Maximiliano Vargas

@bellet @perimosocordiae @wdevazelhes @terrytangyuan There needs to be a broader discussion regarding adding BilinearSimilarity and Bilinear Similarity Learners such as OASIS. By definition bilinear similarity works in the opposite direction...

@wdevazelhes I like the idea of `pairs_remoteness` and `pairs_similarity`, because it adds one layer of abstraction, making the code more generic and being able to support metrics and similarites without...

> This is always a feature we can add later if we feel it is needed. @mvargas33 I think we have some kind of consensus now so you may go...

@bellet @perimosocordiae @wdevazelhes @terrytangyuan This branch is ready for a code review. It would be cool if can spare some time to look at it. Here's a summary of the...

> Here's a review for the documentation part. > > Generally we need to change a bit the vocabulary to accommodate both Mahalanobis and bilinear similarities. As common in the...

- I ran the spellcheck and fixed a few typos. - I moved `test_pair_score_dim` and `test_deprecated_score_pairs_same_result` to `test_base_metric.py` - I refactored `test_same_with_or_without_preprocessor` to run with all metric learners and used...

> Great to see the formal process documented. Let me know when this is ready for review. Hello! The branch is ready for a review. I added a last section...

I just confirmed it, it actually works fixing the bug

I need this !

Same issue here, with 2.52.0 on Vite