DeepSpeed
Default gradient_clipping to 1.0
#8068
Merged

Default gradient_clipping to 1.0 #8068

sfc-gh-truwase merged 4 commits into master from grad-clip-default-1.0
sfc-gh-truwase
sfc-gh-truwase sfc-gh-truwase requested a review from tjruwase tjruwase 20 days ago
sfc-gh-truwase sfc-gh-truwase requested a review from tohtana tohtana 20 days ago
tohtana
tohtana requested changes on 2026-06-23
sfc-gh-truwase sfc-gh-truwase requested a review from loadams loadams 12 days ago
sfc-gh-truwase Default gradient_clipping to 1.0
ded228eb
sfc-gh-truwase Fix docs
7fcc7f9f
sfc-gh-truwase Add tests for default gradient_clipping of 1.0
5df79365
sfc-gh-truwase Improve UT
3cd7246b
sfc-gh-truwase sfc-gh-truwase force pushed from 53ecde5e to 3cd7246b 12 days ago
PKUWZP PKUWZP requested a review from PKUWZP PKUWZP 12 days ago
PKUWZP
sfc-gh-truwase
tohtana
tohtana approved these changes on 2026-06-23
sfc-gh-truwase sfc-gh-truwase merged 44216655 into master 12 days ago
sfc-gh-truwase sfc-gh-truwase deleted the grad-clip-default-1.0 branch 12 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone