trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 4470,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.2107882499694824,
      "learning_rate": 0.0008,
      "loss": 2.436,
      "step": 894
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 1.2680302858352661,
      "eval_rouge1": 0.0305,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.0305,
      "eval_rougeLsum": 0.0305,
      "eval_runtime": 59.1776,
      "eval_samples_per_second": 12.623,
      "eval_steps_per_second": 0.406,
      "step": 894
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.9003331661224365,
      "learning_rate": 0.0006,
      "loss": 1.5097,
      "step": 1788
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.9425563812255859,
      "eval_rouge1": 0.0223,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.0223,
      "eval_rougeLsum": 0.0223,
      "eval_runtime": 62.6611,
      "eval_samples_per_second": 11.921,
      "eval_steps_per_second": 0.383,
      "step": 1788
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.4691579341888428,
      "learning_rate": 0.0004,
      "loss": 1.2078,
      "step": 2682
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.7914846539497375,
      "eval_rouge1": 0.1643,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.1641,
      "eval_rougeLsum": 0.1639,
      "eval_runtime": 62.5819,
      "eval_samples_per_second": 11.936,
      "eval_steps_per_second": 0.383,
      "step": 2682
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.1174557209014893,
      "learning_rate": 0.0002,
      "loss": 1.0371,
      "step": 3576
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.7169661521911621,
      "eval_rouge1": 0.2501,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.2524,
      "eval_rougeLsum": 0.2507,
      "eval_runtime": 61.1078,
      "eval_samples_per_second": 12.224,
      "eval_steps_per_second": 0.393,
      "step": 3576
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.035046100616455,
      "learning_rate": 0.0,
      "loss": 0.9294,
      "step": 4470
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.6790085434913635,
      "eval_rouge1": 0.304,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3076,
      "eval_rougeLsum": 0.3081,
      "eval_runtime": 59.9431,
      "eval_samples_per_second": 12.462,
      "eval_steps_per_second": 0.4,
      "step": 4470
    },
    {
      "epoch": 5.0,
      "step": 4470,
      "total_flos": 5.60207031957504e+16,
      "train_loss": 1.4239941215088436,
      "train_runtime": 6465.3919,
      "train_samples_per_second": 11.051,
      "train_steps_per_second": 0.691
    }
  ],
  "logging_steps": 500,
  "max_steps": 4470,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 5.60207031957504e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}