Executing flare-otf lammps.yaml on the command line results in an error

[915936476]

Error occurred while attempting to train FLARE force field on-the-fly with LAMMPS MD according to the tutorial.

May I ask which step of the configuration error occurred and how to resolve it?

[jonpvandermause]

I'm having trouble reproducing this error. To confirm, you're referring to this tutorial, right? Can you give some details about your setup (lammps version, flare commit, operating system)?

A few general suggestions:

Make sure you're installing the latest release of lammps using the following command:
git clone --depth 1 --branch release https://github.com/lammps/lammps.git lammps

We've had some issues with recent lammps commits, but the release branch should be fine. Make sure you're running install.sh to copy in the flare pair style, as described here, and building lammps with the following flags:
cmake ../cmake -DPKG_MACHDYN=yes -DDOWNLOAD_EIGEN3=yes -DPKG_MANYBODY=ON

Also make sure you are patching ASE as described in #360. This is necessary for recent versions of lammps, which aren't compatible with the ASE 3.22 lammps parser.

If you do all of the above, and are using the master branch of flare, I think the tutorial should work.