Repository navigation
Conversation
…emory, timestamps, max_pixels) + recipe updates (vllm-project#4623) Signed-off-by: lishunyang12 <125541396+lishunyang12@users.noreply.github.com>
Signed-off-by: Maciej Bala <mbala@nvidia.com> Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
…llm-project#1673) Signed-off-by: rongfu.leng <lenronfu@gmail.com> Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>
Signed-off-by: amy-why-3459 <wuhaiyan17@huawei.com>
… Registry Eager (vllm-project#3760) Signed-off-by: Alex Brooks <albrooks@redhat.com>
Signed-off-by: TaffyOfficial <2324465096@qq.com> Co-authored-by: TaffyOfficial <2324465096@qq.com> Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>
Signed-off-by: Mr.Dong <1042542469@qq.com> Signed-off-by: Yueqian Lin <linyueqian@outlook.com> Co-authored-by: Mr.Dong <1042542469@qq.com> Co-authored-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com> Co-authored-by: Yueqian Lin <linyueqian@outlook.com>
…llm-project#4276) Signed-off-by: Chendi Xue <chendi.xue@intel.com>
Signed-off-by: Yuanheng <jonathan.zhaoyh@gmail.com>
…ce mode (vllm-project#4650) Signed-off-by: R2-Y <ruiruyang2@gmail.com>
…ras (vllm-project#4648) Signed-off-by: ForestWisdom <zszhang01@gmail.com>
vllm-project#4148) Signed-off-by: zengchuang <zengchuang3@huawei.com> Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
…SRF (vllm-project#2565) Signed-off-by: Nick Cao <ncao@redhat.com> Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
…m-project#4701) Signed-off-by: wangyu <410167048@qq.com>
Signed-off-by: bjf-frz <frz123db@gmail.com>
…t#4667) Signed-off-by: Ianniu <985683179yzy@gmail.com> Signed-off-by: david6666666 <530634352@qq.com> Co-authored-by: Ianniu <985683179yzy@gmail.com>
…liability/stability test for VoxCPM2 (vllm-project#4600) Signed-off-by: zhumingjue <zhumingjue@huawei.com> Signed-off-by: [Your Name] [Your Email]
Signed-off-by: menjiantong <menjiantong@foxmail.com> Co-authored-by: menjiantong <menjiantong@foxmail.com> Co-authored-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
…-project#3956) Signed-off-by: Daniel Socek <daniel.socek@intel.com> Co-authored-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
Signed-off-by: amy-why-3459 <wuhaiyan17@huawei.com>
Signed-off-by: R2-Y <ruiruyang2@gmail.com>
Signed-off-by: wangyu <410167048@qq.com>
Signed-off-by: gcanlin <canlinguosdu@gmail.com>
Signed-off-by: Acerak01-fy <wfy2003324@163.com> Signed-off-by: zwhzzz0821 <2831474076@qq.com> Signed-off-by: Wang fuyin <wfy2003324@163.com> Co-authored-by: zwhzzz0821 <2831474076@qq.com> Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
…device (vllm-project#4712) Signed-off-by: Weiming Liao <liaowm5@gmail.com>
Signed-off-by: z00806815 <zhuhonghan@huawei.com> Signed-off-by: zhuhh97 <zhuhonghan@huawei.com> Signed-off-by: dph97 <nenbaying@163.com> Signed-off-by: zhh <zhh@users.noreply.github.com> Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com> Co-authored-by: zhuhh97 <zhuhh97@users.noreply.github.com>
…#4341) Signed-off-by: LHXuuu <xulianhao.xlh@antgroup.com> Signed-off-by: Yuanheng Zhao <jonathan.zhaoyh@gmail.com> Signed-off-by: Hongsheng Liu <liuhongsheng4@huawei.com> Co-authored-by: Yuanheng Zhao <jonathan.zhaoyh@gmail.com> Co-authored-by: Yuanheng Zhao <54058983+yuanheng-zhao@users.noreply.github.com> Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
…llm-project#4192) Signed-off-by: Gael Glorian <gael.glorian@gmail.com> Signed-off-by: Ioannis Lamprou <yiannis.lamprou@gmail.com> Signed-off-by: Zhen Zhang <zhen.zhang.fr@huawei.com> Co-authored-by: Ioannis Lamprou <yiannis.lamprou@gmail.com> Co-authored-by: Zhen Zhang <zhen.zhang.fr@huawei.com>
…llm-project#3927) Signed-off-by: Mengkejiergeli Ba <mengkejiergeli.ba@intel.com> Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
…ageConfigFactory (vllm-project#4729) Signed-off-by: Yuanheng Zhao <jonathan.zhaoyh@gmail.com>
Two fixes: 1. Pass bake_targets="cuda cuda-develsdk" to build-image.yml so both image variants are produced in a single bake run. Previously only the cuda target was built, causing ImagePullBackOff for all develsdk smoke tests. 2. Replace createdAt-based polling with vllm_image input matching via the GitHub API. When 14 matrix jobs trigger test runs within the same second, the old sort_by(.createdAt)|first approach grabbed arbitrary runs — a develsdk job could latch onto a cuda run (or vice versa) and falsely report success. Fixes: INFERENG-8005 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Include the cuda-test image variant in full-chain builds so upstream pytest tests can run against the built image. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Use the new omni validation type (nm-cicd PR vllm-project#576) which tests multi-modal input/output: text→text, text→text+audio, image→text, and streaming — instead of the basic single-request chat validation.
The H100 dind runner pool is broken (XFS inode overflow on the only active node). Use k8s-a100-util instead, which supports image builds via the Kubernetes buildx driver (nm-cicd PR vllm-project#537). The dind label can still be used via the build_label_image input override if needed. Ref: INFERENG-8339 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Focus on testing models of interest Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
…m related artifacts together Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
…el of interest Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
Signed-off-by: Clodagh Walsh <clodaghwalsh17@gmail.com>
…turning empty choices Cherry-pick of upstream vllm-project/vllm-omni PRs vllm-project#4718 and vllm-project#4720, adapted for nm-vllm-omni-ent. - Wire up request audio.format to existing encoding infrastructure instead of hardcoding WAV. Validate format against supported set. - Remove AAC from supported formats (unsupported by bundled libsndfile). - Validate requested output modalities against engine capabilities upfront, returning 400 for unsupported modalities instead of silently returning empty choices with HTTP 200. - Validate that modalities is a list of strings. Closes vllm-project#4716 Closes vllm-project#4719 Signed-off-by: Ricardo Noriega De Soto <rnoriega@redhat.com>
…tions Add a field_validator for the prompt field on ImageGenerationRequest that rejects empty and whitespace-only strings with a 400 error. Previously, empty strings passed Pydantic validation and reached the diffusion engine, which would generate an image from nothing instead of returning an error. Unskip the prompt_empty and prompt_whitespace_only test cases that were blocked on this fix. Partial fix for vllm-project#3649 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…ors for /v1/images/generations Addresses additional validation gaps from issue vllm-project#3649: - Add seed range constraint (0 to 2^32-1) - Add output_format validator restricting to png/jpeg/webp - Add negative_prompt/system_prompt empty string rejection (Nick's review feedback) - Unskip seed_negative, seed_above_uint32, output_format_invalid tests Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…/wakeup requests Reject empty stage_ids lists (min_length=1) and negative sleep levels (ge=0) at the entrypoint level. Unskip sleep_empty_stage_ids, sleep_negative_level, and wakeup_empty_stage_ids tests. Partial fix for vllm-project#3649. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: Shaun Walsh <shaunwalsh24@gmail.com>
…erence_steps validators for /v1/audio/generate - Reject empty/whitespace input prompts - Constrain audio_length to gt=0 - Constrain guidance_scale to gt=0, le=1000 - Constrain num_inference_steps to ge=1, le=1000 - Unskip all 8 issue vllm-project#3649 tests and update error expectations to match entrypoint-level Pydantic validation Partial fix for vllm-project#3649. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: Shaun Walsh <shaunwalsh24@gmail.com>
…completions ChatCompletionRequest is imported from upstream vllm and cannot have Pydantic validators added directly. Validate at the handler level by checking the raw JSON body before dispatch: - Reject modalities containing non-string elements (e.g. [123]) - Reject logprobs with non-boolean values (Pydantic v2 coerces "yes" to True instead of returning 400) Unskip modalities_list_bad_element and logprobs_wrong_type tests. Partial fix for vllm-project#3649. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: Shaun Walsh <shaunwalsh24@gmail.com>
vllm-project#3649) Add missing request validators for /v1/audio/speech, /v1/audio/speech/batch, and /v1/audio/voices endpoints that previously allowed invalid inputs to reach the engine, causing garbage output, 500s, or subprocess crashes. Pydantic-level validators (protocol/audio.py): - OpenAICreateSpeechRequest: reject empty/whitespace voice - SpeechBatchItem: reject empty input, empty voice, overlong instructions, out-of-range max_new_tokens, invalid ref_audio URI scheme - BatchSpeechRequest: same field validators as SpeechBatchItem Serving-level validators (serving_speech.py): - _pre_validate_batch: validate voice against supported speakers (CustomVoice task only), empty/unknown language, whitespace-only ref_text for Base task — raises ValueError for HTTP 400 instead of embedding errors in per-item results - upload_voice: validate consent (empty, path separators, length), ref_text length, speaker_description length Unskips 26 test cases in test_invalid_audio_speech.py. Upstream: vllm-project#4749 Signed-off-by: Shaun Walsh <shawalsh@redhat.com> Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: Shaun Walsh <shaunwalsh24@gmail.com>
Sync with upstream PR vllm-project#4720 review feedback: single-stage diffusion engines that advertise output_modalities=["image"] must also accept modalities=["text"] for img2text/text2text. Add regression tests for modality validation. Signed-off-by: Ricardo Noriega De Soto <rnoriega@redhat.com>
The upstream TTS commit removed the definition since upstream has no remaining references. The nm fork has 10 additional voice upload test params that still use this skip marker. Restores the constant to fix Ruff F821. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: Shaun Walsh <shaunwalsh24@gmail.com>
ChatCompletionRequest moved to vllm.entrypoints.openai.chat_completion.protocol ErrorResponse moved to vllm.entrypoints.openai.engine.protocol Signed-off-by: Ricardo Noriega De Soto <rnoriega@redhat.com>
…t completions Signed-off-by: vraiti <vraiti@redhat.com>
Signed-off-by: vraiti <vraiti@redhat.com>
Signed-off-by: vraiti <vraiti@redhat.com>
Signed-off-by: vraiti <vraiti@redhat.com>
oglok
force-pushed
the
rebase/upstream-sync-2026-07-20
branch
from
July 21, 2026 09:20
6d5828a to
b5a5d8c
Compare
dougbtv
approved these changes
Jul 21, 2026
dougbtv
left a comment
dougbtv
left a comment
Collaborator
There was a problem hiding this comment.
Reviewed the midstream-specific files (Dockerfiles, midstream-build.yml, docker-bake.hcl, carry commits). CUDA build path looks clean — workflow chains correctly, bake config is consistent, dropped carries are well-justified, and the 55 preserved carries are all present.
Found a few issues in the ROCm/CPU Dockerfiles (missing predownload_tiktoken_tokenizers.py, wrong package names, glob mismatches) but those aren't on the critical path — ROCm/CPU images aren't actively built yet. Tracked separately.
Post-merge action items filed in INFERENG-9317: verify v0.25.1 wheel, bump FlashInfer pins, add fa3-fwd/av constraints in nm-cicd, full smoke matrix run.
LGTM for merge.
Author
|
Superseded by sync-based PR (merge strategy instead of force-push) |
4 tasks
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
20 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.
Summary
Rebases nm-vllm-omni-ent onto the latest upstream/main (
e24bddfb, 283 commits ahead of our fork pointb1a4b08afrom 2026-06-15).What changed
29766140891fromsync-v0.25.1branch.endpoint rejection (#4762)— merged Jul 8hack for single stage q3omni (PR #3760)— merged Jun 24fix HTTP status codes for audio voice (#3969)— merged Jun 22Centralize audio format constants— fully covered by upstream PR [Bugfix] Fix audio.format parameter silently ignored in chat completions vllm-project/vllm-omni#4718Switch endpoint restrict fail to warn— upstream refactoredStageConfigFactoryintoconfig_factory.pywith gracefulNonehandlingRemove unregister_vllm_metrics monkey-patch warning log— upstream removed the entire monkey-patch (PR Remove unregister_vllm_metrics no-op patch vllm-project/vllm-omni#4396)d1c3e9a1(audio.format + modality): kept modality validation, dropped redundant audio format code (upstream handles via_resolve_audio_format)25c01b59(image validators): used upstream's updatedlayersrange (2,10)87318a70(audio format constants): skipped — empty after resolution0d426169(test helpers): resolved inlined helpers removal against upstream's updated versionDeferred
3 carries from the
v0.21.0branch (lost during v0.23.0 sync) are deferred to a follow-up PR:Open upstream PRs still carried
Test plan
python -c "import vllm_omni"workspytest tests/entrypoints/openai_api/test_audio_format.py tests/entrypoints/openai_api/test_modality_validation.py tests/entrypoints/test_stream_finish_reason.py