-
Notifications
You must be signed in to change notification settings - Fork 1
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
Implement Rotary Positional Embeddings #73
Comments
Open
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Rotary Positional Embeddings are a state of the art positional embedding that encodes both absolute and relative positional encodings using rotations. They seem to help a lot across a bunch of tasks from NLP to Computer Vision. The folks at EPFL have implemented in their tracking library and it could be useful for us to do the same. We should implement it as a function of the
dreem.models.Embedding
class similar to how we did with fixed/learned embeddings.EDIT (Mustafa):
Add unit test in test_models.py
The text was updated successfully, but these errors were encountered: