-
Notifications
You must be signed in to change notification settings - Fork 463
[CI] Update vllm version to 20250922(5aeb925) #3091
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
👋 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. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code Review
This pull request updates the vLLM dependency to a newer commit. The changes primarily involve adapting to API modifications in vLLM, such as the removal of SamplingMetadata
from several function signatures and changes in the sampler module. While most of the changes are correct adaptations, I've found several critical issues where compute_logits
method signatures in various models were not fully updated. This will lead to TypeError
exceptions at runtime. I have provided suggestions to fix these issues.
self, | ||
lm_head: VocabParallelEmbedding, | ||
hidden_states: torch.Tensor, | ||
# keep this for version compatibility |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
it's better to add a todo: delete arg sampling_metadata
when dropping v0.10.2
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ok, let's wait for ci passed, and I'll have a push
Signed-off-by: wangli <wangli858794774@gmail.com>
Signed-off-by: wangli <wangli858794774@gmail.com>
Signed-off-by: wangli <wangli858794774@gmail.com>
Signed-off-by: wangli <wangli858794774@gmail.com>
self.lm_head = PPMissingLayer() | ||
self.logits_processor = LogitsProcessor(config.vocab_size) | ||
self.sampler = get_sampler() | ||
self.sampler = Sampler() |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ditto
What this PR does / why we need it?
This pr bump vllm commit hash to vllm-project/vllm@5aeb925
fix issues:
vllm.model_executor.models
path , which breaks our custom registration of the deepseek_v3 model (it doesn't exist in the vllm model path). so I move deepseek_v3 model registy to deepseek_v2 to solve temporaryDoes this PR introduce any user-facing change?
How was this patch tested?