{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9987760097919217,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004895960832313341,
      "grad_norm": 0.03658522292971611,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.4365,
      "step": 1
    },
    {
      "epoch": 0.009791921664626682,
      "grad_norm": 0.035473328083753586,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.6677,
      "step": 2
    },
    {
      "epoch": 0.014687882496940025,
      "grad_norm": 0.030256139114499092,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.4084,
      "step": 3
    },
    {
      "epoch": 0.019583843329253364,
      "grad_norm": 0.04286854714155197,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.3422,
      "step": 4
    },
    {
      "epoch": 0.02447980416156671,
      "grad_norm": 0.06271344423294067,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.5332,
      "step": 5
    },
    {
      "epoch": 0.02937576499388005,
      "grad_norm": 0.09929833561182022,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.7114,
      "step": 6
    },
    {
      "epoch": 0.03427172582619339,
      "grad_norm": 0.08789193630218506,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.5457,
      "step": 7
    },
    {
      "epoch": 0.03916768665850673,
      "grad_norm": 0.0518026277422905,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.3846,
      "step": 8
    },
    {
      "epoch": 0.044063647490820076,
      "grad_norm": 0.042549822479486465,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.5076,
      "step": 9
    },
    {
      "epoch": 0.04895960832313342,
      "grad_norm": 0.051444098353385925,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.4675,
      "step": 10
    },
    {
      "epoch": 0.05385556915544676,
      "grad_norm": 0.08935613930225372,
      "learning_rate": 0.0001,
      "loss": 0.3412,
      "step": 11
    },
    {
      "epoch": 0.0587515299877601,
      "grad_norm": 0.10168597102165222,
      "learning_rate": 9.99933760728612e-05,
      "loss": 0.4016,
      "step": 12
    },
    {
      "epoch": 0.06364749082007344,
      "grad_norm": 0.1793641448020935,
      "learning_rate": 9.997350604650123e-05,
      "loss": 0.3738,
      "step": 13
    },
    {
      "epoch": 0.06854345165238677,
      "grad_norm": 0.24378074705600739,
      "learning_rate": 9.994039518562432e-05,
      "loss": 0.5781,
      "step": 14
    },
    {
      "epoch": 0.07343941248470012,
      "grad_norm": 0.08958296477794647,
      "learning_rate": 9.989405226318772e-05,
      "loss": 0.4275,
      "step": 15
    },
    {
      "epoch": 0.07833537331701346,
      "grad_norm": 0.16461962461471558,
      "learning_rate": 9.983448955807708e-05,
      "loss": 0.4416,
      "step": 16
    },
    {
      "epoch": 0.0832313341493268,
      "grad_norm": 0.2743546664714813,
      "learning_rate": 9.976172285185314e-05,
      "loss": 0.5598,
      "step": 17
    },
    {
      "epoch": 0.08812729498164015,
      "grad_norm": 0.05548033118247986,
      "learning_rate": 9.967577142457032e-05,
      "loss": 0.3509,
      "step": 18
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.10299059748649597,
      "learning_rate": 9.957665804966829e-05,
      "loss": 0.7287,
      "step": 19
    },
    {
      "epoch": 0.09791921664626684,
      "grad_norm": 0.1017536073923111,
      "learning_rate": 9.946440898793801e-05,
      "loss": 0.638,
      "step": 20
    },
    {
      "epoch": 0.10281517747858017,
      "grad_norm": 0.09810858964920044,
      "learning_rate": 9.933905398056372e-05,
      "loss": 0.4959,
      "step": 21
    },
    {
      "epoch": 0.10771113831089352,
      "grad_norm": 0.07071840763092041,
      "learning_rate": 9.920062624124282e-05,
      "loss": 0.5835,
      "step": 22
    },
    {
      "epoch": 0.11260709914320685,
      "grad_norm": 0.23103618621826172,
      "learning_rate": 9.904916244738571e-05,
      "loss": 0.7785,
      "step": 23
    },
    {
      "epoch": 0.1175030599755202,
      "grad_norm": 0.11760212481021881,
      "learning_rate": 9.888470273039775e-05,
      "loss": 0.3422,
      "step": 24
    },
    {
      "epoch": 0.12239902080783353,
      "grad_norm": 0.07370053976774216,
      "learning_rate": 9.870729066504629e-05,
      "loss": 0.3107,
      "step": 25
    },
    {
      "epoch": 0.12729498164014688,
      "grad_norm": 0.07285131514072418,
      "learning_rate": 9.851697325791505e-05,
      "loss": 0.4473,
      "step": 26
    },
    {
      "epoch": 0.13219094247246022,
      "grad_norm": 0.05355285108089447,
      "learning_rate": 9.831380093494957e-05,
      "loss": 0.3368,
      "step": 27
    },
    {
      "epoch": 0.13708690330477355,
      "grad_norm": 0.07507691532373428,
      "learning_rate": 9.809782752809644e-05,
      "loss": 0.3709,
      "step": 28
    },
    {
      "epoch": 0.1419828641370869,
      "grad_norm": 0.055260878056287766,
      "learning_rate": 9.786911026104007e-05,
      "loss": 0.4986,
      "step": 29
    },
    {
      "epoch": 0.14687882496940025,
      "grad_norm": 0.16646364331245422,
      "learning_rate": 9.762770973404094e-05,
      "loss": 0.6252,
      "step": 30
    },
    {
      "epoch": 0.15177478580171358,
      "grad_norm": 0.22445577383041382,
      "learning_rate": 9.737368990787916e-05,
      "loss": 0.4772,
      "step": 31
    },
    {
      "epoch": 0.15667074663402691,
      "grad_norm": 0.22849321365356445,
      "learning_rate": 9.710711808690754e-05,
      "loss": 0.381,
      "step": 32
    },
    {
      "epoch": 0.16156670746634028,
      "grad_norm": 0.05855480954051018,
      "learning_rate": 9.682806490121885e-05,
      "loss": 0.3429,
      "step": 33
    },
    {
      "epoch": 0.1664626682986536,
      "grad_norm": 0.06263675540685654,
      "learning_rate": 9.653660428793188e-05,
      "loss": 0.3927,
      "step": 34
    },
    {
      "epoch": 0.17135862913096694,
      "grad_norm": 0.277313232421875,
      "learning_rate": 9.623281347160127e-05,
      "loss": 0.9417,
      "step": 35
    },
    {
      "epoch": 0.1762545899632803,
      "grad_norm": 0.07143048942089081,
      "learning_rate": 9.591677294375636e-05,
      "loss": 0.4428,
      "step": 36
    },
    {
      "epoch": 0.18115055079559364,
      "grad_norm": 0.07706379145383835,
      "learning_rate": 9.558856644157432e-05,
      "loss": 0.3614,
      "step": 37
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 0.25005754828453064,
      "learning_rate": 9.52482809256934e-05,
      "loss": 0.6474,
      "step": 38
    },
    {
      "epoch": 0.1909424724602203,
      "grad_norm": 0.06337836384773254,
      "learning_rate": 9.489600655717217e-05,
      "loss": 0.4504,
      "step": 39
    },
    {
      "epoch": 0.19583843329253367,
      "grad_norm": 0.21187523007392883,
      "learning_rate": 9.453183667360062e-05,
      "loss": 0.6336,
      "step": 40
    },
    {
      "epoch": 0.200734394124847,
      "grad_norm": 0.1760697066783905,
      "learning_rate": 9.415586776436973e-05,
      "loss": 0.5665,
      "step": 41
    },
    {
      "epoch": 0.20563035495716034,
      "grad_norm": 0.10999207943677902,
      "learning_rate": 9.376819944510598e-05,
      "loss": 0.3929,
      "step": 42
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.09944123774766922,
      "learning_rate": 9.336893443127738e-05,
      "loss": 0.4195,
      "step": 43
    },
    {
      "epoch": 0.21542227662178703,
      "grad_norm": 0.1161671131849289,
      "learning_rate": 9.295817851097837e-05,
      "loss": 0.4643,
      "step": 44
    },
    {
      "epoch": 0.22031823745410037,
      "grad_norm": 0.11634702235460281,
      "learning_rate": 9.253604051690046e-05,
      "loss": 0.5375,
      "step": 45
    },
    {
      "epoch": 0.2252141982864137,
      "grad_norm": 0.07312487810850143,
      "learning_rate": 9.210263229749626e-05,
      "loss": 0.3291,
      "step": 46
    },
    {
      "epoch": 0.23011015911872704,
      "grad_norm": 0.08545304834842682,
      "learning_rate": 9.165806868734444e-05,
      "loss": 0.5543,
      "step": 47
    },
    {
      "epoch": 0.2350061199510404,
      "grad_norm": 0.15719719231128693,
      "learning_rate": 9.120246747672347e-05,
      "loss": 0.7045,
      "step": 48
    },
    {
      "epoch": 0.23990208078335373,
      "grad_norm": 0.11078042536973953,
      "learning_rate": 9.073594938040231e-05,
      "loss": 0.572,
      "step": 49
    },
    {
      "epoch": 0.24479804161566707,
      "grad_norm": 0.1860225647687912,
      "learning_rate": 9.025863800565613e-05,
      "loss": 0.7102,
      "step": 50
    },
    {
      "epoch": 0.24969400244798043,
      "grad_norm": 0.11442095041275024,
      "learning_rate": 8.977065981951566e-05,
      "loss": 0.5135,
      "step": 51
    },
    {
      "epoch": 0.25458996328029376,
      "grad_norm": 0.17788943648338318,
      "learning_rate": 8.927214411525895e-05,
      "loss": 0.4911,
      "step": 52
    },
    {
      "epoch": 0.2594859241126071,
      "grad_norm": 0.24241389334201813,
      "learning_rate": 8.876322297815405e-05,
      "loss": 0.6657,
      "step": 53
    },
    {
      "epoch": 0.26438188494492043,
      "grad_norm": 0.07881529629230499,
      "learning_rate": 8.824403125046225e-05,
      "loss": 0.4495,
      "step": 54
    },
    {
      "epoch": 0.2692778457772338,
      "grad_norm": 0.12150023877620697,
      "learning_rate": 8.771470649571056e-05,
      "loss": 0.3651,
      "step": 55
    },
    {
      "epoch": 0.2741738066095471,
      "grad_norm": 0.07412393391132355,
      "learning_rate": 8.717538896224332e-05,
      "loss": 0.3277,
      "step": 56
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.06195759400725365,
      "learning_rate": 8.662622154606237e-05,
      "loss": 0.361,
      "step": 57
    },
    {
      "epoch": 0.2839657282741738,
      "grad_norm": 0.06260855495929718,
      "learning_rate": 8.606734975296578e-05,
      "loss": 0.4541,
      "step": 58
    },
    {
      "epoch": 0.28886168910648713,
      "grad_norm": 0.22213032841682434,
      "learning_rate": 8.549892165999505e-05,
      "loss": 0.4133,
      "step": 59
    },
    {
      "epoch": 0.2937576499388005,
      "grad_norm": 0.08584744483232498,
      "learning_rate": 8.492108787620105e-05,
      "loss": 0.3445,
      "step": 60
    },
    {
      "epoch": 0.29865361077111385,
      "grad_norm": 0.07675541192293167,
      "learning_rate": 8.433400150273906e-05,
      "loss": 0.4518,
      "step": 61
    },
    {
      "epoch": 0.30354957160342716,
      "grad_norm": 0.07371577620506287,
      "learning_rate": 8.373781809230355e-05,
      "loss": 0.308,
      "step": 62
    },
    {
      "epoch": 0.3084455324357405,
      "grad_norm": 0.10943123698234558,
      "learning_rate": 8.313269560791342e-05,
      "loss": 0.7011,
      "step": 63
    },
    {
      "epoch": 0.31334149326805383,
      "grad_norm": 0.07427150011062622,
      "learning_rate": 8.251879438105854e-05,
      "loss": 0.3193,
      "step": 64
    },
    {
      "epoch": 0.3182374541003672,
      "grad_norm": 0.06212422996759415,
      "learning_rate": 8.189627706921877e-05,
      "loss": 0.3383,
      "step": 65
    },
    {
      "epoch": 0.32313341493268055,
      "grad_norm": 0.1618916094303131,
      "learning_rate": 8.126530861276677e-05,
      "loss": 0.5653,
      "step": 66
    },
    {
      "epoch": 0.32802937576499386,
      "grad_norm": 0.06333769857883453,
      "learning_rate": 8.062605619126584e-05,
      "loss": 0.3395,
      "step": 67
    },
    {
      "epoch": 0.3329253365973072,
      "grad_norm": 0.16169676184654236,
      "learning_rate": 7.997868917917453e-05,
      "loss": 0.7753,
      "step": 68
    },
    {
      "epoch": 0.3378212974296206,
      "grad_norm": 0.06635843217372894,
      "learning_rate": 7.932337910096961e-05,
      "loss": 0.3176,
      "step": 69
    },
    {
      "epoch": 0.3427172582619339,
      "grad_norm": 0.16557282209396362,
      "learning_rate": 7.866029958569956e-05,
      "loss": 0.4603,
      "step": 70
    },
    {
      "epoch": 0.34761321909424725,
      "grad_norm": 0.08522031456232071,
      "learning_rate": 7.798962632098024e-05,
      "loss": 0.5636,
      "step": 71
    },
    {
      "epoch": 0.3525091799265606,
      "grad_norm": 0.05583783611655235,
      "learning_rate": 7.73115370064452e-05,
      "loss": 0.2692,
      "step": 72
    },
    {
      "epoch": 0.3574051407588739,
      "grad_norm": 0.08181650191545486,
      "learning_rate": 7.6626211306663e-05,
      "loss": 0.6908,
      "step": 73
    },
    {
      "epoch": 0.3623011015911873,
      "grad_norm": 0.07728707045316696,
      "learning_rate": 7.59338308035337e-05,
      "loss": 0.4046,
      "step": 74
    },
    {
      "epoch": 0.3671970624235006,
      "grad_norm": 0.09214276075363159,
      "learning_rate": 7.523457894817745e-05,
      "loss": 0.816,
      "step": 75
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.07536876946687698,
      "learning_rate": 7.452864101232798e-05,
      "loss": 0.334,
      "step": 76
    },
    {
      "epoch": 0.3769889840881273,
      "grad_norm": 0.09890823811292648,
      "learning_rate": 7.381620403924333e-05,
      "loss": 0.4769,
      "step": 77
    },
    {
      "epoch": 0.3818849449204406,
      "grad_norm": 0.09237520396709442,
      "learning_rate": 7.30974567941475e-05,
      "loss": 0.4149,
      "step": 78
    },
    {
      "epoch": 0.386780905752754,
      "grad_norm": 0.14629222452640533,
      "learning_rate": 7.237258971421587e-05,
      "loss": 0.9113,
      "step": 79
    },
    {
      "epoch": 0.39167686658506734,
      "grad_norm": 0.0960475504398346,
      "learning_rate": 7.164179485811727e-05,
      "loss": 0.4643,
      "step": 80
    },
    {
      "epoch": 0.39657282741738065,
      "grad_norm": 0.07956342399120331,
      "learning_rate": 7.090526585512696e-05,
      "loss": 0.3902,
      "step": 81
    },
    {
      "epoch": 0.401468788249694,
      "grad_norm": 0.08580750226974487,
      "learning_rate": 7.016319785382296e-05,
      "loss": 0.5235,
      "step": 82
    },
    {
      "epoch": 0.40636474908200737,
      "grad_norm": 0.12661615014076233,
      "learning_rate": 6.941578747038023e-05,
      "loss": 0.4481,
      "step": 83
    },
    {
      "epoch": 0.4112607099143207,
      "grad_norm": 0.0649070218205452,
      "learning_rate": 6.866323273647563e-05,
      "loss": 0.4189,
      "step": 84
    },
    {
      "epoch": 0.41615667074663404,
      "grad_norm": 0.1020541861653328,
      "learning_rate": 6.79057330468182e-05,
      "loss": 0.4533,
      "step": 85
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.10127013921737671,
      "learning_rate": 6.7143489106318e-05,
      "loss": 0.522,
      "step": 86
    },
    {
      "epoch": 0.4259485924112607,
      "grad_norm": 0.14281871914863586,
      "learning_rate": 6.637670287690799e-05,
      "loss": 0.4772,
      "step": 87
    },
    {
      "epoch": 0.43084455324357407,
      "grad_norm": 0.065117247402668,
      "learning_rate": 6.560557752403277e-05,
      "loss": 0.5043,
      "step": 88
    },
    {
      "epoch": 0.4357405140758874,
      "grad_norm": 0.12292537093162537,
      "learning_rate": 6.483031736281843e-05,
      "loss": 0.4375,
      "step": 89
    },
    {
      "epoch": 0.44063647490820074,
      "grad_norm": 0.06215154007077217,
      "learning_rate": 6.40511278039378e-05,
      "loss": 0.3519,
      "step": 90
    },
    {
      "epoch": 0.4455324357405141,
      "grad_norm": 0.070677250623703,
      "learning_rate": 6.326821529918553e-05,
      "loss": 0.3407,
      "step": 91
    },
    {
      "epoch": 0.4504283965728274,
      "grad_norm": 0.06239693611860275,
      "learning_rate": 6.248178728677711e-05,
      "loss": 0.3799,
      "step": 92
    },
    {
      "epoch": 0.45532435740514077,
      "grad_norm": 0.07950470596551895,
      "learning_rate": 6.16920521363867e-05,
      "loss": 0.3063,
      "step": 93
    },
    {
      "epoch": 0.4602203182374541,
      "grad_norm": 0.06621692329645157,
      "learning_rate": 6.089921909393812e-05,
      "loss": 0.4386,
      "step": 94
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.3681151270866394,
      "learning_rate": 6.0103498226163603e-05,
      "loss": 0.5919,
      "step": 95
    },
    {
      "epoch": 0.4700122399020808,
      "grad_norm": 0.20958742499351501,
      "learning_rate": 5.93051003649452e-05,
      "loss": 0.4636,
      "step": 96
    },
    {
      "epoch": 0.4749082007343941,
      "grad_norm": 0.09456545114517212,
      "learning_rate": 5.850423705145334e-05,
      "loss": 0.4564,
      "step": 97
    },
    {
      "epoch": 0.47980416156670747,
      "grad_norm": 0.06202203407883644,
      "learning_rate": 5.770112048009747e-05,
      "loss": 0.3652,
      "step": 98
    },
    {
      "epoch": 0.4847001223990208,
      "grad_norm": 0.11682210117578506,
      "learning_rate": 5.68959634423037e-05,
      "loss": 0.5179,
      "step": 99
    },
    {
      "epoch": 0.48959608323133413,
      "grad_norm": 0.11552639305591583,
      "learning_rate": 5.60889792701342e-05,
      "loss": 0.467,
      "step": 100
    },
    {
      "epoch": 0.4944920440636475,
      "grad_norm": 0.053526621311903,
      "learning_rate": 5.52803817797633e-05,
      "loss": 0.2341,
      "step": 101
    },
    {
      "epoch": 0.49938800489596086,
      "grad_norm": 0.11124971508979797,
      "learning_rate": 5.4470385214825416e-05,
      "loss": 0.5641,
      "step": 102
    },
    {
      "epoch": 0.5042839657282742,
      "grad_norm": 0.1113283783197403,
      "learning_rate": 5.365920418964973e-05,
      "loss": 0.438,
      "step": 103
    },
    {
      "epoch": 0.5091799265605875,
      "grad_norm": 0.0814870297908783,
      "learning_rate": 5.28470536323965e-05,
      "loss": 0.4562,
      "step": 104
    },
    {
      "epoch": 0.5140758873929009,
      "grad_norm": 0.09781132638454437,
      "learning_rate": 5.2034148728110424e-05,
      "loss": 0.5587,
      "step": 105
    },
    {
      "epoch": 0.5189718482252142,
      "grad_norm": 0.11908842623233795,
      "learning_rate": 5.1220704861705774e-05,
      "loss": 0.7554,
      "step": 106
    },
    {
      "epoch": 0.5238678090575275,
      "grad_norm": 0.06143520399928093,
      "learning_rate": 5.0406937560898646e-05,
      "loss": 0.2544,
      "step": 107
    },
    {
      "epoch": 0.5287637698898409,
      "grad_norm": 0.08653257042169571,
      "learning_rate": 4.9593062439101365e-05,
      "loss": 0.5423,
      "step": 108
    },
    {
      "epoch": 0.5336597307221542,
      "grad_norm": 0.07373999804258347,
      "learning_rate": 4.877929513829424e-05,
      "loss": 0.3577,
      "step": 109
    },
    {
      "epoch": 0.5385556915544676,
      "grad_norm": 0.2530810236930847,
      "learning_rate": 4.796585127188958e-05,
      "loss": 0.454,
      "step": 110
    },
    {
      "epoch": 0.543451652386781,
      "grad_norm": 0.0999133363366127,
      "learning_rate": 4.715294636760352e-05,
      "loss": 0.4601,
      "step": 111
    },
    {
      "epoch": 0.5483476132190942,
      "grad_norm": 0.07089534401893616,
      "learning_rate": 4.634079581035029e-05,
      "loss": 0.2958,
      "step": 112
    },
    {
      "epoch": 0.5532435740514076,
      "grad_norm": 0.15452376008033752,
      "learning_rate": 4.55296147851746e-05,
      "loss": 0.5048,
      "step": 113
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.12181607633829117,
      "learning_rate": 4.471961822023671e-05,
      "loss": 0.5176,
      "step": 114
    },
    {
      "epoch": 0.5630354957160343,
      "grad_norm": 0.07680846750736237,
      "learning_rate": 4.391102072986581e-05,
      "loss": 0.2968,
      "step": 115
    },
    {
      "epoch": 0.5679314565483476,
      "grad_norm": 0.071733757853508,
      "learning_rate": 4.3104036557696295e-05,
      "loss": 0.6076,
      "step": 116
    },
    {
      "epoch": 0.572827417380661,
      "grad_norm": 0.07903078943490982,
      "learning_rate": 4.229887951990255e-05,
      "loss": 0.5459,
      "step": 117
    },
    {
      "epoch": 0.5777233782129743,
      "grad_norm": 0.09677067399024963,
      "learning_rate": 4.149576294854668e-05,
      "loss": 0.44,
      "step": 118
    },
    {
      "epoch": 0.5826193390452876,
      "grad_norm": 0.09369610249996185,
      "learning_rate": 4.069489963505482e-05,
      "loss": 0.6845,
      "step": 119
    },
    {
      "epoch": 0.587515299877601,
      "grad_norm": 0.06507845222949982,
      "learning_rate": 3.98965017738364e-05,
      "loss": 0.3054,
      "step": 120
    },
    {
      "epoch": 0.5924112607099143,
      "grad_norm": 0.1210411936044693,
      "learning_rate": 3.9100780906061896e-05,
      "loss": 0.4176,
      "step": 121
    },
    {
      "epoch": 0.5973072215422277,
      "grad_norm": 0.07214700430631638,
      "learning_rate": 3.83079478636133e-05,
      "loss": 0.4256,
      "step": 122
    },
    {
      "epoch": 0.602203182374541,
      "grad_norm": 0.07553218305110931,
      "learning_rate": 3.7518212713222906e-05,
      "loss": 0.5853,
      "step": 123
    },
    {
      "epoch": 0.6070991432068543,
      "grad_norm": 0.08783773332834244,
      "learning_rate": 3.673178470081448e-05,
      "loss": 0.5636,
      "step": 124
    },
    {
      "epoch": 0.6119951040391677,
      "grad_norm": 0.07783231139183044,
      "learning_rate": 3.594887219606221e-05,
      "loss": 0.6704,
      "step": 125
    },
    {
      "epoch": 0.616891064871481,
      "grad_norm": 0.14530447125434875,
      "learning_rate": 3.516968263718159e-05,
      "loss": 0.4881,
      "step": 126
    },
    {
      "epoch": 0.6217870257037944,
      "grad_norm": 0.23401491343975067,
      "learning_rate": 3.439442247596724e-05,
      "loss": 0.4918,
      "step": 127
    },
    {
      "epoch": 0.6266829865361077,
      "grad_norm": 0.07801464200019836,
      "learning_rate": 3.3623297123092006e-05,
      "loss": 0.5203,
      "step": 128
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.1613989621400833,
      "learning_rate": 3.285651089368202e-05,
      "loss": 0.7026,
      "step": 129
    },
    {
      "epoch": 0.6364749082007344,
      "grad_norm": 0.12846659123897552,
      "learning_rate": 3.209426695318182e-05,
      "loss": 0.4857,
      "step": 130
    },
    {
      "epoch": 0.6413708690330477,
      "grad_norm": 0.11735550314188004,
      "learning_rate": 3.133676726352438e-05,
      "loss": 0.542,
      "step": 131
    },
    {
      "epoch": 0.6462668298653611,
      "grad_norm": 0.09229713678359985,
      "learning_rate": 3.0584212529619775e-05,
      "loss": 0.4807,
      "step": 132
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.1247001513838768,
      "learning_rate": 2.9836802146177034e-05,
      "loss": 0.5123,
      "step": 133
    },
    {
      "epoch": 0.6560587515299877,
      "grad_norm": 0.12723222374916077,
      "learning_rate": 2.9094734144873036e-05,
      "loss": 0.7619,
      "step": 134
    },
    {
      "epoch": 0.6609547123623011,
      "grad_norm": 0.06785371154546738,
      "learning_rate": 2.835820514188273e-05,
      "loss": 0.488,
      "step": 135
    },
    {
      "epoch": 0.6658506731946144,
      "grad_norm": 0.07913102209568024,
      "learning_rate": 2.7627410285784163e-05,
      "loss": 0.5913,
      "step": 136
    },
    {
      "epoch": 0.6707466340269278,
      "grad_norm": 0.10038694739341736,
      "learning_rate": 2.6902543205852492e-05,
      "loss": 0.6774,
      "step": 137
    },
    {
      "epoch": 0.6756425948592412,
      "grad_norm": 0.06108963489532471,
      "learning_rate": 2.618379596075668e-05,
      "loss": 0.304,
      "step": 138
    },
    {
      "epoch": 0.6805385556915544,
      "grad_norm": 0.06344286352396011,
      "learning_rate": 2.5471358987672017e-05,
      "loss": 0.394,
      "step": 139
    },
    {
      "epoch": 0.6854345165238678,
      "grad_norm": 0.1912202686071396,
      "learning_rate": 2.476542105182254e-05,
      "loss": 0.4608,
      "step": 140
    },
    {
      "epoch": 0.6903304773561811,
      "grad_norm": 0.0918666198849678,
      "learning_rate": 2.4066169196466326e-05,
      "loss": 0.7695,
      "step": 141
    },
    {
      "epoch": 0.6952264381884945,
      "grad_norm": 0.09997723251581192,
      "learning_rate": 2.3373788693337024e-05,
      "loss": 0.5197,
      "step": 142
    },
    {
      "epoch": 0.7001223990208079,
      "grad_norm": 0.08384369313716888,
      "learning_rate": 2.268846299355481e-05,
      "loss": 0.4842,
      "step": 143
    },
    {
      "epoch": 0.7050183598531212,
      "grad_norm": 0.14654509723186493,
      "learning_rate": 2.2010373679019776e-05,
      "loss": 0.4083,
      "step": 144
    },
    {
      "epoch": 0.7099143206854345,
      "grad_norm": 0.23128315806388855,
      "learning_rate": 2.133970041430044e-05,
      "loss": 0.6992,
      "step": 145
    },
    {
      "epoch": 0.7148102815177478,
      "grad_norm": 0.09588706493377686,
      "learning_rate": 2.067662089903039e-05,
      "loss": 0.5621,
      "step": 146
    },
    {
      "epoch": 0.7197062423500612,
      "grad_norm": 0.10101126879453659,
      "learning_rate": 2.002131082082549e-05,
      "loss": 0.4135,
      "step": 147
    },
    {
      "epoch": 0.7246022031823746,
      "grad_norm": 0.15346215665340424,
      "learning_rate": 1.937394380873418e-05,
      "loss": 0.5933,
      "step": 148
    },
    {
      "epoch": 0.7294981640146879,
      "grad_norm": 0.07918453961610794,
      "learning_rate": 1.873469138723325e-05,
      "loss": 0.3081,
      "step": 149
    },
    {
      "epoch": 0.7343941248470012,
      "grad_norm": 0.0649547427892685,
      "learning_rate": 1.8103722930781247e-05,
      "loss": 0.407,
      "step": 150
    },
    {
      "epoch": 0.7392900856793145,
      "grad_norm": 0.07419539242982864,
      "learning_rate": 1.748120561894147e-05,
      "loss": 0.7313,
      "step": 151
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.09094121307134628,
      "learning_rate": 1.6867304392086575e-05,
      "loss": 0.3871,
      "step": 152
    },
    {
      "epoch": 0.7490820073439413,
      "grad_norm": 0.10387779027223587,
      "learning_rate": 1.6262181907696454e-05,
      "loss": 0.5026,
      "step": 153
    },
    {
      "epoch": 0.7539779681762546,
      "grad_norm": 0.22015878558158875,
      "learning_rate": 1.5665998497260958e-05,
      "loss": 0.4817,
      "step": 154
    },
    {
      "epoch": 0.758873929008568,
      "grad_norm": 0.06719034910202026,
      "learning_rate": 1.5078912123798961e-05,
      "loss": 0.5186,
      "step": 155
    },
    {
      "epoch": 0.7637698898408812,
      "grad_norm": 0.07599823921918869,
      "learning_rate": 1.4501078340004953e-05,
      "loss": 0.4275,
      "step": 156
    },
    {
      "epoch": 0.7686658506731946,
      "grad_norm": 0.1412460058927536,
      "learning_rate": 1.3932650247034218e-05,
      "loss": 0.626,
      "step": 157
    },
    {
      "epoch": 0.773561811505508,
      "grad_norm": 0.1420402228832245,
      "learning_rate": 1.337377845393763e-05,
      "loss": 0.5193,
      "step": 158
    },
    {
      "epoch": 0.7784577723378213,
      "grad_norm": 0.09665469825267792,
      "learning_rate": 1.2824611037756684e-05,
      "loss": 0.5848,
      "step": 159
    },
    {
      "epoch": 0.7833537331701347,
      "grad_norm": 0.06045060604810715,
      "learning_rate": 1.2285293504289447e-05,
      "loss": 0.2627,
      "step": 160
    },
    {
      "epoch": 0.7882496940024479,
      "grad_norm": 0.08824898302555084,
      "learning_rate": 1.1755968749537754e-05,
      "loss": 0.5949,
      "step": 161
    },
    {
      "epoch": 0.7931456548347613,
      "grad_norm": 0.13809214532375336,
      "learning_rate": 1.1236777021845956e-05,
      "loss": 0.7397,
      "step": 162
    },
    {
      "epoch": 0.7980416156670747,
      "grad_norm": 0.12224618345499039,
      "learning_rate": 1.0727855884741056e-05,
      "loss": 0.553,
      "step": 163
    },
    {
      "epoch": 0.802937576499388,
      "grad_norm": 0.07628399133682251,
      "learning_rate": 1.022934018048432e-05,
      "loss": 0.3447,
      "step": 164
    },
    {
      "epoch": 0.8078335373317014,
      "grad_norm": 0.06734273582696915,
      "learning_rate": 9.741361994343867e-06,
      "loss": 0.7473,
      "step": 165
    },
    {
      "epoch": 0.8127294981640147,
      "grad_norm": 0.0891406238079071,
      "learning_rate": 9.264050619597697e-06,
      "loss": 0.5008,
      "step": 166
    },
    {
      "epoch": 0.817625458996328,
      "grad_norm": 0.08535400032997131,
      "learning_rate": 8.797532523276542e-06,
      "loss": 0.4171,
      "step": 167
    },
    {
      "epoch": 0.8225214198286414,
      "grad_norm": 0.3191221058368683,
      "learning_rate": 8.341931312655582e-06,
      "loss": 0.7633,
      "step": 168
    },
    {
      "epoch": 0.8274173806609547,
      "grad_norm": 0.11288396269083023,
      "learning_rate": 7.897367702503756e-06,
      "loss": 0.4054,
      "step": 169
    },
    {
      "epoch": 0.8323133414932681,
      "grad_norm": 0.09110561013221741,
      "learning_rate": 7.463959483099547e-06,
      "loss": 0.4642,
      "step": 170
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.08345432579517365,
      "learning_rate": 7.041821489021639e-06,
      "loss": 0.3769,
      "step": 171
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.19082055985927582,
      "learning_rate": 6.631065568722633e-06,
      "loss": 0.4487,
      "step": 172
    },
    {
      "epoch": 0.847001223990208,
      "grad_norm": 0.06951060146093369,
      "learning_rate": 6.231800554894029e-06,
      "loss": 0.2915,
      "step": 173
    },
    {
      "epoch": 0.8518971848225214,
      "grad_norm": 0.08803869783878326,
      "learning_rate": 5.844132235630273e-06,
      "loss": 0.5065,
      "step": 174
    },
    {
      "epoch": 0.8567931456548348,
      "grad_norm": 0.06576603651046753,
      "learning_rate": 5.468163326399389e-06,
      "loss": 0.4307,
      "step": 175
    },
    {
      "epoch": 0.8616891064871481,
      "grad_norm": 0.11544258892536163,
      "learning_rate": 5.103993442827831e-06,
      "loss": 0.6356,
      "step": 176
    },
    {
      "epoch": 0.8665850673194615,
      "grad_norm": 0.08324428647756577,
      "learning_rate": 4.751719074306604e-06,
      "loss": 0.5426,
      "step": 177
    },
    {
      "epoch": 0.8714810281517748,
      "grad_norm": 0.0815030187368393,
      "learning_rate": 4.411433558425698e-06,
      "loss": 0.4278,
      "step": 178
    },
    {
      "epoch": 0.8763769889840881,
      "grad_norm": 0.06282901763916016,
      "learning_rate": 4.083227056243644e-06,
      "loss": 0.2847,
      "step": 179
    },
    {
      "epoch": 0.8812729498164015,
      "grad_norm": 0.07095268368721008,
      "learning_rate": 3.767186528398725e-06,
      "loss": 0.3923,
      "step": 180
    },
    {
      "epoch": 0.8861689106487148,
      "grad_norm": 0.09169545769691467,
      "learning_rate": 3.4633957120681293e-06,
      "loss": 0.5155,
      "step": 181
    },
    {
      "epoch": 0.8910648714810282,
      "grad_norm": 0.24409419298171997,
      "learning_rate": 3.1719350987811534e-06,
      "loss": 0.5456,
      "step": 182
    },
    {
      "epoch": 0.8959608323133414,
      "grad_norm": 0.08426636457443237,
      "learning_rate": 2.8928819130924657e-06,
      "loss": 0.5679,
      "step": 183
    },
    {
      "epoch": 0.9008567931456548,
      "grad_norm": 0.23968654870986938,
      "learning_rate": 2.6263100921208482e-06,
      "loss": 0.4841,
      "step": 184
    },
    {
      "epoch": 0.9057527539779682,
      "grad_norm": 0.10230088233947754,
      "learning_rate": 2.372290265959065e-06,
      "loss": 0.6307,
      "step": 185
    },
    {
      "epoch": 0.9106487148102815,
      "grad_norm": 0.07012014836072922,
      "learning_rate": 2.130889738959946e-06,
      "loss": 0.489,
      "step": 186
    },
    {
      "epoch": 0.9155446756425949,
      "grad_norm": 0.0840025320649147,
      "learning_rate": 1.9021724719035628e-06,
      "loss": 0.3811,
      "step": 187
    },
    {
      "epoch": 0.9204406364749081,
      "grad_norm": 0.06925945729017258,
      "learning_rate": 1.6861990650504255e-06,
      "loss": 0.3263,
      "step": 188
    },
    {
      "epoch": 0.9253365973072215,
      "grad_norm": 0.07820533215999603,
      "learning_rate": 1.4830267420849585e-06,
      "loss": 0.5176,
      "step": 189
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.10654879361391068,
      "learning_rate": 1.292709334953729e-06,
      "loss": 0.5375,
      "step": 190
    },
    {
      "epoch": 0.9351285189718482,
      "grad_norm": 0.06305018812417984,
      "learning_rate": 1.1152972696022445e-06,
      "loss": 0.4505,
      "step": 191
    },
    {
      "epoch": 0.9400244798041616,
      "grad_norm": 0.08257196098566055,
      "learning_rate": 9.508375526142976e-07,
      "loss": 0.4773,
      "step": 192
    },
    {
      "epoch": 0.944920440636475,
      "grad_norm": 0.0973181426525116,
      "learning_rate": 7.993737587571826e-07,
      "loss": 0.4163,
      "step": 193
    },
    {
      "epoch": 0.9498164014687882,
      "grad_norm": 0.1023239865899086,
      "learning_rate": 6.609460194362927e-07,
      "loss": 0.4597,
      "step": 194
    },
    {
      "epoch": 0.9547123623011016,
      "grad_norm": 0.09319756925106049,
      "learning_rate": 5.355910120620034e-07,
      "loss": 0.5937,
      "step": 195
    },
    {
      "epoch": 0.9596083231334149,
      "grad_norm": 0.07661579549312592,
      "learning_rate": 4.233419503317182e-07,
      "loss": 0.6014,
      "step": 196
    },
    {
      "epoch": 0.9645042839657283,
      "grad_norm": 0.07042757421731949,
      "learning_rate": 3.242285754296859e-07,
      "loss": 0.4437,
      "step": 197
    },
    {
      "epoch": 0.9694002447980417,
      "grad_norm": 0.07441776990890503,
      "learning_rate": 2.3827714814686486e-07,
      "loss": 0.2822,
      "step": 198
    },
    {
      "epoch": 0.9742962056303549,
      "grad_norm": 0.06588555127382278,
      "learning_rate": 1.655104419229281e-07,
      "loss": 0.3,
      "step": 199
    },
    {
      "epoch": 0.9791921664626683,
      "grad_norm": 0.06164183467626572,
      "learning_rate": 1.059477368122841e-07,
      "loss": 0.3109,
      "step": 200
    },
    {
      "epoch": 0.9840881272949816,
      "grad_norm": 0.13465183973312378,
      "learning_rate": 5.960481437568555e-08,
      "loss": 0.8621,
      "step": 201
    },
    {
      "epoch": 0.988984088127295,
      "grad_norm": 0.06679002195596695,
      "learning_rate": 2.649395349879069e-08,
      "loss": 0.5233,
      "step": 202
    },
    {
      "epoch": 0.9938800489596084,
      "grad_norm": 0.07267452776432037,
      "learning_rate": 6.623927138804664e-09,
      "loss": 0.59,
      "step": 203
    },
    {
      "epoch": 0.9987760097919217,
      "grad_norm": 0.07579752057790756,
      "learning_rate": 0.0,
      "loss": 0.4031,
      "step": 204
    },
    {
      "epoch": 0.9987760097919217,
      "step": 204,
      "total_flos": 1.7231028658783027e+17,
      "train_loss": 0.4889225305295458,
      "train_runtime": 5417.273,
      "train_samples_per_second": 0.151,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 1,
  "max_steps": 204,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7231028658783027e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}