This is my simple implementation of Karpathy’s micrograd in mojo. It is an scalar (1 element tensor) autodiff engine intented for learning purpose while being complete and fast. The goal is to see how far this engine can go while implementing mojo’s performance and in the future, MLIR to describe the operations AST.
This has also been my first proyect in learning mojo so any advice or suggestion is appreciated.
Thanks for your time.
