Add Custom Kernels For LoRA Performance #1884
Merged
Conversation
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
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #1884 +/- ##
==========================================
+ Coverage 72.35% 73.12% +0.76%
==========================================
Files 88 90 +2
Lines 9666 9956 +290
==========================================
+ Hits 6994 7280 +286
- Misses 2672 2676 +4
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
wangxiyuan
reviewed
Jul 23, 2025
| " int added_vocab_end_index) -> (Tensor masked_input, Tensor mask)"); | ||
| ops.impl("get_masked_input_and_mask", torch::kPrivateUse1, &vllm_ascend::get_masked_input_and_mask); | ||
|
|
||
| ops.def("bgmv_shrink(Tensor! x, Tensor! weight, Tensor! indices, Tensor! y, float scale) -> ()"); |
There was a problem hiding this comment.
once the bgmv_shrink and bgmv_expand is exposed, I think we should use it to replace the common one as well. For example here: https://github.com/vllm-project/vllm-ascend/blob/main/vllm_ascend/lora/punica_wrapper/punica_npu.py#L51
There was a problem hiding this comment.
I have already replaced the relevant interface calls in the punica_npu.py
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
Signed-off-by: taoxudonghaha <justsheldon@163.com>
wangxiyuan
approved these changes
Jul 26, 2025
weijinqian0
pushed a commit
to weijinqian0/vllm-ascend
that referenced
this pull request
Jul 30, 2025
### 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>
weijinqian0
pushed a commit
to weijinqian0/vllm-ascend
that referenced
this pull request
Jul 30, 2025
### 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>
chopper0126
pushed a commit
to chopper0126/vllm-ascend
that referenced
this pull request
Sep 26, 2025
### 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>
Angazenn
pushed a commit
to Angazenn/vllm-ascend
that referenced
this pull request
Oct 21, 2025
### 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>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
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%.