JAX
A Python library for numerical computing and machine learning that compiles array code for CPUs, GPUs and TPUs.
In more detail
JAX is a common framework for TPU training. Its programs are compiled for the target chip, which is what lets the same code run on different accelerators.
Related terms
- Tensor Processing Unit — Google's own accelerator chip for training and serving machine learning models built on frameworks such as JAX and PyTorch.
Taught in
- Choosing Compute — Google Cloud Associate Cloud Engineer Exam Prep