{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9952556668423828,
  "eval_steps": 500,
  "global_step": 118,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008434370057986295,
      "grad_norm": 0.08799133449792862,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 1.6351,
      "step": 1
    },
    {
      "epoch": 0.01686874011597259,
      "grad_norm": 0.08821269869804382,
      "learning_rate": 9.999999999999999e-06,
      "loss": 1.6405,
      "step": 2
    },
    {
      "epoch": 0.025303110173958882,
      "grad_norm": 0.028541648760437965,
      "learning_rate": 1.5e-05,
      "loss": 1.6264,
      "step": 3
    },
    {
      "epoch": 0.03373748023194518,
      "grad_norm": 0.016522206366062164,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 1.6233,
      "step": 4
    },
    {
      "epoch": 0.04217185028993147,
      "grad_norm": 0.054906539618968964,
      "learning_rate": 2.5e-05,
      "loss": 1.62,
      "step": 5
    },
    {
      "epoch": 0.050606220347917764,
      "grad_norm": 0.0514790378510952,
      "learning_rate": 3e-05,
      "loss": 1.6134,
      "step": 6
    },
    {
      "epoch": 0.05904059040590406,
      "grad_norm": 0.04156072437763214,
      "learning_rate": 3.5000000000000004e-05,
      "loss": 1.6169,
      "step": 7
    },
    {
      "epoch": 0.06747496046389036,
      "grad_norm": 0.05689298361539841,
      "learning_rate": 3.9999999999999996e-05,
      "loss": 1.6143,
      "step": 8
    },
    {
      "epoch": 0.07590933052187665,
      "grad_norm": 0.041525840759277344,
      "learning_rate": 4.5e-05,
      "loss": 1.6104,
      "step": 9
    },
    {
      "epoch": 0.08434370057986294,
      "grad_norm": 0.031016899272799492,
      "learning_rate": 5e-05,
      "loss": 1.6028,
      "step": 10
    },
    {
      "epoch": 0.09277807063784924,
      "grad_norm": 0.03775344789028168,
      "learning_rate": 5.5e-05,
      "loss": 1.5949,
      "step": 11
    },
    {
      "epoch": 0.10121244069583553,
      "grad_norm": 0.027061201632022858,
      "learning_rate": 6e-05,
      "loss": 1.5966,
      "step": 12
    },
    {
      "epoch": 0.10964681075382182,
      "grad_norm": 0.03555454686284065,
      "learning_rate": 5.998682509526384e-05,
      "loss": 1.601,
      "step": 13
    },
    {
      "epoch": 0.11808118081180811,
      "grad_norm": 0.038648299872875214,
      "learning_rate": 5.994731195292965e-05,
      "loss": 1.6015,
      "step": 14
    },
    {
      "epoch": 0.1265155508697944,
      "grad_norm": 0.03883035108447075,
      "learning_rate": 5.988149527845651e-05,
      "loss": 1.5992,
      "step": 15
    },
    {
      "epoch": 0.13494992092778071,
      "grad_norm": 0.03391977399587631,
      "learning_rate": 5.978943288040551e-05,
      "loss": 1.5932,
      "step": 16
    },
    {
      "epoch": 0.143384290985767,
      "grad_norm": 0.0362255796790123,
      "learning_rate": 5.967120561966492e-05,
      "loss": 1.5873,
      "step": 17
    },
    {
      "epoch": 0.1518186610437533,
      "grad_norm": 0.027403229847550392,
      "learning_rate": 5.952691733842791e-05,
      "loss": 1.5845,
      "step": 18
    },
    {
      "epoch": 0.16025303110173958,
      "grad_norm": 0.02821512520313263,
      "learning_rate": 5.935669476898512e-05,
      "loss": 1.5942,
      "step": 19
    },
    {
      "epoch": 0.16868740115972589,
      "grad_norm": 0.022913869470357895,
      "learning_rate": 5.9160687422412324e-05,
      "loss": 1.5976,
      "step": 20
    },
    {
      "epoch": 0.17712177121771217,
      "grad_norm": 0.02420000359416008,
      "learning_rate": 5.893906745725076e-05,
      "loss": 1.5862,
      "step": 21
    },
    {
      "epoch": 0.18555614127569847,
      "grad_norm": 0.021311871707439423,
      "learning_rate": 5.8692029528295675e-05,
      "loss": 1.5877,
      "step": 22
    },
    {
      "epoch": 0.19399051133368478,
      "grad_norm": 0.024183662608265877,
      "learning_rate": 5.841979061562574e-05,
      "loss": 1.584,
      "step": 23
    },
    {
      "epoch": 0.20242488139167106,
      "grad_norm": 0.02072131633758545,
      "learning_rate": 5.8122589834023634e-05,
      "loss": 1.5841,
      "step": 24
    },
    {
      "epoch": 0.21085925144965736,
      "grad_norm": 0.023273587226867676,
      "learning_rate": 5.7800688222955e-05,
      "loss": 1.5845,
      "step": 25
    },
    {
      "epoch": 0.21929362150764364,
      "grad_norm": 0.0180776659399271,
      "learning_rate": 5.745436851729055e-05,
      "loss": 1.594,
      "step": 26
    },
    {
      "epoch": 0.22772799156562995,
      "grad_norm": 0.018995055928826332,
      "learning_rate": 5.708393489897231e-05,
      "loss": 1.5903,
      "step": 27
    },
    {
      "epoch": 0.23616236162361623,
      "grad_norm": 0.017286648973822594,
      "learning_rate": 5.668971272984242e-05,
      "loss": 1.5804,
      "step": 28
    },
    {
      "epoch": 0.24459673168160254,
      "grad_norm": 0.018625088036060333,
      "learning_rate": 5.6272048265869104e-05,
      "loss": 1.5798,
      "step": 29
    },
    {
      "epoch": 0.2530311017395888,
      "grad_norm": 0.017109202221035957,
      "learning_rate": 5.583130835302066e-05,
      "loss": 1.5848,
      "step": 30
    },
    {
      "epoch": 0.2614654717975751,
      "grad_norm": 0.017000902444124222,
      "learning_rate": 5.536788010505478e-05,
      "loss": 1.5751,
      "step": 31
    },
    {
      "epoch": 0.26989984185556143,
      "grad_norm": 0.018897738307714462,
      "learning_rate": 5.4882170563506055e-05,
      "loss": 1.5799,
      "step": 32
    },
    {
      "epoch": 0.2783342119135477,
      "grad_norm": 0.017153726890683174,
      "learning_rate": 5.437460634017044e-05,
      "loss": 1.5758,
      "step": 33
    },
    {
      "epoch": 0.286768581971534,
      "grad_norm": 0.020006069913506508,
      "learning_rate": 5.3845633242400604e-05,
      "loss": 1.5774,
      "step": 34
    },
    {
      "epoch": 0.2952029520295203,
      "grad_norm": 0.016250574961304665,
      "learning_rate": 5.329571588154127e-05,
      "loss": 1.5748,
      "step": 35
    },
    {
      "epoch": 0.3036373220875066,
      "grad_norm": 0.019675249233841896,
      "learning_rate": 5.2725337264848605e-05,
      "loss": 1.5772,
      "step": 36
    },
    {
      "epoch": 0.3120716921454929,
      "grad_norm": 0.017005721107125282,
      "learning_rate": 5.213499837125182e-05,
      "loss": 1.5697,
      "step": 37
    },
    {
      "epoch": 0.32050606220347916,
      "grad_norm": 0.01664470136165619,
      "learning_rate": 5.152521771132993e-05,
      "loss": 1.5761,
      "step": 38
    },
    {
      "epoch": 0.32894043226146547,
      "grad_norm": 0.01764543540775776,
      "learning_rate": 5.0896530871889914e-05,
      "loss": 1.5793,
      "step": 39
    },
    {
      "epoch": 0.33737480231945177,
      "grad_norm": 0.016753442585468292,
      "learning_rate": 5.024949004554632e-05,
      "loss": 1.5658,
      "step": 40
    },
    {
      "epoch": 0.3458091723774381,
      "grad_norm": 0.019939422607421875,
      "learning_rate": 4.958466354571565e-05,
      "loss": 1.5762,
      "step": 41
    },
    {
      "epoch": 0.35424354243542433,
      "grad_norm": 0.01566561497747898,
      "learning_rate": 4.890263530745134e-05,
      "loss": 1.5703,
      "step": 42
    },
    {
      "epoch": 0.36267791249341064,
      "grad_norm": 0.015579808503389359,
      "learning_rate": 4.8204004374557806e-05,
      "loss": 1.577,
      "step": 43
    },
    {
      "epoch": 0.37111228255139694,
      "grad_norm": 0.016742996871471405,
      "learning_rate": 4.748938437343416e-05,
      "loss": 1.5726,
      "step": 44
    },
    {
      "epoch": 0.37954665260938325,
      "grad_norm": 0.017128925770521164,
      "learning_rate": 4.675940297410958e-05,
      "loss": 1.579,
      "step": 45
    },
    {
      "epoch": 0.38798102266736956,
      "grad_norm": 0.015266829170286655,
      "learning_rate": 4.601470133894373e-05,
      "loss": 1.5611,
      "step": 46
    },
    {
      "epoch": 0.3964153927253558,
      "grad_norm": 0.014922689646482468,
      "learning_rate": 4.525593355947662e-05,
      "loss": 1.5725,
      "step": 47
    },
    {
      "epoch": 0.4048497627833421,
      "grad_norm": 0.01651890017092228,
      "learning_rate": 4.448376608192235e-05,
      "loss": 1.5679,
      "step": 48
    },
    {
      "epoch": 0.4132841328413284,
      "grad_norm": 0.013002808205783367,
      "learning_rate": 4.3698877121811395e-05,
      "loss": 1.5712,
      "step": 49
    },
    {
      "epoch": 0.42171850289931473,
      "grad_norm": 0.013684232719242573,
      "learning_rate": 4.290195606829562e-05,
      "loss": 1.5683,
      "step": 50
    },
    {
      "epoch": 0.430152872957301,
      "grad_norm": 0.01470887940376997,
      "learning_rate": 4.2093702878639174e-05,
      "loss": 1.5784,
      "step": 51
    },
    {
      "epoch": 0.4385872430152873,
      "grad_norm": 0.013774153776466846,
      "learning_rate": 4.127482746342714e-05,
      "loss": 1.5648,
      "step": 52
    },
    {
      "epoch": 0.4470216130732736,
      "grad_norm": 0.01601037010550499,
      "learning_rate": 4.044604906303197e-05,
      "loss": 1.5671,
      "step": 53
    },
    {
      "epoch": 0.4554559831312599,
      "grad_norm": 0.013479109853506088,
      "learning_rate": 3.960809561588513e-05,
      "loss": 1.5759,
      "step": 54
    },
    {
      "epoch": 0.46389035318924615,
      "grad_norm": 0.01525378692895174,
      "learning_rate": 3.876170311910928e-05,
      "loss": 1.5672,
      "step": 55
    },
    {
      "epoch": 0.47232472324723246,
      "grad_norm": 0.013126607052981853,
      "learning_rate": 3.790761498207203e-05,
      "loss": 1.5744,
      "step": 56
    },
    {
      "epoch": 0.48075909330521877,
      "grad_norm": 0.013218970037996769,
      "learning_rate": 3.704658137342952e-05,
      "loss": 1.5688,
      "step": 57
    },
    {
      "epoch": 0.48919346336320507,
      "grad_norm": 0.014142030850052834,
      "learning_rate": 3.617935856223295e-05,
      "loss": 1.5742,
      "step": 58
    },
    {
      "epoch": 0.4976278334211914,
      "grad_norm": 0.013189482502639294,
      "learning_rate": 3.5306708253677186e-05,
      "loss": 1.5615,
      "step": 59
    },
    {
      "epoch": 0.5060622034791776,
      "grad_norm": 0.014055909588932991,
      "learning_rate": 3.442939692007444e-05,
      "loss": 1.5456,
      "step": 60
    },
    {
      "epoch": 0.5144965735371639,
      "grad_norm": 0.011999402195215225,
      "learning_rate": 3.354819512764097e-05,
      "loss": 1.5579,
      "step": 61
    },
    {
      "epoch": 0.5229309435951502,
      "grad_norm": 0.015170286409556866,
      "learning_rate": 3.2663876859688045e-05,
      "loss": 1.5606,
      "step": 62
    },
    {
      "epoch": 0.5313653136531366,
      "grad_norm": 0.013461374677717686,
      "learning_rate": 3.177721883681143e-05,
      "loss": 1.5631,
      "step": 63
    },
    {
      "epoch": 0.5397996837111229,
      "grad_norm": 0.014450161717832088,
      "learning_rate": 3.0888999834676796e-05,
      "loss": 1.5606,
      "step": 64
    },
    {
      "epoch": 0.5482340537691092,
      "grad_norm": 0.014033439569175243,
      "learning_rate": 3e-05,
      "loss": 1.5638,
      "step": 65
    },
    {
      "epoch": 0.5566684238270954,
      "grad_norm": 0.014029957354068756,
      "learning_rate": 2.9111000165323206e-05,
      "loss": 1.5656,
      "step": 66
    },
    {
      "epoch": 0.5651027938850817,
      "grad_norm": 0.016938265413045883,
      "learning_rate": 2.8222781163188573e-05,
      "loss": 1.5595,
      "step": 67
    },
    {
      "epoch": 0.573537163943068,
      "grad_norm": 0.014442404732108116,
      "learning_rate": 2.7336123140311957e-05,
      "loss": 1.5627,
      "step": 68
    },
    {
      "epoch": 0.5819715340010543,
      "grad_norm": 0.015609300695359707,
      "learning_rate": 2.645180487235903e-05,
      "loss": 1.5707,
      "step": 69
    },
    {
      "epoch": 0.5904059040590406,
      "grad_norm": 0.014037694782018661,
      "learning_rate": 2.557060307992557e-05,
      "loss": 1.5635,
      "step": 70
    },
    {
      "epoch": 0.5988402741170269,
      "grad_norm": 0.013035484589636326,
      "learning_rate": 2.469329174632282e-05,
      "loss": 1.5635,
      "step": 71
    },
    {
      "epoch": 0.6072746441750132,
      "grad_norm": 0.013149570673704147,
      "learning_rate": 2.3820641437767053e-05,
      "loss": 1.5607,
      "step": 72
    },
    {
      "epoch": 0.6157090142329995,
      "grad_norm": 0.01272524707019329,
      "learning_rate": 2.2953418626570494e-05,
      "loss": 1.5524,
      "step": 73
    },
    {
      "epoch": 0.6241433842909858,
      "grad_norm": 0.01219966635107994,
      "learning_rate": 2.209238501792798e-05,
      "loss": 1.555,
      "step": 74
    },
    {
      "epoch": 0.632577754348972,
      "grad_norm": 0.01229917537420988,
      "learning_rate": 2.123829688089073e-05,
      "loss": 1.5514,
      "step": 75
    },
    {
      "epoch": 0.6410121244069583,
      "grad_norm": 0.013784164562821388,
      "learning_rate": 2.0391904384114877e-05,
      "loss": 1.5614,
      "step": 76
    },
    {
      "epoch": 0.6494464944649446,
      "grad_norm": 0.010503321886062622,
      "learning_rate": 1.9553950936968042e-05,
      "loss": 1.541,
      "step": 77
    },
    {
      "epoch": 0.6578808645229309,
      "grad_norm": 0.012291346676647663,
      "learning_rate": 1.8725172536572863e-05,
      "loss": 1.556,
      "step": 78
    },
    {
      "epoch": 0.6663152345809172,
      "grad_norm": 0.011516911908984184,
      "learning_rate": 1.7906297121360838e-05,
      "loss": 1.5638,
      "step": 79
    },
    {
      "epoch": 0.6747496046389035,
      "grad_norm": 0.01181780081242323,
      "learning_rate": 1.7098043931704396e-05,
      "loss": 1.5508,
      "step": 80
    },
    {
      "epoch": 0.6831839746968899,
      "grad_norm": 0.010808738879859447,
      "learning_rate": 1.6301122878188607e-05,
      "loss": 1.5567,
      "step": 81
    },
    {
      "epoch": 0.6916183447548762,
      "grad_norm": 0.010649660602211952,
      "learning_rate": 1.551623391807766e-05,
      "loss": 1.5484,
      "step": 82
    },
    {
      "epoch": 0.7000527148128625,
      "grad_norm": 0.010580360889434814,
      "learning_rate": 1.4744066440523391e-05,
      "loss": 1.5591,
      "step": 83
    },
    {
      "epoch": 0.7084870848708487,
      "grad_norm": 0.010917909443378448,
      "learning_rate": 1.3985298661056292e-05,
      "loss": 1.569,
      "step": 84
    },
    {
      "epoch": 0.716921454928835,
      "grad_norm": 0.01177785824984312,
      "learning_rate": 1.324059702589043e-05,
      "loss": 1.5631,
      "step": 85
    },
    {
      "epoch": 0.7253558249868213,
      "grad_norm": 0.009857219643890858,
      "learning_rate": 1.2510615626565844e-05,
      "loss": 1.5561,
      "step": 86
    },
    {
      "epoch": 0.7337901950448076,
      "grad_norm": 0.011106839403510094,
      "learning_rate": 1.1795995625442208e-05,
      "loss": 1.5471,
      "step": 87
    },
    {
      "epoch": 0.7422245651027939,
      "grad_norm": 0.011377968825399876,
      "learning_rate": 1.109736469254867e-05,
      "loss": 1.5583,
      "step": 88
    },
    {
      "epoch": 0.7506589351607802,
      "grad_norm": 0.010118059813976288,
      "learning_rate": 1.0415336454284356e-05,
      "loss": 1.5531,
      "step": 89
    },
    {
      "epoch": 0.7590933052187665,
      "grad_norm": 0.01021275483071804,
      "learning_rate": 9.75050995445369e-06,
      "loss": 1.5559,
      "step": 90
    },
    {
      "epoch": 0.7675276752767528,
      "grad_norm": 0.00994526594877243,
      "learning_rate": 9.103469128110098e-06,
      "loss": 1.5527,
      "step": 91
    },
    {
      "epoch": 0.7759620453347391,
      "grad_norm": 0.01060432381927967,
      "learning_rate": 8.474782288670058e-06,
      "loss": 1.5514,
      "step": 92
    },
    {
      "epoch": 0.7843964153927253,
      "grad_norm": 0.011965557001531124,
      "learning_rate": 7.86500162874818e-06,
      "loss": 1.5536,
      "step": 93
    },
    {
      "epoch": 0.7928307854507116,
      "grad_norm": 0.010221057571470737,
      "learning_rate": 7.274662735151396e-06,
      "loss": 1.5541,
      "step": 94
    },
    {
      "epoch": 0.8012651555086979,
      "grad_norm": 0.01093184296041727,
      "learning_rate": 6.704284118458731e-06,
      "loss": 1.5512,
      "step": 95
    },
    {
      "epoch": 0.8096995255666842,
      "grad_norm": 0.010998157784342766,
      "learning_rate": 6.154366757599399e-06,
      "loss": 1.5492,
      "step": 96
    },
    {
      "epoch": 0.8181338956246705,
      "grad_norm": 0.01003272831439972,
      "learning_rate": 5.625393659829561e-06,
      "loss": 1.5472,
      "step": 97
    },
    {
      "epoch": 0.8265682656826568,
      "grad_norm": 0.010513346642255783,
      "learning_rate": 5.117829436493947e-06,
      "loss": 1.551,
      "step": 98
    },
    {
      "epoch": 0.8350026357406432,
      "grad_norm": 0.01016693189740181,
      "learning_rate": 4.632119894945215e-06,
      "loss": 1.5599,
      "step": 99
    },
    {
      "epoch": 0.8434370057986295,
      "grad_norm": 0.009756877087056637,
      "learning_rate": 4.1686916469793335e-06,
      "loss": 1.5552,
      "step": 100
    },
    {
      "epoch": 0.8518713758566157,
      "grad_norm": 0.010328919626772404,
      "learning_rate": 3.7279517341308977e-06,
      "loss": 1.5645,
      "step": 101
    },
    {
      "epoch": 0.860305745914602,
      "grad_norm": 0.009724525734782219,
      "learning_rate": 3.3102872701575838e-06,
      "loss": 1.5466,
      "step": 102
    },
    {
      "epoch": 0.8687401159725883,
      "grad_norm": 0.009452255442738533,
      "learning_rate": 2.916065101027694e-06,
      "loss": 1.555,
      "step": 103
    },
    {
      "epoch": 0.8771744860305746,
      "grad_norm": 0.009558911435306072,
      "learning_rate": 2.5456314827094463e-06,
      "loss": 1.5479,
      "step": 104
    },
    {
      "epoch": 0.8856088560885609,
      "grad_norm": 0.009129817597568035,
      "learning_rate": 2.1993117770449987e-06,
      "loss": 1.545,
      "step": 105
    },
    {
      "epoch": 0.8940432261465472,
      "grad_norm": 0.00930058490484953,
      "learning_rate": 1.8774101659763731e-06,
      "loss": 1.554,
      "step": 106
    },
    {
      "epoch": 0.9024775962045335,
      "grad_norm": 0.009718949906527996,
      "learning_rate": 1.5802093843742582e-06,
      "loss": 1.5467,
      "step": 107
    },
    {
      "epoch": 0.9109119662625198,
      "grad_norm": 0.009196877479553223,
      "learning_rate": 1.3079704717043273e-06,
      "loss": 1.55,
      "step": 108
    },
    {
      "epoch": 0.9193463363205061,
      "grad_norm": 0.00919976457953453,
      "learning_rate": 1.060932542749241e-06,
      "loss": 1.5558,
      "step": 109
    },
    {
      "epoch": 0.9277807063784923,
      "grad_norm": 0.0089542455971241,
      "learning_rate": 8.393125775876775e-07,
      "loss": 1.5563,
      "step": 110
    },
    {
      "epoch": 0.9362150764364786,
      "grad_norm": 0.009196256287395954,
      "learning_rate": 6.433052310148791e-07,
      "loss": 1.5537,
      "step": 111
    },
    {
      "epoch": 0.9446494464944649,
      "grad_norm": 0.009201628156006336,
      "learning_rate": 4.730826615720951e-07,
      "loss": 1.5567,
      "step": 112
    },
    {
      "epoch": 0.9530838165524512,
      "grad_norm": 0.008883966132998466,
      "learning_rate": 3.28794380335079e-07,
      "loss": 1.5549,
      "step": 113
    },
    {
      "epoch": 0.9615181866104375,
      "grad_norm": 0.009221088141202927,
      "learning_rate": 2.1056711959449247e-07,
      "loss": 1.5585,
      "step": 114
    },
    {
      "epoch": 0.9699525566684238,
      "grad_norm": 0.009092201478779316,
      "learning_rate": 1.1850472154349313e-07,
      "loss": 1.5536,
      "step": 115
    },
    {
      "epoch": 0.9783869267264101,
      "grad_norm": 0.009470025077462196,
      "learning_rate": 5.268804707035946e-08,
      "loss": 1.5705,
      "step": 116
    },
    {
      "epoch": 0.9868212967843965,
      "grad_norm": 0.008715336211025715,
      "learning_rate": 1.3174904736169557e-08,
      "loss": 1.5566,
      "step": 117
    },
    {
      "epoch": 0.9952556668423828,
      "grad_norm": 0.008857190608978271,
      "learning_rate": 0.0,
      "loss": 1.5464,
      "step": 118
    },
    {
      "epoch": 0.9952556668423828,
      "step": 118,
      "total_flos": 1660937136242688.0,
      "train_loss": 1.571211524939133,
      "train_runtime": 47361.2024,
      "train_samples_per_second": 0.641,
      "train_steps_per_second": 0.002
    }
  ],
  "logging_steps": 1,
  "max_steps": 118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1660937136242688.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}