checkpoint-20/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0026171159382360636,
  "eval_steps": 20,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0026171159382360636,
      "grad_norm": 0.7275105118751526,
      "learning_rate": 0.0,
      "log_odds_chosen": -0.11233717203140259,
      "log_odds_ratio": -0.7947841882705688,
      "logits/chosen": -1.6169023513793945,
      "logits/rejected": -1.5169153213500977,
      "logps/chosen": -2.330953359603882,
      "logps/rejected": -2.2381157875061035,
      "loss": 2.907,
      "nll_loss": 2.8274896144866943,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -0.23309533298015594,
      "rewards/margins": -0.009283773601055145,
      "rewards/rejected": -0.2238115817308426,
      "step": 20
    },
    {
      "epoch": 0.0026171159382360636,
      "eval_log_odds_chosen": 0.017083151265978813,
      "eval_log_odds_ratio": -0.7265545129776001,
      "eval_logits/chosen": -1.5783207416534424,
      "eval_logits/rejected": -1.5078614950180054,
      "eval_logps/chosen": -2.304877758026123,
      "eval_logps/rejected": -2.319162130355835,
      "eval_loss": 2.877645969390869,
      "eval_nll_loss": 2.804990768432617,
      "eval_rewards/accuracies": 0.49149999022483826,
      "eval_rewards/chosen": -0.23048776388168335,
      "eval_rewards/margins": 0.0014284582575783134,
      "eval_rewards/rejected": -0.23191621899604797,
      "eval_runtime": 4281.6698,
      "eval_samples_per_second": 0.467,
      "eval_steps_per_second": 0.234,
      "step": 20
    }
  ],
  "logging_steps": 20,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}