This repository is an update to all previous repositories with an implementation of various Quaternion-valued Neural Networks in PyTorch
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
KODAK
utils
LICENSE
README.md
cae.py
convolutional_models.py
copy_task.py
quaternion_layers.py
quaternion_ops.py
recurrent_models.py

README.md

pytorch-quaternion-neural-networks

This repository contains source codes and many examples of quaternion neural networks applications, including convolutional (QCNN) (https://arxiv.org/abs/1806.07789), and gated or non gated recurrent (QRNN / QLSTM) (https://arxiv.org/abs/1806.04418)

Requirements

A GPU with a running CUDA installation is preferable.

pip install torch torchvision matplotlib      

Usage

For an example of QRNN / QLSTM / RNN / LSTM on a simple copy-task:

python copy_task.py {blank_size, default=25}        

or for an example of QCAE / CAE ((Q)-Convolutional autoencoders) on a gray-scale to color image reconstruction with the KODAK dataset:

python cae.py {QCAE,CAE}