{"id":17595151,"url":"https://github.com/takuseno/ppo","last_synced_at":"2025-04-12T07:12:46.720Z","repository":{"id":80512654,"uuid":"108955381","full_name":"takuseno/ppo","owner":"takuseno","description":"Proximal Policy Optimization implementation with TensorFlow","archived":false,"fork":false,"pushed_at":"2018-10-09T12:21:24.000Z","size":62,"stargazers_count":106,"open_issues_count":1,"forks_count":22,"subscribers_count":5,"default_branch":"master","last_synced_at":"2025-04-12T07:12:41.244Z","etag":null,"topics":["reinforcement-learning","tensorflow"],"latest_commit_sha":null,"homepage":"","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/takuseno.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,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2017-10-31T06:35:10.000Z","updated_at":"2025-03-13T00:27:40.000Z","dependencies_parsed_at":"2023-04-01T15:34:46.652Z","dependency_job_id":null,"html_url":"https://github.com/takuseno/ppo","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/takuseno%2Fppo","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/takuseno%2Fppo/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/takuseno%2Fppo/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/takuseno%2Fppo/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/takuseno","download_url":"https://codeload.github.com/takuseno/ppo/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248530571,"owners_count":21119600,"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":["reinforcement-learning","tensorflow"],"created_at":"2024-10-22T07:37:59.922Z","updated_at":"2025-04-12T07:12:46.691Z","avatar_url":"https://github.com/takuseno.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"## PPO\nProximal Policy Optimization implementation with Tensorflow.\n\nhttps://arxiv.org/pdf/1707.06347.pdf\n\nThis repository has been much updated from commit id `a4fbd383f0f89ce2d881a8b78d6b8a03294e5c7c` .\nNew PPO requires a new dependency, [rlsaber](https://github.com/imai-laboratory/rlsaber) which is my utility repository that can be shared across different algorithms.\n\nSome of my design follow [OpenAI baselines](https://github.com/openai/baselines).\nBut, I used as many default tensorflow packages as possible unlike baselines, that makes my codes easier to be read.\n\nIn addition, my PPO automatically switches between continuous action-space and discrete action-space depending on environments.\nIf you want to change hyper parameters, check `atari_constants.py` or `box_constants.py`, which will be loaded depending on environments too.\n\n## requirements\n- Python3\n\n## dependencies\n- tensorflow\n- gym[atari]\n- opencv-python\n- git+https://github.com/imai-laboratory/rlsaber\n\n## usage\n### training\n```\n$ python train.py [--env env-id] [--render] [--logdir log-name]\n```\nexample\n```\n$ python train.py --env BreakoutNoFrameskip-v4 --logdir breakout\n```\n\n### playing\n```\n$ python train.py --demo --load results/path-to-model [--env env-id] [--render]\n```\nexample\n```\n$ python train.py --demo --load results/breakout/model.ckpt-xxxx --env BreakoutNoFrameskip-v4 --render\n```\n\n### performance examples\n#### Pendulumn-v0\n![image](https://user-images.githubusercontent.com/5235131/46388030-e4f72980-c704-11e8-9d76-1790dcb88067.png)\n\n#### BreakoutNoFrameskip-v4\n![image](https://user-images.githubusercontent.com/5235131/46402330-6321f300-c73a-11e8-9b46-46959bce4c3d.png)\n\n\n### implementation\nThis is inspired by following projects.\n\n- [DQN](https://github.com/imai-laboratory/dqn)\n- [OpenAI Baselines](https://github.com/openai/baselines)\n\n## License\nThis repository is MIT-licensed.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftakuseno%2Fppo","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftakuseno%2Fppo","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftakuseno%2Fppo/lists"}