-
Notifications
You must be signed in to change notification settings - Fork 0
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Ase calculator #2
Conversation
Also fixes CI in here |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This looks very good! There are a couple of ways you could integrate better with existing module for logging/cache directory management/... but the code is fine and should work fine!
"ShiftML1.0": "https://tinyurl.com/3xwec68f", | ||
"ShiftML1.1": "https://tinyurl.com/53ymkhvd", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
so you re-created the exact previous models using the new infrastructure? Should these have a different version (like ShiftML1.0+metatensor
) just for clarity or are you sure this will produce the same outputs?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I will discuss this with everyone involved - but in principle I agree, that there should be a ShiftML1.0rev or something that makes it clear.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I will open an issue to remind me of that.
"""Regression test for the ShiftML1.0 model.""" | ||
|
||
frame = bulk("C", "diamond", a=3.566) | ||
model = ShiftML("ShiftML1.0", force_download=True) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why force_download=True
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This was to make sure that the local tests do not resuse a local potentially old cached file.
-uses logging class -uses platformdirs for cache path resolution -asserts correct rascaline.torch version -checks for correct rascaline.torch -model files named after model version -tests for correct assertion in invalid species test
Adds the first very basic functionality to ShiftML: a chemical shielding ase calculator.
The calculator automatically downloads an exported metatensor model from a zenodo archive and caches it locally.
tests. Explanations are added to the README.