Data

Tools

Indexes

Applications

Experiments

Open Source Science

minari

A standard format for offline reinforcement learning datasets, with popular reference datasets and related utilities

https://github.com/farama-foundation/minari

Science Score: 57.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ✓
    DOI references
    Found 5 DOI reference(s) in README
  • ○
    Academic publication links
  • ○
    Committers with academic emails
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (17.5%) to scientific vocabulary

Keywords

datasets gymnasium offline-rl reinforcement-learning

Keywords from Contributors

gym gym-environment gridworld-environment deepmind-control-suite deepmind-lab autonomous-driving
Last synced: 6 months ago · JSON representation ·

Repository

A standard format for offline reinforcement learning datasets, with popular reference datasets and related utilities

Basic Info
  • Host: GitHub
  • Owner: Farama-Foundation
  • License: other
  • Language: Python
  • Default Branch: main
  • Homepage: https://minari.farama.org
  • Size: 1.86 GB
Statistics
  • Stars: 423
  • Watchers: 11
  • Forks: 61
  • Open Issues: 19
  • Releases: 13
Topics
datasets gymnasium offline-rl reinforcement-learning
Created over 3 years ago · Last pushed 6 months ago
Metadata Files
Readme Contributing Funding License Code of conduct Citation

README.md

build pre-commit Code style: black release

Minari is a Python library for conducting research in offline reinforcement learning, akin to an offline version of Gymnasium or an offline RL version of HuggingFace's datasets library.

The documentation website is at minari.farama.org. We also have a public discord server (which we use for Q&A and to coordinate development work) that you can join here: https://discord.gg/bnJ6kubTg6.

Installation

To install Minari from PyPI: bash pip install minari

This will install the minimum required dependencies. Additional dependencies will be prompted for installation based on your use case. To install all dependencies at once, use: bash pip install "minari[all]"

If you'd like to start testing or contribute to Minari please install this project from source with:

git clone https://github.com/Farama-Foundation/Minari.git --single-branch cd Minari pip install -e ".[all]"

Command Line API

To check available remote datasets:

bash minari list remote

To download a dataset:

bash minari download D4RL/door/human-v2

To check available local datasets:

bash minari list local To show the details of a dataset:

bash minari show D4RL/door/human-v2

For the list of commands: bash minari --help

Basic Usage

Reading a Dataset

```python import minari

dataset = minari.load_dataset("D4RL/door/human-v2")

for episodedata in dataset.iterateepisodes(): observations = episodedata.observations actions = episodedata.actions rewards = episodedata.rewards terminations = episodedata.terminations truncations = episodedata.truncations infos = episodedata.infos ... ```

Writing a Dataset

```python import minari import gymnasium as gym from minari import DataCollector

env = gym.make('FrozenLake-v1') env = DataCollector(env)

for _ in range(100): env.reset() done = False while not done: action = env.action_space.sample() # <- use your policy here obs, rew, terminated, truncated, info = env.step(action) done = terminated or truncated

dataset = env.create_dataset("frozenlake/test-v0") ```

For other examples, see Basic Usage. For a complete tutorial on how to create new datasets using Minari, see our Pointmaze D4RL Dataset tutorial, which re-creates the Maze2D datasets from D4RL.

Training Libraries Integrating Minari

Citation

If you use Minari, please consider citing it:

@software{minari, author = {Younis, Omar G. and Perez-Vicente, Rodrigo and Balis, John U. and Dudley, Will and Davey, Alex and Terry, Jordan K}, doi = {10.5281/zenodo.13767625}, month = sep, publisher = {Zenodo}, title = {Minari}, url = {https://doi.org/10.5281/zenodo.13767625}, version = {0.5.0}, year = 2024, bdsk-url-1 = {https://doi.org/10.5281/zenodo.13767625} }

Minari is a shortening of Minarai, the Japanese word for "learning by observation".

Owner

  • Name: Farama Foundation
  • Login: Farama-Foundation
  • Kind: organization
  • Email: contact@farama.org

The Farama foundation is a nonprofit organization working to develop and maintain open source reinforcement learning tools.

Citation (CITATION.cff) 

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
  - name: "Minari Contributors"
title: "Minari: A dataset API for Offline Reinforcement Learning"
date-released: 2023-05-17

GitHub Events

Total
  • Create event: 6
  • Release event: 2
  • Issues event: 31
  • Watch event: 142
  • Delete event: 2
  • Issue comment event: 74
  • Push event: 60
  • Pull request event: 67
  • Pull request review event: 23
  • Pull request review comment event: 13
  • Fork event: 17
Last Year
  • Create event: 6
  • Release event: 2
  • Issues event: 31
  • Watch event: 142
  • Delete event: 2
  • Issue comment event: 74
  • Push event: 60
  • Pull request event: 67
  • Pull request review event: 23
  • Pull request review comment event: 13
  • Fork event: 17

Committers

Last synced: about 2 years ago

All Time
  • Total Commits: 273
  • Total Committers: 24
  • Avg Commits per committer: 11.375
  • Development Distribution Score (DDS): 0.579
Past Year
  • Commits: 111
  • Committers: 20
  • Avg Commits per committer: 5.55
  • Development Distribution Score (DDS): 0.595
Top Committers
Name Email Commits
will w****6@l****k 115
Rodrigo de Lazcano r****6@g****m 46
Will Dudley 1****y 27
Omar Younis 4****k 15
John Balis p****s@g****m 11
Jordan Terry j****0@g****m 7
WillDudley W****6@l****k 7
Alex Davey a****0@g****m 5
Jet 3****s 5
Elliot Tower e****t@e****m 4
Alexander Nikulin h****h@g****m 4
Manuel Goulão m****o@g****m 4
Shreyans Jain j****8@g****m 4
graham g****t@g****m 3
Aaron Marquez a****1@g****m 2
Tosin Oseni o****i@g****m 2
Zhiyuan 3****h 2
Vijay Alagappan 3****e 2
Elliot Tower 3****r 2
Aaron Marquez 1****o 2
Mark Towers m****s@g****m 1
Ariel Kwiatkowski a****i@g****m 1
Anthony Chang 4****s 1
DeeDive 4****e 1
Committer Domains (Top 20 + Academic)
live.co.uk: 2 elliottower.com: 1

Issues and Pull Requests

Last synced: 6 months ago

All Time
  • Total issues: 78
  • Total pull requests: 246
  • Average time to close issues: 2 months
  • Average time to close pull requests: 16 days
  • Total issue authors: 50
  • Total pull request authors: 39
  • Average comments per issue: 2.68
  • Average comments per pull request: 0.86
  • Merged pull requests: 209
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 26
  • Pull requests: 64
  • Average time to close issues: 25 days
  • Average time to close pull requests: 7 days
  • Issue authors: 21
  • Pull request authors: 11
  • Average comments per issue: 2.35
  • Average comments per pull request: 0.48
  • Merged pull requests: 50
  • Bot issues: 0
  • Bot pull requests: 0
View more stats
Top Authors
Issue Authors
  • jamartinh (7)
  • im-Kitsch (6)
  • Howuhh (4)
  • younik (4)
  • WillDudley (4)
  • Kallinteris-Andreas (4)
  • umutucak (3)
  • JustinS6626 (2)
  • kaixin96 (2)
  • rodrigodelazcano (2)
  • GongXudong (1)
  • RedTachyon (1)
  • EmptyJackson (1)
  • carlosgmartin (1)
  • prajjwal1 (1)
Pull Request Authors
  • younik (131)
  • rodrigodelazcano (34)
  • WillDudley (22)
  • alexdavey (16)
  • balisujohn (12)
  • tomekster (6)
  • wakened2024 (6)
  • pseudo-rnd-thoughts (6)
  • elliottower (6)
  • shreyansjainn (6)
  • enerrio (5)
  • Howuhh (4)
  • jamartinh (4)
  • im-Kitsch (3)
  • grahamannett (3)
Top Labels
Issue Labels
good first issue (4) documentation (2) enhancement (1) bug (1)
Pull Request Labels
enhancement (1)

Packages

  • Total packages: 3
  • Total downloads:
    • pypi 10,226 last-month
  • Total docker downloads: 155
  • Total dependent packages: 3
    (may contain duplicates)
  • Total dependent repositories: 1
    (may contain duplicates)
  • Total versions: 34
  • Total maintainers: 3
proxy.golang.org: github.com/farama-foundation/minari
  • Versions: 10
  • Dependent Packages: 0
  • Dependent Repositories: 0
Rankings
Dependent packages count: 5.5%
Average: 5.6%
Dependent repos count: 5.8%
Last synced: 7 months ago
proxy.golang.org: github.com/Farama-Foundation/Minari
  • Versions: 10
  • Dependent Packages: 0
  • Dependent Repositories: 0
Rankings
Dependent packages count: 5.5%
Average: 5.6%
Dependent repos count: 5.8%
Last synced: 7 months ago
pypi.org: minari

A standard format for offline reinforcement learning datasets, with popular reference datasets and related utilities.

  • Versions: 14
  • Dependent Packages: 3
  • Dependent Repositories: 1
  • Downloads: 10,226 Last month
  • Docker Downloads: 155
Rankings
Docker downloads count: 4.6%
Dependent packages count: 4.8%
Stargazers count: 5.5%
Downloads: 5.9%
Forks count: 6.8%
Average: 8.2%
Dependent repos count: 21.6%
Maintainers (3)
jkterry pseudornd WillDudley
Last synced: 7 months ago

Dependencies

docs/requirements.txt pypi
  • furo *
  • myst-parser *
  • sphinx *
setup.py pypi
  • numpy >=1.18.0
.github/workflows/build-docs-version.yml actions
  • JamesIves/github-pages-deploy-action v4 composite
  • actions/checkout v3 composite
  • actions/setup-python v4 composite
  • dawidd6/action-get-tag v1 composite
.github/workflows/build-docs.yml actions
  • JamesIves/github-pages-deploy-action v4 composite
  • actions/checkout v3 composite
  • actions/setup-python v4 composite
.github/workflows/build-publish.yml actions
  • actions/checkout v3 composite
  • actions/download-artifact v2 composite
  • actions/setup-python v2 composite
  • actions/upload-artifact v2 composite
  • pypa/gh-action-pypi-publish release/v1 composite
.github/workflows/linux-tutorials-test-on-pr.yml actions
  • actions/checkout v3 composite
  • actions/setup-python v3 composite
.github/workflows/linux-tutorials-test-on-push.yml actions
  • actions/checkout v3 composite
  • actions/setup-python v3 composite
.github/workflows/manual-build-docs-version copy.yml actions
  • JamesIves/github-pages-deploy-action v4 composite
  • actions/checkout v3 composite
  • actions/setup-python v4 composite
.github/workflows/pre-commit.yml actions
  • actions/checkout v3 composite
  • actions/setup-python v3 composite
tests/DatasetCreation/requirements.txt pypi
  • cython * test
  • gymnasium spec_stack test
  • gymnasium-robotics * test
tests/DatasetUsage/requirements.txt pypi
  • cython * test
  • numpy * test
tests/LocalStorage/requirements.txt pypi
  • cython * test
  • h5py * test
  • structlog * test
  • tensorboardX * test
  • typing_extensions * test
tutorials/DatasetCreation/requirements.txt pypi
  • cython *
  • gymnasium spec_stack
  • gymnasium-robotics *
tutorials/DatasetUsage/requirements.txt pypi
  • cython *
  • numpy *
tutorials/LocalStorage/requirements.txt pypi
  • cython *
  • h5py *
  • structlog *
  • tensorboardX *
  • typing_extensions *
tutorials/RemoteStorage/requirements.txt pypi
  • cython *
  • google-cloud-storage *
  • h5py *
  • structlog *
  • tensorboardX *
  • typing_extensions *