[Bugfix] Fix GPTQ and GPTQ Marlin CPU Offloading #7225
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
|
👋 Hi! Thank you for contributing to the vLLM project. Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge). To run full CI, you can do one of these:
🚀 |
mgoin
changed the title
sfc-gh-mkeralapura
pushed a commit
to sfc-gh-mkeralapura/vllm
that referenced
this pull request
Aug 12, 2024
kylesayrs
pushed a commit
to neuralmagic/vllm
that referenced
this pull request
Aug 17, 2024
fialhocoelho
pushed a commit
to opendatahub-io/vllm
that referenced
this pull request
Aug 22, 2024
Alvant
pushed a commit
to compressa-ai/vllm
that referenced
this pull request
Oct 26, 2024
Signed-off-by: Alvant <alvasian@yandex.ru>
KuntaiDu
pushed a commit
to KuntaiDu/vllm
that referenced
this pull request
Nov 20, 2024
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Outdated suggestions cannot be applied.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.
FIX #7204
GPTQ Marlin was failing with cpu offloading enabled during the initial model creation because
qzeroswas specifically made as a"meta"device parameter. There was no need for this so the simplest method was just to remove it. Another possible fix would be to make cpu offload not touch meta device tensors, but it seems meta isn't used anywhere else at the moment.Upon further testing, GPTQ would produce gibberish with cpu offloading because it shuffled the weights on first forward pass. This was fixed by moving that final preprocessing to
process_weights_after_loading