🐟Tunny🐟 is Grasshopper's optimization component using Optuna, an open source hyperparameter auto-optimization framework.
The following is taken from the official website
Optuna™, an open-source automatic hyperparameter optimization framework, automates the trial-and-error process of optimizing the hyperparameters. It automatically finds optimal hyperparameter values based on an optimization target. Optuna is framework agnostic and can be used with most Python frameworks, including Chainer, Scikit-learn, Pytorch, etc.
Optuna is used in PFN projects with good results. One example is the second place award in the Google AI Open Images 2018 – Object Detection Track competition.
Optuna official site
First, Tunny runs on Windows only.
- Download Tunny from food4rhino or release page
- Right-click the file > Properties > make sure there is no "blocked" text
- In Grasshopper, choose File > Special Folders > Components folder. Move Tunny folder you downloaded there.
- Restart Rhino and Grasshopper
- In Grasshopper, Place the Tunny component and double-click the icon to start downloading the necessary libraries.
- Enjoy!
This software is being updated with your support. If you like this software, please donation.
Tunny is licensed under the MIT license.
Copyright© 2022, hrntsm
Tunny use TT-DesignExplorer & Python runtime & some python packages. These depend on their own licenses. Please see PYTHON_PACKAGE_LICENSE for more license information.
Video.mp4
Please see Tunny documentation page.
- HP : https://hiron.dev/
- Mail : support(at)hrntsm.com
- change (at) to @
- Postings to the discussion page are also welcome.