Skip to content

Occupying-Mars/jaxbolt

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 

Repository files navigation

jaxbolt

A minimal JAX-based autograd engine implementation, inspired by Andrej Karpathy's micrograd. This is a learning project to understand the internals of automatic differentiation and neural network frameworks.

Features

  • JAX-based tensor operations
  • Automatic differentiation
  • Basic neural network operations (relu, sigmoid, tanh)
  • Memory-efficient operations with chunking and checkpointing
  • Multi-device support

Why?

Built this to learn about:

  • How autograd engines work
  • JAX's approach to automatic differentiation
  • Efficient tensor operations
  • GPU acceleration and parallel processing

Feedback Welcome!

This is a learning project and I'm always looking to improve. Feel free to suggest improvements or share your thoughts:

Acknowledgments

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages