Files
Qwen1.5-MOE-sft-ESFT-summary/train_results.json
ModelHub XC 220b277a01 初始化项目,由ModelHub XC社区提供模型
Model: jayzou3773/Qwen1.5-MOE-sft-ESFT-summary
Source: Original Platform
2026-05-05 19:50:49 +08:00

8 lines
217 B
JSON

{
"total_flos": 1.770321796592042e+18,
"train_loss": 1.3923614178746209,
"train_runtime": 2541.2937,
"train_samples": 19587,
"train_samples_per_second": 7.707,
"train_steps_per_second": 0.241
}