all_results.json

{
    "epoch": 3.122342938119981,
    "eval_dpo_loss": 2071.194580078125,
    "eval_logits": -0.5595079660415649,
    "eval_logps": -79.44622039794922,
    "eval_loss": 4901.447265625,
    "eval_objective": 4906.08837890625,
    "eval_ranking_idealized": 0.5212215185165405,
    "eval_ranking_idealized_expo": 0.5212215185165405,
    "eval_ranking_simple": 0.5362318754196167,
    "eval_regularize": 4906.08837890625,
    "eval_runtime": 308.5095,
    "eval_samples": 5790,
    "eval_samples_per_second": 18.768,
    "eval_steps_per_second": 1.566,
    "eval_wo_beta": 14.703009605407715,
    "total_flos": 0.0,
    "train_loss": 1018.9660795454546,
    "train_runtime": 6860.7986,
    "train_samples": 50802,
    "train_samples_per_second": 37.023,
    "train_steps_per_second": 0.257
}