{"id":13935135,"url":"https://github.com/rmst/ddpg","last_synced_at":"2025-08-20T16:31:07.389Z","repository":{"id":46214816,"uuid":"58937063","full_name":"rmst/ddpg","owner":"rmst","description":"TensorFlow implementation of the DDPG algorithm from the paper Continuous Control with Deep Reinforcement Learning (ICLR 2016)","archived":false,"fork":false,"pushed_at":"2018-02-16T16:57:33.000Z","size":14175,"stargazers_count":212,"open_issues_count":4,"forks_count":64,"subscribers_count":16,"default_branch":"master","last_synced_at":"2024-12-11T18:43:28.677Z","etag":null,"topics":["deep-learning","reinforcement-learning","tensorflow"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","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/rmst.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.md","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2016-05-16T14:02:59.000Z","updated_at":"2024-12-01T13:50:03.000Z","dependencies_parsed_at":"2022-09-25T08:38:58.738Z","dependency_job_id":null,"html_url":"https://github.com/rmst/ddpg","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rmst%2Fddpg","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rmst%2Fddpg/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rmst%2Fddpg/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rmst%2Fddpg/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/rmst","download_url":"https://codeload.github.com/rmst/ddpg/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":230438185,"owners_count":18225870,"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":["deep-learning","reinforcement-learning","tensorflow"],"created_at":"2024-08-07T23:01:24.934Z","updated_at":"2024-12-19T13:07:02.819Z","avatar_url":"https://github.com/rmst.png","language":"Jupyter Notebook","funding_links":[],"categories":["Jupyter Notebook"],"sub_categories":[],"readme":"# Deep Deterministic Policy Gradient\n\n__Warning: This repo is no longer maintained. For a more recent (and improved) implementation of DDPG see https://github.com/openai/baselines/tree/master/baselines/ddpg .__\n\nPaper: [\"Continuous control with deep reinforcement learning\" - TP Lillicrap, JJ Hunt et al., 2015](http://arxiv.org/abs/1509.02971)\n\n\u003ctable\u003e\n  \u003ctr\u003e\n    \u003ctd\u003e\u003cimg src=\"./readme/ipend.gif?raw=true\" width=\"200\"\u003e\u003c/td\u003e\n    \u003ctd\u003e\u003cimg src=\"./readme/reacher.gif?raw=true\" width=\"200\"\u003e\u003c/td\u003e\n    \u003ctd\u003e\u003cimg src=\"./readme/pend.gif?raw=true\" width=\"200\"\u003e\u003c/td\u003e\n  \u003c/tr\u003e\n\u003c/table\u003e\n\n### Installation\nInstall [Gym](https://github.com/openai/gym#installation) and [TensorFlow](https://www.tensorflow.org/versions/r0.9/get_started/os_setup.html). Then:\n\n```bash\npip install pyglet # required for gym rendering\npip install jupyter # required only for visualization (see below)\n\ngit clone https://github.com/SimonRamstedt/ddpg.git # get ddpg\n```\n\n### Usage\nExample:\n```bash\npython run.py --outdir ../ddpg-results/experiment1 --env InvertedDoublePendulum-v1\n```\nEnter `python run.py -h` to get a complete overview.\n\nIf you want to run in the cloud or a university cluster [this](https://github.com/SimonRamstedt/ddpg-darmstadt) might contain additional information.\n\n### Visualization\n\n\u003cimg src=\"./readme/db.png\" width=\"800\"\u003e\n\nExample:\n```bash\npython dashboard.py --exdir ../ddpg-results/+\n```\nEnter `python dashboard.py -h` to get a complete overview.\n\n### Known issues\n- No batch normalization yet\n- No conv nets yet (i.e. only learning from low dimensional states)\n- No proper seeding for reproducibilty\n\n*Please write me or open a github issue if you encounter problems! Contributions are welcome!*\n\n### Improvements beyond the original paper\n- [Output normalization](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Publications_files/popart.pdf) – the main reason for divergence are variations in return scales. Output normalization would probably solve this.\n- [Prioritized experience replay](http://arxiv.org/abs/1511.05952) – faster learning, better performance especially with sparse rewards – *Please write if you have/know of an implementation!*\n\n\n### Advaned Usage\nRemote execution:\n```bash\npython run.py --outdir your_username@remotehost.edu:/some/remote/directory/+ --env InvertedDoublePendulum-v1\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frmst%2Fddpg","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frmst%2Fddpg","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frmst%2Fddpg/lists"}