Decompile depyf optional by vincentzed · Pull Request #1 · vincentzed/vllm

vincentzed · 2025-08-02T14:16:39Z

Essential Elements of an Effective PR Description Checklist

The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
The test plan, such as providing test command.
The test results, such as pasting the results comparison before and after, or e2e results
(Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.

Purpose

Test Plan

Test Result

(Optional) Documentation Update

Signed-off-by: Chengji Yao <chengjiyao@google.com>

…d quantization (vllm-project#20766) Signed-off-by: Alex Kogan <alex.kogan@oracle.com>

…17818) Signed-off-by: Farzad Abdolhosseini <farzad@fixie.ai> Signed-off-by: Patrick Li <patrick8289@gmail.com> Co-authored-by: Patrick Li <patrick8289@gmail.com>

…vllm-project#21530) Signed-off-by: David Chen <530634352@qq.com>

…m-project#21641) Signed-off-by: Huy Do <huydhn@gmail.com>

… to use dtype comparison (vllm-project#21612) Signed-off-by: Alexandre Juan <a.juan@netheos.net>

Signed-off-by: Roger Wang <hey@rogerw.me> Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn> Co-authored-by: Your Name <you@example.com> Co-authored-by: Roger Wang <hey@rogerw.me> Co-authored-by: Isotr0py <2037008807@qq.com> Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

…#21618) Signed-off-by: reidliu41 <reid201711@gmail.com>

…#21646) Signed-off-by: Huy Do <huydhn@gmail.com>

…lm-project#21620) Signed-off-by: Benji Beck <benjibeck@meta.com>

vllm-project#21622) Signed-off-by: Benji Beck <benjibeck@meta.com>

Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>

Signed-off-by: Max de Bayser <maxdebayser@gmail.com> Signed-off-by: Max de Bayser <mbayser@br.ibm.com> Co-authored-by: Russell Bryant <rbryant@redhat.com>

…llm-project#21634) Signed-off-by: yewentao256 <zhyanwentao@126.com>

Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>

…er (vllm-project#21428) Signed-off-by: David Chen <530634352@qq.com>

Signed-off-by: yewentao256 <zhyanwentao@126.com>

… Mac with Apple Silicon (vllm-project#21380) Signed-off-by: Yeju Zhou <yejuzhou@outlook.com>

…e to vllm bench CLI (vllm-project#21355) Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>

…e moe fp8 kernels (vllm-project#21411) Signed-off-by: kaixih <kaixih@nvidia.com>

…llm-project#21154) Signed-off-by: Wenchen Lo <charles761013@gmail.com>

…Schema (vllm-project#21656) Signed-off-by: Benji Beck <benjibeck@meta.com>

Signed-off-by: Benji Beck <benjibeck@meta.com>

Signed-off-by: Isotr0py <2037008807@qq.com>

…llm-project#21667) Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>

Signed-off-by: Xiongfei Wei <isaacwxf23@gmail.com>

Signed-off-by: Nick Hill <nhill@redhat.com>

Signed-off-by: mgoin <mgoin64@gmail.com>

…2036) Signed-off-by: yewentao256 <zhyanwentao@126.com>

…ect#21733)

Signed-off-by: kf <kuanfu.liu@embeddedllm.com> Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com> Co-authored-by: kf <kuanfu.liu@embeddedllm.com>

Signed-off-by: NickLucche <nlucches@redhat.com>

…roject#21955) Signed-off-by: Varun Sundar Rabindranath <vsundarr@redhat.com> Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>

…ject#21835) Signed-off-by: Dipika Sikka <dipikasikka1@gmail.com>

Signed-off-by: Nick Hill <nhill@redhat.com>

Signed-off-by: yewentao256 <zhyanwentao@126.com>

…ata (vllm-project#21153) Signed-off-by: Sage Moore <sage@neuralmagic.com>

…wen-VL models on ROCm platform. (vllm-project#22069) Signed-off-by: tjtanaavllm <tunjian.tan@amd.com> Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com> Co-authored-by: tjtanaavllm <tunjian.tan@amd.com>

…2040) Signed-off-by: Rui Qiao <ruisearch42@gmail.com>

Signed-off-by: Yong Hoon Shin <yhshin@meta.com>

…llm-project#22034) Signed-off-by: Chih-Chieh-Yang <7364402+cyang49@users.noreply.github.com>

Signed-off-by: Roger Wang <hey@rogerw.me>

…ist conversion (vllm-project#20000) Signed-off-by: Vadim Gimpelson <vadim.gimpelson@centml.ai>

Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: zRzRzRzRzRzRzR <2448370773@qq.com> Co-authored-by: Isotr0py <2037008807@qq.com>

…ad (vllm-project#21075) Signed-off-by: Chih-Chieh Yang <7364402+cyang49@users.noreply.github.com> Signed-off-by: Chih-Chieh-Yang <7364402+cyang49@users.noreply.github.com>

…22114) Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

…ering at init time (vllm-project#21557) Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>

github-actions · 2025-08-02T14:16:48Z

👋 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 fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

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 ready label to the PR or enable auto-merge.

🚀

…ation Introduce VLLM_COMPILE_DEPYF environment variable to allow users to toggle depyf decompilation during compilation. By default, a placeholder file is written unless VLLM_COMPILE_DEPYF=1 is set. This provides better control over when expensive decompilation operations are performed. Also ensures decompilation always occurs for cudagraph error checking, regardless of the env var setting, to prevent silent errors.

github-actions · 2025-11-01T02:51:00Z

This pull request has been automatically marked as stale because it has not had any activity within 90 days. It will be automatically closed if no further activity occurs within 30 days. Leave a comment if you feel this pull request should remain open. Thank you!

github-actions · 2025-12-01T03:19:38Z

This pull request has been automatically closed due to inactivity. Please feel free to reopen if you intend to continue working on it. Thank you!

Chengji Yao and others added 30 commits July 25, 2025 17:09

[TPU] Update ptxla nightly version to 20250724 (vllm-project#21555)

f1b286b

Signed-off-by: Chengji Yao <chengjiyao@google.com>

[Feature] Add support for MoE models in the calibration-free RTN-base…

7ae75fa

…d quantization (vllm-project#20766) Signed-off-by: Alex Kogan <alex.kogan@oracle.com>

[Model] Ultravox: Support Llama 4 and Gemma 3 backends (vllm-project#…

62965de

…17818) Signed-off-by: Farzad Abdolhosseini <farzad@fixie.ai> Signed-off-by: Patrick Li <patrick8289@gmail.com> Co-authored-by: Patrick Li <patrick8289@gmail.com>

[Docs] add offline serving multi-modal video input expamle Qwen2.5-VL (…

97349fe

…vllm-project#21530) Signed-off-by: David Chen <530634352@qq.com>

Correctly kill vLLM processes after finishing serving benchmarks (vll…

a55c950

…m-project#21641) Signed-off-by: Huy Do <huydhn@gmail.com>

[Bugfix] Fix isinstance check for tensor types in _load_prompt_embeds…

2f6e6b3

… to use dtype comparison (vllm-project#21612) Signed-off-by: Alexandre Juan <a.juan@netheos.net>

[TPU][Test] Divide TPU v1 Test into 2 parts. (vllm-project#21431)

7728dd7

[Misc] remove unused try-except in pooling config check (vllm-project…

05c1126

…#21618) Signed-off-by: reidliu41 <reid201711@gmail.com>

[Take 2] Correctly kill vLLM processes after benchmarks (vllm-project…

e98def4

…#21646) Signed-off-by: Huy Do <huydhn@gmail.com>

Migrate AriaImagePixelInputs to TensorSchema for shape validation (vl…

9d19728

…lm-project#21620) Signed-off-by: Benji Beck <benjibeck@meta.com>

Migrate AyaVisionImagePixelInputs to TensorSchema for shape validation (

de10ff0

vllm-project#21622) Signed-off-by: Benji Beck <benjibeck@meta.com>

[Bugfix] Investigate Qwen2-VL failing test (vllm-project#21527)

f27fdfc

Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>

Support encoder-only models without KV-Cache (vllm-project#21270)

1cd6eab

Signed-off-by: Max de Bayser <maxdebayser@gmail.com> Signed-off-by: Max de Bayser <mbayser@br.ibm.com> Co-authored-by: Russell Bryant <rbryant@redhat.com>

[Bug] Fix has_flashinfer_moe Import Error when it is not installed (v…

c215f5c

…llm-project#21634) Signed-off-by: yewentao256 <zhyanwentao@126.com>

[Misc] Improve memory profiling debug message (vllm-project#21429)

a40a850

Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>

[BugFix] Fix shared storage connector load kv only load attention lay…

97d6c30

…er (vllm-project#21428) Signed-off-by: David Chen <530634352@qq.com>

[Refactor] Remove moe_align_block_size_triton (vllm-project#21335)

56e544f

Signed-off-by: yewentao256 <zhyanwentao@126.com>

[Bugfix][Apple Silicon] fix missing symbols when build from source on…

9094d11

… Mac with Apple Silicon (vllm-project#21380) Signed-off-by: Yeju Zhou <yejuzhou@outlook.com>

[CI/Build][Doc] Move existing benchmark scripts in CI/document/exampl…

e7c4f9e

…e to vllm bench CLI (vllm-project#21355) Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>

[NVIDIA] Explicitly disable shuffled weights for flashinfer blockscal…

de509ae

…e moe fp8 kernels (vllm-project#21411) Signed-off-by: kaixih <kaixih@nvidia.com>

Remove xformers requirement for Mistral-format Pixtral and Mistral3 (v…

6c66f28

…llm-project#21154) Signed-off-by: Wenchen Lo <charles761013@gmail.com>

support torch.compile for bailing moe (vllm-project#21664)

c657369

Migrate Blip2ImagePixelInputs and Blip2ImageEmbeddingInputs to Tensor…

ccf27cc

…Schema (vllm-project#21656) Signed-off-by: Benji Beck <benjibeck@meta.com>

Migrate DeepseekVL2ImageInputs to TensorSchema (vllm-project#21658)

0b8caf9

Signed-off-by: Benji Beck <benjibeck@meta.com>

Migrate FuyuImagePatchInputs to TensorSchema (vllm-project#21662)

3339cba

Signed-off-by: Benji Beck <benjibeck@meta.com>

Migrate ChameleonImagePixelInputs to TensorSchema (vllm-project#21657)

20950b2

Signed-off-by: Benji Beck <benjibeck@meta.com>

[VLM] Support HF format Phi-4-MM model (vllm-project#17121)

eed2f46

Signed-off-by: Isotr0py <2037008807@qq.com>

Handle non-serializable objects in vllm bench (vllm-project#21665)

971948b

[CI/Build][Doc] Clean up more docs that point to old bench scripts (v…

01a395e

…llm-project#21667) Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>

vanbasten23 and others added 23 commits August 1, 2025 18:56

Add lora test for tp>1 case for TPU. (vllm-project#21970)

d84b97a

Signed-off-by: Xiongfei Wei <isaacwxf23@gmail.com>

[BugFix] Harden distributed DP startup (vllm-project#21538)

881e1af

Signed-off-by: Nick Hill <nhill@redhat.com>

[CI] Initial tests for SM100 Blackwell runner (vllm-project#21877)

88faa46

Signed-off-by: mgoin <mgoin64@gmail.com>

[Perf] Optimize reshape_and_cache_flash CUDA Kernel (vllm-project#2…

eefbf4a

…2036) Signed-off-by: yewentao256 <zhyanwentao@126.com>

feat: Add Support GPTQ Quantization MOE on ROCM vllm serve (vllm-proj…

3654847

…ect#21733)

[V1][CUDA] Full cudagraph support for FlashInfer (vllm-project#21367)

2332243

[Model] Qwen2.5 VL SiLU-and-Mul (vllm-project#22066)

ee2eb6e

Signed-off-by: kf <kuanfu.liu@embeddedllm.com> Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com> Co-authored-by: kf <kuanfu.liu@embeddedllm.com>

[Misc] VLLM_TARGET_DEVICE.lower() (vllm-project#22101)

5739371

Signed-off-by: NickLucche <nlucches@redhat.com>

[Misc] DeepGemmExperts : Avoid JIT generation in the hot-path (vllm-p…

a65f46b

…roject#21955) Signed-off-by: Varun Sundar Rabindranath <vsundarr@redhat.com> Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>

[Speculators][Speculative Decoding] Add Qwen Eagle3 Support (vllm-pro…

9f9c38c

…ject#21835) Signed-off-by: Dipika Sikka <dipikasikka1@gmail.com>

[BugFix] Improve internal DP load balancing (vllm-project#21617)

8d524ce

Signed-off-by: Nick Hill <nhill@redhat.com>

[Test] Add Unit Test for Batched DeepGEMM (vllm-project#21559)

6e8d8c4

Signed-off-by: yewentao256 <zhyanwentao@126.com>

[Attention][DBO] Add support for "splitting" the CommonAttentionMetad…

0edaf75

…ata (vllm-project#21153) Signed-off-by: Sage Moore <sage@neuralmagic.com>

[FEAT][ROCm] Enable running Flash Attention as ViT attn backend for Q…

d3a6f21

…wen-VL models on ROCm platform. (vllm-project#22069) Signed-off-by: tjtanaavllm <tunjian.tan@amd.com> Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com> Co-authored-by: tjtanaavllm <tunjian.tan@amd.com>

[Misc] Getting and passing ray runtime_env to workers (vllm-project#2…

4ac8437

…2040) Signed-off-by: Rui Qiao <ruisearch42@gmail.com>

Fix test_kv_sharing_fast_prefill flakiness (vllm-project#22038)

8564dc9

Signed-off-by: Yong Hoon Shin <yhshin@meta.com>

[Bugfix] Mamba2 remove bugged initial state condition in chunk scan (v…

c64861d

…llm-project#22034) Signed-off-by: Chih-Chieh-Yang <7364402+cyang49@users.noreply.github.com>

docs: remove deprecated disable-log-requests flag (vllm-project#22113)

067c34a

Signed-off-by: Roger Wang <hey@rogerw.me>

[PERF] Use faster way of decode in tokenizer: avoid useless list-to-l…

58eee5f

…ist conversion (vllm-project#20000) Signed-off-by: Vadim Gimpelson <vadim.gimpelson@centml.ai>

for glm-4.1V update (vllm-project#22000)

25373b6

Signed-off-by: Isotr0py <2037008807@qq.com> Signed-off-by: zRzRzRzRzRzRzR <2448370773@qq.com> Co-authored-by: Isotr0py <2037008807@qq.com>

[Model] Mamba2 preallocate SSM output tensor to avoid d2d copy overhe…

b690e34

…ad (vllm-project#21075) Signed-off-by: Chih-Chieh Yang <7364402+cyang49@users.noreply.github.com> Signed-off-by: Chih-Chieh-Yang <7364402+cyang49@users.noreply.github.com>

[Frontend] Improve error message for too many mm items (vllm-project#…

f5d0f47

…22114) Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

[V1] [Hybrid] Validate compatibility of attention backend batch reord…

4abfd87

…ering at init time (vllm-project#21557) Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>

vincentzed force-pushed the decompile_depyf_optional branch from 83272a1 to fa7cef7 Compare August 2, 2025 14:17

github-actions bot added the stale label Nov 1, 2025

github-actions bot closed this Dec 1, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Decompile depyf optional#1

Decompile depyf optional#1
vincentzed wants to merge 443 commits intomainfrom
decompile_depyf_optional

vincentzed commented Aug 2, 2025 •

edited by github-actions bot

Loading

Uh oh!

github-actions bot commented Aug 2, 2025

Uh oh!

github-actions bot commented Nov 1, 2025

Uh oh!

github-actions bot commented Dec 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

Conversation

vincentzed commented Aug 2, 2025 • edited by github-actions bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Essential Elements of an Effective PR Description Checklist

Purpose

Test Plan

Test Result

(Optional) Documentation Update

Uh oh!

github-actions bot commented Aug 2, 2025

Uh oh!

github-actions bot commented Nov 1, 2025

Uh oh!

github-actions bot commented Dec 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

vincentzed commented Aug 2, 2025 •

edited by github-actions bot

Loading