Training force with frozen atoms?
Issue #201
resolved
I'm working on surface system, where some of the atoms have to be frozen. The problem is the force for these atoms were set to zero by VASP or other calculators, which will cause serious mistake when training such force data. Can someone solve this problem, or give the guide to do this?
Comments (2)
-
repo owner -
repo owner - changed status to resolved
- Log in to comment
Can you post this question on the amp-users mailing list? If you are not signed up, you can do so here:
https://listserv.brown.edu/?A0=AMP-USERS
I'm closing this issue as it's not a bug with Amp.