DeepSeek-R1-Distill-Qwen-1.5B-GRPO/train_results.json

{
    "total_flos": 0.0,
    "train_loss": 0.0464448650211034,
    "train_runtime": 1230.2242,
    "train_samples": 100,
    "train_samples_per_second": 0.081,
    "train_steps_per_second": 0.081
}