llama-3-8b-base-slic-hf-ultrafeedback-4xh200/train_results.json

{
    "epoch": 0.9989528795811519,
    "total_flos": 0.0,
    "train_loss": 1398.719245958628,
    "train_runtime": 2677.8284,
    "train_samples": 61135,
    "train_samples_per_second": 22.83,
    "train_steps_per_second": 0.178
}