{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 40686,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03686771862557145,
      "grad_norm": 8.30236530303955,
      "learning_rate": 4.938676694686133e-05,
      "loss": 0.7878,
      "step": 500
    },
    {
      "epoch": 0.0737354372511429,
      "grad_norm": 6.442229270935059,
      "learning_rate": 4.8772304969768473e-05,
      "loss": 0.5945,
      "step": 1000
    },
    {
      "epoch": 0.11060315587671435,
      "grad_norm": 5.837690830230713,
      "learning_rate": 4.815784299267562e-05,
      "loss": 0.5546,
      "step": 1500
    },
    {
      "epoch": 0.1474708745022858,
      "grad_norm": 6.377337455749512,
      "learning_rate": 4.7543381015582754e-05,
      "loss": 0.5348,
      "step": 2000
    },
    {
      "epoch": 0.18433859312785725,
      "grad_norm": 6.086786270141602,
      "learning_rate": 4.69289190384899e-05,
      "loss": 0.5203,
      "step": 2500
    },
    {
      "epoch": 0.2212063117534287,
      "grad_norm": 8.050482749938965,
      "learning_rate": 4.631445706139704e-05,
      "loss": 0.506,
      "step": 3000
    },
    {
      "epoch": 0.2580740303790002,
      "grad_norm": 6.228991508483887,
      "learning_rate": 4.5699995084304186e-05,
      "loss": 0.49,
      "step": 3500
    },
    {
      "epoch": 0.2949417490045716,
      "grad_norm": 9.439841270446777,
      "learning_rate": 4.508553310721132e-05,
      "loss": 0.4804,
      "step": 4000
    },
    {
      "epoch": 0.3318094676301431,
      "grad_norm": 8.305259704589844,
      "learning_rate": 4.4471071130118466e-05,
      "loss": 0.4785,
      "step": 4500
    },
    {
      "epoch": 0.3686771862557145,
      "grad_norm": 9.281148910522461,
      "learning_rate": 4.385660915302561e-05,
      "loss": 0.4656,
      "step": 5000
    },
    {
      "epoch": 0.405544904881286,
      "grad_norm": 10.092153549194336,
      "learning_rate": 4.3242147175932754e-05,
      "loss": 0.4575,
      "step": 5500
    },
    {
      "epoch": 0.4424126235068574,
      "grad_norm": 5.881900787353516,
      "learning_rate": 4.26276851988399e-05,
      "loss": 0.4653,
      "step": 6000
    },
    {
      "epoch": 0.47928034213242887,
      "grad_norm": 5.321037769317627,
      "learning_rate": 4.201322322174704e-05,
      "loss": 0.4597,
      "step": 6500
    },
    {
      "epoch": 0.5161480607580003,
      "grad_norm": 3.0871853828430176,
      "learning_rate": 4.1398761244654185e-05,
      "loss": 0.4398,
      "step": 7000
    },
    {
      "epoch": 0.5530157793835717,
      "grad_norm": 6.368654251098633,
      "learning_rate": 4.078429926756133e-05,
      "loss": 0.4489,
      "step": 7500
    },
    {
      "epoch": 0.5898834980091432,
      "grad_norm": 5.613560676574707,
      "learning_rate": 4.0169837290468466e-05,
      "loss": 0.4438,
      "step": 8000
    },
    {
      "epoch": 0.6267512166347147,
      "grad_norm": 6.292891979217529,
      "learning_rate": 3.955537531337561e-05,
      "loss": 0.4335,
      "step": 8500
    },
    {
      "epoch": 0.6636189352602861,
      "grad_norm": 3.034226179122925,
      "learning_rate": 3.8940913336282753e-05,
      "loss": 0.4331,
      "step": 9000
    },
    {
      "epoch": 0.7004866538858575,
      "grad_norm": 3.5885934829711914,
      "learning_rate": 3.83264513591899e-05,
      "loss": 0.4289,
      "step": 9500
    },
    {
      "epoch": 0.737354372511429,
      "grad_norm": 7.238613128662109,
      "learning_rate": 3.7711989382097034e-05,
      "loss": 0.4218,
      "step": 10000
    },
    {
      "epoch": 0.7742220911370005,
      "grad_norm": 6.084052562713623,
      "learning_rate": 3.709752740500418e-05,
      "loss": 0.4196,
      "step": 10500
    },
    {
      "epoch": 0.811089809762572,
      "grad_norm": 8.307782173156738,
      "learning_rate": 3.648306542791132e-05,
      "loss": 0.4255,
      "step": 11000
    },
    {
      "epoch": 0.8479575283881433,
      "grad_norm": 4.905970573425293,
      "learning_rate": 3.5868603450818465e-05,
      "loss": 0.4222,
      "step": 11500
    },
    {
      "epoch": 0.8848252470137148,
      "grad_norm": 4.488884449005127,
      "learning_rate": 3.52541414737256e-05,
      "loss": 0.4149,
      "step": 12000
    },
    {
      "epoch": 0.9216929656392863,
      "grad_norm": 9.317034721374512,
      "learning_rate": 3.4639679496632746e-05,
      "loss": 0.4005,
      "step": 12500
    },
    {
      "epoch": 0.9585606842648577,
      "grad_norm": 3.8271901607513428,
      "learning_rate": 3.402521751953989e-05,
      "loss": 0.4114,
      "step": 13000
    },
    {
      "epoch": 0.9954284028904291,
      "grad_norm": 5.5162482261657715,
      "learning_rate": 3.3410755542447034e-05,
      "loss": 0.4051,
      "step": 13500
    },
    {
      "epoch": 1.0322961215160005,
      "grad_norm": 5.347577095031738,
      "learning_rate": 3.279629356535418e-05,
      "loss": 0.3568,
      "step": 14000
    },
    {
      "epoch": 1.069163840141572,
      "grad_norm": 4.08383321762085,
      "learning_rate": 3.218183158826132e-05,
      "loss": 0.3536,
      "step": 14500
    },
    {
      "epoch": 1.1060315587671434,
      "grad_norm": 3.5846598148345947,
      "learning_rate": 3.1567369611168465e-05,
      "loss": 0.3755,
      "step": 15000
    },
    {
      "epoch": 1.142899277392715,
      "grad_norm": 11.44876480102539,
      "learning_rate": 3.095290763407561e-05,
      "loss": 0.3682,
      "step": 15500
    },
    {
      "epoch": 1.1797669960182864,
      "grad_norm": 4.519806385040283,
      "learning_rate": 3.0338445656982746e-05,
      "loss": 0.3671,
      "step": 16000
    },
    {
      "epoch": 1.2166347146438579,
      "grad_norm": 7.143647193908691,
      "learning_rate": 2.972398367988989e-05,
      "loss": 0.3667,
      "step": 16500
    },
    {
      "epoch": 1.2535024332694293,
      "grad_norm": 10.113604545593262,
      "learning_rate": 2.9109521702797033e-05,
      "loss": 0.3656,
      "step": 17000
    },
    {
      "epoch": 1.2903701518950008,
      "grad_norm": 3.7289226055145264,
      "learning_rate": 2.8495059725704177e-05,
      "loss": 0.3593,
      "step": 17500
    },
    {
      "epoch": 1.3272378705205723,
      "grad_norm": 4.099955081939697,
      "learning_rate": 2.7880597748611314e-05,
      "loss": 0.3634,
      "step": 18000
    },
    {
      "epoch": 1.3641055891461438,
      "grad_norm": 5.327935218811035,
      "learning_rate": 2.7266135771518458e-05,
      "loss": 0.3661,
      "step": 18500
    },
    {
      "epoch": 1.400973307771715,
      "grad_norm": 4.683801174163818,
      "learning_rate": 2.66516737944256e-05,
      "loss": 0.3525,
      "step": 19000
    },
    {
      "epoch": 1.4378410263972865,
      "grad_norm": 6.552036762237549,
      "learning_rate": 2.6037211817332745e-05,
      "loss": 0.3589,
      "step": 19500
    },
    {
      "epoch": 1.474708745022858,
      "grad_norm": 6.014011859893799,
      "learning_rate": 2.5422749840239886e-05,
      "loss": 0.3609,
      "step": 20000
    },
    {
      "epoch": 1.5115764636484295,
      "grad_norm": 4.739955902099609,
      "learning_rate": 2.480828786314703e-05,
      "loss": 0.3574,
      "step": 20500
    },
    {
      "epoch": 1.548444182274001,
      "grad_norm": 8.343321800231934,
      "learning_rate": 2.4193825886054173e-05,
      "loss": 0.3513,
      "step": 21000
    },
    {
      "epoch": 1.5853119008995722,
      "grad_norm": 5.646194934844971,
      "learning_rate": 2.3579363908961314e-05,
      "loss": 0.3614,
      "step": 21500
    },
    {
      "epoch": 1.6221796195251437,
      "grad_norm": 5.9467973709106445,
      "learning_rate": 2.2964901931868457e-05,
      "loss": 0.3548,
      "step": 22000
    },
    {
      "epoch": 1.6590473381507151,
      "grad_norm": 7.322890281677246,
      "learning_rate": 2.2350439954775598e-05,
      "loss": 0.3493,
      "step": 22500
    },
    {
      "epoch": 1.6959150567762866,
      "grad_norm": 6.996487617492676,
      "learning_rate": 2.173597797768274e-05,
      "loss": 0.3569,
      "step": 23000
    },
    {
      "epoch": 1.732782775401858,
      "grad_norm": 6.08064079284668,
      "learning_rate": 2.1121516000589882e-05,
      "loss": 0.3471,
      "step": 23500
    },
    {
      "epoch": 1.7696504940274296,
      "grad_norm": 4.0622711181640625,
      "learning_rate": 2.0507054023497026e-05,
      "loss": 0.3391,
      "step": 24000
    },
    {
      "epoch": 1.806518212653001,
      "grad_norm": 5.080845355987549,
      "learning_rate": 1.989259204640417e-05,
      "loss": 0.3456,
      "step": 24500
    },
    {
      "epoch": 1.8433859312785725,
      "grad_norm": 6.117412090301514,
      "learning_rate": 1.9278130069311313e-05,
      "loss": 0.3486,
      "step": 25000
    },
    {
      "epoch": 1.880253649904144,
      "grad_norm": 5.242644786834717,
      "learning_rate": 1.8663668092218454e-05,
      "loss": 0.3445,
      "step": 25500
    },
    {
      "epoch": 1.9171213685297155,
      "grad_norm": 10.171354293823242,
      "learning_rate": 1.8049206115125597e-05,
      "loss": 0.3462,
      "step": 26000
    },
    {
      "epoch": 1.953989087155287,
      "grad_norm": 4.752163887023926,
      "learning_rate": 1.7434744138032738e-05,
      "loss": 0.3495,
      "step": 26500
    },
    {
      "epoch": 1.9908568057808584,
      "grad_norm": 5.205504417419434,
      "learning_rate": 1.682028216093988e-05,
      "loss": 0.349,
      "step": 27000
    },
    {
      "epoch": 2.02772452440643,
      "grad_norm": 5.852423191070557,
      "learning_rate": 1.6205820183847022e-05,
      "loss": 0.3112,
      "step": 27500
    },
    {
      "epoch": 2.064592243032001,
      "grad_norm": 9.391754150390625,
      "learning_rate": 1.5591358206754166e-05,
      "loss": 0.309,
      "step": 28000
    },
    {
      "epoch": 2.1014599616575724,
      "grad_norm": 7.519664287567139,
      "learning_rate": 1.4976896229661308e-05,
      "loss": 0.2971,
      "step": 28500
    },
    {
      "epoch": 2.138327680283144,
      "grad_norm": 4.092874526977539,
      "learning_rate": 1.4362434252568452e-05,
      "loss": 0.3057,
      "step": 29000
    },
    {
      "epoch": 2.1751953989087154,
      "grad_norm": 8.466115951538086,
      "learning_rate": 1.3747972275475594e-05,
      "loss": 0.2992,
      "step": 29500
    },
    {
      "epoch": 2.212063117534287,
      "grad_norm": 4.548240661621094,
      "learning_rate": 1.3133510298382737e-05,
      "loss": 0.3031,
      "step": 30000
    },
    {
      "epoch": 2.2489308361598583,
      "grad_norm": 8.980031967163086,
      "learning_rate": 1.2519048321289878e-05,
      "loss": 0.2981,
      "step": 30500
    },
    {
      "epoch": 2.28579855478543,
      "grad_norm": 7.443713188171387,
      "learning_rate": 1.1904586344197021e-05,
      "loss": 0.3039,
      "step": 31000
    },
    {
      "epoch": 2.3226662734110013,
      "grad_norm": 4.782769203186035,
      "learning_rate": 1.1290124367104164e-05,
      "loss": 0.3038,
      "step": 31500
    },
    {
      "epoch": 2.3595339920365728,
      "grad_norm": 7.365405082702637,
      "learning_rate": 1.0675662390011307e-05,
      "loss": 0.3013,
      "step": 32000
    },
    {
      "epoch": 2.3964017106621442,
      "grad_norm": 4.001774787902832,
      "learning_rate": 1.006120041291845e-05,
      "loss": 0.3031,
      "step": 32500
    },
    {
      "epoch": 2.4332694292877157,
      "grad_norm": 4.557787895202637,
      "learning_rate": 9.446738435825591e-06,
      "loss": 0.3076,
      "step": 33000
    },
    {
      "epoch": 2.470137147913287,
      "grad_norm": 6.75930118560791,
      "learning_rate": 8.832276458732734e-06,
      "loss": 0.2999,
      "step": 33500
    },
    {
      "epoch": 2.5070048665388587,
      "grad_norm": 9.001293182373047,
      "learning_rate": 8.217814481639877e-06,
      "loss": 0.3024,
      "step": 34000
    },
    {
      "epoch": 2.54387258516443,
      "grad_norm": 9.55263900756836,
      "learning_rate": 7.603352504547019e-06,
      "loss": 0.2994,
      "step": 34500
    },
    {
      "epoch": 2.5807403037900016,
      "grad_norm": 5.571417808532715,
      "learning_rate": 6.9888905274541614e-06,
      "loss": 0.3043,
      "step": 35000
    },
    {
      "epoch": 2.617608022415573,
      "grad_norm": 5.260648727416992,
      "learning_rate": 6.374428550361304e-06,
      "loss": 0.2964,
      "step": 35500
    },
    {
      "epoch": 2.6544757410411446,
      "grad_norm": 9.309138298034668,
      "learning_rate": 5.7599665732684464e-06,
      "loss": 0.302,
      "step": 36000
    },
    {
      "epoch": 2.691343459666716,
      "grad_norm": 6.128640651702881,
      "learning_rate": 5.145504596175589e-06,
      "loss": 0.2999,
      "step": 36500
    },
    {
      "epoch": 2.7282111782922875,
      "grad_norm": 8.606386184692383,
      "learning_rate": 4.5310426190827314e-06,
      "loss": 0.2916,
      "step": 37000
    },
    {
      "epoch": 2.7650788969178586,
      "grad_norm": 5.320854663848877,
      "learning_rate": 3.916580641989874e-06,
      "loss": 0.2931,
      "step": 37500
    },
    {
      "epoch": 2.80194661554343,
      "grad_norm": 8.016927719116211,
      "learning_rate": 3.3021186648970164e-06,
      "loss": 0.2912,
      "step": 38000
    },
    {
      "epoch": 2.8388143341690015,
      "grad_norm": 4.088160514831543,
      "learning_rate": 2.687656687804159e-06,
      "loss": 0.3015,
      "step": 38500
    },
    {
      "epoch": 2.875682052794573,
      "grad_norm": 9.513310432434082,
      "learning_rate": 2.0731947107113014e-06,
      "loss": 0.3003,
      "step": 39000
    },
    {
      "epoch": 2.9125497714201445,
      "grad_norm": 3.679173231124878,
      "learning_rate": 1.4587327336184437e-06,
      "loss": 0.2886,
      "step": 39500
    },
    {
      "epoch": 2.949417490045716,
      "grad_norm": 6.186450481414795,
      "learning_rate": 8.442707565255862e-07,
      "loss": 0.2975,
      "step": 40000
    },
    {
      "epoch": 2.9862852086712874,
      "grad_norm": 7.218328475952148,
      "learning_rate": 2.298087794327287e-07,
      "loss": 0.2925,
      "step": 40500
    }
  ],
  "logging_steps": 500,
  "max_steps": 40686,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9575666049362688.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}