DI-engine
A comprehensive Python-first reinforcement learning framework with modular abstractions for decision intelligence applications.
What is DI-engine?
DI-engine is a comprehensive reinforcement learning framework designed as a generalized decision intelligence engine for PyTorch and JAX. It provides modular abstractions for environments, policies, and models to support a wide variety of deep RL algorithms, from basic methods like DQN and PPO to advanced areas like multi-agent systems, offline RL, and imitation learning. The framework solves the problem of fragmented RL implementations by offering a standardized, extensible foundation for both research and real-world applications.
Researchers and engineers working on decision intelligence problems, including those in reinforcement learning, autonomous systems, game AI, and multi-agent simulations. It's particularly valuable for teams needing a unified framework to experiment with diverse RL algorithms or deploy RL solutions in production.
Developers choose DI-engine for its exceptional algorithm coverage, python-first design, and system optimizations for large-scale training. Unlike many RL libraries, it seamlessly integrates academic research with practical applications through modular abstractions and supports real-world use cases like autonomous driving and StarCraft AI.
Overview
OpenDILab Decision AI Engine. The Most Comprehensive Reinforcement Learning Framework B.P.
Use Cases
Best For
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
