Stars
An automatic, safe, and concurrent garbage collector for Rust
rvLLM: High-performance LLM inference in Rust. Drop-in vLLM replacement.
A Neovim plugin that provides VSCode-style diff rendering with two-tier highlighting (line + character level) in side-by-side and inline layouts, using VSCode's algorithm implemented in C.
Exocompilation for productive programming of hardware accelerators
PyTorchSim is a Comprehensive, Fast, and Accurate NPU Simulation Framework
A machine learning accelerator core designed for energy-efficient AI at the edge.
A menagerie of cute implementations of modern typechecking algorithms
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)
A controlled concurrency testing framework for the JVM
Open-source RTL logic simulator with CUDA acceleration
The "engine" of nissy, including the H48 optimal solver
A lightweight memory allocator for hardware-accelerated machine learning
Framework and Language for Neurosymbolic Programming.
Pen and paper exercises in machine learning
A toy compiler for NumPy array expressions that uses e-graphs and MLIR
graph based intermediate representation and backend for optimising compilers
Puzzles for learning Triton
Home for "How To Scale Your Model", a short blog-style textbook about scaling LLMs on TPUs
Vim plugin for LLM-assisted code/text completion
TensorRight: Automated Verification of Tensor Graph Rewrites
slides for the book "Principles od Abstract Interpretation", P. Cousot, MIT Press, 2021
Official inference framework for 1-bit LLMs