Olivier Grisel

I launched the CUDA GPU CI at: - https://github.com/scikit-learn/scikit-learn/actions/runs/9402808393 EDIT: CUDA tests are green.

Merged `main` and labeled with "CUDA CI" to check that the tests still pass with the recent CI config and dependency updates.

I would convert to float.

Explicit is better than implicit. I guess this applies in this case. +1 for a PR that adds `classes` as an optional hyperparam to all classifiers and a common test...

I would need a changelog entry for 1.5.1 but the changelog file needs to be updated in `main` to add the missing section first. Let's do that in another PR....

In `main`, passing `scoring="roc_auc"` does not result in constant scores as I would have intuitively expected: the ROC AUC metric is evaluated on binary predictions for all possible thresholds instead...

> My main conclusion is that we should make an official distinction of scores/metrics: probabilistic ones and decision metrics. There are also: - non-probabilistic classification confidence scores (e.g. `decision_function` of...

> I don't like that here we're accessing private _response_method as a proxy to see if it's the right kind of scorer or not. Scorers should have a public API...

I updated the PR to issue an informative warning for a [pitfall Christian and I fell into when working on this the release highlights for this new class](https://github.com/scikit-learn/scikit-learn/pull/29056#discussion_r1608095762). Arguably, passing...

I also added a TODO comment on the line in the code that would deserve to be updated once we improve the scoring API in 5a464e4. Please let me know...