Revert "[Refactor] Unify full-graph parameter update logic (#6041)"#6227
Merged
wangxiyuan merged 1 commit intovllm-project:mainfrom Jan 25, 2026
Merged
Revert "[Refactor] Unify full-graph parameter update logic (#6041)"#6227wangxiyuan merged 1 commit intovllm-project:mainfrom
wangxiyuan merged 1 commit intovllm-project:mainfrom
Conversation
wangxiyuan
requested review from
LCAIZJ,
MengqingCao,
weijinqian0 and
yiz-liu
as code owners
wangxiyuan
added
ready
Contributor
|
👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:
If CI fails, you can run linting and testing checks locally according Contributing and Testing. |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Code Review
This pull request reverts commit 8966a99, which refactored the full-graph parameter update logic. The revert moves the parameter update logic from static methods within attention implementation classes back to standalone functions in vllm_ascend/compilation/acl_graph.py. The changes appear to correctly restore the previous implementation across all affected files, including tests. My main feedback is regarding code duplication for the graph parameter update dispatch logic, which is now present in three different locations. Consolidating this into a helper method would improve maintainability.
Comment on lines
+1145
to
+1164
| if self.vllm_config.model_config.use_mla: | ||
| if self.pcp_size * self.dcp_size > 1: | ||
| # FIXME: Try using `auto_dispatch_capture=True` | ||
| update_mla_attn_dcp_pcp_params(self.update_stream, | ||
| forward_context, | ||
| maybe_padded_num_tokens) | ||
| else: | ||
| # FIXME: Try using `auto_dispatch_capture=True` | ||
| update_mla_attn_params(self.update_stream, forward_context, | ||
| maybe_padded_num_tokens, | ||
| self.speculative_config) | ||
| else: | ||
| if self.pcp_size * self.dcp_size > 1: | ||
| update_attn_dcp_pcp_params(self.update_stream, | ||
| forward_context, | ||
| maybe_padded_num_tokens) | ||
| else: | ||
| update_attn_params(self.update_stream, forward_context, | ||
| maybe_padded_num_tokens, | ||
| self.vllm_config) |
Contributor
There was a problem hiding this comment.
This dispatch logic for updating graph parameters is duplicated in _generate_dummy_run_hidden_states (lines 2041-2059) and also in vllm_ascend/spec_decode/eagle_proposer.py (lines 1184-1198). This introduces a maintainability issue, as changes to this logic must be replicated in three places, increasing the risk of inconsistencies.
To address this, consider refactoring this logic into a helper method within the NPUModelRunner class. This would centralize the dispatch logic, making the code cleaner and less prone to errors.
For example, a new method _update_graph_params could be created to encapsulate this logic, and then called from all three locations.
weijinqian0
approved these changes
Jan 25, 2026
weijinqian0
approved these changes
Jan 25, 2026
Collaborator
|
CI all pass after reverting #6041. Do we plan to revert it directly or fix the bug it introduced? |
Collaborator
Author
|
let's merge this first |
wangxiyuan
added a commit
to wangxiyuan/vllm-ascend
that referenced
this pull request
Jan 25, 2026
…ject#6041)" (vllm-project#6227) This reverts commit 9564934.
wangxiyuan
added a commit
that referenced
this pull request
Jan 26, 2026
…6227) (#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
starmountain1997
pushed a commit
to starmountain1997/vllm-ascend
that referenced
this pull request
Jan 31, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
starmountain1997
pushed a commit
to starmountain1997/vllm-ascend
that referenced
this pull request
Jan 31, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
ZRJ026
pushed a commit
to ZRJ026/vllm-ascend
that referenced
this pull request
Feb 28, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094 Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>
ZRJ026
pushed a commit
to ZRJ026/vllm-ascend
that referenced
this pull request
Feb 28, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094 Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>
maoxx241
pushed a commit
to maoxx241/vllm-ascend
that referenced
this pull request
Mar 2, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
maoxx241
pushed a commit
to maoxx241/vllm-ascend
that referenced
this pull request
Mar 2, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
ZRJ026
pushed a commit
to ZRJ026/vllm-ascend
that referenced
this pull request
Mar 4, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094 Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>
ZRJ026
pushed a commit
to ZRJ026/vllm-ascend
that referenced
this pull request
Mar 4, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094 Signed-off-by: zrj026 <zhangrunjiang026@gmail.com>
LCAIZJ
pushed a commit
to LCAIZJ/vllm-ascend
that referenced
this pull request
Mar 7, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
LCAIZJ
pushed a commit
to LCAIZJ/vllm-ascend
that referenced
this pull request
Mar 7, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
jiangyunfan1
pushed a commit
to jiangyunfan1/vllm-ascend
that referenced
this pull request
Apr 9, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
jiangyunfan1
pushed a commit
to jiangyunfan1/vllm-ascend
that referenced
this pull request
Apr 9, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
yangzhe-2026
pushed a commit
to yangzhe-2026/vllm-ascend
that referenced
this pull request
May 6, 2026
…ect#6041)" (vllm-project#6227) This reverts commit 8966a99. It breaks the test `tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]` - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
yangzhe-2026
pushed a commit
to yangzhe-2026/vllm-ascend
that referenced
this pull request
May 6, 2026
…ject#6041)" (vllm-project#6227) (vllm-project#6231) This reverts commit 9564934. The CI failure doesn't related to this change. Let's reapply it. - vLLM version: v0.14.0 - vLLM main: vllm-project/vllm@d682094
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
3 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.
This reverts commit 8966a99.
It breaks the test
tests/e2e/singlecard/spec_decode/test_mtp_eagle_correctness.py::test_deepseek_mtp_correctness[True-FULL_DECODE_ONLY-2-wemaster/deepseek_mtp_main_random_bf16]