feat: Batch embeddings, upgrade model2vec

[Pringled]

This PR upgrades model2vec to 0.10 and embeds new chunks in batches of 1024 across files instead of once per file. This leads to some nice speedups. Putting everything together since 0.6.2 we get these results:

27k files	0.6.2	main (model2vec 0.9)	main + model2vec 0.10	main + model2vec 0.10 + batched embedding (this PR)
Cold build	91s	50s	42s	36s
Warm query	13.4s	1.9s	-	-

Cold builds are ~2.5x faster and warm queries ~7x faster on large repos.

I also reran all the benchmarks since this PR + the previous ones all affect the speed of semble.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/semble/index/create.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.

[greptile-apps]

Confidence Score: 5/5

[High impact] The PR appears safe to merge; the remaining test-coverage suggestion is non-blocking.

_{Reviews (2) · Last reviewed commit: "Check every embedding row across batch b..." · Reviewed by Greptile}