{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 26.182233810424805,
      "learning_rate": 0.0,
      "loss": 0.8433,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 29.490903854370117,
      "learning_rate": 2.5e-06,
      "loss": 0.9102,
      "step": 2
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8433753848075867,
      "eval_runtime": 9.0159,
      "eval_samples_per_second": 4.326,
      "eval_steps_per_second": 0.555,
      "step": 2
    },
    {
      "epoch": 1.5,
      "grad_norm": 25.1413631439209,
      "learning_rate": 5e-06,
      "loss": 0.8534,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.563501358032227,
      "learning_rate": 4.962019382530521e-06,
      "loss": 0.6723,
      "step": 4
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4704136848449707,
      "eval_runtime": 8.3916,
      "eval_samples_per_second": 4.648,
      "eval_steps_per_second": 0.596,
      "step": 4
    },
    {
      "epoch": 2.5,
      "grad_norm": 6.3298869132995605,
      "learning_rate": 4.849231551964771e-06,
      "loss": 0.4748,
      "step": 5
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.5633349418640137,
      "learning_rate": 4.665063509461098e-06,
      "loss": 0.3556,
      "step": 6
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.3564150631427765,
      "eval_runtime": 8.4069,
      "eval_samples_per_second": 4.639,
      "eval_steps_per_second": 0.595,
      "step": 6
    },
    {
      "epoch": 3.5,
      "grad_norm": 2.360266923904419,
      "learning_rate": 4.415111107797445e-06,
      "loss": 0.3381,
      "step": 7
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9188865423202515,
      "learning_rate": 4.106969024216348e-06,
      "loss": 0.3261,
      "step": 8
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.32391461730003357,
      "eval_runtime": 8.4423,
      "eval_samples_per_second": 4.62,
      "eval_steps_per_second": 0.592,
      "step": 8
    },
    {
      "epoch": 4.5,
      "grad_norm": 1.7392710447311401,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.3077,
      "step": 9
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.303384304046631,
      "learning_rate": 3.3550503583141726e-06,
      "loss": 0.2697,
      "step": 10
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.29709815979003906,
      "eval_runtime": 8.4085,
      "eval_samples_per_second": 4.638,
      "eval_steps_per_second": 0.595,
      "step": 10
    },
    {
      "epoch": 5.5,
      "grad_norm": 2.265108346939087,
      "learning_rate": 2.9341204441673267e-06,
      "loss": 0.2657,
      "step": 11
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.813965916633606,
      "learning_rate": 2.5e-06,
      "loss": 0.2508,
      "step": 12
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.27973881363868713,
      "eval_runtime": 8.6049,
      "eval_samples_per_second": 4.532,
      "eval_steps_per_second": 0.581,
      "step": 12
    },
    {
      "epoch": 6.5,
      "grad_norm": 1.516356348991394,
      "learning_rate": 2.0658795558326745e-06,
      "loss": 0.2478,
      "step": 13
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.3250824213027954,
      "learning_rate": 1.6449496416858285e-06,
      "loss": 0.225,
      "step": 14
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.2676139175891876,
      "eval_runtime": 8.4345,
      "eval_samples_per_second": 4.624,
      "eval_steps_per_second": 0.593,
      "step": 14
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.1363078355789185,
      "learning_rate": 1.2500000000000007e-06,
      "loss": 0.2245,
      "step": 15
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.0826901197433472,
      "learning_rate": 8.930309757836517e-07,
      "loss": 0.2327,
      "step": 16
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.2616689205169678,
      "eval_runtime": 8.4051,
      "eval_samples_per_second": 4.64,
      "eval_steps_per_second": 0.595,
      "step": 16
    },
    {
      "epoch": 8.5,
      "grad_norm": 1.1075544357299805,
      "learning_rate": 5.848888922025553e-07,
      "loss": 0.2284,
      "step": 17
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.9888511300086975,
      "learning_rate": 3.3493649053890325e-07,
      "loss": 0.2064,
      "step": 18
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.25935298204421997,
      "eval_runtime": 8.4283,
      "eval_samples_per_second": 4.627,
      "eval_steps_per_second": 0.593,
      "step": 18
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.9419808983802795,
      "learning_rate": 1.507684480352292e-07,
      "loss": 0.2038,
      "step": 19
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.0084599256515503,
      "learning_rate": 3.798061746947995e-08,
      "loss": 0.22,
      "step": 20
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.25876814126968384,
      "eval_runtime": 8.9869,
      "eval_samples_per_second": 4.34,
      "eval_steps_per_second": 0.556,
      "step": 20
    },
    {
      "epoch": 10.0,
      "step": 20,
      "total_flos": 3.3888284102885376e+17,
      "train_loss": 0.3828152969479561,
      "train_runtime": 594.2554,
      "train_samples_per_second": 0.875,
      "train_steps_per_second": 0.034
    }
  ],
  "logging_steps": 1,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 2000000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.3888284102885376e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}