Skip to content

fix for vllm #30623 and #27814#1494

Merged
gpolovets1 merged 7 commits intomainfrom
gpolovets/fix_vllm_30623
Jan 22, 2026
Merged

fix for vllm #30623 and #27814#1494
gpolovets1 merged 7 commits intomainfrom
gpolovets/fix_vllm_30623

Conversation

@gpolovets1
Copy link
Collaborator

@gpolovets1 gpolovets1 commented Jan 20, 2026

Description

Updated imports to use reorganized FusedMoE paths made by vllm-project/vllm#30623.
Overrode initialization logic of fp8 linear layer to workaround refactor in vllm-project/vllm#27814.

Collaborators are kyuyeunk@, lk-chen@, JiriesKaileh@

Tests

Pre-submit build kite passes.

Checklist

Before submitting this PR, please make sure:

  • I have performed a self-review of my code.
  • I have necessary comments in my code, particularly in hard-to-understand areas.
  • I have made or will make corresponding changes to any relevant documentation.

@github-actions
Copy link

Description

Start with a short description of what the PR does and how this is a change from
the past.

The rest of the description includes relevant details and context, examples:

  • why is this change being made,
  • the problem being solved and any relevant context,
  • why this is a good solution,
  • some information about the specific implementation,
  • shortcomings of the solution and possible future improvements.

If the change fixes a Github issue, please include a link, e.g.,:
FIXES: #123456

Tests

Please describe how you tested this change, and include any instructions and/or
commands to reproduce.

Checklist

Before submitting this PR, please make sure:

  • I have performed a self-review of my code.
  • I have necessary comments in my code, particularly in hard-to-understand areas.
  • I have made or will make corresponding changes to any relevant documentation.

@gpolovets1 gpolovets1 added the ready ONLY add when PR is ready to merge/full CI is needed label Jan 21, 2026
@gpolovets1 gpolovets1 changed the title fix for vllm #30623 refactor fix for vllm #30623 and #27814 Jan 21, 2026
Signed-off-by: Kyuyeun Kim <kyuyeunk@google.com>
Co-Authored-By: Kyuyeun Kim <kyuyeunk@google.com>
@gpolovets1 gpolovets1 force-pushed the gpolovets/fix_vllm_30623 branch from 019bf03 to 37a9d15 Compare January 21, 2026 04:16
gpolovets1 and others added 2 commits January 21, 2026 05:14
Co-Authored-By: Kyuyeun Kim <kyuyeunk@google.com>
@gpolovets1 gpolovets1 force-pushed the gpolovets/fix_vllm_30623 branch from 0860c15 to f9a1dc9 Compare January 21, 2026 18:57
@gpolovets1 gpolovets1 merged commit 5742e6f into main Jan 22, 2026
15 of 28 checks passed
@kyuyeunk kyuyeunk deleted the gpolovets/fix_vllm_30623 branch January 30, 2026 09:37
lorriexingfang pushed a commit to lorriexingfang/tpu-inference that referenced this pull request Feb 26, 2026
Signed-off-by: Kyuyeun Kim <kyuyeunk@google.com>
Signed-off-by: JiriesKaileh <jiries@google.com>
Co-authored-by: Kyuyeun Kim <kyuyeunk@google.com>
Co-authored-by: JiriesKaileh <jiries@google.com>
Signed-off-by: Xing Fang <xing.fang@anyscale.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ready ONLY add when PR is ready to merge/full CI is needed

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants