trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.948616600790514,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_bleu": 25.8406,
      "eval_gen_len": 37.7921,
      "eval_loss": 1.0771484375,
      "eval_runtime": 90.1252,
      "eval_samples_per_second": 4.483,
      "eval_steps_per_second": 0.288,
      "step": 303
    },
    {
      "epoch": 1.65,
      "learning_rate": 2e-05,
      "loss": 1.1552,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_bleu": 27.5846,
      "eval_gen_len": 37.9579,
      "eval_loss": 0.99169921875,
      "eval_runtime": 87.9875,
      "eval_samples_per_second": 4.592,
      "eval_steps_per_second": 0.295,
      "step": 606
    },
    {
      "epoch": 3.0,
      "eval_bleu": 28.1347,
      "eval_gen_len": 38.3317,
      "eval_loss": 0.97314453125,
      "eval_runtime": 92.4432,
      "eval_samples_per_second": 4.37,
      "eval_steps_per_second": 0.281,
      "step": 909
    },
    {
      "epoch": 3.3,
      "learning_rate": 2e-05,
      "loss": 0.679,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_bleu": 28.081,
      "eval_gen_len": 37.651,
      "eval_loss": 0.99169921875,
      "eval_runtime": 85.9085,
      "eval_samples_per_second": 4.703,
      "eval_steps_per_second": 0.303,
      "step": 1212
    },
    {
      "epoch": 4.95,
      "learning_rate": 2e-05,
      "loss": 0.5046,
      "step": 1500
    }
  ],
  "max_steps": 1515,
  "num_train_epochs": 5,
  "total_flos": 1.2130831043867443e+17,
  "trial_name": null,
  "trial_params": null
}