[CI] Remove small resolution test in Qwen-Image Perf test when vae patch parallel is enabled#2872
Merged
Merged
Conversation
Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.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. |
wtomin
requested review from
Gaohan123,
david6666666,
gcanlin and
hsliuustc0106
Collaborator
|
The rationale is sound: small resolutions with VAE patch parallel are indeed unstable. Alternative approaches considered:
Removing the test is the pragmatic solution here. |
Collaborator
|
Reasoning makes sense - small resolution + high parallelism = communication overhead dominates. One suggestion: add a comment in the test config explaining why 512x512 is skipped for this parallelization config. Future reviewers may wonder why this resolution is missing otherwise. Example: |
Gaohan123
added
ready
lvliang-intel
pushed a commit
to lvliang-intel/vllm-omni
that referenced
this pull request
Apr 21, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
nainiu258
pushed a commit
to nainiu258/vllm-omni
that referenced
this pull request
Apr 21, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com> Signed-off-by: nainiu258 <cperfect02@163.com>
qinganrice
pushed a commit
to qinganrice/vllm-omni
that referenced
this pull request
Apr 23, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
1 task
lengrongfu
pushed a commit
to lengrongfu/vllm-omni
that referenced
this pull request
May 1, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
clodaghwalsh17
pushed a commit
to clodaghwalsh17/nm-vllm-omni-ent
that referenced
this pull request
May 12, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
daixinning
pushed a commit
to daixinning/vllm-omni
that referenced
this pull request
May 28, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
quyifei23
pushed a commit
to quyifei23/vllm-omni
that referenced
this pull request
Jun 6, 2026
…tch parallel is enabled (vllm-project#2872) Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.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
5 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.
PLEASE FILL IN THE PR DESCRIPTION HERE ENSURING ALL CHECKLIST ITEMS (AT THE BOTTOM) HAVE BEEN CONSIDERED.
Purpose
Solving #2863.
Vae patch parallelism is enabled to reduce the peak memory of long sequence decoding. It may increase the latency when the resolution is too small, because the communication overhead is larger than the parallel computation gains.
Therefore, this PR removes the "512x512_steps20" test in the
Ulysses SP=2 + CFG-parallel=2 + VAE Patch Parallel=4test case. This actually helps to stablize the nightly CI performance, preventing it from failure.Previous CI results
The table above indicates that larger resolution yields more stable performance results.
Test Plan
Test Result
Wait for Nightly-CI results.
cc @yenuo26 @Gaohan123 @hsliuustc0106
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model. Please runmkdocs serveto sync the documentation editions to./docs.BEFORE SUBMITTING, PLEASE READ https://github.com/vllm-project/vllm-omni/blob/main/CONTRIBUTING.md (anything written below this line will be removed by GitHub Actions)