Luthaf / rascaline

Computing representations for atomistic machine learning
https://luthaf.fr/rascaline/
BSD 3-Clause "New" or "Revised" License
44 stars 13 forks source link

Use the input dtype and device for the output in rascaline-torch #278

Closed Luthaf closed 8 months ago

Luthaf commented 8 months ago

All calculations still happen on CPU with 64-bit floats, but the data is automatically converted on input and output as required.

Should help with https://github.com/lab-cosmo/metatensor-models/pull/58


📚 Documentation preview 📚: https://rascaline--278.org.readthedocs.build/en/278/

github-actions[bot] commented 8 months ago

Here is a pre-built version of the code in this pull request: wheels.zip, you can install it locally by unzipping wheels.zip and using pip to install the file matching your system