jph00

marker

Convert PDF to markdown quickly with high accuracy

gemma.cpp

lightweight, standalone C++ inference engine for Google's Gemma models.

LaVague

Large Action Model framework to develop AI Web Agents

moondream

tiny vision language model

nanobind

nanobind: tiny and efficient C++/Python bindings

maxtext

A simple, performant and scalable Jax LLM!

distilabel

⚗️ distilabel is a framework for synthetic data and AI feedback for AI engineers that require high-quality outputs, full data ownership, and overall efficiency.

GPTFast

Accelerate your Hugging Face Transformers 6-8.5x. Native to Hugging Face and PyTorch.

quanto

A pytorch Quantization Toolkit

texify

Math OCR model that outputs LaTeX and markdown

lightning-whisper-mlx

An extremely fast implementation of whisper optimized for Apple Silicon using MLX.

win32yank

Windows clipboard tool

attorch

A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.

open_lm

A repository for research on medium sized language models.

GradCache

Run Effective Large Batch Contrastive Learning Beyond GPU/TPU Memory Constraint

neural-cherche

Neural Search

DHS-LLM-Workshop

DHS 2023 LLM Workshop by Sourab Mangrulkar

micrograd-cuda

cuda_hgemm

Several optimization methods of half-precision general matrix multiplication (HGEMM) using tensor core with WMMA API and MMA PTX instruction.

ao

torchao: PyTorch Architecture Optimization (AO). Performant kernels that work with PyTorch.

ipylab

Control JupyterLab from Python Notebooks with Jupyter Widgets 🧪 ☢️ 🐍

QuaRot

Code for QuaRot, an end-to-end 4-bit inference of large language models.

hippogriff

Griffin MQA + Hawk Linear RNN Hybrid

sudoku-web-app

A Sudoku web app - as used on SudokuExchange.com

torch-cublas-hgemm

PyTorch half precision gemm lib w/ fused optional bias + optional relu/gelu

DipoleAttention

torch-bnb-fp4

Faster Pytorch bitsandbytes 4bit fp4 nn.Linear ops

sudoku

Getting me Sudoku puzzles to solve daily (see Issues!)

driss_torch

Cuda extensions for PyTorch

cuda-experiments