SmolLM3-DPO-Second-Round/train_results.json

{
    "epoch": 1.0,
    "total_flos": 0.0,
    "train_loss": 0.45319234429607314,
    "train_runtime": 13193.0651,
    "train_samples": 161271,
    "train_samples_per_second": 12.224,
    "train_steps_per_second": 0.382
}