trainer_state.json

{
  "best_metric": 5.144372463226318,
  "best_model_checkpoint": "./output_dir/checkpoint-314",
  "epoch": 0.9976171564733916,
  "eval_steps": 50.0,
  "global_step": 314,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 8.471337579617836e-05,
      "loss": 0.881,
      "step": 50
    },
    {
      "epoch": 0.32,
      "learning_rate": 6.878980891719745e-05,
      "loss": 0.244,
      "step": 100
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.286624203821656e-05,
      "loss": 0.1884,
      "step": 150
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.694267515923567e-05,
      "loss": 0.1744,
      "step": 200
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.1019108280254778e-05,
      "loss": 0.1615,
      "step": 250
    },
    {
      "epoch": 0.95,
      "learning_rate": 5.095541401273886e-06,
      "loss": 0.1235,
      "step": 300
    },
    {
      "epoch": 1.0,
      "eval_bleu-1": 82.7612,
      "eval_bleu-2": 79.8854,
      "eval_bleu-3": 79.4646,
      "eval_bleu-4": 79.2979,
      "eval_gen_len": 122.3333,
      "eval_loss": 5.144372463226318,
      "eval_runtime": 4.327,
      "eval_samples_per_second": 2.773,
      "eval_steps_per_second": 0.693,
      "step": 314
    },
    {
      "epoch": 1.0,
      "step": 314,
      "total_flos": 1.8650968363106304e+16,
      "train_loss": 0.2880971112828346,
      "train_runtime": 590.0609,
      "train_samples_per_second": 17.063,
      "train_steps_per_second": 0.532
    }
  ],
  "logging_steps": 50,
  "max_steps": 314,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 1.8650968363106304e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}