{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 538,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00929368029739777,
      "grad_norm": 26.95801544189453,
      "learning_rate": 4.705882352941177e-06,
      "loss": 2.3188411712646486,
      "step": 5
    },
    {
      "epoch": 0.01858736059479554,
      "grad_norm": 9.583927154541016,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 1.9752429962158202,
      "step": 10
    },
    {
      "epoch": 0.027881040892193308,
      "grad_norm": 3.3495802879333496,
      "learning_rate": 1.647058823529412e-05,
      "loss": 1.9044214248657227,
      "step": 15
    },
    {
      "epoch": 0.03717472118959108,
      "grad_norm": 3.7255637645721436,
      "learning_rate": 1.9999272808103276e-05,
      "loss": 1.7568986892700196,
      "step": 20
    },
    {
      "epoch": 0.046468401486988845,
      "grad_norm": 3.0651931762695312,
      "learning_rate": 1.9991093113822542e-05,
      "loss": 1.689285659790039,
      "step": 25
    },
    {
      "epoch": 0.055762081784386616,
      "grad_norm": 2.572429656982422,
      "learning_rate": 1.997383219496441e-05,
      "loss": 1.5579983711242675,
      "step": 30
    },
    {
      "epoch": 0.06505576208178439,
      "grad_norm": 2.868806838989258,
      "learning_rate": 1.9947505740530947e-05,
      "loss": 1.596106719970703,
      "step": 35
    },
    {
      "epoch": 0.07434944237918216,
      "grad_norm": 1.9607709646224976,
      "learning_rate": 1.991213767947991e-05,
      "loss": 1.462496566772461,
      "step": 40
    },
    {
      "epoch": 0.08364312267657993,
      "grad_norm": 2.4217138290405273,
      "learning_rate": 1.986776015897494e-05,
      "loss": 1.4482147216796875,
      "step": 45
    },
    {
      "epoch": 0.09293680297397769,
      "grad_norm": 2.3672616481781006,
      "learning_rate": 1.9814413515165974e-05,
      "loss": 1.4047511100769043,
      "step": 50
    },
    {
      "epoch": 0.10223048327137546,
      "grad_norm": 1.9317628145217896,
      "learning_rate": 1.975214623652643e-05,
      "loss": 1.3898205757141113,
      "step": 55
    },
    {
      "epoch": 0.11152416356877323,
      "grad_norm": 2.4336366653442383,
      "learning_rate": 1.968101491978049e-05,
      "loss": 1.3858367919921875,
      "step": 60
    },
    {
      "epoch": 0.120817843866171,
      "grad_norm": 2.1864516735076904,
      "learning_rate": 1.9601084218460494e-05,
      "loss": 1.3119730949401855,
      "step": 65
    },
    {
      "epoch": 0.13011152416356878,
      "grad_norm": 1.8244357109069824,
      "learning_rate": 1.9512426784141307e-05,
      "loss": 1.39547700881958,
      "step": 70
    },
    {
      "epoch": 0.13940520446096655,
      "grad_norm": 2.205120086669922,
      "learning_rate": 1.9415123200404962e-05,
      "loss": 1.4431995391845702,
      "step": 75
    },
    {
      "epoch": 0.14869888475836432,
      "grad_norm": 2.166412830352783,
      "learning_rate": 1.930926190959566e-05,
      "loss": 1.5699316024780274,
      "step": 80
    },
    {
      "epoch": 0.1579925650557621,
      "grad_norm": 1.9325284957885742,
      "learning_rate": 1.9194939132431682e-05,
      "loss": 1.2972043037414551,
      "step": 85
    },
    {
      "epoch": 0.16728624535315986,
      "grad_norm": 2.332735538482666,
      "learning_rate": 1.9072258780547317e-05,
      "loss": 1.3499565124511719,
      "step": 90
    },
    {
      "epoch": 0.17657992565055763,
      "grad_norm": 2.37465500831604,
      "learning_rate": 1.894133236204423e-05,
      "loss": 1.4370722770690918,
      "step": 95
    },
    {
      "epoch": 0.18587360594795538,
      "grad_norm": 2.0243334770202637,
      "learning_rate": 1.880227888013818e-05,
      "loss": 1.3675311088562012,
      "step": 100
    },
    {
      "epoch": 0.19516728624535315,
      "grad_norm": 2.1604268550872803,
      "learning_rate": 1.8655224724993202e-05,
      "loss": 1.377614402770996,
      "step": 105
    },
    {
      "epoch": 0.20446096654275092,
      "grad_norm": 2.2026524543762207,
      "learning_rate": 1.850030355884151e-05,
      "loss": 1.4125652313232422,
      "step": 110
    },
    {
      "epoch": 0.2137546468401487,
      "grad_norm": 2.1034789085388184,
      "learning_rate": 1.8337656194493637e-05,
      "loss": 1.3413180351257323,
      "step": 115
    },
    {
      "epoch": 0.22304832713754646,
      "grad_norm": 2.506497859954834,
      "learning_rate": 1.8167430467349145e-05,
      "loss": 1.418524932861328,
      "step": 120
    },
    {
      "epoch": 0.23234200743494424,
      "grad_norm": 1.8340388536453247,
      "learning_rate": 1.7989781101024305e-05,
      "loss": 1.1914166450500487,
      "step": 125
    },
    {
      "epoch": 0.241635687732342,
      "grad_norm": 2.2226996421813965,
      "learning_rate": 1.7804869566718833e-05,
      "loss": 1.3447060585021973,
      "step": 130
    },
    {
      "epoch": 0.25092936802973975,
      "grad_norm": 2.4632489681243896,
      "learning_rate": 1.7612863936449568e-05,
      "loss": 1.2475160598754882,
      "step": 135
    },
    {
      "epoch": 0.26022304832713755,
      "grad_norm": 1.781369924545288,
      "learning_rate": 1.7413938730284404e-05,
      "loss": 1.2748435974121093,
      "step": 140
    },
    {
      "epoch": 0.2695167286245353,
      "grad_norm": 1.56698739528656,
      "learning_rate": 1.7208274757715425e-05,
      "loss": 1.2416476249694823,
      "step": 145
    },
    {
      "epoch": 0.2788104089219331,
      "grad_norm": 2.2493855953216553,
      "learning_rate": 1.6996058953315372e-05,
      "loss": 1.2200543403625488,
      "step": 150
    },
    {
      "epoch": 0.28810408921933084,
      "grad_norm": 2.2003560066223145,
      "learning_rate": 1.6777484206826793e-05,
      "loss": 1.411963939666748,
      "step": 155
    },
    {
      "epoch": 0.29739776951672864,
      "grad_norm": 2.178264856338501,
      "learning_rate": 1.6552749187838425e-05,
      "loss": 1.279794406890869,
      "step": 160
    },
    {
      "epoch": 0.3066914498141264,
      "grad_norm": 2.3037335872650146,
      "learning_rate": 1.632205816520799e-05,
      "loss": 1.2985937118530273,
      "step": 165
    },
    {
      "epoch": 0.3159851301115242,
      "grad_norm": 2.1715354919433594,
      "learning_rate": 1.6085620821395722e-05,
      "loss": 1.291085147857666,
      "step": 170
    },
    {
      "epoch": 0.3252788104089219,
      "grad_norm": 2.898193359375,
      "learning_rate": 1.5843652061877245e-05,
      "loss": 1.2797016143798827,
      "step": 175
    },
    {
      "epoch": 0.3345724907063197,
      "grad_norm": 1.671087622642517,
      "learning_rate": 1.5596371819809104e-05,
      "loss": 1.279003620147705,
      "step": 180
    },
    {
      "epoch": 0.34386617100371747,
      "grad_norm": 2.0380825996398926,
      "learning_rate": 1.534400485612449e-05,
      "loss": 1.2591672897338868,
      "step": 185
    },
    {
      "epoch": 0.35315985130111527,
      "grad_norm": 2.0796914100646973,
      "learning_rate": 1.5086780555240803e-05,
      "loss": 1.2169609069824219,
      "step": 190
    },
    {
      "epoch": 0.362453531598513,
      "grad_norm": 2.0199925899505615,
      "learning_rate": 1.482493271656482e-05,
      "loss": 1.2400360107421875,
      "step": 195
    },
    {
      "epoch": 0.37174721189591076,
      "grad_norm": 2.2911055088043213,
      "learning_rate": 1.4558699341984928e-05,
      "loss": 1.163477325439453,
      "step": 200
    },
    {
      "epoch": 0.38104089219330856,
      "grad_norm": 1.932687520980835,
      "learning_rate": 1.4288322419543576e-05,
      "loss": 1.1888216018676758,
      "step": 205
    },
    {
      "epoch": 0.3903345724907063,
      "grad_norm": 2.7049851417541504,
      "learning_rate": 1.40140477034866e-05,
      "loss": 1.2036534309387208,
      "step": 210
    },
    {
      "epoch": 0.3996282527881041,
      "grad_norm": 2.0440292358398438,
      "learning_rate": 1.373612449088931e-05,
      "loss": 1.2139246940612793,
      "step": 215
    },
    {
      "epoch": 0.40892193308550184,
      "grad_norm": 2.1460649967193604,
      "learning_rate": 1.3454805395062386e-05,
      "loss": 1.2134284019470214,
      "step": 220
    },
    {
      "epoch": 0.41821561338289964,
      "grad_norm": 2.0311391353607178,
      "learning_rate": 1.3170346115943575e-05,
      "loss": 1.1781021118164063,
      "step": 225
    },
    {
      "epoch": 0.4275092936802974,
      "grad_norm": 2.272679090499878,
      "learning_rate": 1.2883005207683784e-05,
      "loss": 1.2502960205078124,
      "step": 230
    },
    {
      "epoch": 0.4368029739776952,
      "grad_norm": 2.159701347351074,
      "learning_rate": 1.2593043843638978e-05,
      "loss": 1.1758886337280274,
      "step": 235
    },
    {
      "epoch": 0.44609665427509293,
      "grad_norm": 2.142303705215454,
      "learning_rate": 1.2300725578981308e-05,
      "loss": 1.1891971588134767,
      "step": 240
    },
    {
      "epoch": 0.45539033457249073,
      "grad_norm": 2.0654172897338867,
      "learning_rate": 1.2006316111145401e-05,
      "loss": 1.190941333770752,
      "step": 245
    },
    {
      "epoch": 0.4646840148698885,
      "grad_norm": 2.156872034072876,
      "learning_rate": 1.1710083038327436e-05,
      "loss": 1.1672106742858888,
      "step": 250
    },
    {
      "epoch": 0.4739776951672863,
      "grad_norm": 1.9415847063064575,
      "learning_rate": 1.1412295616256577e-05,
      "loss": 1.1619250297546386,
      "step": 255
    },
    {
      "epoch": 0.483271375464684,
      "grad_norm": 1.659597396850586,
      "learning_rate": 1.1113224513459819e-05,
      "loss": 1.260629940032959,
      "step": 260
    },
    {
      "epoch": 0.49256505576208176,
      "grad_norm": 1.8275004625320435,
      "learning_rate": 1.0813141565242682e-05,
      "loss": 1.2040957450866698,
      "step": 265
    },
    {
      "epoch": 0.5018587360594795,
      "grad_norm": 2.24941086769104,
      "learning_rate": 1.0512319526609405e-05,
      "loss": 1.1944546699523926,
      "step": 270
    },
    {
      "epoch": 0.5111524163568774,
      "grad_norm": 1.6864445209503174,
      "learning_rate": 1.021103182434718e-05,
      "loss": 1.1669962882995606,
      "step": 275
    },
    {
      "epoch": 0.5204460966542751,
      "grad_norm": 1.5734723806381226,
      "learning_rate": 9.909552308499792e-06,
      "loss": 1.283609676361084,
      "step": 280
    },
    {
      "epoch": 0.5297397769516728,
      "grad_norm": 2.204127311706543,
      "learning_rate": 9.608155003456529e-06,
      "loss": 1.1683622360229493,
      "step": 285
    },
    {
      "epoch": 0.5390334572490706,
      "grad_norm": 1.7805500030517578,
      "learning_rate": 9.307113858882664e-06,
      "loss": 1.081281852722168,
      "step": 290
    },
    {
      "epoch": 0.5483271375464684,
      "grad_norm": 2.25589656829834,
      "learning_rate": 9.006702500717786e-06,
      "loss": 1.2235270500183106,
      "step": 295
    },
    {
      "epoch": 0.5576208178438662,
      "grad_norm": 1.6985591650009155,
      "learning_rate": 8.707193982468456e-06,
      "loss": 1.2132750511169434,
      "step": 300
    },
    {
      "epoch": 0.5669144981412639,
      "grad_norm": 2.1660971641540527,
      "learning_rate": 8.408860537021127e-06,
      "loss": 1.2618220329284668,
      "step": 305
    },
    {
      "epoch": 0.5762081784386617,
      "grad_norm": 2.593118667602539,
      "learning_rate": 8.111973329200909e-06,
      "loss": 1.2217357635498047,
      "step": 310
    },
    {
      "epoch": 0.5855018587360595,
      "grad_norm": 3.7704007625579834,
      "learning_rate": 7.816802209301241e-06,
      "loss": 1.117323398590088,
      "step": 315
    },
    {
      "epoch": 0.5947955390334573,
      "grad_norm": 1.7586568593978882,
      "learning_rate": 7.523615467808249e-06,
      "loss": 1.1631814956665039,
      "step": 320
    },
    {
      "epoch": 0.604089219330855,
      "grad_norm": 1.9919712543487549,
      "learning_rate": 7.232679591542979e-06,
      "loss": 1.193962574005127,
      "step": 325
    },
    {
      "epoch": 0.6133828996282528,
      "grad_norm": 1.8056994676589966,
      "learning_rate": 6.944259021442967e-06,
      "loss": 1.0919481277465821,
      "step": 330
    },
    {
      "epoch": 0.6226765799256505,
      "grad_norm": 2.0190021991729736,
      "learning_rate": 6.6586159122033914e-06,
      "loss": 1.2040581703186035,
      "step": 335
    },
    {
      "epoch": 0.6319702602230484,
      "grad_norm": 1.8545655012130737,
      "learning_rate": 6.3760098939962935e-06,
      "loss": 1.2006896018981934,
      "step": 340
    },
    {
      "epoch": 0.6412639405204461,
      "grad_norm": 2.2342867851257324,
      "learning_rate": 6.096697836484382e-06,
      "loss": 1.1592247009277343,
      "step": 345
    },
    {
      "epoch": 0.6505576208178439,
      "grad_norm": 2.185673475265503,
      "learning_rate": 5.820933615343975e-06,
      "loss": 1.192073440551758,
      "step": 350
    },
    {
      "epoch": 0.6598513011152416,
      "grad_norm": 2.9750640392303467,
      "learning_rate": 5.548967881509275e-06,
      "loss": 1.2049386978149415,
      "step": 355
    },
    {
      "epoch": 0.6691449814126395,
      "grad_norm": 1.7897521257400513,
      "learning_rate": 5.281047833347676e-06,
      "loss": 1.200373649597168,
      "step": 360
    },
    {
      "epoch": 0.6784386617100372,
      "grad_norm": 2.0279908180236816,
      "learning_rate": 5.017416991973281e-06,
      "loss": 1.23820219039917,
      "step": 365
    },
    {
      "epoch": 0.6877323420074349,
      "grad_norm": 1.7079287767410278,
      "learning_rate": 4.758314979902734e-06,
      "loss": 1.0726004600524903,
      "step": 370
    },
    {
      "epoch": 0.6970260223048327,
      "grad_norm": 2.112967014312744,
      "learning_rate": 4.503977303254673e-06,
      "loss": 1.096217441558838,
      "step": 375
    },
    {
      "epoch": 0.7063197026022305,
      "grad_norm": 2.745797634124756,
      "learning_rate": 4.25463513769064e-06,
      "loss": 1.1325096130371093,
      "step": 380
    },
    {
      "epoch": 0.7156133828996283,
      "grad_norm": 1.653376817703247,
      "learning_rate": 4.010515118292127e-06,
      "loss": 1.086031436920166,
      "step": 385
    },
    {
      "epoch": 0.724907063197026,
      "grad_norm": 1.4193137884140015,
      "learning_rate": 3.771839133564704e-06,
      "loss": 1.1475345611572265,
      "step": 390
    },
    {
      "epoch": 0.7342007434944238,
      "grad_norm": 1.706320881843567,
      "learning_rate": 3.5388241237564337e-06,
      "loss": 1.1753997802734375,
      "step": 395
    },
    {
      "epoch": 0.7434944237918215,
      "grad_norm": 2.0797157287597656,
      "learning_rate": 3.311681883673937e-06,
      "loss": 1.1733462333679199,
      "step": 400
    },
    {
      "epoch": 0.7527881040892194,
      "grad_norm": 2.26751446723938,
      "learning_rate": 3.0906188701753127e-06,
      "loss": 1.096180534362793,
      "step": 405
    },
    {
      "epoch": 0.7620817843866171,
      "grad_norm": 1.4461183547973633,
      "learning_rate": 2.875836014514867e-06,
      "loss": 1.0737926483154296,
      "step": 410
    },
    {
      "epoch": 0.7713754646840149,
      "grad_norm": 1.5982372760772705,
      "learning_rate": 2.6675285397102856e-06,
      "loss": 1.128117561340332,
      "step": 415
    },
    {
      "epoch": 0.7806691449814126,
      "grad_norm": 1.8420089483261108,
      "learning_rate": 2.465885783098166e-06,
      "loss": 1.1598851203918457,
      "step": 420
    },
    {
      "epoch": 0.7899628252788105,
      "grad_norm": 2.2643508911132812,
      "learning_rate": 2.2710910242392468e-06,
      "loss": 1.1701539039611817,
      "step": 425
    },
    {
      "epoch": 0.7992565055762082,
      "grad_norm": 2.3221635818481445,
      "learning_rate": 2.0833213183297475e-06,
      "loss": 1.0971644401550293,
      "step": 430
    },
    {
      "epoch": 0.8085501858736059,
      "grad_norm": 2.5996909141540527,
      "learning_rate": 1.9027473352702208e-06,
      "loss": 1.142545509338379,
      "step": 435
    },
    {
      "epoch": 0.8178438661710037,
      "grad_norm": 2.051969289779663,
      "learning_rate": 1.729533204538224e-06,
      "loss": 0.9909985542297364,
      "step": 440
    },
    {
      "epoch": 0.8271375464684015,
      "grad_norm": 1.7680370807647705,
      "learning_rate": 1.563836366005782e-06,
      "loss": 1.1032094955444336,
      "step": 445
    },
    {
      "epoch": 0.8364312267657993,
      "grad_norm": 1.7899703979492188,
      "learning_rate": 1.4058074268372224e-06,
      "loss": 1.021013069152832,
      "step": 450
    },
    {
      "epoch": 0.845724907063197,
      "grad_norm": 1.2900737524032593,
      "learning_rate": 1.2555900245975262e-06,
      "loss": 1.1778299331665039,
      "step": 455
    },
    {
      "epoch": 0.8550185873605948,
      "grad_norm": 1.967742681503296,
      "learning_rate": 1.1133206966955213e-06,
      "loss": 1.012075424194336,
      "step": 460
    },
    {
      "epoch": 0.8643122676579925,
      "grad_norm": 2.1728663444519043,
      "learning_rate": 9.79128756280675e-07,
      "loss": 1.0343788146972657,
      "step": 465
    },
    {
      "epoch": 0.8736059479553904,
      "grad_norm": 1.8088945150375366,
      "learning_rate": 8.531361747062272e-07,
      "loss": 1.1092434883117677,
      "step": 470
    },
    {
      "epoch": 0.8828996282527881,
      "grad_norm": 2.509183168411255,
      "learning_rate": 7.354574706655038e-07,
      "loss": 1.0878626823425293,
      "step": 475
    },
    {
      "epoch": 0.8921933085501859,
      "grad_norm": 1.9416303634643555,
      "learning_rate": 6.261996061022335e-07,
      "loss": 1.015376377105713,
      "step": 480
    },
    {
      "epoch": 0.9014869888475836,
      "grad_norm": 1.4156267642974854,
      "learning_rate": 5.254618889893858e-07,
      "loss": 1.1004517555236817,
      "step": 485
    },
    {
      "epoch": 0.9107806691449815,
      "grad_norm": 2.256089210510254,
      "learning_rate": 4.3333588306499584e-07,
      "loss": 1.0936492919921874,
      "step": 490
    },
    {
      "epoch": 0.9200743494423792,
      "grad_norm": 2.2250120639801025,
      "learning_rate": 3.499053246069362e-07,
      "loss": 1.1194836616516113,
      "step": 495
    },
    {
      "epoch": 0.929368029739777,
      "grad_norm": 1.815142035484314,
      "learning_rate": 2.7524604632233054e-07,
      "loss": 1.1793383598327636,
      "step": 500
    },
    {
      "epoch": 0.9386617100371747,
      "grad_norm": 1.8478509187698364,
      "learning_rate": 2.0942590842078503e-07,
      "loss": 1.0344858169555664,
      "step": 505
    },
    {
      "epoch": 0.9479553903345725,
      "grad_norm": 1.5509213209152222,
      "learning_rate": 1.5250473693406486e-07,
      "loss": 0.9757624626159668,
      "step": 510
    },
    {
      "epoch": 0.9572490706319703,
      "grad_norm": 2.8569936752319336,
      "learning_rate": 1.0453426933830002e-07,
      "loss": 1.1587133407592773,
      "step": 515
    },
    {
      "epoch": 0.966542750929368,
      "grad_norm": 2.054386615753174,
      "learning_rate": 6.555810752813308e-08,
      "loss": 1.1704232215881347,
      "step": 520
    },
    {
      "epoch": 0.9758364312267658,
      "grad_norm": 1.4783203601837158,
      "learning_rate": 3.5611678185563106e-08,
      "loss": 1.222896194458008,
      "step": 525
    },
    {
      "epoch": 0.9851301115241635,
      "grad_norm": 1.8598870038986206,
      "learning_rate": 1.4722200579497803e-08,
      "loss": 1.2934889793395996,
      "step": 530
    },
    {
      "epoch": 0.9944237918215614,
      "grad_norm": 2.1733250617980957,
      "learning_rate": 2.9086618252893717e-09,
      "loss": 1.1760504722595215,
      "step": 535
    }
  ],
  "logging_steps": 5,
  "max_steps": 538,
  "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": 2.587637322683044e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}