last-checkpoint/trainer_state.json

{
  "best_metric": 0.39197173714637756,
  "best_model_checkpoint": "../KevinKibe/nllb-200-distilled-1.3B-finetuned-finetuned-finetuned/checkpoint-200",
  "epoch": 200.0,
  "eval_steps": 200,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 200.0,
      "grad_norm": 0.19941186904907227,
      "learning_rate": 1.8e-05,
      "loss": 0.0469,
      "step": 200
    },
    {
      "epoch": 200.0,
      "eval_gen_len": 20.0,
      "eval_loss": 0.39197173714637756,
      "eval_rouge": 0.0476,
      "eval_runtime": 9.9932,
      "eval_samples_per_second": 0.2,
      "eval_steps_per_second": 0.1,
      "step": 200
    }
  ],
  "logging_steps": 200,
  "max_steps": 2000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2000,
  "save_steps": 200,
  "total_flos": 585173630976000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}