last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 150.0,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 50.0,
      "learning_rate": 2.4875e-05,
      "loss": 2.9129,
      "step": 1000
    },
    {
      "epoch": 50.0,
      "eval_bleu": 92.7571,
      "eval_em": 0.2302,
      "eval_gen_len": 79.2806,
      "eval_loss": 0.3861088156700134,
      "eval_rm": 0.2422,
      "eval_runtime": 119.9581,
      "eval_samples_per_second": 3.476,
      "eval_steps_per_second": 0.225,
      "step": 1000
    },
    {
      "epoch": 100.0,
      "learning_rate": 4.9875000000000006e-05,
      "loss": 0.052,
      "step": 2000
    },
    {
      "epoch": 100.0,
      "eval_bleu": 95.3257,
      "eval_em": 0.5156,
      "eval_gen_len": 80.6859,
      "eval_loss": 0.3531915545463562,
      "eval_rm": 0.5635,
      "eval_runtime": 114.8636,
      "eval_samples_per_second": 3.63,
      "eval_steps_per_second": 0.235,
      "step": 2000
    },
    {
      "epoch": 150.0,
      "learning_rate": 3.75625e-05,
      "loss": 0.0155,
      "step": 3000
    },
    {
      "epoch": 150.0,
      "eval_bleu": 94.8017,
      "eval_em": 0.5659,
      "eval_gen_len": 79.3165,
      "eval_loss": 0.4279763698577881,
      "eval_rm": 0.6355,
      "eval_runtime": 113.6979,
      "eval_samples_per_second": 3.668,
      "eval_steps_per_second": 0.237,
      "step": 3000
    }
  ],
  "max_steps": 6000,
  "num_train_epochs": 300,
  "total_flos": 5067042185682432.0,
  "trial_name": null,
  "trial_params": null
}