Xingtian

Overview

XingTian (刑天) is a componentized library for the development and verification of reinforcement learning algorithms. It supports multiple algorithms, including DQN, DDPG, PPO, and IMPALA, and can train agents in multiple environments such as Gym, Atari, Torcs, and StarCraftII. XingTian abstracts four modules - Algorithm, Model, Agent, and Environment - to provide users with quick verification and problem-solving capabilities. The library works in a similar way to building blocks, allowing users to easily configure and customize various components.

Features

• Componentized library for reinforcement learning algorithm development and verification
• Supports algorithms such as DQN, DDPG, PPO, and IMPALA
• Multiple environment support including Gym, Atari, Torcs, and StarCraftII

Summary

XingTian is a powerful library for developing and verifying reinforcement learning algorithms. With support for multiple algorithms and environments, it provides users with flexibility and ease-of-use. The componentized architecture allows for easy configuration and customization, making it a valuable tool for solving RL problems.