trainer_state.json

{
  "best_metric": 0.5116,
  "best_model_checkpoint": "outputs/checkpoint-29586",
  "epoch": 2.0,
  "global_step": 59172,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 0.00027307465111127524,
      "loss": 3.6072,
      "step": 29586
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 15.1014,
      "eval_loss": 2.7608768939971924,
      "eval_runtime": 362.214,
      "eval_sacrebleu": 0.5116,
      "eval_samples_per_second": 6.286,
      "eval_steps_per_second": 3.145,
      "step": 29586
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0,
      "loss": 2.7156,
      "step": 59172
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 14.5924,
      "eval_loss": 2.6520941257476807,
      "eval_runtime": 359.3759,
      "eval_sacrebleu": 0.4482,
      "eval_samples_per_second": 6.336,
      "eval_steps_per_second": 3.169,
      "step": 59172
    },
    {
      "epoch": 2.0,
      "step": 59172,
      "total_flos": 8.028993014273434e+17,
      "train_loss": 3.16140578525316,
      "train_runtime": 93885.6881,
      "train_samples_per_second": 20.168,
      "train_steps_per_second": 0.63
    }
  ],
  "max_steps": 59172,
  "num_train_epochs": 2,
  "total_flos": 8.028993014273434e+17,
  "trial_name": null,
  "trial_params": null
}