-
Notifications
You must be signed in to change notification settings - Fork 28.6k
add DeepSpeed tensor parallel initialization. #36114
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, cc @ArthurZucker
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
thanks for the review! @muellerzr |
thanks for the review. @muellerzr! The related PR has already been merged: huggingface/accelerate#3390. Is this PR ready for merge? Thanks! |
sorry to bother you again and I hope this doesn't disturb you : ) the doc deepspeedai/DeepSpeed#7151 is ready but pending with this pr, I would appreciate it if this PR can be merged. Thank you very much! |
d3c2d1c
to
24f1be4
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM ! Is the CI stalling @muellerzr ?
Not quite sure, it's been hanging on Checking for the ability to merge automatically... for several hours. move to #36825 fix this hang issue. |
Hi 👋, thank you for opening this pull request! The pull request is converted to draft by default. When it is ready for review, please click the |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Happy to merge let's make sure CIs work
@inkcherry I think we can close this one in favor of #36825 yes? |
yes, still hang now, let me close this one. |
Initialization for DeepSpeed tensor parallel+zero optimizer, which needs to be done before the optimizer initializes the model weights. The related APIs are already supported in DeepSpeed. deepspeedai/DeepSpeed#6922 @muellerzr
cc DeepSpeed team: @tjruwase
move to #36825