[main][Feature]Moe alltoallv communication optimization for unquantized RL training sence #2088

weijinqian0 · 2025-07-29T10:57:17Z

It comes from 0.9.1dev
[0.9.1][Feature]Moe alltoallv communication optimization for unquantized RL training sence & alltoallv support dpo (#1547)

vLLM version: v0.10.0
vLLM main: vllm-project/vllm@97608dc

…zed RL training sence & alltoallv support dpo (vllm-project#1547) Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

github-actions · 2025-07-29T15:57:33Z

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

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

…llm-project#1890) Optimize number of index selections of sin/cos cache. - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@656c24f Signed-off-by: whx-sjtu <2952154980@qq.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

…ect#1994) Support the inference of the Deepseekr1-w8a8-mtp model with statically-quantized shared_head in MTP layers. - vLLM version: v0.9.2 - vLLM main: vllm-project/vllm@6eca337 Signed-off-by: curryliu <120010041@link.cuhk.edu.cn> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

… up CI (vllm-project#2065) ### What this PR does / why we need it? Currently our workflow run time takes about 3 hours in total, which seriously affects the developer experience, so it is urgent to have a optimization, after this pr, It is expected that the running time of the full CI can be shortened to 1h40min. - Enable linux-aarch64-a2 (64GB) to replace linux-arm64-npu (32GB) - Change TP4 ---> TP2 * 2 max-parallel - Move DeepSeek-V2-Lite-W8A8 to single card test ### Does this PR introduce _any_ user-facing change? No - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@a248025 --------- Signed-off-by: wangli <wangli858794774@gmail.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

### What this PR does / why we need it? Bump default python version to 3.11, see vllm-project#1980 ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? pass CI - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@12a223e Signed-off-by: ChenTaoyu-SJTU <ctynb@qq.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

### What this PR does / why we need it? Add two custom kernels(bgmv_shrink and bgmv expand) to solve the performance of LoRA ### Does this PR introduce _any_ user-facing change? no user-facing change ### How was this patch tested? we add Unit Test file to test the custom ascendc kernel. See vllm-ascend/tests/e2e/singlecard/ops/test_bgmv_expand.py and vllm-ascend/tests/e2e/singlecard/ops/test_bgmv_expand.py Based on the actual test of the QWen2.5 7B model using vllm-ascend version v0.9.2.rc1, the TTFT, TPOT and throughput have increased by about 70%. - vLLM version: v0.9.2 - vLLM main: vllm-project/vllm@40d86ee --------- Signed-off-by: taoxudonghaha <justsheldon@163.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

### What this PR does / why we need it? Add performance tuning doc to main. Closes: vllm-project#1387 - vLLM version: v0.9.1 - vLLM main: vllm-project/vllm@923147b --------- Signed-off-by: shen-shanshan <467638484@qq.com> Signed-off-by: Shanshan Shen <87969357+shen-shanshan@users.noreply.github.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

…er tests (vllm-project#1246) ### What this PR does / why we need it? 1.Fixed the issue that pyhccl e2e cannot run continuously with other tests. 2.Cleaned up the resources occupied by the dynamic_npugraph_batchsize e2e test. ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? This is a e2e test e2e multi-cards tests local running successfully. - vLLM version: v0.9.2 - vLLM main: vllm-project/vllm@0df4d9b Signed-off-by: leo-pony <nengjunma@outlook.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

…t#1891) This PR designs the shared expert multi-stream parallelism of w8a8-dynamic-quantized MoE stage in more detail to achieve better performance. - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@2cc5711 Signed-off-by: whx-sjtu <2952154980@qq.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

Refactor Sampler implementation from patch way to inherit from vLLM Sampler interface. Next step: Make the op `TopKTopPSampler` in vLLM support custom ops register mechanism - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@61a6905 Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

### What this PR does / why we need it? Fix test on pyhccl to 2 cards ### Does this PR introduce _any_ user-facing change? N/A ### How was this patch tested? CI passed with existing test. - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@0d0cc9e Signed-off-by: MengqingCao <cmq0113@163.com> Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

MengqingCao

Nit: we could do the following in the following up pr

MengqingCao · 2025-08-01T06:13:50Z

tests/ut/test_distributed_tensor_parallel.py

@@ -0,0 +1,139 @@
+#


plz move this to tests/distributed/test_tensor_paralle.py

MengqingCao · 2025-08-01T06:14:57Z

tests/ut/test_token_dispatcher.py

@@ -0,0 +1,65 @@
+#


plz move this to tests/moe_dispatcher/test_token_dispatcher.py

vllm_ascend/ascend_forward_context.py

MengqingCao · 2025-08-01T06:32:46Z

vllm_ascend/distributed/tensor_parallel.py

+import torch
+
+
+def _gather_along_first_dim(input_, group, output_split_sizes=None):


I think we can refactor the ops in this file into distributed/communicator, and call them through get_tp_group().gather_along_first_dim for example.

(Maybe the refactor could be done in the following up pr

ok, we'll consider it

I think this need GroupCoordinator to support gather_along_first_dim as well. It's need vLLM support it first. we can do it in the future.

wangxiyuan

@jianzs Please help review. Thanks. This is a large PR, while I think it's good to go, because it's a single feature and most of the code doesn't affect other function.

wangxiyuan · 2025-08-01T10:24:12Z

tests/e2e/multicard/test_offline_inference_distributed.py



+@patch.dict(os.environ, {"VLLM_ASCEND_ENABLE_MOE_ALL2ALL_SEQ": "1"})
+def test_models_distributed_alltoallv() -> None:


This test is not ran in CI. You should enable it here as well https://github.com/vllm-project/vllm-ascend/blob/main/.github/workflows/vllm_ascend_test.yaml#L277-L281

wangxiyuan · 2025-08-01T10:26:07Z

vllm_ascend/distributed/tensor_parallel.py

+import torch
+
+
+def _gather_along_first_dim(input_, group, output_split_sizes=None):


I think this need GroupCoordinator to support gather_along_first_dim as well. It's need vLLM support it first. we can do it in the future.

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

github-actions · 2025-08-01T11:25:34Z

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

wangxiyuan · 2025-08-02T01:25:38Z

tests/ut/ops/test_token_dispatcher.py

@@ -0,0 +1,65 @@
+#


this should be in ops/moe_dispatcher folder, it can be done in a follow up PR.

wangxiyuan · 2025-08-02T01:49:59Z

Let's merge this first to unblock other cherry-pick action. @jianzs any comment is welcome. It can be don in the follow up PR.

jianzs · 2025-08-02T02:28:55Z

Let's merge this first to unblock other cherry-pick action. @jianzs any comment is welcome. It can be don in the follow up PR.

Got it.

Yikun · 2025-08-04T00:41:49Z

This PR break the Qwen3-30B-A3B accuracy test: https://github.com/vllm-project/vllm-ascend/actions/runs/16690820541/job/47248288106

 torch._dynamo.exc.Unsupported: dynamic shape operator: aten.bincount.default; Operator does not have a meta kernel that supports dynamic output shapes, please report an issue to PyTorch
 
 from user code:
    File "/__w/vllm-ascend/vllm-ascend/vllm-empty/vllm/model_executor/models/qwen3_moe.py", line 369, in forward
     hidden_states, residual = layer(positions, hidden_states, residual)
   File "/__w/vllm-ascend/vllm-ascend/vllm-empty/vllm/model_executor/models/qwen3_moe.py", line 313, in forward
     hidden_states = self.mlp(hidden_states)
   File "/__w/vllm-ascend/vllm-ascend/vllm_ascend/ops/fused_moe.py", line 1615, in forward
     hidden_states = self.experts(
   File "/__w/vllm-ascend/vllm-ascend/vllm_ascend/ops/fused_moe.py", line 1446, in forward
     e_hidden_states = self.quant_method.apply(
   File "/__w/vllm-ascend/vllm-ascend/vllm_ascend/ops/fused_moe.py", line 1159, in apply
     return fused_experts_with_all2all(hidden_states=x,
   File "/__w/vllm-ascend/vllm-ascend/vllm_ascend/ops/fused_moe.py", line 371, in fused_experts_with_all2all
     global_expert_tokens = torch.bincount(expanded_expert_idx,
 
 Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information
 
 
 You can suppress this exception and fall back to eager by setting:
     import torch._dynamo
     torch._dynamo.config.suppress_errors = True

…ed RL training sence (vllm-project#2088) It comes from 0.9.1dev [0.9.1][Feature]Moe alltoallv communication optimization for unquantized RL training sence & alltoallv support dpo (vllm-project#1547) - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@97608dc --------- Signed-off-by: weijinqian_v1 <weijinqian@huawei.com> Signed-off-by: whx-sjtu <2952154980@qq.com> Signed-off-by: curryliu <120010041@link.cuhk.edu.cn> Signed-off-by: wangli <wangli858794774@gmail.com> Signed-off-by: ChenTaoyu-SJTU <ctynb@qq.com> Signed-off-by: taoxudonghaha <justsheldon@163.com> Signed-off-by: shen-shanshan <467638484@qq.com> Signed-off-by: Shanshan Shen <87969357+shen-shanshan@users.noreply.github.com> Signed-off-by: leo-pony <nengjunma@outlook.com> Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com> Signed-off-by: MengqingCao <cmq0113@163.com> Co-authored-by: weijinqian_v1 <weijinqian@huawei.com> Co-authored-by: whx <56632993+whx-sjtu@users.noreply.github.com> Co-authored-by: curryliu <99582471+Irving11-BKN@users.noreply.github.com> Co-authored-by: Li Wang <wangli858794774@gmail.com> Co-authored-by: TaoYu Chen <ctynb@qq.com> Co-authored-by: taoxudonghaha <justsheldon@163.com> Co-authored-by: Shanshan Shen <467638484@qq.com> Co-authored-by: leo-pony <nengjunma@outlook.com> Co-authored-by: wangxiyuan <wangxiyuan1007@gmail.com> Co-authored-by: Mengqing Cao <cmq0113@163.com>

@MengqingCao

I would like to nominate Mengqing Cao (@MengqingCao https://github.com/MengqingCao) as a maintainer, starting with my +1. ## Reason Review Quality‌: She has completed [120+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao) since Feb. 2025, include [#review-3077842852](#2088 (review)), [comment-2990074116](#1032 (comment)), [comment-2921063723](#1013 (comment)) high quality review. Sustained and Quality Contributions: She has Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases and solid contributions include The vLLM contributions and help vLLM Ascend release is the main reason I nominated her: - vLLM: Things worth mentioning that she completed [28+ PR contributions](https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+) in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support. She is one of the important co-authors of [vllm#8054](vllm-project/vllm#8054) and hardware plugin RFC, this makes vllm-ascend plugin possible. Community Involvement: She is also very active and involved in [60+ issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao). So I think she's a great addition to the vLLM Ascend Maintainer team. - ✅**Review Quality‌:** She has completed 120+ reviews since Feb. 2025. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao, include #2088 (review), #1446 (comment), #1032 (comment), #1013 (comment) quality review. - ✅**Sustained Contributions:** 99+ PR merged in vllm-project/vllm-ascend https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged - ✅**Quality Contribution‌:** She is one of the important co-authors of vllm-project/vllm#8054 , this makes vllm-ascend plugin possible. Things worth mentioning that she complete 28+ PR contributions in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support: https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+. At 2025 Q2, She also lead the [[RFC]: E2E CI test for key features](#413) and [[RFC]: Unit test coverage improvement](#1298) to help vllm ascend improve the coverage. Her main contributions focus on the adaptation of parallel strategies and communicator, such as #1800, #1856. These contributions are sufficient to prove she has “Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases” - ✅**Community Involvement‌:** Involved in 63+ issue reviewer https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao She led the v0.10.1 release as release manager - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@78dba40 Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>

Yikun · 2025-08-26T01:18:41Z

This PR break the Qwen3-30B-A3B accuracy test: https://github.com/vllm-project/vllm-ascend/actions/runs/16690820541/job/47248288106

issue: #2226
v0.9.1-dev: resolved since v0.9.1rc3: #2478
main: resolved since v0.10.0rc1: #2183

@MengqingCao

I would like to nominate Mengqing Cao (@MengqingCao https://github.com/MengqingCao) as a maintainer, starting with my +1. ## Reason Review Quality‌: She has completed [120+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao) since Feb. 2025, include [#review-3077842852](vllm-project#2088 (review)), [comment-2990074116](vllm-project#1032 (comment)), [comment-2921063723](vllm-project#1013 (comment)) high quality review. Sustained and Quality Contributions: She has Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases and solid contributions include The vLLM contributions and help vLLM Ascend release is the main reason I nominated her: - vLLM: Things worth mentioning that she completed [28+ PR contributions](https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+) in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support. She is one of the important co-authors of [vllm#8054](vllm-project/vllm#8054) and hardware plugin RFC, this makes vllm-ascend plugin possible. Community Involvement: She is also very active and involved in [60+ issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao). So I think she's a great addition to the vLLM Ascend Maintainer team. - ✅**Review Quality‌:** She has completed 120+ reviews since Feb. 2025. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao, include vllm-project#2088 (review), vllm-project#1446 (comment), vllm-project#1032 (comment), vllm-project#1013 (comment) quality review. - ✅**Sustained Contributions:** 99+ PR merged in vllm-project/vllm-ascend https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged - ✅**Quality Contribution‌:** She is one of the important co-authors of vllm-project/vllm#8054 , this makes vllm-ascend plugin possible. Things worth mentioning that she complete 28+ PR contributions in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support: https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+. At 2025 Q2, She also lead the [[RFC]: E2E CI test for key features](vllm-project#413) and [[RFC]: Unit test coverage improvement](vllm-project#1298) to help vllm ascend improve the coverage. Her main contributions focus on the adaptation of parallel strategies and communicator, such as vllm-project#1800, vllm-project#1856. These contributions are sufficient to prove she has “Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases” - ✅**Community Involvement‌:** Involved in 63+ issue reviewer https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao She led the v0.10.1 release as release manager - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@78dba40 Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>

…ed RL training sence (vllm-project#2088) It comes from 0.9.1dev [0.9.1][Feature]Moe alltoallv communication optimization for unquantized RL training sence & alltoallv support dpo (vllm-project#1547) - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@97608dc --------- Signed-off-by: weijinqian_v1 <weijinqian@huawei.com> Signed-off-by: whx-sjtu <2952154980@qq.com> Signed-off-by: curryliu <120010041@link.cuhk.edu.cn> Signed-off-by: wangli <wangli858794774@gmail.com> Signed-off-by: ChenTaoyu-SJTU <ctynb@qq.com> Signed-off-by: taoxudonghaha <justsheldon@163.com> Signed-off-by: shen-shanshan <467638484@qq.com> Signed-off-by: Shanshan Shen <87969357+shen-shanshan@users.noreply.github.com> Signed-off-by: leo-pony <nengjunma@outlook.com> Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com> Signed-off-by: MengqingCao <cmq0113@163.com> Co-authored-by: weijinqian_v1 <weijinqian@huawei.com> Co-authored-by: whx <56632993+whx-sjtu@users.noreply.github.com> Co-authored-by: curryliu <99582471+Irving11-BKN@users.noreply.github.com> Co-authored-by: Li Wang <wangli858794774@gmail.com> Co-authored-by: TaoYu Chen <ctynb@qq.com> Co-authored-by: taoxudonghaha <justsheldon@163.com> Co-authored-by: Shanshan Shen <467638484@qq.com> Co-authored-by: leo-pony <nengjunma@outlook.com> Co-authored-by: wangxiyuan <wangxiyuan1007@gmail.com> Co-authored-by: Mengqing Cao <cmq0113@163.com>

@MengqingCao

I would like to nominate Mengqing Cao (@MengqingCao https://github.com/MengqingCao) as a maintainer, starting with my +1. ## Reason Review Quality‌: She has completed [120+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao) since Feb. 2025, include [#review-3077842852](vllm-project#2088 (review)), [comment-2990074116](vllm-project#1032 (comment)), [comment-2921063723](vllm-project#1013 (comment)) high quality review. Sustained and Quality Contributions: She has Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases and solid contributions include The vLLM contributions and help vLLM Ascend release is the main reason I nominated her: - vLLM: Things worth mentioning that she completed [28+ PR contributions](https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+) in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support. She is one of the important co-authors of [vllm#8054](vllm-project/vllm#8054) and hardware plugin RFC, this makes vllm-ascend plugin possible. Community Involvement: She is also very active and involved in [60+ issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao). So I think she's a great addition to the vLLM Ascend Maintainer team. - ✅**Review Quality‌:** She has completed 120+ reviews since Feb. 2025. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao, include vllm-project#2088 (review), vllm-project#1446 (comment), vllm-project#1032 (comment), vllm-project#1013 (comment) quality review. - ✅**Sustained Contributions:** 99+ PR merged in vllm-project/vllm-ascend https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged - ✅**Quality Contribution‌:** She is one of the important co-authors of vllm-project/vllm#8054 , this makes vllm-ascend plugin possible. Things worth mentioning that she complete 28+ PR contributions in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support: https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+. At 2025 Q2, She also lead the [[RFC]: E2E CI test for key features](vllm-project#413) and [[RFC]: Unit test coverage improvement](vllm-project#1298) to help vllm ascend improve the coverage. Her main contributions focus on the adaptation of parallel strategies and communicator, such as vllm-project#1800, vllm-project#1856. These contributions are sufficient to prove she has “Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases” - ✅**Community Involvement‌:** Involved in 63+ issue reviewer https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao She led the v0.10.1 release as release manager - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@78dba40 Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>

…ed RL training sence (vllm-project#2088) It comes from 0.9.1dev [0.9.1][Feature]Moe alltoallv communication optimization for unquantized RL training sence & alltoallv support dpo (vllm-project#1547) - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@97608dc --------- Signed-off-by: weijinqian_v1 <weijinqian@huawei.com> Signed-off-by: whx-sjtu <2952154980@qq.com> Signed-off-by: curryliu <120010041@link.cuhk.edu.cn> Signed-off-by: wangli <wangli858794774@gmail.com> Signed-off-by: ChenTaoyu-SJTU <ctynb@qq.com> Signed-off-by: taoxudonghaha <justsheldon@163.com> Signed-off-by: shen-shanshan <467638484@qq.com> Signed-off-by: Shanshan Shen <87969357+shen-shanshan@users.noreply.github.com> Signed-off-by: leo-pony <nengjunma@outlook.com> Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com> Signed-off-by: MengqingCao <cmq0113@163.com> Co-authored-by: weijinqian_v1 <weijinqian@huawei.com> Co-authored-by: whx <56632993+whx-sjtu@users.noreply.github.com> Co-authored-by: curryliu <99582471+Irving11-BKN@users.noreply.github.com> Co-authored-by: Li Wang <wangli858794774@gmail.com> Co-authored-by: TaoYu Chen <ctynb@qq.com> Co-authored-by: taoxudonghaha <justsheldon@163.com> Co-authored-by: Shanshan Shen <467638484@qq.com> Co-authored-by: leo-pony <nengjunma@outlook.com> Co-authored-by: wangxiyuan <wangxiyuan1007@gmail.com> Co-authored-by: Mengqing Cao <cmq0113@163.com>

@MengqingCao

I would like to nominate Mengqing Cao (@MengqingCao https://github.com/MengqingCao) as a maintainer, starting with my +1. ## Reason Review Quality‌: She has completed [120+ reviews](https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao) since Feb. 2025, include [#review-3077842852](vllm-project#2088 (review)), [comment-2990074116](vllm-project#1032 (comment)), [comment-2921063723](vllm-project#1013 (comment)) high quality review. Sustained and Quality Contributions: She has Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases and solid contributions include The vLLM contributions and help vLLM Ascend release is the main reason I nominated her: - vLLM: Things worth mentioning that she completed [28+ PR contributions](https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+) in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support. She is one of the important co-authors of [vllm#8054](vllm-project/vllm#8054) and hardware plugin RFC, this makes vllm-ascend plugin possible. Community Involvement: She is also very active and involved in [60+ issues](https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao). So I think she's a great addition to the vLLM Ascend Maintainer team. - ✅**Review Quality‌:** She has completed 120+ reviews since Feb. 2025. https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+is%3Aclosed+commenter%3Amengqingcao+-author%3Amengqingcao, include vllm-project#2088 (review), vllm-project#1446 (comment), vllm-project#1032 (comment), vllm-project#1013 (comment) quality review. - ✅**Sustained Contributions:** 99+ PR merged in vllm-project/vllm-ascend https://github.com/vllm-project/vllm-ascend/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged - ✅**Quality Contribution‌:** She is one of the important co-authors of vllm-project/vllm#8054 , this makes vllm-ascend plugin possible. Things worth mentioning that she complete 28+ PR contributions in vllm-project/vllm, especially for vLLM platform module to improve vLLM mult hardware support: https://github.com/vllm-project/vllm/pulls?q=is%3Apr+author%3AMengqingCao+is%3Amerged+. At 2025 Q2, She also lead the [[RFC]: E2E CI test for key features](vllm-project#413) and [[RFC]: Unit test coverage improvement](vllm-project#1298) to help vllm ascend improve the coverage. Her main contributions focus on the adaptation of parallel strategies and communicator, such as vllm-project#1800, vllm-project#1856. These contributions are sufficient to prove she has “Deep understanding of ‌vLLM‌ and ‌vLLM Ascend‌ codebases” - ✅**Community Involvement‌:** Involved in 63+ issue reviewer https://github.com/vllm-project/vllm-ascend/issues?q=is%3Aissue%20state%3Aclosed%20-author%3AMengqingCao%20commenter%3AMengqingCao She led the v0.10.1 release as release manager - vLLM version: v0.10.0 - vLLM main: vllm-project/vllm@78dba40 Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>

weijinqian0 and others added 5 commits July 29, 2025 10:26

[0.9.1][Feature]Moe alltoallv communication optimization for unquanti…

75503d3

…zed RL training sence & alltoallv support dpo (vllm-project#1547) Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add Moe alltoallv.

63cb062

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add Moe alltoallv.

715e6f1

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

7c7e4e9

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

7106d77

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

github-actions bot added module:tests module:ops module:core merge-conflicts labels Jul 29, 2025

weijinqian_v1 and others added 15 commits July 30, 2025 09:39

[v0.9.1][Feature] add moe alltoallv.

1b53047

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

a863507

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

a702414

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

5841dc8

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

868aa2f

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

weijinqian0 force-pushed the main_merge_qwen3 branch from 7d74fc8 to 868aa2f Compare July 30, 2025 01:39

weijinqian_v1 and others added 4 commits July 30, 2025 09:46

[v0.9.1][Feature] add moe alltoallv.

71bc50b

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

b118bbd

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

978f430

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

Merge branch 'main' into main_merge_qwen3

c7cc22a

weijinqian_v1 added 4 commits July 31, 2025 21:10

[v0.9.1][Feature] add moe alltoallv.

f8aa32b

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

01ebd07

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

57b5378

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

[v0.9.1][Feature] add moe alltoallv.

aa26b19

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

ganyi1996ppo approved these changes Aug 1, 2025

View reviewed changes

MengqingCao reviewed Aug 1, 2025

View reviewed changes

wangxiyuan reviewed Aug 1, 2025

View reviewed changes

[v0.9.1][Feature] add moe alltoallv.

c4993df

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>

github-actions bot added the merge-conflicts label Aug 1, 2025

Merge branch 'main' into main_merge_qwen3

5932033

github-actions bot removed the merge-conflicts label Aug 1, 2025

wangxiyuan reviewed Aug 2, 2025

View reviewed changes

wangxiyuan merged commit 6e00aed into vllm-project:main Aug 2, 2025
25 checks passed

Yikun mentioned this pull request Aug 5, 2025

[Bug]: Qwen3 MoE aclgraph mode with tp failed when enbale ep due to bincount error #2226

Open

jianzs mentioned this pull request Aug 19, 2025

Nominate Mengqing Cao as vllm-ascend maintainer #2433

Merged

weijinqian0 deleted the main_merge_qwen3 branch September 8, 2025 01:11

wangxiyuan mentioned this pull request Oct 13, 2025

[Community] Nominate new maintainers: @yiz-liu @paulyu12 @weijinqian0 @nalinaly #3406

Merged

		import torch


		def _gather_along_first_dim(input_, group, output_split_sizes=None):



		@patch.dict(os.environ, {"VLLM_ASCEND_ENABLE_MOE_ALL2ALL_SEQ": "1"})
		def test_models_distributed_alltoallv() -> None:

[main][Feature]Moe alltoallv communication optimization for unquantized RL training sence #2088

[main][Feature]Moe alltoallv communication optimization for unquantized RL training sence #2088

Uh oh!

Conversation

weijinqian0 commented Jul 29, 2025 • edited by github-actions bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot commented Jul 29, 2025

Uh oh!

MengqingCao left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

wangxiyuan left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Aug 1, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

wangxiyuan commented Aug 2, 2025

Uh oh!

jianzs commented Aug 2, 2025

Uh oh!

Yikun commented Aug 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Yikun commented Aug 26, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

13 participants

weijinqian0 commented Jul 29, 2025 •

edited by github-actions bot

Loading

Yikun commented Aug 4, 2025 •

edited

Loading