Contents

• Contents
• Overview
• Requirements
• Quick Start
• Acknowledgments

Overview

Requirements

conda env create -f conda_env.yml
source activate erqt

Then add your conda environment include to CPATH (put this in your .bashrc to make it permanent):

export CPATH=$CONDA_PREFIX/include

If you encounter the libstdc++.so.6: version 'GLIBCXX_3.4.xx' not found error, the following command might help:

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<path-to-your-conda-env>/lib

I have also found that tensorboard wants protobuf version to be 3.20.x, and this helped

# you might need to uninstall dm-control
pip3 install --upgrade protobuf==3.20.0

Quick Start

When your environment is ready, you could run scripts in the "run.sh". For example:

## AUG
python main.py run_params=aug env_params=halfcheetah_medium env_params.pct_traj=0.1

## BC
python main.py agent_params=bc 

## filtered BC
python main.py agent_params=bc env_params.pct_traj=0.1

## ERQT
python main.py agent_params.behavior_ckpt_file=./save/ env_params.use_aug=true env_params.pct_traj=0.1 10%_bc_stochastic-halfcheetah-medium-replay-123-250324-112957/epoch_15.pth

## QT
python main.py agent_params=qt

Acknowledgments

This repo benefits from DT and Diffusion-QL. Thanks for their wonderful works!