all_results.json

{
    "epoch": 3.689182805857345,
    "eval_dpo_loss": 2703.03125,
    "eval_logits": -0.4645061194896698,
    "eval_logps": -85.19200134277344,
    "eval_loss": 5280.861328125,
    "eval_objective": 5329.05712890625,
    "eval_ranking_idealized": 0.5212215185165405,
    "eval_ranking_idealized_expo": 0.5212215185165405,
    "eval_ranking_simple": 0.5263975262641907,
    "eval_regularize": 5329.05712890625,
    "eval_runtime": 308.7203,
    "eval_samples": 5790,
    "eval_samples_per_second": 18.755,
    "eval_steps_per_second": 1.565,
    "eval_wo_beta": 14.025749206542969,
    "total_flos": 0.0,
    "train_loss": 202.7064813701923,
    "train_runtime": 6887.9795,
    "train_samples": 50802,
    "train_samples_per_second": 36.877,
    "train_steps_per_second": 0.256
}