{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9923838537699923,
  "eval_steps": 500,
  "global_step": 218,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00913937547600914,
      "grad_norm": 5.803379058837891,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.7449,
      "step": 1
    },
    {
      "epoch": 0.01827875095201828,
      "grad_norm": 5.878463268280029,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.7528,
      "step": 2
    },
    {
      "epoch": 0.027418126428027417,
      "grad_norm": 5.657175064086914,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.7296,
      "step": 3
    },
    {
      "epoch": 0.03655750190403656,
      "grad_norm": 5.441561698913574,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.7337,
      "step": 4
    },
    {
      "epoch": 0.0456968773800457,
      "grad_norm": 4.161638259887695,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.7005,
      "step": 5
    },
    {
      "epoch": 0.05483625285605483,
      "grad_norm": 2.691371202468872,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.666,
      "step": 6
    },
    {
      "epoch": 0.06397562833206398,
      "grad_norm": 2.3983771800994873,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.6625,
      "step": 7
    },
    {
      "epoch": 0.07311500380807312,
      "grad_norm": 4.262558460235596,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.6645,
      "step": 8
    },
    {
      "epoch": 0.08225437928408226,
      "grad_norm": 4.646894931793213,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.6624,
      "step": 9
    },
    {
      "epoch": 0.0913937547600914,
      "grad_norm": 3.8496811389923096,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.6186,
      "step": 10
    },
    {
      "epoch": 0.10053313023610053,
      "grad_norm": 4.285327911376953,
      "learning_rate": 1e-05,
      "loss": 0.6059,
      "step": 11
    },
    {
      "epoch": 0.10967250571210967,
      "grad_norm": 3.3189685344696045,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.5763,
      "step": 12
    },
    {
      "epoch": 0.1188118811881188,
      "grad_norm": 2.195256233215332,
      "learning_rate": 1.181818181818182e-05,
      "loss": 0.5776,
      "step": 13
    },
    {
      "epoch": 0.12795125666412796,
      "grad_norm": 2.200354814529419,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 0.5703,
      "step": 14
    },
    {
      "epoch": 0.1370906321401371,
      "grad_norm": 3.10870099067688,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.5697,
      "step": 15
    },
    {
      "epoch": 0.14623000761614624,
      "grad_norm": 2.311767101287842,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.5148,
      "step": 16
    },
    {
      "epoch": 0.15536938309215537,
      "grad_norm": 1.341518759727478,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.5402,
      "step": 17
    },
    {
      "epoch": 0.16450875856816452,
      "grad_norm": 1.3586676120758057,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 0.5223,
      "step": 18
    },
    {
      "epoch": 0.17364813404417365,
      "grad_norm": 1.9843041896820068,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 0.4927,
      "step": 19
    },
    {
      "epoch": 0.1827875095201828,
      "grad_norm": 1.5958917140960693,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.5186,
      "step": 20
    },
    {
      "epoch": 0.19192688499619193,
      "grad_norm": 0.8092266321182251,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 0.4834,
      "step": 21
    },
    {
      "epoch": 0.20106626047220105,
      "grad_norm": 1.3990743160247803,
      "learning_rate": 2e-05,
      "loss": 0.4879,
      "step": 22
    },
    {
      "epoch": 0.2102056359482102,
      "grad_norm": 1.377229928970337,
      "learning_rate": 1.9998715457999313e-05,
      "loss": 0.493,
      "step": 23
    },
    {
      "epoch": 0.21934501142421933,
      "grad_norm": 0.9455161094665527,
      "learning_rate": 1.999486216200688e-05,
      "loss": 0.477,
      "step": 24
    },
    {
      "epoch": 0.2284843869002285,
      "grad_norm": 0.8552016615867615,
      "learning_rate": 1.9988441101966807e-05,
      "loss": 0.4986,
      "step": 25
    },
    {
      "epoch": 0.2376237623762376,
      "grad_norm": 0.9342105984687805,
      "learning_rate": 1.9979453927503366e-05,
      "loss": 0.5122,
      "step": 26
    },
    {
      "epoch": 0.24676313785224677,
      "grad_norm": 0.8206979632377625,
      "learning_rate": 1.9967902947497158e-05,
      "loss": 0.4918,
      "step": 27
    },
    {
      "epoch": 0.2559025133282559,
      "grad_norm": 0.8245558142662048,
      "learning_rate": 1.9953791129491985e-05,
      "loss": 0.4861,
      "step": 28
    },
    {
      "epoch": 0.26504188880426505,
      "grad_norm": 0.6446841359138489,
      "learning_rate": 1.9937122098932428e-05,
      "loss": 0.4633,
      "step": 29
    },
    {
      "epoch": 0.2741812642802742,
      "grad_norm": 0.7485717535018921,
      "learning_rate": 1.991790013823246e-05,
      "loss": 0.4867,
      "step": 30
    },
    {
      "epoch": 0.2833206397562833,
      "grad_norm": 0.8434760570526123,
      "learning_rate": 1.9896130185675263e-05,
      "loss": 0.4833,
      "step": 31
    },
    {
      "epoch": 0.2924600152322925,
      "grad_norm": 0.6975553631782532,
      "learning_rate": 1.9871817834144506e-05,
      "loss": 0.4642,
      "step": 32
    },
    {
      "epoch": 0.3015993907083016,
      "grad_norm": 0.7539258003234863,
      "learning_rate": 1.9844969329687526e-05,
      "loss": 0.4796,
      "step": 33
    },
    {
      "epoch": 0.31073876618431073,
      "grad_norm": 0.661018967628479,
      "learning_rate": 1.9815591569910654e-05,
      "loss": 0.4542,
      "step": 34
    },
    {
      "epoch": 0.31987814166031986,
      "grad_norm": 0.7433336973190308,
      "learning_rate": 1.9783692102207156e-05,
      "loss": 0.4845,
      "step": 35
    },
    {
      "epoch": 0.32901751713632904,
      "grad_norm": 0.6025668382644653,
      "learning_rate": 1.9749279121818235e-05,
      "loss": 0.424,
      "step": 36
    },
    {
      "epoch": 0.33815689261233817,
      "grad_norm": 0.7638101577758789,
      "learning_rate": 1.971236146972764e-05,
      "loss": 0.4267,
      "step": 37
    },
    {
      "epoch": 0.3472962680883473,
      "grad_norm": 0.6455888748168945,
      "learning_rate": 1.9672948630390296e-05,
      "loss": 0.4232,
      "step": 38
    },
    {
      "epoch": 0.3564356435643564,
      "grad_norm": 0.7659748792648315,
      "learning_rate": 1.9631050729295705e-05,
      "loss": 0.4801,
      "step": 39
    },
    {
      "epoch": 0.3655750190403656,
      "grad_norm": 0.7797123193740845,
      "learning_rate": 1.9586678530366607e-05,
      "loss": 0.4514,
      "step": 40
    },
    {
      "epoch": 0.3747143945163747,
      "grad_norm": 0.5871541500091553,
      "learning_rate": 1.953984343319364e-05,
      "loss": 0.433,
      "step": 41
    },
    {
      "epoch": 0.38385376999238385,
      "grad_norm": 0.668601930141449,
      "learning_rate": 1.949055747010669e-05,
      "loss": 0.4583,
      "step": 42
    },
    {
      "epoch": 0.392993145468393,
      "grad_norm": 0.7493565082550049,
      "learning_rate": 1.9438833303083677e-05,
      "loss": 0.459,
      "step": 43
    },
    {
      "epoch": 0.4021325209444021,
      "grad_norm": 0.6386315226554871,
      "learning_rate": 1.9384684220497605e-05,
      "loss": 0.4239,
      "step": 44
    },
    {
      "epoch": 0.4112718964204113,
      "grad_norm": 0.7803364396095276,
      "learning_rate": 1.932812413370265e-05,
      "loss": 0.475,
      "step": 45
    },
    {
      "epoch": 0.4204112718964204,
      "grad_norm": 0.6278897523880005,
      "learning_rate": 1.926916757346022e-05,
      "loss": 0.433,
      "step": 46
    },
    {
      "epoch": 0.42955064737242954,
      "grad_norm": 0.6488973498344421,
      "learning_rate": 1.9207829686205882e-05,
      "loss": 0.4262,
      "step": 47
    },
    {
      "epoch": 0.43869002284843867,
      "grad_norm": 0.7175928950309753,
      "learning_rate": 1.9144126230158127e-05,
      "loss": 0.447,
      "step": 48
    },
    {
      "epoch": 0.44782939832444785,
      "grad_norm": 0.7046955823898315,
      "learning_rate": 1.9078073571269922e-05,
      "loss": 0.4256,
      "step": 49
    },
    {
      "epoch": 0.456968773800457,
      "grad_norm": 0.6800966262817383,
      "learning_rate": 1.900968867902419e-05,
      "loss": 0.4295,
      "step": 50
    },
    {
      "epoch": 0.4661081492764661,
      "grad_norm": 0.6840780973434448,
      "learning_rate": 1.8938989122074195e-05,
      "loss": 0.4383,
      "step": 51
    },
    {
      "epoch": 0.4752475247524752,
      "grad_norm": 0.85346519947052,
      "learning_rate": 1.8865993063730003e-05,
      "loss": 0.456,
      "step": 52
    },
    {
      "epoch": 0.4843869002284844,
      "grad_norm": 0.627661943435669,
      "learning_rate": 1.8790719257292175e-05,
      "loss": 0.4598,
      "step": 53
    },
    {
      "epoch": 0.49352627570449353,
      "grad_norm": 0.865216851234436,
      "learning_rate": 1.8713187041233896e-05,
      "loss": 0.4547,
      "step": 54
    },
    {
      "epoch": 0.5026656511805027,
      "grad_norm": 0.7087027430534363,
      "learning_rate": 1.8633416334232754e-05,
      "loss": 0.4261,
      "step": 55
    },
    {
      "epoch": 0.5118050266565118,
      "grad_norm": 0.8220586180686951,
      "learning_rate": 1.8551427630053464e-05,
      "loss": 0.4589,
      "step": 56
    },
    {
      "epoch": 0.5209444021325209,
      "grad_norm": 0.6183983087539673,
      "learning_rate": 1.8467241992282842e-05,
      "loss": 0.4105,
      "step": 57
    },
    {
      "epoch": 0.5300837776085301,
      "grad_norm": 0.7341344952583313,
      "learning_rate": 1.8380881048918406e-05,
      "loss": 0.4435,
      "step": 58
    },
    {
      "epoch": 0.5392231530845393,
      "grad_norm": 0.86666339635849,
      "learning_rate": 1.8292366986811952e-05,
      "loss": 0.4148,
      "step": 59
    },
    {
      "epoch": 0.5483625285605483,
      "grad_norm": 0.6903016567230225,
      "learning_rate": 1.820172254596956e-05,
      "loss": 0.4356,
      "step": 60
    },
    {
      "epoch": 0.5575019040365575,
      "grad_norm": 0.7348374128341675,
      "learning_rate": 1.8108971013709512e-05,
      "loss": 0.419,
      "step": 61
    },
    {
      "epoch": 0.5666412795125666,
      "grad_norm": 0.6142125725746155,
      "learning_rate": 1.8014136218679566e-05,
      "loss": 0.4238,
      "step": 62
    },
    {
      "epoch": 0.5757806549885758,
      "grad_norm": 0.689704954624176,
      "learning_rate": 1.79172425247352e-05,
      "loss": 0.4208,
      "step": 63
    },
    {
      "epoch": 0.584920030464585,
      "grad_norm": 0.6363848447799683,
      "learning_rate": 1.78183148246803e-05,
      "loss": 0.4212,
      "step": 64
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 0.6809417605400085,
      "learning_rate": 1.771737853387202e-05,
      "loss": 0.4396,
      "step": 65
    },
    {
      "epoch": 0.6031987814166032,
      "grad_norm": 0.6080987453460693,
      "learning_rate": 1.7614459583691346e-05,
      "loss": 0.4359,
      "step": 66
    },
    {
      "epoch": 0.6123381568926123,
      "grad_norm": 0.5947669744491577,
      "learning_rate": 1.7509584414881114e-05,
      "loss": 0.4328,
      "step": 67
    },
    {
      "epoch": 0.6214775323686215,
      "grad_norm": 0.6680779457092285,
      "learning_rate": 1.7402779970753156e-05,
      "loss": 0.4424,
      "step": 68
    },
    {
      "epoch": 0.6306169078446306,
      "grad_norm": 0.5437312722206116,
      "learning_rate": 1.7294073690266343e-05,
      "loss": 0.4445,
      "step": 69
    },
    {
      "epoch": 0.6397562833206397,
      "grad_norm": 0.5836828947067261,
      "learning_rate": 1.7183493500977277e-05,
      "loss": 0.4296,
      "step": 70
    },
    {
      "epoch": 0.6488956587966489,
      "grad_norm": 0.6663245558738708,
      "learning_rate": 1.7071067811865477e-05,
      "loss": 0.4309,
      "step": 71
    },
    {
      "epoch": 0.6580350342726581,
      "grad_norm": 0.5611643195152283,
      "learning_rate": 1.6956825506034866e-05,
      "loss": 0.4192,
      "step": 72
    },
    {
      "epoch": 0.6671744097486672,
      "grad_norm": 0.63041752576828,
      "learning_rate": 1.6840795933293464e-05,
      "loss": 0.4201,
      "step": 73
    },
    {
      "epoch": 0.6763137852246763,
      "grad_norm": 0.6765708923339844,
      "learning_rate": 1.672300890261317e-05,
      "loss": 0.4121,
      "step": 74
    },
    {
      "epoch": 0.6854531607006854,
      "grad_norm": 0.5874064564704895,
      "learning_rate": 1.6603494674471595e-05,
      "loss": 0.4133,
      "step": 75
    },
    {
      "epoch": 0.6945925361766946,
      "grad_norm": 0.6408722996711731,
      "learning_rate": 1.6482283953077887e-05,
      "loss": 0.4423,
      "step": 76
    },
    {
      "epoch": 0.7037319116527038,
      "grad_norm": 0.6493033170700073,
      "learning_rate": 1.635940787848455e-05,
      "loss": 0.4148,
      "step": 77
    },
    {
      "epoch": 0.7128712871287128,
      "grad_norm": 0.6096062064170837,
      "learning_rate": 1.6234898018587336e-05,
      "loss": 0.4189,
      "step": 78
    },
    {
      "epoch": 0.722010662604722,
      "grad_norm": 0.639083981513977,
      "learning_rate": 1.6108786361015145e-05,
      "loss": 0.4076,
      "step": 79
    },
    {
      "epoch": 0.7311500380807312,
      "grad_norm": 0.611152708530426,
      "learning_rate": 1.598110530491216e-05,
      "loss": 0.4584,
      "step": 80
    },
    {
      "epoch": 0.7402894135567403,
      "grad_norm": 0.665632426738739,
      "learning_rate": 1.5851887652614238e-05,
      "loss": 0.4325,
      "step": 81
    },
    {
      "epoch": 0.7494287890327495,
      "grad_norm": 0.5669299960136414,
      "learning_rate": 1.5721166601221697e-05,
      "loss": 0.4484,
      "step": 82
    },
    {
      "epoch": 0.7585681645087585,
      "grad_norm": 0.6546853184700012,
      "learning_rate": 1.5588975734070717e-05,
      "loss": 0.412,
      "step": 83
    },
    {
      "epoch": 0.7677075399847677,
      "grad_norm": 0.49923935532569885,
      "learning_rate": 1.5455349012105488e-05,
      "loss": 0.4289,
      "step": 84
    },
    {
      "epoch": 0.7768469154607769,
      "grad_norm": 0.6773508191108704,
      "learning_rate": 1.5320320765153367e-05,
      "loss": 0.4269,
      "step": 85
    },
    {
      "epoch": 0.785986290936786,
      "grad_norm": 0.5767094492912292,
      "learning_rate": 1.5183925683105254e-05,
      "loss": 0.4285,
      "step": 86
    },
    {
      "epoch": 0.7951256664127951,
      "grad_norm": 0.6376355886459351,
      "learning_rate": 1.504619880700346e-05,
      "loss": 0.4258,
      "step": 87
    },
    {
      "epoch": 0.8042650418888042,
      "grad_norm": 0.5988445281982422,
      "learning_rate": 1.4907175520039381e-05,
      "loss": 0.4192,
      "step": 88
    },
    {
      "epoch": 0.8134044173648134,
      "grad_norm": 0.633305013179779,
      "learning_rate": 1.4766891538463255e-05,
      "loss": 0.4166,
      "step": 89
    },
    {
      "epoch": 0.8225437928408226,
      "grad_norm": 0.5571799278259277,
      "learning_rate": 1.4625382902408356e-05,
      "loss": 0.4125,
      "step": 90
    },
    {
      "epoch": 0.8316831683168316,
      "grad_norm": 0.5751418471336365,
      "learning_rate": 1.448268596663197e-05,
      "loss": 0.4044,
      "step": 91
    },
    {
      "epoch": 0.8408225437928408,
      "grad_norm": 0.5511994957923889,
      "learning_rate": 1.4338837391175582e-05,
      "loss": 0.4227,
      "step": 92
    },
    {
      "epoch": 0.84996191926885,
      "grad_norm": 0.5298687219619751,
      "learning_rate": 1.419387413194657e-05,
      "loss": 0.4073,
      "step": 93
    },
    {
      "epoch": 0.8591012947448591,
      "grad_norm": 0.5533376932144165,
      "learning_rate": 1.4047833431223938e-05,
      "loss": 0.4132,
      "step": 94
    },
    {
      "epoch": 0.8682406702208683,
      "grad_norm": 0.5788342356681824,
      "learning_rate": 1.390075280809047e-05,
      "loss": 0.4135,
      "step": 95
    },
    {
      "epoch": 0.8773800456968773,
      "grad_norm": 0.5516043901443481,
      "learning_rate": 1.3752670048793744e-05,
      "loss": 0.4144,
      "step": 96
    },
    {
      "epoch": 0.8865194211728865,
      "grad_norm": 0.5432876348495483,
      "learning_rate": 1.3603623197038536e-05,
      "loss": 0.4211,
      "step": 97
    },
    {
      "epoch": 0.8956587966488957,
      "grad_norm": 0.6421664953231812,
      "learning_rate": 1.3453650544213078e-05,
      "loss": 0.4329,
      "step": 98
    },
    {
      "epoch": 0.9047981721249048,
      "grad_norm": 0.5505258440971375,
      "learning_rate": 1.3302790619551673e-05,
      "loss": 0.4012,
      "step": 99
    },
    {
      "epoch": 0.913937547600914,
      "grad_norm": 0.6094185709953308,
      "learning_rate": 1.315108218023621e-05,
      "loss": 0.4206,
      "step": 100
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.5962590575218201,
      "learning_rate": 1.2998564201439117e-05,
      "loss": 0.4206,
      "step": 101
    },
    {
      "epoch": 0.9322162985529322,
      "grad_norm": 0.576652467250824,
      "learning_rate": 1.2845275866310325e-05,
      "loss": 0.4181,
      "step": 102
    },
    {
      "epoch": 0.9413556740289414,
      "grad_norm": 0.5562915205955505,
      "learning_rate": 1.2691256555910769e-05,
      "loss": 0.4134,
      "step": 103
    },
    {
      "epoch": 0.9504950495049505,
      "grad_norm": 0.5346901416778564,
      "learning_rate": 1.2536545839095074e-05,
      "loss": 0.4363,
      "step": 104
    },
    {
      "epoch": 0.9596344249809596,
      "grad_norm": 0.5291918516159058,
      "learning_rate": 1.2381183462345983e-05,
      "loss": 0.4189,
      "step": 105
    },
    {
      "epoch": 0.9687738004569688,
      "grad_norm": 0.5179046988487244,
      "learning_rate": 1.2225209339563144e-05,
      "loss": 0.4131,
      "step": 106
    },
    {
      "epoch": 0.9779131759329779,
      "grad_norm": 0.50760418176651,
      "learning_rate": 1.206866354180891e-05,
      "loss": 0.4145,
      "step": 107
    },
    {
      "epoch": 0.9870525514089871,
      "grad_norm": 0.483079195022583,
      "learning_rate": 1.1911586287013726e-05,
      "loss": 0.4208,
      "step": 108
    },
    {
      "epoch": 0.9961919268849961,
      "grad_norm": 0.5721336007118225,
      "learning_rate": 1.1754017929643818e-05,
      "loss": 0.4141,
      "step": 109
    },
    {
      "epoch": 1.0053313023610053,
      "grad_norm": 0.9483687281608582,
      "learning_rate": 1.1595998950333794e-05,
      "loss": 0.5968,
      "step": 110
    },
    {
      "epoch": 1.0144706778370145,
      "grad_norm": 0.6470979452133179,
      "learning_rate": 1.143756994548682e-05,
      "loss": 0.3799,
      "step": 111
    },
    {
      "epoch": 1.0236100533130237,
      "grad_norm": 0.5971588492393494,
      "learning_rate": 1.1278771616845061e-05,
      "loss": 0.3711,
      "step": 112
    },
    {
      "epoch": 1.0327494287890326,
      "grad_norm": 0.5583009123802185,
      "learning_rate": 1.1119644761033079e-05,
      "loss": 0.3657,
      "step": 113
    },
    {
      "epoch": 1.0418888042650418,
      "grad_norm": 0.5742236971855164,
      "learning_rate": 1.0960230259076819e-05,
      "loss": 0.3561,
      "step": 114
    },
    {
      "epoch": 1.051028179741051,
      "grad_norm": 0.6248900294303894,
      "learning_rate": 1.0800569065900935e-05,
      "loss": 0.3861,
      "step": 115
    },
    {
      "epoch": 1.0601675552170602,
      "grad_norm": 0.6363320350646973,
      "learning_rate": 1.064070219980713e-05,
      "loss": 0.3392,
      "step": 116
    },
    {
      "epoch": 1.0693069306930694,
      "grad_norm": 0.5456411838531494,
      "learning_rate": 1.0480670731936209e-05,
      "loss": 0.3361,
      "step": 117
    },
    {
      "epoch": 1.0784463061690786,
      "grad_norm": 0.510452389717102,
      "learning_rate": 1.0320515775716556e-05,
      "loss": 0.3245,
      "step": 118
    },
    {
      "epoch": 1.0875856816450875,
      "grad_norm": 0.5856946110725403,
      "learning_rate": 1.0160278476301739e-05,
      "loss": 0.409,
      "step": 119
    },
    {
      "epoch": 1.0967250571210967,
      "grad_norm": 0.5407898426055908,
      "learning_rate": 1e-05,
      "loss": 0.3577,
      "step": 120
    },
    {
      "epoch": 1.1058644325971059,
      "grad_norm": 0.4385759234428406,
      "learning_rate": 9.839721523698265e-06,
      "loss": 0.2774,
      "step": 121
    },
    {
      "epoch": 1.115003808073115,
      "grad_norm": 0.5878434777259827,
      "learning_rate": 9.67948422428345e-06,
      "loss": 0.3834,
      "step": 122
    },
    {
      "epoch": 1.1241431835491242,
      "grad_norm": 0.5798720121383667,
      "learning_rate": 9.519329268063795e-06,
      "loss": 0.3729,
      "step": 123
    },
    {
      "epoch": 1.1332825590251332,
      "grad_norm": 0.5163602828979492,
      "learning_rate": 9.359297800192873e-06,
      "loss": 0.335,
      "step": 124
    },
    {
      "epoch": 1.1424219345011424,
      "grad_norm": 0.5222927331924438,
      "learning_rate": 9.199430934099068e-06,
      "loss": 0.3317,
      "step": 125
    },
    {
      "epoch": 1.1515613099771516,
      "grad_norm": 0.5279880166053772,
      "learning_rate": 9.039769740923183e-06,
      "loss": 0.3216,
      "step": 126
    },
    {
      "epoch": 1.1607006854531607,
      "grad_norm": 0.5063551068305969,
      "learning_rate": 8.880355238966923e-06,
      "loss": 0.3386,
      "step": 127
    },
    {
      "epoch": 1.16984006092917,
      "grad_norm": 0.5172789692878723,
      "learning_rate": 8.721228383154939e-06,
      "loss": 0.3507,
      "step": 128
    },
    {
      "epoch": 1.178979436405179,
      "grad_norm": 0.49883145093917847,
      "learning_rate": 8.562430054513184e-06,
      "loss": 0.3447,
      "step": 129
    },
    {
      "epoch": 1.188118811881188,
      "grad_norm": 0.503009021282196,
      "learning_rate": 8.404001049666211e-06,
      "loss": 0.348,
      "step": 130
    },
    {
      "epoch": 1.1972581873571972,
      "grad_norm": 0.5342569351196289,
      "learning_rate": 8.245982070356186e-06,
      "loss": 0.3237,
      "step": 131
    },
    {
      "epoch": 1.2063975628332064,
      "grad_norm": 0.49484220147132874,
      "learning_rate": 8.08841371298628e-06,
      "loss": 0.3438,
      "step": 132
    },
    {
      "epoch": 1.2155369383092156,
      "grad_norm": 0.48607689142227173,
      "learning_rate": 7.931336458191092e-06,
      "loss": 0.3744,
      "step": 133
    },
    {
      "epoch": 1.2246763137852246,
      "grad_norm": 0.46569332480430603,
      "learning_rate": 7.774790660436857e-06,
      "loss": 0.3359,
      "step": 134
    },
    {
      "epoch": 1.2338156892612338,
      "grad_norm": 0.46949270367622375,
      "learning_rate": 7.618816537654018e-06,
      "loss": 0.3473,
      "step": 135
    },
    {
      "epoch": 1.242955064737243,
      "grad_norm": 0.43653199076652527,
      "learning_rate": 7.463454160904928e-06,
      "loss": 0.3022,
      "step": 136
    },
    {
      "epoch": 1.2520944402132521,
      "grad_norm": 0.5573130249977112,
      "learning_rate": 7.308743444089232e-06,
      "loss": 0.3571,
      "step": 137
    },
    {
      "epoch": 1.2612338156892613,
      "grad_norm": 0.4045010805130005,
      "learning_rate": 7.154724133689677e-06,
      "loss": 0.322,
      "step": 138
    },
    {
      "epoch": 1.2703731911652705,
      "grad_norm": 0.5753940939903259,
      "learning_rate": 7.001435798560884e-06,
      "loss": 0.348,
      "step": 139
    },
    {
      "epoch": 1.2795125666412794,
      "grad_norm": 0.5088015794754028,
      "learning_rate": 6.848917819763794e-06,
      "loss": 0.3265,
      "step": 140
    },
    {
      "epoch": 1.2886519421172886,
      "grad_norm": 0.4818384051322937,
      "learning_rate": 6.697209380448333e-06,
      "loss": 0.3585,
      "step": 141
    },
    {
      "epoch": 1.2977913175932978,
      "grad_norm": 0.5760669708251953,
      "learning_rate": 6.546349455786926e-06,
      "loss": 0.3969,
      "step": 142
    },
    {
      "epoch": 1.306930693069307,
      "grad_norm": 0.46207141876220703,
      "learning_rate": 6.396376802961468e-06,
      "loss": 0.3448,
      "step": 143
    },
    {
      "epoch": 1.316070068545316,
      "grad_norm": 0.423732727766037,
      "learning_rate": 6.24732995120626e-06,
      "loss": 0.3273,
      "step": 144
    },
    {
      "epoch": 1.3252094440213251,
      "grad_norm": 0.42002105712890625,
      "learning_rate": 6.099247191909532e-06,
      "loss": 0.3247,
      "step": 145
    },
    {
      "epoch": 1.3343488194973343,
      "grad_norm": 0.43163520097732544,
      "learning_rate": 5.952166568776062e-06,
      "loss": 0.3673,
      "step": 146
    },
    {
      "epoch": 1.3434881949733435,
      "grad_norm": 0.44529059529304504,
      "learning_rate": 5.806125868053433e-06,
      "loss": 0.3081,
      "step": 147
    },
    {
      "epoch": 1.3526275704493527,
      "grad_norm": 0.44915682077407837,
      "learning_rate": 5.66116260882442e-06,
      "loss": 0.3201,
      "step": 148
    },
    {
      "epoch": 1.3617669459253618,
      "grad_norm": 0.45789527893066406,
      "learning_rate": 5.517314033368031e-06,
      "loss": 0.3533,
      "step": 149
    },
    {
      "epoch": 1.370906321401371,
      "grad_norm": 0.4622362554073334,
      "learning_rate": 5.37461709759165e-06,
      "loss": 0.355,
      "step": 150
    },
    {
      "epoch": 1.38004569687738,
      "grad_norm": 0.45579296350479126,
      "learning_rate": 5.233108461536749e-06,
      "loss": 0.3515,
      "step": 151
    },
    {
      "epoch": 1.3891850723533892,
      "grad_norm": 0.42939919233322144,
      "learning_rate": 5.092824479960625e-06,
      "loss": 0.3592,
      "step": 152
    },
    {
      "epoch": 1.3983244478293984,
      "grad_norm": 0.48685169219970703,
      "learning_rate": 4.9538011929965436e-06,
      "loss": 0.373,
      "step": 153
    },
    {
      "epoch": 1.4074638233054075,
      "grad_norm": 0.4406936466693878,
      "learning_rate": 4.81607431689475e-06,
      "loss": 0.3346,
      "step": 154
    },
    {
      "epoch": 1.4166031987814165,
      "grad_norm": 0.46851012110710144,
      "learning_rate": 4.679679234846636e-06,
      "loss": 0.3602,
      "step": 155
    },
    {
      "epoch": 1.4257425742574257,
      "grad_norm": 0.40592843294143677,
      "learning_rate": 4.544650987894514e-06,
      "loss": 0.3088,
      "step": 156
    },
    {
      "epoch": 1.4348819497334349,
      "grad_norm": 0.4123000204563141,
      "learning_rate": 4.411024265929283e-06,
      "loss": 0.3343,
      "step": 157
    },
    {
      "epoch": 1.444021325209444,
      "grad_norm": 0.395128071308136,
      "learning_rate": 4.278833398778306e-06,
      "loss": 0.2924,
      "step": 158
    },
    {
      "epoch": 1.4531607006854532,
      "grad_norm": 0.4521404206752777,
      "learning_rate": 4.148112347385762e-06,
      "loss": 0.3816,
      "step": 159
    },
    {
      "epoch": 1.4623000761614624,
      "grad_norm": 0.41636931896209717,
      "learning_rate": 4.01889469508784e-06,
      "loss": 0.3352,
      "step": 160
    },
    {
      "epoch": 1.4714394516374714,
      "grad_norm": 0.43509963154792786,
      "learning_rate": 3.891213638984858e-06,
      "loss": 0.3559,
      "step": 161
    },
    {
      "epoch": 1.4805788271134805,
      "grad_norm": 0.37414830923080444,
      "learning_rate": 3.7651019814126656e-06,
      "loss": 0.3027,
      "step": 162
    },
    {
      "epoch": 1.4897182025894897,
      "grad_norm": 0.44813913106918335,
      "learning_rate": 3.6405921215154492e-06,
      "loss": 0.3676,
      "step": 163
    },
    {
      "epoch": 1.498857578065499,
      "grad_norm": 0.4593113958835602,
      "learning_rate": 3.5177160469221184e-06,
      "loss": 0.3432,
      "step": 164
    },
    {
      "epoch": 1.5079969535415079,
      "grad_norm": 0.3725077211856842,
      "learning_rate": 3.3965053255284085e-06,
      "loss": 0.2981,
      "step": 165
    },
    {
      "epoch": 1.517136329017517,
      "grad_norm": 0.4420298933982849,
      "learning_rate": 3.2769910973868314e-06,
      "loss": 0.3699,
      "step": 166
    },
    {
      "epoch": 1.5262757044935262,
      "grad_norm": 0.3754097819328308,
      "learning_rate": 3.1592040667065393e-06,
      "loss": 0.3133,
      "step": 167
    },
    {
      "epoch": 1.5354150799695354,
      "grad_norm": 0.44490984082221985,
      "learning_rate": 3.0431744939651365e-06,
      "loss": 0.3577,
      "step": 168
    },
    {
      "epoch": 1.5445544554455446,
      "grad_norm": 0.3956274688243866,
      "learning_rate": 2.9289321881345257e-06,
      "loss": 0.3087,
      "step": 169
    },
    {
      "epoch": 1.5536938309215538,
      "grad_norm": 0.44501689076423645,
      "learning_rate": 2.8165064990227255e-06,
      "loss": 0.3393,
      "step": 170
    },
    {
      "epoch": 1.562833206397563,
      "grad_norm": 0.45090392231941223,
      "learning_rate": 2.7059263097336595e-06,
      "loss": 0.3807,
      "step": 171
    },
    {
      "epoch": 1.5719725818735721,
      "grad_norm": 0.42932793498039246,
      "learning_rate": 2.597220029246846e-06,
      "loss": 0.369,
      "step": 172
    },
    {
      "epoch": 1.581111957349581,
      "grad_norm": 0.3550121784210205,
      "learning_rate": 2.490415585118887e-06,
      "loss": 0.29,
      "step": 173
    },
    {
      "epoch": 1.5902513328255903,
      "grad_norm": 0.4236467182636261,
      "learning_rate": 2.3855404163086558e-06,
      "loss": 0.3518,
      "step": 174
    },
    {
      "epoch": 1.5993907083015992,
      "grad_norm": 0.37905916571617126,
      "learning_rate": 2.282621466127982e-06,
      "loss": 0.297,
      "step": 175
    },
    {
      "epoch": 1.6085300837776084,
      "grad_norm": 0.3965452313423157,
      "learning_rate": 2.1816851753197023e-06,
      "loss": 0.3166,
      "step": 176
    },
    {
      "epoch": 1.6176694592536176,
      "grad_norm": 0.43022918701171875,
      "learning_rate": 2.082757475264804e-06,
      "loss": 0.3552,
      "step": 177
    },
    {
      "epoch": 1.6268088347296268,
      "grad_norm": 0.40345683693885803,
      "learning_rate": 1.9858637813204352e-06,
      "loss": 0.2897,
      "step": 178
    },
    {
      "epoch": 1.635948210205636,
      "grad_norm": 0.4786721169948578,
      "learning_rate": 1.8910289862904917e-06,
      "loss": 0.414,
      "step": 179
    },
    {
      "epoch": 1.6450875856816451,
      "grad_norm": 0.3699612319469452,
      "learning_rate": 1.7982774540304404e-06,
      "loss": 0.288,
      "step": 180
    },
    {
      "epoch": 1.6542269611576543,
      "grad_norm": 0.41491949558258057,
      "learning_rate": 1.7076330131880525e-06,
      "loss": 0.3297,
      "step": 181
    },
    {
      "epoch": 1.6633663366336635,
      "grad_norm": 0.40379223227500916,
      "learning_rate": 1.6191189510815942e-06,
      "loss": 0.3822,
      "step": 182
    },
    {
      "epoch": 1.6725057121096725,
      "grad_norm": 0.34889909625053406,
      "learning_rate": 1.5327580077171589e-06,
      "loss": 0.305,
      "step": 183
    },
    {
      "epoch": 1.6816450875856817,
      "grad_norm": 0.3765786588191986,
      "learning_rate": 1.4485723699465392e-06,
      "loss": 0.3624,
      "step": 184
    },
    {
      "epoch": 1.6907844630616908,
      "grad_norm": 0.36284613609313965,
      "learning_rate": 1.3665836657672493e-06,
      "loss": 0.3178,
      "step": 185
    },
    {
      "epoch": 1.6999238385376998,
      "grad_norm": 0.3935321867465973,
      "learning_rate": 1.286812958766106e-06,
      "loss": 0.3743,
      "step": 186
    },
    {
      "epoch": 1.709063214013709,
      "grad_norm": 0.3890056312084198,
      "learning_rate": 1.209280742707828e-06,
      "loss": 0.3397,
      "step": 187
    },
    {
      "epoch": 1.7182025894897182,
      "grad_norm": 0.40561848878860474,
      "learning_rate": 1.134006936269999e-06,
      "loss": 0.3465,
      "step": 188
    },
    {
      "epoch": 1.7273419649657273,
      "grad_norm": 0.43405306339263916,
      "learning_rate": 1.0610108779258043e-06,
      "loss": 0.3369,
      "step": 189
    },
    {
      "epoch": 1.7364813404417365,
      "grad_norm": 0.418133407831192,
      "learning_rate": 9.903113209758098e-07,
      "loss": 0.3499,
      "step": 190
    },
    {
      "epoch": 1.7456207159177457,
      "grad_norm": 0.36436179280281067,
      "learning_rate": 9.2192642873008e-07,
      "loss": 0.3311,
      "step": 191
    },
    {
      "epoch": 1.7547600913937549,
      "grad_norm": 0.3649395704269409,
      "learning_rate": 8.558737698418762e-07,
      "loss": 0.3279,
      "step": 192
    },
    {
      "epoch": 1.763899466869764,
      "grad_norm": 0.3702605962753296,
      "learning_rate": 7.921703137941172e-07,
      "loss": 0.335,
      "step": 193
    },
    {
      "epoch": 1.773038842345773,
      "grad_norm": 0.3535550534725189,
      "learning_rate": 7.308324265397837e-07,
      "loss": 0.3227,
      "step": 194
    },
    {
      "epoch": 1.7821782178217822,
      "grad_norm": 0.3407094478607178,
      "learning_rate": 6.718758662973524e-07,
      "loss": 0.325,
      "step": 195
    },
    {
      "epoch": 1.7913175932977912,
      "grad_norm": 0.3515715003013611,
      "learning_rate": 6.153157795023956e-07,
      "loss": 0.3623,
      "step": 196
    },
    {
      "epoch": 1.8004569687738003,
      "grad_norm": 0.3598881959915161,
      "learning_rate": 5.611666969163243e-07,
      "loss": 0.3614,
      "step": 197
    },
    {
      "epoch": 1.8095963442498095,
      "grad_norm": 0.34573909640312195,
      "learning_rate": 5.094425298933136e-07,
      "loss": 0.341,
      "step": 198
    },
    {
      "epoch": 1.8187357197258187,
      "grad_norm": 0.3632952868938446,
      "learning_rate": 4.6015656680636234e-07,
      "loss": 0.3511,
      "step": 199
    },
    {
      "epoch": 1.827875095201828,
      "grad_norm": 0.35777807235717773,
      "learning_rate": 4.133214696333943e-07,
      "loss": 0.343,
      "step": 200
    },
    {
      "epoch": 1.837014470677837,
      "grad_norm": 0.3659260869026184,
      "learning_rate": 3.6894927070429744e-07,
      "loss": 0.324,
      "step": 201
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.37839674949645996,
      "learning_rate": 3.2705136960970554e-07,
      "loss": 0.3629,
      "step": 202
    },
    {
      "epoch": 1.8552932216298554,
      "grad_norm": 0.36181095242500305,
      "learning_rate": 2.8763853027236277e-07,
      "loss": 0.3303,
      "step": 203
    },
    {
      "epoch": 1.8644325971058644,
      "grad_norm": 0.36267733573913574,
      "learning_rate": 2.507208781817638e-07,
      "loss": 0.3305,
      "step": 204
    },
    {
      "epoch": 1.8735719725818736,
      "grad_norm": 0.37985163927078247,
      "learning_rate": 2.1630789779284677e-07,
      "loss": 0.3638,
      "step": 205
    },
    {
      "epoch": 1.8827113480578828,
      "grad_norm": 0.3690941333770752,
      "learning_rate": 1.844084300893456e-07,
      "loss": 0.297,
      "step": 206
    },
    {
      "epoch": 1.8918507235338917,
      "grad_norm": 0.3706071674823761,
      "learning_rate": 1.55030670312476e-07,
      "loss": 0.3285,
      "step": 207
    },
    {
      "epoch": 1.900990099009901,
      "grad_norm": 0.38501277565956116,
      "learning_rate": 1.2818216585549824e-07,
      "loss": 0.3927,
      "step": 208
    },
    {
      "epoch": 1.91012947448591,
      "grad_norm": 0.35342392325401306,
      "learning_rate": 1.0386981432474075e-07,
      "loss": 0.3389,
      "step": 209
    },
    {
      "epoch": 1.9192688499619193,
      "grad_norm": 0.34960195422172546,
      "learning_rate": 8.209986176753947e-08,
      "loss": 0.3235,
      "step": 210
    },
    {
      "epoch": 1.9284082254379284,
      "grad_norm": 0.37650802731513977,
      "learning_rate": 6.287790106757396e-08,
      "loss": 0.3658,
      "step": 211
    },
    {
      "epoch": 1.9375476009139376,
      "grad_norm": 0.34741005301475525,
      "learning_rate": 4.6208870508017703e-08,
      "loss": 0.3335,
      "step": 212
    },
    {
      "epoch": 1.9466869763899468,
      "grad_norm": 0.3389255702495575,
      "learning_rate": 3.2097052502843005e-08,
      "loss": 0.304,
      "step": 213
    },
    {
      "epoch": 1.955826351865956,
      "grad_norm": 0.3613230586051941,
      "learning_rate": 2.054607249663665e-08,
      "loss": 0.3668,
      "step": 214
    },
    {
      "epoch": 1.964965727341965,
      "grad_norm": 0.36797016859054565,
      "learning_rate": 1.1558898033191545e-08,
      "loss": 0.3406,
      "step": 215
    },
    {
      "epoch": 1.9741051028179741,
      "grad_norm": 0.3596664071083069,
      "learning_rate": 5.137837993121064e-09,
      "loss": 0.3294,
      "step": 216
    },
    {
      "epoch": 1.983244478293983,
      "grad_norm": 0.3750191628932953,
      "learning_rate": 1.2845420006879494e-09,
      "loss": 0.3285,
      "step": 217
    },
    {
      "epoch": 1.9923838537699923,
      "grad_norm": 0.41900479793548584,
      "learning_rate": 0.0,
      "loss": 0.3674,
      "step": 218
    },
    {
      "epoch": 1.9923838537699923,
      "step": 218,
      "total_flos": 165914164723712.0,
      "train_loss": 0.4081570637335471,
      "train_runtime": 15049.1519,
      "train_samples_per_second": 1.395,
      "train_steps_per_second": 0.014
    }
  ],
  "logging_steps": 1,
  "max_steps": 218,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 165914164723712.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}