[Feat] Refactor rejection sampler by realliujiaxu · Pull Request #4975 · vllm-project/vllm-ascend

realliujiaxu · 2025-12-12T13:10:05Z

What this PR does / why we need it?

Currently, we are using AscendRejctionSampler that extends from RejctionSampler in spec decoding. AscendRejctionSampler override forward of RejctionSampler, only aming to replace rejection_sample func. This
causes a lot of code of RejctionSampler cannot be reused, for example:

Proposed Change:

Delete AscendRejctionSampler and use RejctionSampler directly in model runner.
Patch RejctionSampler.expand_batch_to_tokens and RejctionSampler.rejection_sample, maybe a better way is to make them as custom ops.
Modify NPUModelRunner following [V1][spec decode] return logprobs for spec decoding vllm#26060

Does this PR introduce any user-facing change?

No

How was this patch tested?

test logits processor for spec decoding
test logprobs for spec decoding
test logprobs for spec decoding + async shcheduling (test with [Feat] Support async_scheduler and disable_padded_drafter_batch in eagle #4893)
vLLM version: v0.12.0
vLLM main: vllm-project/vllm@ad32e3e

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

github-actions · 2025-12-12T13:10:13Z

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

A PR should do only one thing, smaller PRs enable faster reviews.
Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.

gemini-code-assist

Code Review

This pull request refactors the rejection sampler by removing the custom AscendRejectionSampler and using the upstream vllm.RejectionSampler with monkey-patching for Ascend-specific functions. This is a good change that reduces code duplication and aligns with the upstream implementation. The code is well-refactored into smaller, more readable methods like _sample and _bookkeeping_sync. My main concern is a potential issue with the log probability calculation for speculative decoding. The logic for handling logprobs seems incomplete, which might lead to incorrect results when logprobs are requested.

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

github-actions · 2025-12-13T09:09:08Z

This pull request has conflicts, please resolve those before we can evaluate the pull request.

…mpler-rebase # Conflicts: # vllm_ascend/worker/model_runner_v1.py

zhenwenqi2024 · 2025-12-15T03:12:06Z

+            list[int],
+    ]:
+        num_reqs = self.input_batch.num_reqs
+        discard_sampled_tokens_req_indices = np.nonzero(


can some attrs like discard_requests_mask be deleted in prepare_inputs?

we need to implement vllm-project/vllm#28597 first.

github-actions · 2025-12-15T03:18:35Z

This pull request has conflicts, please resolve those before we can evaluate the pull request.

…mpler-rebase # Conflicts: # vllm_ascend/worker/model_runner_v1.py

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

wangxiyuan · 2025-12-15T08:10:09Z

@@ -33,3 +33,4 @@
 import vllm_ascend.patch.worker.patch_qwen3_vl  # noqa


update this file please https://github.com/vllm-project/vllm-ascend/blob/main/vllm_ascend/patch/__init__.py

whx-sjtu

LGTM

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

@zzzzwwjj

I'd like to nominate @zzzzwwjj @realliujiaxu @LCAIZJ to join vLLM Ascend committer team. @zzzzwwjj --- - Review Quality‌: He has completed 80+reviews since April. 2025, include #3232 (comment), #4822 (comment), #4768 (comment) high quality review. - Sustained Contributions 15+ Valuable bug fix and refactor is very good. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Azzzzwwjj+is%3Aclosed+review%3Aapproved Continuous optimization of code architecture https://github.com/vllm-project/vllm-ascend/pulls?q=author%3Azzzzwwjj+is%3Amerged - Quality Contribution‌: #1229 #1979 #4359 #4878 - Community Involvement‌: He lead the #1147, to refactor AscendFusedMoE at the first time. He shared topics about large-scale distributed inference and reinforcement learning on vLLM-Ascend meetup on August 2nd. @realliujiaxu --- - Review Quality‌: He has completed about [40+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3Arealliujiaxu+-author%3Arealliujiaxu+) since September, include #4868 (comment), #2275 (comment). - Sustained Contributions He has completed (17 commits)[https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged], continuously optimizing the performance of the MoE model. - Quality Contribution‌: Contributed the Flash Comm1 feature to the community, supporting both eager and aclgraph execution modes, while compatible with multiple MoE models including DeepSeek and GLM4.5. - #3334 - #3420 - #3015 co-author: - #3495 - #4868 - Community Involvement‌: 1. Completed two major refactors, enabling vllm-ascend to evolve more rapidly and robustly: [Linear module](#2867) and [rejection sampler](#4975) 2. [fixed 8 bugs](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged+bugfix+) in graph mode, spec decoding and async scheduling. @LCAIZJ --- - Review Quality‌: He's been the go-to reviewer for virtually all PD disaggregation and KV Pool related PRs, having completed [30+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3ALCAIZJ+is%3Aopen+-author%3ALCAIZJ+) since May 2025. Notable examples include [discussion_r2553887360](#4345 (comment)), [issuecomment-3540994801](#4161 (comment)), and [discussion_r2492593988](#3981 (comment)), all demonstrating thorough and insightful feedback. - Sustained and Quality Contributions: His contributions reflect a strong grasp of both ‌vLLM‌ and ‌vLLM Ascend‌ codebases, particularly in prefill-decode disaggregation and KV pool areas ([7 PRs merged](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+)). Prefill-Decode Disaggregation: Delivered KV transfer functionality using Mooncake TransferEngine and enabled layerwise KV transfer #1568 #2602 KV Pool: Developed the foundational KV Pool infrastructure and migrated it to the latest ADXL stack #2913 #3350 - Quality Contribution‌: #1568 #2602 #2913 #3350 - Community Involvement‌: He actively responds to [community issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20commenter%3ALCAIZJ%20is%3Aopen%20-author%3ALCAIZJ), continuously monitors functionality and accuracy issues related to PD disaggregation and KV Pool, and proactively delivers [bug fixes](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+bugfix). - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>

### What this PR does / why we need it? Currently, we are using `AscendRejctionSampler` that extends from `RejctionSampler` in spec decoding. `AscendRejctionSampler` override `forward` of `RejctionSampler`, only aming to replace `rejection_sample` func. This causes a lot of code of `RejctionSampler` cannot be reused, for example: - vllm-project/vllm#19482 - vllm-project/vllm#26060 - vllm-project/vllm#29223 #### Proposed Change: - Delete `AscendRejctionSampler` and use `RejctionSampler` directly in model runner. - Patch `RejctionSampler.expand_batch_to_tokens` and `RejctionSampler.rejection_sample`, maybe a better way is to make them as custom ops. - Modify `NPUModelRunner` following vllm-project/vllm#26060 ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? - [x] test logits processor for spec decoding - [x] test logprobs for spec decoding - [x] test logprobs for spec decoding + async shcheduling (test with vllm-project#4893) - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e --------- Signed-off-by: realliujiaxu <realliujiaxu@163.com>

@zzzzwwjj

…t#5152) I'd like to nominate @zzzzwwjj @realliujiaxu @LCAIZJ to join vLLM Ascend committer team. @zzzzwwjj --- - Review Quality‌: He has completed 80+reviews since April. 2025, include vllm-project#3232 (comment), vllm-project#4822 (comment), vllm-project#4768 (comment) high quality review. - Sustained Contributions 15+ Valuable bug fix and refactor is very good. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Azzzzwwjj+is%3Aclosed+review%3Aapproved Continuous optimization of code architecture https://github.com/vllm-project/vllm-ascend/pulls?q=author%3Azzzzwwjj+is%3Amerged - Quality Contribution‌: vllm-project#1229 vllm-project#1979 vllm-project#4359 vllm-project#4878 - Community Involvement‌: He lead the vllm-project#1147, to refactor AscendFusedMoE at the first time. He shared topics about large-scale distributed inference and reinforcement learning on vLLM-Ascend meetup on August 2nd. @realliujiaxu --- - Review Quality‌: He has completed about [40+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3Arealliujiaxu+-author%3Arealliujiaxu+) since September, include vllm-project#4868 (comment), vllm-project#2275 (comment). - Sustained Contributions He has completed (17 commits)[https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged], continuously optimizing the performance of the MoE model. - Quality Contribution‌: Contributed the Flash Comm1 feature to the community, supporting both eager and aclgraph execution modes, while compatible with multiple MoE models including DeepSeek and GLM4.5. - vllm-project#3334 - vllm-project#3420 - vllm-project#3015 co-author: - vllm-project#3495 - vllm-project#4868 - Community Involvement‌: 1. Completed two major refactors, enabling vllm-ascend to evolve more rapidly and robustly: [Linear module](vllm-project#2867) and [rejection sampler](vllm-project#4975) 2. [fixed 8 bugs](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged+bugfix+) in graph mode, spec decoding and async scheduling. @LCAIZJ --- - Review Quality‌: He's been the go-to reviewer for virtually all PD disaggregation and KV Pool related PRs, having completed [30+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3ALCAIZJ+is%3Aopen+-author%3ALCAIZJ+) since May 2025. Notable examples include [discussion_r2553887360](vllm-project#4345 (comment)), [issuecomment-3540994801](vllm-project#4161 (comment)), and [discussion_r2492593988](vllm-project#3981 (comment)), all demonstrating thorough and insightful feedback. - Sustained and Quality Contributions: His contributions reflect a strong grasp of both ‌vLLM‌ and ‌vLLM Ascend‌ codebases, particularly in prefill-decode disaggregation and KV pool areas ([7 PRs merged](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+)). Prefill-Decode Disaggregation: Delivered KV transfer functionality using Mooncake TransferEngine and enabled layerwise KV transfer vllm-project#1568 vllm-project#2602 KV Pool: Developed the foundational KV Pool infrastructure and migrated it to the latest ADXL stack vllm-project#2913 vllm-project#3350 - Quality Contribution‌: vllm-project#1568 vllm-project#2602 vllm-project#2913 vllm-project#3350 - Community Involvement‌: He actively responds to [community issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20commenter%3ALCAIZJ%20is%3Aopen%20-author%3ALCAIZJ), continuously monitors functionality and accuracy issues related to PD disaggregation and KV Pool, and proactively delivers [bug fixes](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+bugfix). - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>

### What this PR does / why we need it? Currently, we are using `AscendRejctionSampler` that extends from `RejctionSampler` in spec decoding. `AscendRejctionSampler` override `forward` of `RejctionSampler`, only aming to replace `rejection_sample` func. This causes a lot of code of `RejctionSampler` cannot be reused, for example: - vllm-project/vllm#19482 - vllm-project/vllm#26060 - vllm-project/vllm#29223 #### Proposed Change: - Delete `AscendRejctionSampler` and use `RejctionSampler` directly in model runner. - Patch `RejctionSampler.expand_batch_to_tokens` and `RejctionSampler.rejection_sample`, maybe a better way is to make them as custom ops. - Modify `NPUModelRunner` following vllm-project/vllm#26060 ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? - [x] test logits processor for spec decoding - [x] test logprobs for spec decoding - [x] test logprobs for spec decoding + async shcheduling (test with vllm-project#4893) - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e --------- Signed-off-by: realliujiaxu <realliujiaxu@163.com> Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>

@zzzzwwjj

…t#5152) I'd like to nominate @zzzzwwjj @realliujiaxu @LCAIZJ to join vLLM Ascend committer team. @zzzzwwjj --- - Review Quality‌: He has completed 80+reviews since April. 2025, include vllm-project#3232 (comment), vllm-project#4822 (comment), vllm-project#4768 (comment) high quality review. - Sustained Contributions 15+ Valuable bug fix and refactor is very good. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Azzzzwwjj+is%3Aclosed+review%3Aapproved Continuous optimization of code architecture https://github.com/vllm-project/vllm-ascend/pulls?q=author%3Azzzzwwjj+is%3Amerged - Quality Contribution‌: vllm-project#1229 vllm-project#1979 vllm-project#4359 vllm-project#4878 - Community Involvement‌: He lead the vllm-project#1147, to refactor AscendFusedMoE at the first time. He shared topics about large-scale distributed inference and reinforcement learning on vLLM-Ascend meetup on August 2nd. @realliujiaxu --- - Review Quality‌: He has completed about [40+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3Arealliujiaxu+-author%3Arealliujiaxu+) since September, include vllm-project#4868 (comment), vllm-project#2275 (comment). - Sustained Contributions He has completed (17 commits)[https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged], continuously optimizing the performance of the MoE model. - Quality Contribution‌: Contributed the Flash Comm1 feature to the community, supporting both eager and aclgraph execution modes, while compatible with multiple MoE models including DeepSeek and GLM4.5. - vllm-project#3334 - vllm-project#3420 - vllm-project#3015 co-author: - vllm-project#3495 - vllm-project#4868 - Community Involvement‌: 1. Completed two major refactors, enabling vllm-ascend to evolve more rapidly and robustly: [Linear module](vllm-project#2867) and [rejection sampler](vllm-project#4975) 2. [fixed 8 bugs](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged+bugfix+) in graph mode, spec decoding and async scheduling. @LCAIZJ --- - Review Quality‌: He's been the go-to reviewer for virtually all PD disaggregation and KV Pool related PRs, having completed [30+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3ALCAIZJ+is%3Aopen+-author%3ALCAIZJ+) since May 2025. Notable examples include [discussion_r2553887360](vllm-project#4345 (comment)), [issuecomment-3540994801](vllm-project#4161 (comment)), and [discussion_r2492593988](vllm-project#3981 (comment)), all demonstrating thorough and insightful feedback. - Sustained and Quality Contributions: His contributions reflect a strong grasp of both ‌vLLM‌ and ‌vLLM Ascend‌ codebases, particularly in prefill-decode disaggregation and KV pool areas ([7 PRs merged](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+)). Prefill-Decode Disaggregation: Delivered KV transfer functionality using Mooncake TransferEngine and enabled layerwise KV transfer vllm-project#1568 vllm-project#2602 KV Pool: Developed the foundational KV Pool infrastructure and migrated it to the latest ADXL stack vllm-project#2913 vllm-project#3350 - Quality Contribution‌: vllm-project#1568 vllm-project#2602 vllm-project#2913 vllm-project#3350 - Community Involvement‌: He actively responds to [community issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20commenter%3ALCAIZJ%20is%3Aopen%20-author%3ALCAIZJ), continuously monitors functionality and accuracy issues related to PD disaggregation and KV Pool, and proactively delivers [bug fixes](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+bugfix). - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com> Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>

### What this PR does / why we need it? Currently, we are using `AscendRejctionSampler` that extends from `RejctionSampler` in spec decoding. `AscendRejctionSampler` override `forward` of `RejctionSampler`, only aming to replace `rejection_sample` func. This causes a lot of code of `RejctionSampler` cannot be reused, for example: - vllm-project/vllm#19482 - vllm-project/vllm#26060 - vllm-project/vllm#29223 #### Proposed Change: - Delete `AscendRejctionSampler` and use `RejctionSampler` directly in model runner. - Patch `RejctionSampler.expand_batch_to_tokens` and `RejctionSampler.rejection_sample`, maybe a better way is to make them as custom ops. - Modify `NPUModelRunner` following vllm-project/vllm#26060 ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? - [x] test logits processor for spec decoding - [x] test logprobs for spec decoding - [x] test logprobs for spec decoding + async shcheduling (test with vllm-project#4893) - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e --------- Signed-off-by: realliujiaxu <realliujiaxu@163.com> Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>

@zzzzwwjj

…t#5152) I'd like to nominate @zzzzwwjj @realliujiaxu @LCAIZJ to join vLLM Ascend committer team. @zzzzwwjj --- - Review Quality‌: He has completed 80+reviews since April. 2025, include vllm-project#3232 (comment), vllm-project#4822 (comment), vllm-project#4768 (comment) high quality review. - Sustained Contributions 15+ Valuable bug fix and refactor is very good. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Azzzzwwjj+is%3Aclosed+review%3Aapproved Continuous optimization of code architecture https://github.com/vllm-project/vllm-ascend/pulls?q=author%3Azzzzwwjj+is%3Amerged - Quality Contribution‌: vllm-project#1229 vllm-project#1979 vllm-project#4359 vllm-project#4878 - Community Involvement‌: He lead the vllm-project#1147, to refactor AscendFusedMoE at the first time. He shared topics about large-scale distributed inference and reinforcement learning on vLLM-Ascend meetup on August 2nd. @realliujiaxu --- - Review Quality‌: He has completed about [40+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3Arealliujiaxu+-author%3Arealliujiaxu+) since September, include vllm-project#4868 (comment), vllm-project#2275 (comment). - Sustained Contributions He has completed (17 commits)[https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged], continuously optimizing the performance of the MoE model. - Quality Contribution‌: Contributed the Flash Comm1 feature to the community, supporting both eager and aclgraph execution modes, while compatible with multiple MoE models including DeepSeek and GLM4.5. - vllm-project#3334 - vllm-project#3420 - vllm-project#3015 co-author: - vllm-project#3495 - vllm-project#4868 - Community Involvement‌: 1. Completed two major refactors, enabling vllm-ascend to evolve more rapidly and robustly: [Linear module](vllm-project#2867) and [rejection sampler](vllm-project#4975) 2. [fixed 8 bugs](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3Arealliujiaxu+is%3Amerged+bugfix+) in graph mode, spec decoding and async scheduling. @LCAIZJ --- - Review Quality‌: He's been the go-to reviewer for virtually all PD disaggregation and KV Pool related PRs, having completed [30+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+commenter%3ALCAIZJ+is%3Aopen+-author%3ALCAIZJ+) since May 2025. Notable examples include [discussion_r2553887360](vllm-project#4345 (comment)), [issuecomment-3540994801](vllm-project#4161 (comment)), and [discussion_r2492593988](vllm-project#3981 (comment)), all demonstrating thorough and insightful feedback. - Sustained and Quality Contributions: His contributions reflect a strong grasp of both ‌vLLM‌ and ‌vLLM Ascend‌ codebases, particularly in prefill-decode disaggregation and KV pool areas ([7 PRs merged](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+)). Prefill-Decode Disaggregation: Delivered KV transfer functionality using Mooncake TransferEngine and enabled layerwise KV transfer vllm-project#1568 vllm-project#2602 KV Pool: Developed the foundational KV Pool infrastructure and migrated it to the latest ADXL stack vllm-project#2913 vllm-project#3350 - Quality Contribution‌: vllm-project#1568 vllm-project#2602 vllm-project#2913 vllm-project#3350 - Community Involvement‌: He actively responds to [community issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20commenter%3ALCAIZJ%20is%3Aopen%20-author%3ALCAIZJ), continuously monitors functionality and accuracy issues related to PD disaggregation and KV Pool, and proactively delivers [bug fixes](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3ALCAIZJ+is%3Amerged+bugfix). - vLLM version: v0.12.0 - vLLM main: vllm-project/vllm@ad32e3e Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com> Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>

realliujiaxu added 6 commits December 12, 2025 21:09

extract _sample

dd08323

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

extract _bookkeeping_sync

ba3360b

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

refactor rejection sampler

e95e7fb

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

fix lint

a54e195

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

delete AscendRejectionSampler

c8a92e3

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

fix eagle proposer

876ee5c

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

gemini-code-assist bot reviewed Dec 12, 2025

View reviewed changes

Comment thread vllm_ascend/worker/model_runner_v1.py

wangxiyuan added ready read for review ready-for-test start test by label for PR labels Dec 12, 2025

realliujiaxu added 4 commits December 13, 2025 11:00

delete _bookkeeping_sync

5cba4c7

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

update bookkeeping

44b4582

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

fix async scheduling

ad659f1

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

fix lint

8882134

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

github-actions bot added the merge-conflicts label Dec 13, 2025

Merge remote-tracking branch 'origin/main' into refactor-rejection-sa…

9b1b816

…mpler-rebase # Conflicts: # vllm_ascend/worker/model_runner_v1.py

github-actions bot removed the merge-conflicts label Dec 14, 2025

realliujiaxu added 2 commits December 14, 2025 12:50

Merge branch 'main' into refactor-rejection-sampler-rebase

8cddedb

Merge branch 'main' into refactor-rejection-sampler-rebase

8886f90

zhenwenqi2024 reviewed Dec 15, 2025

View reviewed changes

github-actions bot added the merge-conflicts label Dec 15, 2025

Merge remote-tracking branch 'origin/main' into refactor-rejection-sa…

4369d54

…mpler-rebase # Conflicts: # vllm_ascend/worker/model_runner_v1.py

github-actions bot removed the merge-conflicts label Dec 15, 2025

realliujiaxu added 2 commits December 15, 2025 13:56

fix bookkeeping

0b9ad8b

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

rm unused prev_sampled_token_ids_invalid_indices

5c61b3e

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

wangxiyuan reviewed Dec 15, 2025

View reviewed changes

whx-sjtu approved these changes Dec 15, 2025

View reviewed changes

add comment and e2e UT

ee2ea15

Signed-off-by: realliujiaxu <realliujiaxu@163.com>

realliujiaxu added 2 commits December 15, 2025 22:27

Merge branch 'main' into refactor-rejection-sampler-rebase

6a36ee8

Merge branch 'main' into refactor-rejection-sampler-rebase

2b85726

wangxiyuan merged commit 9e24bdd into vllm-project:main Dec 16, 2025
25 checks passed

wangxiyuan mentioned this pull request Dec 18, 2025

Nominate new maintainers @zzzzwwjj @realliujiaxu @LCAIZJ #5152

Merged

realliujiaxu mentioned this pull request Jan 19, 2026

[RFC]: refactor sample of model runner #5987

Open

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Feat] Refactor rejection sampler#4975

[Feat] Refactor rejection sampler#4975
wangxiyuan merged 19 commits intovllm-project:mainfrom
realliujiaxu:refactor-rejection-sampler-rebase

realliujiaxu commented Dec 12, 2025 •

edited

Loading

Uh oh!

github-actions bot commented Dec 12, 2025

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

Uh oh!

github-actions bot commented Dec 13, 2025

Uh oh!

zhenwenqi2024 Dec 15, 2025

Uh oh!

realliujiaxu Dec 15, 2025

Uh oh!

github-actions bot commented Dec 15, 2025

Uh oh!

wangxiyuan Dec 15, 2025

Uh oh!

whx-sjtu left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

		@@ -33,3 +33,4 @@
		import vllm_ascend.patch.worker.patch_qwen3_vl # noqa

Conversation

realliujiaxu commented Dec 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What this PR does / why we need it?

Proposed Change:

Does this PR introduce any user-facing change?

How was this patch tested?

Uh oh!

github-actions bot commented Dec 12, 2025

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

github-actions bot commented Dec 13, 2025

Uh oh!

zhenwenqi2024 Dec 15, 2025

Choose a reason for hiding this comment

Uh oh!

realliujiaxu Dec 15, 2025

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Dec 15, 2025

Uh oh!

wangxiyuan Dec 15, 2025

Choose a reason for hiding this comment

Uh oh!

whx-sjtu left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

realliujiaxu commented Dec 12, 2025 •

edited

Loading