Models

The models submodule implements common neural network architectures in pytorch to work with eyetracking data. There are two main classes - Classifier and Regressor, to solve classification and regression tasks, respectively. They are wrappers upon the backbone parameter, which could be built using pytorch-implemented blocks.

Building Blocks

The blocks implemented in models to construct the backbone:

Backbones

Models

Functions

Methods that combine above classes and wrappers.