Introducing Flax NNX (Part 1)

From the creator

We'll be diving into Flax NNX, a new neural network library for JAX that aims to simplify machine learning development. We will explore its core concepts, compare it with PyTorch, and see how it fits into the broader JAX ecosystem. This is the first episode of a three part series introducing Flax NNX. In this episode we’ll explain its core philosophy, how models are structured, and how state is managed. We’ll focus on the "what" and "why" of NNX from a Python developer's perspective. Resources: Learn more → https://goo.gle/learning-jax Subscribe to Google for Developers → https://goo.gle/developers Speaker: Robert Crowe

Published 2025-12-04 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.