{"id":32587033,"url":"https://github.com/mattjmattj/php-rl","last_synced_at":"2026-04-18T03:33:01.718Z","repository":{"id":78891151,"uuid":"259303469","full_name":"mattjmattj/php-rl","owner":"mattjmattj","description":"A basic reinforcement learning library in PHP","archived":false,"fork":false,"pushed_at":"2020-05-26T22:34:44.000Z","size":52,"stargazers_count":4,"open_issues_count":0,"forks_count":2,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-10-29T22:53:49.254Z","etag":null,"topics":["artificial-intelligence","ddqn","double-dqn","dqn","machine-learning","neural-network","prioritized-experience-replay","qlearning","reinforcement-learning","rl","sarsa"],"latest_commit_sha":null,"homepage":"","language":"PHP","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/mattjmattj.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,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2020-04-27T12:09:29.000Z","updated_at":"2024-05-08T20:22:45.000Z","dependencies_parsed_at":"2023-03-05T16:15:48.066Z","dependency_job_id":null,"html_url":"https://github.com/mattjmattj/php-rl","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/mattjmattj/php-rl","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mattjmattj%2Fphp-rl","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mattjmattj%2Fphp-rl/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mattjmattj%2Fphp-rl/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mattjmattj%2Fphp-rl/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/mattjmattj","download_url":"https://codeload.github.com/mattjmattj/php-rl/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mattjmattj%2Fphp-rl/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31955740,"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":"online","status_checked_at":"2026-04-18T02:00:07.018Z","response_time":103,"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":["artificial-intelligence","ddqn","double-dqn","dqn","machine-learning","neural-network","prioritized-experience-replay","qlearning","reinforcement-learning","rl","sarsa"],"created_at":"2025-10-29T22:53:43.018Z","updated_at":"2026-04-18T03:33:01.712Z","avatar_url":"https://github.com/mattjmattj.png","language":"PHP","funding_links":[],"categories":[],"sub_categories":[],"readme":"# php-rl\n\nA reinforcement learning library in PHP\n\n## Disclaimer\n\nThis library is basically me reimplementing well-known RL algorithms in order to better\nunderstand them.\n\n## Algorithms\n\n### Value-based algorithms\n\n#### SARSA\n\nA standard state-action-reward-state-action implementation based on a Q table\n\n#### Q-Learning\n\nBased on a Q-table implemented as a \"max\" policy SARSA.\nCurrent API provides a basic epsilon-greedy agent.\nSee the Tic-Tac-Toe example for some details\n\n\n#### Deep Q-Learning\n\nCurrent API provides a basic epsilon-greedy agent, with separated target model, as described\nin Mnih, V., Kavukcuoglu, K., Silver, D. _et al_. Human-level control through deep reinforcement learning. _Nature_ **518**, 529–533 (2015). https://doi.org/10.1038/nature14236.\n\nUser can choose between a Vanilla DQN ou a Double DQN, (see Hado van Hasselt, Arthur Guez, David Silver. Deep Reinforcement Learning with Double Q-learning. [arXiv:1509.06461](https://arxiv.org/abs/1509.06461) [cs.LG])\n\nExperience replay is available as 2 distinct implementations:\n- random minibatch\n- prioritized experience replay (Tom Schaul, John Quan, Ioannis Antonoglou, David Silver - Prioritized Experience Replay, [arXiv:1511.05952](https://arxiv.org/abs/1511.05952) [cs.LG], 2015)\n\n### Policy-based algorithms\n\nTODO\n\n## TODO\n- ~~Q-learning~~\n- ~~SARSA~~\n- ~~DQN~~\n- ~~Double DQN~~\n- ~~[DQN] prioritized experience replay~~\n- Vanilla Policy Gradient (REINFORCE)\n- Actor-Critic\n- real documentation :)\n- more examples\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmattjmattj%2Fphp-rl","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmattjmattj%2Fphp-rl","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmattjmattj%2Fphp-rl/lists"}