Skip to content

arnab39/Symmetry-RL

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

EqR: Equivariant Representations for Data-Efficient Reinforcement Learning

This repo provides the official implementation of the paper. The code is based on original spr implementation.

Installation

To install the requirements, follow these steps:

# PyTorch
pip install pytorch torchvision

# Install requirements
pip install -r requirements.txt

How to run?

  • To run Eqr with group equivariant loss
python -m scripts.run --game krull --momentum-tau 1. --seed 0 --group-type 'En' --num-blocks 12 --projection-type 'q_shared' --second-projection-type 'mlp'  --acteqv-loss-weight 1. --groupeqv-loss-weight 1. --reward-loss-weight 1.
  • To run Eqr with just action equivariant loss
python -m scripts.run --game krull --momentum-tau 1. --seed 0 --group-type 'En' --num-blocks 12 --projection-type 'q_shared' --second-projection-type 'mlp'  --acteqv-loss-weight 1. --groupeqv-loss-weight 0. --reward-loss-weight 1.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages