Differentiable controlled differential equation solvers for PyTorch with GPU support and memory-efficient adjoint backpropagation.
neural-ode time-series pytorch torchcde code library

This library provides differentiable GPU-capable solvers for controlled differential equations (CDEs). Backpropagation through the solver or via the adjoint method is supported; the latter allows for improved memory efficiency.

In particular this allows for building Neural Controlled Differential Equation models, which are state-of-the-art models for (arbitrarily irregular!) time series. Neural CDEs can be thought of as a "continuous time RNN".

Powered by the torchdiffeq library.

Don't forget to tag @patrick-kidger in your comment, otherwise they may not be notified.

Authors community post
A mathematician who enjoys martial arts, ice skating and scuba diving.
Share this project
Similar projects
Forecasting the weather with neural ODEs
Interesting blog post applying neural ODEs to the problem of weather forecasting.
Neural ODE Explained
Explains "Neural Ordinary Differential Equations", a very interesting idea came out in NIPS 2018.
Top collections