PyTorch native quantization and sparsity for training and inference

depyf is a tool to help you understand and adapt to PyTorch compiler torch.compile.

A native PyTorch Library for large model training

PyTorch native finetuning library

LLM training in simple, raw C/CUDA

A PyTorch-based Speech Toolkit

Building a quick conversation-based search demo with Lepton AI.

TensorDict is a pytorch dedicated tensor container.

A Gradio web UI for Large Language Models.

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

An open-source framework for training large multimodal models.

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

Tensors and Dynamic neural networks in Python with strong GPU acceleration

ASDL: Automatic Second-order Differentiation Library for PyTorch

AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.

torch::deploy (multipy for non-torch uses) is a system that lets you get around the GIL problem by running multiple Python interpreters in a single C++ process.

Containers for machine learning

ROS package that can filter geometry defined in URDF models from Kinect depth images. Can also preprocess data for the OpenNI tracker, to remove backgrounds, robots etc.

[CVPR'22] NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

The standard data-centric AI package for data quality and machine learning with messy, real-world data and labels.

C++ wrappers for SIMD intrinsics and parallelized, optimized mathematical functions (SSE, AVX, AVX512, NEON, SVE))

Visual Inertial Odometry with SLAM capabilities and 3D Mesh generation.

Create agents that monitor and act on your behalf. Your agents are standing by!

Structured state space sequence models

Robotics with GPU computing

Multithreaded Python without the GIL

vinca configuration files for ros-noetic