grad_norm": 0.39966338872909546,
+      "learning_rate": 9.888174223773325e-05,
+      "loss": 2.0768,
+      "step": 316
+    },
+    {
+      "epoch": 0.07966325312558899,
+      "grad_norm": 0.3038378357887268,
+      "learning_rate": 9.887331853199546e-05,
+      "loss": 2.4825,
+      "step": 317
+    },
+    {
+      "epoch": 0.07991455676320915,
+      "grad_norm": 0.36689433455467224,
+      "learning_rate": 9.886486357931767e-05,
+      "loss": 2.3153,
+      "step": 318
+    },
+    {
+      "epoch": 0.0801658604008293,
+      "grad_norm": 0.3301517963409424,
+      "learning_rate": 9.885637738510551e-05,
+      "loss": 2.4927,
+      "step": 319
+    },
+    {
+      "epoch": 0.08041716403844945,
+      "grad_norm": 0.26406022906303406,
+      "learning_rate": 9.884785995478458e-05,
+      "loss": 2.3793,
+      "step": 320
+    },
+    {
+      "epoch": 0.08066846767606961,
+      "grad_norm": 0.3765343427658081,
+      "learning_rate": 9.883931129380049e-05,
+      "loss": 2.3813,
+      "step": 321
+    },
+    {
+      "epoch": 0.08091977131368977,
+      "grad_norm": 0.42643749713897705,
+      "learning_rate": 9.883073140761876e-05,
+      "loss": 2.444,
+      "step": 322
+    },
+    {
+      "epoch": 0.08117107495130992,
+      "grad_norm": 0.42273804545402527,
+      "learning_rate": 9.882212030172493e-05,
+      "loss": 2.3634,
+      "step": 323
+    },
+    {
+      "epoch": 0.08142237858893007,
+      "grad_norm": 0.3113279938697815,
+      "learning_rate": 9.881347798162443e-05,
+      "loss": 2.236,
+      "step": 324
+    },
+    {
+      "epoch": 0.08167368222655023,
+      "grad_norm": 0.4435007870197296,
+      "learning_rate": 9.880480445284274e-05,
+      "loss": 2.3091,
+      "step": 325
+    },
+    {
+      "epoch": 0.08192498586417038,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 9.879609972092522e-05,
+      "loss": 1.7405,
+      "step": 326
+    },
+    {
+      "epoch": 0.08217628950179054,
+      "grad_norm": 0.3486250638961792,
+      "learning_rate": 9.878736379143719e-05,
+      "loss": 1.9632,
+      "step": 327
+    },
+    {
+      "epoch": 0.0824275931394107,
+      "grad_norm": 0.33690646290779114,
+      "learning_rate": 9.877859666996395e-05,
+      "loss": 2.454,
+      "step": 328
+    },
+    {
+      "epoch": 0.08267889677703084,
+      "grad_norm": 0.4438174068927765,
+      "learning_rate": 9.876979836211069e-05,
+      "loss": 2.0577,
+      "step": 329
+    },
+    {
+      "epoch": 0.082930200414651,
+      "grad_norm": 0.13432446122169495,
+      "learning_rate": 9.87609688735026e-05,
+      "loss": 0.8846,
+      "step": 330
+    },
+    {
+      "epoch": 0.08318150405227116,
+      "grad_norm": 0.4359930753707886,
+      "learning_rate": 9.875210820978475e-05,
+      "loss": 2.1561,
+      "step": 331
+    },
+    {
+      "epoch": 0.08343280768989131,
+      "grad_norm": 0.38827580213546753,
+      "learning_rate": 9.87432163766222e-05,
+      "loss": 2.2525,
+      "step": 332
+    },
+    {
+      "epoch": 0.08368411132751147,
+      "grad_norm": 0.3862955868244171,
+      "learning_rate": 9.873429337969985e-05,
+      "loss": 2.3337,
+      "step": 333
+    },
+    {
+      "epoch": 0.08393541496513161,
+      "grad_norm": 0.3881922960281372,
+      "learning_rate": 9.872533922472264e-05,
+      "loss": 2.3624,
+      "step": 334
+    },
+    {
+      "epoch": 0.08418671860275177,
+      "grad_norm": 0.21210597455501556,
+      "learning_rate": 9.871635391741533e-05,
+      "loss": 2.0269,
+      "step": 335
+    },
+    {
+      "epoch": 0.08443802224037193,
+      "grad_norm": 0.38510724902153015,
+      "learning_rate": 9.870733746352265e-05,
+      "loss": 2.1879,
+      "step": 336
+    },
+    {
+      "epoch": 0.08468932587799208,
+      "grad_norm": 0.31321218609809875,
+      "learning_rate": 9.869828986880924e-05,
+      "loss": 1.8977,
+      "step": 337
+    },
+    {
+      "epoch": 0.08494062951561224,
+      "grad_norm": 0.3835349678993225,
+      "learning_rate": 9.868921113905961e-05,
+      "loss": 2.5596,
+      "step": 338
+    },
+    {
+      "epoch": 0.0851919331532324,
+      "grad_norm": 0.2992005944252014,
+      "learning_rate": 9.868010128007823e-05,
+      "loss": 2.3065,
+      "step": 339
+    },
+    {
+      "epoch": 0.08544323679085254,
+      "grad_norm": 0.31576502323150635,
+      "learning_rate": 9.867096029768943e-05,
+      "loss": 2.4326,
+      "step": 340
+    },
+    {
+      "epoch": 0.0856945404284727,
+      "grad_norm": 0.265114426612854,
+      "learning_rate": 9.866178819773747e-05,
+      "loss": 2.4567,
+      "step": 341
+    },
+    {
+      "epoch": 0.08594584406609286,
+      "grad_norm": 0.40755563974380493,
+      "learning_rate": 9.86525849860865e-05,
+      "loss": 2.4134,
+      "step": 342
+    },
+    {
+      "epoch": 0.086197147703713,
+      "grad_norm": 0.35845133662223816,
+      "learning_rate": 9.864335066862054e-05,
+      "loss": 2.1912,
+      "step": 343
+    },
+    {
+      "epoch": 0.08644845134133317,
+      "grad_norm": 0.2640887498855591,
+      "learning_rate": 9.863408525124349e-05,
+      "loss": 2.3818,
+      "step": 344
+    },
+    {
+      "epoch": 0.08669975497895333,
+      "grad_norm": 0.19019848108291626,
+      "learning_rate": 9.862478873987919e-05,
+      "loss": 1.2064,
+      "step": 345
+    },
+    {
+      "epoch": 0.08695105861657347,
+      "grad_norm": 0.3398876488208771,
+      "learning_rate": 9.861546114047131e-05,
+      "loss": 2.513,
+      "step": 346
+    },
+    {
+      "epoch": 0.08720236225419363,
+      "grad_norm": 0.3897905945777893,
+      "learning_rate": 9.86061024589834e-05,
+      "loss": 2.3802,
+      "step": 347
+    },
+    {
+      "epoch": 0.08745366589181379,
+      "grad_norm": 0.5403574109077454,
+      "learning_rate": 9.859671270139892e-05,
+      "loss": 2.492,
+      "step": 348
+    },
+    {
+      "epoch": 0.08770496952943393,
+      "grad_norm": 0.38538798689842224,
+      "learning_rate": 9.858729187372114e-05,
+      "loss": 2.6697,
+      "step": 349
+    },
+    {
+      "epoch": 0.0879562731670541,
+      "grad_norm": 0.47243911027908325,
+      "learning_rate": 9.857783998197321e-05,
+      "loss": 2.5493,
+      "step": 350
+    },
+    {
+      "epoch": 0.08820757680467425,
+      "grad_norm": 0.24692249298095703,
+      "learning_rate": 9.85683570321982e-05,
+      "loss": 1.9117,
+      "step": 351
+    },
+    {
+      "epoch": 0.0884588804422944,
+      "grad_norm": 0.35984736680984497,
+      "learning_rate": 9.855884303045897e-05,
+      "loss": 2.0952,
+      "step": 352
+    },
+    {
+      "epoch": 0.08871018407991456,
+      "grad_norm": 0.4461250901222229,
+      "learning_rate": 9.854929798283826e-05,
+      "loss": 2.1836,
+      "step": 353
+    },
+    {
+      "epoch": 0.08896148771753472,
+      "grad_norm": 0.4806773066520691,
+      "learning_rate": 9.853972189543864e-05,
+      "loss": 2.0937,
+      "step": 354
+    },
+    {
+      "epoch": 0.08921279135515486,
+      "grad_norm": 0.5709269642829895,
+      "learning_rate": 9.853011477438254e-05,
+      "loss": 1.9677,
+      "step": 355
+    },
+    {
+      "epoch": 0.08946409499277502,
+      "grad_norm": 0.3929988443851471,
+      "learning_rate": 9.852047662581225e-05,
+      "loss": 2.5909,
+      "step": 356
+    },
+    {
+      "epoch": 0.08971539863039517,
+      "grad_norm": 0.33998608589172363,
+      "learning_rate": 9.851080745588987e-05,
+      "loss": 2.0388,
+      "step": 357
+    },
+    {
+      "epoch": 0.08996670226801533,
+      "grad_norm": 0.4756825268268585,
+      "learning_rate": 9.850110727079735e-05,
+      "loss": 2.1712,
+      "step": 358
+    },
+    {
+      "epoch": 0.09021800590563549,
+      "grad_norm": 0.3752896785736084,
+      "learning_rate": 9.849137607673643e-05,
+      "loss": 2.4498,
+      "step": 359
+    },
+    {
+      "epoch": 0.09046930954325563,
+      "grad_norm": 0.4232666790485382,
+      "learning_rate": 9.848161387992874e-05,
+      "loss": 2.0123,
+      "step": 360
+    },
+    {
+      "epoch": 0.09072061318087579,
+      "grad_norm": 0.4741215705871582,
+      "learning_rate": 9.847182068661567e-05,
+      "loss": 2.0118,
+      "step": 361
+    },
+    {
+      "epoch": 0.09097191681849595,
+      "grad_norm": 0.41064271330833435,
+      "learning_rate": 9.846199650305846e-05,
+      "loss": 2.2685,
+      "step": 362
+    },
+    {
+      "epoch": 0.0912232204561161,
+      "grad_norm": 0.31049537658691406,
+      "learning_rate": 9.845214133553817e-05,
+      "loss": 1.9623,
+      "step": 363
+    },
+    {
+      "epoch": 0.09147452409373626,
+      "grad_norm": 0.2547619342803955,
+      "learning_rate": 9.844225519035565e-05,
+      "loss": 1.8739,
+      "step": 364
+    },
+    {
+      "epoch": 0.09172582773135642,
+      "grad_norm": 0.26991865038871765,
+      "learning_rate": 9.843233807383159e-05,
+      "loss": 2.2907,
+      "step": 365
+    },
+    {
+      "epoch": 0.09197713136897656,
+      "grad_norm": 0.3443757891654968,
+      "learning_rate": 9.84223899923064e-05,
+      "loss": 2.1534,
+      "step": 366
+    },
+    {
+      "epoch": 0.09222843500659672,
+      "grad_norm": 0.2624412477016449,
+      "learning_rate": 9.841241095214038e-05,
+      "loss": 2.4799,
+      "step": 367
+    },
+    {
+      "epoch": 0.09247973864421688,
+      "grad_norm": 0.34518447518348694,
+      "learning_rate": 9.840240095971358e-05,
+      "loss": 2.2512,
+      "step": 368
+    },
+    {
+      "epoch": 0.09273104228183703,
+      "grad_norm": 0.4448896646499634,
+      "learning_rate": 9.839236002142584e-05,
+      "loss": 2.3634,
+      "step": 369
+    },
+    {
+      "epoch": 0.09298234591945718,
+      "grad_norm": 0.1713920682668686,
+      "learning_rate": 9.83822881436968e-05,
+      "loss": 1.9883,
+      "step": 370
+    },
+    {
+      "epoch": 0.09323364955707734,
+      "grad_norm": 0.26538336277008057,
+      "learning_rate": 9.837218533296587e-05,
+      "loss": 2.3259,
+      "step": 371
+    },
+    {
+      "epoch": 0.09348495319469749,
+      "grad_norm": 0.3171433210372925,
+      "learning_rate": 9.83620515956922e-05,
+      "loss": 2.3368,
+      "step": 372
+    },
+    {
+      "epoch": 0.09373625683231765,
+      "grad_norm": 0.47525712847709656,
+      "learning_rate": 9.83518869383548e-05,
+      "loss": 2.0206,
+      "step": 373
+    },
+    {
+      "epoch": 0.09398756046993781,
+      "grad_norm": 0.6884750723838806,
+      "learning_rate": 9.834169136745237e-05,
+      "loss": 2.2423,
+      "step": 374
+    },
+    {
+      "epoch": 0.09423886410755795,
+      "grad_norm": 0.3786754012107849,
+      "learning_rate": 9.833146488950342e-05,
+      "loss": 2.5151,
+      "step": 375
+    },
+    {
+      "epoch": 0.09449016774517811,
+      "grad_norm": 0.15857549011707306,
+      "learning_rate": 9.832120751104617e-05,
+      "loss": 1.4356,
+      "step": 376
+    },
+    {
+      "epoch": 0.09474147138279826,
+      "grad_norm": 0.6414403915405273,
+      "learning_rate": 9.831091923863868e-05,
+      "loss": 2.2734,
+      "step": 377
+    },
+    {
+      "epoch": 0.09499277502041842,
+      "grad_norm": 0.25082263350486755,
+      "learning_rate": 9.830060007885868e-05,
+      "loss": 2.8101,
+      "step": 378
+    },
+    {
+      "epoch": 0.09524407865803858,
+      "grad_norm": 0.29801589250564575,
+      "learning_rate": 9.829025003830368e-05,
+      "loss": 1.8527,
+      "step": 379
+    },
+    {
+      "epoch": 0.09549538229565872,
+      "grad_norm": 0.293905109167099,
+      "learning_rate": 9.827986912359094e-05,
+      "loss": 2.1692,
+      "step": 380
+    },
+    {
+      "epoch": 0.09574668593327888,
+      "grad_norm": 0.5134365558624268,
+      "learning_rate": 9.826945734135744e-05,
+      "loss": 2.1052,
+      "step": 381
+    },
+    {
+      "epoch": 0.09599798957089904,
+      "grad_norm": 0.24011516571044922,
+      "learning_rate": 9.825901469825994e-05,
+      "loss": 1.8188,
+      "step": 382
+    },
+    {
+      "epoch": 0.09624929320851919,
+      "grad_norm": 0.39140889048576355,
+      "learning_rate": 9.824854120097485e-05,
+      "loss": 2.1372,
+      "step": 383
+    },
+    {
+      "epoch": 0.09650059684613935,
+      "grad_norm": 0.35121777653694153,
+      "learning_rate": 9.82380368561984e-05,
+      "loss": 2.2186,
+      "step": 384
+    },
+    {
+      "epoch": 0.0967519004837595,
+      "grad_norm": 0.3034665286540985,
+      "learning_rate": 9.822750167064645e-05,
+      "loss": 2.2167,
+      "step": 385
+    },
+    {
+      "epoch": 0.09700320412137965,
+      "grad_norm": 0.18376107513904572,
+      "learning_rate": 9.821693565105465e-05,
+      "loss": 1.573,
+      "step": 386
+    },
+    {
+      "epoch": 0.09725450775899981,
+      "grad_norm": 0.18543782830238342,
+      "learning_rate": 9.820633880417836e-05,
+      "loss": 2.0766,
+      "step": 387
+    },
+    {
+      "epoch": 0.09750581139661997,
+      "grad_norm": 0.3020473122596741,
+      "learning_rate": 9.819571113679258e-05,
+      "loss": 2.4098,
+      "step": 388
+    },
+    {
+      "epoch": 0.09775711503424012,
+      "grad_norm": 0.43857908248901367,
+      "learning_rate": 9.818505265569209e-05,
+      "loss": 2.3636,
+      "step": 389
+    },
+    {
+      "epoch": 0.09800841867186028,
+      "grad_norm": 0.31539830565452576,
+      "learning_rate": 9.817436336769135e-05,
+      "loss": 2.3027,
+      "step": 390
+    },
+    {
+      "epoch": 0.09825972230948044,
+      "grad_norm": 0.38662660121917725,
+      "learning_rate": 9.816364327962449e-05,
+      "loss": 2.7305,
+      "step": 391
+    },
+    {
+      "epoch": 0.09851102594710058,
+      "grad_norm": 0.6164030432701111,
+      "learning_rate": 9.815289239834536e-05,
+      "loss": 1.8858,
+      "step": 392
+    },
+    {
+      "epoch": 0.09876232958472074,
+      "grad_norm": 0.2521904408931732,
+      "learning_rate": 9.814211073072748e-05,
+      "loss": 1.9274,
+      "step": 393
+    },
+    {
+      "epoch": 0.0990136332223409,
+      "grad_norm": 0.35303381085395813,
+      "learning_rate": 9.813129828366407e-05,
+      "loss": 2.1945,
+      "step": 394
+    },
+    {
+      "epoch": 0.09926493685996104,
+      "grad_norm": 0.276737242937088,
+      "learning_rate": 9.812045506406803e-05,
+      "loss": 2.2649,
+      "step": 395
+    },
+    {
+      "epoch": 0.0995162404975812,
+      "grad_norm": 0.45490264892578125,
+      "learning_rate": 9.81095810788719e-05,
+      "loss": 2.37,
+      "step": 396
+    },
+    {
+      "epoch": 0.09976754413520136,
+      "grad_norm": 0.5738433599472046,
+      "learning_rate": 9.809867633502794e-05,
+      "loss": 2.1393,
+      "step": 397
+    },
+    {
+      "epoch": 0.10001884777282151,
+      "grad_norm": 0.2685263454914093,
+      "learning_rate": 9.808774083950802e-05,
+      "loss": 2.6325,
+      "step": 398
+    },
+    {
+      "epoch": 0.10027015141044167,
+      "grad_norm": 0.41066989302635193,
+      "learning_rate": 9.807677459930374e-05,
+      "loss": 2.1897,
+      "step": 399
+    },
+    {
+      "epoch": 0.10052145504806181,
+      "grad_norm": 0.41453301906585693,
+      "learning_rate": 9.806577762142628e-05,
+      "loss": 2.243,
+      "step": 400
+    },
+    {
+      "epoch": 0.10077275868568197,
+      "grad_norm": 0.39212143421173096,
+      "learning_rate": 9.805474991290652e-05,
+      "loss": 2.502,
+      "step": 401
+    },
+    {
+      "epoch": 0.10102406232330213,
+      "grad_norm": 0.23721270263195038,
+      "learning_rate": 9.804369148079498e-05,
+      "loss": 1.7259,
+      "step": 402
+    },
+    {
+      "epoch": 0.10127536596092228,
+      "grad_norm": 0.22118382155895233,
+      "learning_rate": 9.803260233216184e-05,
+      "loss": 1.1238,
+      "step": 403
+    },
+    {
+      "epoch": 0.10152666959854244,
+      "grad_norm": 0.386078804731369,
+      "learning_rate": 9.802148247409686e-05,
+      "loss": 2.1155,
+      "step": 404
+    },
+    {
+      "epoch": 0.1017779732361626,
+      "grad_norm": 0.3332570195198059,
+      "learning_rate": 9.80103319137095e-05,
+      "loss": 2.5907,
+      "step": 405
+    },
+    {
+      "epoch": 0.10202927687378274,
+      "grad_norm": 0.20631489157676697,
+      "learning_rate": 9.799915065812882e-05,
+      "loss": 2.3401,
+      "step": 406
+    },
+    {
+      "epoch": 0.1022805805114029,
+      "grad_norm": 0.529591977596283,
+      "learning_rate": 9.798793871450346e-05,
+      "loss": 2.3916,
+      "step": 407
+    },
+    {
+      "epoch": 0.10253188414902306,
+      "grad_norm": 0.32578417658805847,
+      "learning_rate": 9.79766960900018e-05,
+      "loss": 2.3661,
+      "step": 408
+    },
+    {
+      "epoch": 0.10278318778664321,
+      "grad_norm": 0.4134072959423065,
+      "learning_rate": 9.796542279181172e-05,
+      "loss": 2.0656,
+      "step": 409
+    },
+    {
+      "epoch": 0.10303449142426337,
+      "grad_norm": 0.3845951557159424,
+      "learning_rate": 9.795411882714076e-05,
+      "loss": 2.3176,
+      "step": 410
+    },
+    {
+      "epoch": 0.10328579506188353,
+      "grad_norm": 0.6262491345405579,
+      "learning_rate": 9.794278420321605e-05,
+      "loss": 2.42,
+      "step": 411
+    },
+    {
+      "epoch": 0.10353709869950367,
+      "grad_norm": 0.4291568696498871,
+      "learning_rate": 9.793141892728436e-05,
+      "loss": 1.9455,
+      "step": 412
+    },
+    {
+      "epoch": 0.10378840233712383,
+      "grad_norm": 0.24446240067481995,
+      "learning_rate": 9.792002300661201e-05,
+      "loss": 2.6102,
+      "step": 413
+    },
+    {
+      "epoch": 0.10403970597474399,
+      "grad_norm": 0.4551761746406555,
+      "learning_rate": 9.79085964484849e-05,
+      "loss": 2.0785,
+      "step": 414
+    },
+    {
+      "epoch": 0.10429100961236414,
+      "grad_norm": 0.384036123752594,
+      "learning_rate": 9.789713926020863e-05,
+      "loss": 2.3401,
+      "step": 415
+    },
+    {
+      "epoch": 0.1045423132499843,
+      "grad_norm": 0.2475469559431076,
+      "learning_rate": 9.788565144910822e-05,
+      "loss": 2.2387,
+      "step": 416
+    },
+    {
+      "epoch": 0.10479361688760445,
+      "grad_norm": 0.269940048456192,
+      "learning_rate": 9.78741330225284e-05,
+      "loss": 2.2894,
+      "step": 417
+    },
+    {
+      "epoch": 0.1050449205252246,
+      "grad_norm": 0.27409201860427856,
+      "learning_rate": 9.786258398783341e-05,
+      "loss": 2.2478,
+      "step": 418
+    },
+    {
+      "epoch": 0.10529622416284476,
+      "grad_norm": 0.28689828515052795,
+      "learning_rate": 9.785100435240706e-05,
+      "loss": 1.9406,
+      "step": 419
+    },
+    {
+      "epoch": 0.10554752780046492,
+      "grad_norm": 0.5282906293869019,
+      "learning_rate": 9.783939412365278e-05,
+      "loss": 2.1423,
+      "step": 420
+    },
+    {
+      "epoch": 0.10579883143808506,
+      "grad_norm": 0.2722564935684204,
+      "learning_rate": 9.782775330899347e-05,
+      "loss": 2.3398,
+      "step": 421
+    },
+    {
+      "epoch": 0.10605013507570522,
+      "grad_norm": 0.22443710267543793,
+      "learning_rate": 9.781608191587166e-05,
+      "loss": 2.2233,
+      "step": 422
+    },
+    {
+      "epoch": 0.10630143871332537,
+      "grad_norm": 1.7654945850372314,
+      "learning_rate": 9.78043799517494e-05,
+      "loss": 2.2441,
+      "step": 423
+    },
+    {
+      "epoch": 0.10655274235094553,
+      "grad_norm": 0.272491455078125,
+      "learning_rate": 9.779264742410829e-05,
+      "loss": 2.4267,
+      "step": 424
+    },
+    {
+      "epoch": 0.10680404598856569,
+      "grad_norm": 0.4860396683216095,
+      "learning_rate": 9.778088434044945e-05,
+      "loss": 2.3139,
+      "step": 425
+    },
+    {
+      "epoch": 0.10705534962618583,
+      "grad_norm": 0.5187298059463501,
+      "learning_rate": 9.77690907082936e-05,
+      "loss": 2.0393,
+      "step": 426
+    },
+    {
+      "epoch": 0.10730665326380599,
+      "grad_norm": 0.25615808367729187,
+      "learning_rate": 9.775726653518091e-05,
+      "loss": 2.4811,
+      "step": 427
+    },
+    {
+      "epoch": 0.10755795690142615,
+      "grad_norm": 0.26843705773353577,
+      "learning_rate": 9.774541182867112e-05,
+      "loss": 2.0025,
+      "step": 428
+    },
+    {
+      "epoch": 0.1078092605390463,
+      "grad_norm": 0.42573508620262146,
+      "learning_rate": 9.773352659634348e-05,
+      "loss": 2.3841,
+      "step": 429
+    },
+    {
+      "epoch": 0.10806056417666646,
+      "grad_norm": 0.34817081689834595,
+      "learning_rate": 9.772161084579679e-05,
+      "loss": 2.4695,
+      "step": 430
+    },
+    {
+      "epoch": 0.10831186781428662,
+      "grad_norm": 0.3741927742958069,
+      "learning_rate": 9.770966458464927e-05,
+      "loss": 2.099,
+      "step": 431
+    },
+    {
+      "epoch": 0.10856317145190676,
+      "grad_norm": 0.4013387858867645,
+      "learning_rate": 9.769768782053879e-05,
+      "loss": 2.2978,
+      "step": 432
+    },
+    {
+      "epoch": 0.10881447508952692,
+      "grad_norm": 0.4488285779953003,
+      "learning_rate": 9.768568056112258e-05,
+      "loss": 2.064,
+      "step": 433
+    },
+    {
+      "epoch": 0.10906577872714708,
+      "grad_norm": 0.43745100498199463,
+      "learning_rate": 9.767364281407745e-05,
+      "loss": 2.0517,
+      "step": 434
+    },
+    {
+      "epoch": 0.10931708236476723,
+      "grad_norm": 0.2905375361442566,
+      "learning_rate": 9.766157458709967e-05,
+      "loss": 2.43,
+      "step": 435
+    },
+    {
+      "epoch": 0.10956838600238739,
+      "grad_norm": 0.3817865252494812,
+      "learning_rate": 9.764947588790502e-05,
+      "loss": 2.439,
+      "step": 436
+    },
+    {
+      "epoch": 0.10981968964000755,
+      "grad_norm": 0.44527363777160645,
+      "learning_rate": 9.763734672422876e-05,
+      "loss": 2.4307,
+      "step": 437
+    },
+    {
+      "epoch": 0.11007099327762769,
+      "grad_norm": 0.5595135688781738,
+      "learning_rate": 9.76251871038256e-05,
+      "loss": 2.1648,
+      "step": 438
+    },
+    {
+      "epoch": 0.11032229691524785,
+      "grad_norm": 0.4044279456138611,
+      "learning_rate": 9.761299703446973e-05,
+      "loss": 2.3435,
+      "step": 439
+    },
+    {
+      "epoch": 0.11057360055286801,
+      "grad_norm": 0.1558169573545456,
+      "learning_rate": 9.760077652395483e-05,
+      "loss": 0.9334,
+      "step": 440
+    },
+    {
+      "epoch": 0.11082490419048815,
+      "grad_norm": 0.42228519916534424,
+      "learning_rate": 9.758852558009404e-05,
+      "loss": 2.4764,
+      "step": 441
+    },
+    {
+      "epoch": 0.11107620782810831,
+      "grad_norm": 0.18882694840431213,
+      "learning_rate": 9.757624421071993e-05,
+      "loss": 1.3726,
+      "step": 442
+    },
+    {
+      "epoch": 0.11132751146572846,
+      "grad_norm": 0.37905827164649963,
+      "learning_rate": 9.756393242368453e-05,
+      "loss": 1.8814,
+      "step": 443
+    },
+    {
+      "epoch": 0.11157881510334862,
+      "grad_norm": 0.347260445356369,
+      "learning_rate": 9.755159022685936e-05,
+      "loss": 2.6184,
+      "step": 444
+    },
+    {
+      "epoch": 0.11183011874096878,
+      "grad_norm": 0.4600488543510437,
+      "learning_rate": 9.753921762813534e-05,
+      "loss": 2.2642,
+      "step": 445
+    },
+    {
+      "epoch": 0.11208142237858892,
+      "grad_norm": 0.3841269612312317,
+      "learning_rate": 9.75268146354228e-05,
+      "loss": 2.2473,
+      "step": 446
+    },
+    {
+      "epoch": 0.11233272601620908,
+      "grad_norm": 0.4765447676181793,
+      "learning_rate": 9.751438125665158e-05,
+      "loss": 2.3474,
+      "step": 447
+    },
+    {
+      "epoch": 0.11258402965382924,
+      "grad_norm": 0.3766055405139923,
+      "learning_rate": 9.750191749977089e-05,
+      "loss": 1.809,
+      "step": 448
+    },
+    {
+      "epoch": 0.11283533329144939,
+      "grad_norm": 0.3452647626399994,
+      "learning_rate": 9.748942337274938e-05,
+      "loss": 2.5878,
+      "step": 449
+    },
+    {
+      "epoch": 0.11308663692906955,
+      "grad_norm": 0.21866516768932343,
+      "learning_rate": 9.747689888357509e-05,
+      "loss": 0.992,
+      "step": 450
+    },
+    {
+      "epoch": 0.11333794056668971,
+      "grad_norm": 0.36812421679496765,
+      "learning_rate": 9.746434404025555e-05,
+      "loss": 2.0005,
+      "step": 451
+    },
+    {
+      "epoch": 0.11358924420430985,
+      "grad_norm": 0.44335392117500305,
+      "learning_rate": 9.74517588508176e-05,
+      "loss": 2.4726,
+      "step": 452
+    },
+    {
+      "epoch": 0.11384054784193001,
+      "grad_norm": 0.448779433965683,
+      "learning_rate": 9.743914332330754e-05,
+      "loss": 2.2657,
+      "step": 453
+    },
+    {
+      "epoch": 0.11409185147955017,
+      "grad_norm": 0.5500572323799133,
+      "learning_rate": 9.742649746579105e-05,
+      "loss": 2.4726,
+      "step": 454
+    },
+    {
+      "epoch": 0.11434315511717032,
+      "grad_norm": 0.47495442628860474,
+      "learning_rate": 9.741382128635321e-05,
+      "loss": 2.4044,
+      "step": 455
+    },
+    {
+      "epoch": 0.11459445875479048,
+      "grad_norm": 0.4240530729293823,
+      "learning_rate": 9.740111479309847e-05,
+      "loss": 2.6976,
+      "step": 456
+    },
+    {
+      "epoch": 0.11484576239241064,
+      "grad_norm": 0.48315781354904175,
+      "learning_rate": 9.738837799415067e-05,
+      "loss": 2.4035,
+      "step": 457
+    },
+    {
+      "epoch": 0.11509706603003078,
+      "grad_norm": 0.28931689262390137,
+      "learning_rate": 9.737561089765303e-05,
+      "loss": 1.9762,
+      "step": 458
+    },
+    {
+      "epoch": 0.11534836966765094,
+      "grad_norm": 0.2892288267612457,
+      "learning_rate": 9.736281351176813e-05,
+      "loss": 2.1718,
+      "step": 459
+    },
+    {
+      "epoch": 0.1155996733052711,
+      "grad_norm": 0.4304305911064148,
+      "learning_rate": 9.734998584467794e-05,
+      "loss": 2.2799,
+      "step": 460
+    },
+    {
+      "epoch": 0.11585097694289125,
+      "grad_norm": 0.2887556552886963,
+      "learning_rate": 9.733712790458375e-05,
+      "loss": 2.4617,
+      "step": 461
+    },
+    {
+      "epoch": 0.1161022805805114,
+      "grad_norm": 0.37903836369514465,
+      "learning_rate": 9.732423969970626e-05,
+      "loss": 2.4154,
+      "step": 462
+    },
+    {
+      "epoch": 0.11635358421813156,
+      "grad_norm": 0.34230297803878784,
+      "learning_rate": 9.731132123828543e-05,
+      "loss": 1.9664,
+      "step": 463
+    },
+    {
+      "epoch": 0.11660488785575171,
+      "grad_norm": 0.4099084138870239,
+      "learning_rate": 9.729837252858067e-05,
+      "loss": 2.473,
+      "step": 464
+    },
+    {
+      "epoch": 0.11685619149337187,
+      "grad_norm": 0.4079136252403259,
+      "learning_rate": 9.728539357887068e-05,
+      "loss": 1.8914,
+      "step": 465
+    },
+    {
+      "epoch": 0.11710749513099201,
+      "grad_norm": 0.2663559317588806,
+      "learning_rate": 9.727238439745346e-05,
+      "loss": 2.1168,
+      "step": 466
+    },
+    {
+      "epoch": 0.11735879876861217,
+      "grad_norm": 0.35732075572013855,
+      "learning_rate": 9.72593449926464e-05,
+      "loss": 2.3938,
+      "step": 467
+    },
+    {
+      "epoch": 0.11761010240623233,
+      "grad_norm": 0.26939693093299866,
+      "learning_rate": 9.724627537278616e-05,
+      "loss": 1.9446,
+      "step": 468
+    },
+    {
+      "epoch": 0.11786140604385248,
+      "grad_norm": 0.3639311194419861,
+      "learning_rate": 9.72331755462288e-05,
+      "loss": 2.8929,
+      "step": 469
+    },
+    {
+      "epoch": 0.11811270968147264,
+      "grad_norm": 0.26295900344848633,
+      "learning_rate": 9.722004552134956e-05,
+      "loss": 2.3815,
+      "step": 470
+    },
+    {
+      "epoch": 0.1183640133190928,
+      "grad_norm": 0.4366108179092407,
+      "learning_rate": 9.720688530654311e-05,
+      "loss": 2.1099,
+      "step": 471
+    },
+    {
+      "epoch": 0.11861531695671294,
+      "grad_norm": 0.5005189776420593,
+      "learning_rate": 9.719369491022339e-05,
+      "loss": 2.7562,
+      "step": 472
+    },
+    {
+      "epoch": 0.1188666205943331,
+      "grad_norm": 0.3057880401611328,
+      "learning_rate": 9.718047434082357e-05,
+      "loss": 2.5181,
+      "step": 473
+    },
+    {
+      "epoch": 0.11911792423195326,
+      "grad_norm": 0.2341679483652115,
+      "learning_rate": 9.716722360679619e-05,
+      "loss": 1.2066,
+      "step": 474
+    },
+    {
+      "epoch": 0.11936922786957341,
+      "grad_norm": 0.27327990531921387,
+      "learning_rate": 9.715394271661306e-05,
+      "loss": 2.3469,
+      "step": 475
+    },
+    {
+      "epoch": 0.11962053150719357,
+      "grad_norm": 0.39172980189323425,
+      "learning_rate": 9.714063167876527e-05,
+      "loss": 2.2347,
+      "step": 476
+    },
+    {
+      "epoch": 0.11987183514481373,
+      "grad_norm": 0.49665653705596924,
+      "learning_rate": 9.71272905017631e-05,
+      "loss": 1.954,
+      "step": 477
+    },
+    {
+      "epoch": 0.12012313878243387,
+      "grad_norm": 0.2704184949398041,
+      "learning_rate": 9.711391919413626e-05,
+      "loss": 2.3236,
+      "step": 478
+    },
+    {
+      "epoch": 0.12037444242005403,
+      "grad_norm": 0.17571642994880676,
+      "learning_rate": 9.710051776443358e-05,
+      "loss": 1.277,
+      "step": 479
+    },
+    {
+      "epoch": 0.12062574605767419,
+      "grad_norm": 0.2642328143119812,
+      "learning_rate": 9.708708622122322e-05,
+      "loss": 2.4537,
+      "step": 480
+    },
+    {
+      "epoch": 0.12087704969529434,
+      "grad_norm": 0.21277543902397156,
+      "learning_rate": 9.707362457309261e-05,
+      "loss": 1.3385,
+      "step": 481
+    },
+    {
+      "epoch": 0.1211283533329145,
+      "grad_norm": 0.3256551921367645,
+      "learning_rate": 9.706013282864834e-05,
+      "loss": 2.1567,
+      "step": 482
+    },
+    {
+      "epoch": 0.12137965697053466,
+      "grad_norm": 0.4194876551628113,
+      "learning_rate": 9.704661099651633e-05,
+      "loss": 2.2617,
+      "step": 483
+    },
+    {
+      "epoch": 0.1216309606081548,
+      "grad_norm": 0.13989388942718506,
+      "learning_rate": 9.70330590853417e-05,
+      "loss": 0.8074,
+      "step": 484
+    },
+    {
+      "epoch": 0.12188226424577496,
+      "grad_norm": 0.2916125953197479,
+      "learning_rate": 9.701947710378881e-05,
+      "loss": 2.0217,
+      "step": 485
+    },
+    {
+      "epoch": 0.1221335678833951,
+      "grad_norm": 0.17038998007774353,
+      "learning_rate": 9.700586506054121e-05,
+      "loss": 0.8153,
+      "step": 486
+    },
+    {
+      "epoch": 0.12238487152101526,
+      "grad_norm": 0.39059990644454956,
+      "learning_rate": 9.699222296430172e-05,
+      "loss": 2.161,
+      "step": 487
+    },
+    {
+      "epoch": 0.12263617515863542,
+      "grad_norm": 0.5335647463798523,
+      "learning_rate": 9.697855082379239e-05,
+      "loss": 2.2604,
+      "step": 488
+    },
+    {
+      "epoch": 0.12288747879625557,
+      "grad_norm": 0.4468785524368286,
+      "learning_rate": 9.696484864775437e-05,
+      "loss": 2.3022,
+      "step": 489
+    },
+    {
+      "epoch": 0.12313878243387573,
+      "grad_norm": 0.38395336270332336,
+      "learning_rate": 9.695111644494814e-05,
+      "loss": 2.2467,
+      "step": 490
+    },
+    {
+      "epoch": 0.12339008607149589,
+      "grad_norm": 0.35352465510368347,
+      "learning_rate": 9.693735422415332e-05,
+      "loss": 2.7816,
+      "step": 491
+    },
+    {
+      "epoch": 0.12364138970911603,
+      "grad_norm": 0.5462369322776794,
+      "learning_rate": 9.692356199416868e-05,
+      "loss": 2.2723,
+      "step": 492
+    },
+    {
+      "epoch": 0.1238926933467362,
+      "grad_norm": 0.24535368382930756,
+      "learning_rate": 9.690973976381228e-05,
+      "loss": 2.0469,
+      "step": 493
+    },
+    {
+      "epoch": 0.12414399698435635,
+      "grad_norm": 0.3857629597187042,
+      "learning_rate": 9.689588754192126e-05,
+      "loss": 2.1396,
+      "step": 494
+    },
+    {
+      "epoch": 0.1243953006219765,
+      "grad_norm": 0.4369061291217804,
+      "learning_rate": 9.688200533735199e-05,
+      "loss": 2.3099,
+      "step": 495
+    },
+    {
+      "epoch": 0.12464660425959666,
+      "grad_norm": 0.22359801828861237,
+      "learning_rate": 9.686809315898e-05,
+      "loss": 2.113,
+      "step": 496
+    },
+    {
+      "epoch": 0.12489790789721682,
+      "grad_norm": 0.3624935746192932,
+      "learning_rate": 9.685415101569999e-05,
+      "loss": 2.2168,
+      "step": 497
+    },
+    {
+      "epoch": 0.12514921153483696,
+      "grad_norm": 0.4480370581150055,
+      "learning_rate": 9.684017891642578e-05,
+      "loss": 2.5805,
+      "step": 498
+    },
+    {
+      "epoch": 0.12540051517245712,
+      "grad_norm": 0.47670865058898926,
+      "learning_rate": 9.682617687009039e-05,
+      "loss": 1.5993,
+      "step": 499
+    },
+    {
+      "epoch": 0.12565181881007728,
+      "grad_norm": 0.37690502405166626,
+      "learning_rate": 9.681214488564596e-05,
+      "loss": 2.125,
+      "step": 500
+    },
+    {
+      "epoch": 0.12590312244769744,
+      "grad_norm": 0.29597729444503784,
+      "learning_rate": 9.679808297206377e-05,
+      "loss": 2.6068,
+      "step": 501
+    },
+    {
+      "epoch": 0.12615442608531757,
+      "grad_norm": 0.33130696415901184,
+      "learning_rate": 9.678399113833425e-05,
+      "loss": 2.2649,
+      "step": 502
+    },
+    {
+      "epoch": 0.12640572972293773,
+      "grad_norm": 0.4030790627002716,
+      "learning_rate": 9.676986939346696e-05,
+      "loss": 1.8589,
+      "step": 503
+    },
+    {
+      "epoch": 0.1266570333605579,
+      "grad_norm": 0.37136292457580566,
+      "learning_rate": 9.675571774649057e-05,
+      "loss": 2.1863,
+      "step": 504
+    },
+    {
+      "epoch": 0.12690833699817805,
+      "grad_norm": 0.439748615026474,
+      "learning_rate": 9.674153620645287e-05,
+      "loss": 2.6519,
+      "step": 505
+    },
+    {
+      "epoch": 0.1271596406357982,
+      "grad_norm": 0.3839961588382721,
+      "learning_rate": 9.672732478242075e-05,
+      "loss": 2.4758,
+      "step": 506
+    },
+    {
+      "epoch": 0.12741094427341837,
+      "grad_norm": 0.2992021441459656,
+      "learning_rate": 9.671308348348025e-05,
+      "loss": 2.4348,
+      "step": 507
+    },
+    {
+      "epoch": 0.1276622479110385,
+      "grad_norm": 0.3640328645706177,
+      "learning_rate": 9.669881231873646e-05,
+      "loss": 2.6048,
+      "step": 508
+    },
+    {
+      "epoch": 0.12791355154865866,
+      "grad_norm": 0.27911099791526794,
+      "learning_rate": 9.66845112973136e-05,
+      "loss": 2.4449,
+      "step": 509
+    },
+    {
+      "epoch": 0.12816485518627882,
+      "grad_norm": 0.4001411199569702,
+      "learning_rate": 9.667018042835496e-05,
+      "loss": 3.0271,
+      "step": 510
+    },
+    {
+      "epoch": 0.12841615882389898,
+      "grad_norm": 0.4469778537750244,
+      "learning_rate": 9.665581972102291e-05,
+      "loss": 2.192,
+      "step": 511
+    },
+    {
+      "epoch": 0.12866746246151914,
+      "grad_norm": 0.4696493446826935,
+      "learning_rate": 9.66414291844989e-05,
+      "loss": 2.0927,
+      "step": 512
+    },
+    {
+      "epoch": 0.1289187660991393,
+      "grad_norm": 0.485324889421463,
+      "learning_rate": 9.662700882798348e-05,
+      "loss": 1.8437,
+      "step": 513
+    },
+    {
+      "epoch": 0.12917006973675943,
+      "grad_norm": 0.2740216851234436,
+      "learning_rate": 9.661255866069622e-05,
+      "loss": 2.0895,
+      "step": 514
+    },
+    {
+      "epoch": 0.1294213733743796,
+      "grad_norm": 0.48760735988616943,
+      "learning_rate": 9.659807869187578e-05,
+      "loss": 2.3416,
+      "step": 515
+    },
+    {
+      "epoch": 0.12967267701199975,
+      "grad_norm": 0.14928022027015686,
+      "learning_rate": 9.658356893077987e-05,
+      "loss": 1.1877,
+      "step": 516
+    },
+    {
+      "epoch": 0.1299239806496199,
+      "grad_norm": 0.46163851022720337,
+      "learning_rate": 9.656902938668524e-05,
+      "loss": 2.6305,
+      "step": 517
+    },
+    {
+      "epoch": 0.13017528428724007,
+      "grad_norm": 0.30622944235801697,
+      "learning_rate": 9.655446006888766e-05,
+      "loss": 1.5633,
+      "step": 518
+    },
+    {
+      "epoch": 0.1304265879248602,
+      "grad_norm": 0.30690157413482666,
+      "learning_rate": 9.653986098670198e-05,
+      "loss": 2.5689,
+      "step": 519
+    },
+    {
+      "epoch": 0.13067789156248036,
+      "grad_norm": 0.3105219602584839,
+      "learning_rate": 9.652523214946205e-05,
+      "loss": 2.564,
+      "step": 520
+    },
+    {
+      "epoch": 0.13092919520010052,
+      "grad_norm": 0.2955935299396515,
+      "learning_rate": 9.651057356652077e-05,
+      "loss": 2.5073,
+      "step": 521
+    },
+    {
+      "epoch": 0.13118049883772068,
+      "grad_norm": 0.19297459721565247,
+      "learning_rate": 9.649588524725002e-05,
+      "loss": 1.8412,
+      "step": 522
+    },
+    {
+      "epoch": 0.13143180247534084,
+      "grad_norm": 0.3480035066604614,
+      "learning_rate": 9.64811672010407e-05,
+      "loss": 2.5528,
+      "step": 523
+    },
+    {
+      "epoch": 0.131683106112961,
+      "grad_norm": 0.15509484708309174,
+      "learning_rate": 9.646641943730277e-05,
+      "loss": 0.7748,
+      "step": 524
+    },
+    {
+      "epoch": 0.13193440975058113,
+      "grad_norm": 0.33523187041282654,
+      "learning_rate": 9.645164196546512e-05,
+      "loss": 1.939,
+      "step": 525
+    },
+    {
+      "epoch": 0.1321857133882013,
+      "grad_norm": 0.3940199613571167,
+      "learning_rate": 9.643683479497567e-05,
+      "loss": 2.3232,
+      "step": 526
+    },
+    {
+      "epoch": 0.13243701702582145,
+      "grad_norm": 0.3470746576786041,
+      "learning_rate": 9.64219979353013e-05,
+      "loss": 1.933,
+      "step": 527
+    },
+    {
+      "epoch": 0.1326883206634416,
+      "grad_norm": 0.3262689709663391,
+      "learning_rate": 9.640713139592792e-05,
+      "loss": 2.3454,
+      "step": 528
+    },
+    {
+      "epoch": 0.13293962430106177,
+      "grad_norm": 0.5404649972915649,
+      "learning_rate": 9.639223518636036e-05,
+      "loss": 2.4712,
+      "step": 529
+    },
+    {
+      "epoch": 0.13319092793868192,
+      "grad_norm": 0.30311527848243713,
+      "learning_rate": 9.637730931612245e-05,
+      "loss": 2.0744,
+      "step": 530
+    },
+    {
+      "epoch": 0.13344223157630206,
+      "grad_norm": 0.9992802739143372,
+      "learning_rate": 9.6362353794757e-05,
+      "loss": 2.1378,
+      "step": 531
+    },
+    {
+      "epoch": 0.13369353521392222,
+      "grad_norm": 0.2992432415485382,
+      "learning_rate": 9.634736863182574e-05,
+      "loss": 2.4675,
+      "step": 532
+    },
+    {
+      "epoch": 0.13394483885154237,
+      "grad_norm": 0.3760550618171692,
+      "learning_rate": 9.633235383690937e-05,
+      "loss": 2.3762,
+      "step": 533
+    },
+    {
+      "epoch": 0.13419614248916253,
+      "grad_norm": 0.3664836287498474,
+      "learning_rate": 9.631730941960752e-05,
+      "loss": 2.2417,
+      "step": 534
+    },
+    {
+      "epoch": 0.1344474461267827,
+      "grad_norm": 0.4195888042449951,
+      "learning_rate": 9.630223538953881e-05,
+      "loss": 1.9261,
+      "step": 535
+    },
+    {
+      "epoch": 0.13469874976440285,
+      "grad_norm": 0.39428257942199707,
+      "learning_rate": 9.628713175634072e-05,
+      "loss": 2.2189,
+      "step": 536
+    },
+    {
+      "epoch": 0.13495005340202298,
+      "grad_norm": 0.3145690858364105,
+      "learning_rate": 9.627199852966969e-05,
+      "loss": 2.1252,
+      "step": 537
+    },
+    {
+      "epoch": 0.13520135703964314,
+      "grad_norm": 0.41197821497917175,
+      "learning_rate": 9.625683571920108e-05,
+      "loss": 2.0061,
+      "step": 538
+    },
+    {
+      "epoch": 0.1354526606772633,
+      "grad_norm": 0.4640690088272095,
+      "learning_rate": 9.62416433346292e-05,
+      "loss": 1.9037,
+      "step": 539
+    },
+    {
+      "epoch": 0.13570396431488346,
+      "grad_norm": 0.39173510670661926,
+      "learning_rate": 9.62264213856672e-05,
+      "loss": 2.1323,
+      "step": 540
+    },
+    {
+      "epoch": 0.13595526795250362,
+      "grad_norm": 0.33818063139915466,
+      "learning_rate": 9.62111698820472e-05,
+      "loss": 2.4662,
+      "step": 541
+    },
+    {
+      "epoch": 0.13620657159012375,
+      "grad_norm": 0.3888367712497711,
+      "learning_rate": 9.619588883352011e-05,
+      "loss": 1.987,
+      "step": 542
+    },
+    {
+      "epoch": 0.1364578752277439,
+      "grad_norm": 0.49469324946403503,
+      "learning_rate": 9.61805782498559e-05,
+      "loss": 2.075,
+      "step": 543
+    },
+    {
+      "epoch": 0.13670917886536407,
+      "grad_norm": 0.3214045464992523,
+      "learning_rate": 9.616523814084324e-05,
+      "loss": 2.6687,
+      "step": 544
+    },
+    {
+      "epoch": 0.13696048250298423,
+      "grad_norm": 0.3943631649017334,
+      "learning_rate": 9.61498685162898e-05,
+      "loss": 1.9817,
+      "step": 545
+    },
+    {
+      "epoch": 0.1372117861406044,
+      "grad_norm": 0.46393269300460815,
+      "learning_rate": 9.613446938602209e-05,
+      "loss": 2.2147,
+      "step": 546
+    },
+    {
+      "epoch": 0.13746308977822455,
+      "grad_norm": 0.3647920489311218,
+      "learning_rate": 9.611904075988544e-05,
+      "loss": 1.9163,
+      "step": 547
+    },
+    {
+      "epoch": 0.13771439341584468,
+      "grad_norm": 0.3861480951309204,
+      "learning_rate": 9.610358264774411e-05,
+      "loss": 2.7924,
+      "step": 548
+    },
+    {
+      "epoch": 0.13796569705346484,
+      "grad_norm": 0.4985499083995819,
+      "learning_rate": 9.608809505948114e-05,
+      "loss": 2.0298,
+      "step": 549
+    },
+    {
+      "epoch": 0.138217000691085,
+      "grad_norm": 0.4473382532596588,
+      "learning_rate": 9.607257800499849e-05,
+      "loss": 2.0965,
+      "step": 550
+    },
+    {
+      "epoch": 0.13846830432870516,
+      "grad_norm": 0.4468131959438324,
+      "learning_rate": 9.60570314942169e-05,
+      "loss": 2.311,
+      "step": 551
+    },
+    {
+      "epoch": 0.13871960796632532,
+      "grad_norm": 0.1555897295475006,
+      "learning_rate": 9.604145553707595e-05,
+      "loss": 0.8849,
+      "step": 552
+    },
+    {
+      "epoch": 0.13897091160394548,
+      "grad_norm": 0.17197562754154205,
+      "learning_rate": 9.602585014353409e-05,
+      "loss": 1.0516,
+      "step": 553
+    },
+    {
+      "epoch": 0.1392222152415656,
+      "grad_norm": 0.22020426392555237,
+      "learning_rate": 9.601021532356854e-05,
+      "loss": 1.9116,
+      "step": 554
+    },
+    {
+      "epoch": 0.13947351887918577,
+      "grad_norm": 0.33761653304100037,
+      "learning_rate": 9.599455108717535e-05,
+      "loss": 2.3363,
+      "step": 555
+    },
+    {
+      "epoch": 0.13972482251680593,
+      "grad_norm": 0.5171981453895569,
+      "learning_rate": 9.59788574443694e-05,
+      "loss": 2.0701,
+      "step": 556
+    },
+    {
+      "epoch": 0.1399761261544261,
+      "grad_norm": 0.22373630106449127,
+      "learning_rate": 9.596313440518432e-05,
+      "loss": 1.0946,
+      "step": 557
+    },
+    {
+      "epoch": 0.14022742979204625,
+      "grad_norm": 0.5129598379135132,
+      "learning_rate": 9.594738197967259e-05,
+      "loss": 2.3418,
+      "step": 558
+    },
+    {
+      "epoch": 0.1404787334296664,
+      "grad_norm": 0.4349968433380127,
+      "learning_rate": 9.593160017790546e-05,
+      "loss": 2.0397,
+      "step": 559
+    },
+    {
+      "epoch": 0.14073003706728654,
+      "grad_norm": 0.39511239528656006,
+      "learning_rate": 9.591578900997292e-05,
+      "loss": 2.2823,
+      "step": 560
+    },
+    {
+      "epoch": 0.1409813407049067,
+      "grad_norm": 0.25913137197494507,
+      "learning_rate": 9.58999484859838e-05,
+      "loss": 2.3511,
+      "step": 561
+    },
+    {
+      "epoch": 0.14123264434252686,
+      "grad_norm": 0.19200782477855682,
+      "learning_rate": 9.588407861606566e-05,
+      "loss": 1.6859,
+      "step": 562
+    },
+    {
+      "epoch": 0.14148394798014702,
+      "grad_norm": 0.30504798889160156,
+      "learning_rate": 9.586817941036483e-05,
+      "loss": 2.1132,
+      "step": 563
+    },
+    {
+      "epoch": 0.14173525161776718,
+      "grad_norm": 0.46058428287506104,
+      "learning_rate": 9.585225087904641e-05,
+      "loss": 2.4114,
+      "step": 564
+    },
+    {
+      "epoch": 0.1419865552553873,
+      "grad_norm": 0.2597549557685852,
+      "learning_rate": 9.583629303229423e-05,
+      "loss": 2.4956,
+      "step": 565
+    },
+    {
+      "epoch": 0.14223785889300747,
+      "grad_norm": 0.35064586997032166,
+      "learning_rate": 9.582030588031084e-05,
+      "loss": 2.4188,
+      "step": 566
+    },
+    {
+      "epoch": 0.14248916253062763,
+      "grad_norm": 0.21857944130897522,
+      "learning_rate": 9.580428943331758e-05,
+      "loss": 1.4052,
+      "step": 567
+    },
+    {
+      "epoch": 0.1427404661682478,
+      "grad_norm": 0.4945662021636963,
+      "learning_rate": 9.578824370155451e-05,
+      "loss": 2.6965,
+      "step": 568
+    },
+    {
+      "epoch": 0.14299176980586795,
+      "grad_norm": 0.37999972701072693,
+      "learning_rate": 9.577216869528038e-05,
+      "loss": 2.5524,
+      "step": 569
+    },
+    {
+      "epoch": 0.1432430734434881,
+      "grad_norm": 0.47479113936424255,
+      "learning_rate": 9.575606442477267e-05,
+      "loss": 2.166,
+      "step": 570
+    },
+    {
+      "epoch": 0.14349437708110824,
+      "grad_norm": 0.42284709215164185,
+      "learning_rate": 9.573993090032758e-05,
+      "loss": 1.7279,
+      "step": 571
+    },
+    {
+      "epoch": 0.1437456807187284,
+      "grad_norm": 0.412218302488327,
+      "learning_rate": 9.572376813225999e-05,
+      "loss": 2.1049,
+      "step": 572
+    },
+    {
+      "epoch": 0.14399698435634856,
+      "grad_norm": 0.6507567167282104,
+      "learning_rate": 9.570757613090353e-05,
+      "loss": 2.5453,
+      "step": 573
+    },
+    {
+      "epoch": 0.14424828799396872,
+      "grad_norm": 0.31104111671447754,
+      "learning_rate": 9.569135490661046e-05,
+      "loss": 2.6578,
+      "step": 574
+    },
+    {
+      "epoch": 0.14449959163158888,
+      "grad_norm": 0.3139590620994568,
+      "learning_rate": 9.567510446975176e-05,
+      "loss": 2.5413,
+      "step": 575
+    },
+    {
+      "epoch": 0.14475089526920903,
+      "grad_norm": 0.5592718720436096,
+      "learning_rate": 9.565882483071706e-05,
+      "loss": 2.3341,
+      "step": 576
+    },
+    {
+      "epoch": 0.14500219890682917,
+      "grad_norm": 0.4113386869430542,
+      "learning_rate": 9.564251599991467e-05,
+      "loss": 2.7414,
+      "step": 577
+    },
+    {
+      "epoch": 0.14525350254444933,
+      "grad_norm": 0.2924419045448303,
+      "learning_rate": 9.56261779877716e-05,
+      "loss": 1.882,
+      "step": 578
+    },
+    {
+      "epoch": 0.14550480618206948,
+      "grad_norm": 0.3123188316822052,
+      "learning_rate": 9.560981080473346e-05,
+      "loss": 2.2884,
+      "step": 579
+    },
+    {
+      "epoch": 0.14575610981968964,
+      "grad_norm": 0.3279459476470947,
+      "learning_rate": 9.559341446126455e-05,
+      "loss": 1.8245,
+      "step": 580
+    },
+    {
+      "epoch": 0.1460074134573098,
+      "grad_norm": 0.2936881184577942,
+      "learning_rate": 9.55769889678478e-05,
+      "loss": 2.2238,
+      "step": 581
+    },
+    {
+      "epoch": 0.14625871709492994,
+      "grad_norm": 0.3411659598350525,
+      "learning_rate": 9.556053433498475e-05,
+      "loss": 2.1781,
+      "step": 582
+    },
+    {
+      "epoch": 0.1465100207325501,
+      "grad_norm": 0.42901405692100525,
+      "learning_rate": 9.554405057319565e-05,
+      "loss": 2.5198,
+      "step": 583
+    },
+    {
+      "epoch": 0.14676132437017025,
+      "grad_norm": 0.37800028920173645,
+      "learning_rate": 9.552753769301925e-05,
+      "loss": 2.009,
+      "step": 584
+    },
+    {
+      "epoch": 0.1470126280077904,
+      "grad_norm": 0.35843050479888916,
+      "learning_rate": 9.551099570501305e-05,
+      "loss": 2.0567,
+      "step": 585
+    },
+    {
+      "epoch": 0.14726393164541057,
+      "grad_norm": 0.4403095245361328,
+      "learning_rate": 9.549442461975306e-05,
+      "loss": 2.242,
+      "step": 586
+    },
+    {
+      "epoch": 0.14751523528303073,
+      "grad_norm": 0.3680509626865387,
+      "learning_rate": 9.547782444783393e-05,
+      "loss": 1.9327,
+      "step": 587
+    },
+    {
+      "epoch": 0.14776653892065086,
+      "grad_norm": 0.26541033387184143,
+      "learning_rate": 9.546119519986894e-05,
+      "loss": 2.5879,
+      "step": 588
+    },
+    {
+      "epoch": 0.14801784255827102,
+      "grad_norm": 0.3599735200405121,
+      "learning_rate": 9.544453688648989e-05,
+      "loss": 2.448,
+      "step": 589
+    },
+    {
+      "epoch": 0.14826914619589118,
+      "grad_norm": 0.46123915910720825,
+      "learning_rate": 9.542784951834721e-05,
+      "loss": 2.1933,
+      "step": 590
+    },
+    {
+      "epoch": 0.14852044983351134,
+      "grad_norm": 0.3718903064727783,
+      "learning_rate": 9.54111331061099e-05,
+      "loss": 2.0662,
+      "step": 591
+    },
+    {
+      "epoch": 0.1487717534711315,
+      "grad_norm": 0.5671700239181519,
+      "learning_rate": 9.539438766046554e-05,
+      "loss": 2.2881,
+      "step": 592
+    },
+    {
+      "epoch": 0.14902305710875166,
+      "grad_norm": 0.415791779756546,
+      "learning_rate": 9.537761319212021e-05,
+      "loss": 2.1724,
+      "step": 593
+    },
+    {
+      "epoch": 0.1492743607463718,
+      "grad_norm": 0.33377009630203247,
+      "learning_rate": 9.536080971179864e-05,
+      "loss": 2.2665,
+      "step": 594
+    },
+    {
+      "epoch": 0.14952566438399195,
+      "grad_norm": 0.48479607701301575,
+      "learning_rate": 9.534397723024402e-05,
+      "loss": 1.9262,
+      "step": 595
+    },
+    {
+      "epoch": 0.1497769680216121,
+      "grad_norm": 0.4698795974254608,
+      "learning_rate": 9.532711575821816e-05,
+      "loss": 2.3923,
+      "step": 596
+    },
+    {
+      "epoch": 0.15002827165923227,
+      "grad_norm": 0.45373061299324036,
+      "learning_rate": 9.531022530650135e-05,
+      "loss": 2.2584,
+      "step": 597
+    },
+    {
+      "epoch": 0.15027957529685243,
+      "grad_norm": 0.25719451904296875,
+      "learning_rate": 9.529330588589243e-05,
+      "loss": 2.1529,
+      "step": 598
+    },
+    {
+      "epoch": 0.1505308789344726,
+      "grad_norm": 0.27690425515174866,
+      "learning_rate": 9.527635750720875e-05,
+      "loss": 2.1076,
+      "step": 599
+    },
+    {
+      "epoch": 0.15078218257209272,
+      "grad_norm": 0.11603706330060959,
+      "learning_rate": 9.525938018128617e-05,
+      "loss": 0.5727,
+      "step": 600
+    },
+    {
+      "epoch": 0.15103348620971288,
+      "grad_norm": 0.39744696021080017,
+      "learning_rate": 9.524237391897909e-05,
+      "loss": 2.1513,
+      "step": 601
+    },
+    {
+      "epoch": 0.15128478984733304,
+      "grad_norm": 0.3867836594581604,
+      "learning_rate": 9.522533873116041e-05,
+      "loss": 2.3665,
+      "step": 602
+    },
+    {
+      "epoch": 0.1515360934849532,
+      "grad_norm": 0.16841183602809906,
+      "learning_rate": 9.520827462872144e-05,
+      "loss": 1.1614,
+      "step": 603
+    },
+    {
+      "epoch": 0.15178739712257336,
+      "grad_norm": 0.34245559573173523,
+      "learning_rate": 9.519118162257209e-05,
+      "loss": 2.2823,
+      "step": 604
+    },
+    {
+      "epoch": 0.1520387007601935,
+      "grad_norm": 0.47992077469825745,
+      "learning_rate": 9.517405972364067e-05,
+      "loss": 1.9997,
+      "step": 605
+    },
+    {
+      "epoch": 0.15229000439781365,
+      "grad_norm": 0.2888699471950531,
+      "learning_rate": 9.5156908942874e-05,
+      "loss": 2.2298,
+      "step": 606
+    },
+    {
+      "epoch": 0.1525413080354338,
+      "grad_norm": 0.1741029918193817,
+      "learning_rate": 9.513972929123737e-05,
+      "loss": 0.7344,
+      "step": 607
+    },
+    {
+      "epoch": 0.15279261167305397,
+      "grad_norm": 0.3610229790210724,
+      "learning_rate": 9.512252077971448e-05,
+      "loss": 2.0037,
+      "step": 608
+    },
+    {
+      "epoch": 0.15304391531067413,
+      "grad_norm": 0.30153217911720276,
+      "learning_rate": 9.510528341930756e-05,
+      "loss": 2.1396,
+      "step": 609
+    },
+    {
+      "epoch": 0.1532952189482943,
+      "grad_norm": 0.5049226880073547,
+      "learning_rate": 9.50880172210372e-05,
+      "loss": 2.1197,
+      "step": 610
+    },
+    {
+      "epoch": 0.15354652258591442,
+      "grad_norm": 0.27209773659706116,
+      "learning_rate": 9.507072219594249e-05,
+      "loss": 2.37,
+      "step": 611
+    },
+    {
+      "epoch": 0.15379782622353458,
+      "grad_norm": 0.33535251021385193,
+      "learning_rate": 9.505339835508091e-05,
+      "loss": 2.4393,
+      "step": 612
+    },
+    {
+      "epoch": 0.15404912986115474,
+      "grad_norm": 0.4262019991874695,
+      "learning_rate": 9.50360457095284e-05,
+      "loss": 2.0283,
+      "step": 613
+    },
+    {
+      "epoch": 0.1543004334987749,
+      "grad_norm": 0.5614545941352844,
+      "learning_rate": 9.50186642703793e-05,
+      "loss": 2.0785,
+      "step": 614
+    },
+    {
+      "epoch": 0.15455173713639506,
+      "grad_norm": 0.3979303240776062,
+      "learning_rate": 9.500125404874631e-05,
+      "loss": 2.6146,
+      "step": 615
+    },
+    {
+      "epoch": 0.15480304077401522,
+      "grad_norm": 0.39508217573165894,
+      "learning_rate": 9.498381505576064e-05,
+      "loss": 2.2959,
+      "step": 616
+    },
+    {
+      "epoch": 0.15505434441163535,
+      "grad_norm": 0.36526814103126526,
+      "learning_rate": 9.49663473025718e-05,
+      "loss": 2.3149,
+      "step": 617
+    },
+    {
+      "epoch": 0.1553056480492555,
+      "grad_norm": 0.6422840356826782,
+      "learning_rate": 9.494885080034774e-05,
+      "loss": 2.0929,
+      "step": 618
+    },
+    {
+      "epoch": 0.15555695168687567,
+      "grad_norm": 0.32328036427497864,
+      "learning_rate": 9.493132556027475e-05,
+      "loss": 2.2233,
+      "step": 619
+    },
+    {
+      "epoch": 0.15580825532449583,
+      "grad_norm": 0.25220245122909546,
+      "learning_rate": 9.491377159355752e-05,
+      "loss": 2.5095,
+      "step": 620
+    },
+    {
+      "epoch": 0.15605955896211599,
+      "grad_norm": 0.32473161816596985,
+      "learning_rate": 9.489618891141911e-05,
+      "loss": 2.6889,
+      "step": 621
+    },
+    {
+      "epoch": 0.15631086259973614,
+      "grad_norm": 0.4301775395870209,
+      "learning_rate": 9.487857752510093e-05,
+      "loss": 2.227,
+      "step": 622
+    },
+    {
+      "epoch": 0.15656216623735628,
+      "grad_norm": 0.4302082061767578,
+      "learning_rate": 9.486093744586271e-05,
+      "loss": 2.2468,
+      "step": 623
+    },
+    {
+      "epoch": 0.15681346987497644,
+      "grad_norm": 0.39020925760269165,
+      "learning_rate": 9.484326868498261e-05,
+      "loss": 2.3726,
+      "step": 624
+    },
+    {
+      "epoch": 0.1570647735125966,
+      "grad_norm": 0.3023484945297241,
+      "learning_rate": 9.482557125375704e-05,
+      "loss": 2.4235,
+      "step": 625
+    },
+    {
+      "epoch": 0.15731607715021675,
+      "grad_norm": 0.3555540442466736,
+      "learning_rate": 9.480784516350079e-05,
+      "loss": 2.4532,
+      "step": 626
+    },
+    {
+      "epoch": 0.1575673807878369,
+      "grad_norm": 0.3152056634426117,
+      "learning_rate": 9.479009042554694e-05,
+      "loss": 2.7172,
+      "step": 627
+    },
+    {
+      "epoch": 0.15781868442545705,
+      "grad_norm": 0.5882668495178223,
+      "learning_rate": 9.477230705124692e-05,
+      "loss": 2.5912,
+      "step": 628
+    },
+    {
+      "epoch": 0.1580699880630772,
+      "grad_norm": 0.36808767914772034,
+      "learning_rate": 9.475449505197043e-05,
+      "loss": 2.2798,
+      "step": 629
+    },
+    {
+      "epoch": 0.15832129170069736,
+      "grad_norm": 0.44441547989845276,
+      "learning_rate": 9.473665443910551e-05,
+      "loss": 1.9648,
+      "step": 630
+    },
+    {
+      "epoch": 0.15857259533831752,
+      "grad_norm": 0.40813305974006653,
+      "learning_rate": 9.471878522405849e-05,
+      "loss": 1.7885,
+      "step": 631
+    },
+    {
+      "epoch": 0.15882389897593768,
+      "grad_norm": 0.32029062509536743,
+      "learning_rate": 9.470088741825394e-05,
+      "loss": 2.6975,
+      "step": 632
+    },
+    {
+      "epoch": 0.15907520261355784,
+      "grad_norm": 0.270475834608078,
+      "learning_rate": 9.468296103313476e-05,
+      "loss": 0.8822,
+      "step": 633
+    },
+    {
+      "epoch": 0.15932650625117797,
+      "grad_norm": 0.3409525454044342,
+      "learning_rate": 9.46650060801621e-05,
+      "loss": 1.7725,
+      "step": 634
+    },
+    {
+      "epoch": 0.15957780988879813,
+      "grad_norm": 0.4669889807701111,
+      "learning_rate": 9.464702257081539e-05,
+      "loss": 2.379,
+      "step": 635
+    },
+    {
+      "epoch": 0.1598291135264183,
+      "grad_norm": 0.5293301939964294,
+      "learning_rate": 9.462901051659232e-05,
+      "loss": 2.3118,
+      "step": 636
+    },
+    {
+      "epoch": 0.16008041716403845,
+      "grad_norm": 0.38586729764938354,
+      "learning_rate": 9.461096992900879e-05,
+      "loss": 2.0312,
+      "step": 637
+    },
+    {
+      "epoch": 0.1603317208016586,
+      "grad_norm": 0.3819951117038727,
+      "learning_rate": 9.459290081959897e-05,
+      "loss": 1.973,
+      "step": 638
+    },
+    {
+      "epoch": 0.16058302443927877,
+      "grad_norm": 0.37507110834121704,
+      "learning_rate": 9.457480319991529e-05,
+      "loss": 2.3668,
+      "step": 639
+    },
+    {
+      "epoch": 0.1608343280768989,
+      "grad_norm": 0.3354203701019287,
+      "learning_rate": 9.455667708152836e-05,
+      "loss": 1.7392,
+      "step": 640
+    },
+    {
+      "epoch": 0.16108563171451906,
+      "grad_norm": 0.31567585468292236,
+      "learning_rate": 9.453852247602704e-05,
+      "loss": 2.2258,
+      "step": 641
+    },
+    {
+      "epoch": 0.16133693535213922,
+      "grad_norm": 0.35445067286491394,
+      "learning_rate": 9.452033939501839e-05,
+      "loss": 2.5792,
+      "step": 642
+    },
+    {
+      "epoch": 0.16158823898975938,
+      "grad_norm": 0.3172896206378937,
+      "learning_rate": 9.45021278501277e-05,
+      "loss": 2.4493,
+      "step": 643
+    },
+    {
+      "epoch": 0.16183954262737954,
+      "grad_norm": 0.4169299602508545,
+      "learning_rate": 9.448388785299842e-05,
+      "loss": 2.3125,
+      "step": 644
+    },
+    {
+      "epoch": 0.1620908462649997,
+      "grad_norm": 0.5329498648643494,
+      "learning_rate": 9.446561941529224e-05,
+      "loss": 2.1722,
+      "step": 645
+    },
+    {
+      "epoch": 0.16234214990261983,
+      "grad_norm": 0.21547356247901917,
+      "learning_rate": 9.444732254868898e-05,
+      "loss": 2.1758,
+      "step": 646
+    },
+    {
+      "epoch": 0.16259345354024,
+      "grad_norm": 0.29865655303001404,
+      "learning_rate": 9.442899726488665e-05,
+      "loss": 1.5479,
+      "step": 647
+    },
+    {
+      "epoch": 0.16284475717786015,
+      "grad_norm": 0.38035672903060913,
+      "learning_rate": 9.441064357560147e-05,
+      "loss": 2.8413,
+      "step": 648
+    },
+    {
+      "epoch": 0.1630960608154803,
+      "grad_norm": 0.6323506832122803,
+      "learning_rate": 9.439226149256779e-05,
+      "loss": 2.1557,
+      "step": 649
+    },
+    {
+      "epoch": 0.16334736445310047,
+      "grad_norm": 0.5052780508995056,
+      "learning_rate": 9.43738510275381e-05,
+      "loss": 1.7929,
+      "step": 650
+    },
+    {
+      "epoch": 0.1635986680907206,
+      "grad_norm": 0.3237319588661194,
+      "learning_rate": 9.435541219228303e-05,
+      "loss": 2.336,
+      "step": 651
+    },
+    {
+      "epoch": 0.16384997172834076,
+      "grad_norm": 0.18192382156848907,
+      "learning_rate": 9.433694499859141e-05,
+      "loss": 0.9273,
+      "step": 652
+    },
+    {
+      "epoch": 0.16410127536596092,
+      "grad_norm": 0.31093844771385193,
+      "learning_rate": 9.431844945827014e-05,
+      "loss": 2.633,
+      "step": 653
+    },
+    {
+      "epoch": 0.16435257900358108,
+      "grad_norm": 0.40673384070396423,
+      "learning_rate": 9.429992558314423e-05,
+      "loss": 1.8868,
+      "step": 654
+    },
+    {
+      "epoch": 0.16460388264120124,
+      "grad_norm": 0.30386146903038025,
+      "learning_rate": 9.428137338505687e-05,
+      "loss": 2.1055,
+      "step": 655
+    },
+    {
+      "epoch": 0.1648551862788214,
+      "grad_norm": 0.3704879581928253,
+      "learning_rate": 9.426279287586934e-05,
+      "loss": 1.9851,
+      "step": 656
+    },
+    {
+      "epoch": 0.16510648991644153,
+      "grad_norm": 0.17675453424453735,
+      "learning_rate": 9.424418406746098e-05,
+      "loss": 1.0456,
+      "step": 657
+    },
+    {
+      "epoch": 0.1653577935540617,
+      "grad_norm": 0.2338314950466156,
+      "learning_rate": 9.422554697172925e-05,
+      "loss": 1.8722,
+      "step": 658
+    },
+    {
+      "epoch": 0.16560909719168185,
+      "grad_norm": 0.2262151688337326,
+      "learning_rate": 9.420688160058972e-05,
+      "loss": 1.2118,
+      "step": 659
+    },
+    {
+      "epoch": 0.165860400829302,
+      "grad_norm": 0.32973191142082214,
+      "learning_rate": 9.418818796597597e-05,
+      "loss": 1.974,
+      "step": 660
+    },
+    {
+      "epoch": 0.16611170446692217,
+      "grad_norm": 0.3016055226325989,
+      "learning_rate": 9.416946607983975e-05,
+      "loss": 1.8877,
+      "step": 661
+    },
+    {
+      "epoch": 0.16636300810454233,
+      "grad_norm": 0.39721089601516724,
+      "learning_rate": 9.415071595415075e-05,
+      "loss": 2.1139,
+      "step": 662
+    },
+    {
+      "epoch": 0.16661431174216246,
+      "grad_norm": 0.3398868143558502,
+      "learning_rate": 9.413193760089682e-05,
+      "loss": 1.9081,
+      "step": 663
+    },
+    {
+      "epoch": 0.16686561537978262,
+      "grad_norm": 0.3406602144241333,
+      "learning_rate": 9.411313103208382e-05,
+      "loss": 2.1456,
+      "step": 664
+    },
+    {
+      "epoch": 0.16711691901740278,
+      "grad_norm": 0.4096097946166992,
+      "learning_rate": 9.409429625973563e-05,
+      "loss": 2.294,
+      "step": 665
+    },
+    {
+      "epoch": 0.16736822265502294,
+      "grad_norm": 0.2669360041618347,
+      "learning_rate": 9.407543329589418e-05,
+      "loss": 1.8453,
+      "step": 666
+    },
+    {
+      "epoch": 0.1676195262926431,
+      "grad_norm": 0.5418170690536499,
+      "learning_rate": 9.405654215261944e-05,
+      "loss": 2.3642,
+      "step": 667
+    },
+    {
+      "epoch": 0.16787082993026323,
+      "grad_norm": 0.32345104217529297,
+      "learning_rate": 9.403762284198936e-05,
+      "loss": 2.1211,
+      "step": 668
+    },
+    {
+      "epoch": 0.1681221335678834,
+      "grad_norm": 0.49764448404312134,
+      "learning_rate": 9.401867537609991e-05,
+      "loss": 1.9661,
+      "step": 669
+    },
+    {
+      "epoch": 0.16837343720550355,
+      "grad_norm": 1.5639890432357788,
+      "learning_rate": 9.399969976706509e-05,
+      "loss": 1.9938,
+      "step": 670
+    },
+    {
+      "epoch": 0.1686247408431237,
+      "grad_norm": 1.2142359018325806,
+      "learning_rate": 9.398069602701687e-05,
+      "loss": 2.148,
+      "step": 671
+    },
+    {
+      "epoch": 0.16887604448074386,
+      "grad_norm": 0.2211383879184723,
+      "learning_rate": 9.396166416810519e-05,
+      "loss": 2.0333,
+      "step": 672
+    },
+    {
+      "epoch": 0.16912734811836402,
+      "grad_norm": 0.3100007176399231,
+      "learning_rate": 9.394260420249801e-05,
+      "loss": 2.529,
+      "step": 673
+    },
+    {
+      "epoch": 0.16937865175598416,
+      "grad_norm": 0.39237165451049805,
+      "learning_rate": 9.39235161423812e-05,
+      "loss": 2.2316,
+      "step": 674
+    },
+    {
+      "epoch": 0.16962995539360431,
+      "grad_norm": 0.553925096988678,
+      "learning_rate": 9.390439999995865e-05,
+      "loss": 2.1865,
+      "step": 675
+    },
+    {
+      "epoch": 0.16988125903122447,
+      "grad_norm": 0.45522618293762207,
+      "learning_rate": 9.38852557874522e-05,
+      "loss": 2.346,
+      "step": 676
+    },
+    {
+      "epoch": 0.17013256266884463,
+      "grad_norm": 0.4678781032562256,
+      "learning_rate": 9.386608351710157e-05,
+      "loss": 1.9335,
+      "step": 677
+    },
+    {
+      "epoch": 0.1703838663064648,
+      "grad_norm": 0.3529011905193329,
+      "learning_rate": 9.38468832011645e-05,
+      "loss": 2.6414,
+      "step": 678
+    },
+    {
+      "epoch": 0.17063516994408495,
+      "grad_norm": 0.26705560088157654,
+      "learning_rate": 9.382765485191662e-05,
+      "loss": 2.3328,
+      "step": 679
+    },
+    {
+      "epoch": 0.17088647358170508,
+      "grad_norm": 0.3495092988014221,
+      "learning_rate": 9.380839848165149e-05,
+      "loss": 2.3455,
+      "step": 680
+    },
+    {
+      "epoch": 0.17113777721932524,
+      "grad_norm": 0.32814642786979675,
+      "learning_rate": 9.378911410268058e-05,
+      "loss": 2.4395,
+      "step": 681
+    },
+    {
+      "epoch": 0.1713890808569454,
+      "grad_norm": 0.33247315883636475,
+      "learning_rate": 9.376980172733329e-05,
+      "loss": 1.6115,
+      "step": 682
+    },
+    {
+      "epoch": 0.17164038449456556,
+      "grad_norm": 0.44238927960395813,
+      "learning_rate": 9.375046136795686e-05,
+      "loss": 2.1676,
+      "step": 683
+    },
+    {
+      "epoch": 0.17189168813218572,
+      "grad_norm": 0.3289899528026581,
+      "learning_rate": 9.373109303691652e-05,
+      "loss": 2.3906,
+      "step": 684
+    },
+    {
+      "epoch": 0.17214299176980588,
+      "grad_norm": 0.3287547826766968,
+      "learning_rate": 9.371169674659529e-05,
+      "loss": 2.1608,
+      "step": 685
+    },
+    {
+      "epoch": 0.172394295407426,
+      "grad_norm": 0.2546299397945404,
+      "learning_rate": 9.36922725093941e-05,
+      "loss": 2.4981,
+      "step": 686
+    },
+    {
+      "epoch": 0.17264559904504617,
+      "grad_norm": 0.2905830144882202,
+      "learning_rate": 9.367282033773177e-05,
+      "loss": 2.3735,
+      "step": 687
+    },
+    {
+      "epoch": 0.17289690268266633,
+      "grad_norm": 0.5941200256347656,
+      "learning_rate": 9.365334024404495e-05,
+      "loss": 2.7027,
+      "step": 688
+    },
+    {
+      "epoch": 0.1731482063202865,
+      "grad_norm": 0.3145928680896759,
+      "learning_rate": 9.363383224078814e-05,
+      "loss": 2.1827,
+      "step": 689
+    },
+    {
+      "epoch": 0.17339950995790665,
+      "grad_norm": 0.3063415586948395,
+      "learning_rate": 9.361429634043372e-05,
+      "loss": 2.0369,
+      "step": 690
+    },
+    {
+      "epoch": 0.17365081359552678,
+      "grad_norm": 0.40637439489364624,
+      "learning_rate": 9.359473255547186e-05,
+      "loss": 2.4201,
+      "step": 691
+    },
+    {
+      "epoch": 0.17390211723314694,
+      "grad_norm": 0.5028929710388184,
+      "learning_rate": 9.357514089841061e-05,
+      "loss": 1.7536,
+      "step": 692
+    },
+    {
+      "epoch": 0.1741534208707671,
+      "grad_norm": 0.19582106173038483,
+      "learning_rate": 9.355552138177577e-05,
+      "loss": 1.6803,
+      "step": 693
+    },
+    {
+      "epoch": 0.17440472450838726,
+      "grad_norm": 0.42780154943466187,
+      "learning_rate": 9.353587401811101e-05,
+      "loss": 2.0536,
+      "step": 694
+    },
+    {
+      "epoch": 0.17465602814600742,
+      "grad_norm": 0.41308510303497314,
+      "learning_rate": 9.351619881997779e-05,
+      "loss": 2.1088,
+      "step": 695
+    },
+    {
+      "epoch": 0.17490733178362758,
+      "grad_norm": 0.37217557430267334,
+      "learning_rate": 9.349649579995536e-05,
+      "loss": 2.3313,
+      "step": 696
+    },
+    {
+      "epoch": 0.1751586354212477,
+      "grad_norm": 0.49139076471328735,
+      "learning_rate": 9.347676497064074e-05,
+      "loss": 2.048,
+      "step": 697
+    },
+    {
+      "epoch": 0.17540993905886787,
+      "grad_norm": 0.49698978662490845,
+      "learning_rate": 9.345700634464876e-05,
+      "loss": 2.2736,
+      "step": 698
+    },
+    {
+      "epoch": 0.17566124269648803,
+      "grad_norm": 0.3499569296836853,
+      "learning_rate": 9.343721993461203e-05,
+      "loss": 2.4421,
+      "step": 699
+    },
+    {
+      "epoch": 0.1759125463341082,
+      "grad_norm": 0.2008151412010193,
+      "learning_rate": 9.341740575318088e-05,
+      "loss": 1.8728,
+      "step": 700
+    },
+    {
+      "epoch": 0.17616384997172835,
+      "grad_norm": 0.43613699078559875,
+      "learning_rate": 9.339756381302341e-05,
+      "loss": 2.1268,
+      "step": 701
+    },
+    {
+      "epoch": 0.1764151536093485,
+      "grad_norm": 0.24621760845184326,
+      "learning_rate": 9.337769412682551e-05,
+      "loss": 2.3428,
+      "step": 702
+    },
+    {
+      "epoch": 0.17666645724696864,
+      "grad_norm": 0.6878277659416199,
+      "learning_rate": 9.335779670729075e-05,
+      "loss": 2.1971,
+      "step": 703
+    },
+    {
+      "epoch": 0.1769177608845888,
+      "grad_norm": 0.3094477653503418,
+      "learning_rate": 9.333787156714047e-05,
+      "loss": 1.6276,
+      "step": 704
+    },
+    {
+      "epoch": 0.17716906452220896,
+      "grad_norm": 0.2712198495864868,
+      "learning_rate": 9.331791871911371e-05,
+      "loss": 2.1207,
+      "step": 705
+    },
+    {
+      "epoch": 0.17742036815982912,
+      "grad_norm": 0.40410006046295166,
+      "learning_rate": 9.329793817596724e-05,
+      "loss": 2.5001,
+      "step": 706
+    },
+    {
+      "epoch": 0.17767167179744928,
+      "grad_norm": 0.37285852432250977,
+      "learning_rate": 9.327792995047553e-05,
+      "loss": 2.513,
+      "step": 707
+    },
+    {
+      "epoch": 0.17792297543506944,
+      "grad_norm": 0.3589307963848114,
+      "learning_rate": 9.325789405543075e-05,
+      "loss": 2.3407,
+      "step": 708
+    },
+    {
+      "epoch": 0.17817427907268957,
+      "grad_norm": 0.16013433039188385,
+      "learning_rate": 9.323783050364276e-05,
+      "loss": 1.0699,
+      "step": 709
+    },
+    {
+      "epoch": 0.17842558271030973,
+      "grad_norm": 0.3747367858886719,
+      "learning_rate": 9.321773930793914e-05,
+      "loss": 2.2041,
+      "step": 710
+    },
+    {
+      "epoch": 0.1786768863479299,
+      "grad_norm": 0.48945263028144836,
+      "learning_rate": 9.319762048116503e-05,
+      "loss": 1.8284,
+      "step": 711
+    },
+    {
+      "epoch": 0.17892818998555005,
+      "grad_norm": 0.5572097897529602,
+      "learning_rate": 9.317747403618337e-05,
+      "loss": 1.8432,
+      "step": 712
+    },
+    {
+      "epoch": 0.1791794936231702,
+      "grad_norm": 0.4351899027824402,
+      "learning_rate": 9.31572999858747e-05,
+      "loss": 2.0469,
+      "step": 713
+    },
+    {
+      "epoch": 0.17943079726079034,
+      "grad_norm": 0.747898519039154,
+      "learning_rate": 9.31370983431372e-05,
+      "loss": 2.3444,
+      "step": 714
+    },
+    {
+      "epoch": 0.1796821008984105,
+      "grad_norm": 0.3677506148815155,
+      "learning_rate": 9.311686912088669e-05,
+      "loss": 1.7389,
+      "step": 715
+    },
+    {
+      "epoch": 0.17993340453603066,
+      "grad_norm": 0.7408022880554199,
+      "learning_rate": 9.309661233205663e-05,
+      "loss": 2.3839,
+      "step": 716
+    },
+    {
+      "epoch": 0.18018470817365081,
+      "grad_norm": 0.46684297919273376,
+      "learning_rate": 9.307632798959813e-05,
+      "loss": 2.4899,
+      "step": 717
+    },
+    {
+      "epoch": 0.18043601181127097,
+      "grad_norm": 0.5166415572166443,
+      "learning_rate": 9.305601610647989e-05,
+      "loss": 2.2496,
+      "step": 718
+    },
+    {
+      "epoch": 0.18068731544889113,
+      "grad_norm": 0.17087407410144806,
+      "learning_rate": 9.30356766956882e-05,
+      "loss": 0.624,
+      "step": 719
+    },
+    {
+      "epoch": 0.18093861908651127,
+      "grad_norm": 0.3293837904930115,
+      "learning_rate": 9.301530977022701e-05,
+      "loss": 2.1589,
+      "step": 720
+    },
+    {
+      "epoch": 0.18118992272413142,
+      "grad_norm": 0.5174190402030945,
+      "learning_rate": 9.29949153431178e-05,
+      "loss": 1.9908,
+      "step": 721
+    },
+    {
+      "epoch": 0.18144122636175158,
+      "grad_norm": 0.6492531895637512,
+      "learning_rate": 9.297449342739964e-05,
+      "loss": 2.3542,
+      "step": 722
+    },
+    {
+      "epoch": 0.18169252999937174,
+      "grad_norm": 0.3231172263622284,
+      "learning_rate": 9.295404403612924e-05,
+      "loss": 2.0346,
+      "step": 723
+    },
+    {
+      "epoch": 0.1819438336369919,
+      "grad_norm": 0.3993067145347595,
+      "learning_rate": 9.293356718238077e-05,
+      "loss": 2.0821,
+      "step": 724
+    },
+    {
+      "epoch": 0.18219513727461206,
+      "grad_norm": 0.3920503854751587,
+      "learning_rate": 9.291306287924608e-05,
+      "loss": 2.1273,
+      "step": 725
+    },
+    {
+      "epoch": 0.1824464409122322,
+      "grad_norm": 0.3167310655117035,
+      "learning_rate": 9.289253113983444e-05,
+      "loss": 2.0075,
+      "step": 726
+    },
+    {
+      "epoch": 0.18269774454985235,
+      "grad_norm": 0.3710818290710449,
+      "learning_rate": 9.287197197727277e-05,
+      "loss": 2.0204,
+      "step": 727
+    },
+    {
+      "epoch": 0.1829490481874725,
+      "grad_norm": 0.18097934126853943,
+      "learning_rate": 9.285138540470546e-05,
+      "loss": 0.749,
+      "step": 728
+    },
+    {
+      "epoch": 0.18320035182509267,
+      "grad_norm": 0.18432289361953735,
+      "learning_rate": 9.283077143529446e-05,
+      "loss": 0.8822,
+      "step": 729
+    },
+    {
+      "epoch": 0.18345165546271283,
+      "grad_norm": 0.3951958417892456,
+      "learning_rate": 9.281013008221921e-05,
+      "loss": 2.1613,
+      "step": 730
+    },
+    {
+      "epoch": 0.183702959100333,
+      "grad_norm": 0.3688110113143921,
+      "learning_rate": 9.278946135867665e-05,
+      "loss": 2.8058,
+      "step": 731
+    },
+    {
+      "epoch": 0.18395426273795312,
+      "grad_norm": 0.29696959257125854,
+      "learning_rate": 9.276876527788127e-05,
+      "loss": 1.7129,
+      "step": 732
+    },
+    {
+      "epoch": 0.18420556637557328,
+      "grad_norm": 0.405823677778244,
+      "learning_rate": 9.274804185306503e-05,
+      "loss": 2.334,
+      "step": 733
+    },
+    {
+      "epoch": 0.18445687001319344,
+      "grad_norm": 0.4110073447227478,
+      "learning_rate": 9.27272910974773e-05,
+      "loss": 2.4929,
+      "step": 734
+    },
+    {
+      "epoch": 0.1847081736508136,
+      "grad_norm": 0.3715936541557312,
+      "learning_rate": 9.270651302438502e-05,
+      "loss": 1.7891,
+      "step": 735
+    },
+    {
+      "epoch": 0.18495947728843376,
+      "grad_norm": 0.1682804673910141,
+      "learning_rate": 9.268570764707257e-05,
+      "loss": 0.726,
+      "step": 736
+    },
+    {
+      "epoch": 0.1852107809260539,
+      "grad_norm": 0.2395019680261612,
+      "learning_rate": 9.266487497884176e-05,
+      "loss": 2.4391,
+      "step": 737
+    },
+    {
+      "epoch": 0.18546208456367405,
+      "grad_norm": 0.3601885437965393,
+      "learning_rate": 9.264401503301185e-05,
+      "loss": 2.5297,
+      "step": 738
+    },
+    {
+      "epoch": 0.1857133882012942,
+      "grad_norm": 0.26369959115982056,
+      "learning_rate": 9.262312782291959e-05,
+      "loss": 1.7459,
+      "step": 739
+    },
+    {
+      "epoch": 0.18596469183891437,
+      "grad_norm": 0.15436404943466187,
+      "learning_rate": 9.26022133619191e-05,
+      "loss": 1.0022,
+      "step": 740
+    },
+    {
+      "epoch": 0.18621599547653453,
+      "grad_norm": 0.3303896486759186,
+      "learning_rate": 9.258127166338196e-05,
+      "loss": 2.3745,
+      "step": 741
+    },
+    {
+      "epoch": 0.1864672991141547,
+      "grad_norm": 0.3118177056312561,
+      "learning_rate": 9.256030274069713e-05,
+      "loss": 2.8046,
+      "step": 742
+    },
+    {
+      "epoch": 0.18671860275177482,
+      "grad_norm": 0.17974944412708282,
+      "learning_rate": 9.253930660727104e-05,
+      "loss": 0.8354,
+      "step": 743
+    },
+    {
+      "epoch": 0.18696990638939498,
+      "grad_norm": 0.2209557741880417,
+      "learning_rate": 9.251828327652742e-05,
+      "loss": 2.1792,
+      "step": 744
+    },
+    {
+      "epoch": 0.18722121002701514,
+      "grad_norm": 0.4132773280143738,
+      "learning_rate": 9.24972327619075e-05,
+      "loss": 2.3944,
+      "step": 745
+    },
+    {
+      "epoch": 0.1874725136646353,
+      "grad_norm": 0.47139856219291687,
+      "learning_rate": 9.24761550768698e-05,
+      "loss": 2.3182,
+      "step": 746
+    },
+    {
+      "epoch": 0.18772381730225546,
+      "grad_norm": 0.40354403853416443,
+      "learning_rate": 9.245505023489024e-05,
+      "loss": 2.1719,
+      "step": 747
+    },
+    {
+      "epoch": 0.18797512093987562,
+      "grad_norm": 0.23668596148490906,
+      "learning_rate": 9.243391824946213e-05,
+      "loss": 1.9976,
+      "step": 748
+    },
+    {
+      "epoch": 0.18822642457749575,
+      "grad_norm": 0.48701080679893494,
+      "learning_rate": 9.24127591340961e-05,
+      "loss": 2.3892,
+      "step": 749
+    },
+    {
+      "epoch": 0.1884777282151159,
+      "grad_norm": 0.48195892572402954,
+      "learning_rate": 9.239157290232014e-05,
+      "loss": 2.2488,
+      "step": 750
+    },
+    {
+      "epoch": 0.18872903185273607,
+      "grad_norm": 0.3957456946372986,
+      "learning_rate": 9.237035956767956e-05,
+      "loss": 2.2675,
+      "step": 751
+    },
+    {
+      "epoch": 0.18898033549035623,
+      "grad_norm": 0.419040709733963,
+      "learning_rate": 9.234911914373702e-05,
+      "loss": 1.9331,
+      "step": 752
+    },
+    {
+      "epoch": 0.1892316391279764,
+      "grad_norm": 0.3198854327201843,
+      "learning_rate": 9.23278516440725e-05,
+      "loss": 1.9211,
+      "step": 753
+    },
+    {
+      "epoch": 0.18948294276559652,
+      "grad_norm": 0.4320249855518341,
+      "learning_rate": 9.230655708228328e-05,
+      "loss": 1.9932,
+      "step": 754
+    },
+    {
+      "epoch": 0.18973424640321668,
+      "grad_norm": 0.34588703513145447,
+      "learning_rate": 9.228523547198393e-05,
+      "loss": 1.6818,
+      "step": 755
+    },
+    {
+      "epoch": 0.18998555004083684,
+      "grad_norm": 0.17924979329109192,
+      "learning_rate": 9.226388682680633e-05,
+      "loss": 1.2715,
+      "step": 756
+    },
+    {
+      "epoch": 0.190236853678457,
+      "grad_norm": 0.3479664921760559,
+      "learning_rate": 9.224251116039965e-05,
+      "loss": 2.6595,
+      "step": 757
+    },
+    {
+      "epoch": 0.19048815731607716,
+      "grad_norm": 0.9396395087242126,
+      "learning_rate": 9.222110848643035e-05,
+      "loss": 2.1373,
+      "step": 758
+    },
+    {
+      "epoch": 0.19073946095369732,
+      "grad_norm": 0.423880934715271,
+      "learning_rate": 9.219967881858209e-05,
+      "loss": 2.0013,
+      "step": 759
+    },
+    {
+      "epoch": 0.19099076459131745,
+      "grad_norm": 0.18442866206169128,
+      "learning_rate": 9.217822217055586e-05,
+      "loss": 1.1016,
+      "step": 760
+    },
+    {
+      "epoch": 0.1912420682289376,
+      "grad_norm": 0.33031755685806274,
+      "learning_rate": 9.215673855606986e-05,
+      "loss": 2.208,
+      "step": 761
+    },
+    {
+      "epoch": 0.19149337186655777,
+      "grad_norm": 0.5207613706588745,
+      "learning_rate": 9.213522798885956e-05,
+      "loss": 2.0212,
+      "step": 762
+    },
+    {
+      "epoch": 0.19174467550417792,
+      "grad_norm": 0.29409703612327576,
+      "learning_rate": 9.211369048267764e-05,
+      "loss": 2.5577,
+      "step": 763
+    },
+    {
+      "epoch": 0.19199597914179808,
+      "grad_norm": 0.44755882024765015,
+      "learning_rate": 9.2092126051294e-05,
+      "loss": 2.1216,
+      "step": 764
+    },
+    {
+      "epoch": 0.19224728277941824,
+      "grad_norm": 0.33680227398872375,
+      "learning_rate": 9.207053470849576e-05,
+      "loss": 2.5058,
+      "step": 765
+    },
+    {
+      "epoch": 0.19249858641703838,
+      "grad_norm": 0.41669735312461853,
+      "learning_rate": 9.204891646808726e-05,
+      "loss": 2.5137,
+      "step": 766
+    },
+    {
+      "epoch": 0.19274989005465853,
+      "grad_norm": 0.4869091808795929,
+      "learning_rate": 9.202727134389004e-05,
+      "loss": 2.2094,
+      "step": 767
+    },
+    {
+      "epoch": 0.1930011936922787,
+      "grad_norm": 0.3771580159664154,
+      "learning_rate": 9.20055993497428e-05,
+      "loss": 2.5748,
+      "step": 768
+    },
+    {
+      "epoch": 0.19325249732989885,
+      "grad_norm": 0.4663945734500885,
+      "learning_rate": 9.198390049950143e-05,
+      "loss": 2.6845,
+      "step": 769
+    },
+    {
+      "epoch": 0.193503800967519,
+      "grad_norm": 0.6000380516052246,
+      "learning_rate": 9.196217480703899e-05,
+      "loss": 2.4598,
+      "step": 770
+    },
+    {
+      "epoch": 0.19375510460513917,
+      "grad_norm": 0.4322783946990967,
+      "learning_rate": 9.194042228624572e-05,
+      "loss": 2.5049,
+      "step": 771
+    },
+    {
+      "epoch": 0.1940064082427593,
+      "grad_norm": 0.194077730178833,
+      "learning_rate": 9.191864295102899e-05,
+      "loss": 1.5018,
+      "step": 772
+    },
+    {
+      "epoch": 0.19425771188037946,
+      "grad_norm": 0.28692805767059326,
+      "learning_rate": 9.189683681531333e-05,
+      "loss": 1.8483,
+      "step": 773
+    },
+    {
+      "epoch": 0.19450901551799962,
+      "grad_norm": 0.8345639109611511,
+      "learning_rate": 9.187500389304037e-05,
+      "loss": 1.8403,
+      "step": 774
+    },
+    {
+      "epoch": 0.19476031915561978,
+      "grad_norm": 0.3533509373664856,
+      "learning_rate": 9.185314419816892e-05,
+      "loss": 2.4375,
+      "step": 775
+    },
+    {
+      "epoch": 0.19501162279323994,
+      "grad_norm": 0.40252941846847534,
+      "learning_rate": 9.18312577446749e-05,
+      "loss": 2.1452,
+      "step": 776
+    },
+    {
+      "epoch": 0.19526292643086007,
+      "grad_norm": 0.4904803931713104,
+      "learning_rate": 9.180934454655126e-05,
+      "loss": 2.2475,
+      "step": 777
+    },
+    {
+      "epoch": 0.19551423006848023,
+      "grad_norm": 0.4086427688598633,
+      "learning_rate": 9.178740461780812e-05,
+      "loss": 1.9234,
+      "step": 778
+    },
+    {
+      "epoch": 0.1957655337061004,
+      "grad_norm": 0.32106295228004456,
+      "learning_rate": 9.176543797247271e-05,
+      "loss": 2.3433,
+      "step": 779
+    },
+    {
+      "epoch": 0.19601683734372055,
+      "grad_norm": 0.45663875341415405,
+      "learning_rate": 9.17434446245893e-05,
+      "loss": 2.4102,
+      "step": 780
+    },
+    {
+      "epoch": 0.1962681409813407,
+      "grad_norm": 0.16669417917728424,
+      "learning_rate": 9.17214245882192e-05,
+      "loss": 0.9148,
+      "step": 781
+    },
+    {
+      "epoch": 0.19651944461896087,
+      "grad_norm": 0.21185293793678284,
+      "learning_rate": 9.169937787744088e-05,
+      "loss": 2.032,
+      "step": 782
+    },
+    {
+      "epoch": 0.196770748256581,
+      "grad_norm": 0.28057172894477844,
+      "learning_rate": 9.167730450634975e-05,
+      "loss": 2.3357,
+      "step": 783
+    },
+    {
+      "epoch": 0.19702205189420116,
+      "grad_norm": 0.30073508620262146,
+      "learning_rate": 9.165520448905835e-05,
+      "loss": 1.9842,
+      "step": 784
+    },
+    {
+      "epoch": 0.19727335553182132,
+      "grad_norm": 0.5807662606239319,
+      "learning_rate": 9.163307783969624e-05,
+      "loss": 2.1852,
+      "step": 785
+    },
+    {
+      "epoch": 0.19752465916944148,
+      "grad_norm": 0.43151628971099854,
+      "learning_rate": 9.161092457240999e-05,
+      "loss": 2.3249,
+      "step": 786
+    },
+    {
+      "epoch": 0.19777596280706164,
+      "grad_norm": 0.31615135073661804,
+      "learning_rate": 9.158874470136319e-05,
+      "loss": 2.3183,
+      "step": 787
+    },
+    {
+      "epoch": 0.1980272664446818,
+      "grad_norm": 0.4318180978298187,
+      "learning_rate": 9.156653824073642e-05,
+      "loss": 2.0189,
+      "step": 788
+    },
+    {
+      "epoch": 0.19827857008230193,
+      "grad_norm": 0.6335446834564209,
+      "learning_rate": 9.154430520472731e-05,
+      "loss": 1.8264,
+      "step": 789
+    },
+    {
+      "epoch": 0.1985298737199221,
+      "grad_norm": 0.1641846001148224,
+      "learning_rate": 9.152204560755045e-05,
+      "loss": 1.3867,
+      "step": 790
+    },
+    {
+      "epoch": 0.19878117735754225,
+      "grad_norm": 0.44742926955223083,
+      "learning_rate": 9.149975946343741e-05,
+      "loss": 1.9269,
+      "step": 791
+    },
+    {
+      "epoch": 0.1990324809951624,
+      "grad_norm": 0.438804566860199,
+      "learning_rate": 9.147744678663672e-05,
+      "loss": 1.8561,
+      "step": 792
+    },
+    {
+      "epoch": 0.19928378463278257,
+      "grad_norm": 0.6063904166221619,
+      "learning_rate": 9.145510759141393e-05,
+      "loss": 2.1038,
+      "step": 793
+    },
+    {
+      "epoch": 0.19953508827040273,
+      "grad_norm": 0.3686808943748474,
+      "learning_rate": 9.143274189205147e-05,
+      "loss": 2.811,
+      "step": 794
+    },
+    {
+      "epoch": 0.19978639190802286,
+      "grad_norm": 0.45831429958343506,
+      "learning_rate": 9.141034970284877e-05,
+      "loss": 2.1029,
+      "step": 795
+    },
+    {
+      "epoch": 0.20003769554564302,
+      "grad_norm": 0.4418196678161621,
+      "learning_rate": 9.138793103812218e-05,
+      "loss": 1.9126,
+      "step": 796
+    },
+    {
+      "epoch": 0.20028899918326318,
+      "grad_norm": 0.23358654975891113,
+      "learning_rate": 9.136548591220495e-05,
+      "loss": 2.0087,
+      "step": 797
+    },
+    {
+      "epoch": 0.20054030282088334,
+      "grad_norm": 0.5014088749885559,
+      "learning_rate": 9.134301433944731e-05,
+      "loss": 2.1698,
+      "step": 798
+    },
+    {
+      "epoch": 0.2007916064585035,
+      "grad_norm": 0.48934677243232727,
+      "learning_rate": 9.132051633421632e-05,
+      "loss": 1.9628,
+      "step": 799
+    },
+    {
+      "epoch": 0.20104291009612363,
+      "grad_norm": 0.46975913643836975,
+      "learning_rate": 9.129799191089601e-05,
+      "loss": 2.1432,
+      "step": 800
+    },
+    {
+      "epoch": 0.2012942137337438,
+      "grad_norm": 0.33380022644996643,
+      "learning_rate": 9.127544108388725e-05,
+      "loss": 1.7332,
+      "step": 801
+    },
+    {
+      "epoch": 0.20154551737136395,
+      "grad_norm": 0.6146292090415955,
+      "learning_rate": 9.125286386760785e-05,
+      "loss": 2.2721,
+      "step": 802
+    },
+    {
+      "epoch": 0.2017968210089841,
+      "grad_norm": 0.14950276911258698,
+      "learning_rate": 9.12302602764924e-05,
+      "loss": 1.2602,
+      "step": 803
+    },
+    {
+      "epoch": 0.20204812464660427,
+      "grad_norm": 0.4614298641681671,
+      "learning_rate": 9.120763032499242e-05,
+      "loss": 2.2327,
+      "step": 804
+    },
+    {
+      "epoch": 0.20229942828422443,
+      "grad_norm": 0.249393031001091,
+      "learning_rate": 9.118497402757631e-05,
+      "loss": 1.8148,
+      "step": 805
+    },
+    {
+      "epoch": 0.20255073192184456,
+      "grad_norm": 0.41946738958358765,
+      "learning_rate": 9.116229139872922e-05,
+      "loss": 2.5221,
+      "step": 806
+    },
+    {
+      "epoch": 0.20280203555946472,
+      "grad_norm": 0.4390411078929901,
+      "learning_rate": 9.113958245295321e-05,
+      "loss": 2.1989,
+      "step": 807
+    },
+    {
+      "epoch": 0.20305333919708488,
+      "grad_norm": 0.5600268244743347,
+      "learning_rate": 9.111684720476717e-05,
+      "loss": 2.6773,
+      "step": 808
+    },
+    {
+      "epoch": 0.20330464283470503,
+      "grad_norm": 0.2843821346759796,
+      "learning_rate": 9.109408566870673e-05,
+      "loss": 1.9472,
+      "step": 809
+    },
+    {
+      "epoch": 0.2035559464723252,
+      "grad_norm": 0.3715212643146515,
+      "learning_rate": 9.107129785932443e-05,
+      "loss": 2.1466,
+      "step": 810
+    },
+    {
+      "epoch": 0.20380725010994535,
+      "grad_norm": 0.8551393151283264,
+      "learning_rate": 9.10484837911895e-05,
+      "loss": 2.199,
+      "step": 811
+    },
+    {
+      "epoch": 0.20405855374756549,
+      "grad_norm": 0.5052684545516968,
+      "learning_rate": 9.102564347888806e-05,
+      "loss": 1.9972,
+      "step": 812
+    },
+    {
+      "epoch": 0.20430985738518564,
+      "grad_norm": 0.24479907751083374,
+      "learning_rate": 9.100277693702294e-05,
+      "loss": 2.3708,
+      "step": 813
+    },
+    {
+      "epoch": 0.2045611610228058,
+      "grad_norm": 0.42139294743537903,
+      "learning_rate": 9.097988418021377e-05,
+      "loss": 1.9225,
+      "step": 814
+    },
+    {
+      "epoch": 0.20481246466042596,
+      "grad_norm": 0.342489629983902,
+      "learning_rate": 9.095696522309693e-05,
+      "loss": 2.7236,
+      "step": 815
+    },
+    {
+      "epoch": 0.20506376829804612,
+      "grad_norm": 0.47755831480026245,
+      "learning_rate": 9.093402008032554e-05,
+      "loss": 2.2168,
+      "step": 816
+    },
+    {
+      "epoch": 0.20531507193566628,
+      "grad_norm": 0.32807457447052,
+      "learning_rate": 9.09110487665695e-05,
+      "loss": 2.4625,
+      "step": 817
+    },
+    {
+      "epoch": 0.20556637557328641,
+      "grad_norm": 0.3581337034702301,
+      "learning_rate": 9.088805129651542e-05,
+      "loss": 2.6607,
+      "step": 818
+    },
+    {
+      "epoch": 0.20581767921090657,
+      "grad_norm": 0.24006719887256622,
+      "learning_rate": 9.08650276848666e-05,
+      "loss": 2.0933,
+      "step": 819
+    },
+    {
+      "epoch": 0.20606898284852673,
+      "grad_norm": 0.22163568437099457,
+      "learning_rate": 9.084197794634312e-05,
+      "loss": 2.0709,
+      "step": 820
+    },
+    {
+      "epoch": 0.2063202864861469,
+      "grad_norm": 0.20495516061782837,
+      "learning_rate": 9.081890209568169e-05,
+      "loss": 1.8137,
+      "step": 821
+    },
+    {
+      "epoch": 0.20657159012376705,
+      "grad_norm": 0.33006590604782104,
+      "learning_rate": 9.079580014763579e-05,
+      "loss": 2.242,
+      "step": 822
+    },
+    {
+      "epoch": 0.20682289376138718,
+      "grad_norm": 0.31708237528800964,
+      "learning_rate": 9.077267211697554e-05,
+      "loss": 2.5707,
+      "step": 823
+    },
+    {
+      "epoch": 0.20707419739900734,
+      "grad_norm": 0.3039303719997406,
+      "learning_rate": 9.07495180184877e-05,
+      "loss": 2.4013,
+      "step": 824
+    },
+    {
+      "epoch": 0.2073255010366275,
+      "grad_norm": 0.3243713974952698,
+      "learning_rate": 9.072633786697581e-05,
+      "loss": 1.9324,
+      "step": 825
+    },
+    {
+      "epoch": 0.20757680467424766,
+      "grad_norm": 0.376941055059433,
+      "learning_rate": 9.070313167725995e-05,
+      "loss": 1.9856,
+      "step": 826
+    },
+    {
+      "epoch": 0.20782810831186782,
+      "grad_norm": 0.4256725013256073,
+      "learning_rate": 9.06798994641769e-05,
+      "loss": 1.9451,
+      "step": 827
+    },
+    {
+      "epoch": 0.20807941194948798,
+      "grad_norm": 0.21601825952529907,
+      "learning_rate": 9.06566412425801e-05,
+      "loss": 0.9348,
+      "step": 828
+    },
+    {
+      "epoch": 0.2083307155871081,
+      "grad_norm": 0.5165765881538391,
+      "learning_rate": 9.063335702733958e-05,
+      "loss": 2.3604,
+      "step": 829
+    },
+    {
+      "epoch": 0.20858201922472827,
+      "grad_norm": 0.4365144670009613,
+      "learning_rate": 9.061004683334196e-05,
+      "loss": 2.0167,
+      "step": 830
+    },
+    {
+      "epoch": 0.20883332286234843,
+      "grad_norm": 0.2237500250339508,
+      "learning_rate": 9.058671067549056e-05,
+      "loss": 1.7844,
+      "step": 831
+    },
+    {
+      "epoch": 0.2090846264999686,
+      "grad_norm": 0.26887792348861694,
+      "learning_rate": 9.056334856870522e-05,
+      "loss": 2.3547,
+      "step": 832
+    },
+    {
+      "epoch": 0.20933593013758875,
+      "grad_norm": 0.21619755029678345,
+      "learning_rate": 9.053996052792244e-05,
+      "loss": 2.0557,
+      "step": 833
+    },
+    {
+      "epoch": 0.2095872337752089,
+      "grad_norm": 0.5002549886703491,
+      "learning_rate": 9.051654656809521e-05,
+      "loss": 1.962,
+      "step": 834
+    },
+    {
+      "epoch": 0.20983853741282904,
+      "grad_norm": 0.3360225260257721,
+      "learning_rate": 9.049310670419316e-05,
+      "loss": 1.9531,
+      "step": 835
+    },
+    {
+      "epoch": 0.2100898410504492,
+      "grad_norm": 0.24657025933265686,
+      "learning_rate": 9.046964095120248e-05,
+      "loss": 1.2244,
+      "step": 836
+    },
+    {
+      "epoch": 0.21034114468806936,
+      "grad_norm": 0.29951533675193787,
+      "learning_rate": 9.044614932412587e-05,
+      "loss": 1.9471,
+      "step": 837
+    },
+    {
+      "epoch": 0.21059244832568952,
+      "grad_norm": 0.3678789734840393,
+      "learning_rate": 9.04226318379826e-05,
+      "loss": 2.3963,
+      "step": 838
+    },
+    {
+      "epoch": 0.21084375196330968,
+      "grad_norm": 0.45650580525398254,
+      "learning_rate": 9.03990885078085e-05,
+      "loss": 2.342,
+      "step": 839
+    },
+    {
+      "epoch": 0.21109505560092984,
+      "grad_norm": 0.4444562792778015,
+      "learning_rate": 9.037551934865587e-05,
+      "loss": 1.9851,
+      "step": 840
+    },
+    {
+      "epoch": 0.21134635923854997,
+      "grad_norm": 0.2063484787940979,
+      "learning_rate": 9.035192437559354e-05,
+      "loss": 2.2532,
+      "step": 841
+    },
+    {
+      "epoch": 0.21159766287617013,
+      "grad_norm": 0.3520076274871826,
+      "learning_rate": 9.032830360370688e-05,
+      "loss": 1.6042,
+      "step": 842
+    },
+    {
+      "epoch": 0.2118489665137903,
+      "grad_norm": 0.4106435477733612,
+      "learning_rate": 9.03046570480977e-05,
+      "loss": 2.0197,
+      "step": 843
+    },
+    {
+      "epoch": 0.21210027015141045,
+      "grad_norm": 0.37212634086608887,
+      "learning_rate": 9.028098472388433e-05,
+      "loss": 2.1224,
+      "step": 844
+    },
+    {
+      "epoch": 0.2123515737890306,
+      "grad_norm": 0.36991527676582336,
+      "learning_rate": 9.025728664620157e-05,
+      "loss": 2.5759,
+      "step": 845
+    },
+    {
+      "epoch": 0.21260287742665074,
+      "grad_norm": 0.3646388649940491,
+      "learning_rate": 9.023356283020067e-05,
+      "loss": 2.2962,
+      "step": 846
+    },
+    {
+      "epoch": 0.2128541810642709,
+      "grad_norm": 0.21754935383796692,
+      "learning_rate": 9.020981329104936e-05,
+      "loss": 1.7198,
+      "step": 847
+    },
+    {
+      "epoch": 0.21310548470189106,
+      "grad_norm": 0.24025185406208038,
+      "learning_rate": 9.01860380439318e-05,
+      "loss": 1.271,
+      "step": 848
+    },
+    {
+      "epoch": 0.21335678833951122,
+      "grad_norm": 0.42848479747772217,
+      "learning_rate": 9.016223710404856e-05,
+      "loss": 2.3559,
+      "step": 849
+    },
+    {
+      "epoch": 0.21360809197713138,
+      "grad_norm": 0.4237200915813446,
+      "learning_rate": 9.013841048661673e-05,
+      "loss": 2.2002,
+      "step": 850
+    },
+    {
+      "epoch": 0.21385939561475154,
+      "grad_norm": 0.31660404801368713,
+      "learning_rate": 9.01145582068697e-05,
+      "loss": 2.441,
+      "step": 851
+    },
+    {
+      "epoch": 0.21411069925237167,
+      "grad_norm": 0.4812658131122589,
+      "learning_rate": 9.009068028005732e-05,
+      "loss": 2.2861,
+      "step": 852
+    },
+    {
+      "epoch": 0.21436200288999183,
+      "grad_norm": 0.37174031138420105,
+      "learning_rate": 9.006677672144586e-05,
+      "loss": 2.4933,
+      "step": 853
+    },
+    {
+      "epoch": 0.21461330652761199,
+      "grad_norm": 0.5248540639877319,
+      "learning_rate": 9.004284754631793e-05,
+      "loss": 1.9397,
+      "step": 854
+    },
+    {
+      "epoch": 0.21486461016523214,
+      "grad_norm": 0.2802974283695221,
+      "learning_rate": 9.001889276997258e-05,
+      "loss": 2.2688,
+      "step": 855
+    },
+    {
+      "epoch": 0.2151159138028523,
+      "grad_norm": 0.4122345745563507,
+      "learning_rate": 8.999491240772516e-05,
+      "loss": 2.1688,
+      "step": 856
+    },
+    {
+      "epoch": 0.21536721744047246,
+      "grad_norm": 0.49358898401260376,
+      "learning_rate": 8.99709064749074e-05,
+      "loss": 2.0717,
+      "step": 857
+    },
+    {
+      "epoch": 0.2156185210780926,
+      "grad_norm": 0.415002703666687,
+      "learning_rate": 8.994687498686742e-05,
+      "loss": 2.4572,
+      "step": 858
+    },
+    {
+      "epoch": 0.21586982471571275,
+      "grad_norm": 0.3565453290939331,
+      "learning_rate": 8.992281795896962e-05,
+      "loss": 2.2275,
+      "step": 859
+    },
+    {
+      "epoch": 0.21612112835333291,
+      "grad_norm": 0.25147542357444763,
+      "learning_rate": 8.989873540659476e-05,
+      "loss": 1.6368,
+      "step": 860
+    },
+    {
+      "epoch": 0.21637243199095307,
+      "grad_norm": 0.330954372882843,
+      "learning_rate": 8.987462734513993e-05,
+      "loss": 1.6743,
+      "step": 861
+    },
+    {
+      "epoch": 0.21662373562857323,
+      "grad_norm": 0.3177106976509094,
+      "learning_rate": 8.985049379001849e-05,
+      "loss": 2.373,
+      "step": 862
+    },
+    {
+      "epoch": 0.21687503926619336,
+      "grad_norm": 0.4051658511161804,
+      "learning_rate": 8.982633475666014e-05,
+      "loss": 2.504,
+      "step": 863
+    },
+    {
+      "epoch": 0.21712634290381352,
+      "grad_norm": 0.1954205185174942,
+      "learning_rate": 8.980215026051083e-05,
+      "loss": 1.8698,
+      "step": 864
+    },
+    {
+      "epoch": 0.21737764654143368,
+      "grad_norm": 0.31402403116226196,
+      "learning_rate": 8.977794031703282e-05,
+      "loss": 2.2363,
+      "step": 865
+    },
+    {
+      "epoch": 0.21762895017905384,
+      "grad_norm": 0.28115931153297424,
+      "learning_rate": 8.975370494170463e-05,
+      "loss": 2.3547,
+      "step": 866
+    },
+    {
+      "epoch": 0.217880253816674,
+      "grad_norm": 0.4603864252567291,
+      "learning_rate": 8.972944415002105e-05,
+      "loss": 2.3678,
+      "step": 867
+    },
+    {
+      "epoch": 0.21813155745429416,
+      "grad_norm": 0.21994365751743317,
+      "learning_rate": 8.97051579574931e-05,
+      "loss": 2.2028,
+      "step": 868
+    },
+    {
+      "epoch": 0.2183828610919143,
+      "grad_norm": 0.3067527711391449,
+      "learning_rate": 8.968084637964804e-05,
+      "loss": 2.6017,
+      "step": 869
+    },
+    {
+      "epoch": 0.21863416472953445,
+      "grad_norm": 0.33018767833709717,
+      "learning_rate": 8.96565094320294e-05,
+      "loss": 2.3301,
+      "step": 870
+    },
+    {
+      "epoch": 0.2188854683671546,
+      "grad_norm": 0.6793331503868103,
+      "learning_rate": 8.963214713019687e-05,
+      "loss": 2.7803,
+      "step": 871
+    },
+    {
+      "epoch": 0.21913677200477477,
+      "grad_norm": 0.3478843569755554,
+      "learning_rate": 8.96077594897264e-05,
+      "loss": 2.2915,
+      "step": 872
+    },
+    {
+      "epoch": 0.21938807564239493,
+      "grad_norm": 0.400673508644104,
+      "learning_rate": 8.95833465262101e-05,
+      "loss": 1.2611,
+      "step": 873
+    },
+    {
+      "epoch": 0.2196393792800151,
+      "grad_norm": 0.27117395401000977,
+      "learning_rate": 8.955890825525631e-05,
+      "loss": 2.6418,
+      "step": 874
+    },
+    {
+      "epoch": 0.21989068291763522,
+      "grad_norm": 0.41636621952056885,
+      "learning_rate": 8.953444469248952e-05,
+      "loss": 2.0555,
+      "step": 875
+    },
+    {
+      "epoch": 0.22014198655525538,
+      "grad_norm": 0.5339227318763733,
+      "learning_rate": 8.95099558535504e-05,
+      "loss": 2.4363,
+      "step": 876
+    },
+    {
+      "epoch": 0.22039329019287554,
+      "grad_norm": 0.4329914450645447,
+      "learning_rate": 8.948544175409579e-05,
+      "loss": 2.198,
+      "step": 877
+    },
+    {
+      "epoch": 0.2206445938304957,
+      "grad_norm": 0.377668559551239,
+      "learning_rate": 8.946090240979865e-05,
+      "loss": 2.2962,
+      "step": 878
+    },
+    {
+      "epoch": 0.22089589746811586,
+      "grad_norm": 0.3951661288738251,
+      "learning_rate": 8.943633783634813e-05,
+      "loss": 2.1264,
+      "step": 879
+    },
+    {
+      "epoch": 0.22114720110573602,
+      "grad_norm": 0.36469566822052,
+      "learning_rate": 8.941174804944948e-05,
+      "loss": 2.6947,
+      "step": 880
+    },
+    {
+      "epoch": 0.22139850474335615,
+      "grad_norm": 0.21884319186210632,
+      "learning_rate": 8.938713306482403e-05,
+      "loss": 1.9526,
+      "step": 881
+    },
+    {
+      "epoch": 0.2216498083809763,
+      "grad_norm": 0.38323378562927246,
+      "learning_rate": 8.936249289820931e-05,
+      "loss": 2.1726,
+      "step": 882
+    },
+    {
+      "epoch": 0.22190111201859647,
+      "grad_norm": 0.653200089931488,
+      "learning_rate": 8.933782756535887e-05,
+      "loss": 2.505,
+      "step": 883
+    },
+    {
+      "epoch": 0.22215241565621663,
+      "grad_norm": 0.41666847467422485,
+      "learning_rate": 8.931313708204239e-05,
+      "loss": 2.218,
+      "step": 884
+    },
+    {
+      "epoch": 0.2224037192938368,
+      "grad_norm": 0.3992173373699188,
+      "learning_rate": 8.928842146404562e-05,
+      "loss": 2.4002,
+      "step": 885
+    },
+    {
+      "epoch": 0.22265502293145692,
+      "grad_norm": 0.5349919199943542,
+      "learning_rate": 8.92636807271704e-05,
+      "loss": 1.5443,
+      "step": 886
+    },
+    {
+      "epoch": 0.22290632656907708,
+      "grad_norm": 0.3590314984321594,
+      "learning_rate": 8.923891488723459e-05,
+      "loss": 2.3424,
+      "step": 887
+    },
+    {
+      "epoch": 0.22315763020669724,
+      "grad_norm": 0.4399639666080475,
+      "learning_rate": 8.921412396007212e-05,
+      "loss": 2.3039,
+      "step": 888
+    },
+    {
+      "epoch": 0.2234089338443174,
+      "grad_norm": 0.476307213306427,
+      "learning_rate": 8.918930796153297e-05,
+      "loss": 2.0807,
+      "step": 889
+    },
+    {
+      "epoch": 0.22366023748193756,
+      "grad_norm": 0.40012243390083313,
+      "learning_rate": 8.916446690748315e-05,
+      "loss": 2.2535,
+      "step": 890
+    },
+    {
+      "epoch": 0.22391154111955772,
+      "grad_norm": 0.4795278012752533,
+      "learning_rate": 8.913960081380465e-05,
+      "loss": 1.9693,
+      "step": 891
+    },
+    {
+      "epoch": 0.22416284475717785,
+      "grad_norm": 0.1798836588859558,
+      "learning_rate": 8.911470969639551e-05,
+      "loss": 0.6201,
+      "step": 892
+    },
+    {
+      "epoch": 0.224414148394798,
+      "grad_norm": 0.8967116475105286,
+      "learning_rate": 8.908979357116976e-05,
+      "loss": 2.3321,
+      "step": 893
+    },
+    {
+      "epoch": 0.22466545203241817,
+      "grad_norm": 0.363643616437912,
+      "learning_rate": 8.90648524540574e-05,
+      "loss": 2.2787,
+      "step": 894
+    },
+    {
+      "epoch": 0.22491675567003833,
+      "grad_norm": 0.46721211075782776,
+      "learning_rate": 8.903988636100445e-05,
+      "loss": 2.05,
+      "step": 895
+    },
+    {
+      "epoch": 0.22516805930765849,
+      "grad_norm": 0.41919320821762085,
+      "learning_rate": 8.901489530797282e-05,
+      "loss": 2.2821,
+      "step": 896
+    },
+    {
+      "epoch": 0.22541936294527865,
+      "grad_norm": 0.29892247915267944,
+      "learning_rate": 8.898987931094049e-05,
+      "loss": 2.3336,
+      "step": 897
+    },
+    {
+      "epoch": 0.22567066658289878,
+      "grad_norm": 0.41926395893096924,
+      "learning_rate": 8.896483838590131e-05,
+      "loss": 2.1726,
+      "step": 898
+    },
+    {
+      "epoch": 0.22592197022051894,
+      "grad_norm": 0.3506767451763153,
+      "learning_rate": 8.893977254886505e-05,
+      "loss": 1.8011,
+      "step": 899
+    },
+    {
+      "epoch": 0.2261732738581391,
+      "grad_norm": 0.39030522108078003,
+      "learning_rate": 8.891468181585747e-05,
+      "loss": 1.962,
+      "step": 900
+    },
+    {
+      "epoch": 0.22642457749575925,
+      "grad_norm": 0.40221068263053894,
+      "learning_rate": 8.888956620292022e-05,
+      "loss": 2.2385,
+      "step": 901
+    },
+    {
+      "epoch": 0.22667588113337941,
+      "grad_norm": 0.312210351228714,
+      "learning_rate": 8.886442572611087e-05,
+      "loss": 2.3336,
+      "step": 902
+    },
+    {
+      "epoch": 0.22692718477099957,
+      "grad_norm": 0.1972939372062683,
+      "learning_rate": 8.883926040150283e-05,
+      "loss": 1.8672,
+      "step": 903
+    },
+    {
+      "epoch": 0.2271784884086197,
+      "grad_norm": 0.30837222933769226,
+      "learning_rate": 8.881407024518548e-05,
+      "loss": 2.1923,
+      "step": 904
+    },
+    {
+      "epoch": 0.22742979204623986,
+      "grad_norm": 0.32954657077789307,
+      "learning_rate": 8.8788855273264e-05,
+      "loss": 2.53,
+      "step": 905
+    },
+    {
+      "epoch": 0.22768109568386002,
+      "grad_norm": 0.4983203113079071,
+      "learning_rate": 8.87636155018595e-05,
+      "loss": 1.7686,
+      "step": 906
+    },
+    {
+      "epoch": 0.22793239932148018,
+      "grad_norm": 0.29157817363739014,
+      "learning_rate": 8.873835094710891e-05,
+      "loss": 2.0444,
+      "step": 907
+    },
+    {
+      "epoch": 0.22818370295910034,
+      "grad_norm": 0.7063765525817871,
+      "learning_rate": 8.8713061625165e-05,
+      "loss": 2.3444,
+      "step": 908
+    },
+    {
+      "epoch": 0.22843500659672047,
+      "grad_norm": 0.2752906382083893,
+      "learning_rate": 8.868774755219641e-05,
+      "loss": 2.2363,
+      "step": 909
+    },
+    {
+      "epoch": 0.22868631023434063,
+      "grad_norm": 0.4020974338054657,
+      "learning_rate": 8.866240874438755e-05,
+      "loss": 2.2792,
+      "step": 910
+    },
+    {
+      "epoch": 0.2289376138719608,
+      "grad_norm": 0.3203171193599701,
+      "learning_rate": 8.863704521793869e-05,
+      "loss": 2.1895,
+      "step": 911
+    },
+    {
+      "epoch": 0.22918891750958095,
+      "grad_norm": 0.7605288028717041,
+      "learning_rate": 8.861165698906589e-05,
+      "loss": 2.2526,
+      "step": 912
+    },
+    {
+      "epoch": 0.2294402211472011,
+      "grad_norm": 0.4583891034126282,
+      "learning_rate": 8.8586244074001e-05,
+      "loss": 2.5682,
+      "step": 913
+    },
+    {
+      "epoch": 0.22969152478482127,
+      "grad_norm": 0.29976171255111694,
+      "learning_rate": 8.856080648899163e-05,
+      "loss": 1.8839,
+      "step": 914
+    },
+    {
+      "epoch": 0.2299428284224414,
+      "grad_norm": 0.4007084369659424,
+      "learning_rate": 8.853534425030123e-05,
+      "loss": 2.2063,
+      "step": 915
+    },
+    {
+      "epoch": 0.23019413206006156,
+      "grad_norm": 0.37610113620758057,
+      "learning_rate": 8.850985737420896e-05,
+      "loss": 2.5636,
+      "step": 916
+    },
+    {
+      "epoch": 0.23044543569768172,
+      "grad_norm": 0.3937523663043976,
+      "learning_rate": 8.84843458770097e-05,
+      "loss": 2.2865,
+      "step": 917
+    },
+    {
+      "epoch": 0.23069673933530188,
+      "grad_norm": 0.44338393211364746,
+      "learning_rate": 8.845880977501419e-05,
+      "loss": 2.2287,
+      "step": 918
+    },
+    {
+      "epoch": 0.23094804297292204,
+      "grad_norm": 0.4235187768936157,
+      "learning_rate": 8.843324908454875e-05,
+      "loss": 2.3839,
+      "step": 919
+    },
+    {
+      "epoch": 0.2311993466105422,
+      "grad_norm": 0.5745797157287598,
+      "learning_rate": 8.840766382195553e-05,
+      "loss": 1.9735,
+      "step": 920
+    },
+    {
+      "epoch": 0.23145065024816233,
+      "grad_norm": 0.35707423090934753,
+      "learning_rate": 8.838205400359234e-05,
+      "loss": 1.8084,
+      "step": 921
+    },
+    {
+      "epoch": 0.2317019538857825,
+      "grad_norm": 0.323047935962677,
+      "learning_rate": 8.835641964583272e-05,
+      "loss": 2.422,
+      "step": 922
+    },
+    {
+      "epoch": 0.23195325752340265,
+      "grad_norm": 0.4362463355064392,
+      "learning_rate": 8.833076076506588e-05,
+      "loss": 2.5153,
+      "step": 923
+    },
+    {
+      "epoch": 0.2322045611610228,
+      "grad_norm": 0.3632015883922577,
+      "learning_rate": 8.830507737769669e-05,
+      "loss": 2.7776,
+      "step": 924
+    },
+    {
+      "epoch": 0.23245586479864297,
+      "grad_norm": 0.48653021454811096,
+      "learning_rate": 8.827936950014573e-05,
+      "loss": 2.3679,
+      "step": 925
+    },
+    {
+      "epoch": 0.23270716843626313,
+      "grad_norm": 0.37184858322143555,
+      "learning_rate": 8.825363714884922e-05,
+      "loss": 1.9735,
+      "step": 926
+    },
+    {
+      "epoch": 0.23295847207388326,
+      "grad_norm": 0.38167500495910645,
+      "learning_rate": 8.822788034025903e-05,
+      "loss": 2.8957,
+      "step": 927
+    },
+    {
+      "epoch": 0.23320977571150342,
+      "grad_norm": 0.4858744144439697,
+      "learning_rate": 8.820209909084265e-05,
+      "loss": 2.4265,
+      "step": 928
+    },
+    {
+      "epoch": 0.23346107934912358,
+      "grad_norm": 0.4078523516654968,
+      "learning_rate": 8.81762934170832e-05,
+      "loss": 2.0574,
+      "step": 929
+    },
+    {
+      "epoch": 0.23371238298674374,
+      "grad_norm": 0.3559398353099823,
+      "learning_rate": 8.815046333547943e-05,
+      "loss": 2.2669,
+      "step": 930
+    },
+    {
+      "epoch": 0.2339636866243639,
+      "grad_norm": 0.4299301207065582,
+      "learning_rate": 8.81246088625457e-05,
+      "loss": 2.1053,
+      "step": 931
+    },
+    {
+      "epoch": 0.23421499026198403,
+      "grad_norm": 0.40756756067276,
+      "learning_rate": 8.809873001481193e-05,
+      "loss": 2.3635,
+      "step": 932
+    },
+    {
+      "epoch": 0.2344662938996042,
+      "grad_norm": 0.37873560190200806,
+      "learning_rate": 8.807282680882367e-05,
+      "loss": 1.0149,
+      "step": 933
+    },
+    {
+      "epoch": 0.23471759753722435,
+      "grad_norm": 0.40375521779060364,
+      "learning_rate": 8.8046899261142e-05,
+      "loss": 2.5615,
+      "step": 934
+    },
+    {
+      "epoch": 0.2349689011748445,
+      "grad_norm": 0.4414771497249603,
+      "learning_rate": 8.802094738834361e-05,
+      "loss": 2.513,
+      "step": 935
+    },
+    {
+      "epoch": 0.23522020481246467,
+      "grad_norm": 0.40930548310279846,
+      "learning_rate": 8.799497120702069e-05,
+      "loss": 1.5781,
+      "step": 936
+    },
+    {
+      "epoch": 0.23547150845008483,
+      "grad_norm": 0.4570627808570862,
+      "learning_rate": 8.7968970733781e-05,
+      "loss": 2.2293,
+      "step": 937
+    },
+    {
+      "epoch": 0.23572281208770496,
+      "grad_norm": 0.38974353671073914,
+      "learning_rate": 8.794294598524784e-05,
+      "loss": 1.959,
+      "step": 938
+    },
+    {
+      "epoch": 0.23597411572532512,
+      "grad_norm": 0.5479612946510315,
+      "learning_rate": 8.791689697806e-05,
+      "loss": 2.0345,
+      "step": 939
+    },
+    {
+      "epoch": 0.23622541936294528,
+      "grad_norm": 0.35705479979515076,
+      "learning_rate": 8.789082372887183e-05,
+      "loss": 2.2542,
+      "step": 940
+    },
+    {
+      "epoch": 0.23647672300056544,
+      "grad_norm": 0.4570086896419525,
+      "learning_rate": 8.786472625435311e-05,
+      "loss": 2.4306,
+      "step": 941
+    },
+    {
+      "epoch": 0.2367280266381856,
+      "grad_norm": 1.2224934101104736,
+      "learning_rate": 8.783860457118918e-05,
+      "loss": 2.031,
+      "step": 942
+    },
+    {
+      "epoch": 0.23697933027580576,
+      "grad_norm": 0.47019338607788086,
+      "learning_rate": 8.781245869608077e-05,
+      "loss": 2.5845,
+      "step": 943
+    },
+    {
+      "epoch": 0.2372306339134259,
+      "grad_norm": 0.4843159019947052,
+      "learning_rate": 8.778628864574419e-05,
+      "loss": 2.2976,
+      "step": 944
+    },
+    {
+      "epoch": 0.23748193755104605,
+      "grad_norm": 0.44156232476234436,
+      "learning_rate": 8.776009443691109e-05,
+      "loss": 1.7753,
+      "step": 945
+    },
+    {
+      "epoch": 0.2377332411886662,
+      "grad_norm": 0.4849649667739868,
+      "learning_rate": 8.773387608632867e-05,
+      "loss": 2.4643,
+      "step": 946
+    },
+    {
+      "epoch": 0.23798454482628636,
+      "grad_norm": 0.47490194439888,
+      "learning_rate": 8.770763361075949e-05,
+      "loss": 1.9425,
+      "step": 947
+    },
+    {
+      "epoch": 0.23823584846390652,
+      "grad_norm": 0.42135703563690186,
+      "learning_rate": 8.768136702698158e-05,
+      "loss": 2.2653,
+      "step": 948
+    },
+    {
+      "epoch": 0.23848715210152666,
+      "grad_norm": 0.3938981294631958,
+      "learning_rate": 8.765507635178832e-05,
+      "loss": 1.8336,
+      "step": 949
+    },
+    {
+      "epoch": 0.23873845573914682,
+      "grad_norm": 0.3590041399002075,
+      "learning_rate": 8.762876160198858e-05,
+      "loss": 2.083,
+      "step": 950
+    },
+    {
+      "epoch": 0.23898975937676697,
+      "grad_norm": 10.042701721191406,
+      "learning_rate": 8.760242279440657e-05,
+      "loss": 1.9433,
+      "step": 951
+    },
+    {
+      "epoch": 0.23924106301438713,
+      "grad_norm": 0.4117979407310486,
+      "learning_rate": 8.75760599458819e-05,
+      "loss": 1.9981,
+      "step": 952
+    },
+    {
+      "epoch": 0.2394923666520073,
+      "grad_norm": 0.3976224660873413,
+      "learning_rate": 8.754967307326951e-05,
+      "loss": 1.9221,
+      "step": 953
+    },
+    {
+      "epoch": 0.23974367028962745,
+      "grad_norm": 0.4007420241832733,
+      "learning_rate": 8.752326219343977e-05,
+      "loss": 2.1583,
+      "step": 954
+    },
+    {
+      "epoch": 0.23999497392724758,
+      "grad_norm": 0.43828290700912476,
+      "learning_rate": 8.74968273232783e-05,
+      "loss": 2.0958,
+      "step": 955
+    },
+    {
+      "epoch": 0.24024627756486774,
+      "grad_norm": 0.3099430501461029,
+      "learning_rate": 8.747036847968618e-05,
+      "loss": 2.0985,
+      "step": 956
+    },
+    {
+      "epoch": 0.2404975812024879,
+      "grad_norm": 0.2864762842655182,
+      "learning_rate": 8.744388567957971e-05,
+      "loss": 1.9034,
+      "step": 957
+    },
+    {
+      "epoch": 0.24074888484010806,
+      "grad_norm": 0.38019657135009766,
+      "learning_rate": 8.741737893989058e-05,
+      "loss": 2.2289,
+      "step": 958
+    },
+    {
+      "epoch": 0.24100018847772822,
+      "grad_norm": 0.506572425365448,
+      "learning_rate": 8.739084827756575e-05,
+      "loss": 2.3025,
+      "step": 959
+    },
+    {
+      "epoch": 0.24125149211534838,
+      "grad_norm": 0.4378896653652191,
+      "learning_rate": 8.736429370956746e-05,
+      "loss": 1.9396,
+      "step": 960
+    },
+    {
+      "epoch": 0.2415027957529685,
+      "grad_norm": 0.36668267846107483,
+      "learning_rate": 8.733771525287331e-05,
+      "loss": 2.7244,
+      "step": 961
+    },
+    {
+      "epoch": 0.24175409939058867,
+      "grad_norm": 0.3157023787498474,
+      "learning_rate": 8.731111292447605e-05,
+      "loss": 2.2407,
+      "step": 962
+    },
+    {
+      "epoch": 0.24200540302820883,
+      "grad_norm": 0.2563331127166748,
+      "learning_rate": 8.728448674138381e-05,
+      "loss": 2.4275,
+      "step": 963
+    },
+    {
+      "epoch": 0.242256706665829,
+      "grad_norm": 0.49870792031288147,
+      "learning_rate": 8.72578367206199e-05,
+      "loss": 2.0975,
+      "step": 964
+    },
+    {
+      "epoch": 0.24250801030344915,
+      "grad_norm": 0.328143835067749,
+      "learning_rate": 8.723116287922288e-05,
+      "loss": 2.0371,
+      "step": 965
+    },
+    {
+      "epoch": 0.2427593139410693,
+      "grad_norm": 0.40886473655700684,
+      "learning_rate": 8.72044652342466e-05,
+      "loss": 2.294,
+      "step": 966
+    },
+    {
+      "epoch": 0.24301061757868944,
+      "grad_norm": 0.39602747559547424,
+      "learning_rate": 8.717774380276002e-05,
+      "loss": 2.067,
+      "step": 967
+    },
+    {
+      "epoch": 0.2432619212163096,
+      "grad_norm": 0.33269646763801575,
+      "learning_rate": 8.715099860184743e-05,
+      "loss": 2.2003,
+      "step": 968
+    },
+    {
+      "epoch": 0.24351322485392976,
+      "grad_norm": 0.4260612726211548,
+      "learning_rate": 8.712422964860822e-05,
+      "loss": 2.4808,
+      "step": 969
+    },
+    {
+      "epoch": 0.24376452849154992,
+      "grad_norm": 0.3412139415740967,
+      "learning_rate": 8.7097436960157e-05,
+      "loss": 2.3457,
+      "step": 970
+    },
+    {
+      "epoch": 0.24401583212917008,
+      "grad_norm": 0.402170866727829,
+      "learning_rate": 8.707062055362359e-05,
+      "loss": 2.4638,
+      "step": 971
+    },
+    {
+      "epoch": 0.2442671357667902,
+      "grad_norm": 0.2597676217556,
+      "learning_rate": 8.70437804461529e-05,
+      "loss": 1.7648,
+      "step": 972
+    },
+    {
+      "epoch": 0.24451843940441037,
+      "grad_norm": 0.424844354391098,
+      "learning_rate": 8.701691665490504e-05,
+      "loss": 2.276,
+      "step": 973
+    },
+    {
+      "epoch": 0.24476974304203053,
+      "grad_norm": 0.1249924823641777,
+      "learning_rate": 8.699002919705527e-05,
+      "loss": 0.5754,
+      "step": 974
+    },
+    {
+      "epoch": 0.2450210466796507,
+      "grad_norm": 0.14425934851169586,
+      "learning_rate": 8.696311808979397e-05,
+      "loss": 0.9444,
+      "step": 975
+    },
+    {
+      "epoch": 0.24527235031727085,
+      "grad_norm": 0.5581234693527222,
+      "learning_rate": 8.693618335032663e-05,
+      "loss": 2.2831,
+      "step": 976
+    },
+    {
+      "epoch": 0.245523653954891,
+      "grad_norm": 0.17069493234157562,
+      "learning_rate": 8.690922499587387e-05,
+      "loss": 1.0646,
+      "step": 977
+    },
+    {
+      "epoch": 0.24577495759251114,
+      "grad_norm": 0.3384755551815033,
+      "learning_rate": 8.688224304367137e-05,
+      "loss": 1.8558,
+      "step": 978
+    },
+    {
+      "epoch": 0.2460262612301313,
+      "grad_norm": 0.3919682204723358,
+      "learning_rate": 8.685523751096994e-05,
+      "loss": 2.5674,
+      "step": 979
+    },
+    {
+      "epoch": 0.24627756486775146,
+      "grad_norm": 0.25237560272216797,
+      "learning_rate": 8.682820841503542e-05,
+      "loss": 1.7601,
+      "step": 980
+    },
+    {
+      "epoch": 0.24652886850537162,
+      "grad_norm": 0.3378541171550751,
+      "learning_rate": 8.680115577314877e-05,
+      "loss": 2.5141,
+      "step": 981
+    },
+    {
+      "epoch": 0.24678017214299178,
+      "grad_norm": 0.4153686761856079,
+      "learning_rate": 8.6774079602606e-05,
+      "loss": 1.1121,
+      "step": 982
+    },
+    {
+      "epoch": 0.24703147578061194,
+      "grad_norm": 0.3883324861526489,
+      "learning_rate": 8.67469799207181e-05,
+      "loss": 1.8358,
+      "step": 983
+    },
+    {
+      "epoch": 0.24728277941823207,
+      "grad_norm": 0.35025763511657715,
+      "learning_rate": 8.671985674481113e-05,
+      "loss": 1.7407,
+      "step": 984
+    },
+    {
+      "epoch": 0.24753408305585223,
+      "grad_norm": 0.1769835203886032,
+      "learning_rate": 8.66927100922262e-05,
+      "loss": 1.1858,
+      "step": 985
+    },
+    {
+      "epoch": 0.2477853866934724,
+      "grad_norm": 0.4593118131160736,
+      "learning_rate": 8.66655399803194e-05,
+      "loss": 2.4193,
+      "step": 986
+    },
+    {
+      "epoch": 0.24803669033109255,
+      "grad_norm": 0.24635930359363556,
+      "learning_rate": 8.663834642646178e-05,
+      "loss": 2.3587,
+      "step": 987
+    },
+    {
+      "epoch": 0.2482879939687127,
+      "grad_norm": 0.3709847927093506,
+      "learning_rate": 8.661112944803946e-05,
+      "loss": 2.5285,
+      "step": 988
+    },
+    {
+      "epoch": 0.24853929760633287,
+      "grad_norm": 0.3426840603351593,
+      "learning_rate": 8.65838890624535e-05,
+      "loss": 1.9006,
+      "step": 989
+    },
+    {
+      "epoch": 0.248790601243953,
+      "grad_norm": 0.29686790704727173,
+      "learning_rate": 8.655662528711987e-05,
+      "loss": 1.9789,
+      "step": 990
+    },
+    {
+      "epoch": 0.24904190488157316,
+      "grad_norm": 0.15335550904273987,
+      "learning_rate": 8.65293381394696e-05,
+      "loss": 1.379,
+      "step": 991
+    },
+    {
+      "epoch": 0.24929320851919332,
+      "grad_norm": 0.31200841069221497,
+      "learning_rate": 8.650202763694856e-05,
+      "loss": 2.6586,
+      "step": 992
+    },
+    {
+      "epoch": 0.24954451215681347,
+      "grad_norm": 0.28087717294692993,
+      "learning_rate": 8.647469379701765e-05,
+      "loss": 2.1206,
+      "step": 993
+    },
+    {
+      "epoch": 0.24979581579443363,
+      "grad_norm": 0.4623563289642334,
+      "learning_rate": 8.64473366371526e-05,
+      "loss": 2.3116,
+      "step": 994
+    },
+    {
+      "epoch": 0.25004711943205377,
+      "grad_norm": 0.3620995879173279,
+      "learning_rate": 8.641995617484411e-05,
+      "loss": 2.1746,
+      "step": 995
+    },
+    {
+      "epoch": 0.2502984230696739,
+      "grad_norm": 0.42614445090293884,
+      "learning_rate": 8.639255242759773e-05,
+      "loss": 2.1065,
+      "step": 996
+    },
+    {
+      "epoch": 0.2505497267072941,
+      "grad_norm": 0.5489148497581482,
+      "learning_rate": 8.636512541293396e-05,
+      "loss": 1.6889,
+      "step": 997
+    },
+    {
+      "epoch": 0.25080103034491424,
+      "grad_norm": 0.5294693112373352,
+      "learning_rate": 8.633767514838811e-05,
+      "loss": 1.9634,
+      "step": 998
+    },
+    {
+      "epoch": 0.2510523339825344,
+      "grad_norm": 0.2602441906929016,
+      "learning_rate": 8.631020165151041e-05,
+      "loss": 2.0264,
+      "step": 999
+    },
+    {
+      "epoch": 0.25130363762015456,
+      "grad_norm": 0.5113153457641602,
+      "learning_rate": 8.62827049398659e-05,
+      "loss": 2.47,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3979,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.536126594304095e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}