Python Package isn't downloaded

[netsonicyxf]

Win10 x64, Rhino 7.30, Tunny 0.80
After double click Tunny component, it shows this window and I wait for ever

There is only one new file called "Settings.json" in the
C:\Users\user name\AppData\Roaming\Grasshopper\Libraries\Tunny
The python package is never be download.

Could you please provide offline python package?

[hrntsm]

Hi @netsonicyxf

I've created a modified version, could you try this one? This one contains the Python Package inside.
https://github.com/hrntsm/Tunny/releases/tag/v0.8.1

[netsonicyxf]

Thanks.
When I run built-in "Karamba3d_column_position_opt.gh", it shows

[hrntsm]

I checked the file you gave me and it was created properly as a sqlite file.
Possible problems are as follows.
When Human-in-the-loop is run, it automatically creates a folder to store the results, but if the directory does not have permissions to create new files, it may not work.
The Human-in-the-loop files are created in the same folder as the db file.

[netsonicyxf]

The sample.gh and Vase_Shape_opt.gh are in the same folder.
The sample.gh can be run without any problem, which means the directory do have permissions to create new files.

[hrntsm]

It doesn't matter where the gh file is, it's where the db file is. db files are created in a place where it doesn't matter to create new files as well?

[netsonicyxf]

I mean the db file can be created for sample.gh, which means the directory do have permissions to create new files.
Since Vase_Shape_opt.gh and sample.gh are in the same folder, Human-in-the-loop can automatically creates a folder to store the results in it.

[hrntsm]

Thanks for the response.
I do not know how to fix the current situation. I am currently working on a support for the latest version of optuna. I will let you know when it is done, as I plan to improve the ease of use of human-in-the-loop with this support.

[netsonicyxf]

I have some question about sample.gh. Everytime I ran it, I got different answer
1st time

2nd time

Are they suppoed to be the same?

[hrntsm]

All optimizations handled by Tunny involve randomness in intermediate steps of the algorithm. Therefore, unless you directly specify the seed value of the random, the result of the optimization will change each time.

Also, sample.gh performs multi-objective optimization. In the case of multi-objective optimization, multiple solutions (Pareto solution) are obtained instead of a single optimal solution, so the optimization results are not reflected in Grasshopper's slider. The results can be reflected on the Grasshopper side by specifying the necessary values from the Output tab.

Check the following for seed value settings.

• Reproductibility - Is there a way to specify the seed number as in Optuna? #169

For information on how to check and output results, see below

• https://tunny-docs.deno.dev/docs/optimize-window/visualize-tab
• https://tunny-docs.deno.dev/docs/optimize-window/output-tab