remove num_candidates #301
Open
Conversation
This file contains 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
|
This actually supports an important feature - it's not publicly documented yet, but will be soon. |
|
Ok thanks @maciejkula would you like me to close it then? |
|
Hi @maciejkula, friendly nudge here... seems like this change makes things a bit clearer (since it seems like num_candidates == num_queires) but understood that you're suggesting that will eventually not be the case and you want to keep this. |
|
Yes, please! |
|
@evanfwelch with this proposed change we wouldn't be able to do Mixed Negative Sampling. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Context
One nice feature of the
tfrs.tasks.retrievaltask is how simple it is. Especially how it dot products equal-sized of query and candidate embeddings and internally computes the "labels" as the identity matrix.This PR
num_candidatesasscores.shape[1]. The task doesn't support any kind of non-square score matrix as far as I can tell and having this extra degree of freedom actually made it harder for me to understand what was happening.I would humbly suggest removing these bits for now and adding back in if there becomes a more general way to pass in candidates of a different shape.