Open-source Packages

State-of-the-art Machine Learning for PyTorch, TensorFlow, and JAX. Provides thousands of pre-trained models and integrated optimization tools.

An efficient library for differentiable optimization in PyTorch. It enables gradient-based meta-learning and hyperparameter optimization.

A deep learning optimization library that makes distributed training easy, efficient, and effective. Features ZeRO technology for massive model scaling.

A gradient processing and optimization library for JAX. Designed to facilitate research by providing composable building blocks.

A distributed deep learning training framework for TensorFlow, Keras, PyTorch, and Apache MXNet. Makes distributed training easy and fast.

An automatic hyperparameter optimization software framework, particularly designed for machine learning. Featuring an define-by-run API.

A scalable hyperparameter tuning library. It is compatible with any machine learning framework and provides state-of-the-art search algorithms.

An efficient and generalized black-box optimization system. It supports multiple optimization tasks including hyperparameter optimization and neural architecture search.

A gradient-free optimization platform. It contains a wide range of algorithms for parameter tuning and works well for non-differentiable or noisy functions.

Hardware accelerated, batchable and differentiable optimizers in JAX. Enforces distinct separation between objective functions and solvers.

A toolkit for Lagrangian-based constrained optimization in PyTorch. Handles formulation and solving of constrained optimization problems.

An automatic differentiation library for multilevel optimization. It provides a unified framework for generalized meta-learning and hyperparameter optimization.

A Python-embedded modeling language for convex optimization problems. It allows you to express problems in a natural way that follows the math.

A library for higher-order optimization in PyTorch. It facilitates the implementation of "unrolled" optimization loops for meta-learning.

A comprehensive collection of optimizer implementations for PyTorch, including AdaBound, diffGrad, LAMB, and many others not in the official library.