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": 1.616557002067566,
      "learning_rate": 4e-05,
      "loss": 3.5006,
      "step": 1783
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 2.5352871417999268,
      "eval_rouge1": 0.571,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.5726,
      "eval_rougeLsum": 0.5713,
      "eval_runtime": 167.1974,
      "eval_samples_per_second": 4.486,
      "eval_steps_per_second": 0.144,
      "step": 1783
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2016897201538086,
      "learning_rate": 3e-05,
      "loss": 2.9963,
      "step": 3566
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 2.3938610553741455,
      "eval_rouge1": 0.4906,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.4911,
      "eval_rougeLsum": 0.4909,
      "eval_runtime": 166.4171,
      "eval_samples_per_second": 4.507,
      "eval_steps_per_second": 0.144,
      "step": 3566
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.6185603141784668,
      "learning_rate": 2e-05,
      "loss": 2.8832,
      "step": 5349
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 2.3137385845184326,
      "eval_rouge1": 0.4144,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.4158,
      "eval_rougeLsum": 0.4144,
      "eval_runtime": 166.4301,
      "eval_samples_per_second": 4.506,
      "eval_steps_per_second": 0.144,
      "step": 5349
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2302087545394897,
      "learning_rate": 1e-05,
      "loss": 2.8145,
      "step": 7132
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 2.2872378826141357,
      "eval_rouge1": 0.4356,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.435,
      "eval_rougeLsum": 0.4361,
      "eval_runtime": 167.356,
      "eval_samples_per_second": 4.481,
      "eval_steps_per_second": 0.143,
      "step": 7132
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.441249966621399,
      "learning_rate": 0.0,
      "loss": 2.7759,
      "step": 8915
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 2.2388505935668945,
      "eval_rouge1": 0.4284,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.431,
      "eval_rougeLsum": 0.4286,
      "eval_runtime": 167.1969,
      "eval_samples_per_second": 4.486,
      "eval_steps_per_second": 0.144,
      "step": 8915
    },
    {
      "epoch": 5.0,
      "step": 8915,
      "total_flos": 8.512744818393907e+16,
      "train_loss": 2.9941017925371565,
      "train_runtime": 5831.7853,
      "train_samples_per_second": 12.228,
      "train_steps_per_second": 1.529
    }
  ],
  "logging_steps": 500,
  "max_steps": 8915,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 8.512744818393907e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}