TorchDyn reference documentation

(New release 0.2.0 is finally out!)

End-to-end Pytorch suite for continuous neural architectures featuring several models, training methods and visualization tools for research, industry and amateurs.

git clone https://github.com/DiffEqML/torchdyn.git

cd torchdyn

python setup.py install

Please visit our GitHub page.

Note

• This library is developed and maintained by Michael Poli & Stefano Massaroli.
• The work has been supported by Prof. Jinkyoo Park (KAIST), Prof. Atsushi Yamashita (The University of Tokyo) and Prof. Hajime Asama (The University of Tokyo).

Basic Tutorials

• TorchDyn Quickstart
• The Cookbook of Neural ODEs
• Image Classification
• Approximating a Reflection Map

Advanced Tutorials

• Augmentation Strategies
• Trajectory Tracking with Integral Loss
• Higher-Order Neural ODEs

Generative Models

• Density Estimation with Neural ODEs
• Density Estimation with FFJORDs

Energy-Based Models

• Hamiltonian Neural Networks
• Lagrangian Neural Networks
• Stable Neural ODEs (Stable Neural Flows)

Graph Models

• Neural Graph Differential Equations

API documentation

• Public API: torchdyn package

Notes

• Change Log

Indices and tables

• Index
• Module Index
• Search Page