all_results.json

{
    "epoch": 0.997638167217761,
    "eval_dpo_loss": 12906.1103515625,
    "eval_logits": -1.3459484577178955,
    "eval_logps": -80.87725067138672,
    "eval_loss": 12497.908203125,
    "eval_objective": 12906.1103515625,
    "eval_ranking_idealized": 0.5092975497245789,
    "eval_ranking_idealized_expo": 0.5092975497245789,
    "eval_ranking_simple": 0.5258264541625977,
    "eval_regularize": 12906.1103515625,
    "eval_runtime": 258.4891,
    "eval_samples": 5790,
    "eval_samples_per_second": 22.399,
    "eval_steps_per_second": 0.936,
    "total_flos": 0.0,
    "train_loss": 8072.862625997175,
    "train_runtime": 10876.2953,
    "train_samples": 50802,
    "train_samples_per_second": 4.671,
    "train_steps_per_second": 0.016
}