Qwen2.5-7B-Whitebox-GSM8k-Exp/train_results.json

{
    "total_flos": 0.0,
    "train_loss": 0.07978848305841287,
    "train_runtime": 7290.3624,
    "train_samples": 7473,
    "train_samples_per_second": 1.027,
    "train_steps_per_second": 0.032
}