trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 8915,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.1129989624023438,
      "learning_rate": 4e-05,
      "loss": 1.2298,
      "step": 1783
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.6183177828788757,
      "eval_rouge1": 0.5036,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.5051,
      "eval_rougeLsum": 0.5002,
      "eval_runtime": 113.5797,
      "eval_samples_per_second": 6.594,
      "eval_steps_per_second": 0.211,
      "step": 1783
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2814314365386963,
      "learning_rate": 3e-05,
      "loss": 0.7893,
      "step": 3566
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5936142802238464,
      "eval_rouge1": 0.5166,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.5199,
      "eval_rougeLsum": 0.5139,
      "eval_runtime": 113.8839,
      "eval_samples_per_second": 6.577,
      "eval_steps_per_second": 0.211,
      "step": 3566
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9882182478904724,
      "learning_rate": 2e-05,
      "loss": 0.7368,
      "step": 5349
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5787354707717896,
      "eval_rouge1": 0.517,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.5231,
      "eval_rougeLsum": 0.516,
      "eval_runtime": 114.8491,
      "eval_samples_per_second": 6.522,
      "eval_steps_per_second": 0.209,
      "step": 5349
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3597564697265625,
      "learning_rate": 1e-05,
      "loss": 0.7107,
      "step": 7132
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5670214295387268,
      "eval_rouge1": 0.5105,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.5148,
      "eval_rougeLsum": 0.5089,
      "eval_runtime": 113.942,
      "eval_samples_per_second": 6.574,
      "eval_steps_per_second": 0.211,
      "step": 7132
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.5828347206115723,
      "learning_rate": 0.0,
      "loss": 0.6989,
      "step": 8915
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5651674866676331,
      "eval_rouge1": 0.5057,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.5089,
      "eval_rougeLsum": 0.5041,
      "eval_runtime": 113.7562,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 0.211,
      "step": 8915
    },
    {
      "epoch": 5.0,
      "step": 8915,
      "total_flos": 7.495546629705523e+16,
      "train_loss": 0.8330970066688868,
      "train_runtime": 5165.2629,
      "train_samples_per_second": 13.807,
      "train_steps_per_second": 1.726
    }
  ],
  "logging_steps": 500,
  "max_steps": 8915,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 7.495546629705523e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}