Files
P9-split3_only_answer_Qwen3…/train_results.json
ModelHub XC f5b7175b01 初始化项目,由ModelHub XC社区提供模型
Model: Hyeongwon/P9-split3_only_answer_Qwen3-4B-Base_0402-01-5e-6
Source: Original Platform
2026-06-06 12:08:34 +08:00

9 lines
232 B
JSON

{
"epoch": 6.0,
"total_flos": 513726926487552.0,
"train_loss": 0.5520123199349848,
"train_runtime": 70216.6152,
"train_samples": 41102,
"train_samples_per_second": 3.512,
"train_steps_per_second": 0.008
}