all_results.json

{
    "epoch": 2.975909305621162,
    "eval_dpo_loss": 0.7205542922019958,
    "eval_logits": -0.5665243268013,
    "eval_logps": -79.37840270996094,
    "eval_loss": 0.6457626223564148,
    "eval_objective": 0.6297247409820557,
    "eval_ranking_idealized": 0.6030020713806152,
    "eval_ranking_idealized_expo": 0.5222567319869995,
    "eval_ranking_simple": 0.534679114818573,
    "eval_regularize": 0.6297247409820557,
    "eval_runtime": 309.533,
    "eval_samples": 5790,
    "eval_samples_per_second": 18.706,
    "eval_steps_per_second": 1.56,
    "eval_wo_beta": 14.316285133361816,
    "total_flos": 0.0,
    "train_loss": 0.3115594020343962,
    "train_runtime": 28931.7373,
    "train_samples": 50802,
    "train_samples_per_second": 8.78,
    "train_steps_per_second": 0.061
}