Files
Qwen3-4B-Instruct-2507-User…/all_results.json
ModelHub XC e270a365e2 初始化项目,由ModelHub XC社区提供模型
Model: davidanugraha/Qwen3-4B-Instruct-2507-UserSim-SFT-Factored
Source: Original Platform
2026-07-26 18:40:16 +08:00

8 lines
207 B
JSON

{
"epoch": 5.0,
"total_flos": 6.322240708254106e+17,
"train_loss": 1.4024348258972168,
"train_runtime": 1552.6736,
"train_samples_per_second": 2.273,
"train_steps_per_second": 0.039
}