Files
Qwen3-4B_Paper_Impact_model…/train_results.json
ModelHub XC 683943d8c2 初始化项目,由ModelHub XC社区提供模型
Model: FlyPig23/Qwen3-4B_Paper_Impact_model_SFT_1ep
Source: Original Platform
2026-04-29 16:35:06 +08:00

8 lines
208 B
JSON

{
"epoch": 1.0,
"total_flos": 2.743502659167191e+17,
"train_loss": 0.12591312176141983,
"train_runtime": 1812.1381,
"train_samples_per_second": 2.749,
"train_steps_per_second": 0.043
}