Welcome to MultiVae’s documentation!

This library implements some of the most common Multimodal Variational Autoencoders methods in a unifying framework for effective benchmarking and development. Multimodal Variational Autoencoders are versatile models that can be used to learn joint representations of multiple modalities and generate new samples for prediction or data-augmentations tasks. Our library provides a simple and flexible way to define, train and evaluate these models. It is designed to be easy to use, even by non-specialists and even on incomplete datasets. For easy benchmarking, we include ready-to-use datasets and metrics modules. It integrates model monitoring with Wandb and a quick way to save/load model from HuggingFaceHub 🤗. To improve joint generation of multimodal samples, we also propose samplers to explore the latent space of your model.

_images/diagram2.png

Basics

Documentation

Setup

To install the latest stable release of this library run the following :

$ pip install multivae

To install the latest version of this library run the following using pip

$ pip install git+https://github.com/AgatheSenellart/MultiVae.git

or alternatively you can clone the github repo to access to tests, tutorials and scripts.

$ git clone https://github.com/AgatheSenellart/MultiVae.git

and install the library

$ cd MultiVae
$ pip install -e .

If you clone the MultiVae’s repository you will access to the following:

  • docs: The folder in which the documentation can be retrieved.

  • tests: multivae’s unit-testing using pytest.

  • examples: A list of ipynb tutorials and scripts describing the main functionalities of multivae.

  • src/multivae: The main library which can be installed with pip.