A python implementation of differentiable quality diversity.

Related tags

Miscellaneousdqd
Overview

Differentiable Quality Diversity

This repository is the official implementation of Differentiable Quality Diversity.

The project contains a modified version of pyribs a quality diversity optimization library. All MEGA variants are implemented in pyribs. The GradientEmitter implements both the OG-MAP-Elites and the OMG-MEGA algorithms. The GradientImprovementEmitter implements the CMA-MEGA algorithm.

See ribs/emitters/_gradient_emitter.py and ribs/emitters/_gradient_improvement_emitter.py.

Requirements

The project builds in Anaconda.

Here are the instructions to create the conda environment:

conda env create -f experiments/environment.yml

Next install the local copy of pyribs after activating conda:

conda activate dqdexps
pip3 install -e .[all]

Pretrained Models

You can download the StyleGAN pretrained models from the StyleGAN repo. Place the .pt file in the folder experiments/lsi_clip.

CLIP automatically installs with the conda environment.

Running Experiments

For each experiment you pick an identifier for the algorithm you want to run.

Quality Diversity Algorithm Identifier
MAP-Elites map_elites
MAP-Elites (line) map_elites_line
CMA-ME cma_me_imp
OG-MAP-Elites og_map_elites
OMG-MEGA omg_mega
CMA-MEGA cma_mega
CMA-MEGA (Adam) cma_mega_adam

Linear Projection (sphere)

To run an experiment with MAP-Elites:

conda activate dqdexps
cd experiments/lin_proj

python3 lin_proj.py map_elites --objective sphere

To run a different algorithm replace map_elites with another identifier from the above table.

For additional options see:

python3 lin_proj.py --help

Linear Projection (Rastrigin)

To run an experiment with MAP-Elites:

conda activate dqdexps
cd experiments/lin_proj

python3 lin_proj.py map_elites --objective Rastrigin

To run a different algorithm replace map_elites with another identifier from the above table.

For additional options see:

python3 lin_proj.py --help

Arm Repertoire

To run an experiment with MAP-Elites:

conda activate dqdexps
cd experiments/arm

python3 arm.py map_elites

To run a different algorithm replace map_elites with another identifier from the above table.

For additional options see:

python3 arm.py --help

Latent Space Illumination (LSI)

To run an experiment with MAP-Elites:

conda activate dqdexps
cd experiments/lsi_clip

python3 lsi.py map_elites 

To run a different algorithm replace map_elites with another identifier from the above table.

For additional options see:

python3 lsi.py --help

Results

The following tables contain the reported results from the DQD paper.

Linear Projection (sphere)

Quality Diversity Algorithms QD-score Coverage
MAP-Elites 1.04 1.17%
MAP-Elites (line) 12.21 14.32%
CMA-ME 1.08 1.21%
OG-MAP-Elites 1.52 1.67%
OMG-MEGA 71.58 92.09%
CMA-MEGA 75.29 100.00%
CMA-MEGA (Adam) 75.3 100.00%

Linear Projection (Rastrigin)

Quality Diversity Algorithms QD-score Coverage
MAP-Elites 1.18 1.72%
MAP-Elites (line) 8.12 11.79%
CMA-ME 1.21 1.76%
OG-MAP-Elites 0.83 1.26%
OMG-MEGA 55.90 77.00%
CMA-MEGA 62.54 100.00%
CMA-MEGA (Adam) 62.58 100.00%

Arm Repertoire

Quality Diversity Algorithms QD-score Coverage
MAP-Elites 1.97 8.06%
MAP-Elites (line) 33.51 35.79%
CMA-ME 55.98 56.95%
OG-MAP-Elites 57.17 58.08%
OMG-MEGA 44.12 44.13%
CMA-MEGA 74.18 74.18%
CMA-MEGA (Adam) 73.82 73.82%

Latent Space Illumination (LSI)

Quality Diversity Algorithms QD-score Coverage
MAP-Elites 13.88 23.15%
MAP-Elites (line) 16.54 25.73%
CMA-ME 18.96 26.18%
CMA-MEGA 5.36 8.61%
CMA-MEGA (Adam) 21.82 30.73%

See the paper and supplementary materials for full data and standard error bars.

License

pyribs and this project are both released under the MIT License.

pyribs MIT License

Owner
ICAROS
Interactive and Collaborative Autonomous Robotic Systems
ICAROS
A Python library for inspecting JVM class files (.class)

lawu Lawu is a human-friendly library for assembling, disassembling, and exploring JVM class files. It's highly suitable for automation tasks. Documen

Tyler Kennedy 45 Oct 23, 2022
📙 Super lightweight function registries for your library

catalogue: Super lightweight function registries for your library catalogue is a tiny, zero-dependencies library that makes it easy to add function (o

Explosion 139 Jan 02, 2023
Med to csv - A simple way to parse MedAssociate output file in tidy data

MedAssociates to CSV file A simple way to parse MedAssociate output file in tidy

Jean-Emmanuel Longueville 5 Sep 09, 2022
Demo code for "Logs in distributed systems" webinar

Hexlet Logs Demo Пререквизиты docker-compose python3 Учетка в DataDog Базовое понимание, что такое логи (можно почитать гайд

Anton Markelov 1 Dec 01, 2021
A python tool used for hacking WhatsApp by diverting otp

W-HACK A python tool used for hacking WhatsApp by diverting otp You can hack WhatsApp easily with this tool Note:OTP expires after 5 seconds HOW TO IN

Spider Anongreyhat 3 Oct 17, 2021
Python implementation of the ASFLIP advection method

This is a python implementation of the ASFLIP advection method . We would like to hear from you if you appreciate this work.

Raymond Yun Fei 133 Nov 13, 2022
A python script for compiling and executing .cc files

Debug And Run A python script for compiling and executing .cc files Example dbrun fname.cc [DEBUG MODE] Compiling fname.cc with C++17 ------------

1 May 28, 2022
WinBoost: Boost your windows system.

Winboost runs a complete checkup of your entire system locating junk files, speed-reducing issues and causes of any system or application glitches or crashes. Through a lot of research and testing, w

Smit Parmar 4 Oct 01, 2021
a simple proof system I made to learn math without any mistakes

math_up a simple proof system I made to learn math without any mistakes 0. Short Introduction test yourself, enjoy your math! math_up is an NBG-based,

양현우 5 Jun 04, 2021
This code makes the logs provided by Fiddler proxy of the Google Analytics events coming from iOS more readable.

GA-beautifier-iOS This code makes the logs provided by Fiddler proxy of the Google Analytics events coming from iOS more readable. To run it, create a

Rafael Machado 3 Feb 02, 2022
A python script that fetches the grades of a student from a WAEC result in pdf format.

About waec-result-analyzer A python script that fetches the grades of a student from a WAEC result in pdf format. Built for federal government college

Oshodi Kolapo 2 Dec 04, 2021
Visualization of COVID-19 Omicron wave data in Seoul, Osaka, Tokyo, Hong Kong and Shanghai. 首尔、大阪、东京、香港、上海由新冠病毒 Omicron 变异株引起的本轮疫情数据可视化分析。

COVID-19 in East Asian Megacities This repository holds original Python code for processing and visualization COVID-19 data in East Asian megacities a

STONE 10 May 18, 2022
Python script for diving image data to train test and val

dataset-division-to-train-val-test-python python script for dividing image data to train test and val If you have an image dataset in the following st

Muhammad Zeeshan 1 Nov 14, 2022
🏃 Python3 Solutions of All Problems in GKS 2022 (In Progress)

GoogleKickStart 2022 Python3 solutions of Google Kick Start 2022. Solution begins with * means it will get TLE in the largest data set. Total computat

kamyu 38 Dec 29, 2022
Hacking and Learning consistently for 100 days straight af.

#100DaysOfHacking Hacking and Learning consistently for 100 days straight af. [yes, no breaks except mental-break ones, Obviously.] This Repo is one s

FENIL SHAH 17 Sep 09, 2022
Python interface to IEX and IEX cloud APIs

Python interface to IEX Cloud Referral Please subscribe to IEX Cloud using this referral code. Getting Started Install Install from pip pip install py

IEX Cloud 41 Dec 21, 2022
Learn Python tips, tools, and techniques in around 5 minutes each.

Python shorts Learn Python tips, tools, and techniques in around 5 minutes each. Watch on YouTube Subscribe on YouTube to keep up with all the videos.

Michael Kennedy 28 Jan 01, 2023
The bidirectional mapping library for Python.

bidict The bidirectional mapping library for Python. Status bidict: has been used for many years by several teams at Google, Venmo, CERN, Bank of Amer

Joshua Bronson 1.2k Dec 31, 2022
Streamlit apps done following data professor's course on YouTube

streamlit-twelve-apps Streamlit apps done following data professor's course on YouTube Español Curso de apps de data science hecho por Data Professor

Federico Bravin 1 Jan 10, 2022