[fix] blocksparse sanity checks

[blefaudeux]

What does this PR do?

Fixes #206, constraints the block sizes to being the ones we actually uni tests. Other sizes probably work but bring no benefit really, since you can express the pattern with the smaller blocks

Before submitting

• Did you have fun?
  • Make sure you had fun coding 🙃
• Did you read the contributor guideline?
• Was this discussed/approved via a Github issue? (no need for typos, doc improvements)
  • N/A
• Did you make sure to update the docs?
  • N/A
• Did you write any new necessary tests?
  • N/A
• Did you update the changelog? (if needed)
  • N/A

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

[codecov-commenter]

Codecov Report

Merging #207 (d992171) into main (8008a3f) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #207   +/-   ##
=======================================
  Coverage   91.62%   91.62%           
=======================================
  Files          59       59           
  Lines        2998     2998           
=======================================
  Hits         2747     2747           
  Misses        251      251           

Flag	Coverage Δ
Python	91.62% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
xformers/components/attention/blocksparse.py	94.20% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 8008a3f...d992171. Read the comment docs.

[fmassa]

LGTM!

With #202 merged, my idea would be to eventually deprecate this module, as it is already covered by the scaled_dot_product_attention. But we can discuss about this next week

[blefaudeux]

LGTM!

With #202 merged, my idea would be to eventually deprecate this module, as it is already covered by the scaled_dot_product_attention. But we can discuss about this next week

It's not there yet, and the issue was present right now..