{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012578616352201259,
      "grad_norm": 24.235414505004883,
      "learning_rate": 0.0,
      "loss": 5.9429,
      "step": 1
    },
    {
      "epoch": 0.025157232704402517,
      "grad_norm": 24.20835304260254,
      "learning_rate": 2.5e-06,
      "loss": 5.9877,
      "step": 2
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 23.525680541992188,
      "learning_rate": 5e-06,
      "loss": 5.9906,
      "step": 3
    },
    {
      "epoch": 0.050314465408805034,
      "grad_norm": 25.947757720947266,
      "learning_rate": 7.500000000000001e-06,
      "loss": 5.9738,
      "step": 4
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 23.722261428833008,
      "learning_rate": 1e-05,
      "loss": 5.9398,
      "step": 5
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 23.66267204284668,
      "learning_rate": 1.25e-05,
      "loss": 5.8582,
      "step": 6
    },
    {
      "epoch": 0.0880503144654088,
      "grad_norm": 22.97976303100586,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 5.7527,
      "step": 7
    },
    {
      "epoch": 0.10062893081761007,
      "grad_norm": 21.711450576782227,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 5.644,
      "step": 8
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 19.465877532958984,
      "learning_rate": 2e-05,
      "loss": 5.5225,
      "step": 9
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 17.033899307250977,
      "learning_rate": 1.999048221581858e-05,
      "loss": 5.2972,
      "step": 10
    },
    {
      "epoch": 0.13836477987421383,
      "grad_norm": 15.466821670532227,
      "learning_rate": 1.9961946980917457e-05,
      "loss": 5.1174,
      "step": 11
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 14.021515846252441,
      "learning_rate": 1.9914448613738107e-05,
      "loss": 4.9761,
      "step": 12
    },
    {
      "epoch": 0.16352201257861634,
      "grad_norm": 13.911083221435547,
      "learning_rate": 1.9848077530122083e-05,
      "loss": 4.8011,
      "step": 13
    },
    {
      "epoch": 0.1761006289308176,
      "grad_norm": 12.739690780639648,
      "learning_rate": 1.9762960071199334e-05,
      "loss": 4.6854,
      "step": 14
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 11.953996658325195,
      "learning_rate": 1.9659258262890683e-05,
      "loss": 4.5724,
      "step": 15
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 11.002123832702637,
      "learning_rate": 1.953716950748227e-05,
      "loss": 4.4691,
      "step": 16
    },
    {
      "epoch": 0.2138364779874214,
      "grad_norm": 10.387340545654297,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 4.3876,
      "step": 17
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 10.04249382019043,
      "learning_rate": 1.9238795325112867e-05,
      "loss": 4.2144,
      "step": 18
    },
    {
      "epoch": 0.2389937106918239,
      "grad_norm": 9.719793319702148,
      "learning_rate": 1.9063077870366504e-05,
      "loss": 4.0801,
      "step": 19
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 9.387718200683594,
      "learning_rate": 1.887010833178222e-05,
      "loss": 3.9868,
      "step": 20
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 9.012751579284668,
      "learning_rate": 1.866025403784439e-05,
      "loss": 3.9461,
      "step": 21
    },
    {
      "epoch": 0.27672955974842767,
      "grad_norm": 9.132621765136719,
      "learning_rate": 1.843391445812886e-05,
      "loss": 3.881,
      "step": 22
    },
    {
      "epoch": 0.2893081761006289,
      "grad_norm": 8.821338653564453,
      "learning_rate": 1.819152044288992e-05,
      "loss": 3.7709,
      "step": 23
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 8.580257415771484,
      "learning_rate": 1.7933533402912354e-05,
      "loss": 3.7456,
      "step": 24
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 7.926499843597412,
      "learning_rate": 1.766044443118978e-05,
      "loss": 3.6151,
      "step": 25
    },
    {
      "epoch": 0.3270440251572327,
      "grad_norm": 7.801671028137207,
      "learning_rate": 1.737277336810124e-05,
      "loss": 3.6009,
      "step": 26
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 7.5823283195495605,
      "learning_rate": 1.7071067811865477e-05,
      "loss": 3.5085,
      "step": 27
    },
    {
      "epoch": 0.3522012578616352,
      "grad_norm": 7.164687633514404,
      "learning_rate": 1.6755902076156606e-05,
      "loss": 3.4462,
      "step": 28
    },
    {
      "epoch": 0.36477987421383645,
      "grad_norm": 6.993910312652588,
      "learning_rate": 1.6427876096865394e-05,
      "loss": 3.4085,
      "step": 29
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 6.862181663513184,
      "learning_rate": 1.608761429008721e-05,
      "loss": 3.3406,
      "step": 30
    },
    {
      "epoch": 0.389937106918239,
      "grad_norm": 6.750012397766113,
      "learning_rate": 1.573576436351046e-05,
      "loss": 3.2278,
      "step": 31
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 6.868540287017822,
      "learning_rate": 1.5372996083468242e-05,
      "loss": 3.1773,
      "step": 32
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 6.78640604019165,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 3.1175,
      "step": 33
    },
    {
      "epoch": 0.4276729559748428,
      "grad_norm": 7.065934181213379,
      "learning_rate": 1.4617486132350343e-05,
      "loss": 3.0835,
      "step": 34
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 7.395648956298828,
      "learning_rate": 1.4226182617406996e-05,
      "loss": 3.0579,
      "step": 35
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 7.377147197723389,
      "learning_rate": 1.3826834323650899e-05,
      "loss": 3.001,
      "step": 36
    },
    {
      "epoch": 0.46540880503144655,
      "grad_norm": 7.350756645202637,
      "learning_rate": 1.342020143325669e-05,
      "loss": 2.9134,
      "step": 37
    },
    {
      "epoch": 0.4779874213836478,
      "grad_norm": 7.21485710144043,
      "learning_rate": 1.300705799504273e-05,
      "loss": 2.8916,
      "step": 38
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 7.14323616027832,
      "learning_rate": 1.2588190451025209e-05,
      "loss": 2.8545,
      "step": 39
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 7.045004844665527,
      "learning_rate": 1.2164396139381029e-05,
      "loss": 2.7926,
      "step": 40
    },
    {
      "epoch": 0.5157232704402516,
      "grad_norm": 7.102219104766846,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 2.7661,
      "step": 41
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 7.334245681762695,
      "learning_rate": 1.130526192220052e-05,
      "loss": 2.7134,
      "step": 42
    },
    {
      "epoch": 0.5408805031446541,
      "grad_norm": 7.159778594970703,
      "learning_rate": 1.0871557427476585e-05,
      "loss": 2.6547,
      "step": 43
    },
    {
      "epoch": 0.5534591194968553,
      "grad_norm": 7.077873706817627,
      "learning_rate": 1.0436193873653362e-05,
      "loss": 2.607,
      "step": 44
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 7.217611789703369,
      "learning_rate": 1e-05,
      "loss": 2.6118,
      "step": 45
    },
    {
      "epoch": 0.5786163522012578,
      "grad_norm": 6.973113059997559,
      "learning_rate": 9.563806126346643e-06,
      "loss": 2.5558,
      "step": 46
    },
    {
      "epoch": 0.5911949685534591,
      "grad_norm": 7.162886142730713,
      "learning_rate": 9.128442572523418e-06,
      "loss": 2.5148,
      "step": 47
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 7.065187931060791,
      "learning_rate": 8.694738077799487e-06,
      "loss": 2.4874,
      "step": 48
    },
    {
      "epoch": 0.6163522012578616,
      "grad_norm": 7.004238605499268,
      "learning_rate": 8.263518223330698e-06,
      "loss": 2.4503,
      "step": 49
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 7.0252790451049805,
      "learning_rate": 7.835603860618973e-06,
      "loss": 2.4348,
      "step": 50
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 6.905361175537109,
      "learning_rate": 7.411809548974792e-06,
      "loss": 2.377,
      "step": 51
    },
    {
      "epoch": 0.6540880503144654,
      "grad_norm": 6.864150524139404,
      "learning_rate": 6.992942004957271e-06,
      "loss": 2.3397,
      "step": 52
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 6.928408145904541,
      "learning_rate": 6.579798566743314e-06,
      "loss": 2.331,
      "step": 53
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 6.910766124725342,
      "learning_rate": 6.173165676349103e-06,
      "loss": 2.3138,
      "step": 54
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 6.662342071533203,
      "learning_rate": 5.773817382593008e-06,
      "loss": 2.2694,
      "step": 55
    },
    {
      "epoch": 0.7044025157232704,
      "grad_norm": 6.670816421508789,
      "learning_rate": 5.382513867649663e-06,
      "loss": 2.2476,
      "step": 56
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": 6.450694561004639,
      "learning_rate": 5.000000000000003e-06,
      "loss": 2.23,
      "step": 57
    },
    {
      "epoch": 0.7295597484276729,
      "grad_norm": 6.611984729766846,
      "learning_rate": 4.627003916531761e-06,
      "loss": 2.1873,
      "step": 58
    },
    {
      "epoch": 0.7421383647798742,
      "grad_norm": 6.599513053894043,
      "learning_rate": 4.264235636489542e-06,
      "loss": 2.1645,
      "step": 59
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 6.474591255187988,
      "learning_rate": 3.912385709912794e-06,
      "loss": 2.132,
      "step": 60
    },
    {
      "epoch": 0.7672955974842768,
      "grad_norm": 6.484611988067627,
      "learning_rate": 3.5721239031346067e-06,
      "loss": 2.1703,
      "step": 61
    },
    {
      "epoch": 0.779874213836478,
      "grad_norm": 6.360864162445068,
      "learning_rate": 3.2440979238433977e-06,
      "loss": 2.1489,
      "step": 62
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 6.536640644073486,
      "learning_rate": 2.9289321881345257e-06,
      "loss": 2.1015,
      "step": 63
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 6.510339736938477,
      "learning_rate": 2.6272266318987606e-06,
      "loss": 2.0919,
      "step": 64
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 6.453073501586914,
      "learning_rate": 2.339555568810221e-06,
      "loss": 2.1152,
      "step": 65
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 6.530029773712158,
      "learning_rate": 2.0664665970876496e-06,
      "loss": 2.1062,
      "step": 66
    },
    {
      "epoch": 0.8427672955974843,
      "grad_norm": 6.49432897567749,
      "learning_rate": 1.808479557110081e-06,
      "loss": 2.1046,
      "step": 67
    },
    {
      "epoch": 0.8553459119496856,
      "grad_norm": 6.546372890472412,
      "learning_rate": 1.566085541871145e-06,
      "loss": 2.0763,
      "step": 68
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 6.68710994720459,
      "learning_rate": 1.339745962155613e-06,
      "loss": 2.0718,
      "step": 69
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 6.77480936050415,
      "learning_rate": 1.129891668217783e-06,
      "loss": 2.0944,
      "step": 70
    },
    {
      "epoch": 0.8930817610062893,
      "grad_norm": 6.603630065917969,
      "learning_rate": 9.369221296335007e-07,
      "loss": 2.0652,
      "step": 71
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 6.616003513336182,
      "learning_rate": 7.612046748871327e-07,
      "loss": 2.0317,
      "step": 72
    },
    {
      "epoch": 0.9182389937106918,
      "grad_norm": 6.576902389526367,
      "learning_rate": 6.030737921409169e-07,
      "loss": 2.0268,
      "step": 73
    },
    {
      "epoch": 0.9308176100628931,
      "grad_norm": 6.478517055511475,
      "learning_rate": 4.628304925177318e-07,
      "loss": 2.0539,
      "step": 74
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 6.657393455505371,
      "learning_rate": 3.4074173710931804e-07,
      "loss": 2.0288,
      "step": 75
    },
    {
      "epoch": 0.9559748427672956,
      "grad_norm": 6.596218585968018,
      "learning_rate": 2.370399288006664e-07,
      "loss": 2.0511,
      "step": 76
    },
    {
      "epoch": 0.9685534591194969,
      "grad_norm": 6.601206302642822,
      "learning_rate": 1.519224698779198e-07,
      "loss": 2.0497,
      "step": 77
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 6.5980401039123535,
      "learning_rate": 8.555138626189619e-08,
      "loss": 2.0436,
      "step": 78
    },
    {
      "epoch": 0.9937106918238994,
      "grad_norm": 6.563250541687012,
      "learning_rate": 3.805301908254455e-08,
      "loss": 2.0327,
      "step": 79
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.6095662117004395,
      "learning_rate": 9.517784181422018e-09,
      "loss": 1.9954,
      "step": 80
    }
  ],
  "logging_steps": 1,
  "max_steps": 80,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 9.836271296493304e+17,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}