Simplify hyperparameter optimization for MBTR

globus-labs / faster-molecular-hessians

Reduce the amount of computation needed for Hessians using machine learning

MIT License

0 stars 1 forks source link

Simplify hyperparameter optimization for MBTR #9

Closed WardLT closed 7 months ago

WardLT commented 9 months ago

We should combine the model and descriptor step into a single Pipeline model so that you can run hyperparameter optimization using standard sklearn tools. This will require making MBTR into a BaseEstimator class

WardLT commented 7 months ago

Fixed in newer notebooks