DI-engine docs (Chinese and English)

OpenDILab Decision AI Engine. The Most Comprehensive Reinforcement Learning Framework B.P.

A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.

Minimal model-based RL algorithm implementations

Official repository for "iVideoGPT: Interactive VideoGPTs are Scalable World Models" (NeurIPS 2024), https://arxiv.org/abs/2405.15223

NeurIPS'24 Exploring the Edges of Latent State Clusters for Goal-Conditioned Reinforcement Learning

NeurIPS'24 Learning World Models for Unconstrained Goal Navigation

A curated list of awesome model based RL resources (continually updated)

Official implementation of DICL (Disentangled In-Context Learning), featured in the paper Zero-shot Model-based Reinforcement Learning using Large Language Models.

Symbolic Model-Based Reinforcement Learning

Simple world models lead to good abstractions, Google Cerebra internship 2020/master thesis at EPFL LCN 2021 ⬛◼️▪️🔦

Learning Discrete World Models for Heuristic Search

This repository offers implementations of classic and deep reinforcement learning algorithms, including dynamic programming, monte carlo methods, td-learning, and also both q-function-based and policy gradient approaches with deep nerual networks.

Numerical Evidence for Sample Efficiency of Model-Based over Model-Free Reinforcement Learning Control of Partial Differential Equations [ECC'24]

Code for Tackling Long-Horizon Tasks with Model-based Offline Reinforcement Learning

Library for Model Based RL

Code release for "HarmonyDream: Task Harmonization Inside World Models" (ICML 2024), https://arxiv.org/abs/2310.00344

Recall to Imagine, a model-based RL algorithm with superhuman memory. Oral (1.2%) @ ICLR 2024

Official codebase for "Privileged Sensing Scaffolds Reinforcement Learning", contains the Scaffolder algorithm and Sensory Scaffolding Suite.

This project focuses on implementing a novel approach to Risk-Aware Transfer in Reinforcement Learning (RL). This project introduces a unique perspective by incorporating risk at the test level rather than during training.