feat: add regression

[stephantul]

This PR adds a regression capability to model2vec training. This is actually just what we used to call a similarity pipeline, but with an MSE loss. Because the pipelines are the same except for the training loss, it's a relatively small change.

Of note is that I also refactored some things:

1. Regression is now the base case, similarity is a case of regression.
2. The cosine loss is now a module, not a simple function.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
model2vec/train/__init__.py	100.00% <100.00%> (ø)
model2vec/train/classifier.py	98.13% <100.00%> (ø)
model2vec/train/lightning_modules.py	98.57% <100.00%> (+0.15%)	⬆️
model2vec/train/regression.py	100.00% <100.00%> (ø)
model2vec/train/similarity.py	100.00% <100.00%> (ø)
model2vec/train/utils.py	100.00% <100.00%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Pringled]

Clean, LGTM!