Intro to JAX: Accelerating Machine Learning research

JAX is a Python package that combines a NumPy-like API with a set of powerful composable transformations for automatic differentiation, vectorization, parallelization, and JIT compilation. Your code can run on CPU, GPU or TPU. This talk will get you started accelerating your ML with JAX! Resources: JAX reference documentation → Speaker: Jake VanderPlas (Software Engineer) Watch all Google’s Machine Learning Virtual Community Day sessions → Subscribe to the TensorFlow channel → #MLCommunityDay product: TensorFlow - General; event: ML Community Day 2021; fullname: Jake VanderPlas; re_ty: Publish;