-
-
Notifications
You must be signed in to change notification settings - Fork 9.4k
New moe quant config #22537
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
New moe quant config #22537
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code Review
This pull request introduces a significant refactoring of the Mixture of Experts (MoE) quantization configuration by introducing a new FusedMoEQuantConfig
structure. This is a positive change towards a more structured and extensible configuration. However, the refactoring appears to be incomplete, as there are several critical issues, including assert False
statements, NotImplementedError
s, and usage of undefined variables in the new code paths. These issues will cause runtime failures and need to be addressed before this PR can be considered for merging. My review focuses on these critical issues.
vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_moe.py
Show resolved
Hide resolved
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add 🚀 |
This pull request has merge conflicts that must be resolved before it can be |
27a4513
to
688374b
Compare
This pull request has merge conflicts that must be resolved before it can be |
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Bill Nell <bnell@redhat.com>
35f5c34
to
81de521
Compare
Purpose
FusedMoEQuantConfig
objects to the subclass ofFusedMoEMethodBase
that will use that info.FusedMoEQuantConfig
and make it more uniform.fused_experts
functions by putting all/most quantization information intoFusedMoEQuantConfig
. This eliminates the varioususe_
bool flags and various quantization scales_scales
,_zp
,_bias
,_gscale
, etc. as arguments to the variousfused_experts
.Note: this is branched from #22035 so contains the same changes. I'll rebase once #22035 lands.
Test Plan
Test Result
(Optional) Documentation Update
cc @varun-sundar-rabindranath , @LucasWilkinson , @jeejeelee , @wenscarl , @nvpohanh