[{"_1":2,"_176":-5,"_177":-5},"loaderData",{"_3":4,"_17":18,"_25":26,"_151":152},"root",{"_5":-5,"_6":-7,"_7":8,"_9":-7,"_10":11,"_12":-7,"_13":14,"_15":16},"currentUser","primaryEmail","siteVerification","6LeIxAcTAAAAAJcZVRqyHh71UMIEGNQ_MXjiZKhI","toast","enableDev",false,"incidentStatus","theme","system","contrast","normal","routes/_layout",{"_19":-5,"_20":21,"_22":21,"_23":11,"_24":21},"username","isAskGordonEnabled",true,"isUsingNewTemplate","showMinimalLanding","showHoneypotTrap","routes/_layout.r.$namespace.$reponame",{"_27":28,"_29":-5,"_30":11,"_31":11,"_32":33,"_34":35,"_36":-5,"_37":11,"_38":39,"_88":-5,"_89":90,"_139":-5,"_140":11,"_141":21,"_142":11,"_143":11,"_144":-5,"_145":11,"_146":-5,"_147":148,"_149":91,"_150":-7},"canonicalUrl","https://hub.docker.com/r/sbugallo/ttt/tags","mcpServer","canSeeDockerfile","canSeeAutobuilds","namespace","sbugallo","reponame","ttt","cloudBuildSource","isAutobuildsAvailable","repo",{"_40":41,"_42":43,"_45":46,"_47":48,"_49":11,"_50":51,"_52":53,"_54":55,"_57":35,"_32":33,"_58":59,"_60":44,"_61":62,"_63":64,"_65":66,"_67":-5,"_68":62,"_69":70,"_71":11,"_72":33,"_73":74,"_79":11,"_80":81,"_85":33,"_86":87},"categories",[],"content_types",[44],"image","date_registered","2019-08-24T09:21:34.040395Z","description","Python implementation of Tic-Tac-Toe game alongside a Markov Decision Process-based AI","is_private","last_updated","2019-08-28T22:42:16.888071Z","last_modified","2024-10-16T13:48:34.145251Z","media_types",[56],"application/vnd.docker.container.image.v1+json","name","pull_count",1032,"repository_type","star_count",0,"status",1,"status_description","active","source","collaborator_count","full_description","# Tic-Tac-Toe AI - Markov Decision Process\n\n[![CodeFactor](https://www.codefactor.io/repository/github/sbugallo/tictactoe-mdp/badge)](https://www.codefactor.io/repository/github/sbugallo/tictactoe-mdp)\n[![Build Status](https://travis-ci.com/sbugallo/TicTacToe-MDP.svg)](https://travis-ci.com/sbugallo/TicTacToe-MDP)\n[![Coverage Status](https://coveralls.io/repos/github/sbugallo/TicTacToe-MDP/badge.svg?branch=dev)](https://coveralls.io/github/sbugallo/TicTacToe-MDP?branch=dev)\n\n

\n\nSee [documentation](https://sbugallo.github.io/TicTacToe-MDP/)\n## Game description\n\n\"Tic-tac-toe (American English), noughts and crosses (British English), or Xs and Os is a paper-and-pencil game for two\nplayers, X and O, who take turns marking the spaces in a 3×3 grid. The player who succeeds in placing three of their\nmarks in a horizontal, vertical, or diagonal row wins the game.\"\n\n-- Wikipedia\n\n## Installation\n\n### Prerequisites\n\n- `conda` (included in Miniconda 3 and Anaconda 3 distributions)\n\n### Steps\n\n#### Clone the ttt's repository\n\n```bash\ngit clone https://github.com/sbugallo/TicTacToe-MDP.git\ncd TicTacToe-MDP\n```\n\nThe following instructions assume that you are located in the root of `TicTacToe-MDP'`.\n\n#### Install requirements\n\n`conda` is the recommended way for installing all the requirements. You can install `conda` from\n[this link](https://docs.conda.io/en/latest/miniconda.html). The following instructions assume that you have\nconda installed on your system.\n\nPredefined conda environments for different use cases are provided under `./envs`. There are 2\navailable environments:\n\n- **dev.yml**: for simple execution usage.\n- **prod.yml**: to generate docs, test and develope new features.\n\nYou can create a new conda environment as follows:\n\n```bash\nconda env create -f ./envs/{dev.yml | prod.yml}\n```\n\nUsage examples:\n\n```bash\nconda env create -f ./envs/dev.yml\nconda env create -f ./envs/prod.yml\n```\n\nOnce our environment is installed, we have to activate it running the following:\n\n```bash\nsource activate ttt\n```\n\nAssuming that you have the environment activated, we install `ttt` module running:\n\n```bash\npython setup.py install\n```\n\nIf you plan to modify de source code, run some tests or generate documentation you should install the module in\ndevelopment mode:\n\n```bash\npython setup.py develop\n``` \n\nThis will also install some scripts that can be run directly in your terminal:\n\n- `ttt-human-vs-human`\n- `ttt-human-vs-cpu`\n- `ttt-cpu-vs-cpu`\n- `ttt-train`\n\nTo see how they work, please go to Usage section\n\n### Docker\n\nTo facilitate the usage, a docker image is available at\n[https://hub.docker.com/r/sbugallo/ttt](https://hub.docker.com/r/sbugallo/ttt).\n\n#### Prerequisites\n\n - `docker` (see [https://docs.docker.com/install/](https://docs.docker.com/install/))\n\n## Usage\n\n### Train agent\n\nTo train an agent simply run the following command:\n\n```bash\nttt-train \\\n --output_path \\\n --lr \\\n --exploration_iterations \\\n --exploitation_iterations \\\n --exploration_exploitation_iterations \n\n```\n\nExample:\n\n```bash\nttt-train \\\n --output_path /home/sbugallo/Documents/weights.json \\\n --lr 0.1 \\\n --exploration_iterations 2500 \\\n --exploitation_iterations 1500 \\\n --exploration_exploitation_iterations 1000\n```\n\n### Play game CPU vs CPU\n\nTo make 2 CPU agents play a game, run the following command:\n\n```bash\nttt-cpu-vs-cpu \\\n ---cpu_1_weights_path \\\n ---cpu_1_mode \\\n ---cpu_2_weights_path \\\n ---cpu_2_mode \\\n ---num_rounds \\\n ---display_board \n```\n\nExample:\n\n```bash\nttt-cpu-vs-cpu \\\n ---cpu_1_weights_path /home/sbugallo/Documents/weights_1.json \\\n ---cpu_1_mode best \\\n ---cpu_2_weights_path /home/sbugallo/Documents/weights_2.json \\\n ---cpu_2_mode best \\\n ---num_rounds 5 \\\n ---display_board True\n```\n\n### Play game Human vs CPU\n\nTo play a game against a CPU simply run the following command:\n\n```bash\nttt-human-vs-cpu \\\n --cpu_weights_path \\\n --cpu_mode \n```\n\nExample:\n\n```bash\nttt-human-vs-cpu \\\n --cpu_weights_path /home/sbugallo/Documents/weights.json \\\n --cpu_mode best\n```\n\n### Play game Human vs Human\n\n```bash\nttt-human-vs-human\n```\n\n### Docker\n\nFirst, we run the app with an interactive terminal as follows:\n\n```bash\ndocker run -ti sbugallo/ttt\n```\n\nThis will let you execute some commands in the running Docker container. This terminal will already have the conda\nenvironment activated and will be inside the project's root. In this interactive terminal you will be able to run the\nexamples showed above.\n\n## AI description\n\n### Introduction\n\nCPU agents are a simple implementation of a Markov Decision Process (MDP). In this algorithm, we have an agent that \ninteracts with it's environment. These interactions occur sequentially over time. At each time step, the agent will get \nsome representation of the environment’s. Then, the agent selects an action to take. The environment is then \ntransitioned into a new state and the agent is given a reward as a consequence of the previous action.\n\nComponents:\n\n- Agent\n- State\n- Action\n- Reward\n\nThe agent’s goal is to maximize the rewards that it receives from taking actions.\n\nIn an MDP, we have a set of states S, a set of actions A, and a set of rewards R.\n\nAt each time step t=0,1,2,..., the agent receives some representation of the environment’s state\nS_t ∈ S. Based on this state, the agent selects an action A_t ∈ A.\n\nTime is then incremented to the next time step t+1., and the environment is transitioned to a new state\nS_{t+1} ∈ S. At this time, the agent receives a numerical reward R_{t+1} ∈ R for the action\nA_t taken from state S_t.\n\n### Episodic tasks\n\nTicTacToe is an episodic task, being each episode a round. In episodic tasks, each episode ends in a terminal state at\ntime T, which is followed by resetting the environment to its initial state. The next episode then begins\nindependently from how the previous episode ended.\n\n### Exploration VS Exploitation\n\nThink of a row of slot machines. Each machine has its own probability of winning. As a player you want to\nmake as much money as possible. How do you figure out which of the machine has the best odds, while\nat the same time maximizing your profit? If you constantly played one of the machines (exploitation) you would never\nlearn anything about the odds of the other machines. If you always picked a machine at random (exploration) you would\nlearn a lot about the odds of each machine but probably wouldn’t make as much money as you could have always playing the\n“best” machine.\n\nAgents are trained using a mixed approach:\n\n- First, an agent is trained playing against itself with random moves.\n- Then, the agent is trained playing against itself with the best moves it learnt.\n- Finally, the agent is trained playing against itself but this time it choses both random and best moves.\n\n**Note**: Transition weights are updated at each round.\n","has_starred","hub_user","immutable_tags_settings",{"_75":11,"_76":77},"enabled","rules",[78],".*","is_automated","permissions",{"_82":21,"_83":11,"_84":11},"read","write","admin","user","storage_size",3782761511,"orgData","recentTags",[91,129],{"_92":44,"_93":94,"_95":96,"_97":98,"_99":100,"_50":118,"_119":94,"_120":33,"_121":56,"_57":122,"_123":124,"_125":109,"_126":118,"_127":66,"_128":21},"content_type","creator",7399423,"full_size",859354854,"id",66660590,"images",[101],{"_102":103,"_104":105,"_106":107,"_108":109,"_110":111,"_112":113,"_114":107,"_115":-5,"_116":96,"_63":66,"_117":-5},"architecture","amd64","digest","sha256:122228ea0b1ec56f5360f1b3b43394bcc84a47bc77d483a5c4bcc46a7f242eff","features","","last_pulled","2026-10-19T15:53:47.262498305Z","last_pushed","2019-08-28T22:42:17Z","os","linux","os_features","os_version","size","variant","2019-08-28T22:42:16.507729Z","last_updater","last_updater_username","media_type","latest","repository",7598332,"tag_last_pulled","tag_last_pushed","tag_status","v2",{"_92":44,"_93":94,"_95":130,"_97":131,"_99":132,"_50":137,"_119":94,"_120":33,"_121":56,"_57":138,"_123":124,"_125":135,"_126":137,"_127":66,"_128":21},858953244,66989136,[133],{"_102":103,"_104":134,"_106":107,"_108":135,"_110":136,"_112":113,"_114":107,"_115":-5,"_116":130,"_63":66,"_117":-5},"sha256:4e696b0ff3744b47f87a2e999035911f8da24a59a8214812332640c391762eed","2026-10-19T15:53:47.049843793Z","2019-08-27T23:08:01Z","2019-08-27T23:08:00.627668Z","2019.08","sbxKitMetadata","isPotentialSpam","isHubSbxKitsEnabled","showSbxRunInCloud","showSbxLogoOverride","sbxTrademarkText","displaySecurityMetadata","tagAttestations","tagsHaveAttestations",[],"selectedTag","attestationsUrl","routes/_layout.r.$namespace.$reponame.tags",{"_38":153,"_160":161,"_147":174,"_175":11},{"_40":154,"_42":155,"_45":46,"_47":48,"_49":11,"_50":51,"_52":53,"_54":156,"_57":35,"_32":33,"_58":59,"_60":44,"_61":62,"_63":64,"_65":66,"_67":-5,"_68":62,"_69":70,"_71":11,"_72":33,"_73":157,"_79":11,"_80":159,"_85":33,"_86":87},[],[44],[56],{"_75":11,"_76":158},[78],{"_82":21,"_83":11,"_84":11},"tags",{"_162":163,"_164":-5,"_165":-5,"_166":167},"count",2,"next","previous","results",[168,171],{"_92":44,"_93":94,"_95":96,"_97":98,"_99":169,"_50":118,"_119":94,"_120":33,"_121":56,"_57":122,"_123":124,"_125":109,"_126":118,"_127":66,"_128":21},[170],{"_102":103,"_104":105,"_106":107,"_108":109,"_110":111,"_112":113,"_114":107,"_115":-5,"_116":96,"_63":66,"_117":-5},{"_92":44,"_93":94,"_95":130,"_97":131,"_99":172,"_50":137,"_119":94,"_120":33,"_121":56,"_57":138,"_123":124,"_125":135,"_126":137,"_127":66,"_128":21},[173],{"_102":103,"_104":134,"_106":107,"_108":135,"_110":136,"_112":113,"_114":107,"_115":-5,"_116":130,"_63":66,"_117":-5},[],"canDeleteTags","actionData","errors"]

sbugallo/ttt

Tags for sbugallo/ttt