Qwen-2.5-7B-GRPO-NoBaseline-Adam-FisherMaskToken-1e-5-HessianMaskToken-0.01-v2_4270
/
train_results.json
| { | |
| "total_flos": 0.0, | |
| "train_loss": -0.7133333477377891, | |
| "train_runtime": 21711.9057, | |
| "train_samples": 7500, | |
| "train_samples_per_second": 0.442, | |
| "train_steps_per_second": 0.005 | |
| } |