all_results.json

{
    "epoch": 5.0,
    "eval_logps_logqs/chosen": 0.025634765625,
    "eval_logqs/chosen": -4.0625,
    "eval_logqs/rejected": -3.8125,
    "eval_logqs_logps/rejected": -0.002044677734375,
    "eval_loss": 0.3089521825313568,
    "eval_runtime": 582.1707,
    "eval_samples": 1034,
    "eval_samples_per_second": 1.776,
    "eval_steps_per_second": 0.888,
    "total_flos": 0.0,
    "train_loss": 0.16590411390888093,
    "train_runtime": 31619.711,
    "train_samples": 7000,
    "train_samples_per_second": 1.107,
    "train_steps_per_second": 0.035
}