TedNet: A Pytorch Toolkit for Tensor Decomposition Networks

TL;DR

TedNet is a flexible PyTorch toolkit that enables easy construction and experimentation with various Tensor Decomposition Networks (TDNs) for deep learning models, promoting research in compact neural architectures.

Contribution

The paper introduces TedNet, a PyTorch toolkit implementing five tensor decomposition methods to facilitate the development of diverse TDN architectures.

Findings

Supports five tensor decomposition types: CP, BTT, Tucker-2, TT, TR

Enables easy construction of TDNs using basic layers

Promotes research in compact neural network architectures

Abstract

Tensor Decomposition Networks (TDNs) prevail for their inherent compact architectures. To give more researchers a flexible way to exploit TDNs, we present a Pytorch toolkit named TedNet. TedNet implements 5 kinds of tensor decomposition(i.e., CANDECOMP/PARAFAC (CP), Block-Term Tucker (BTT), Tucker-2, Tensor Train (TT) and Tensor Ring (TR) on traditional deep neural layers, the convolutional layer and the fully-connected layer. By utilizing the basic layers, it is simple to construct a variety of TDNs. TedNet is available at https://github.com/tnbar/tednet.