Skip to content

Continual Transformer Encoder #317

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

Merged
merged 49 commits into from
Dec 8, 2022
Merged

Conversation

LukasHedegaard
Copy link
Collaborator

New tool for efficient application of Transformer Encoders to time-series data.
Note: this does not come with pre-trained models. The user must fit the learner on her own features.

@LukasHedegaard LukasHedegaard changed the base branch from master to develop September 26, 2022 11:21
@LukasHedegaard LukasHedegaard added enhancement New feature or request test sources Run style checks test tools Test the toolkit methods labels Sep 26, 2022
@ad-daniel ad-daniel marked this pull request as draft October 10, 2022 06:41
@ad-daniel
Copy link
Collaborator

I assume this PR is still ongoing, I've converted it to draft for now

@LukasHedegaard LukasHedegaard marked this pull request as ready for review October 10, 2022 06:54
@LukasHedegaard
Copy link
Collaborator Author

I assume this PR is still ongoing, I've converted it to draft for now

It is ready for review

Copy link
Collaborator

@ad-daniel ad-daniel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you, only a few points

  • I've already committed some formatting adjustments.
  • ensure the default values in the code are indeed the desired ones and adapt the docs/comments accordingly
  • add a changelog entry mentioning the added tool
  • if possible, provide a demo that shows how to use it (like you did in the docs, can be the same one)
  • Concerning the project structure, my understanding is that it's a self-contained module. You've made (rightly so) an ad hoc folder for this new tool in perception/activity_recognition and test folders, but documentation wise you've put everything in the same existing document. I think it would be better to split the documentation of x3d and this tool into 2 separate documents. Basically as we did for other stuff like speech recognition (see link), and each document should be referenced in the index here : https://github.com/opendr-eu/opendr/blob/master/docs/reference/index.md
  • I don't think this was among the planned tools, so there isn't a strict requirement, but having a ROS/ROS2 node would be a plus. Up to you.

@LukasHedegaard
Copy link
Collaborator Author

My bad. I've updated the CHANGELOG and added a demo. I will not be providing a ROS node at this time.

Copy link
Collaborator

@ad-daniel ad-daniel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sorry, last detail, could you add also a README.md file to the demos folder with some instructions? other than that it's fine for me

@ad-daniel ad-daniel mentioned this pull request Dec 7, 2022
ad-daniel
ad-daniel previously approved these changes Dec 7, 2022
Copy link
Collaborator

@ad-daniel ad-daniel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you!

@LukasHedegaard
Copy link
Collaborator Author

Thank you!

Likewise :-)

Copy link
Collaborator

@tsampazk tsampazk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you Lukas! I left some comments regarding some minor typos, etc.

…l_transformer_encoder/README.md

Co-authored-by: Kostas Tsampazis <[email protected]>
@LukasHedegaard
Copy link
Collaborator Author

Thank you Lukas! I left some comments regarding some minor typos, etc.

Thanks for the thorough review, @tsampazk . I've accepted all your suggestions and fixed typos as requested.
The code itself was left unchanged, so these changes should not affect anything you reviewed earlier, @ad-daniel.

Copy link
Collaborator

@tsampazk tsampazk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you!

Copy link
Collaborator

@ad-daniel ad-daniel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you!

@LukasHedegaard LukasHedegaard merged commit 935bc3c into develop Dec 8, 2022
@LukasHedegaard LukasHedegaard deleted the continual-transformer branch December 8, 2022 09:39
@ad-daniel ad-daniel mentioned this pull request Dec 8, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request test sources Run style checks test tools Test the toolkit methods
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants