Skip to content

howsmyanimeprofilepicture/image-tokenizer-for-mbrl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

README

# First of all, extract dataset as follows
tar xf traj.tar.xz
  • This repo is an implementation of VQ-VAE-based image tokenizer for a model-based reinforcement learning.
  • In the MBRL like World Model, dynamic and behavior are estimated through the encoded observations.
  • So, to implement the World Model-like MBRL, it is very important to encoding observations effectively.
  • The dataset is collected from the Car Racing environment.

References

  1. https://github.com/CompVis/taming-transformers
  2. https://github.com/eloialonso/iris
  3. https://github.com/nadavbh12/VQ-VAE