all_results.json

{
    "epoch": 0.99,
    "eval_logits/chosen": 0.8014350533485413,
    "eval_logits/rejected": 0.81965172290802,
    "eval_logps/chosen": -91.72908020019531,
    "eval_logps/rejected": -94.21379089355469,
    "eval_loss": 2499.1455078125,
    "eval_rewards/accuracies": 0.25,
    "eval_rewards/chosen": -1.710148535494227e-05,
    "eval_rewards/margins": 0.00015072095266077667,
    "eval_rewards/rejected": -0.00016782239254098386,
    "eval_runtime": 274.4423,
    "eval_samples": 2000,
    "eval_samples_per_second": 7.288,
    "eval_steps_per_second": 0.455,
    "train_loss": 2499.3611225328946,
    "train_runtime": 1147.2033,
    "train_samples": 6113,
    "train_samples_per_second": 5.329,
    "train_steps_per_second": 0.083
}