{"id":21739566,"url":"https://github.com/berkeleyautomation/skillhub","last_synced_at":"2026-05-08T19:32:34.409Z","repository":{"id":148172798,"uuid":"175808844","full_name":"BerkeleyAutomation/SkillHub","owner":"BerkeleyAutomation","description":null,"archived":false,"fork":false,"pushed_at":"2019-03-15T11:37:48.000Z","size":494,"stargazers_count":2,"open_issues_count":0,"forks_count":2,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-10-06T01:24:08.040Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"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/BerkeleyAutomation.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,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2019-03-15T11:36:31.000Z","updated_at":"2020-07-15T06:22:11.000Z","dependencies_parsed_at":"2023-05-19T08:45:26.053Z","dependency_job_id":null,"html_url":"https://github.com/BerkeleyAutomation/SkillHub","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/BerkeleyAutomation/SkillHub","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BerkeleyAutomation%2FSkillHub","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BerkeleyAutomation%2FSkillHub/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BerkeleyAutomation%2FSkillHub/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BerkeleyAutomation%2FSkillHub/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/BerkeleyAutomation","download_url":"https://codeload.github.com/BerkeleyAutomation/SkillHub/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BerkeleyAutomation%2FSkillHub/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32794633,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-08T08:22:46.396Z","status":"ssl_error","status_checked_at":"2026-05-08T08:22:45.650Z","response_time":54,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"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":[],"created_at":"2024-11-26T06:09:27.149Z","updated_at":"2026-05-08T19:32:34.392Z","avatar_url":"https://github.com/BerkeleyAutomation.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Multi-Task Hierarchical Imitation Learning of Robot Skills\n\nCode for running the *SkillHub* server, generating annotated demonstrations, learning hierarchical controllers, and rolloing learned controllers on an HSR robot.\n\n## Installation\n\nIn addition to the dependencies in [requirements.txt](requirements.txt), the [`HSREnv`](envs/hsr.py#L202) depends on [pyyolo](https://github.com/digitalbrain79/pyyolo).\nPlease follow [these instructions](https://github.com/digitalbrain79/pyyolo#building) to install pyyolo, and then change the paths in [vision.py](envs/vision.py) to your installation path.\n\n## Running SkillHub\n\nTo run the SkillHub server:\n\n`python skillhub/server.py`\n\nTo run the server in debug mode, set the `DEBUG` flag to `True` in [server.py](skillhub/server.py#L21).\n\n## Providing demonstrations\n\nTo rollout a demonstration of the `SetTable` task:\n\n`python rollout.py --domain dishes --task SetTable --data data --teacher`\n\nAvailable tasks are: `ClearTable` and `SetTable` in the `dishes` domain; and `Pyramid\u003cn\u003e` (pyramid of height `n`) in the `pyramid` domain.\nAnnotated demonstrations will be saved in the path provided to `--data`.\n\nTo implement new domains, inherit from [`Env`](envs/env.py#L5) (or its subclass [`HSREnv`](envs/hsr.py#L202)) -- see [`PyramidEnv`](envs/pyramid.py#L4) for example.\nTo implement new tasks, inherit from [`Agent`](agents/agent.py#L6) (or its subclass [`HierarchicalAgent`](agents/hierarchical.py#L8)) -- see [`PyramidAgent`](agents/pyramid.py#L229) for example.\n\nHuman teleoperation is performed by actions starting with `Record_`, which pause the script to allow human control and record it.\nBefore training, the recorded demonstrations must be fixed to format the recorded control as robot control.\nTo fix all demonstrations in a path:\n\n`python -c \"import fix; fix.fix_record('data/dishes/SetTable_fix', 'data/dishes/SetTable_rec')\"`\n\n## Training hierarchical controller\n\nTo train controllers for the `ClearTable` and `SetTable` tasks:\n\n`python train.py --domain dishes --tasks ClearTable SetTable --data data`\n\nAll controllers but the last will be trained independently from all their available data.\nThe last controller (`SetTable` in this case) will be trained with *detailed mode selection* with the data of past tasks and with new demonstrations added one by one.\nThe results of training will be saved in `results/dishes/ClearTable.SetTable`.\n\nUse `--runs \u003cn\u003e` to repeat training for `n` independent trials.\nUse `--independent` to train the last controller only independently of the other tasks.\nUse `--full-batch` to train the last controller from all available data, rather than stopping when enough demonstrations were given to successfully train and validate the controller.\n\n## Evaluating a trained controller\n\nTo rollout a trained controller:\n\n`python rollout.py --domain dishes --task SetTable --model model --data eval`\n\nData from the evaluation experiment will be saved in the path provided to `--data`.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fberkeleyautomation%2Fskillhub","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fberkeleyautomation%2Fskillhub","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fberkeleyautomation%2Fskillhub/lists"}