{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.40210627094303497,
  "eval_steps": 500,
  "global_step": 210,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0019147917663954045,
      "grad_norm": 2.9491562843322754,
      "learning_rate": 0.0,
      "loss": 0.6229,
      "step": 1
    },
    {
      "epoch": 0.003829583532790809,
      "grad_norm": 3.0646867752075195,
      "learning_rate": 3.846153846153847e-07,
      "loss": 0.6119,
      "step": 2
    },
    {
      "epoch": 0.0057443752991862135,
      "grad_norm": 3.0737922191619873,
      "learning_rate": 7.692307692307694e-07,
      "loss": 0.6582,
      "step": 3
    },
    {
      "epoch": 0.007659167065581618,
      "grad_norm": 2.9172728061676025,
      "learning_rate": 1.153846153846154e-06,
      "loss": 0.6209,
      "step": 4
    },
    {
      "epoch": 0.009573958831977022,
      "grad_norm": 2.668588161468506,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.5589,
      "step": 5
    },
    {
      "epoch": 0.011488750598372427,
      "grad_norm": 3.2810585498809814,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 0.5968,
      "step": 6
    },
    {
      "epoch": 0.013403542364767831,
      "grad_norm": 2.434365749359131,
      "learning_rate": 2.307692307692308e-06,
      "loss": 0.5636,
      "step": 7
    },
    {
      "epoch": 0.015318334131163236,
      "grad_norm": 2.060615301132202,
      "learning_rate": 2.6923076923076923e-06,
      "loss": 0.5661,
      "step": 8
    },
    {
      "epoch": 0.01723312589755864,
      "grad_norm": 1.8817814588546753,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.5817,
      "step": 9
    },
    {
      "epoch": 0.019147917663954045,
      "grad_norm": 1.766438603401184,
      "learning_rate": 3.4615384615384617e-06,
      "loss": 0.5529,
      "step": 10
    },
    {
      "epoch": 0.02106270943034945,
      "grad_norm": 1.5240556001663208,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.5207,
      "step": 11
    },
    {
      "epoch": 0.022977501196744854,
      "grad_norm": 1.5381622314453125,
      "learning_rate": 4.230769230769231e-06,
      "loss": 0.5171,
      "step": 12
    },
    {
      "epoch": 0.02489229296314026,
      "grad_norm": 1.4144328832626343,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.5612,
      "step": 13
    },
    {
      "epoch": 0.026807084729535663,
      "grad_norm": 1.282257318496704,
      "learning_rate": 5e-06,
      "loss": 0.493,
      "step": 14
    },
    {
      "epoch": 0.028721876495931067,
      "grad_norm": 1.3273121118545532,
      "learning_rate": 5.384615384615385e-06,
      "loss": 0.4723,
      "step": 15
    },
    {
      "epoch": 0.030636668262326472,
      "grad_norm": 1.1829627752304077,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.4675,
      "step": 16
    },
    {
      "epoch": 0.032551460028721876,
      "grad_norm": 1.0885576009750366,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.4275,
      "step": 17
    },
    {
      "epoch": 0.03446625179511728,
      "grad_norm": 0.9974104762077332,
      "learning_rate": 6.538461538461539e-06,
      "loss": 0.4709,
      "step": 18
    },
    {
      "epoch": 0.036381043561512685,
      "grad_norm": 1.0769761800765991,
      "learning_rate": 6.923076923076923e-06,
      "loss": 0.4916,
      "step": 19
    },
    {
      "epoch": 0.03829583532790809,
      "grad_norm": 0.967096745967865,
      "learning_rate": 7.307692307692308e-06,
      "loss": 0.4785,
      "step": 20
    },
    {
      "epoch": 0.040210627094303494,
      "grad_norm": 1.0460747480392456,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.4653,
      "step": 21
    },
    {
      "epoch": 0.0421254188606989,
      "grad_norm": 1.0114920139312744,
      "learning_rate": 8.076923076923077e-06,
      "loss": 0.4648,
      "step": 22
    },
    {
      "epoch": 0.0440402106270943,
      "grad_norm": 1.1619290113449097,
      "learning_rate": 8.461538461538462e-06,
      "loss": 0.4833,
      "step": 23
    },
    {
      "epoch": 0.04595500239348971,
      "grad_norm": 0.9872665405273438,
      "learning_rate": 8.846153846153847e-06,
      "loss": 0.4545,
      "step": 24
    },
    {
      "epoch": 0.04786979415988511,
      "grad_norm": 0.9702840447425842,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.4651,
      "step": 25
    },
    {
      "epoch": 0.04978458592628052,
      "grad_norm": 0.9493695497512817,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.477,
      "step": 26
    },
    {
      "epoch": 0.05169937769267592,
      "grad_norm": 0.9152507185935974,
      "learning_rate": 1e-05,
      "loss": 0.4499,
      "step": 27
    },
    {
      "epoch": 0.053614169459071326,
      "grad_norm": 1.0640617609024048,
      "learning_rate": 9.999899706000774e-06,
      "loss": 0.4853,
      "step": 28
    },
    {
      "epoch": 0.05552896122546673,
      "grad_norm": 0.9641034603118896,
      "learning_rate": 9.999598828026644e-06,
      "loss": 0.475,
      "step": 29
    },
    {
      "epoch": 0.057443752991862135,
      "grad_norm": 0.8927161693572998,
      "learning_rate": 9.999097378148116e-06,
      "loss": 0.4448,
      "step": 30
    },
    {
      "epoch": 0.05935854475825754,
      "grad_norm": 0.881844699382782,
      "learning_rate": 9.998395376482152e-06,
      "loss": 0.4327,
      "step": 31
    },
    {
      "epoch": 0.061273336524652944,
      "grad_norm": 0.8794113993644714,
      "learning_rate": 9.99749285119138e-06,
      "loss": 0.4294,
      "step": 32
    },
    {
      "epoch": 0.06318812829104835,
      "grad_norm": 0.9898825287818909,
      "learning_rate": 9.996389838482942e-06,
      "loss": 0.5294,
      "step": 33
    },
    {
      "epoch": 0.06510292005744375,
      "grad_norm": 0.9184749126434326,
      "learning_rate": 9.995086382607064e-06,
      "loss": 0.4774,
      "step": 34
    },
    {
      "epoch": 0.06701771182383916,
      "grad_norm": 0.9067336320877075,
      "learning_rate": 9.993582535855265e-06,
      "loss": 0.4569,
      "step": 35
    },
    {
      "epoch": 0.06893250359023456,
      "grad_norm": 0.8807307481765747,
      "learning_rate": 9.991878358558267e-06,
      "loss": 0.478,
      "step": 36
    },
    {
      "epoch": 0.07084729535662997,
      "grad_norm": 0.9359887838363647,
      "learning_rate": 9.989973919083576e-06,
      "loss": 0.4659,
      "step": 37
    },
    {
      "epoch": 0.07276208712302537,
      "grad_norm": 0.9008484482765198,
      "learning_rate": 9.987869293832727e-06,
      "loss": 0.4659,
      "step": 38
    },
    {
      "epoch": 0.07467687888942078,
      "grad_norm": 0.8065485954284668,
      "learning_rate": 9.985564567238237e-06,
      "loss": 0.4441,
      "step": 39
    },
    {
      "epoch": 0.07659167065581618,
      "grad_norm": 0.9766021966934204,
      "learning_rate": 9.983059831760205e-06,
      "loss": 0.4834,
      "step": 40
    },
    {
      "epoch": 0.07850646242221158,
      "grad_norm": 0.8222993016242981,
      "learning_rate": 9.980355187882606e-06,
      "loss": 0.443,
      "step": 41
    },
    {
      "epoch": 0.08042125418860699,
      "grad_norm": 0.8215630054473877,
      "learning_rate": 9.977450744109258e-06,
      "loss": 0.4219,
      "step": 42
    },
    {
      "epoch": 0.0823360459550024,
      "grad_norm": 0.8324375748634338,
      "learning_rate": 9.974346616959476e-06,
      "loss": 0.4362,
      "step": 43
    },
    {
      "epoch": 0.0842508377213978,
      "grad_norm": 0.9242782592773438,
      "learning_rate": 9.97104293096339e-06,
      "loss": 0.4738,
      "step": 44
    },
    {
      "epoch": 0.0861656294877932,
      "grad_norm": 0.9275208711624146,
      "learning_rate": 9.967539818656953e-06,
      "loss": 0.4571,
      "step": 45
    },
    {
      "epoch": 0.0880804212541886,
      "grad_norm": 0.876868724822998,
      "learning_rate": 9.96383742057662e-06,
      "loss": 0.5172,
      "step": 46
    },
    {
      "epoch": 0.08999521302058401,
      "grad_norm": 0.8446276783943176,
      "learning_rate": 9.959935885253715e-06,
      "loss": 0.4457,
      "step": 47
    },
    {
      "epoch": 0.09191000478697942,
      "grad_norm": 0.8077015280723572,
      "learning_rate": 9.955835369208475e-06,
      "loss": 0.4234,
      "step": 48
    },
    {
      "epoch": 0.09382479655337482,
      "grad_norm": 0.7882896065711975,
      "learning_rate": 9.951536036943753e-06,
      "loss": 0.4264,
      "step": 49
    },
    {
      "epoch": 0.09573958831977022,
      "grad_norm": 0.8539751768112183,
      "learning_rate": 9.94703806093845e-06,
      "loss": 0.461,
      "step": 50
    },
    {
      "epoch": 0.09765438008616563,
      "grad_norm": 0.8285911679267883,
      "learning_rate": 9.942341621640558e-06,
      "loss": 0.4379,
      "step": 51
    },
    {
      "epoch": 0.09956917185256103,
      "grad_norm": 0.8029133081436157,
      "learning_rate": 9.937446907459954e-06,
      "loss": 0.4565,
      "step": 52
    },
    {
      "epoch": 0.10148396361895644,
      "grad_norm": 0.7964851260185242,
      "learning_rate": 9.932354114760819e-06,
      "loss": 0.4262,
      "step": 53
    },
    {
      "epoch": 0.10339875538535184,
      "grad_norm": 0.9846324920654297,
      "learning_rate": 9.92706344785377e-06,
      "loss": 0.5302,
      "step": 54
    },
    {
      "epoch": 0.10531354715174725,
      "grad_norm": 0.7648650407791138,
      "learning_rate": 9.921575118987672e-06,
      "loss": 0.4066,
      "step": 55
    },
    {
      "epoch": 0.10722833891814265,
      "grad_norm": 0.83173668384552,
      "learning_rate": 9.915889348341098e-06,
      "loss": 0.4438,
      "step": 56
    },
    {
      "epoch": 0.10914313068453806,
      "grad_norm": 0.7968882322311401,
      "learning_rate": 9.910006364013522e-06,
      "loss": 0.407,
      "step": 57
    },
    {
      "epoch": 0.11105792245093346,
      "grad_norm": 0.8423118591308594,
      "learning_rate": 9.903926402016153e-06,
      "loss": 0.4174,
      "step": 58
    },
    {
      "epoch": 0.11297271421732887,
      "grad_norm": 0.9054727554321289,
      "learning_rate": 9.897649706262474e-06,
      "loss": 0.4764,
      "step": 59
    },
    {
      "epoch": 0.11488750598372427,
      "grad_norm": 0.8318431973457336,
      "learning_rate": 9.891176528558451e-06,
      "loss": 0.4326,
      "step": 60
    },
    {
      "epoch": 0.11680229775011967,
      "grad_norm": 0.8409565687179565,
      "learning_rate": 9.884507128592435e-06,
      "loss": 0.4451,
      "step": 61
    },
    {
      "epoch": 0.11871708951651508,
      "grad_norm": 0.8471431136131287,
      "learning_rate": 9.877641773924748e-06,
      "loss": 0.4217,
      "step": 62
    },
    {
      "epoch": 0.12063188128291048,
      "grad_norm": 0.8495103120803833,
      "learning_rate": 9.870580739976936e-06,
      "loss": 0.421,
      "step": 63
    },
    {
      "epoch": 0.12254667304930589,
      "grad_norm": 0.8164567947387695,
      "learning_rate": 9.863324310020735e-06,
      "loss": 0.4266,
      "step": 64
    },
    {
      "epoch": 0.12446146481570129,
      "grad_norm": 0.8732247948646545,
      "learning_rate": 9.855872775166696e-06,
      "loss": 0.4661,
      "step": 65
    },
    {
      "epoch": 0.1263762565820967,
      "grad_norm": 0.8157728910446167,
      "learning_rate": 9.848226434352513e-06,
      "loss": 0.4401,
      "step": 66
    },
    {
      "epoch": 0.12829104834849211,
      "grad_norm": 0.8860891461372375,
      "learning_rate": 9.840385594331022e-06,
      "loss": 0.4748,
      "step": 67
    },
    {
      "epoch": 0.1302058401148875,
      "grad_norm": 0.8987312316894531,
      "learning_rate": 9.83235056965791e-06,
      "loss": 0.4881,
      "step": 68
    },
    {
      "epoch": 0.13212063188128292,
      "grad_norm": 0.8786044716835022,
      "learning_rate": 9.824121682679072e-06,
      "loss": 0.4417,
      "step": 69
    },
    {
      "epoch": 0.13403542364767831,
      "grad_norm": 0.8325650691986084,
      "learning_rate": 9.815699263517712e-06,
      "loss": 0.4377,
      "step": 70
    },
    {
      "epoch": 0.13595021541407373,
      "grad_norm": 0.8149142861366272,
      "learning_rate": 9.807083650061063e-06,
      "loss": 0.4496,
      "step": 71
    },
    {
      "epoch": 0.13786500718046912,
      "grad_norm": 0.8394611477851868,
      "learning_rate": 9.798275187946859e-06,
      "loss": 0.4394,
      "step": 72
    },
    {
      "epoch": 0.13977979894686454,
      "grad_norm": 0.7746449112892151,
      "learning_rate": 9.789274230549456e-06,
      "loss": 0.4039,
      "step": 73
    },
    {
      "epoch": 0.14169459071325993,
      "grad_norm": 0.7592336535453796,
      "learning_rate": 9.780081138965663e-06,
      "loss": 0.3788,
      "step": 74
    },
    {
      "epoch": 0.14360938247965535,
      "grad_norm": 0.9066088199615479,
      "learning_rate": 9.770696282000245e-06,
      "loss": 0.4541,
      "step": 75
    },
    {
      "epoch": 0.14552417424605074,
      "grad_norm": 0.8512394428253174,
      "learning_rate": 9.761120036151138e-06,
      "loss": 0.4217,
      "step": 76
    },
    {
      "epoch": 0.14743896601244616,
      "grad_norm": 0.795378565788269,
      "learning_rate": 9.751352785594337e-06,
      "loss": 0.4014,
      "step": 77
    },
    {
      "epoch": 0.14935375777884155,
      "grad_norm": 0.9467825293540955,
      "learning_rate": 9.741394922168495e-06,
      "loss": 0.4855,
      "step": 78
    },
    {
      "epoch": 0.15126854954523697,
      "grad_norm": 0.7824875712394714,
      "learning_rate": 9.731246845359187e-06,
      "loss": 0.4088,
      "step": 79
    },
    {
      "epoch": 0.15318334131163236,
      "grad_norm": 0.7557615637779236,
      "learning_rate": 9.720908962282893e-06,
      "loss": 0.4023,
      "step": 80
    },
    {
      "epoch": 0.15509813307802778,
      "grad_norm": 0.8093947768211365,
      "learning_rate": 9.710381687670675e-06,
      "loss": 0.4345,
      "step": 81
    },
    {
      "epoch": 0.15701292484442317,
      "grad_norm": 0.8901275396347046,
      "learning_rate": 9.699665443851518e-06,
      "loss": 0.4444,
      "step": 82
    },
    {
      "epoch": 0.1589277166108186,
      "grad_norm": 0.7518415451049805,
      "learning_rate": 9.688760660735403e-06,
      "loss": 0.4024,
      "step": 83
    },
    {
      "epoch": 0.16084250837721398,
      "grad_norm": 0.7495772242546082,
      "learning_rate": 9.677667775796052e-06,
      "loss": 0.4005,
      "step": 84
    },
    {
      "epoch": 0.1627573001436094,
      "grad_norm": 0.8903560638427734,
      "learning_rate": 9.666387234053385e-06,
      "loss": 0.4495,
      "step": 85
    },
    {
      "epoch": 0.1646720919100048,
      "grad_norm": 0.8854427933692932,
      "learning_rate": 9.654919488055656e-06,
      "loss": 0.4381,
      "step": 86
    },
    {
      "epoch": 0.1665868836764002,
      "grad_norm": 0.8393151164054871,
      "learning_rate": 9.643264997861312e-06,
      "loss": 0.4177,
      "step": 87
    },
    {
      "epoch": 0.1685016754427956,
      "grad_norm": 0.8448845148086548,
      "learning_rate": 9.631424231020523e-06,
      "loss": 0.4437,
      "step": 88
    },
    {
      "epoch": 0.170416467209191,
      "grad_norm": 0.8987253904342651,
      "learning_rate": 9.619397662556434e-06,
      "loss": 0.4479,
      "step": 89
    },
    {
      "epoch": 0.1723312589755864,
      "grad_norm": 0.9512760639190674,
      "learning_rate": 9.607185774946106e-06,
      "loss": 0.5188,
      "step": 90
    },
    {
      "epoch": 0.17424605074198182,
      "grad_norm": 0.9057194590568542,
      "learning_rate": 9.594789058101154e-06,
      "loss": 0.4448,
      "step": 91
    },
    {
      "epoch": 0.1761608425083772,
      "grad_norm": 0.8147549033164978,
      "learning_rate": 9.582208009348104e-06,
      "loss": 0.4106,
      "step": 92
    },
    {
      "epoch": 0.17807563427477263,
      "grad_norm": 0.8666926622390747,
      "learning_rate": 9.569443133408434e-06,
      "loss": 0.4558,
      "step": 93
    },
    {
      "epoch": 0.17999042604116802,
      "grad_norm": 0.8677969574928284,
      "learning_rate": 9.556494942378328e-06,
      "loss": 0.4379,
      "step": 94
    },
    {
      "epoch": 0.18190521780756344,
      "grad_norm": 0.8896477222442627,
      "learning_rate": 9.543363955708124e-06,
      "loss": 0.4498,
      "step": 95
    },
    {
      "epoch": 0.18382000957395883,
      "grad_norm": 0.7357858419418335,
      "learning_rate": 9.530050700181499e-06,
      "loss": 0.3666,
      "step": 96
    },
    {
      "epoch": 0.18573480134035425,
      "grad_norm": 0.7851715683937073,
      "learning_rate": 9.5165557098943e-06,
      "loss": 0.411,
      "step": 97
    },
    {
      "epoch": 0.18764959310674964,
      "grad_norm": 0.8098123669624329,
      "learning_rate": 9.502879526233151e-06,
      "loss": 0.4023,
      "step": 98
    },
    {
      "epoch": 0.18956438487314506,
      "grad_norm": 0.8245725631713867,
      "learning_rate": 9.48902269785371e-06,
      "loss": 0.423,
      "step": 99
    },
    {
      "epoch": 0.19147917663954045,
      "grad_norm": 0.8497715592384338,
      "learning_rate": 9.47498578065867e-06,
      "loss": 0.4125,
      "step": 100
    },
    {
      "epoch": 0.19339396840593587,
      "grad_norm": 0.8205481171607971,
      "learning_rate": 9.460769337775461e-06,
      "loss": 0.4312,
      "step": 101
    },
    {
      "epoch": 0.19530876017233126,
      "grad_norm": 0.8062931299209595,
      "learning_rate": 9.446373939533642e-06,
      "loss": 0.3961,
      "step": 102
    },
    {
      "epoch": 0.19722355193872668,
      "grad_norm": 0.8209528923034668,
      "learning_rate": 9.431800163442043e-06,
      "loss": 0.4121,
      "step": 103
    },
    {
      "epoch": 0.19913834370512207,
      "grad_norm": 0.8154571652412415,
      "learning_rate": 9.417048594165572e-06,
      "loss": 0.4475,
      "step": 104
    },
    {
      "epoch": 0.20105313547151749,
      "grad_norm": 0.8546404838562012,
      "learning_rate": 9.402119823501787e-06,
      "loss": 0.4293,
      "step": 105
    },
    {
      "epoch": 0.20296792723791288,
      "grad_norm": 0.8470130562782288,
      "learning_rate": 9.387014450357128e-06,
      "loss": 0.4139,
      "step": 106
    },
    {
      "epoch": 0.2048827190043083,
      "grad_norm": 0.9199275970458984,
      "learning_rate": 9.371733080722911e-06,
      "loss": 0.4825,
      "step": 107
    },
    {
      "epoch": 0.20679751077070369,
      "grad_norm": 0.9049551486968994,
      "learning_rate": 9.356276327651006e-06,
      "loss": 0.4378,
      "step": 108
    },
    {
      "epoch": 0.2087123025370991,
      "grad_norm": 0.8089979887008667,
      "learning_rate": 9.340644811229243e-06,
      "loss": 0.4027,
      "step": 109
    },
    {
      "epoch": 0.2106270943034945,
      "grad_norm": 0.7452864050865173,
      "learning_rate": 9.324839158556542e-06,
      "loss": 0.3795,
      "step": 110
    },
    {
      "epoch": 0.2125418860698899,
      "grad_norm": 0.8286869525909424,
      "learning_rate": 9.308860003717748e-06,
      "loss": 0.4137,
      "step": 111
    },
    {
      "epoch": 0.2144566778362853,
      "grad_norm": 0.8634768724441528,
      "learning_rate": 9.292707987758202e-06,
      "loss": 0.445,
      "step": 112
    },
    {
      "epoch": 0.21637146960268072,
      "grad_norm": 0.8329188227653503,
      "learning_rate": 9.27638375865801e-06,
      "loss": 0.4307,
      "step": 113
    },
    {
      "epoch": 0.2182862613690761,
      "grad_norm": 0.8780718445777893,
      "learning_rate": 9.259887971306064e-06,
      "loss": 0.4863,
      "step": 114
    },
    {
      "epoch": 0.22020105313547153,
      "grad_norm": 0.9007835388183594,
      "learning_rate": 9.243221287473755e-06,
      "loss": 0.4482,
      "step": 115
    },
    {
      "epoch": 0.22211584490186692,
      "grad_norm": 0.8163229823112488,
      "learning_rate": 9.226384375788435e-06,
      "loss": 0.4168,
      "step": 116
    },
    {
      "epoch": 0.22403063666826234,
      "grad_norm": 0.8288677334785461,
      "learning_rate": 9.209377911706585e-06,
      "loss": 0.4038,
      "step": 117
    },
    {
      "epoch": 0.22594542843465773,
      "grad_norm": 0.8035851716995239,
      "learning_rate": 9.192202577486725e-06,
      "loss": 0.3922,
      "step": 118
    },
    {
      "epoch": 0.22786022020105315,
      "grad_norm": 0.8203516006469727,
      "learning_rate": 9.174859062162037e-06,
      "loss": 0.3971,
      "step": 119
    },
    {
      "epoch": 0.22977501196744854,
      "grad_norm": 0.8246352076530457,
      "learning_rate": 9.157348061512728e-06,
      "loss": 0.4433,
      "step": 120
    },
    {
      "epoch": 0.23168980373384396,
      "grad_norm": 0.8655344247817993,
      "learning_rate": 9.139670278038109e-06,
      "loss": 0.4405,
      "step": 121
    },
    {
      "epoch": 0.23360459550023935,
      "grad_norm": 0.7439157366752625,
      "learning_rate": 9.121826420928421e-06,
      "loss": 0.3683,
      "step": 122
    },
    {
      "epoch": 0.23551938726663477,
      "grad_norm": 0.817434549331665,
      "learning_rate": 9.103817206036383e-06,
      "loss": 0.4034,
      "step": 123
    },
    {
      "epoch": 0.23743417903303016,
      "grad_norm": 0.8455221056938171,
      "learning_rate": 9.085643355848468e-06,
      "loss": 0.4418,
      "step": 124
    },
    {
      "epoch": 0.23934897079942558,
      "grad_norm": 0.8356925845146179,
      "learning_rate": 9.06730559945592e-06,
      "loss": 0.4012,
      "step": 125
    },
    {
      "epoch": 0.24126376256582097,
      "grad_norm": 0.8181227445602417,
      "learning_rate": 9.048804672525513e-06,
      "loss": 0.4174,
      "step": 126
    },
    {
      "epoch": 0.24317855433221638,
      "grad_norm": 0.8010542988777161,
      "learning_rate": 9.030141317270026e-06,
      "loss": 0.3952,
      "step": 127
    },
    {
      "epoch": 0.24509334609861178,
      "grad_norm": 0.8500829935073853,
      "learning_rate": 9.011316282418474e-06,
      "loss": 0.4123,
      "step": 128
    },
    {
      "epoch": 0.2470081378650072,
      "grad_norm": 0.8971666693687439,
      "learning_rate": 8.992330323186069e-06,
      "loss": 0.4451,
      "step": 129
    },
    {
      "epoch": 0.24892292963140258,
      "grad_norm": 0.9065473079681396,
      "learning_rate": 8.973184201243922e-06,
      "loss": 0.4821,
      "step": 130
    },
    {
      "epoch": 0.250837721397798,
      "grad_norm": 0.8722876906394958,
      "learning_rate": 8.953878684688492e-06,
      "loss": 0.4204,
      "step": 131
    },
    {
      "epoch": 0.2527525131641934,
      "grad_norm": 0.8343362808227539,
      "learning_rate": 8.934414548010764e-06,
      "loss": 0.408,
      "step": 132
    },
    {
      "epoch": 0.2546673049305888,
      "grad_norm": 0.8162686824798584,
      "learning_rate": 8.914792572065178e-06,
      "loss": 0.416,
      "step": 133
    },
    {
      "epoch": 0.25658209669698423,
      "grad_norm": 0.9116921424865723,
      "learning_rate": 8.89501354403831e-06,
      "loss": 0.4589,
      "step": 134
    },
    {
      "epoch": 0.2584968884633796,
      "grad_norm": 0.9577599763870239,
      "learning_rate": 8.875078257417294e-06,
      "loss": 0.4654,
      "step": 135
    },
    {
      "epoch": 0.260411680229775,
      "grad_norm": 0.8709072470664978,
      "learning_rate": 8.854987511957974e-06,
      "loss": 0.4395,
      "step": 136
    },
    {
      "epoch": 0.26232647199617043,
      "grad_norm": 0.8386030197143555,
      "learning_rate": 8.834742113652835e-06,
      "loss": 0.4281,
      "step": 137
    },
    {
      "epoch": 0.26424126376256585,
      "grad_norm": 0.7646230459213257,
      "learning_rate": 8.81434287469866e-06,
      "loss": 0.3804,
      "step": 138
    },
    {
      "epoch": 0.2661560555289612,
      "grad_norm": 0.8096075057983398,
      "learning_rate": 8.793790613463956e-06,
      "loss": 0.4112,
      "step": 139
    },
    {
      "epoch": 0.26807084729535663,
      "grad_norm": 0.8051929473876953,
      "learning_rate": 8.773086154456106e-06,
      "loss": 0.4172,
      "step": 140
    },
    {
      "epoch": 0.26998563906175205,
      "grad_norm": 0.9208196401596069,
      "learning_rate": 8.752230328288314e-06,
      "loss": 0.4768,
      "step": 141
    },
    {
      "epoch": 0.27190043082814747,
      "grad_norm": 0.7890869975090027,
      "learning_rate": 8.731223971646261e-06,
      "loss": 0.3915,
      "step": 142
    },
    {
      "epoch": 0.27381522259454283,
      "grad_norm": 0.786723792552948,
      "learning_rate": 8.710067927254555e-06,
      "loss": 0.3844,
      "step": 143
    },
    {
      "epoch": 0.27573001436093825,
      "grad_norm": 0.791117250919342,
      "learning_rate": 8.688763043842916e-06,
      "loss": 0.4065,
      "step": 144
    },
    {
      "epoch": 0.27764480612733367,
      "grad_norm": 0.8172312378883362,
      "learning_rate": 8.66731017611213e-06,
      "loss": 0.4337,
      "step": 145
    },
    {
      "epoch": 0.2795595978937291,
      "grad_norm": 0.8335762023925781,
      "learning_rate": 8.645710184699756e-06,
      "loss": 0.4182,
      "step": 146
    },
    {
      "epoch": 0.28147438966012445,
      "grad_norm": 0.8034957051277161,
      "learning_rate": 8.6239639361456e-06,
      "loss": 0.4097,
      "step": 147
    },
    {
      "epoch": 0.28338918142651986,
      "grad_norm": 0.8107390403747559,
      "learning_rate": 8.602072302856961e-06,
      "loss": 0.4055,
      "step": 148
    },
    {
      "epoch": 0.2853039731929153,
      "grad_norm": 0.8442232012748718,
      "learning_rate": 8.580036163073615e-06,
      "loss": 0.4307,
      "step": 149
    },
    {
      "epoch": 0.2872187649593107,
      "grad_norm": 0.8290265202522278,
      "learning_rate": 8.5578564008326e-06,
      "loss": 0.3892,
      "step": 150
    },
    {
      "epoch": 0.28913355672570606,
      "grad_norm": 0.8057438731193542,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.4042,
      "step": 151
    },
    {
      "epoch": 0.2910483484921015,
      "grad_norm": 0.8582248091697693,
      "learning_rate": 8.513069573898944e-06,
      "loss": 0.4149,
      "step": 152
    },
    {
      "epoch": 0.2929631402584969,
      "grad_norm": 0.8402311205863953,
      "learning_rate": 8.490464305946296e-06,
      "loss": 0.4243,
      "step": 153
    },
    {
      "epoch": 0.2948779320248923,
      "grad_norm": 0.812869668006897,
      "learning_rate": 8.467719008943886e-06,
      "loss": 0.4134,
      "step": 154
    },
    {
      "epoch": 0.2967927237912877,
      "grad_norm": 0.8431028723716736,
      "learning_rate": 8.444834595378434e-06,
      "loss": 0.4185,
      "step": 155
    },
    {
      "epoch": 0.2987075155576831,
      "grad_norm": 0.802760899066925,
      "learning_rate": 8.421811983317682e-06,
      "loss": 0.4011,
      "step": 156
    },
    {
      "epoch": 0.3006223073240785,
      "grad_norm": 0.814274251461029,
      "learning_rate": 8.398652096373566e-06,
      "loss": 0.4194,
      "step": 157
    },
    {
      "epoch": 0.30253709909047394,
      "grad_norm": 0.8286414742469788,
      "learning_rate": 8.375355863665155e-06,
      "loss": 0.4044,
      "step": 158
    },
    {
      "epoch": 0.3044518908568693,
      "grad_norm": 0.8244617581367493,
      "learning_rate": 8.351924219781393e-06,
      "loss": 0.4415,
      "step": 159
    },
    {
      "epoch": 0.3063666826232647,
      "grad_norm": 0.8288456201553345,
      "learning_rate": 8.328358104743588e-06,
      "loss": 0.4143,
      "step": 160
    },
    {
      "epoch": 0.30828147438966014,
      "grad_norm": 0.7895364165306091,
      "learning_rate": 8.304658463967705e-06,
      "loss": 0.4122,
      "step": 161
    },
    {
      "epoch": 0.31019626615605556,
      "grad_norm": 0.7923944592475891,
      "learning_rate": 8.28082624822645e-06,
      "loss": 0.3812,
      "step": 162
    },
    {
      "epoch": 0.3121110579224509,
      "grad_norm": 0.7424578666687012,
      "learning_rate": 8.256862413611113e-06,
      "loss": 0.3883,
      "step": 163
    },
    {
      "epoch": 0.31402584968884634,
      "grad_norm": 0.8261198401451111,
      "learning_rate": 8.232767921493216e-06,
      "loss": 0.432,
      "step": 164
    },
    {
      "epoch": 0.31594064145524176,
      "grad_norm": 0.8710785508155823,
      "learning_rate": 8.20854373848595e-06,
      "loss": 0.4508,
      "step": 165
    },
    {
      "epoch": 0.3178554332216372,
      "grad_norm": 0.7583726048469543,
      "learning_rate": 8.184190836405394e-06,
      "loss": 0.3709,
      "step": 166
    },
    {
      "epoch": 0.31977022498803254,
      "grad_norm": 0.7795834541320801,
      "learning_rate": 8.15971019223152e-06,
      "loss": 0.4055,
      "step": 167
    },
    {
      "epoch": 0.32168501675442795,
      "grad_norm": 0.7580612897872925,
      "learning_rate": 8.135102788069015e-06,
      "loss": 0.3605,
      "step": 168
    },
    {
      "epoch": 0.3235998085208234,
      "grad_norm": 0.7536636590957642,
      "learning_rate": 8.110369611107869e-06,
      "loss": 0.3656,
      "step": 169
    },
    {
      "epoch": 0.3255146002872188,
      "grad_norm": 0.8029680252075195,
      "learning_rate": 8.085511653583772e-06,
      "loss": 0.3819,
      "step": 170
    },
    {
      "epoch": 0.32742939205361415,
      "grad_norm": 0.8548794388771057,
      "learning_rate": 8.060529912738316e-06,
      "loss": 0.4449,
      "step": 171
    },
    {
      "epoch": 0.3293441838200096,
      "grad_norm": 0.877955436706543,
      "learning_rate": 8.035425390778975e-06,
      "loss": 0.4504,
      "step": 172
    },
    {
      "epoch": 0.331258975586405,
      "grad_norm": 0.8173900246620178,
      "learning_rate": 8.010199094838915e-06,
      "loss": 0.4211,
      "step": 173
    },
    {
      "epoch": 0.3331737673528004,
      "grad_norm": 0.8715358972549438,
      "learning_rate": 7.984852036936578e-06,
      "loss": 0.3909,
      "step": 174
    },
    {
      "epoch": 0.3350885591191958,
      "grad_norm": 0.8475743532180786,
      "learning_rate": 7.959385233935087e-06,
      "loss": 0.4416,
      "step": 175
    },
    {
      "epoch": 0.3370033508855912,
      "grad_norm": 0.7483753561973572,
      "learning_rate": 7.933799707501448e-06,
      "loss": 0.351,
      "step": 176
    },
    {
      "epoch": 0.3389181426519866,
      "grad_norm": 0.8065423965454102,
      "learning_rate": 7.908096484065569e-06,
      "loss": 0.4085,
      "step": 177
    },
    {
      "epoch": 0.340832934418382,
      "grad_norm": 0.8215972185134888,
      "learning_rate": 7.88227659477908e-06,
      "loss": 0.4132,
      "step": 178
    },
    {
      "epoch": 0.3427477261847774,
      "grad_norm": 0.7788512706756592,
      "learning_rate": 7.856341075473963e-06,
      "loss": 0.3828,
      "step": 179
    },
    {
      "epoch": 0.3446625179511728,
      "grad_norm": 0.7943012118339539,
      "learning_rate": 7.830290966620997e-06,
      "loss": 0.3737,
      "step": 180
    },
    {
      "epoch": 0.3465773097175682,
      "grad_norm": 0.8680888414382935,
      "learning_rate": 7.804127313288023e-06,
      "loss": 0.4019,
      "step": 181
    },
    {
      "epoch": 0.34849210148396365,
      "grad_norm": 0.8370754718780518,
      "learning_rate": 7.777851165098012e-06,
      "loss": 0.4202,
      "step": 182
    },
    {
      "epoch": 0.350406893250359,
      "grad_norm": 0.7426475882530212,
      "learning_rate": 7.751463576186957e-06,
      "loss": 0.378,
      "step": 183
    },
    {
      "epoch": 0.3523216850167544,
      "grad_norm": 0.827038586139679,
      "learning_rate": 7.72496560516159e-06,
      "loss": 0.415,
      "step": 184
    },
    {
      "epoch": 0.35423647678314985,
      "grad_norm": 0.8714759349822998,
      "learning_rate": 7.6983583150569e-06,
      "loss": 0.4204,
      "step": 185
    },
    {
      "epoch": 0.35615126854954526,
      "grad_norm": 0.8127462863922119,
      "learning_rate": 7.671642773293506e-06,
      "loss": 0.3904,
      "step": 186
    },
    {
      "epoch": 0.3580660603159406,
      "grad_norm": 0.8972522020339966,
      "learning_rate": 7.644820051634813e-06,
      "loss": 0.4168,
      "step": 187
    },
    {
      "epoch": 0.35998085208233604,
      "grad_norm": 0.9051675200462341,
      "learning_rate": 7.617891226144034e-06,
      "loss": 0.4742,
      "step": 188
    },
    {
      "epoch": 0.36189564384873146,
      "grad_norm": 0.8041402101516724,
      "learning_rate": 7.59085737714101e-06,
      "loss": 0.3916,
      "step": 189
    },
    {
      "epoch": 0.3638104356151269,
      "grad_norm": 0.9296969175338745,
      "learning_rate": 7.563719589158874e-06,
      "loss": 0.4198,
      "step": 190
    },
    {
      "epoch": 0.36572522738152224,
      "grad_norm": 0.8441433310508728,
      "learning_rate": 7.536478950900537e-06,
      "loss": 0.4094,
      "step": 191
    },
    {
      "epoch": 0.36764001914791766,
      "grad_norm": 0.8146634101867676,
      "learning_rate": 7.509136555195025e-06,
      "loss": 0.398,
      "step": 192
    },
    {
      "epoch": 0.3695548109143131,
      "grad_norm": 0.8095076680183411,
      "learning_rate": 7.481693498953621e-06,
      "loss": 0.4121,
      "step": 193
    },
    {
      "epoch": 0.3714696026807085,
      "grad_norm": 0.8033435344696045,
      "learning_rate": 7.4541508831258695e-06,
      "loss": 0.3912,
      "step": 194
    },
    {
      "epoch": 0.37338439444710386,
      "grad_norm": 0.7945087552070618,
      "learning_rate": 7.4265098126554065e-06,
      "loss": 0.3784,
      "step": 195
    },
    {
      "epoch": 0.3752991862134993,
      "grad_norm": 0.858241081237793,
      "learning_rate": 7.3987713964356335e-06,
      "loss": 0.451,
      "step": 196
    },
    {
      "epoch": 0.3772139779798947,
      "grad_norm": 0.9208387136459351,
      "learning_rate": 7.370936747265226e-06,
      "loss": 0.4539,
      "step": 197
    },
    {
      "epoch": 0.3791287697462901,
      "grad_norm": 0.775140643119812,
      "learning_rate": 7.3430069818035e-06,
      "loss": 0.3956,
      "step": 198
    },
    {
      "epoch": 0.3810435615126855,
      "grad_norm": 0.7926008105278015,
      "learning_rate": 7.314983220525604e-06,
      "loss": 0.4044,
      "step": 199
    },
    {
      "epoch": 0.3829583532790809,
      "grad_norm": 0.7891693711280823,
      "learning_rate": 7.286866587677576e-06,
      "loss": 0.3881,
      "step": 200
    },
    {
      "epoch": 0.3848731450454763,
      "grad_norm": 0.8547941446304321,
      "learning_rate": 7.2586582112312355e-06,
      "loss": 0.4289,
      "step": 201
    },
    {
      "epoch": 0.38678793681187174,
      "grad_norm": 0.7894405722618103,
      "learning_rate": 7.230359222838939e-06,
      "loss": 0.3886,
      "step": 202
    },
    {
      "epoch": 0.3887027285782671,
      "grad_norm": 0.9024775624275208,
      "learning_rate": 7.201970757788172e-06,
      "loss": 0.4586,
      "step": 203
    },
    {
      "epoch": 0.3906175203446625,
      "grad_norm": 0.7940675616264343,
      "learning_rate": 7.173493954956012e-06,
      "loss": 0.3905,
      "step": 204
    },
    {
      "epoch": 0.39253231211105793,
      "grad_norm": 0.8231476545333862,
      "learning_rate": 7.144929956763438e-06,
      "loss": 0.4044,
      "step": 205
    },
    {
      "epoch": 0.39444710387745335,
      "grad_norm": 0.9094031453132629,
      "learning_rate": 7.116279909129492e-06,
      "loss": 0.4502,
      "step": 206
    },
    {
      "epoch": 0.3963618956438487,
      "grad_norm": 0.843540608882904,
      "learning_rate": 7.087544961425317e-06,
      "loss": 0.4037,
      "step": 207
    },
    {
      "epoch": 0.39827668741024413,
      "grad_norm": 0.8074728846549988,
      "learning_rate": 7.058726266428042e-06,
      "loss": 0.405,
      "step": 208
    },
    {
      "epoch": 0.40019147917663955,
      "grad_norm": 0.7620254755020142,
      "learning_rate": 7.029824980274536e-06,
      "loss": 0.3727,
      "step": 209
    },
    {
      "epoch": 0.40210627094303497,
      "grad_norm": 0.8311992883682251,
      "learning_rate": 7.0008422624150285e-06,
      "loss": 0.4172,
      "step": 210
    }
  ],
  "logging_steps": 1,
  "max_steps": 522,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 105,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1029906388628275e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}