DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
33.6k
stars
3.94k
forks
source link
[XPU] adapt lazy_call func to different versions #5670
Closed
YizhouZ closed 1 week ago
Previously, lay_call function was wrapped by torch.xpu.lay_init._lazy_call, which is now changed to torch.xpu._lazy_call.
Thus we change this function to adapt different versions.