Open MatSci ML Toolkit is a framework for prototyping and scaling out deep learning models for materials discovery supporting widely used materials science datasets, and built on top of PyTorch Lightning, the Deep Graph Library, and PyTorch Geometric.
This PR refactors the experiment configs. Previously, the config files needed to be present in their respective directories, making it hard to separate out config files for different groups of experiments. This PR moves all the default configs into a folder called configs with subfolders for each config type. Now, the user may create folders for their own specific experiments, use a combination of default config files and custom configs by specifying a path to each respective config in the cli arguments.
This PR refactors the experiment configs. Previously, the config files needed to be present in their respective directories, making it hard to separate out config files for different groups of experiments. This PR moves all the default configs into a folder called
configs
with subfolders for each config type. Now, the user may create folders for their own specific experiments, use a combination of default config files and custom configs by specifying a path to each respective config in the cli arguments.