ActsPlugins::OnnxEdgeClassifier::Config Struct Reference

Configuration for OnnxEdgeClassifier. More...

#include <ActsPlugins/Gnn/OnnxEdgeClassifier.hpp>

Collaboration diagram for ActsPlugins::OnnxEdgeClassifier::Config:

Public Attributes
float	cut = 0.5
	Classification threshold cut.
Device	device = Device::Cuda()
	Device to allocate the model on.
std::vector< float >	featureScales = {}
	Vector of per-feature scale factors applied to selected features.
std::string	modelPath
	Path to the ONNX model file.
std::vector< int >	selectedFeatures = {}
	Vector of selected node feature indices to use as input for the model.

Detailed Description

Configuration for OnnxEdgeClassifier.