concepts

Description

Models and handles for concept embedding analysis and model extension. This includes handles for

• Defining concept models

• Training and evaluating concept models

• Conducting a series of concept embedding analysis experiments

Sub-modules

analysis	Handles for conducting concept embedding analysis experiments.
concepts	Concept base class, concept types, and corresponding sub-classes.
models	Models and handles to define and store concept embeddings.
train_eval	Routines for training and evaluation of pytorch DNNs with callbacks and calibration.