Add support for Marlin 2:4 sparsity #2102
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
Narsil
reviewed
Jun 24, 2024
| @@ -19,6 +19,23 @@ def gptq_marlin_gemm( | |||
| """ | |||
| ... | |||
|
|
|||
| def gptq_marlin_24_gemm( | |||
There was a problem hiding this comment.
Are we running a fork of marlin ?
Shouldn't we use the classic makefile + commit approach instead (or already made releases if possible) ?
There was a problem hiding this comment.
These come from VLLM. We could also directly import them from VLLM if we prefer.
Narsil
previously approved these changes
Jun 24, 2024
Comment on lines
+217
to
+221
| GPTQ_MARLIN_24_MIN_THREAD_N = 128 | ||
| GPTQ_MARLIN_24_MIN_THREAD_K = 128 | ||
| GPTQ_MARLIN_24_MAX_PARALLEL = 64 | ||
| GPTQ_MARLIN_24_SUPPORTED_NUM_BITS = [4, 8] | ||
| GPTQ_MARLIN_24_SUPPORTED_GROUP_SIZES = [-1, 128] |
There was a problem hiding this comment.
This is coming from marlin directly I guess ?
There was a problem hiding this comment.
Yes, they come from upstream.
This change adds support for 2:4 sparsity when using Marlin quantization. The 2:4 kernel is used when: * The quantizer is `marlin`; * the quantizer checkpoint format is `marlin_24`. Fixes #2098.
danieldk
force-pushed
the
feature/marlin-24
branch
from
8f541bc to
b06dda9
Compare
glegendre01
pushed a commit
that referenced
this pull request
Jul 2, 2024
This change adds support for 2:4 sparsity when using Marlin quantization. The 2:4 kernel is used when: * The quantizer is `marlin`; * the quantizer checkpoint format is `marlin_24`. Fixes #2098.
yuanwu2017
pushed a commit
to yuanwu2017/tgi-gaudi
that referenced
this pull request
Sep 26, 2024
This change adds support for 2:4 sparsity when using Marlin quantization. The 2:4 kernel is used when: * The quantizer is `marlin`; * the quantizer checkpoint format is `marlin_24`. Fixes huggingface#2098.
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?
This change adds support for 2:4 sparsity when using Marlin quantization. The 2:4 kernel is used when:
marlin;marlin_24.Fixes #2098.
Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.