{"id":23927523,"url":"https://github.com/bhansconnect/alpha_zero_othello","last_synced_at":"2025-07-11T07:06:25.097Z","repository":{"id":28108727,"uuid":"114480151","full_name":"bhansconnect/alpha_zero_othello","owner":"bhansconnect","description":"A functional Alpha Zero that plays Othello using Keras","archived":false,"fork":false,"pushed_at":"2022-12-07T23:48:01.000Z","size":21659,"stargazers_count":116,"open_issues_count":9,"forks_count":23,"subscribers_count":15,"default_branch":"master","last_synced_at":"2025-07-04T18:58:39.361Z","etag":null,"topics":["alphazero","deep-learning","machine-learning","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":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/bhansconnect.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":"2017-12-16T17:58:40.000Z","updated_at":"2025-04-14T03:04:45.000Z","dependencies_parsed_at":"2022-08-30T09:21:49.297Z","dependency_job_id":null,"html_url":"https://github.com/bhansconnect/alpha_zero_othello","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/bhansconnect/alpha_zero_othello","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bhansconnect%2Falpha_zero_othello","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bhansconnect%2Falpha_zero_othello/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bhansconnect%2Falpha_zero_othello/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bhansconnect%2Falpha_zero_othello/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/bhansconnect","download_url":"https://codeload.github.com/bhansconnect/alpha_zero_othello/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bhansconnect%2Falpha_zero_othello/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":264753075,"owners_count":23658790,"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":["alphazero","deep-learning","machine-learning","reinforcement-learning"],"created_at":"2025-01-05T22:43:54.076Z","updated_at":"2025-07-11T07:06:25.075Z","avatar_url":"https://github.com/bhansconnect.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Alpha Zero Othello\n\n\u003e There are 2 newer versions of this project that are both significantly faster: \u003cbr\u003e - [bhansconnect/fast-alphazero-general](https://github.com/bhansconnect/fast-alphazero-general) is completely written in python, but is much faster and more extensible than this library. \u003cbr\u003e - [bhansconnect/alphazero-pybind11](https://github.com/bhansconnect/alphazero-pybind11) has large parts written in C++ and may be less beginner friendly, but totally crushes previous versions in performance. \n\nIf you like this project, please look at my other project, [Expert Iterations General](https://github.com/bhansconnect/expert-iteration-general). It is similar to this one, but uses Pytorch and is built better.\n\nA Python program for learning to play Othello/Reversi from zero. The best part is that it is easy\nto plug this same algorithm into many other games. If you have any questions. feel free to contact me:\nbrendan.hansknecht@gmail.com\n\nThe current best version is in the root directory as model-best.h5 To use it, simply copy it to into data\\models.\nOnce it is copied over, you can train based off of it or play with it using the commands below. This model was only\ntrained for 64 iterations. This equates to approximately 30000 games, which is not a lot in terms of rl. It is definitely\ngetting better, but is not super good.\n\nTo install required libraries run: `pip install -r requirements.txt`\n\n\nThis version currently uses a smaller neural network for testing purposes as things are confirmed\nto be working, I will upgrade to full size.\n\nCurrently run by calling: `python run.py {self, opt, eval, play, rank, compile_rank}`\n\n* self for it to generate self play games.\n* opt for it to update the network.\n* eval to compare different versions.\n* play to play against it yourself.\n* rank to compare multiple models\n* compile_rank to load all ranking/eval history into one comparison\n\nI generally run 2 instances self playing games and 1 optimizing while training it.\nConfig.py has all of the options. \n\nAs a side note, it is possible to train on multiple servers by setting up a shared drive between multiple computers\nfrom that shared drive, you can run the program on multiple computers and they will all generate self play games.\nThat being said, you would still be limited to one optimizer. Luckily, it takes much less time to optimize than\nself play games.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbhansconnect%2Falpha_zero_othello","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbhansconnect%2Falpha_zero_othello","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbhansconnect%2Falpha_zero_othello/lists"}