all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -1.7287349700927734,
    "eval_logits/rejected": -1.7011512517929077,
    "eval_logps/chosen": -453.2426452636719,
    "eval_logps/rejected": -481.8094787597656,
    "eval_loss": 0.6444016695022583,
    "eval_rewards/accuracies": 0.6666666865348816,
    "eval_rewards/chosen": -2.479323625564575,
    "eval_rewards/margins": 0.47671714425086975,
    "eval_rewards/mix_margin": 0.17488937079906464,
    "eval_rewards/rejected": -2.9560413360595703,
    "eval_runtime": 716.6697,
    "eval_samples": 1522,
    "eval_samples_per_second": 2.124,
    "eval_steps_per_second": 0.532,
    "train_loss": 0.5442263817735228,
    "train_runtime": 33688.2171,
    "train_samples": 29394,
    "train_samples_per_second": 0.873,
    "train_steps_per_second": 0.014
}