# Notebooks for Differentiable Programming Tutorial

Christian Legaard published on

1 min,
197 words

Over the summer I was employed as a PhD Intern at *Pacific Northwest National Laboratory* in Washington.
They were kind enough to sponsor my attendance to Annual Modeling and Simulation Conference 2022 where I delivered a tutorial on differentiable programming.

As part of the tutorial I wrote a couple of workbooks as an introduction to differentiable programming. Specifically the notebooks cover:

- Basic use of
`jax.grad`

,`jax.vmap`

, and how to implement a simple gradient descent optimization scheme to solve a linear regression problem. - An exercise in how to use differentiable programming to solve geometric constraints.
- How to define ordinary differential equations in Python, how to implement a solver in Jax, how to tune the parameters of an ODE using gradient descent, and how to implement and train a neural ODEs model in pure Jax.

The notebooks can be opened in Google Colaboratory through the links: