Support AIMNet2 #64
Draft
Support AIMNet2 #64
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Is this PR blocked by something on the AIMNet2 or TorchMD-Net sides? |
|
It's blocked on the issues described above. AIMNet2 doesn't yet support cutoffs or periodic boundary conditions, and it isn't yet conda installable. |
|
@peastman I have recently added aimnet2 to conda-forge as from pyaimnet2 import load_model
model = load_model("wb97m-d3") # can also load b973cI hope this helps! |
|
Thanks! I'll give it a try. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Implements #63.
This is a first draft of AIMNet2 support. The current implementation assumes that you have downloaded the file aimnet2_wb97m-d3_ens.jpt from https://github.com/isayevlab/AIMNet2 and it's in your current working directory. Once conda packages become available, I'll modify it to use them instead. AIMNet2 does not currently support cutoffs or periodic boundary conditions. Hopefully they will be coming soon.
To test the performance, I tried simulating alanine dipeptide in vacuum (22 atoms) and compared it to ANI-2x. It takes 25.9 ms per step, compared to 0.97 ms for ANI. Hopefully there's room for it to be optimized.