Hi, JAX: An introduction to vanilla JAX for deep learning research (github.com)

The new course "Hi, JAX!" serves as a comprehensive introduction to JAX, a Python library designed for high-performance deep learning research. It emphasizes JAX's unique benefits, including its speed through XLA compilation, portability across different hardware (e.g., GPUs and TPUs), and a strong focus on elegant mathematical abstractions. By offering recorded lectures, coding demonstrations, and programming challenges, the course aims to bridge the knowledge gap for users transitioning from libraries like PyTorch to JAX, particularly addressing the learning curve associated with JAX's distinctive programming paradigm. This initiative is significant for the AI/ML community as it fosters greater adoption of JAX, enhancing the ecosystem with educational resources that are currently limited. Through practical lessons and a strong emphasis on foundational concepts, the course equips researchers with the skills needed to leverage JAX’s powerful features for advanced neural network training and experimentation. The course content includes carefully structured lessons, from basic JAX concepts to advanced topics like just-in-time compilation, culminating in independent project work that encourages participants to engage deeply with JAX’s capabilities and contribute to its growing ecosystem.