{"id":13668839,"url":"https://github.com/danaugrs/huskarl","last_synced_at":"2025-04-04T11:16:24.947Z","repository":{"id":34204471,"uuid":"171322847","full_name":"danaugrs/huskarl","owner":"danaugrs","description":"Deep Reinforcement Learning Framework + Algorithms","archived":false,"fork":false,"pushed_at":"2023-03-25T00:07:33.000Z","size":44166,"stargazers_count":414,"open_issues_count":14,"forks_count":51,"subscribers_count":38,"default_branch":"master","last_synced_at":"2025-03-28T05:32:19.932Z","etag":null,"topics":["algorithms","artificial-intelligence","deep-learning","python","reinforcement-learning","tensorflow"],"latest_commit_sha":null,"homepage":"https://medium.com/@tensorflow/introducing-huskarl-the-modular-deep-reinforcement-learning-framework-e47d4b228dd3","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/danaugrs.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2019-02-18T17:01:53.000Z","updated_at":"2025-03-21T16:08:10.000Z","dependencies_parsed_at":"2023-12-09T11:43:31.248Z","dependency_job_id":null,"html_url":"https://github.com/danaugrs/huskarl","commit_stats":{"total_commits":44,"total_committers":4,"mean_commits":11.0,"dds":0.06818181818181823,"last_synced_commit":"8708d967dc7db785d5aac65c28b896a6d0739e45"},"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/danaugrs%2Fhuskarl","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/danaugrs%2Fhuskarl/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/danaugrs%2Fhuskarl/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/danaugrs%2Fhuskarl/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/danaugrs","download_url":"https://codeload.github.com/danaugrs/huskarl/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247166169,"owners_count":20894654,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["algorithms","artificial-intelligence","deep-learning","python","reinforcement-learning","tensorflow"],"created_at":"2024-08-02T08:00:50.514Z","updated_at":"2025-04-04T11:16:24.925Z","avatar_url":"https://github.com/danaugrs.png","language":"Python","funding_links":[],"categories":["Libraries made for TensorFlow 2.0","Python","Other 💛💛💛💛💛\u003ca name=\"Other\" /\u003e"],"sub_categories":["强化学习"],"readme":"\u003cimg align=\"left\" src=\"https://github.com/danaugrs/huskarl/blob/master/logo.png\"\u003e\n\n# Huskarl [![PyPI version](https://badge.fury.io/py/huskarl.svg)](https://badge.fury.io/py/huskarl)\n\nHuskarl is a framework for deep reinforcement learning focused on modularity and fast prototyping.\nIt's built on TensorFlow 2.0 and uses the `tf.keras` API when possible for conciseness and readability.\n\nHuskarl makes it easy to parallelize computation of environment dynamics across multiple CPU cores.\nThis is useful for speeding up on-policy learning algorithms that benefit from multiple concurrent sources of experience such as A2C or PPO.\nIt is especially useful for computationally intensive environments such as physics-based ones.\n\nHuskarl works seamlessly with [OpenAI Gym](https://gym.openai.com/) environments.\n\nThere are plans to support multi-agent environments and [Unity3D environments](https://unity3d.ai).\n\n## Algorithms\n\nSeveral algorithms are implemented and more are planned.\n\n* [x] Deep Q-Learning Network (DQN)\n* [x] Multi-step DQN\n* [x] Double DQN\n* [x] Dueling Architecture DQN\n* [x] Advantage Actor-Critic (A2C)\n* [x] Deep Deterministic Policy Gradient (DDPG)\n* [x] Prioritized Experience Replay\n* [ ] Proximal Policy Optimization (PPO)\n* [ ] Curiosity-Driven Exploration\n\n\n## Installation\nYou can install the latest version from source with:\n```\ngit clone https://github.com/danaugrs/huskarl.git\ncd huskarl\npip install -e .\n```\nIf you prefer, you can get the packaged version from [PyPI](https://pypi.org/project/huskarl/):\n```\npip install huskarl\n```\n\n## Examples\nThere are three examples included - one for each implemented agent type. To run the examples you will need [`matplotlib`](https://github.com/matplotlib/matplotlib) and [`gym`](https://github.com/openai/gym) installed. \n\n### [dqn-cartpole.py](https://github.com/danaugrs/huskarl/blob/master/examples/dqn-cartpole.py)\n![dqn-cartpole.gif](examples/dqn-cartpole.gif)\n### [ddpg-pendulum.py](https://github.com/danaugrs/huskarl/blob/master/examples/ddpg-pendulum.py)\n![ddpg-pendulum.gif](examples/ddpg-pendulum.gif)\n### [a2c-cartpole.py](https://github.com/danaugrs/huskarl/blob/master/examples/a2c-cartpole.py)\n![a2c-cartpole.gif](examples/a2c-cartpole.gif)\n\n## Citing\n\nIf you use Huskarl in your research, you can cite it as follows:\n```\n@misc{salvadori2019huskarl,\n    author = {Daniel Salvadori},\n    title = {huskarl},\n    year = {2019},\n    publisher = {GitHub},\n    journal = {GitHub repository},\n    howpublished = {\\url{https://github.com/danaugrs/huskarl}},\n}\n```\n\n## About\n\n_hùskarl_ in Old Norse means a warrior who works in his/her lord's service.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdanaugrs%2Fhuskarl","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdanaugrs%2Fhuskarl","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdanaugrs%2Fhuskarl/lists"}