DeepSpeed
Fixed the issue that universal checkpoint cannot be loaded for stage3 when world size expansion.
#7599

Merged

Fixed the issue that universal checkpoint cannot be loaded for stage3 when world size expansion. #7599

sfc-gh-truwase merged 5 commits into deepspeedai:master from zhengchenyu:fix.load.universal

Fixed the issue that universal checkpoint cannot be loaded for stage3…

1b29e297

zhengchenyu requested a review from

tjruwase 84 days ago

zhengchenyu requested a review from

tohtana 84 days ago

zhengchenyu marked this pull request as draft 84 days ago

zhengchenyu marked this pull request as ready for review 84 days ago

zhengchenyu marked this pull request as draft 84 days ago

fix load checkpoint when disable univeral

707a1d08

zhengchenyu marked this pull request as ready for review 83 days ago

Merge branch 'master' into fix.load.universal

21f448c6

sfc-gh-truwase commented on 2025-09-30

Merge branch 'master' into fix.load.universal

97a05869

sfc-gh-truwase enabled auto-merge (squash) 81 days ago

Merge branch 'master' into fix.load.universal

1129787a

sfc-gh-truwase approved these changes on 2025-10-01

sfc-gh-truwase merged 07e76bd4 into master 81 days ago

zhengchenyu deleted the fix.load.universal branch 81 days ago

Reviewers

sfc-gh-truwase

tjruwase

tohtana

Assignees

No one assigned

Labels

None yet

Milestone

No milestone