all_results.json

{
    "epoch": 4.988190836088805,
    "eval_dpo_loss": 3.090646982192993,
    "eval_logits": -1.2660622596740723,
    "eval_logps": -88.38468170166016,
    "eval_loss": 5.930107116699219,
    "eval_objective": 5.97521448135376,
    "eval_ranking_idealized": 0.5092975497245789,
    "eval_ranking_idealized_expo": 0.5092975497245789,
    "eval_ranking_simple": 0.5134297609329224,
    "eval_regularize": 5.97521448135376,
    "eval_runtime": 258.9712,
    "eval_samples": 5790,
    "eval_samples_per_second": 22.358,
    "eval_steps_per_second": 0.934,
    "total_flos": 0.0,
    "train_loss": 2.940262828902765,
    "train_runtime": 35074.095,
    "train_samples": 50802,
    "train_samples_per_second": 7.242,
    "train_steps_per_second": 0.025
}