Closed InfluenceFunctional closed 3 months ago
Currently we pass many configs explicitly through the various models, into molecule_graph_model, and onto GraphNeuralNetwork. Probably this could be simplified with kwargs.
Not sure if kwargs is actually a responsible approach as it may leave room for mistakes & omissions. Need to be careful.
Fixed in equivariant_implement
Currently we pass many configs explicitly through the various models, into molecule_graph_model, and onto GraphNeuralNetwork. Probably this could be simplified with kwargs.