[Bugfix] Fix output tensor shape in vanilla_chunked_prefill and update import paths for model_loader #773
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: Yizhou Liu <[email protected]>
…oader in NPUModelRunnerBase Signed-off-by: Yizhou Liu <[email protected]>
yiz-liu
changed the title
jianzs
reviewed
May 7, 2025
Collaborator
jianzs
left a comment
jianzs
left a comment
There was a problem hiding this comment.
We should make sure the vllm-ascend main branch works with vllm versions 0.8.5 and 0.8.5.post1.
yiz-liu
force-pushed
the
fix-output-shape
branch
2 times, most recently
from
3e53cb9 to
87badbf
Compare
Collaborator
|
Looks CI is incomplete, can you trigger it again with a new commit? |
yiz-liu
force-pushed
the
fix-output-shape
branch
4 times, most recently
from
10a64bc to
ef7fb88
Compare
Collaborator
Author
yiz-liu
force-pushed
the
fix-output-shape
branch
10 times, most recently
from
046deab to
eb8ccd2
Compare
Signed-off-by: Yizhou Liu <[email protected]>
Collaborator
Author
MengqingCao
reviewed
May 8, 2025
| ) -> None: | ||
| from vllm.model_executor.model_loader.loader import ShardedStateLoader | ||
| if vllm_version_is("0.8.5") or vllm_version_is("0.8.5.post1"): | ||
| from vllm.model_executor.model_loader.loader import ShardedStateLoader # type: ignore[import] # isort: skip # noqa |
Collaborator
There was a problem hiding this comment.
I think this is a reasonable approach to skip code static checking. we could not install 2 version of vllm in the same python env, thus I agree to skip this in v0.8.5
Collaborator
There was a problem hiding this comment.
Should we ensure vllm-ascend main branch is compatible with both vllm version 0.8.5 and 0.8.5.post1?
Collaborator
There was a problem hiding this comment.
Yeah, but this is just skip code format static check of 0.8.5 in CI of main branch. This has no impact on the features.
Collaborator
There was a problem hiding this comment.
Using vLLM v0.8.5 without checking this condition for v0.8.5 will trigger the else branch, causing a problem.
Collaborator
|
Merge this pull request ASAP; many others are blocked. @MengqingCao @Yikun |
ganyi1996ppo
approved these changes
May 8, 2025
chopper0126
pushed a commit
to chopper0126/vllm-ascend
that referenced
this pull request
Oct 16, 2025
…e import paths for model_loader (vllm-project#773) <!-- Thanks for sending a pull request! BEFORE SUBMITTING, PLEASE READ https://docs.vllm.ai/en/latest/contributing/overview.html --> ### What this PR does / why we need it? <!-- - Please clarify what changes you are proposing. The purpose of this section is to outline the changes and how this PR fixes the issue. If possible, please consider writing useful notes for better and faster reviews in your PR. - Please clarify why the changes are needed. For instance, the use case and bug description. - Fixes # --> Fix output tensor shape in vanilla_chunked_prefill function. ### Does this PR introduce _any_ user-facing change? <!-- Note that it means *any* user-facing change including all aspects such as API, interface or other behavior changes. Documentation-only updates are not considered user-facing changes. --> None. ### How was this patch tested? <!-- CI passed with new added/existing test. If it was tested in a way different from regular unit tests, please clarify how you tested step by step, ideally copy and paste-able, so that other reviewers can test and check, and descendants can verify in the future. If tests were not added, please describe why they were not added and/or why it was difficult to add. --> Run offline inference on DeepSeek models. --------- Signed-off-by: Yizhou Liu <[email protected]>
Angazenn
pushed a commit
to Angazenn/vllm-ascend
that referenced
this pull request
Oct 21, 2025
…e import paths for model_loader (vllm-project#773) <!-- Thanks for sending a pull request! BEFORE SUBMITTING, PLEASE READ https://docs.vllm.ai/en/latest/contributing/overview.html --> ### What this PR does / why we need it? <!-- - Please clarify what changes you are proposing. The purpose of this section is to outline the changes and how this PR fixes the issue. If possible, please consider writing useful notes for better and faster reviews in your PR. - Please clarify why the changes are needed. For instance, the use case and bug description. - Fixes # --> Fix output tensor shape in vanilla_chunked_prefill function. ### Does this PR introduce _any_ user-facing change? <!-- Note that it means *any* user-facing change including all aspects such as API, interface or other behavior changes. Documentation-only updates are not considered user-facing changes. --> None. ### How was this patch tested? <!-- CI passed with new added/existing test. If it was tested in a way different from regular unit tests, please clarify how you tested step by step, ideally copy and paste-able, so that other reviewers can test and check, and descendants can verify in the future. If tests were not added, please describe why they were not added and/or why it was difficult to add. --> Run offline inference on DeepSeek models. --------- Signed-off-by: Yizhou Liu <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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?
Fix output tensor shape in vanilla_chunked_prefill function.
Does this PR introduce any user-facing change?
None.
How was this patch tested?
Run offline inference on DeepSeek models.