{"id":13648354,"url":"https://github.com/hzwer/NIPS2017-LearningToRunACE","last_synced_at":"2025-04-22T07:31:22.487Z","repository":{"id":98075407,"uuid":"111886128","full_name":"hzwer/NIPS2017-LearningToRunACE","owner":"hzwer","description":"2nd place solution of NIPS2017 LearningToRun Competition.","archived":false,"fork":false,"pushed_at":"2022-05-29T10:09:07.000Z","size":14935,"stargazers_count":124,"open_issues_count":1,"forks_count":28,"subscribers_count":7,"default_branch":"master","last_synced_at":"2025-04-12T17:37:14.384Z","etag":null,"topics":["deep-learning","keras","reinforcement-learning"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/hzwer.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null}},"created_at":"2017-11-24T06:54:22.000Z","updated_at":"2025-01-21T03:56:28.000Z","dependencies_parsed_at":null,"dependency_job_id":"f9cfe4a2-ea3b-493b-b2c6-761ec54f84cf","html_url":"https://github.com/hzwer/NIPS2017-LearningToRunACE","commit_stats":null,"previous_names":["hzwer/nips2017-learningtorunace","megvii-research/nips2017-learningtorunace"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FNIPS2017-LearningToRunACE","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FNIPS2017-LearningToRunACE/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FNIPS2017-LearningToRunACE/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FNIPS2017-LearningToRunACE/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hzwer","download_url":"https://codeload.github.com/hzwer/NIPS2017-LearningToRunACE/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":250195033,"owners_count":21390230,"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","keras","reinforcement-learning"],"created_at":"2024-08-02T01:04:10.280Z","updated_at":"2025-04-22T07:31:21.438Z","avatar_url":"https://github.com/hzwer.png","language":"Python","readme":"# NIPS2017-LearningToRun with ACE\n[Zhewei Huang](https://scholar.google.com/citations?user=zJEkaG8AAAAJ\u0026hl=zh-CN\u0026oi=sra), [Shuchang Zhou](https://scholar.google.com/citations?user=zYI0rysAAAAJ\u0026hl=zh-CN\u0026oi=sra), BoEr Zhuang, [Xinyu Zhou](https://scholar.google.com/citations?user=Jv4LCj8AAAAJ\u0026hl=zh-CN\u0026oi=ao)\n\n![Demo](https://github.com/hzwer/NIPS2017-LearningToRun/raw/master/demo/hzwer-NIPS2017-LearningToRun-small.gif)\n\nA keras solution for 2nd place [NIPS RL 2017 challenge](https://www.crowdai.org/challenges/nips-2017-learning-to-run/leaderboards?challenge_round_id=12).\n\nThere is a [slide](https://docs.google.com/presentation/d/1dgXDFlr62jQ-OdEoYVCGwuUgux3u-jrMaXVp94OVOSk/edit?usp=sharing), a [lecture](https://drive.google.com/open?id=15_XBOms-T1G1jeiDm7xGTn2JGQ2FviT5) and a [writeup(arxiv)](https://arxiv.org/abs/1712.08987) about our work.\n\n## To Run\n### preparation\n\nThese instructions expect that opensim-rl conda environment is already setup as described in : https://github.com/stanfordnmbl/osim-rl/ .\n\n```\n$ source activate opensim-rl\n```\n\nOther dependencies is needed as follow\n* Keras(since old version does not support selu activation)\n* TensorFlow\n* matplotlib\n* numpy\n* Pyro4\n* parse\n* pymsgbox(optional)\n\n### parallelism\n\nThis version requires farming, before starting `train.py`, you should first start some farms by running `python farm.py` on each \u003cu\u003eSLAVE\u003c/u\u003e machine you own. Then  create a `farmlist.py` in the working directory (on the \u003cu\u003eHOST\u003c/u\u003e machine) with the following content :\n\n```\nfarmlist_base = [('127.0.0.1', 4), ('192.168.1.1', 8)]\n\n# a farm of 4 cores is available on localhost, while a farm of 8 is available on another machine.\n\n# expand the list if you have more machines.\n\n# this file will be consumed by the host to find the slaves.\n```\nTry `python farm.py --help` to get more information about how to set the environment.\n\nMore information can be found in https://github.com/ctmakro/stanford-osrl .\n\nThanks to @ctmakro for providing us with this frame.\n\n### test\n\nTest the model in parallel and calculate the average score.\n\nWe provide you with some [trained parameters](https://drive.google.com/open?id=10RDVQA5zjUjNXz7Igak3k92_s_XKI2Uw).\n\n```\npython test.py -a=10 -c=5 -t=200 -p logs\n\n# test the model for 200 times with 10 actor networks and 5 critic networks ensemble\n\n# the network parameters should be placed as logs/actormodel1.h5 ... logs/actormodel10.h5\n```\n\nTry `python test.py --help` to get more information .\n\n## Contributors\n\n- [hzwer](https://github.com/hzwer)\n- [floz](https://github.com/NewGod)\n\n## Resources\n- [Official YouTube demos](https://www.youtube.com/watch?v=rhNxt0VccsE)\n- [赛后专访](https://www.leiphone.com/news/201711/b2OfTdcMUmpYKx6S.html)\n- Also a teaser video from the winning team https://www.bilibili.com/video/BV1jE411B74u\n","funding_links":[],"categories":["Python"],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhzwer%2FNIPS2017-LearningToRunACE","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhzwer%2FNIPS2017-LearningToRunACE","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhzwer%2FNIPS2017-LearningToRunACE/lists"}