Jax – Composable transformations of Python and NumPy programs

Hello from the JAX team! We'd like to take this opportunity to give a shout out to some of the awesome projects folks are building on top of JAX, e.g., * Flax, a neural network library for JAX (https://github.com/google/flax) * Haiku, a neural network library for JAX inspired by Sonnet (https://github.com/deepmind/dm-haiku) * RLax, a library for building reinforcement learning agents (https://github.com/deepmind/rlax) * NumPyro, a probabilistic programming library on top of JAX…

For anyone else outside of machine learning who was wondering what all of this is, here is my best explanation: The inferencing phase of a neural network attempts to minimize error or loss as defined by the user. This is done by iteratively applying gradient descent to the error function. Thus, the error function must have a known derivative, which can be difficult if the function has loops and conditionals. Autograd is a software package that produces derivatives of a…