show_progress_bar as parameter on predict and predict_prob

[davidsbatista]

• I want to deploy Setfit in production without having the progress bar shown when any of the predict methods is called, it seems there's no explicit way to do this unless changing the logging level which sentence-transformer
  uses to decide on the show_progress_bar boolean value

• Adding show_progress_bar as a parameter on SetFitModel.predict and SetFitModel.predict_prob methods

[davidsbatista]

The PR is now failing due to the message from flake8

src/setfit/exporters/onnx.py:156:8: E721 do not compare types, for exact checks use is / is not, for instance checks use isinstance()

which was not something my PR introduced, I can nevertheless fix it

    # If the datatype of the model is double we need to cast the outputs
    # from the setfit model to doubles for compatibility inside of ONNX.
    if type(dtype) == onnxconverter_common.data_types.DoubleTensorType:
        sklearn_model = Pipeline([("castdouble", CastTransformer(dtype=np.double)), ("head", model_head)])
    else:
        sklearn_model = model_head

[tomaarsen]

I'll tackle the unrelated quality check in a separate PR. Thanks for this!

[davidsbatista]

OK, if you want I can then merge/sync this with main, just let me know when that separate quality check issue is solved/merged

[tomaarsen]

No worries, I'll handle it from here. I'll merge this as soon as the tests pass (which they should), despite the quality failure.

[tomaarsen]

Thanks again!

[davidsbatista]

happy to have contributed :)