Welcome to MiniTorch! (https://github.com/minitorch/minitorch.github.io)
MiniTorch is a teaching library for machine learning engineers who wish to learn about the internal concepts underlying deep learning systems. Specifically, it is a pure Python re-implementation of the Torch API designed to be simple, easy-to-read, tested, and incremental. The final library can run Torch code with minimal changes (at some efficiency cost). The project was developed for the course Machine Learning Engineering at Cornell Tech.
MiniTorch is strictly DIY. It is not a library, per se, but a series of practically-oriented modules. These modules aim to provide the user with guidance on re-building the codebase by themselves, through a series of incremental implementations. These implementations can be done with a course or by an interested learner.
To get started, first read Workspace Setup and Contributing to build your workspace. Then follow through each of the modules to the right. Minimal computational resources are required. Module starting code is available on GitHub, and each proceeds incrementally from past modules.
Sasha Rush (@srush_nlp) and Ge Gao