{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 270,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.4779861330986024,
      "epoch": 0.11173184357541899,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002,
      "loss": 1.4731324195861817,
      "mean_token_accuracy": 0.7202368557453156,
      "num_tokens": 139114.0,
      "step": 10
    },
    {
      "entropy": 1.0133022338151931,
      "epoch": 0.22346368715083798,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0001992764570419069,
      "loss": 0.8902671813964844,
      "mean_token_accuracy": 0.8063317209482193,
      "num_tokens": 287784.0,
      "step": 20
    },
    {
      "entropy": 0.7525231212377548,
      "epoch": 0.33519553072625696,
      "grad_norm": 0.328125,
      "learning_rate": 0.00019711629845587164,
      "loss": 0.7365246772766113,
      "mean_token_accuracy": 0.8312019616365433,
      "num_tokens": 425110.0,
      "step": 30
    },
    {
      "entropy": 0.7569459870457649,
      "epoch": 0.44692737430167595,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0001935507835925601,
      "loss": 0.7281070709228515,
      "mean_token_accuracy": 0.8265180081129074,
      "num_tokens": 557368.0,
      "step": 40
    },
    {
      "entropy": 0.7587758377194405,
      "epoch": 0.5586592178770949,
      "grad_norm": 0.298828125,
      "learning_rate": 0.00018863150851539877,
      "loss": 0.7540057182312012,
      "mean_token_accuracy": 0.8272259771823883,
      "num_tokens": 697056.0,
      "step": 50
    },
    {
      "epoch": 0.5586592178770949,
      "eval_entropy": 0.7372037768363953,
      "eval_loss": 0.6630074381828308,
      "eval_mean_token_accuracy": 0.8417112976312637,
      "eval_num_tokens": 697056.0,
      "eval_runtime": 2.5833,
      "eval_samples_per_second": 24.774,
      "eval_steps_per_second": 3.097,
      "step": 50
    },
    {
      "entropy": 0.6524852216243744,
      "epoch": 0.6703910614525139,
      "grad_norm": 0.298828125,
      "learning_rate": 0.00018242965936120768,
      "loss": 0.6453401565551757,
      "mean_token_accuracy": 0.8488216072320938,
      "num_tokens": 839286.0,
      "step": 60
    },
    {
      "entropy": 0.6424317806959152,
      "epoch": 0.7821229050279329,
      "grad_norm": 0.271484375,
      "learning_rate": 0.00017503498221564025,
      "loss": 0.6219259262084961,
      "mean_token_accuracy": 0.8490033924579621,
      "num_tokens": 975302.0,
      "step": 70
    },
    {
      "entropy": 0.6861401617527008,
      "epoch": 0.8938547486033519,
      "grad_norm": 0.310546875,
      "learning_rate": 0.00016655448441021747,
      "loss": 0.6655363082885742,
      "mean_token_accuracy": 0.8398628294467926,
      "num_tokens": 1124078.0,
      "step": 80
    },
    {
      "entropy": 0.6849280376183359,
      "epoch": 1.0,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00015711088603430405,
      "loss": 0.6517285823822021,
      "mean_token_accuracy": 0.8404176392053303,
      "num_tokens": 1253938.0,
      "step": 90
    },
    {
      "entropy": 0.5663869976997375,
      "epoch": 1.111731843575419,
      "grad_norm": 0.3515625,
      "learning_rate": 0.00014684084406997903,
      "loss": 0.5269561767578125,
      "mean_token_accuracy": 0.867387443780899,
      "num_tokens": 1393724.0,
      "step": 100
    },
    {
      "epoch": 1.111731843575419,
      "eval_entropy": 0.5684726685285568,
      "eval_loss": 0.6147818565368652,
      "eval_mean_token_accuracy": 0.8511798083782196,
      "eval_num_tokens": 1393724.0,
      "eval_runtime": 2.5766,
      "eval_samples_per_second": 24.839,
      "eval_steps_per_second": 3.105,
      "step": 100
    },
    {
      "entropy": 0.6197814658284188,
      "epoch": 1.223463687150838,
      "grad_norm": 0.4375,
      "learning_rate": 0.0001358929748480946,
      "loss": 0.5889564514160156,
      "mean_token_accuracy": 0.8541492283344269,
      "num_tokens": 1532534.0,
      "step": 110
    },
    {
      "entropy": 0.6445826917886734,
      "epoch": 1.3351955307262569,
      "grad_norm": 0.419921875,
      "learning_rate": 0.00012442570344228313,
      "loss": 0.6050686836242676,
      "mean_token_accuracy": 0.8532525420188903,
      "num_tokens": 1672541.0,
      "step": 120
    },
    {
      "entropy": 0.6313177704811096,
      "epoch": 1.446927374301676,
      "grad_norm": 0.408203125,
      "learning_rate": 0.00011260497112202895,
      "loss": 0.5981513500213623,
      "mean_token_accuracy": 0.8510892212390899,
      "num_tokens": 1814087.0,
      "step": 130
    },
    {
      "entropy": 0.6326019644737244,
      "epoch": 1.558659217877095,
      "grad_norm": 0.330078125,
      "learning_rate": 0.00010060183403992856,
      "loss": 0.5860264301300049,
      "mean_token_accuracy": 0.8547371089458465,
      "num_tokens": 1955877.0,
      "step": 140
    },
    {
      "entropy": 0.563032440841198,
      "epoch": 1.670391061452514,
      "grad_norm": 0.365234375,
      "learning_rate": 8.858998790219753e-05,
      "loss": 0.5231016159057618,
      "mean_token_accuracy": 0.8669817239046097,
      "num_tokens": 2094363.0,
      "step": 150
    },
    {
      "epoch": 1.670391061452514,
      "eval_entropy": 0.5885124132037163,
      "eval_loss": 0.5921480655670166,
      "eval_mean_token_accuracy": 0.8505344167351723,
      "eval_num_tokens": 2094363.0,
      "eval_runtime": 2.5789,
      "eval_samples_per_second": 24.817,
      "eval_steps_per_second": 3.102,
      "step": 150
    },
    {
      "entropy": 0.5723796486854553,
      "epoch": 1.7821229050279328,
      "grad_norm": 0.388671875,
      "learning_rate": 7.674325444256899e-05,
      "loss": 0.5367129802703857,
      "mean_token_accuracy": 0.8645603656768799,
      "num_tokens": 2229344.0,
      "step": 160
    },
    {
      "entropy": 0.5252880439162254,
      "epoch": 1.893854748603352,
      "grad_norm": 0.373046875,
      "learning_rate": 6.523306607246527e-05,
      "loss": 0.4896131992340088,
      "mean_token_accuracy": 0.8740466505289077,
      "num_tokens": 2376372.0,
      "step": 170
    },
    {
      "entropy": 0.5986712182822981,
      "epoch": 2.0,
      "grad_norm": 0.5703125,
      "learning_rate": 5.422598510671666e-05,
      "loss": 0.5518558502197266,
      "mean_token_accuracy": 0.859065808747944,
      "num_tokens": 2507876.0,
      "step": 180
    },
    {
      "entropy": 0.58402778506279,
      "epoch": 2.111731843575419,
      "grad_norm": 0.3828125,
      "learning_rate": 4.388129346376178e-05,
      "loss": 0.5097923278808594,
      "mean_token_accuracy": 0.8693755626678467,
      "num_tokens": 2647895.0,
      "step": 190
    },
    {
      "entropy": 0.5195507079362869,
      "epoch": 2.223463687150838,
      "grad_norm": 0.310546875,
      "learning_rate": 3.4348687719438665e-05,
      "loss": 0.4614274024963379,
      "mean_token_accuracy": 0.8798785299062729,
      "num_tokens": 2802239.0,
      "step": 200
    },
    {
      "epoch": 2.223463687150838,
      "eval_entropy": 0.5370538644492626,
      "eval_loss": 0.5916606187820435,
      "eval_mean_token_accuracy": 0.8553198575973511,
      "eval_num_tokens": 2802239.0,
      "eval_runtime": 2.5788,
      "eval_samples_per_second": 24.818,
      "eval_steps_per_second": 3.102,
      "step": 200
    },
    {
      "entropy": 0.513118077814579,
      "epoch": 2.335195530726257,
      "grad_norm": 0.353515625,
      "learning_rate": 2.576611286891901e-05,
      "loss": 0.47298493385314944,
      "mean_token_accuracy": 0.8792857259511948,
      "num_tokens": 2945305.0,
      "step": 210
    },
    {
      "entropy": 0.5385309860110283,
      "epoch": 2.446927374301676,
      "grad_norm": 0.400390625,
      "learning_rate": 1.825776614411082e-05,
      "loss": 0.48717317581176756,
      "mean_token_accuracy": 0.8722086817026138,
      "num_tokens": 3079330.0,
      "step": 220
    },
    {
      "entropy": 0.5207595437765121,
      "epoch": 2.558659217877095,
      "grad_norm": 0.404296875,
      "learning_rate": 1.1932299773007228e-05,
      "loss": 0.4776895523071289,
      "mean_token_accuracy": 0.8777944326400757,
      "num_tokens": 3218477.0,
      "step": 230
    },
    {
      "entropy": 0.5544492155313492,
      "epoch": 2.6703910614525137,
      "grad_norm": 0.431640625,
      "learning_rate": 6.881248688597553e-06,
      "loss": 0.5149590969085693,
      "mean_token_accuracy": 0.8693302899599076,
      "num_tokens": 3364372.0,
      "step": 240
    },
    {
      "entropy": 0.5477708056569099,
      "epoch": 2.782122905027933,
      "grad_norm": 0.423828125,
      "learning_rate": 3.1777059397436692e-06,
      "loss": 0.4804663181304932,
      "mean_token_accuracy": 0.8743831604719162,
      "num_tokens": 3501064.0,
      "step": 250
    },
    {
      "epoch": 2.782122905027933,
      "eval_entropy": 0.5382514670491219,
      "eval_loss": 0.5897685289382935,
      "eval_mean_token_accuracy": 0.8542963713407516,
      "eval_num_tokens": 3501064.0,
      "eval_runtime": 2.5772,
      "eval_samples_per_second": 24.834,
      "eval_steps_per_second": 3.104,
      "step": 250
    },
    {
      "entropy": 0.5103608518838882,
      "epoch": 2.893854748603352,
      "grad_norm": 0.337890625,
      "learning_rate": 8.752649719641848e-07,
      "loss": 0.456165075302124,
      "mean_token_accuracy": 0.882489138841629,
      "num_tokens": 3648558.0,
      "step": 260
    },
    {
      "entropy": 0.5144368755189996,
      "epoch": 3.0,
      "grad_norm": 0.703125,
      "learning_rate": 7.244084232338466e-09,
      "loss": 0.4693727493286133,
      "mean_token_accuracy": 0.8807045284070467,
      "num_tokens": 3761814.0,
      "step": 270
    },
    {
      "epoch": 3.0,
      "eval_entropy": 0.5382274091243744,
      "eval_loss": 0.5899057388305664,
      "eval_mean_token_accuracy": 0.854370690882206,
      "eval_num_tokens": 3761814.0,
      "eval_runtime": 2.574,
      "eval_samples_per_second": 24.864,
      "eval_steps_per_second": 3.108,
      "step": 270
    }
  ],
  "logging_steps": 10,
  "max_steps": 270,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.189400112079462e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}