[Bugfix] Truncate mimo-audio code2wav prompt to MAX_CODE2WAV_TOKENS#2693
Merged
lishunyang12 merged 5 commits intoApr 20, 2026
Merged
Conversation
Fixes vllm-project#2683 Signed-off-by: lishunyang <lishunyang12@163.com>
|
Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits. |
1 task
Collaborator
Author
|
@qibaoyuan PTAL |
Contributor
|
LGTM |
lishunyang12
commented
Apr 16, 2026
…s captured Signed-off-by: lishunyang <lishunyang12@163.com>
Signed-off-by: lishunyang <lishunyang12@163.com>
lvliang-intel
pushed a commit
to lvliang-intel/vllm-omni
that referenced
this pull request
Apr 21, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com>
nainiu258
pushed a commit
to nainiu258/vllm-omni
that referenced
this pull request
Apr 21, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com> Signed-off-by: nainiu258 <cperfect02@163.com>
qinganrice
pushed a commit
to qinganrice/vllm-omni
that referenced
this pull request
Apr 23, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com>
lengrongfu
pushed a commit
to lengrongfu/vllm-omni
that referenced
this pull request
May 1, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com>
clodaghwalsh17
pushed a commit
to clodaghwalsh17/nm-vllm-omni-ent
that referenced
this pull request
May 12, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com>
daixinning
pushed a commit
to daixinning/vllm-omni
that referenced
this pull request
May 28, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com>
quyifei23
pushed a commit
to quyifei23/vllm-omni
that referenced
this pull request
Jun 6, 2026
…llm-project#2693) Signed-off-by: lishunyang <lishunyang12@163.com>
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Purpose
Fixes #2683.
Stage-1 (
code2wav) crashes withwhenever the online
llm2code2wavinput processor flattens a long stage-0 talker output and hands it to stage-1. Each talker frame expands to 36 ids afterprepend_and_flatten_colmajor(pad row + col-major reshape of an (8, 4) codec block), so ~1546 frames already produce 55656 ids — well pastmimo_audio.yaml's stage-1max_model_len: 18192, and far past whatever lower cap the underlying code2wav model enforces.The offline path (
examples/offline_inference/mimo_audio/end2end.py) already guards against this exact failure withMAX_CODE2WAV_TOKENS = 18192+ a warn-and-truncate. The online stage input processor was missing the same safeguard, so any long-enough request killed the engine core.Test Result