Why does the latest code seem to have slower training speed and increased memory usage?

[xdaiycl]

Describe the bug
What the bug is, and how to reproduce, better with screenshots(描述bug以及复现过程，最好有截图)

Your hardware and system info
Write your system info like CUDA version/system/GPU/torch version here(在这里给出硬件信息和系统信息，如CUDA版本，系统，GPU型号和torch版本等)

Additional context
Add any other context about the problem here(在这里补充其他信息)

[tastelikefeet]

Can you share me give a detail description(like startup shell, logs) please?
We cannot reproduce this problem.