PR #28364 Fix for checkpoint rename race condition

Fix for checkpoint rename race condition #28364

ArthurZucker merged 5 commits into huggingface:main from tblattner:fix-checkpoint-rename-race-condition

muellerzr commented on 2024-01-08

siddartha-RE commented on 2024-01-08

Changed logic for renaming staging directory when saving checkpoint t…

728231ce

Updated styling using make fixup

f2452fda

Updated check for main process to use built-in versions from trainer

3c1c92fa

Fixed incorrect usage of trainer main process checks

df9f6e9a

Removed "with open" due to not working with directory. os.open seems …

eb586987

tblattner force pushed to eb586987 2 years ago

muellerzr approved these changes on 2024-01-09

muellerzr requested a review from

ArthurZucker 2 years ago

ArthurZucker approved these changes on 2024-01-10

ArthurZucker merged cef2e40e into main 2 years ago

Reviewers

ArthurZucker

muellerzr

siddartha-RE

Assignees

No one assigned

Labels

None yet

Milestone

No milestone