DOOM Open Source Release

GPU-Parallelizing Arbitrary Python Code By Running 1 Million Python Interpreters on a GPU 🐍

Official Codebase For paper "One-step Language Modeling via Continuous Denoising"

talkie is a vintage language model from 1930

+256,000,000 points per plot, +60 Fps on shity laptop. Only limit is the size of your RAM.

surround.vim: Delete/change/add parentheses/quotes/XML-tags/much more with ease

My Solutions to K. N. King's "C Programming: A Modern Approach", second edition

Open-weights voice acting pipeline combining zero-shot voice cloning with natural-language direction. Provide a reference voice (or generate one) and describe how the line should be performed. Prod…

A user-friendly, lightweight TUI for disk imaging

A good looking terminal emulator which mimics the old cathode display...

Scripts for computing the Intelligibility and CLVP scores for evaluating TTS models

Important concepts in numerical linear algebra and related areas

Flow-based model for learning non-gradient field dynamics

A tiny CPU simulator written in Python

torchax is a PyTorch frontend for JAX. It gives JAX the ability to author JAX programs using familiar PyTorch syntax. It also provides JAX-Pytorch interoperability, meaning, one can mix JAX & Pytor…

Master programming by recreating your favorite technologies from scratch.

Official Repository of Paper: "Emilia-NV: A Non-Verbal Speech Dataset with Word-Level Annotation for Human-Like Speech Modeling"

graftr: an interactive shell to view and edit PyTorch checkpoints.

Simple implementation of adjoint sampling in JAX

Minimal reproduction of DeepSeek R1-Zero

An extremely fast Python package and project manager, written in Rust.

Minimal and annotated implementations of key ideas from modern deep learning research.

The Modular Platform (includes MAX & Mojo)

Numbers every LLM researchers should know

code for "Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching"

Datasets to protect Earth's forests and biodiversity