Enable the graph mode and add the full warmup logic for Deepseek OCR model #2138
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
HeJunyan
requested review from
PatrykWo,
afierka-intel,
jikunshang,
kzawora-intel,
madamczyk-intel,
mgawarkiewicz-intel,
michalkuligowski,
mswiniarsk,
vivekgoe and
xuechendi
as code owners
HeJunyan
force-pushed
the
improvement_performance_deepseek_ocr
branch
from
dd04d22 to
8f1c5fa
Compare
HeJunyan
force-pushed
the
improvement_performance_deepseek_ocr
branch
2 times, most recently
from
8c391a6 to
b560b50
Compare
HeJunyan
force-pushed
the
improvement_performance_deepseek_ocr
branch
from
b560b50 to
ebf2220
Compare
czhu15
reviewed
Dec 8, 2025
vllm/worker/hpu_model_runner.py
Outdated
| ''' | ||
|
|
||
| def __init__(self, is_batch_based): | ||
| def __init__(self, is_batch_based, sub_image_list=None): |
There was a problem hiding this comment.
suggest to add the description of the argument of is_batch_based and sub_image_list in the defination.
sub_image_list seems like a list of image objects. But from line 117, maybe it is a list of integers?
HeJunyan
force-pushed
the
improvement_performance_deepseek_ocr
branch
from
ebf2220 to
51dfa56
Compare
Author
|
I verified the PR in full warmup mode, the result is unchanged. So it is OK. |
HeJunyan
force-pushed
the
improvement_performance_deepseek_ocr
branch
2 times, most recently
from
44f67ad to
e1ab76b
Compare
HeJunyan
force-pushed
the
improvement_performance_deepseek_ocr
branch
from
e1ab76b to
363a85a
Compare
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Outdated suggestions cannot be applied.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.
The Deepseek OCR model uses fixed input tensor shape, so there is no need for padding.
Its input is different from the other MM models so a new fake data set is added to warm it up.