Training strategies in MLFF

Question on input files/tags, interpreting output, etc.

Please check whether the answer to your question is given in the VASP online manual or has been discussed in this forum previously!

Moderators: Global Moderator, Moderator

Locked
Message
Author
jun_yin2
Newbie
Newbie
Posts: 25
Joined: Wed Jul 20, 2022 7:18 am

Training strategies in MLFF

#1 Post by jun_yin2 » Mon Jun 05, 2023 8:20 am

Dear all,

I have questions about training in Machine learning force field. If I want to explore more local reference configurations during temperature ramp from 300 K to 500 K, in two strategies, which training strategy is a better choice?
1. First train a force field from 300K to 400K by 10 ps with 0.5 fs per step, then continue training from 400K to 500K by 10 ps with 0.5 fs step.
2. First train a force field from 300K to 500K by 10 ps with 0.5 fs per step, then continue training from 300 K to 500K again with the same initial structure by 10 ps with 0.5 fs per step.

I hope you could help me. For now, I think the first may be a better choice. But I could not confirm.

marie-therese.huebsch
Full Member
Full Member
Posts: 175
Joined: Tue Jan 19, 2021 12:01 am

Re: Training strategies in MLFF

#2 Post by marie-therese.huebsch » Mon Jun 05, 2023 12:09 pm

Hi,

Dividing long trajectories into smaller parts is always a good idea, i.e., use option 1. One reason is in an NpT ensemble, the volume would likely change during the simulation, and it is good to reinitialize the PAW basis for the electronic minimization.

But this won't significantly affect exploring more local reference configurations. Instead, you should look into parameters such as ML_SCLC_CTIFOR, ML_CX, ML_EPS_REG, etc.

Marie-Therese

jun_yin2
Newbie
Newbie
Posts: 25
Joined: Wed Jul 20, 2022 7:18 am

Re: Training strategies in MLFF

#3 Post by jun_yin2 » Mon Jun 05, 2023 12:53 pm

Very thanks to your reply!

jun_yin2
Newbie
Newbie
Posts: 25
Joined: Wed Jul 20, 2022 7:18 am

Re: Training strategies in MLFF

#4 Post by jun_yin2 » Mon Jun 05, 2023 1:08 pm

Hi,

I also want to ask about that whether ML_SCLC_CTIFOR can only be used in ML_MODE=select?

ferenc_karsai
Global Moderator
Global Moderator
Posts: 422
Joined: Mon Nov 04, 2019 12:44 pm

Re: Training strategies in MLFF

#5 Post by ferenc_karsai » Mon Jun 05, 2023 1:51 pm

ML_SCLC_CTIFOR is always applied when local reference configurations are selected.
That means for ML_MODE=TRAIN and ML_MODE=SELECT.

Locked