Evaluate test sets separately for different heads

ACEsuit / mace

MACE - Fast and accurate machine learning interatomic potentials with higher order equivariant message passing.

Other

550 stars 205 forks source link

Evaluate test sets separately for different heads #681

Closed ThomasWarford closed 1 week ago

ThomasWarford commented 1 week ago

See issue.

I am not sure what the desired behaviour is here. Do we want the user to be able to specify 'head_key'='head_name' in the input xyz files? This fix would override that.

I also need to check that the correct heads are being evaluated in the tables - has this previously been tested?