-
Notifications
You must be signed in to change notification settings - Fork 489
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 for custom input shapes in exporters onnx #575
Support for custom input shapes in exporters onnx #575
Conversation
The documentation is not available anymore as the PR was closed or merged. |
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.
Thanks @fxmarty!!
Just left a couple of nits.
Let's wait for the stable diffusion PR to be merged before this. |
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
…b.com/fxmarty/optimum into exporters-onnx-variable-inputs-shapes
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.
Left a few comments, but besides that it looks great, thanks @fxmarty !
@@ -31,4 +31,4 @@ jobs: | |||
- name: Test with unittest | |||
working-directory: tests | |||
run: | | |||
RUN_SLOW=1 pytest exporters --durations=0 | |||
pytest exporters -s --durations=0 |
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.
Are you sure you want to output everything in the log?
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.
yes, why not?
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!
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. |
This will be mainly useful for testing, but could also solve temporarily issues on models where the export is dependent on the input shape (for example longformer (but it's actually a bug in PyTorch), and longt5.
Related issues are linked in #503
This PR follows #522
Edit: let's wait for #573 to be merged to clean up
Before submitting