TorchWM Documentation

TorchWM is a modular PyTorch library for world models, latent-dynamics planning, and representation learning workflows (Dreamer, PlaNet/RSSM, JEPA, and diffusion components).

Documentation

• Getting Started
  • Installation
  • Quick Start: Dreamer
  • Quick Start: JEPA
  • Environment Backends
  • Typical Training Flow
• Package Overview
  • Core APIs
  • Environment Integration
  • World Model Building Blocks
  • Representation Learning and Diffusion
  • Data and Memory
  • Utilities
  • Which API Should I Use?
• API Reference
  • Core Public APIs
  • Dreamer
  • JEPA and ViT
  • Diffusion
  • Datasets and Transforms
  • Memory and Controllers
  • Utilities