Matthews correlation coefficient

[sanderland]

See #317

• Refrained from adding 'micro' as an option for averaging, as I think it has too many pitfalls.
• Would like to test a bit better, but not sure where to add this

[HuggingFaceDocBuilderDev]

The documentation is not available anymore as the PR was closed or merged.

[lvwerra]

Thanks for adding this! The tests for metrics are in the docstrings itself. So if you add an example or two to the docstring the will be tested by the CI. Also to fix the style tests you can run make style && make quality.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint.

[sanderland]

Should be all ready now @lvwerra

[lvwerra]

Code looks good! Would you mind also updating the README with the new options?

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint.

[sanderland]

@lvwerra gave it a shot, but there's not a lot of examples to go on as the multilabel config option is quite under-documented in general.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint.

[lvwerra]

Thank you for adding this!