move tensor if necessary (like device_map: auto with >1 gpu)
#2
by
lodrick-the-lafted
- opened
No description provided.
RangiLyu
changed pull request status to
merged