Handle set_to_none when using DeepSpeed optimizer in Lite
#16275
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
for more information, see https://pre-commit.ci
… lite/deepspeed-set-to-none
for more information, see https://pre-commit.ci
awaelchli
changed the title
set_to_none when using DeepSpeed in Liteset_to_none when using DeepSpeed in Lite
for more information, see https://pre-commit.ci
awaelchli
added
the
fun
… lite/deepspeed-set-to-none
for more information, see https://pre-commit.ci
… lite/deepspeed-set-to-none
awaelchli
added
feature
awaelchli
changed the title
set_to_none when using DeepSpeed in Liteset_to_none when using DeepSpeed optimizer in Lite
Borda
approved these changes
Jan 8, 2023
carmocca
approved these changes
Jan 9, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
The deepspeed optimizer does not accept the argument
optimizer.zero_grad(set_to_none=...). In their optimizer, the name isoptimizer.zero_grad(set_grads_to_None=...)🤣 We are translating this inconsistency for the user.Perhaps DeepSpeed should change their name to be consistent with PyTorch: deepspeedai/DeepSpeed#2678
cc @Borda @carmocca @justusschock @awaelchli
Before submitting
PR review
Anyone in the community is free to review the PR once the tests have passed.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:
Did you have fun?
I made sure I had fun coding 🙃