On-device AI across mobile, embedded and edge for PyTorch

A modular, easy to extend GFlowNet library

Experiments used in "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning"

LeanRL is a fork of CleanRL, where selected PyTorch scripts optimized for performance using compile and cudagraphs.

A networkx backend that uses joblib to run graph algorithms in parallel.

Fast and Accurate ML in 3 Lines of Code

Install and Run Python Applications in Isolated Environments

Python packaging and dependency management made easy

Simulation-based inference toolkit

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

Laplace approximations for Deep Learning.

An extremely fast Python linter and code formatter, written in Rust.

Bare-bones implementations of some generative models in Jax: diffusion, normalizing flows, consistency models, flow matching, (beta)-VAEs, etc

PyTorch implementation of normalizing flow models

Normalizing flows in PyTorch

BlackJAX is a Bayesian Inference library designed for ease of use, speed and modularity.

EasyTPP: Towards Open Benchmarking Temporal Point Processes

Trax — Deep Learning with Clear Code and Speed

Hackable and optimized Transformers building blocks, supporting a composable construction.

The AI developer platform. Use Weights & Biases to train and fine-tune models, and manage models from experimentation to production.

One of the "BEST" markdown preview extensions for Visual Studio Code

Implementation of E(n)-Equivariant Graph Neural Networks, in Pytorch

A modular graph-based Retrieval-Augmented Generation (RAG) system

Graph Neural PDEs

Code for CRATE (Coding RAte reduction TransformEr).

PyTorch native quantization and sparsity for training and inference