tnguyen21

etcd

Distributed reliable key-value store for the most critical data of a distributed system

Detectron

FAIR's research platform for object detection research, implementing popular algorithms like Mask R-CNN and RetinaNet.

llm.c

LLM training in simple, raw C/CUDA

valkey

A new project to resume development on the formerly open-source Redis project. We're calling it Valkey, since it's a twist on the key-value datastore.

SWE-agent

SWE-agent takes a GitHub issue and tries to automatically fix it, using GPT-4, or your LM of choice. It solves 12.47% of bugs in the SWE-bench evaluation set and takes just 1 minute to run.

LMFlow

An Extensible Toolkit for Finetuning and Inference of Large Foundation Models. Large Models for All.

shattered-pixel-dungeon

Shattered Pixel Dungeon is an open-source traditional roguelike dungeon crawler with randomized levels and enemies, and hundreds of items to collect and use. It's based on the source code of Pixel Dungeon, by Watabou.

grok

dlrm

An implementation of a deep learning recommendation model (DLRM)

maturin

Build and publish crates with pyo3, cffi and uniffi bindings as well as rust binaries as python packages

KataGo

GTP engine and self-play learning in Go

t5x

nginx-tuning

NGINX tuning for best performance

tinyssh

TinySSH is small server (less than 100000 words of code)

TransformerLens

A library for mechanistic interpretability of GPT-style language models

evo

DNA foundation modeling from molecular to genome scale

ludic

🌳 Web Development in Pure Python with Type-Guided Components.

prize

A prize for finding tasks that cause large language models to show inverse scaling

rive-renderer

Downstream repo for Rive renderer

MyViewOfLinuxSystems

flash-attention-minimal

Flash Attention in ~100 lines of CUDA (forward pass only)

algebraic-nnhw

AI acceleration using matrix multiplication with half the multiplications

nano-web

Hyper-minimal low-latency webserver for serving SPAs and static content based on fasthttp.

micrograd-cuda

bespoke_automata

Bespoke Automata is a GUI and deployment pipline for making complex AI agents locally and offline

ConstitutionalHarmlessnessPaper

lenssort

Making some cool pfps 👩‍🎤

ARENA_3.0

tinygrad-notes

Tutorials on tinygrad

sae_vis

Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).