petrex

alphatensor

Auto-GPT

An experimental open-source attempt to make GPT-4 fully autonomous.

FasterTransformer

Transformer related optimization, including BERT, GPT

flash-attention

Fast and memory-efficient exact attention

pytorch

Tensors and Dynamic neural networks in Python with strong GPU acceleration

triton

Development repository for the Triton language and compiler

tvm

Open deep learning compiler stack for cpu, gpu and specialized accelerators

vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

ChatDev

Create Customized Software using Natural Language Idea (through LLM-powered Multi-Agent Collaboration)

composable_kernel

Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators

DeepSpeed

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

gpt-fast

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

gpt-researcher

GPT based autonomous agent that does online comprehensive research on any given topic

human-eval

Code for the paper "Evaluating Large Language Models Trained on Code"

jax

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

llama

Inference code for LLaMA models

llama-recipes

Examples and recipes for Llama 2 model

llama.cpp

Port of Facebook's LLaMA model in C/C++

llm.c

LLM training in simple, raw C/CUDA

llvm-project

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies. Note: the repository does not accept github pull requests at this moment. Please submit your patches at http://reviews.llvm.org.

nanoGPT

The simplest, fastest repository for training/finetuning medium-sized GPTs.

onnx

Open Neural Network Exchange

OpenChatKit

SpaceVim

A community-driven modular vim distribution - The ultimate vim configuration

stable-diffusion

A latent text-to-image diffusion model

TensorRT-LLM

TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT-LLM also contains components to create Python and C++ runtimes that execute those TensorRT engines.

torchtune

A Native-PyTorch Library for LLM Fine-tuning

training

Reference implementations of training benchmarks

transformers

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

tvm-1

TVM integration into PyTorch