DeepSpeed
Fix zero stage2 cpu_offload when some model trainable parameters skipped in training
#861
Merged

Fix zero stage2 cpu_offload when some model trainable parameters skipped in training #861

ghosthamlet
ghosthamlet Merge pull request #1 from microsoft/master
517357e7
ghosthamlet Fix zero stage2 cpu_offload when some model trainable parameters skip…
d8f1dcd3
ghosthamlet ghosthamlet requested a review from arashashari arashashari 4 years ago
ghosthamlet ghosthamlet requested a review from awan-10 awan-10 4 years ago
ghosthamlet ghosthamlet requested a review from cli99 cli99 4 years ago
ghosthamlet ghosthamlet requested a review from conglongli conglongli 4 years ago
ghosthamlet ghosthamlet requested a review from eltonzheng eltonzheng 4 years ago
ghosthamlet ghosthamlet requested a review from jeffra jeffra 4 years ago
ghosthamlet ghosthamlet requested a review from minjiaz minjiaz 4 years ago
ghosthamlet ghosthamlet requested a review from niumanar niumanar 4 years ago
ghosthamlet ghosthamlet requested a review from RezaYazdaniAminabadi RezaYazdaniAminabadi 4 years ago
ghosthamlet ghosthamlet requested a review from samyam samyam 4 years ago
ghosthamlet ghosthamlet requested a review from ShadenSmith ShadenSmith 4 years ago
ghosthamlet ghosthamlet requested a review from tjruwase tjruwase 4 years ago
ghosthamlet Trim space
da595c2a
ghosthamlet Trim space
e6a46c37
tjruwase
tjruwase approved these changes on 2021-03-27
tjruwase Merge branch 'master' into ghosthamlet-stage2-bug
190d250f
tjruwase tjruwase merged 7fcc8911 into master 4 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone