{"id":22921034,"url":"https://github.com/yuulis/el_school","last_synced_at":"2026-04-18T01:01:39.616Z","repository":{"id":45190392,"uuid":"410175628","full_name":"Yuulis/EL_School","owner":"Yuulis","description":"Learning how to escape from school as fast as possible.","archived":false,"fork":false,"pushed_at":"2022-01-22T14:51:46.000Z","size":29429,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-06-08T09:45:26.628Z","etag":null,"topics":["csharp","machine-learning","ml","mlagents","python","reinforcement-learning","reinforcement-learning-agent","reinforcement-learning-environments","rl","unity"],"latest_commit_sha":null,"homepage":"","language":"C#","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/Yuulis.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}},"created_at":"2021-09-25T04:32:49.000Z","updated_at":"2022-01-02T12:02:47.000Z","dependencies_parsed_at":"2022-08-30T02:20:58.270Z","dependency_job_id":null,"html_url":"https://github.com/Yuulis/EL_School","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Yuulis/EL_School","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yuulis%2FEL_School","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yuulis%2FEL_School/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yuulis%2FEL_School/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yuulis%2FEL_School/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Yuulis","download_url":"https://codeload.github.com/Yuulis/EL_School/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Yuulis%2FEL_School/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31952206,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-18T00:39:45.007Z","status":"ssl_error","status_checked_at":"2026-04-18T00:39:20.671Z","response_time":62,"last_error":"SSL_read: 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":["csharp","machine-learning","ml","mlagents","python","reinforcement-learning","reinforcement-learning-agent","reinforcement-learning-environments","rl","unity"],"created_at":"2024-12-14T07:18:22.014Z","updated_at":"2026-04-18T01:01:39.591Z","avatar_url":"https://github.com/Yuulis.png","language":"C#","funding_links":[],"categories":[],"sub_categories":[],"readme":"# EL_School\n**Goal**: To learn how to escape from school as fast as possible.  \nUsing [ML-Agents](https://github.com/Unity-Technologies/ml-agents) in [Release 17](https://github.com/Unity-Technologies/ml-agents/tree/release_17).  \n\n## Environments\n1. **School_Only1F** (Curriculum Learning)\n1. **School_From2F** (Curriculum Learning) \u003c\u003c Training now\n\n---\n## Env-1. School_Only1F\n\n### Image\n![Stage](https://user-images.githubusercontent.com/79734873/150643360-931afeef-303c-4b48-a7ca-15c2fe9c80f8.png)\n\n### Environment\n* **Exit1 ~ 3** : When Agent touched them, one episode will success and be ended. \n* **Obstacles** : When Agent touched them, one episode will fail and be ended.\n\n### Agent\n* It spawns random place in this stage.  \n[!] There is spawnable area, made by collider, over the floor. When Agent land on the floor without touching it, Agent spawns again.\n* It uses different brains depending on where it spawns. \n* It can move forward and back and turn around right and left direction (Discrete Action) .\n* It can observe around with ray sensor. This ray is fired at 360 degrees.\n\n### How to train\n* Set three different brains to Agent where it spawns.\n\n\\\u003cCurriculum Training\u003e  \nThere two curriculum parameters : ``SpawnableAreaNum`` and ``StepReward``.  \ncf. ``\\config\\AgentManagerCurriculum.yaml``  \nCurriculum settings is below: \n\n| SpawnableAreaNum | StepReward | Using Behavior | threshold |  \n|:----------------:|:----------:|:--------------:|:---------:|\n| 0.0 (B_StairSide)| -0.0002    | EL_B_StairSide | 0.6       |\n| 1.0 (A_StairSide)| -0.0002    | EL_A_StairSide | 0.5       |\n| 2.0 (C_StairSide)| -0.0002    | EL_C_StairSide | 0.5       |\n| 3.0 (All)        | -0.00025   | One of three   | -         |\n\nTraining starts from ``SpawnableAreaNum = 0``.\n\nMax step of each Behavior is below :  \n\n| Behavior Name | Max Step   |\n|:-------------:|:----------:|\n|EL_B_StairSide | 1,000,000  |\n|EL_A_StairSide | 10,000,000 |\n|EL_C_StairSide | 10,000,000 |\n\n### Rewards\n* Agent gets ``StepReward`` set by Curriculum training at every step.\n* When Agent touches Obstacles, it gets ``-1.0``.\n* If Agent reaches Exit which is closest from where it spawned, it gets ``1.5``. Else, it gets ``0.75``.\n\n### Result\nHere is the result video. *The video is slow, this is due to the specs of my PC :(  \n![result video](https://user-images.githubusercontent.com/79734873/147444470-8b665edb-289f-4361-b69f-fd716cac849f.mp4)  \nLook it on [My Twitter](https://twitter.com/Yuulis04/status/1475024424101621761).\n\nHere are graph :\n\n![Reward graph](/images/graph_reward.png)  \n![Reward graph](/images/graph_episode-length.png)\n\nHere is the scatter plot. Please compare environment map.\n\n![scatter plot](/images/Only1F_result_2021-12-26-16-19-30.png)\n\nFinally here is result of each value.\n\n![scatter plot](/images/Only1F_result_value_2021-12-26-16-19-30.png)\n\n**Agent has a 90% chance to evacuate from 1F of this school!**\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyuulis%2Fel_school","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fyuulis%2Fel_school","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyuulis%2Fel_school/lists"}