Skip to content

Genetic algorithm control input bot for Bizhawk for the creation of tools-assisted speedruns.

License

Notifications You must be signed in to change notification settings

tommai78101/Bizhawk-GeneticAlgorithmBot

Repository files navigation

Bizhawk - Genetic Algorithm Bot

Control inputs generation bot using genetic algorithm for BizHawk for the creation and assistance of tools-assisted speedruns.

Also includes an experimental neuro-evolution augmented topology (NEAT) for control inputs generation as an alternative generation feature. (Pre-release v1.0.4-dev)

Requirements

  • Runs in BizHawk Emulator v2.9 and above.

Development Requirements

Source Code Installation

  1. Make sure you unzip a working copy of BizHawk (the full folder) and copy the entire folder to this project's root directory. File structure should look something like this:
GeneticAlgorithmBot
 ├ BizHawk/         <-----    This is where you put the full BizHawk release folder at.
 ├ src/
 ├ run_build.cmd    <-----    For development. This is the file you execute after putting the BizHawk folder.
 ├ build_only.cmd   <-----    For release. This is the file you execute after putting the BizHawk folder.
 ├ LICENSE
 ├ README.md
 └ .gitignore

2a. Run the run_build.cmd batch file.
2b. Or for distribution only, run the build_only.cmd batch file.

Click the video below to see the installation process using Windows Sandbox:

Bizhawk Genetic Algorithm Bot Source Installation

Release Installation

  1. Make sure you back up your BizHawk's config.ini file before you attempt to load the bot (See TASEmulators/BizHawk#3337 for more info).
  2. Download and unzip the contents to BizHawk/ExternalTools. If this folder directory doesn't exist, you will need to make a new folder.
  3. Run EmuHawk.
  4. Make sure you have U+D/L+R controller buttons set to Allow.
  5. Open TAStudio. (Required)
  6. In the toolbar menu: Tools -> External Tools -> Genetic Algorithm Bot.

Credits

About

Genetic algorithm control input bot for Bizhawk for the creation of tools-assisted speedruns.

Resources

License

Stars

Watchers

Forks

Packages

No packages published