[Enhance]: Add validation for expert parallelism settings #1199
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
Add validation to prevent simultaneous use of `--enable-expert-parallel` and `expert-tensor-parallel-size` configurations. These settings are mutually exclusive. Implementing this check prevents unexpected behavior and improves error tracing. If both settings are enabled concurrently, the system now throws an error, making it easier to identify and resolve configuration issues. Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>
Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>
Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>
ApsarasX
approved these changes
Jun 13, 2025
There was a problem hiding this comment.
Pull Request Overview
Adds a validation step in check_ascend_config to prevent simultaneous use of --enable-expert-parallel and an expert_tensor_parallel_size greater than 1, avoiding conflicting configurations.
- Import
TYPE_CHECKINGand annotatevllm_configwithVllmConfigfor better type safety. - Introduce a runtime check that raises an error when expert parallelism flags conflict.
- Preserve existing experimental ACL graph warnings.
Comments suppressed due to low confidence (1)
vllm_ascend/ascend_config.py:171
- Consider adding a unit test to verify that enabling both
enable_expert_parallelandexpert_tensor_parallel_size > 1reliably raises the intended error.
# for expert parallelism
Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
ApsarasX
reviewed
Jun 16, 2025
| "raise an issue on https://github.com/vllm-project/vllm-ascend/issues" | ||
| " if you encourage any Error") | ||
|
|
||
| # for expert parallelism |
There was a problem hiding this comment.
move this validation into check_and_update_config in platform.py?
There was a problem hiding this comment.
This function will be called within check_and_update_config
|
@Yikun @wangxiyuan ready to merge. |
wangxiyuan
approved these changes
Jun 17, 2025
|
can you update the tests/ut/test_ascend_config.py as well? |
|
This pull request has conflicts, please resolve those before we can evaluate the pull request. |
|
ETP has been removed already. |
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.
Add validation to prevent simultaneous use of
--enable-expert-parallelandexpert-tensor-parallel-sizeconfigurations. These settings are mutually exclusive. Implementing this check prevents unexpected behavior and improves error tracing.If both settings are enabled concurrently, the system now throws an error, making it easier to identify and resolve configuration issues.