Explore projects

Optimized primitives for collective multi-GPU communication

SpikingJelly is an open-source deep learning framework for Spiking Neural Network (SNN) based on PyTorch.

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

CUDA Templates for Linear Algebra Subroutines

Fast Hadamard transform in CUDA, with a PyTorch interface

The official repository of Quamba

Mamba SSM architecture

an easy-to-use knn-mt toolkit

Detectron2 is FAIR's next-generation platform for object detection, segmentation and other visual recognition tasks.

physics meets neural networks

Distributed training framework for TensorFlow, Keras, PyTorch, and Apache MXNet.

A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch