{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9961919268849961,
  "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": 3.0662612915039062,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.6539,
      "step": 1
    },
    {
      "epoch": 0.01827875095201828,
      "grad_norm": 3.0926544666290283,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.6614,
      "step": 2
    },
    {
      "epoch": 0.027418126428027417,
      "grad_norm": 3.0256407260894775,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.6435,
      "step": 3
    },
    {
      "epoch": 0.03655750190403656,
      "grad_norm": 2.8626863956451416,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.6528,
      "step": 4
    },
    {
      "epoch": 0.0456968773800457,
      "grad_norm": 2.3245227336883545,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.6292,
      "step": 5
    },
    {
      "epoch": 0.05483625285605483,
      "grad_norm": 1.6400835514068604,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.6013,
      "step": 6
    },
    {
      "epoch": 0.06397562833206398,
      "grad_norm": 1.4912341833114624,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.6013,
      "step": 7
    },
    {
      "epoch": 0.07311500380807312,
      "grad_norm": 1.28671395778656,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.588,
      "step": 8
    },
    {
      "epoch": 0.08225437928408226,
      "grad_norm": 2.198812246322632,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.5794,
      "step": 9
    },
    {
      "epoch": 0.0913937547600914,
      "grad_norm": 0.9983230233192444,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.5403,
      "step": 10
    },
    {
      "epoch": 0.10053313023610053,
      "grad_norm": 1.5041189193725586,
      "learning_rate": 1e-05,
      "loss": 0.526,
      "step": 11
    },
    {
      "epoch": 0.10967250571210967,
      "grad_norm": 1.1437222957611084,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.5001,
      "step": 12
    },
    {
      "epoch": 0.1188118811881188,
      "grad_norm": 0.75607830286026,
      "learning_rate": 1.181818181818182e-05,
      "loss": 0.5017,
      "step": 13
    },
    {
      "epoch": 0.12795125666412796,
      "grad_norm": 1.0801345109939575,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 0.4927,
      "step": 14
    },
    {
      "epoch": 0.1370906321401371,
      "grad_norm": 1.3382116556167603,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.4917,
      "step": 15
    },
    {
      "epoch": 0.14623000761614624,
      "grad_norm": 0.9588184952735901,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.4444,
      "step": 16
    },
    {
      "epoch": 0.15536938309215537,
      "grad_norm": 0.7174947261810303,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.4665,
      "step": 17
    },
    {
      "epoch": 0.16450875856816452,
      "grad_norm": 0.6924068331718445,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 0.454,
      "step": 18
    },
    {
      "epoch": 0.17364813404417365,
      "grad_norm": 0.8726766109466553,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 0.4245,
      "step": 19
    },
    {
      "epoch": 0.1827875095201828,
      "grad_norm": 0.9118347764015198,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.4473,
      "step": 20
    },
    {
      "epoch": 0.19192688499619193,
      "grad_norm": 0.6092010736465454,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 0.4181,
      "step": 21
    },
    {
      "epoch": 0.20106626047220105,
      "grad_norm": 0.7477689981460571,
      "learning_rate": 2e-05,
      "loss": 0.4225,
      "step": 22
    },
    {
      "epoch": 0.2102056359482102,
      "grad_norm": 0.7922837138175964,
      "learning_rate": 1.9998715457999313e-05,
      "loss": 0.4286,
      "step": 23
    },
    {
      "epoch": 0.21934501142421933,
      "grad_norm": 0.6749143600463867,
      "learning_rate": 1.999486216200688e-05,
      "loss": 0.4125,
      "step": 24
    },
    {
      "epoch": 0.2284843869002285,
      "grad_norm": 0.6132336258888245,
      "learning_rate": 1.9988441101966807e-05,
      "loss": 0.4321,
      "step": 25
    },
    {
      "epoch": 0.2376237623762376,
      "grad_norm": 0.6256394982337952,
      "learning_rate": 1.9979453927503366e-05,
      "loss": 0.4401,
      "step": 26
    },
    {
      "epoch": 0.24676313785224677,
      "grad_norm": 0.5225641131401062,
      "learning_rate": 1.9967902947497158e-05,
      "loss": 0.4255,
      "step": 27
    },
    {
      "epoch": 0.2559025133282559,
      "grad_norm": 0.4767383635044098,
      "learning_rate": 1.9953791129491985e-05,
      "loss": 0.42,
      "step": 28
    },
    {
      "epoch": 0.26504188880426505,
      "grad_norm": 0.5424817204475403,
      "learning_rate": 1.9937122098932428e-05,
      "loss": 0.4023,
      "step": 29
    },
    {
      "epoch": 0.2741812642802742,
      "grad_norm": 0.5565168857574463,
      "learning_rate": 1.991790013823246e-05,
      "loss": 0.4182,
      "step": 30
    },
    {
      "epoch": 0.2833206397562833,
      "grad_norm": 0.4491325318813324,
      "learning_rate": 1.9896130185675263e-05,
      "loss": 0.4142,
      "step": 31
    },
    {
      "epoch": 0.2924600152322925,
      "grad_norm": 0.465501606464386,
      "learning_rate": 1.9871817834144506e-05,
      "loss": 0.402,
      "step": 32
    },
    {
      "epoch": 0.3015993907083016,
      "grad_norm": 0.43807297945022583,
      "learning_rate": 1.9844969329687526e-05,
      "loss": 0.4112,
      "step": 33
    },
    {
      "epoch": 0.31073876618431073,
      "grad_norm": 0.4875776469707489,
      "learning_rate": 1.9815591569910654e-05,
      "loss": 0.3917,
      "step": 34
    },
    {
      "epoch": 0.31987814166031986,
      "grad_norm": 0.48279792070388794,
      "learning_rate": 1.9783692102207156e-05,
      "loss": 0.4108,
      "step": 35
    },
    {
      "epoch": 0.32901751713632904,
      "grad_norm": 0.4583301544189453,
      "learning_rate": 1.9749279121818235e-05,
      "loss": 0.3623,
      "step": 36
    },
    {
      "epoch": 0.33815689261233817,
      "grad_norm": 0.3826312720775604,
      "learning_rate": 1.971236146972764e-05,
      "loss": 0.3637,
      "step": 37
    },
    {
      "epoch": 0.3472962680883473,
      "grad_norm": 0.32831740379333496,
      "learning_rate": 1.9672948630390296e-05,
      "loss": 0.3627,
      "step": 38
    },
    {
      "epoch": 0.3564356435643564,
      "grad_norm": 0.4264345169067383,
      "learning_rate": 1.9631050729295705e-05,
      "loss": 0.4095,
      "step": 39
    },
    {
      "epoch": 0.3655750190403656,
      "grad_norm": 0.37228643894195557,
      "learning_rate": 1.9586678530366607e-05,
      "loss": 0.3846,
      "step": 40
    },
    {
      "epoch": 0.3747143945163747,
      "grad_norm": 0.2934049665927887,
      "learning_rate": 1.953984343319364e-05,
      "loss": 0.3688,
      "step": 41
    },
    {
      "epoch": 0.38385376999238385,
      "grad_norm": 0.30369582772254944,
      "learning_rate": 1.949055747010669e-05,
      "loss": 0.3885,
      "step": 42
    },
    {
      "epoch": 0.392993145468393,
      "grad_norm": 0.3635307848453522,
      "learning_rate": 1.9438833303083677e-05,
      "loss": 0.3908,
      "step": 43
    },
    {
      "epoch": 0.4021325209444021,
      "grad_norm": 0.3347395360469818,
      "learning_rate": 1.9384684220497605e-05,
      "loss": 0.3628,
      "step": 44
    },
    {
      "epoch": 0.4112718964204113,
      "grad_norm": 0.3150138258934021,
      "learning_rate": 1.932812413370265e-05,
      "loss": 0.4,
      "step": 45
    },
    {
      "epoch": 0.4204112718964204,
      "grad_norm": 0.2996457815170288,
      "learning_rate": 1.926916757346022e-05,
      "loss": 0.3682,
      "step": 46
    },
    {
      "epoch": 0.42955064737242954,
      "grad_norm": 0.3075307607650757,
      "learning_rate": 1.9207829686205882e-05,
      "loss": 0.3611,
      "step": 47
    },
    {
      "epoch": 0.43869002284843867,
      "grad_norm": 0.3118605613708496,
      "learning_rate": 1.9144126230158127e-05,
      "loss": 0.38,
      "step": 48
    },
    {
      "epoch": 0.44782939832444785,
      "grad_norm": 0.32702377438545227,
      "learning_rate": 1.9078073571269922e-05,
      "loss": 0.3617,
      "step": 49
    },
    {
      "epoch": 0.456968773800457,
      "grad_norm": 0.29042407870292664,
      "learning_rate": 1.900968867902419e-05,
      "loss": 0.3647,
      "step": 50
    },
    {
      "epoch": 0.4661081492764661,
      "grad_norm": 0.32697346806526184,
      "learning_rate": 1.8938989122074195e-05,
      "loss": 0.3712,
      "step": 51
    },
    {
      "epoch": 0.4752475247524752,
      "grad_norm": 0.3000766634941101,
      "learning_rate": 1.8865993063730003e-05,
      "loss": 0.3846,
      "step": 52
    },
    {
      "epoch": 0.4843869002284844,
      "grad_norm": 0.3716011047363281,
      "learning_rate": 1.8790719257292175e-05,
      "loss": 0.3888,
      "step": 53
    },
    {
      "epoch": 0.49352627570449353,
      "grad_norm": 0.3023267984390259,
      "learning_rate": 1.8713187041233896e-05,
      "loss": 0.3819,
      "step": 54
    },
    {
      "epoch": 0.5026656511805027,
      "grad_norm": 0.33082687854766846,
      "learning_rate": 1.8633416334232754e-05,
      "loss": 0.3617,
      "step": 55
    },
    {
      "epoch": 0.5118050266565118,
      "grad_norm": 0.3457026481628418,
      "learning_rate": 1.8551427630053464e-05,
      "loss": 0.388,
      "step": 56
    },
    {
      "epoch": 0.5209444021325209,
      "grad_norm": 0.30542826652526855,
      "learning_rate": 1.8467241992282842e-05,
      "loss": 0.347,
      "step": 57
    },
    {
      "epoch": 0.5300837776085301,
      "grad_norm": 0.345498651266098,
      "learning_rate": 1.8380881048918406e-05,
      "loss": 0.3736,
      "step": 58
    },
    {
      "epoch": 0.5392231530845393,
      "grad_norm": 0.35136133432388306,
      "learning_rate": 1.8292366986811952e-05,
      "loss": 0.3512,
      "step": 59
    },
    {
      "epoch": 0.5483625285605483,
      "grad_norm": 0.2956295609474182,
      "learning_rate": 1.820172254596956e-05,
      "loss": 0.368,
      "step": 60
    },
    {
      "epoch": 0.5575019040365575,
      "grad_norm": 0.33257603645324707,
      "learning_rate": 1.8108971013709512e-05,
      "loss": 0.3564,
      "step": 61
    },
    {
      "epoch": 0.5666412795125666,
      "grad_norm": 0.2967292368412018,
      "learning_rate": 1.8014136218679566e-05,
      "loss": 0.3577,
      "step": 62
    },
    {
      "epoch": 0.5757806549885758,
      "grad_norm": 0.3108254075050354,
      "learning_rate": 1.79172425247352e-05,
      "loss": 0.3557,
      "step": 63
    },
    {
      "epoch": 0.584920030464585,
      "grad_norm": 0.3251456916332245,
      "learning_rate": 1.78183148246803e-05,
      "loss": 0.355,
      "step": 64
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 0.2736187279224396,
      "learning_rate": 1.771737853387202e-05,
      "loss": 0.3674,
      "step": 65
    },
    {
      "epoch": 0.6031987814166032,
      "grad_norm": 0.3374134600162506,
      "learning_rate": 1.7614459583691346e-05,
      "loss": 0.3697,
      "step": 66
    },
    {
      "epoch": 0.6123381568926123,
      "grad_norm": 0.2792860269546509,
      "learning_rate": 1.7509584414881114e-05,
      "loss": 0.3642,
      "step": 67
    },
    {
      "epoch": 0.6214775323686215,
      "grad_norm": 0.31335848569869995,
      "learning_rate": 1.7402779970753156e-05,
      "loss": 0.3709,
      "step": 68
    },
    {
      "epoch": 0.6306169078446306,
      "grad_norm": 0.31955182552337646,
      "learning_rate": 1.7294073690266343e-05,
      "loss": 0.3777,
      "step": 69
    },
    {
      "epoch": 0.6397562833206397,
      "grad_norm": 0.2629980444908142,
      "learning_rate": 1.7183493500977277e-05,
      "loss": 0.363,
      "step": 70
    },
    {
      "epoch": 0.6488956587966489,
      "grad_norm": 0.3005543053150177,
      "learning_rate": 1.7071067811865477e-05,
      "loss": 0.3622,
      "step": 71
    },
    {
      "epoch": 0.6580350342726581,
      "grad_norm": 0.27785226702690125,
      "learning_rate": 1.6956825506034866e-05,
      "loss": 0.3548,
      "step": 72
    },
    {
      "epoch": 0.6671744097486672,
      "grad_norm": 0.31932327151298523,
      "learning_rate": 1.6840795933293464e-05,
      "loss": 0.3542,
      "step": 73
    },
    {
      "epoch": 0.6763137852246763,
      "grad_norm": 0.3696252703666687,
      "learning_rate": 1.672300890261317e-05,
      "loss": 0.352,
      "step": 74
    },
    {
      "epoch": 0.6854531607006854,
      "grad_norm": 0.2692716121673584,
      "learning_rate": 1.6603494674471595e-05,
      "loss": 0.349,
      "step": 75
    },
    {
      "epoch": 0.6945925361766946,
      "grad_norm": 0.35122308135032654,
      "learning_rate": 1.6482283953077887e-05,
      "loss": 0.3724,
      "step": 76
    },
    {
      "epoch": 0.7037319116527038,
      "grad_norm": 0.3385901153087616,
      "learning_rate": 1.635940787848455e-05,
      "loss": 0.351,
      "step": 77
    },
    {
      "epoch": 0.7128712871287128,
      "grad_norm": 0.28408318758010864,
      "learning_rate": 1.6234898018587336e-05,
      "loss": 0.3545,
      "step": 78
    },
    {
      "epoch": 0.722010662604722,
      "grad_norm": 0.32605817914009094,
      "learning_rate": 1.6108786361015145e-05,
      "loss": 0.3457,
      "step": 79
    },
    {
      "epoch": 0.7311500380807312,
      "grad_norm": 0.2926092743873596,
      "learning_rate": 1.598110530491216e-05,
      "loss": 0.3836,
      "step": 80
    },
    {
      "epoch": 0.7402894135567403,
      "grad_norm": 0.3354974389076233,
      "learning_rate": 1.5851887652614238e-05,
      "loss": 0.3657,
      "step": 81
    },
    {
      "epoch": 0.7494287890327495,
      "grad_norm": 0.3095051646232605,
      "learning_rate": 1.5721166601221697e-05,
      "loss": 0.378,
      "step": 82
    },
    {
      "epoch": 0.7585681645087585,
      "grad_norm": 0.31830886006355286,
      "learning_rate": 1.5588975734070717e-05,
      "loss": 0.3493,
      "step": 83
    },
    {
      "epoch": 0.7677075399847677,
      "grad_norm": 0.28574857115745544,
      "learning_rate": 1.5455349012105488e-05,
      "loss": 0.3632,
      "step": 84
    },
    {
      "epoch": 0.7768469154607769,
      "grad_norm": 0.2779458463191986,
      "learning_rate": 1.5320320765153367e-05,
      "loss": 0.3601,
      "step": 85
    },
    {
      "epoch": 0.785986290936786,
      "grad_norm": 0.2946210205554962,
      "learning_rate": 1.5183925683105254e-05,
      "loss": 0.3614,
      "step": 86
    },
    {
      "epoch": 0.7951256664127951,
      "grad_norm": 0.29695403575897217,
      "learning_rate": 1.504619880700346e-05,
      "loss": 0.3585,
      "step": 87
    },
    {
      "epoch": 0.8042650418888042,
      "grad_norm": 0.31594327092170715,
      "learning_rate": 1.4907175520039381e-05,
      "loss": 0.3538,
      "step": 88
    },
    {
      "epoch": 0.8134044173648134,
      "grad_norm": 0.2952496409416199,
      "learning_rate": 1.4766891538463255e-05,
      "loss": 0.3529,
      "step": 89
    },
    {
      "epoch": 0.8225437928408226,
      "grad_norm": 0.3008720576763153,
      "learning_rate": 1.4625382902408356e-05,
      "loss": 0.3488,
      "step": 90
    },
    {
      "epoch": 0.8316831683168316,
      "grad_norm": 0.27587708830833435,
      "learning_rate": 1.448268596663197e-05,
      "loss": 0.343,
      "step": 91
    },
    {
      "epoch": 0.8408225437928408,
      "grad_norm": 0.2870708405971527,
      "learning_rate": 1.4338837391175582e-05,
      "loss": 0.3577,
      "step": 92
    },
    {
      "epoch": 0.84996191926885,
      "grad_norm": 0.2745375335216522,
      "learning_rate": 1.419387413194657e-05,
      "loss": 0.3437,
      "step": 93
    },
    {
      "epoch": 0.8591012947448591,
      "grad_norm": 0.30285215377807617,
      "learning_rate": 1.4047833431223938e-05,
      "loss": 0.3504,
      "step": 94
    },
    {
      "epoch": 0.8682406702208683,
      "grad_norm": 0.331877738237381,
      "learning_rate": 1.390075280809047e-05,
      "loss": 0.3517,
      "step": 95
    },
    {
      "epoch": 0.8773800456968773,
      "grad_norm": 0.28984907269477844,
      "learning_rate": 1.3752670048793744e-05,
      "loss": 0.3507,
      "step": 96
    },
    {
      "epoch": 0.8865194211728865,
      "grad_norm": 0.27279210090637207,
      "learning_rate": 1.3603623197038536e-05,
      "loss": 0.3563,
      "step": 97
    },
    {
      "epoch": 0.8956587966488957,
      "grad_norm": 0.2931230664253235,
      "learning_rate": 1.3453650544213078e-05,
      "loss": 0.3648,
      "step": 98
    },
    {
      "epoch": 0.9047981721249048,
      "grad_norm": 0.304097443819046,
      "learning_rate": 1.3302790619551673e-05,
      "loss": 0.3407,
      "step": 99
    },
    {
      "epoch": 0.913937547600914,
      "grad_norm": 0.27398836612701416,
      "learning_rate": 1.315108218023621e-05,
      "loss": 0.355,
      "step": 100
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.28750985860824585,
      "learning_rate": 1.2998564201439117e-05,
      "loss": 0.3581,
      "step": 101
    },
    {
      "epoch": 0.9322162985529322,
      "grad_norm": 0.2681604027748108,
      "learning_rate": 1.2845275866310325e-05,
      "loss": 0.3549,
      "step": 102
    },
    {
      "epoch": 0.9413556740289414,
      "grad_norm": 0.26904138922691345,
      "learning_rate": 1.2691256555910769e-05,
      "loss": 0.3503,
      "step": 103
    },
    {
      "epoch": 0.9504950495049505,
      "grad_norm": 0.26759082078933716,
      "learning_rate": 1.2536545839095074e-05,
      "loss": 0.3678,
      "step": 104
    },
    {
      "epoch": 0.9596344249809596,
      "grad_norm": 0.2824430465698242,
      "learning_rate": 1.2381183462345983e-05,
      "loss": 0.3529,
      "step": 105
    },
    {
      "epoch": 0.9687738004569688,
      "grad_norm": 0.29599159955978394,
      "learning_rate": 1.2225209339563144e-05,
      "loss": 0.3513,
      "step": 106
    },
    {
      "epoch": 0.9779131759329779,
      "grad_norm": 0.2748142182826996,
      "learning_rate": 1.206866354180891e-05,
      "loss": 0.3508,
      "step": 107
    },
    {
      "epoch": 0.9870525514089871,
      "grad_norm": 0.27814149856567383,
      "learning_rate": 1.1911586287013726e-05,
      "loss": 0.3553,
      "step": 108
    },
    {
      "epoch": 0.9961919268849961,
      "grad_norm": 0.3061699867248535,
      "learning_rate": 1.1754017929643818e-05,
      "loss": 0.3504,
      "step": 109
    },
    {
      "epoch": 1.0053313023610053,
      "grad_norm": 0.42901527881622314,
      "learning_rate": 1.1595998950333794e-05,
      "loss": 0.504,
      "step": 110
    },
    {
      "epoch": 1.0144706778370145,
      "grad_norm": 0.3380506634712219,
      "learning_rate": 1.143756994548682e-05,
      "loss": 0.3248,
      "step": 111
    },
    {
      "epoch": 1.0236100533130237,
      "grad_norm": 0.27545130252838135,
      "learning_rate": 1.1278771616845061e-05,
      "loss": 0.3143,
      "step": 112
    },
    {
      "epoch": 1.0327494287890326,
      "grad_norm": 0.30115991830825806,
      "learning_rate": 1.1119644761033079e-05,
      "loss": 0.3108,
      "step": 113
    },
    {
      "epoch": 1.0418888042650418,
      "grad_norm": 0.33423128724098206,
      "learning_rate": 1.0960230259076819e-05,
      "loss": 0.3024,
      "step": 114
    },
    {
      "epoch": 1.051028179741051,
      "grad_norm": 0.37499961256980896,
      "learning_rate": 1.0800569065900935e-05,
      "loss": 0.3275,
      "step": 115
    },
    {
      "epoch": 1.0601675552170602,
      "grad_norm": 0.3214980363845825,
      "learning_rate": 1.064070219980713e-05,
      "loss": 0.2905,
      "step": 116
    },
    {
      "epoch": 1.0693069306930694,
      "grad_norm": 0.343065470457077,
      "learning_rate": 1.0480670731936209e-05,
      "loss": 0.2874,
      "step": 117
    },
    {
      "epoch": 1.0784463061690786,
      "grad_norm": 0.30497705936431885,
      "learning_rate": 1.0320515775716556e-05,
      "loss": 0.2786,
      "step": 118
    },
    {
      "epoch": 1.0875856816450875,
      "grad_norm": 0.3317449390888214,
      "learning_rate": 1.0160278476301739e-05,
      "loss": 0.3472,
      "step": 119
    },
    {
      "epoch": 1.0967250571210967,
      "grad_norm": 0.32423466444015503,
      "learning_rate": 1e-05,
      "loss": 0.3031,
      "step": 120
    },
    {
      "epoch": 1.1058644325971059,
      "grad_norm": 0.288601815700531,
      "learning_rate": 9.839721523698265e-06,
      "loss": 0.2416,
      "step": 121
    },
    {
      "epoch": 1.115003808073115,
      "grad_norm": 0.33957648277282715,
      "learning_rate": 9.67948422428345e-06,
      "loss": 0.329,
      "step": 122
    },
    {
      "epoch": 1.1241431835491242,
      "grad_norm": 0.31219515204429626,
      "learning_rate": 9.519329268063795e-06,
      "loss": 0.3184,
      "step": 123
    },
    {
      "epoch": 1.1332825590251332,
      "grad_norm": 0.2770800292491913,
      "learning_rate": 9.359297800192873e-06,
      "loss": 0.287,
      "step": 124
    },
    {
      "epoch": 1.1424219345011424,
      "grad_norm": 0.29623621702194214,
      "learning_rate": 9.199430934099068e-06,
      "loss": 0.2848,
      "step": 125
    },
    {
      "epoch": 1.1515613099771516,
      "grad_norm": 0.3068810999393463,
      "learning_rate": 9.039769740923183e-06,
      "loss": 0.2797,
      "step": 126
    },
    {
      "epoch": 1.1607006854531607,
      "grad_norm": 0.2660495936870575,
      "learning_rate": 8.880355238966923e-06,
      "loss": 0.2905,
      "step": 127
    },
    {
      "epoch": 1.16984006092917,
      "grad_norm": 0.31384557485580444,
      "learning_rate": 8.721228383154939e-06,
      "loss": 0.2996,
      "step": 128
    },
    {
      "epoch": 1.178979436405179,
      "grad_norm": 0.26587456464767456,
      "learning_rate": 8.562430054513184e-06,
      "loss": 0.2955,
      "step": 129
    },
    {
      "epoch": 1.188118811881188,
      "grad_norm": 0.2939557731151581,
      "learning_rate": 8.404001049666211e-06,
      "loss": 0.2981,
      "step": 130
    },
    {
      "epoch": 1.1972581873571972,
      "grad_norm": 0.3092159330844879,
      "learning_rate": 8.245982070356186e-06,
      "loss": 0.2779,
      "step": 131
    },
    {
      "epoch": 1.2063975628332064,
      "grad_norm": 0.263584703207016,
      "learning_rate": 8.08841371298628e-06,
      "loss": 0.296,
      "step": 132
    },
    {
      "epoch": 1.2155369383092156,
      "grad_norm": 0.3118206262588501,
      "learning_rate": 7.931336458191092e-06,
      "loss": 0.3203,
      "step": 133
    },
    {
      "epoch": 1.2246763137852246,
      "grad_norm": 0.278341680765152,
      "learning_rate": 7.774790660436857e-06,
      "loss": 0.2915,
      "step": 134
    },
    {
      "epoch": 1.2338156892612338,
      "grad_norm": 0.27913302183151245,
      "learning_rate": 7.618816537654018e-06,
      "loss": 0.2997,
      "step": 135
    },
    {
      "epoch": 1.242955064737243,
      "grad_norm": 0.2791356146335602,
      "learning_rate": 7.463454160904928e-06,
      "loss": 0.2625,
      "step": 136
    },
    {
      "epoch": 1.2520944402132521,
      "grad_norm": 0.2794390022754669,
      "learning_rate": 7.308743444089232e-06,
      "loss": 0.3042,
      "step": 137
    },
    {
      "epoch": 1.2612338156892613,
      "grad_norm": 0.2647666335105896,
      "learning_rate": 7.154724133689677e-06,
      "loss": 0.2775,
      "step": 138
    },
    {
      "epoch": 1.2703731911652705,
      "grad_norm": 0.3084835708141327,
      "learning_rate": 7.001435798560884e-06,
      "loss": 0.2996,
      "step": 139
    },
    {
      "epoch": 1.2795125666412794,
      "grad_norm": 0.33654364943504333,
      "learning_rate": 6.848917819763794e-06,
      "loss": 0.2816,
      "step": 140
    },
    {
      "epoch": 1.2886519421172886,
      "grad_norm": 0.26153555512428284,
      "learning_rate": 6.697209380448333e-06,
      "loss": 0.3098,
      "step": 141
    },
    {
      "epoch": 1.2977913175932978,
      "grad_norm": 0.35121041536331177,
      "learning_rate": 6.546349455786926e-06,
      "loss": 0.341,
      "step": 142
    },
    {
      "epoch": 1.306930693069307,
      "grad_norm": 0.2708624601364136,
      "learning_rate": 6.396376802961468e-06,
      "loss": 0.2959,
      "step": 143
    },
    {
      "epoch": 1.316070068545316,
      "grad_norm": 0.2626864016056061,
      "learning_rate": 6.24732995120626e-06,
      "loss": 0.2829,
      "step": 144
    },
    {
      "epoch": 1.3252094440213251,
      "grad_norm": 0.29574039578437805,
      "learning_rate": 6.099247191909532e-06,
      "loss": 0.281,
      "step": 145
    },
    {
      "epoch": 1.3343488194973343,
      "grad_norm": 0.26153430342674255,
      "learning_rate": 5.952166568776062e-06,
      "loss": 0.3167,
      "step": 146
    },
    {
      "epoch": 1.3434881949733435,
      "grad_norm": 0.2655847668647766,
      "learning_rate": 5.806125868053433e-06,
      "loss": 0.2687,
      "step": 147
    },
    {
      "epoch": 1.3526275704493527,
      "grad_norm": 0.2613089382648468,
      "learning_rate": 5.66116260882442e-06,
      "loss": 0.2788,
      "step": 148
    },
    {
      "epoch": 1.3617669459253618,
      "grad_norm": 0.27271604537963867,
      "learning_rate": 5.517314033368031e-06,
      "loss": 0.3039,
      "step": 149
    },
    {
      "epoch": 1.370906321401371,
      "grad_norm": 0.28087660670280457,
      "learning_rate": 5.37461709759165e-06,
      "loss": 0.3076,
      "step": 150
    },
    {
      "epoch": 1.38004569687738,
      "grad_norm": 0.2736712098121643,
      "learning_rate": 5.233108461536749e-06,
      "loss": 0.3041,
      "step": 151
    },
    {
      "epoch": 1.3891850723533892,
      "grad_norm": 0.2977502942085266,
      "learning_rate": 5.092824479960625e-06,
      "loss": 0.31,
      "step": 152
    },
    {
      "epoch": 1.3983244478293984,
      "grad_norm": 0.3099557161331177,
      "learning_rate": 4.9538011929965436e-06,
      "loss": 0.3212,
      "step": 153
    },
    {
      "epoch": 1.4074638233054075,
      "grad_norm": 0.3034738004207611,
      "learning_rate": 4.81607431689475e-06,
      "loss": 0.2922,
      "step": 154
    },
    {
      "epoch": 1.4166031987814165,
      "grad_norm": 0.2969326674938202,
      "learning_rate": 4.679679234846636e-06,
      "loss": 0.3107,
      "step": 155
    },
    {
      "epoch": 1.4257425742574257,
      "grad_norm": 0.32106441259384155,
      "learning_rate": 4.544650987894514e-06,
      "loss": 0.2697,
      "step": 156
    },
    {
      "epoch": 1.4348819497334349,
      "grad_norm": 0.2815849483013153,
      "learning_rate": 4.411024265929283e-06,
      "loss": 0.2905,
      "step": 157
    },
    {
      "epoch": 1.444021325209444,
      "grad_norm": 0.29976797103881836,
      "learning_rate": 4.278833398778306e-06,
      "loss": 0.2548,
      "step": 158
    },
    {
      "epoch": 1.4531607006854532,
      "grad_norm": 0.3238127827644348,
      "learning_rate": 4.148112347385762e-06,
      "loss": 0.3296,
      "step": 159
    },
    {
      "epoch": 1.4623000761614624,
      "grad_norm": 0.3087165355682373,
      "learning_rate": 4.01889469508784e-06,
      "loss": 0.2928,
      "step": 160
    },
    {
      "epoch": 1.4714394516374714,
      "grad_norm": 0.3463199734687805,
      "learning_rate": 3.891213638984858e-06,
      "loss": 0.3121,
      "step": 161
    },
    {
      "epoch": 1.4805788271134805,
      "grad_norm": 0.25795602798461914,
      "learning_rate": 3.7651019814126656e-06,
      "loss": 0.267,
      "step": 162
    },
    {
      "epoch": 1.4897182025894897,
      "grad_norm": 0.3257160186767578,
      "learning_rate": 3.6405921215154492e-06,
      "loss": 0.3175,
      "step": 163
    },
    {
      "epoch": 1.498857578065499,
      "grad_norm": 0.3168756365776062,
      "learning_rate": 3.5177160469221184e-06,
      "loss": 0.2985,
      "step": 164
    },
    {
      "epoch": 1.5079969535415079,
      "grad_norm": 0.27244633436203003,
      "learning_rate": 3.3965053255284085e-06,
      "loss": 0.2619,
      "step": 165
    },
    {
      "epoch": 1.517136329017517,
      "grad_norm": 0.3427713215351105,
      "learning_rate": 3.2769910973868314e-06,
      "loss": 0.3191,
      "step": 166
    },
    {
      "epoch": 1.5262757044935262,
      "grad_norm": 0.2740377187728882,
      "learning_rate": 3.1592040667065393e-06,
      "loss": 0.2744,
      "step": 167
    },
    {
      "epoch": 1.5354150799695354,
      "grad_norm": 0.3417263925075531,
      "learning_rate": 3.0431744939651365e-06,
      "loss": 0.3138,
      "step": 168
    },
    {
      "epoch": 1.5445544554455446,
      "grad_norm": 0.2803570032119751,
      "learning_rate": 2.9289321881345257e-06,
      "loss": 0.2748,
      "step": 169
    },
    {
      "epoch": 1.5536938309215538,
      "grad_norm": 0.34591829776763916,
      "learning_rate": 2.8165064990227255e-06,
      "loss": 0.2931,
      "step": 170
    },
    {
      "epoch": 1.562833206397563,
      "grad_norm": 0.30100783705711365,
      "learning_rate": 2.7059263097336595e-06,
      "loss": 0.3322,
      "step": 171
    },
    {
      "epoch": 1.5719725818735721,
      "grad_norm": 0.30982816219329834,
      "learning_rate": 2.597220029246846e-06,
      "loss": 0.3208,
      "step": 172
    },
    {
      "epoch": 1.581111957349581,
      "grad_norm": 0.2882677912712097,
      "learning_rate": 2.490415585118887e-06,
      "loss": 0.2542,
      "step": 173
    },
    {
      "epoch": 1.5902513328255903,
      "grad_norm": 0.2950691878795624,
      "learning_rate": 2.3855404163086558e-06,
      "loss": 0.3102,
      "step": 174
    },
    {
      "epoch": 1.5993907083015992,
      "grad_norm": 0.2920461595058441,
      "learning_rate": 2.282621466127982e-06,
      "loss": 0.2608,
      "step": 175
    },
    {
      "epoch": 1.6085300837776084,
      "grad_norm": 0.29719457030296326,
      "learning_rate": 2.1816851753197023e-06,
      "loss": 0.2785,
      "step": 176
    },
    {
      "epoch": 1.6176694592536176,
      "grad_norm": 0.30973735451698303,
      "learning_rate": 2.082757475264804e-06,
      "loss": 0.3116,
      "step": 177
    },
    {
      "epoch": 1.6268088347296268,
      "grad_norm": 0.2760176956653595,
      "learning_rate": 1.9858637813204352e-06,
      "loss": 0.256,
      "step": 178
    },
    {
      "epoch": 1.635948210205636,
      "grad_norm": 0.37629491090774536,
      "learning_rate": 1.8910289862904917e-06,
      "loss": 0.3598,
      "step": 179
    },
    {
      "epoch": 1.6450875856816451,
      "grad_norm": 0.22734475135803223,
      "learning_rate": 1.7982774540304404e-06,
      "loss": 0.2526,
      "step": 180
    },
    {
      "epoch": 1.6542269611576543,
      "grad_norm": 0.31396934390068054,
      "learning_rate": 1.7076330131880525e-06,
      "loss": 0.2885,
      "step": 181
    },
    {
      "epoch": 1.6633663366336635,
      "grad_norm": 0.32990527153015137,
      "learning_rate": 1.6191189510815942e-06,
      "loss": 0.3314,
      "step": 182
    },
    {
      "epoch": 1.6725057121096725,
      "grad_norm": 0.26883870363235474,
      "learning_rate": 1.5327580077171589e-06,
      "loss": 0.2691,
      "step": 183
    },
    {
      "epoch": 1.6816450875856817,
      "grad_norm": 0.2914712131023407,
      "learning_rate": 1.4485723699465392e-06,
      "loss": 0.3184,
      "step": 184
    },
    {
      "epoch": 1.6907844630616908,
      "grad_norm": 0.2758725881576538,
      "learning_rate": 1.3665836657672493e-06,
      "loss": 0.2783,
      "step": 185
    },
    {
      "epoch": 1.6999238385376998,
      "grad_norm": 0.3416280150413513,
      "learning_rate": 1.286812958766106e-06,
      "loss": 0.3252,
      "step": 186
    },
    {
      "epoch": 1.709063214013709,
      "grad_norm": 0.30903884768486023,
      "learning_rate": 1.209280742707828e-06,
      "loss": 0.2988,
      "step": 187
    },
    {
      "epoch": 1.7182025894897182,
      "grad_norm": 0.2973853349685669,
      "learning_rate": 1.134006936269999e-06,
      "loss": 0.3029,
      "step": 188
    },
    {
      "epoch": 1.7273419649657273,
      "grad_norm": 0.29858219623565674,
      "learning_rate": 1.0610108779258043e-06,
      "loss": 0.2904,
      "step": 189
    },
    {
      "epoch": 1.7364813404417365,
      "grad_norm": 0.28917691111564636,
      "learning_rate": 9.903113209758098e-07,
      "loss": 0.3059,
      "step": 190
    },
    {
      "epoch": 1.7456207159177457,
      "grad_norm": 0.2837529480457306,
      "learning_rate": 9.2192642873008e-07,
      "loss": 0.292,
      "step": 191
    },
    {
      "epoch": 1.7547600913937549,
      "grad_norm": 0.27116718888282776,
      "learning_rate": 8.558737698418762e-07,
      "loss": 0.287,
      "step": 192
    },
    {
      "epoch": 1.763899466869764,
      "grad_norm": 0.2641710042953491,
      "learning_rate": 7.921703137941172e-07,
      "loss": 0.2916,
      "step": 193
    },
    {
      "epoch": 1.773038842345773,
      "grad_norm": 0.27609920501708984,
      "learning_rate": 7.308324265397837e-07,
      "loss": 0.2847,
      "step": 194
    },
    {
      "epoch": 1.7821782178217822,
      "grad_norm": 0.2574136555194855,
      "learning_rate": 6.718758662973524e-07,
      "loss": 0.2842,
      "step": 195
    },
    {
      "epoch": 1.7913175932977912,
      "grad_norm": 0.26150667667388916,
      "learning_rate": 6.153157795023956e-07,
      "loss": 0.3169,
      "step": 196
    },
    {
      "epoch": 1.8004569687738003,
      "grad_norm": 0.26446864008903503,
      "learning_rate": 5.611666969163243e-07,
      "loss": 0.3155,
      "step": 197
    },
    {
      "epoch": 1.8095963442498095,
      "grad_norm": 0.2812100946903229,
      "learning_rate": 5.094425298933136e-07,
      "loss": 0.2952,
      "step": 198
    },
    {
      "epoch": 1.8187357197258187,
      "grad_norm": 0.29095372557640076,
      "learning_rate": 4.6015656680636234e-07,
      "loss": 0.3076,
      "step": 199
    },
    {
      "epoch": 1.827875095201828,
      "grad_norm": 0.3154431879520416,
      "learning_rate": 4.133214696333943e-07,
      "loss": 0.3001,
      "step": 200
    },
    {
      "epoch": 1.8408225437928407,
      "grad_norm": 0.26161229610443115,
      "learning_rate": 3.6894927070429744e-07,
      "loss": 0.3034,
      "step": 201
    },
    {
      "epoch": 1.84996191926885,
      "grad_norm": 0.275260329246521,
      "learning_rate": 3.2705136960970554e-07,
      "loss": 0.2986,
      "step": 202
    },
    {
      "epoch": 1.859101294744859,
      "grad_norm": 0.28469160199165344,
      "learning_rate": 2.8763853027236277e-07,
      "loss": 0.3022,
      "step": 203
    },
    {
      "epoch": 1.8682406702208683,
      "grad_norm": 0.28028687834739685,
      "learning_rate": 2.507208781817638e-07,
      "loss": 0.2878,
      "step": 204
    },
    {
      "epoch": 1.8773800456968774,
      "grad_norm": 0.27624866366386414,
      "learning_rate": 2.1630789779284677e-07,
      "loss": 0.2957,
      "step": 205
    },
    {
      "epoch": 1.8865194211728866,
      "grad_norm": 0.25148093700408936,
      "learning_rate": 1.844084300893456e-07,
      "loss": 0.2913,
      "step": 206
    },
    {
      "epoch": 1.8956587966488958,
      "grad_norm": 0.3340187966823578,
      "learning_rate": 1.55030670312476e-07,
      "loss": 0.3051,
      "step": 207
    },
    {
      "epoch": 1.9047981721249048,
      "grad_norm": 0.28733426332473755,
      "learning_rate": 1.2818216585549824e-07,
      "loss": 0.2968,
      "step": 208
    },
    {
      "epoch": 1.913937547600914,
      "grad_norm": 0.2607288062572479,
      "learning_rate": 1.0386981432474075e-07,
      "loss": 0.3034,
      "step": 209
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.27604418992996216,
      "learning_rate": 8.209986176753947e-08,
      "loss": 0.2934,
      "step": 210
    },
    {
      "epoch": 1.932216298552932,
      "grad_norm": 0.26872286200523376,
      "learning_rate": 6.287790106757396e-08,
      "loss": 0.3107,
      "step": 211
    },
    {
      "epoch": 1.9413556740289413,
      "grad_norm": 0.2715740203857422,
      "learning_rate": 4.6208870508017703e-08,
      "loss": 0.2925,
      "step": 212
    },
    {
      "epoch": 1.9504950495049505,
      "grad_norm": 0.24944356083869934,
      "learning_rate": 3.2097052502843005e-08,
      "loss": 0.2911,
      "step": 213
    },
    {
      "epoch": 1.9596344249809596,
      "grad_norm": 0.2840694487094879,
      "learning_rate": 2.054607249663665e-08,
      "loss": 0.2823,
      "step": 214
    },
    {
      "epoch": 1.9687738004569688,
      "grad_norm": 0.28054678440093994,
      "learning_rate": 1.1558898033191545e-08,
      "loss": 0.3068,
      "step": 215
    },
    {
      "epoch": 1.977913175932978,
      "grad_norm": 0.3040056526660919,
      "learning_rate": 5.137837993121064e-09,
      "loss": 0.3039,
      "step": 216
    },
    {
      "epoch": 1.9870525514089872,
      "grad_norm": 0.2787896990776062,
      "learning_rate": 1.2845420006879494e-09,
      "loss": 0.2965,
      "step": 217
    },
    {
      "epoch": 1.9961919268849961,
      "grad_norm": 0.2908172905445099,
      "learning_rate": 0.0,
      "loss": 0.298,
      "step": 218
    },
    {
      "epoch": 1.9961919268849961,
      "step": 218,
      "total_flos": 556668498083840.0,
      "train_loss": 0.02458530199637107,
      "train_runtime": 2561.7302,
      "train_samples_per_second": 8.198,
      "train_steps_per_second": 0.085
    }
  ],
  "logging_steps": 1,
  "max_steps": 218,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 556668498083840.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}