-
Notifications
You must be signed in to change notification settings - Fork 124
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
Support deeepseek models export #1155
Conversation
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
f8cb843
to
b1a2950
Compare
@echarlaix @IlyasMoutawwakil could you please take a look? |
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.
LGTM
@@ -3575,6 +3575,299 @@ def __exit__(self, exc_type, exc_value, traceback): | |||
block.self_attn.forward = block.self_attn._orig_forward | |||
|
|||
|
|||
class DeepseekPatcher(DecoderModelPatcher): |
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.
deepseek-v3 will be included in transformers v4.50 (huggingface/transformers#35926) could make sense to make sure everything is compatible before merging
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.
the pr stated "code relies heavily on original remote code.", it is exact the same reference code that I used
@IlyasMoutawwakil @echarlaix could you please merge if there is no any additional comments? |
What does this PR do?
Fixes # (issue)
Before submitting