Benjamin Bossan

> So instead of adding it automatically, we could have a method like `hub_utils.add_file` which can also accept `__file__` as input value. I'll take a look at that

See the comment linked above. So for example for text classification and text regression, we currently only allow lists of str, but scikit-learn is more flexible there, allowing, for instance,...

I don't quite understand the issue. `ColumnTransformer` would just be part of the model `Pipeline` and as such of the final model object, no?

I think since `init` already creates files, it would not hurt to create a minimal README.md. If users want to add a proper model card, it wouldn't prevent them from...

Issue opened here: https://github.com/huggingface/api-inference-community/issues/94

From a user perspective, I would probably not want to change my model pipeline to accommodate for potential dtype issues when the model is hosted on the hub. Ideally, I...

Okay, I think we can agree on having both options (which should share most of the important code anyway). Just another reason for not modifying the pipeline: Say I create...

@skops-dev/maintainers Could you help me figure out why the new tests are failing? The offending code seems to stem from this fixutre: https://github.com/BenjaminBossan/skops/blob/adding-files/skops/hub_utils/tests/test_hf_hub.py#L497-L510 which results in the error `OSError: Model...

@skops-dev/maintainers The PR is ready for review now. The error mentioned above was caused by `test_inference` doing some cleanup that would disrupt tests running later. We use some session-scoped fixtures,...

@adrinjalali Yes, I was thinking about making files `*args` too, I changed it as suggested. Furthermore, I added an `exist_ok` argument (default `True`). If the file exists and `exist_ok=True`, I'll...