all_results.json

{
    "epoch": 4.988190836088805,
    "eval_dpo_loss": 0.6843491792678833,
    "eval_logits": -1.6921485662460327,
    "eval_logps": -95.97679138183594,
    "eval_loss": 0.4083395004272461,
    "eval_objective": 0.4120636284351349,
    "eval_ranking_idealized": 0.6570248007774353,
    "eval_ranking_idealized_expo": 0.5113636255264282,
    "eval_ranking_simple": 0.5206611752510071,
    "eval_regularize": 0.4120636284351349,
    "eval_runtime": 260.3366,
    "eval_samples": 5790,
    "eval_samples_per_second": 22.24,
    "eval_steps_per_second": 0.93,
    "total_flos": 0.0,
    "train_loss": 0.1521414952352643,
    "train_runtime": 35252.4203,
    "train_samples": 50802,
    "train_samples_per_second": 7.205,
    "train_steps_per_second": 0.025
}