This repository is an official PyTorch implementation of the paper, DISCO-DANCE: Learning to Discover Skills through Guidance, NeurIPS 2023. This codebase was adapted from URLB.
arXiv / project page / poster
Authors: Hyunseung Kim*, Byungkun Lee*, Hojoon Lee, Dongyoon Hwang, Sejik Park, Kyushik Min, and Jaegul Choo.
We assume you have access to a GPU that can run CUDA 11.1 and CUDNN 8. Then, the simplest way to install all required dependencies is to create an anaconda environment by running
conda env create -f requirements.yaml
After the instalation ends you can activate your environment with
conda activate DISCO-DANCE
To run pre-training use the pretrain_maze.py
script
python pretrain_maze.py maze_type=square_bottleneck
This script will produce several agent snapshots after training for 1M
, 2M
, ..., 5M
frames. The snapshots will be stored under the following directory:
./models/<agent_name>/<maze_type>/
The console output is also available in a form:
| train | F: 6000 | S: 3000 | E: 6 | L: 1000 | R: 5.5177 | FPS: 96.7586 | T: 0:00:42
a training entry decodes as
F : total number of environment frames
S : total number of agent steps
E : total number of episodes
R : episode return
FPS: training throughput (frames per second)
T : total training time
@article{kim2024learning,
title={Learning to discover skills through guidance},
author={Kim, Hyunseung and LEE, BYUNG KUN and Lee, Hojoon and Hwang, Dongyoon and Park, Sejik and Min, Kyushik and Choo, Jaegul},
journal={Advances in Neural Information Processing Systems},
volume={36},
year={2024}
}
For personal communication, please contact Hyunseung Kim, or Byungkun Lee at
{mynsng, byungkun.lee}@kaist.ac.kr
.