llama-3-8b-base-new-dpo-ultrafeedback-4xh200-batch-128-q_t-0.43-s_star-0.4/train_results.json

{
    "epoch": 0.9989528795811519,
    "total_flos": 0.0,
    "train_loss": 4.88172167052263,
    "train_runtime": 7274.147,
    "train_samples": 61135,
    "train_samples_per_second": 8.404,
    "train_steps_per_second": 0.066
}