fix optimizer reset for relora sft #1414
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
winglian
force-pushed
the
relora-fix-optimizer-reset
branch
from
571cdd4 to
c318c76
Compare
winglian
force-pushed
the
relora-fix-optimizer-reset
branch
from
c318c76 to
490ecba
Compare
|
hmm, still doesn't seem to be correct, also, ooms after the first relora restart |
|
reference implementation for Relora is here https://github.com/Guitaricet/relora/blob/main/peft_pretraining/training_utils.py#L267-L364 |
winglian
force-pushed
the
relora-fix-optimizer-reset
branch
from
490ecba to
e0b26f0
Compare
winglian
force-pushed
the
relora-fix-optimizer-reset
branch
from
04121f5 to
cc438c4
Compare
winglian
force-pushed
the
relora-fix-optimizer-reset
branch
from
cc438c4 to
13b5865
Compare
bursteratom
pushed a commit
that referenced
this pull request
Dec 4, 2024
* fix optimizer reset * set states to reset for 8bit optimizers and handle quantile runtime error for embeddings * fix relora test to check grad_norm * use flash attn for relora and tweak hyperparams for test * fix messages field for test dataset
bursteratom
pushed a commit
that referenced
this pull request
Dec 4, 2024
* fix optimizer reset * set states to reset for 8bit optimizers and handle quantile runtime error for embeddings * fix relora test to check grad_norm * use flash attn for relora and tweak hyperparams for test * fix messages field for test dataset
3 tasks
djsaunde
pushed a commit
that referenced
this pull request
Dec 16, 2024
* fix optimizer reset * set states to reset for 8bit optimizers and handle quantile runtime error for embeddings * fix relora test to check grad_norm * use flash attn for relora and tweak hyperparams for test * fix messages field for test dataset
djsaunde
pushed a commit
that referenced
this pull request
Dec 17, 2024
* fix optimizer reset * set states to reset for 8bit optimizers and handle quantile runtime error for embeddings * fix relora test to check grad_norm * use flash attn for relora and tweak hyperparams for test * fix messages field for test dataset
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.
the relora optimizer reset wasn't pruning anything, this fixes that.
what seemed to be wrong with the original implementation is that iterating over the optimizer states isn't sufficient to determine the optimizer keys.