neuralop.models.base_model.get_model

neuralop.models.base_model.get_model(config)[source]

Returns an instantiated model for the given config

• Reads the model to be used from config[‘arch’]

• Adjusts config[“arch”][“data_channels”] accordingly if multi-grid patching is used

Also prints warnings for safety, in case:: * some given arguments aren’t actually used by the model * some keyword arguments of the models aren’t provided by the config

Parameters:

configBunch or dict-like

configuration, must have arch = config[‘arch’] (string) and the corresponding config[arch] (a subdict with the kwargs of the model)

Returns:

modelnn.Module

the instanciated module