all_results.json

{
    "epoch": 4.988190836088805,
    "eval_dpo_loss": 0.6770760416984558,
    "eval_logits": -3.2698676586151123,
    "eval_logps": -195.45721435546875,
    "eval_loss": 0.39512959122657776,
    "eval_objective": 0.3955548405647278,
    "eval_ranking_idealized": 0.9194214940071106,
    "eval_ranking_idealized_expo": 0.5309917330741882,
    "eval_ranking_simple": 0.56611567735672,
    "eval_regularize": 0.3955548405647278,
    "eval_runtime": 259.4514,
    "eval_samples": 5790,
    "eval_samples_per_second": 22.316,
    "eval_steps_per_second": 0.933,
    "total_flos": 0.0,
    "train_loss": 0.1496292933313684,
    "train_runtime": 35141.4223,
    "train_samples": 50802,
    "train_samples_per_second": 7.228,
    "train_steps_per_second": 0.025
}