{
  "best_metric": 0.45661377906799316,
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s4000/checkpoint-1000",
  "epoch": 22.22222222222222,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 4.398683071136475,
      "learning_rate": 2.4500000000000004e-07,
      "loss": 0.8108,
      "step": 50
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 3.726436138153076,
      "learning_rate": 4.95e-07,
      "loss": 0.792,
      "step": 100
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 2.0379676818847656,
      "learning_rate": 7.450000000000001e-07,
      "loss": 0.7413,
      "step": 150
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 2.0460364818573,
      "learning_rate": 9.950000000000002e-07,
      "loss": 0.7342,
      "step": 200
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 2.803539276123047,
      "learning_rate": 1.2450000000000002e-06,
      "loss": 0.7147,
      "step": 250
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 1.9818241596221924,
      "learning_rate": 1.495e-06,
      "loss": 0.7072,
      "step": 300
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 3.081399917602539,
      "learning_rate": 1.745e-06,
      "loss": 0.6843,
      "step": 350
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 2.4465785026550293,
      "learning_rate": 1.9950000000000004e-06,
      "loss": 0.657,
      "step": 400
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.929323434829712,
      "learning_rate": 2.245e-06,
      "loss": 0.6597,
      "step": 450
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 1.419311285018921,
      "learning_rate": 2.4950000000000003e-06,
      "loss": 0.6246,
      "step": 500
    },
    {
      "epoch": 11.11111111111111,
      "eval_loss": 0.5375720262527466,
      "eval_runtime": 9.2568,
      "eval_samples_per_second": 17.177,
      "eval_steps_per_second": 2.161,
      "step": 500
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 2.3438374996185303,
      "learning_rate": 2.7450000000000004e-06,
      "loss": 0.6018,
      "step": 550
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 1.7441222667694092,
      "learning_rate": 2.995e-06,
      "loss": 0.5547,
      "step": 600
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 1.5592021942138672,
      "learning_rate": 3.2450000000000003e-06,
      "loss": 0.5466,
      "step": 650
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 1.305444359779358,
      "learning_rate": 3.495e-06,
      "loss": 0.535,
      "step": 700
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 2.2067477703094482,
      "learning_rate": 3.745e-06,
      "loss": 0.5343,
      "step": 750
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 1.667480707168579,
      "learning_rate": 3.995000000000001e-06,
      "loss": 0.5296,
      "step": 800
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 1.5465797185897827,
      "learning_rate": 4.245e-06,
      "loss": 0.5198,
      "step": 850
    },
    {
      "epoch": 20.0,
      "grad_norm": 2.760589361190796,
      "learning_rate": 4.495e-06,
      "loss": 0.5193,
      "step": 900
    },
    {
      "epoch": 21.11111111111111,
      "grad_norm": 1.4570050239562988,
      "learning_rate": 4.745e-06,
      "loss": 0.5124,
      "step": 950
    },
    {
      "epoch": 22.22222222222222,
      "grad_norm": 2.1914265155792236,
      "learning_rate": 4.9950000000000005e-06,
      "loss": 0.5034,
      "step": 1000
    },
    {
      "epoch": 22.22222222222222,
      "eval_loss": 0.45661377906799316,
      "eval_runtime": 9.2158,
      "eval_samples_per_second": 17.253,
      "eval_steps_per_second": 2.17,
      "step": 1000
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 89,
  "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": 6889857259725792.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}