{"id":31003968,"url":"https://github.com/timotheemathieu/adaptive_stopping_mc_rl","last_synced_at":"2025-09-13T01:52:06.225Z","repository":{"id":195803356,"uuid":"509134410","full_name":"TimotheeMathieu/Adaptive_stopping_MC_RL","owner":"TimotheeMathieu","description":null,"archived":false,"fork":false,"pushed_at":"2024-02-26T15:23:45.000Z","size":382595,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":4,"default_branch":"paper_branch","last_synced_at":"2024-02-26T16:40:00.506Z","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/TimotheeMathieu.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}},"created_at":"2022-06-30T15:27:32.000Z","updated_at":"2023-09-19T13:52:56.000Z","dependencies_parsed_at":null,"dependency_job_id":"ba6ef3ee-2dad-4246-b763-16ce2f02caf4","html_url":"https://github.com/TimotheeMathieu/Adaptive_stopping_MC_RL","commit_stats":null,"previous_names":["timotheemathieu/adaptive_stopping_mc_rl"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/TimotheeMathieu/Adaptive_stopping_MC_RL","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TimotheeMathieu%2FAdaptive_stopping_MC_RL","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TimotheeMathieu%2FAdaptive_stopping_MC_RL/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TimotheeMathieu%2FAdaptive_stopping_MC_RL/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TimotheeMathieu%2FAdaptive_stopping_MC_RL/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/TimotheeMathieu","download_url":"https://codeload.github.com/TimotheeMathieu/Adaptive_stopping_MC_RL/tar.gz/refs/heads/paper_branch","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TimotheeMathieu%2FAdaptive_stopping_MC_RL/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":274907868,"owners_count":25371822,"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","status":"online","status_checked_at":"2025-09-12T02:00:09.324Z","response_time":60,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"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":"2025-09-13T01:51:42.530Z","updated_at":"2025-09-13T01:52:06.206Z","avatar_url":"https://github.com/TimotheeMathieu.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"This repository can be used to reproduce all the experiments in the paper [AdaStop: sequential testing for efficient and reliable comparisons of Deep RL Agents](https://arxiv.org/abs/2306.10882).\n\n### Structure of the repository\n\n- Folder `toy_examples_xp` contains code for Section 6.1 \"Toy examples\"\n- Folder `Non_adaptive_comparison_xp` contains code for Section 6.2 \"Comparison with non-adaptive approach\"\n- Folder `Deep_RL_xp` contains code for Section 6.3 \"AdaStop for Deep Reinforcement Learning\"\n\nWe provide a `requirements.txt` file which contains the dependencies and their versions at the time of the article.\n\n### What is reproducible?\nWe aim at having a completely reproducible paper. All the scripts for all the experiments are available and using the makefile all the figures and results of the paper can be reproduced. One can use `make all` (on linux) to do the figures from the article with a reduced size of the experiments so that it runs in a reasonable amount of time, using a virtual environment that is constructed in tmp folder. FOr more courageous people, you can use `make ARG=--full-xp all` which does the exact same figures as the ones in the article.\n\nRemark: due to its complexity, the deep rl experiments from Section 6.3. are not automatically recomputed, we only recompute the adastop algorithm and not the actual training of the agents.\n\n### Using Guix\nFor (almost) perfect reproducibility, you can use guix with the following commands to fix the versions of the shared libraries (e.g. C libraries)\n```\nguix time-machine --channels=channels.scm --  shell -CF -m manifest.scm\nexport LD_LIBRARY_PATH=/lib\nmake all \n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftimotheemathieu%2Fadaptive_stopping_mc_rl","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftimotheemathieu%2Fadaptive_stopping_mc_rl","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftimotheemathieu%2Fadaptive_stopping_mc_rl/lists"}