{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 18.0,
  "eval_steps": 500,
  "global_step": 3375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.2409825623035431,
      "learning_rate": 0.0001,
      "loss": 1.7886,
      "step": 100
    },
    {
      "epoch": 0.9973333333333333,
      "eval_accuracy": 0.6061076233183856,
      "eval_loss": 1.6901254653930664,
      "eval_runtime": 5.5484,
      "eval_samples_per_second": 90.116,
      "eval_steps_per_second": 11.355,
      "step": 187
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.2729445993900299,
      "learning_rate": 0.0001,
      "loss": 1.7283,
      "step": 200
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.37678101658821106,
      "learning_rate": 0.0001,
      "loss": 1.6544,
      "step": 300
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.607677130044843,
      "eval_loss": 1.6766204833984375,
      "eval_runtime": 6.4294,
      "eval_samples_per_second": 77.768,
      "eval_steps_per_second": 9.799,
      "step": 375
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.4604090750217438,
      "learning_rate": 0.0001,
      "loss": 1.6338,
      "step": 400
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5606330633163452,
      "learning_rate": 0.0001,
      "loss": 1.5273,
      "step": 500
    },
    {
      "epoch": 2.997333333333333,
      "eval_accuracy": 0.6079910313901346,
      "eval_loss": 1.6929064989089966,
      "eval_runtime": 5.431,
      "eval_samples_per_second": 92.063,
      "eval_steps_per_second": 11.6,
      "step": 562
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.6363856196403503,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 600
    },
    {
      "epoch": 3.7333333333333334,
      "grad_norm": 0.5985731482505798,
      "learning_rate": 0.0001,
      "loss": 1.3871,
      "step": 700
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.606896860986547,
      "eval_loss": 1.7257370948791504,
      "eval_runtime": 5.8318,
      "eval_samples_per_second": 85.737,
      "eval_steps_per_second": 10.803,
      "step": 750
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.7769015431404114,
      "learning_rate": 0.0001,
      "loss": 1.3013,
      "step": 800
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.708043098449707,
      "learning_rate": 0.0001,
      "loss": 1.23,
      "step": 900
    },
    {
      "epoch": 4.997333333333334,
      "eval_accuracy": 0.6061076233183856,
      "eval_loss": 1.7812739610671997,
      "eval_runtime": 5.9335,
      "eval_samples_per_second": 84.267,
      "eval_steps_per_second": 10.618,
      "step": 937
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.8580867648124695,
      "learning_rate": 0.0001,
      "loss": 1.1011,
      "step": 1000
    },
    {
      "epoch": 5.866666666666667,
      "grad_norm": 0.9232878684997559,
      "learning_rate": 0.0001,
      "loss": 1.0749,
      "step": 1100
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6018026905829597,
      "eval_loss": 1.877648949623108,
      "eval_runtime": 6.474,
      "eval_samples_per_second": 77.231,
      "eval_steps_per_second": 9.731,
      "step": 1125
    },
    {
      "epoch": 6.4,
      "grad_norm": 1.0364516973495483,
      "learning_rate": 0.0001,
      "loss": 0.9164,
      "step": 1200
    },
    {
      "epoch": 6.933333333333334,
      "grad_norm": 1.0082701444625854,
      "learning_rate": 0.0001,
      "loss": 0.8957,
      "step": 1300
    },
    {
      "epoch": 6.997333333333334,
      "eval_accuracy": 0.5997937219730942,
      "eval_loss": 1.978247046470642,
      "eval_runtime": 5.9491,
      "eval_samples_per_second": 84.047,
      "eval_steps_per_second": 10.59,
      "step": 1312
    },
    {
      "epoch": 7.466666666666667,
      "grad_norm": 1.0375924110412598,
      "learning_rate": 0.0001,
      "loss": 0.7382,
      "step": 1400
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.115308403968811,
      "learning_rate": 0.0001,
      "loss": 0.729,
      "step": 1500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5965919282511211,
      "eval_loss": 2.097362756729126,
      "eval_runtime": 6.5905,
      "eval_samples_per_second": 75.867,
      "eval_steps_per_second": 9.559,
      "step": 1500
    },
    {
      "epoch": 8.533333333333333,
      "grad_norm": 1.3251475095748901,
      "learning_rate": 0.0001,
      "loss": 0.5643,
      "step": 1600
    },
    {
      "epoch": 8.997333333333334,
      "eval_accuracy": 0.5931300448430493,
      "eval_loss": 2.2553179264068604,
      "eval_runtime": 6.1864,
      "eval_samples_per_second": 80.823,
      "eval_steps_per_second": 10.184,
      "step": 1687
    },
    {
      "epoch": 9.066666666666666,
      "grad_norm": 1.0753710269927979,
      "learning_rate": 0.0001,
      "loss": 0.5706,
      "step": 1700
    },
    {
      "epoch": 9.6,
      "grad_norm": 1.5658849477767944,
      "learning_rate": 0.0001,
      "loss": 0.4538,
      "step": 1800
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.5900627802690583,
      "eval_loss": 2.408897876739502,
      "eval_runtime": 6.4582,
      "eval_samples_per_second": 77.422,
      "eval_steps_per_second": 9.755,
      "step": 1875
    },
    {
      "epoch": 10.133333333333333,
      "grad_norm": 1.6254231929779053,
      "learning_rate": 0.0001,
      "loss": 0.4286,
      "step": 1900
    },
    {
      "epoch": 10.666666666666666,
      "grad_norm": 1.187232255935669,
      "learning_rate": 0.0001,
      "loss": 0.3563,
      "step": 2000
    },
    {
      "epoch": 10.997333333333334,
      "eval_accuracy": 0.588914798206278,
      "eval_loss": 2.529802083969116,
      "eval_runtime": 5.9806,
      "eval_samples_per_second": 83.604,
      "eval_steps_per_second": 10.534,
      "step": 2062
    },
    {
      "epoch": 11.2,
      "grad_norm": 1.3838512897491455,
      "learning_rate": 0.0001,
      "loss": 0.3256,
      "step": 2100
    },
    {
      "epoch": 11.733333333333333,
      "grad_norm": 1.249315619468689,
      "learning_rate": 0.0001,
      "loss": 0.2787,
      "step": 2200
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.5870582959641255,
      "eval_loss": 2.684791088104248,
      "eval_runtime": 5.7871,
      "eval_samples_per_second": 86.399,
      "eval_steps_per_second": 10.886,
      "step": 2250
    },
    {
      "epoch": 12.266666666666667,
      "grad_norm": 1.163707971572876,
      "learning_rate": 0.0001,
      "loss": 0.2558,
      "step": 2300
    },
    {
      "epoch": 12.8,
      "grad_norm": 1.092466950416565,
      "learning_rate": 0.0001,
      "loss": 0.2314,
      "step": 2400
    },
    {
      "epoch": 12.997333333333334,
      "eval_accuracy": 0.5863408071748879,
      "eval_loss": 2.7943203449249268,
      "eval_runtime": 6.5027,
      "eval_samples_per_second": 76.891,
      "eval_steps_per_second": 9.688,
      "step": 2437
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 1.1065410375595093,
      "learning_rate": 0.0001,
      "loss": 0.2029,
      "step": 2500
    },
    {
      "epoch": 13.866666666666667,
      "grad_norm": 1.1960605382919312,
      "learning_rate": 0.0001,
      "loss": 0.1923,
      "step": 2600
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.5856502242152466,
      "eval_loss": 2.8624069690704346,
      "eval_runtime": 5.5405,
      "eval_samples_per_second": 90.244,
      "eval_steps_per_second": 11.371,
      "step": 2625
    },
    {
      "epoch": 14.4,
      "grad_norm": 1.014957070350647,
      "learning_rate": 0.0001,
      "loss": 0.1661,
      "step": 2700
    },
    {
      "epoch": 14.933333333333334,
      "grad_norm": 1.2782268524169922,
      "learning_rate": 0.0001,
      "loss": 0.1687,
      "step": 2800
    },
    {
      "epoch": 14.997333333333334,
      "eval_accuracy": 0.5847533632286995,
      "eval_loss": 2.978261947631836,
      "eval_runtime": 6.5241,
      "eval_samples_per_second": 76.639,
      "eval_steps_per_second": 9.656,
      "step": 2812
    },
    {
      "epoch": 15.466666666666667,
      "grad_norm": 1.0024491548538208,
      "learning_rate": 0.0001,
      "loss": 0.1464,
      "step": 2900
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.9853018522262573,
      "learning_rate": 0.0001,
      "loss": 0.1514,
      "step": 3000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.5849865470852018,
      "eval_loss": 3.0238428115844727,
      "eval_runtime": 5.8457,
      "eval_samples_per_second": 85.533,
      "eval_steps_per_second": 10.777,
      "step": 3000
    },
    {
      "epoch": 16.533333333333335,
      "grad_norm": 1.2373902797698975,
      "learning_rate": 0.0001,
      "loss": 0.1282,
      "step": 3100
    },
    {
      "epoch": 16.997333333333334,
      "eval_accuracy": 0.5841614349775784,
      "eval_loss": 3.091371774673462,
      "eval_runtime": 5.5018,
      "eval_samples_per_second": 90.879,
      "eval_steps_per_second": 11.451,
      "step": 3187
    },
    {
      "epoch": 17.066666666666666,
      "grad_norm": 0.9496106505393982,
      "learning_rate": 0.0001,
      "loss": 0.1358,
      "step": 3200
    },
    {
      "epoch": 17.6,
      "grad_norm": 1.0109747648239136,
      "learning_rate": 0.0001,
      "loss": 0.121,
      "step": 3300
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.5847982062780269,
      "eval_loss": 3.143228530883789,
      "eval_runtime": 5.4395,
      "eval_samples_per_second": 91.919,
      "eval_steps_per_second": 11.582,
      "step": 3375
    }
  ],
  "logging_steps": 100,
  "max_steps": 9350,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.399578187296932e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}