{ "info": { "author": "Seungjae (Ryan) Lee", "author_email": "seungjaeryanlee@gmail.com", "bugtrack_url": null, "classifiers": [ "License :: OSI Approved :: MIT License", "Operating System :: OS Independent", "Programming Language :: Python :: 3" ], "description": "\n\n# rldb\n\n[![Build Status](https://travis-ci.com/seungjaeryanlee/rldb.svg?branch=master)](https://travis-ci.com/seungjaeryanlee/rldb)\n\n![Environments tracked in rldb](https://img.shields.io/badge/environments-114-blue.svg)\n![Papers tracked in rldb](https://img.shields.io/badge/papers-22-blue.svg)\n![Repos tracked in rldb](https://img.shields.io/badge/repos-2-blue.svg)\n![Algorithms tracked in rldb](https://img.shields.io/badge/algorithms-59-blue.svg)\n![Entries tracked in rldb](https://img.shields.io/badge/entries-3266-blue.svg)\n\nDatabase of RL algorithms\n\n| Atari Space Invaders Scores | MuJoCo Walker2d Scores |\n|:-:|:-:|\n| ![Atari Space Invaders Scores](/docs/atari-space-invaders.png) | ![MuJoCo Walker2d Scores](/docs/mujoco-walker2d.png) |\n\n## Examples\n\nYou can use `rldb.find_all({})` to retrieve all existing entries in `rldb`.\n\n```python\nimport rldb\n\n\nall_entries = rldb.find_all({})\n```\n\nYou can also filter entries by specifying key-value pairs that the entry must match:\n\n```python\nimport rldb\n\n\ndqn_entries = rldb.find_all({'algo-nickname': 'DQN'})\nbreakout_noop_entries = rldb.find_all({\n 'env-title': 'atari-breakout',\n 'env-variant': 'No-op start',\n})\n```\n\nYou can also use `rldbl.find_one(filter_dict)` to find one entry that matches the key-value pair specified in `filter_dict`:\n\n```python\nimport rldb\nimport pprint\n\n\nentry = rldb.find_one({\n 'env-title': 'atari-pong',\n 'algo-title': 'Human',\n})\npprint.pprint(entry)\n```\n\n\n
Output\n

\n\n```python\n{\n 'algo-nickname': 'Human',\n 'algo-title': 'Human',\n 'env-title': 'atari-pong',\n 'env-variant': 'No-op start',\n 'score': 14.6,\n 'source-arxiv-id': '1511.06581',\n 'source-arxiv-version': 3,\n 'source-authors': [ 'Ziyu Wang',\n 'Tom Schaul',\n 'Matteo Hessel',\n 'Hado van Hasselt',\n 'Marc Lanctot',\n 'Nando de Freitas'],\n 'source-bibtex': '@article{DBLP:journals/corr/WangFL15,\\n'\n ' author = {Ziyu Wang and\\n'\n ' Nando de Freitas and\\n'\n ' Marc Lanctot},\\n'\n ' title = {Dueling Network Architectures for Deep '\n 'Reinforcement Learning},\\n'\n ' journal = {CoRR},\\n'\n ' volume = {abs/1511.06581},\\n'\n ' year = {2015},\\n'\n ' url = {http://arxiv.org/abs/1511.06581},\\n'\n ' archivePrefix = {arXiv},\\n'\n ' eprint = {1511.06581},\\n'\n ' timestamp = {Mon, 13 Aug 2018 16:48:17 +0200},\\n'\n ' biburl = '\n '{https://dblp.org/rec/bib/journals/corr/WangFL15},\\n'\n ' bibsource = {dblp computer science bibliography, '\n 'https://dblp.org}\\n'\n '}',\n 'source-nickname': 'DuDQN',\n 'source-title': 'Dueling Network Architectures for Deep Reinforcement '\n 'Learning'\n}\n```\n\n

\n
\n\n## Entry Structure\n\nHere is the format of every entry:\n\n```python3\n{\n # BASICS\n \"source-title\": \"\",\n \"source-nickname\": \"\",\n \"source-authors\": [],\n\n # MISC.\n \"source-bibtex\": \"\",\n\n # ALGORITHM\n \"algo-title\": \"\",\n \"algo-nickname\": \"\",\n \"algo-source-title\": \"\",\n\n # SCORE\n \"env-title\": \"\",\n \"score\": 0,\n}\n```\n\n- `source-title` is the full title of the source of the score: it can be the title of the paper or GitHub repository title. `source-nickname` is a popular nickname or acronym for that title if it exists, otherwise it is the same as `source-title`. \n- `source-authors` are a list of authors or contributors.\n- `source-bibtex` is a BibTeX-format citation.\n- `algo-title` is the full title of the algorithm used. `algo-nickname` is the nickname or acronym for that algorithm if it exists, otherwise it is the same as `algo-nickname`.\n- `algo-source-title` is the title of the source of the **algorithm**. It can and often is different from `source-title`.\n\nFor example, the **Space Invaders** score of **Asynchronous Advantage Actor Critic (A3C)** algorithm in the **Noisy Networks for Exploration (NoisyNet)** paper is represented by the following entry:\n\n```python3\n{\n # BASICS\n \"source-title\": \"Noisy Networks for Exploration\",\n \"source-nickname\": \"NoisyNet\",\n \"source-authors\": [\n \"Meire Fortunato\",\n \"Mohammad Gheshlaghi Azar\",\n \"Bilal Piot\",\n \"Jacob Menick\",\n \"Ian Osband\",\n \"Alex Graves\",\n \"Vlad Mnih\",\n \"Remi Munos\",\n \"Demis Hassabis\",\n \"Olivier Pietquin\",\n \"Charles Blundell\",\n \"Shane Legg\",\n ],\n\n # ARXIV\n \"source-arxiv-id\": \"1706.10295\",\n \"source-arxiv-version\": 2,\n\n # MISC.\n \"source-bibtex\": \"\"\"\n@article{DBLP:journals/corr/FortunatoAPMOGM17,\n author = {Meire Fortunato and\n Mohammad Gheshlaghi Azar and\n Bilal Piot and\n Jacob Menick and\n Ian Osband and\n Alex Graves and\n Vlad Mnih and\n R{\\'{e}}mi Munos and\n Demis Hassabis and\n Olivier Pietquin and\n Charles Blundell and\n Shane Legg},\n title = {Noisy Networks for Exploration},\n journal = {CoRR},\n volume = {abs/1706.10295},\n year = {2017},\n url = {http://arxiv.org/abs/1706.10295},\n archivePrefix = {arXiv},\n eprint = {1706.10295},\n timestamp = {Mon, 13 Aug 2018 16:46:11 +0200},\n biburl = {https://dblp.org/rec/bib/journals/corr/FortunatoAPMOGM17},\n bibsource = {dblp computer science bibliography, https://dblp.org}\n}\"\"\",\n\n # ALGORITHM\n \"algo-title\": \"Asynchronous Advantage Actor Critic\",\n \"algo-nickname\": \"A3C\",\n \"algo-source-title\": \"Asynchronous Methods for Deep Reinforcement Learning\",\n\n # HYPERPARAMETERS\n \"algo-frames\": 320 * 1000 * 1000, # Number of frames\n\n # SCORE\n \"env-title\": \"atari-space-invaders\",\n \"env-variant\": \"No-op start\",\n \"score\": 1034,\n \"stddev\": 49,\n}\n```\n\nNote that, as shown here, the entry can contain additional information.\n\n## Sources\n\n### Papers\n\n#### Deep Q-Networks\n\n- [x] [Playing Atari with Deep Reinforcement Learning (Mnih et al., 2013)](https://arxiv.org/abs/1312.5602)\n- [x] [Human-level control through deep reinforcement learning (Mnih et al., 2015)](https://deepmind.com/research/dqn/)\n- [x] [Deep Recurrent Q-Learning for Partially Observable MDPs (Hausknecht and Stone, 2015)](https://arxiv.org/abs/1507.06527)\n- [x] [Massively Parallel Methods for Deep Reinforcement Learning (Nair et al., 2015)](https://arxiv.org/abs/1507.04296)\n- [x] [Deep Reinforcement Learning with Double Q-learning (Hasselt et al., 2015)](https://arxiv.org/abs/1509.06461)\n- [x] [Prioritized Experience Replay (Schaul et al., 2015)](https://arxiv.org/abs/1511.05952)\n- [x] [Dueling Network Architectures for Deep Reinforcement Learning (Wang et al., 2015)](https://arxiv.org/abs/1511.06581)\n- [x] [Noisy Networks for Exploration (Fortunato et al., 2017)](https://arxiv.org/abs/1706.10295)\n- [x] [A Distributional Perspective on Reinforcement Learning (Bellemare et al., 2017)](https://arxiv.org/abs/1707.06887)\n- [x] [Rainbow: Combining Improvements in Deep Reinforcement Learning (Hessel et al., 2017)](https://arxiv.org/abs/1710.02298)\n- [x] [Distributional Reinforcement Learning with Quantile Regression (Dabney et al., 2017)](https://arxiv.org/abs/1710.10044)\n- [x] [Implicit Quantile Networks for Distributional Reinforcement Learning (Dabney et al., 2018)](https://arxiv.org/abs/1806.06923)\n\n#### Policy Gradients\n\n- [x] [Asynchronous Methods for Deep Reinforcement Learning (Mnih et al., 2016)](https://arxiv.org/abs/1602.01783)\n- [x] [Trust Region Policy Optimization (Schulman et al., 2015)](https://arxiv.org/abs/1502.05477)\n- [x] [Proximal Policy Optimization Algorithms (Schulman et al., 2017)](https://arxiv.org/abs/1707.06347)\n- [x] [Scalable trust-region method for deep reinforcement learning using Kronecker-factored approximation (Wu et al., 2017)](https://arxiv.org/abs/1708.05144)\n- [x] [Addressing Function Approximation Error in Actor-Critic Methods (Fujimoto et al., 2018)](https://arxiv.org/abs/1802.09477)\n- [x] [IMPALA: Importance Weighted Actor-Learner Architectures (Espeholt et al., 2018)](https://arxiv.org/abs/1802.01561)\n- [x] [The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning (Gruslys et al., 2017)](https://arxiv.org/abs/1704.04651)\n\n#### Exploration\n\n- [x] [Exploration by Random Network Distillation (Burda et al., 2018)](https://arxiv.org/abs/1810.12894)\n\n#### Misc.\n\n- [x] [Trust-PCL: An Off-Policy Trust Region Method for Continuous Control (Nachum et al., 2017)](https://arxiv.org/abs/1707.01891)\n\n### Repositories\n\n- [x] [OpenAI Baselines](https://github.com/openai/baselines)\n- [x] [RL Baselines Zoo](https://github.com/araffin/rl-baselines-zoo)\n\n\n", "description_content_type": "text/markdown", "docs_url": null, "download_url": "", "downloads": { "last_day": -1, "last_month": -1, "last_week": -1 }, "home_page": "https://github.com/seungjaeryanlee/rldb", "keywords": "", "license": "", "maintainer": "", "maintainer_email": "", "name": "rldb", "package_url": "https://pypi.org/project/rldb/", "platform": "", "project_url": "https://pypi.org/project/rldb/", "project_urls": { "Homepage": "https://github.com/seungjaeryanlee/rldb" }, "release_url": "https://pypi.org/project/rldb/0.0.0/", "requires_dist": null, "requires_python": "", "summary": "Performances of Reinforcement Learning Agents", "version": "0.0.0" }, "last_serial": 5275040, "releases": { "0.0.0": [ { "comment_text": "", "digests": { "md5": "3c33243225adc13b22171d58b7a51343", "sha256": "ec2ebf98140759edf515c7f2274b802ea4c12483dd109f4a814ba04aa9a51cfb" }, "downloads": -1, "filename": "rldb-0.0.0-py3-none-any.whl", "has_sig": false, "md5_digest": "3c33243225adc13b22171d58b7a51343", "packagetype": "bdist_wheel", "python_version": "py3", "requires_python": null, "size": 146571, "upload_time": "2019-05-16T02:06:01", "url": "https://files.pythonhosted.org/packages/c7/ab/95db1d80c06b53b232569953cf12baa4ba3f3f60666124ae98583b713be2/rldb-0.0.0-py3-none-any.whl" }, { "comment_text": "", "digests": { "md5": "f662d95c7f10bf62184972bd8acfbbb7", "sha256": "39a7499a04882800c2421a30af2efebb2854c69b75629ff5058084c4766d0817" }, "downloads": -1, "filename": "rldb-0.0.0.tar.gz", "has_sig": false, "md5_digest": "f662d95c7f10bf62184972bd8acfbbb7", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 59185, "upload_time": "2019-05-16T02:06:05", "url": "https://files.pythonhosted.org/packages/6d/59/ece9626bc9e433562490fefe956c09f097bd4b142b37475c21e7af3dd85e/rldb-0.0.0.tar.gz" } ] }, "urls": [ { "comment_text": "", "digests": { "md5": "3c33243225adc13b22171d58b7a51343", "sha256": "ec2ebf98140759edf515c7f2274b802ea4c12483dd109f4a814ba04aa9a51cfb" }, "downloads": -1, "filename": "rldb-0.0.0-py3-none-any.whl", "has_sig": false, "md5_digest": "3c33243225adc13b22171d58b7a51343", "packagetype": "bdist_wheel", "python_version": "py3", "requires_python": null, "size": 146571, "upload_time": "2019-05-16T02:06:01", "url": "https://files.pythonhosted.org/packages/c7/ab/95db1d80c06b53b232569953cf12baa4ba3f3f60666124ae98583b713be2/rldb-0.0.0-py3-none-any.whl" }, { "comment_text": "", "digests": { "md5": "f662d95c7f10bf62184972bd8acfbbb7", "sha256": "39a7499a04882800c2421a30af2efebb2854c69b75629ff5058084c4766d0817" }, "downloads": -1, "filename": "rldb-0.0.0.tar.gz", "has_sig": false, "md5_digest": "f662d95c7f10bf62184972bd8acfbbb7", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 59185, "upload_time": "2019-05-16T02:06:05", "url": "https://files.pythonhosted.org/packages/6d/59/ece9626bc9e433562490fefe956c09f097bd4b142b37475c21e7af3dd85e/rldb-0.0.0.tar.gz" } ] }