Set use_cache to True, otherwise inference performance is poor
#2
by
TheBloke
- opened
No description provided.
winglian
changed pull request status to
merged