Files
PS_only_answer_Qwen3-4B-Bas…/train_results.json
ModelHub XC b7edf659df 初始化项目,由ModelHub XC社区提供模型
Model: Hyeongwon/PS_only_answer_Qwen3-4B-Base_0328-01-1e-5-seed45
Source: Original Platform
2026-05-13 17:16:28 +08:00

9 lines
232 B
JSON

{
"epoch": 6.0,
"total_flos": 475900034088960.0,
"train_loss": 0.4346669979916086,
"train_runtime": 96086.8853,
"train_samples": 18626,
"train_samples_per_second": 1.163,
"train_steps_per_second": 0.006
}