Remove the "extra_state" for TE DPA module #14
Merged
Conversation
This file contains 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
wangye805
force-pushed
the
dist_ckpt_dpa_fix
branch
from
871c161 to
e727c56
Compare
wangye805
force-pushed
the
dist_ckpt_dpa_fix
branch
from
e727c56 to
da46f7a
Compare
wangye805
force-pushed
the
dist_ckpt_dpa_fix
branch
2 times, most recently
from
d21a895 to
746e080
Compare
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Outdated suggestions cannot be applied.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.
Remove the "extra_state" from the shared_state_dict of TEDotProductAttention module which was added from NVTE (ver >=1.9) , so that the dist checkpointing can ignore the weights for this extra_state for some earlier ckpts
Passed the manual pytest:
torchrun --nproc_per_node=8 -m pytest --color=yes -m "not flaky and not internal and not failing_on_rocm" --csv output/test_report.csv tests/unit_tests/