Adding ContrastiveOutput #1191
Merged
Adding ContrastiveOutput #1191
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
Documentation preview |
gabrielspmoreira
approved these changes
Jul 11, 2023
| assert isinstance(dot.to_call, DotProduct) | ||
|
|
||
| target = ContrastiveOutput(schema=Schema([item_id_col_schema])) | ||
| assert isinstance(target.to_call, CategoricalTarget) |
There was a problem hiding this comment.
This is neat! Same interface for contrastive for two-tower like architectures and for sampled softmax.
| from merlin.models.torch.block import registry | ||
|
|
||
|
|
||
| class LogUniformSampler(object): |
There was a problem hiding this comment.
@marcromeyn Have you checked the LogUniformSampler class I created in T4Rec based on this one?
I adds some additional options like returning unique samples and also provides the probs for the items for logQ correction (considering whether returning unique samples or popularity biased).
It matches the implementation of sampling and probs from tf.random.log_uniform_candidate_sampler
There was a problem hiding this comment.
I did not, yours looks much better! I can port that
| self.false_negative_score = false_negative_score | ||
|
|
||
| @classmethod | ||
| def with_weight_tying( |
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Outdated suggestions cannot be applied.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.
Goals ⚽
This PR builds on the
CategoricalOutputand addsContrastiveOutput. This can be used with dot-product, categorical-prediction or weight-tying.Implementation Details 🚧
The contrastive part of
CategoricalOutputdoesn't work with torch-script, this is fine since it's during training only.