{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 234,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008547008547008548,
      "grad_norm": 0.6883277297019958,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.0565,
      "step": 1
    },
    {
      "epoch": 0.017094017094017096,
      "grad_norm": 0.5167361497879028,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.8421,
      "step": 2
    },
    {
      "epoch": 0.02564102564102564,
      "grad_norm": 0.8402963876724243,
      "learning_rate": 3e-06,
      "loss": 1.1245,
      "step": 3
    },
    {
      "epoch": 0.03418803418803419,
      "grad_norm": 0.930655300617218,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.432,
      "step": 4
    },
    {
      "epoch": 0.042735042735042736,
      "grad_norm": 0.5283745527267456,
      "learning_rate": 5e-06,
      "loss": 0.941,
      "step": 5
    },
    {
      "epoch": 0.05128205128205128,
      "grad_norm": 1.1349669694900513,
      "learning_rate": 6e-06,
      "loss": 1.3429,
      "step": 6
    },
    {
      "epoch": 0.05982905982905983,
      "grad_norm": 1.173917293548584,
      "learning_rate": 7e-06,
      "loss": 0.9637,
      "step": 7
    },
    {
      "epoch": 0.06837606837606838,
      "grad_norm": 0.6507728099822998,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.0163,
      "step": 8
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 0.6534399390220642,
      "learning_rate": 9e-06,
      "loss": 0.9108,
      "step": 9
    },
    {
      "epoch": 0.08547008547008547,
      "grad_norm": 0.8090460300445557,
      "learning_rate": 1e-05,
      "loss": 1.1224,
      "step": 10
    },
    {
      "epoch": 0.09401709401709402,
      "grad_norm": 1.183127760887146,
      "learning_rate": 9.999882372979835e-06,
      "loss": 1.1556,
      "step": 11
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 1.1587895154953003,
      "learning_rate": 9.999529497453782e-06,
      "loss": 0.9223,
      "step": 12
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.7878014445304871,
      "learning_rate": 9.998941390024924e-06,
      "loss": 1.0363,
      "step": 13
    },
    {
      "epoch": 0.11965811965811966,
      "grad_norm": 1.0422732830047607,
      "learning_rate": 9.998118078364186e-06,
      "loss": 1.1158,
      "step": 14
    },
    {
      "epoch": 0.1282051282051282,
      "grad_norm": 0.8618931174278259,
      "learning_rate": 9.99705960120905e-06,
      "loss": 1.1986,
      "step": 15
    },
    {
      "epoch": 0.13675213675213677,
      "grad_norm": 0.7314261198043823,
      "learning_rate": 9.99576600836172e-06,
      "loss": 0.9607,
      "step": 16
    },
    {
      "epoch": 0.1452991452991453,
      "grad_norm": 0.8001905679702759,
      "learning_rate": 9.994237360686784e-06,
      "loss": 1.3201,
      "step": 17
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 0.6340293288230896,
      "learning_rate": 9.992473730108354e-06,
      "loss": 0.8039,
      "step": 18
    },
    {
      "epoch": 0.1623931623931624,
      "grad_norm": 1.0305331945419312,
      "learning_rate": 9.990475199606672e-06,
      "loss": 0.987,
      "step": 19
    },
    {
      "epoch": 0.17094017094017094,
      "grad_norm": 0.5756571292877197,
      "learning_rate": 9.988241863214212e-06,
      "loss": 0.856,
      "step": 20
    },
    {
      "epoch": 0.1794871794871795,
      "grad_norm": 0.7210500836372375,
      "learning_rate": 9.985773826011256e-06,
      "loss": 0.9009,
      "step": 21
    },
    {
      "epoch": 0.18803418803418803,
      "grad_norm": 0.6321185231208801,
      "learning_rate": 9.98307120412095e-06,
      "loss": 0.9718,
      "step": 22
    },
    {
      "epoch": 0.19658119658119658,
      "grad_norm": 0.7177990078926086,
      "learning_rate": 9.980134124703837e-06,
      "loss": 1.0357,
      "step": 23
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 0.695940375328064,
      "learning_rate": 9.976962725951878e-06,
      "loss": 1.0613,
      "step": 24
    },
    {
      "epoch": 0.21367521367521367,
      "grad_norm": 0.7316240072250366,
      "learning_rate": 9.973557157081946e-06,
      "loss": 1.5432,
      "step": 25
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.1439138650894165,
      "learning_rate": 9.969917578328808e-06,
      "loss": 1.3765,
      "step": 26
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 0.7460082173347473,
      "learning_rate": 9.966044160937588e-06,
      "loss": 0.8814,
      "step": 27
    },
    {
      "epoch": 0.23931623931623933,
      "grad_norm": 0.6455249190330505,
      "learning_rate": 9.961937087155697e-06,
      "loss": 0.889,
      "step": 28
    },
    {
      "epoch": 0.24786324786324787,
      "grad_norm": 0.7218654155731201,
      "learning_rate": 9.957596550224285e-06,
      "loss": 1.1877,
      "step": 29
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 0.7643616795539856,
      "learning_rate": 9.953022754369115e-06,
      "loss": 1.133,
      "step": 30
    },
    {
      "epoch": 0.26495726495726496,
      "grad_norm": 0.4708094298839569,
      "learning_rate": 9.94821591479098e-06,
      "loss": 0.8308,
      "step": 31
    },
    {
      "epoch": 0.27350427350427353,
      "grad_norm": 0.5885545611381531,
      "learning_rate": 9.943176257655567e-06,
      "loss": 0.8915,
      "step": 32
    },
    {
      "epoch": 0.28205128205128205,
      "grad_norm": 0.7514286637306213,
      "learning_rate": 9.937904020082815e-06,
      "loss": 1.063,
      "step": 33
    },
    {
      "epoch": 0.2905982905982906,
      "grad_norm": 0.541725754737854,
      "learning_rate": 9.932399450135765e-06,
      "loss": 0.9508,
      "step": 34
    },
    {
      "epoch": 0.29914529914529914,
      "grad_norm": 0.5545334815979004,
      "learning_rate": 9.92666280680888e-06,
      "loss": 0.8066,
      "step": 35
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.47642382979393005,
      "learning_rate": 9.920694360015864e-06,
      "loss": 0.755,
      "step": 36
    },
    {
      "epoch": 0.3162393162393162,
      "grad_norm": 0.5091294050216675,
      "learning_rate": 9.914494390576958e-06,
      "loss": 0.7879,
      "step": 37
    },
    {
      "epoch": 0.3247863247863248,
      "grad_norm": 0.46325746178627014,
      "learning_rate": 9.908063190205739e-06,
      "loss": 0.83,
      "step": 38
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5515424609184265,
      "learning_rate": 9.901401061495379e-06,
      "loss": 0.8476,
      "step": 39
    },
    {
      "epoch": 0.3418803418803419,
      "grad_norm": 0.5145699977874756,
      "learning_rate": 9.894508317904418e-06,
      "loss": 0.9449,
      "step": 40
    },
    {
      "epoch": 0.3504273504273504,
      "grad_norm": 0.46632590889930725,
      "learning_rate": 9.88738528374201e-06,
      "loss": 0.9886,
      "step": 41
    },
    {
      "epoch": 0.358974358974359,
      "grad_norm": 0.41940009593963623,
      "learning_rate": 9.880032294152673e-06,
      "loss": 0.7544,
      "step": 42
    },
    {
      "epoch": 0.36752136752136755,
      "grad_norm": 0.3798862099647522,
      "learning_rate": 9.872449695100503e-06,
      "loss": 0.77,
      "step": 43
    },
    {
      "epoch": 0.37606837606837606,
      "grad_norm": 0.571315348148346,
      "learning_rate": 9.864637843352916e-06,
      "loss": 1.1002,
      "step": 44
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.44546273350715637,
      "learning_rate": 9.856597106463847e-06,
      "loss": 0.8818,
      "step": 45
    },
    {
      "epoch": 0.39316239316239315,
      "grad_norm": 0.6359449028968811,
      "learning_rate": 9.848327862756466e-06,
      "loss": 0.8167,
      "step": 46
    },
    {
      "epoch": 0.4017094017094017,
      "grad_norm": 0.5933560729026794,
      "learning_rate": 9.839830501305371e-06,
      "loss": 0.873,
      "step": 47
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 0.36119118332862854,
      "learning_rate": 9.831105421918287e-06,
      "loss": 0.776,
      "step": 48
    },
    {
      "epoch": 0.4188034188034188,
      "grad_norm": 0.4318462312221527,
      "learning_rate": 9.822153035117246e-06,
      "loss": 0.7745,
      "step": 49
    },
    {
      "epoch": 0.42735042735042733,
      "grad_norm": 0.5515265464782715,
      "learning_rate": 9.812973762119282e-06,
      "loss": 1.0686,
      "step": 50
    },
    {
      "epoch": 0.4358974358974359,
      "grad_norm": 0.3906237781047821,
      "learning_rate": 9.803568034816606e-06,
      "loss": 0.7159,
      "step": 51
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.3262993395328522,
      "learning_rate": 9.793936295756292e-06,
      "loss": 0.7008,
      "step": 52
    },
    {
      "epoch": 0.452991452991453,
      "grad_norm": 0.43187564611434937,
      "learning_rate": 9.784078998119442e-06,
      "loss": 0.7201,
      "step": 53
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.3680849075317383,
      "learning_rate": 9.773996605699876e-06,
      "loss": 1.0274,
      "step": 54
    },
    {
      "epoch": 0.4700854700854701,
      "grad_norm": 0.32845616340637207,
      "learning_rate": 9.763689592882307e-06,
      "loss": 0.6843,
      "step": 55
    },
    {
      "epoch": 0.47863247863247865,
      "grad_norm": 0.5680167078971863,
      "learning_rate": 9.753158444620013e-06,
      "loss": 1.1483,
      "step": 56
    },
    {
      "epoch": 0.48717948717948717,
      "grad_norm": 0.4027453660964966,
      "learning_rate": 9.742403656412033e-06,
      "loss": 0.6624,
      "step": 57
    },
    {
      "epoch": 0.49572649572649574,
      "grad_norm": 0.42083829641342163,
      "learning_rate": 9.73142573427984e-06,
      "loss": 0.8074,
      "step": 58
    },
    {
      "epoch": 0.5042735042735043,
      "grad_norm": 0.43723517656326294,
      "learning_rate": 9.720225194743544e-06,
      "loss": 0.7623,
      "step": 59
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.37138086557388306,
      "learning_rate": 9.70880256479758e-06,
      "loss": 0.7541,
      "step": 60
    },
    {
      "epoch": 0.5213675213675214,
      "grad_norm": 0.38942328095436096,
      "learning_rate": 9.697158381885915e-06,
      "loss": 0.7369,
      "step": 61
    },
    {
      "epoch": 0.5299145299145299,
      "grad_norm": 0.35463273525238037,
      "learning_rate": 9.685293193876766e-06,
      "loss": 0.6687,
      "step": 62
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 0.443660706281662,
      "learning_rate": 9.673207559036817e-06,
      "loss": 1.3078,
      "step": 63
    },
    {
      "epoch": 0.5470085470085471,
      "grad_norm": 0.42827773094177246,
      "learning_rate": 9.660902046004954e-06,
      "loss": 0.7356,
      "step": 64
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.6239178776741028,
      "learning_rate": 9.648377233765507e-06,
      "loss": 0.6916,
      "step": 65
    },
    {
      "epoch": 0.5641025641025641,
      "grad_norm": 0.40673717856407166,
      "learning_rate": 9.635633711621014e-06,
      "loss": 0.728,
      "step": 66
    },
    {
      "epoch": 0.5726495726495726,
      "grad_norm": 0.4105391800403595,
      "learning_rate": 9.622672079164487e-06,
      "loss": 0.811,
      "step": 67
    },
    {
      "epoch": 0.5811965811965812,
      "grad_norm": 0.37009334564208984,
      "learning_rate": 9.60949294625121e-06,
      "loss": 0.6723,
      "step": 68
    },
    {
      "epoch": 0.5897435897435898,
      "grad_norm": 0.37860628962516785,
      "learning_rate": 9.596096932970035e-06,
      "loss": 0.7644,
      "step": 69
    },
    {
      "epoch": 0.5982905982905983,
      "grad_norm": 0.36861270666122437,
      "learning_rate": 9.582484669614212e-06,
      "loss": 0.7353,
      "step": 70
    },
    {
      "epoch": 0.6068376068376068,
      "grad_norm": 0.3790634274482727,
      "learning_rate": 9.568656796651733e-06,
      "loss": 0.8376,
      "step": 71
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.5848673582077026,
      "learning_rate": 9.554613964695189e-06,
      "loss": 1.3309,
      "step": 72
    },
    {
      "epoch": 0.6239316239316239,
      "grad_norm": 0.3627384305000305,
      "learning_rate": 9.540356834471178e-06,
      "loss": 0.6774,
      "step": 73
    },
    {
      "epoch": 0.6324786324786325,
      "grad_norm": 0.37787535786628723,
      "learning_rate": 9.525886076789195e-06,
      "loss": 0.703,
      "step": 74
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 0.32273605465888977,
      "learning_rate": 9.511202372510083e-06,
      "loss": 0.7019,
      "step": 75
    },
    {
      "epoch": 0.6495726495726496,
      "grad_norm": 0.30288276076316833,
      "learning_rate": 9.496306412513989e-06,
      "loss": 0.7098,
      "step": 76
    },
    {
      "epoch": 0.6581196581196581,
      "grad_norm": 0.47629785537719727,
      "learning_rate": 9.481198897667875e-06,
      "loss": 0.8417,
      "step": 77
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.29766926169395447,
      "learning_rate": 9.465880538792519e-06,
      "loss": 0.6709,
      "step": 78
    },
    {
      "epoch": 0.6752136752136753,
      "grad_norm": 0.33883240818977356,
      "learning_rate": 9.450352056629083e-06,
      "loss": 0.6712,
      "step": 79
    },
    {
      "epoch": 0.6837606837606838,
      "grad_norm": 0.3106386959552765,
      "learning_rate": 9.434614181805203e-06,
      "loss": 0.6535,
      "step": 80
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.38624322414398193,
      "learning_rate": 9.418667654800607e-06,
      "loss": 0.7493,
      "step": 81
    },
    {
      "epoch": 0.7008547008547008,
      "grad_norm": 0.46079033613204956,
      "learning_rate": 9.402513225912273e-06,
      "loss": 1.1914,
      "step": 82
    },
    {
      "epoch": 0.7094017094017094,
      "grad_norm": 0.4166659414768219,
      "learning_rate": 9.386151655219137e-06,
      "loss": 1.1341,
      "step": 83
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 0.3459385931491852,
      "learning_rate": 9.369583712546322e-06,
      "loss": 1.1233,
      "step": 84
    },
    {
      "epoch": 0.7264957264957265,
      "grad_norm": 0.303739994764328,
      "learning_rate": 9.352810177428917e-06,
      "loss": 0.6361,
      "step": 85
    },
    {
      "epoch": 0.7350427350427351,
      "grad_norm": 0.31175675988197327,
      "learning_rate": 9.335831839075303e-06,
      "loss": 0.5938,
      "step": 86
    },
    {
      "epoch": 0.7435897435897436,
      "grad_norm": 0.3335458040237427,
      "learning_rate": 9.318649496330021e-06,
      "loss": 0.673,
      "step": 87
    },
    {
      "epoch": 0.7521367521367521,
      "grad_norm": 0.5561854839324951,
      "learning_rate": 9.30126395763618e-06,
      "loss": 1.0438,
      "step": 88
    },
    {
      "epoch": 0.7606837606837606,
      "grad_norm": 0.39674779772758484,
      "learning_rate": 9.283676040997426e-06,
      "loss": 1.2274,
      "step": 89
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.46839889883995056,
      "learning_rate": 9.265886573939448e-06,
      "loss": 1.0736,
      "step": 90
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.329444020986557,
      "learning_rate": 9.247896393471045e-06,
      "loss": 0.6996,
      "step": 91
    },
    {
      "epoch": 0.7863247863247863,
      "grad_norm": 0.37539413571357727,
      "learning_rate": 9.229706346044749e-06,
      "loss": 0.6772,
      "step": 92
    },
    {
      "epoch": 0.7948717948717948,
      "grad_norm": 0.3232697546482086,
      "learning_rate": 9.211317287516985e-06,
      "loss": 0.6433,
      "step": 93
    },
    {
      "epoch": 0.8034188034188035,
      "grad_norm": 0.4283379912376404,
      "learning_rate": 9.19273008310782e-06,
      "loss": 0.808,
      "step": 94
    },
    {
      "epoch": 0.811965811965812,
      "grad_norm": 0.40039879083633423,
      "learning_rate": 9.173945607360238e-06,
      "loss": 0.6781,
      "step": 95
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 0.421421617269516,
      "learning_rate": 9.154964744099006e-06,
      "loss": 1.1649,
      "step": 96
    },
    {
      "epoch": 0.8290598290598291,
      "grad_norm": 0.37563416361808777,
      "learning_rate": 9.135788386389077e-06,
      "loss": 0.6748,
      "step": 97
    },
    {
      "epoch": 0.8376068376068376,
      "grad_norm": 0.34847089648246765,
      "learning_rate": 9.116417436493574e-06,
      "loss": 1.2002,
      "step": 98
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 0.38143283128738403,
      "learning_rate": 9.096852805831348e-06,
      "loss": 0.8034,
      "step": 99
    },
    {
      "epoch": 0.8547008547008547,
      "grad_norm": 0.43068060278892517,
      "learning_rate": 9.077095414934076e-06,
      "loss": 0.7409,
      "step": 100
    },
    {
      "epoch": 0.8632478632478633,
      "grad_norm": 0.4279479384422302,
      "learning_rate": 9.057146193402968e-06,
      "loss": 1.0627,
      "step": 101
    },
    {
      "epoch": 0.8717948717948718,
      "grad_norm": 0.4032224416732788,
      "learning_rate": 9.037006079865017e-06,
      "loss": 1.1393,
      "step": 102
    },
    {
      "epoch": 0.8803418803418803,
      "grad_norm": 0.36322587728500366,
      "learning_rate": 9.016676021928838e-06,
      "loss": 0.9575,
      "step": 103
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.42848172783851624,
      "learning_rate": 8.996156976140088e-06,
      "loss": 1.1044,
      "step": 104
    },
    {
      "epoch": 0.8974358974358975,
      "grad_norm": 0.38128426671028137,
      "learning_rate": 8.975449907936447e-06,
      "loss": 1.2012,
      "step": 105
    },
    {
      "epoch": 0.905982905982906,
      "grad_norm": 0.8348135948181152,
      "learning_rate": 8.95455579160221e-06,
      "loss": 1.1161,
      "step": 106
    },
    {
      "epoch": 0.9145299145299145,
      "grad_norm": 0.599600613117218,
      "learning_rate": 8.933475610222435e-06,
      "loss": 0.8809,
      "step": 107
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.34604817628860474,
      "learning_rate": 8.91221035563669e-06,
      "loss": 1.1079,
      "step": 108
    },
    {
      "epoch": 0.9316239316239316,
      "grad_norm": 0.6436942219734192,
      "learning_rate": 8.890761028392385e-06,
      "loss": 1.136,
      "step": 109
    },
    {
      "epoch": 0.9401709401709402,
      "grad_norm": 0.44971659779548645,
      "learning_rate": 8.869128637697702e-06,
      "loss": 0.8062,
      "step": 110
    },
    {
      "epoch": 0.9487179487179487,
      "grad_norm": 0.3893284201622009,
      "learning_rate": 8.847314201374102e-06,
      "loss": 0.7011,
      "step": 111
    },
    {
      "epoch": 0.9572649572649573,
      "grad_norm": 0.39437901973724365,
      "learning_rate": 8.82531874580844e-06,
      "loss": 0.6845,
      "step": 112
    },
    {
      "epoch": 0.9658119658119658,
      "grad_norm": 0.39099910855293274,
      "learning_rate": 8.803143305904676e-06,
      "loss": 0.6957,
      "step": 113
    },
    {
      "epoch": 0.9743589743589743,
      "grad_norm": 0.3814919590950012,
      "learning_rate": 8.780788925035178e-06,
      "loss": 0.8374,
      "step": 114
    },
    {
      "epoch": 0.9829059829059829,
      "grad_norm": 0.31528154015541077,
      "learning_rate": 8.758256654991627e-06,
      "loss": 0.601,
      "step": 115
    },
    {
      "epoch": 0.9914529914529915,
      "grad_norm": 0.45662426948547363,
      "learning_rate": 8.735547555935538e-06,
      "loss": 0.7883,
      "step": 116
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3865978419780731,
      "learning_rate": 8.712662696348371e-06,
      "loss": 0.6754,
      "step": 117
    },
    {
      "epoch": 1.0085470085470085,
      "grad_norm": 0.337187260389328,
      "learning_rate": 8.689603152981262e-06,
      "loss": 0.6326,
      "step": 118
    },
    {
      "epoch": 1.017094017094017,
      "grad_norm": 0.38046014308929443,
      "learning_rate": 8.666370010804361e-06,
      "loss": 0.6708,
      "step": 119
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 0.42673853039741516,
      "learning_rate": 8.642964362955781e-06,
      "loss": 0.6928,
      "step": 120
    },
    {
      "epoch": 1.0341880341880343,
      "grad_norm": 0.45885011553764343,
      "learning_rate": 8.619387310690167e-06,
      "loss": 0.6886,
      "step": 121
    },
    {
      "epoch": 1.0427350427350428,
      "grad_norm": 0.4303334057331085,
      "learning_rate": 8.59563996332688e-06,
      "loss": 1.3497,
      "step": 122
    },
    {
      "epoch": 1.0512820512820513,
      "grad_norm": 0.5063712000846863,
      "learning_rate": 8.5717234381978e-06,
      "loss": 1.1424,
      "step": 123
    },
    {
      "epoch": 1.0598290598290598,
      "grad_norm": 0.43861711025238037,
      "learning_rate": 8.547638860594765e-06,
      "loss": 1.1289,
      "step": 124
    },
    {
      "epoch": 1.0683760683760684,
      "grad_norm": 0.43634119629859924,
      "learning_rate": 8.523387363716611e-06,
      "loss": 0.7524,
      "step": 125
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 0.3733837604522705,
      "learning_rate": 8.498970088615861e-06,
      "loss": 0.6589,
      "step": 126
    },
    {
      "epoch": 1.0854700854700854,
      "grad_norm": 0.32617077231407166,
      "learning_rate": 8.474388184145043e-06,
      "loss": 1.2309,
      "step": 127
    },
    {
      "epoch": 1.0940170940170941,
      "grad_norm": 0.35106804966926575,
      "learning_rate": 8.449642806902623e-06,
      "loss": 0.6126,
      "step": 128
    },
    {
      "epoch": 1.1025641025641026,
      "grad_norm": 0.4258238971233368,
      "learning_rate": 8.424735121178598e-06,
      "loss": 0.6661,
      "step": 129
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.4120415151119232,
      "learning_rate": 8.399666298899706e-06,
      "loss": 0.7212,
      "step": 130
    },
    {
      "epoch": 1.1196581196581197,
      "grad_norm": 0.4039503037929535,
      "learning_rate": 8.374437519574296e-06,
      "loss": 1.0448,
      "step": 131
    },
    {
      "epoch": 1.1282051282051282,
      "grad_norm": 0.33159151673316956,
      "learning_rate": 8.349049970236822e-06,
      "loss": 1.1204,
      "step": 132
    },
    {
      "epoch": 1.1367521367521367,
      "grad_norm": 0.4609539210796356,
      "learning_rate": 8.32350484539199e-06,
      "loss": 0.7522,
      "step": 133
    },
    {
      "epoch": 1.1452991452991452,
      "grad_norm": 0.34498193860054016,
      "learning_rate": 8.29780334695857e-06,
      "loss": 1.0665,
      "step": 134
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 0.3104630410671234,
      "learning_rate": 8.271946684212832e-06,
      "loss": 0.5928,
      "step": 135
    },
    {
      "epoch": 1.1623931623931625,
      "grad_norm": 0.4486801326274872,
      "learning_rate": 8.245936073731654e-06,
      "loss": 0.6778,
      "step": 136
    },
    {
      "epoch": 1.170940170940171,
      "grad_norm": 0.41299891471862793,
      "learning_rate": 8.219772739335272e-06,
      "loss": 1.6928,
      "step": 137
    },
    {
      "epoch": 1.1794871794871795,
      "grad_norm": 0.41245394945144653,
      "learning_rate": 8.193457912029713e-06,
      "loss": 0.6847,
      "step": 138
    },
    {
      "epoch": 1.188034188034188,
      "grad_norm": 0.3258431553840637,
      "learning_rate": 8.166992829948868e-06,
      "loss": 0.5718,
      "step": 139
    },
    {
      "epoch": 1.1965811965811965,
      "grad_norm": 0.5331162214279175,
      "learning_rate": 8.140378738296233e-06,
      "loss": 0.7571,
      "step": 140
    },
    {
      "epoch": 1.205128205128205,
      "grad_norm": 0.36795511841773987,
      "learning_rate": 8.113616889286325e-06,
      "loss": 0.6596,
      "step": 141
    },
    {
      "epoch": 1.2136752136752136,
      "grad_norm": 0.35999539494514465,
      "learning_rate": 8.086708542085769e-06,
      "loss": 1.1737,
      "step": 142
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 0.4550798237323761,
      "learning_rate": 8.05965496275404e-06,
      "loss": 0.6575,
      "step": 143
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.43874284625053406,
      "learning_rate": 8.032457424183909e-06,
      "loss": 0.7127,
      "step": 144
    },
    {
      "epoch": 1.2393162393162394,
      "grad_norm": 0.39959272742271423,
      "learning_rate": 8.005117206041544e-06,
      "loss": 0.7145,
      "step": 145
    },
    {
      "epoch": 1.2478632478632479,
      "grad_norm": 0.4271208941936493,
      "learning_rate": 7.977635594706298e-06,
      "loss": 0.7244,
      "step": 146
    },
    {
      "epoch": 1.2564102564102564,
      "grad_norm": 0.4204410910606384,
      "learning_rate": 7.950013883210198e-06,
      "loss": 0.6295,
      "step": 147
    },
    {
      "epoch": 1.264957264957265,
      "grad_norm": 0.40335509181022644,
      "learning_rate": 7.922253371177081e-06,
      "loss": 0.7212,
      "step": 148
    },
    {
      "epoch": 1.2735042735042734,
      "grad_norm": 0.3271823227405548,
      "learning_rate": 7.894355364761476e-06,
      "loss": 0.9603,
      "step": 149
    },
    {
      "epoch": 1.282051282051282,
      "grad_norm": 0.4051213562488556,
      "learning_rate": 7.866321176587129e-06,
      "loss": 1.1063,
      "step": 150
    },
    {
      "epoch": 1.2905982905982907,
      "grad_norm": 0.3575092852115631,
      "learning_rate": 7.838152125685245e-06,
      "loss": 0.5582,
      "step": 151
    },
    {
      "epoch": 1.2991452991452992,
      "grad_norm": 0.39023974537849426,
      "learning_rate": 7.809849537432432e-06,
      "loss": 0.6651,
      "step": 152
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.45742174983024597,
      "learning_rate": 7.781414743488338e-06,
      "loss": 0.7104,
      "step": 153
    },
    {
      "epoch": 1.3162393162393162,
      "grad_norm": 0.3916301727294922,
      "learning_rate": 7.752849081732993e-06,
      "loss": 0.7525,
      "step": 154
    },
    {
      "epoch": 1.3247863247863247,
      "grad_norm": 0.41341787576675415,
      "learning_rate": 7.724153896203868e-06,
      "loss": 0.6589,
      "step": 155
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.36725375056266785,
      "learning_rate": 7.695330537032629e-06,
      "loss": 0.6316,
      "step": 156
    },
    {
      "epoch": 1.341880341880342,
      "grad_norm": 0.39961159229278564,
      "learning_rate": 7.666380360381616e-06,
      "loss": 1.2004,
      "step": 157
    },
    {
      "epoch": 1.3504273504273505,
      "grad_norm": 0.5076507925987244,
      "learning_rate": 7.637304728380036e-06,
      "loss": 0.7745,
      "step": 158
    },
    {
      "epoch": 1.358974358974359,
      "grad_norm": 0.47983452677726746,
      "learning_rate": 7.608105009059867e-06,
      "loss": 0.8066,
      "step": 159
    },
    {
      "epoch": 1.3675213675213675,
      "grad_norm": 0.4021775722503662,
      "learning_rate": 7.578782576291501e-06,
      "loss": 0.5962,
      "step": 160
    },
    {
      "epoch": 1.376068376068376,
      "grad_norm": 0.5335017442703247,
      "learning_rate": 7.5493388097190915e-06,
      "loss": 0.8891,
      "step": 161
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 0.42396119236946106,
      "learning_rate": 7.51977509469565e-06,
      "loss": 0.6718,
      "step": 162
    },
    {
      "epoch": 1.393162393162393,
      "grad_norm": 0.4155985713005066,
      "learning_rate": 7.490092822217856e-06,
      "loss": 0.9571,
      "step": 163
    },
    {
      "epoch": 1.4017094017094016,
      "grad_norm": 0.5259201526641846,
      "learning_rate": 7.460293388860616e-06,
      "loss": 0.8977,
      "step": 164
    },
    {
      "epoch": 1.4102564102564101,
      "grad_norm": 0.4060882031917572,
      "learning_rate": 7.4303781967113494e-06,
      "loss": 0.7018,
      "step": 165
    },
    {
      "epoch": 1.4188034188034189,
      "grad_norm": 0.32119300961494446,
      "learning_rate": 7.400348653304022e-06,
      "loss": 0.5588,
      "step": 166
    },
    {
      "epoch": 1.4273504273504274,
      "grad_norm": 0.42005738615989685,
      "learning_rate": 7.370206171552914e-06,
      "loss": 0.9614,
      "step": 167
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 0.3684864938259125,
      "learning_rate": 7.3399521696861505e-06,
      "loss": 0.6402,
      "step": 168
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.4385385811328888,
      "learning_rate": 7.309588071178968e-06,
      "loss": 0.6126,
      "step": 169
    },
    {
      "epoch": 1.452991452991453,
      "grad_norm": 0.414637953042984,
      "learning_rate": 7.2791153046867344e-06,
      "loss": 0.7569,
      "step": 170
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.3897780179977417,
      "learning_rate": 7.248535303977739e-06,
      "loss": 0.6537,
      "step": 171
    },
    {
      "epoch": 1.4700854700854702,
      "grad_norm": 0.37271955609321594,
      "learning_rate": 7.217849507865724e-06,
      "loss": 0.6451,
      "step": 172
    },
    {
      "epoch": 1.4786324786324787,
      "grad_norm": 0.4022608697414398,
      "learning_rate": 7.187059360142194e-06,
      "loss": 0.7482,
      "step": 173
    },
    {
      "epoch": 1.4871794871794872,
      "grad_norm": 0.3631649315357208,
      "learning_rate": 7.156166309508482e-06,
      "loss": 0.9156,
      "step": 174
    },
    {
      "epoch": 1.4957264957264957,
      "grad_norm": 0.4740133583545685,
      "learning_rate": 7.125171809507581e-06,
      "loss": 0.6974,
      "step": 175
    },
    {
      "epoch": 1.5042735042735043,
      "grad_norm": 0.49716681241989136,
      "learning_rate": 7.094077318455762e-06,
      "loss": 1.2114,
      "step": 176
    },
    {
      "epoch": 1.5128205128205128,
      "grad_norm": 0.449844628572464,
      "learning_rate": 7.062884299373955e-06,
      "loss": 0.6517,
      "step": 177
    },
    {
      "epoch": 1.5213675213675213,
      "grad_norm": 0.38638660311698914,
      "learning_rate": 7.031594219918916e-06,
      "loss": 0.6244,
      "step": 178
    },
    {
      "epoch": 1.5299145299145298,
      "grad_norm": 0.44147396087646484,
      "learning_rate": 7.000208552314166e-06,
      "loss": 0.5929,
      "step": 179
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.36652877926826477,
      "learning_rate": 6.96872877328073e-06,
      "loss": 0.6289,
      "step": 180
    },
    {
      "epoch": 1.547008547008547,
      "grad_norm": 0.353456974029541,
      "learning_rate": 6.937156363967647e-06,
      "loss": 0.5993,
      "step": 181
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.4075149893760681,
      "learning_rate": 6.905492809882286e-06,
      "loss": 1.1381,
      "step": 182
    },
    {
      "epoch": 1.564102564102564,
      "grad_norm": 0.375399649143219,
      "learning_rate": 6.873739600820457e-06,
      "loss": 0.5813,
      "step": 183
    },
    {
      "epoch": 1.5726495726495726,
      "grad_norm": 0.5181817412376404,
      "learning_rate": 6.841898230796302e-06,
      "loss": 0.7546,
      "step": 184
    },
    {
      "epoch": 1.5811965811965814,
      "grad_norm": 0.40129345655441284,
      "learning_rate": 6.809970197972014e-06,
      "loss": 0.666,
      "step": 185
    },
    {
      "epoch": 1.5897435897435899,
      "grad_norm": 0.44013726711273193,
      "learning_rate": 6.777957004587332e-06,
      "loss": 0.6635,
      "step": 186
    },
    {
      "epoch": 1.5982905982905984,
      "grad_norm": 0.9672113060951233,
      "learning_rate": 6.745860156888878e-06,
      "loss": 1.0801,
      "step": 187
    },
    {
      "epoch": 1.606837606837607,
      "grad_norm": 0.6321570873260498,
      "learning_rate": 6.713681165059271e-06,
      "loss": 0.6552,
      "step": 188
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.42415156960487366,
      "learning_rate": 6.68142154314608e-06,
      "loss": 0.7572,
      "step": 189
    },
    {
      "epoch": 1.623931623931624,
      "grad_norm": 0.41570088267326355,
      "learning_rate": 6.6490828089905854e-06,
      "loss": 0.9682,
      "step": 190
    },
    {
      "epoch": 1.6324786324786325,
      "grad_norm": 0.7180127501487732,
      "learning_rate": 6.616666484156358e-06,
      "loss": 0.9209,
      "step": 191
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 0.41402408480644226,
      "learning_rate": 6.584174093857676e-06,
      "loss": 0.6809,
      "step": 192
    },
    {
      "epoch": 1.6495726495726495,
      "grad_norm": 0.4954575002193451,
      "learning_rate": 6.551607166887761e-06,
      "loss": 0.7514,
      "step": 193
    },
    {
      "epoch": 1.658119658119658,
      "grad_norm": 0.9597253799438477,
      "learning_rate": 6.5189672355468415e-06,
      "loss": 0.5929,
      "step": 194
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.49050456285476685,
      "learning_rate": 6.486255835570063e-06,
      "loss": 0.6365,
      "step": 195
    },
    {
      "epoch": 1.6752136752136753,
      "grad_norm": 0.3644927144050598,
      "learning_rate": 6.453474506055228e-06,
      "loss": 0.9474,
      "step": 196
    },
    {
      "epoch": 1.6837606837606838,
      "grad_norm": 0.41037657856941223,
      "learning_rate": 6.420624789390378e-06,
      "loss": 0.7692,
      "step": 197
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 0.33042111992836,
      "learning_rate": 6.387708231181229e-06,
      "loss": 0.644,
      "step": 198
    },
    {
      "epoch": 1.7008547008547008,
      "grad_norm": 0.4650563597679138,
      "learning_rate": 6.354726380178442e-06,
      "loss": 1.209,
      "step": 199
    },
    {
      "epoch": 1.7094017094017095,
      "grad_norm": 0.41142725944519043,
      "learning_rate": 6.3216807882047585e-06,
      "loss": 0.7169,
      "step": 200
    },
    {
      "epoch": 1.717948717948718,
      "grad_norm": 0.48036524653434753,
      "learning_rate": 6.288573010081984e-06,
      "loss": 0.7699,
      "step": 201
    },
    {
      "epoch": 1.7264957264957266,
      "grad_norm": 0.6143119931221008,
      "learning_rate": 6.255404603557833e-06,
      "loss": 0.7177,
      "step": 202
    },
    {
      "epoch": 1.735042735042735,
      "grad_norm": 0.42116302251815796,
      "learning_rate": 6.222177129232634e-06,
      "loss": 0.6262,
      "step": 203
    },
    {
      "epoch": 1.7435897435897436,
      "grad_norm": 0.42195364832878113,
      "learning_rate": 6.188892150485904e-06,
      "loss": 0.9916,
      "step": 204
    },
    {
      "epoch": 1.7521367521367521,
      "grad_norm": 0.46677255630493164,
      "learning_rate": 6.155551233402789e-06,
      "loss": 1.2428,
      "step": 205
    },
    {
      "epoch": 1.7606837606837606,
      "grad_norm": 0.5056412816047668,
      "learning_rate": 6.122155946700381e-06,
      "loss": 0.744,
      "step": 206
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.5227958559989929,
      "learning_rate": 6.088707861653904e-06,
      "loss": 0.7133,
      "step": 207
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.4398983418941498,
      "learning_rate": 6.0552085520227875e-06,
      "loss": 0.5776,
      "step": 208
    },
    {
      "epoch": 1.7863247863247862,
      "grad_norm": 0.42121821641921997,
      "learning_rate": 6.021659593976621e-06,
      "loss": 0.6745,
      "step": 209
    },
    {
      "epoch": 1.7948717948717947,
      "grad_norm": 0.4671107232570648,
      "learning_rate": 5.988062566020987e-06,
      "loss": 0.7452,
      "step": 210
    },
    {
      "epoch": 1.8034188034188035,
      "grad_norm": 0.45300018787384033,
      "learning_rate": 5.954419048923202e-06,
      "loss": 0.7965,
      "step": 211
    },
    {
      "epoch": 1.811965811965812,
      "grad_norm": 0.4954420030117035,
      "learning_rate": 5.920730625637934e-06,
      "loss": 0.8834,
      "step": 212
    },
    {
      "epoch": 1.8205128205128205,
      "grad_norm": 0.5425894260406494,
      "learning_rate": 5.886998881232715e-06,
      "loss": 0.7124,
      "step": 213
    },
    {
      "epoch": 1.8290598290598292,
      "grad_norm": 0.40424826741218567,
      "learning_rate": 5.853225402813381e-06,
      "loss": 0.713,
      "step": 214
    },
    {
      "epoch": 1.8376068376068377,
      "grad_norm": 0.3879939019680023,
      "learning_rate": 5.819411779449381e-06,
      "loss": 0.5891,
      "step": 215
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.44357284903526306,
      "learning_rate": 5.785559602099019e-06,
      "loss": 0.6287,
      "step": 216
    },
    {
      "epoch": 1.8547008547008548,
      "grad_norm": 0.3938916325569153,
      "learning_rate": 5.751670463534594e-06,
      "loss": 0.7215,
      "step": 217
    },
    {
      "epoch": 1.8632478632478633,
      "grad_norm": 0.39076554775238037,
      "learning_rate": 5.7177459582674595e-06,
      "loss": 0.6089,
      "step": 218
    },
    {
      "epoch": 1.8717948717948718,
      "grad_norm": 0.43660053610801697,
      "learning_rate": 5.683787682473003e-06,
      "loss": 0.6207,
      "step": 219
    },
    {
      "epoch": 1.8803418803418803,
      "grad_norm": 0.46270671486854553,
      "learning_rate": 5.649797233915539e-06,
      "loss": 0.6384,
      "step": 220
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 0.5016070604324341,
      "learning_rate": 5.615776211873142e-06,
      "loss": 0.7007,
      "step": 221
    },
    {
      "epoch": 1.8974358974358974,
      "grad_norm": 0.4464798867702484,
      "learning_rate": 5.5817262170623865e-06,
      "loss": 0.6267,
      "step": 222
    },
    {
      "epoch": 1.9059829059829059,
      "grad_norm": 0.47871559858322144,
      "learning_rate": 5.547648851563047e-06,
      "loss": 0.6108,
      "step": 223
    },
    {
      "epoch": 1.9145299145299144,
      "grad_norm": 0.4208378791809082,
      "learning_rate": 5.513545718742702e-06,
      "loss": 0.6503,
      "step": 224
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.4062391519546509,
      "learning_rate": 5.479418423181311e-06,
      "loss": 0.922,
      "step": 225
    },
    {
      "epoch": 1.9316239316239316,
      "grad_norm": 0.4971669018268585,
      "learning_rate": 5.4452685705957084e-06,
      "loss": 0.6235,
      "step": 226
    },
    {
      "epoch": 1.9401709401709402,
      "grad_norm": 0.45603546500205994,
      "learning_rate": 5.411097767764053e-06,
      "loss": 0.9878,
      "step": 227
    },
    {
      "epoch": 1.9487179487179487,
      "grad_norm": 0.419859915971756,
      "learning_rate": 5.376907622450229e-06,
      "loss": 0.5956,
      "step": 228
    },
    {
      "epoch": 1.9572649572649574,
      "grad_norm": 0.5258283615112305,
      "learning_rate": 5.342699743328203e-06,
      "loss": 0.6999,
      "step": 229
    },
    {
      "epoch": 1.965811965811966,
      "grad_norm": 0.46300017833709717,
      "learning_rate": 5.308475739906329e-06,
      "loss": 0.7178,
      "step": 230
    },
    {
      "epoch": 1.9743589743589745,
      "grad_norm": 0.5326732993125916,
      "learning_rate": 5.2742372224516235e-06,
      "loss": 0.6377,
      "step": 231
    },
    {
      "epoch": 1.982905982905983,
      "grad_norm": 0.4621569812297821,
      "learning_rate": 5.2399858019140005e-06,
      "loss": 0.6213,
      "step": 232
    },
    {
      "epoch": 1.9914529914529915,
      "grad_norm": 0.43373093008995056,
      "learning_rate": 5.205723089850472e-06,
      "loss": 0.6279,
      "step": 233
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.738778829574585,
      "learning_rate": 5.171450698349329e-06,
      "loss": 0.7957,
      "step": 234
    }
  ],
  "logging_steps": 1,
  "max_steps": 468,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 117,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.702116131294413e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}