Disable TF-32 tensor cores for VAMPNET training

[yuxuanzhuang]

Fix #220

Here's a quick checklist that should be present in PRs:

• Make sure to include one or more tests for your change
• Add yourself to AUTHORS
• Add a new entry to the doc/source/CHANGELOG. Decide whether your change is a fix or a new feature.

[codecov]

Codecov Report

Merging #222 (ba294d5) into main (590f67e) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #222   +/-   ##
=======================================
  Coverage   93.74%   93.75%           
=======================================
  Files         136      136           
  Lines       10807    10816    +9     
=======================================
+ Hits        10131    10140    +9     
  Misses        676      676           

Impacted Files	Coverage Δ
deeptime/decomposition/deep/_vampnet.py	94.73% <100.00%> (+0.04%)	⬆️
deeptime/util/torch.py	93.49% <100.00%> (+0.39%)	⬆️

---

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 590f67e...ba294d5. Read the comment docs.

[clonker]

Hey thank you very much for the contribution! 🎉

Two small suggestions: can we call it disable_TF32 instead of the good old camel case disableTF32? Also you'll have to add a

pytest.importorskip("torch")

at the top of the unit testing file, otherwise the tests fail for jobs which do not have torch installed (py310 currently)

[clonker]

Thanks!