Integrate mme5 models

[Aradhye2002]

Changes/additions

mteb/models/mme5_models.py
mteb/models/mme5

Explanation for adding the mme5 folder:

haon-chen/mmE5 doesn't expose the the repository as a pip installable package, hence it is necessary to have the required model classes and functions (along with their dependencies) available locally. One option could have been to dump all these files in mme5_models.py itself, however, due to the large number of lines of code, this approach has been adopted.

Closes #2079

Code Quality

• Code Formatted: Format the code using make lint to maintain consistent style.

Documentation

• Updated Documentation: Add or update documentation to reflect the changes introduced in this PR.

Testing

• New Tests Added: Write tests to cover new functionality. Validate with make test-with-coverage.
• Tests Passed: Run tests locally using make test or make test-with-coverage to ensure no existing functionality is broken.

Adding a model checklist

• I have filled out the ModelMeta object to the extent possible
• I have ensured that my model can be loaded using
  • mteb.get_model(model_name, revision) and
  • mteb.get_model_meta(model_name, revision)
• I have tested the implementation works on a representative set of tasks.

[Aradhye2002]

It seems the mmE5 repo is not installable via pip install. Can't seem to think of a way of using the MMEBModel, ModelArguments and load_processor from their repository (which is required for loading the model) without explicitly adding their repository to the mteb repo and then changing the PYHTHONPATH. This seems quite inelegant and I'm open to possible alternative solutions.

[Samoed]

You can copy their model class and other functions from library

[Aradhye2002]

You can copy their model class and other functions from library

I see. Should I add the helper functions and model classes to the mme5_models.py (created) file itself?

[Samoed]

Yes, I think this is best option

[Samoed]

You should copy only required classes, not full repo

[Aradhye2002]

You should copy only required classes, not full repo

I've only kept the required classes and functions and deleted the rest (only the directory structure has been maintained). Let me know if you still want them in the same file ie. mme5_models.py

[Samoed]

This seems a lot of files. I think we can wait until they convert to library or integrated to sentence transformers

[Aradhye2002]

This seems a lot of files. I think we can wait until they convert to library

Noted.