Manager to make it more "Friendly"

[mgatelabs]

Ok, so over in my fork/branch https://github.com/mgatelabs/VQGAN-CLIP/tree/dev I have been building buddy.py, which is a manager for controlling your renders. The only main issue I see when playing with this verses the online tools, is the online tools offer to further "enhance" images to see it with more detail, so that is what I sort of did, it keeps track of the seed used so it van rerun the image through the processor one more time.

How to use it

First off you need to be in your right, I don't know the name, I just type "conda activate vqgan", if you followed the instructions.

Then python buddy.py

It will startup and try to load in the buddy.json file, if that doesn't exist, it will default everything. If you use the Config command you can Set the values.

So with the buddy python script, you can Add a package, give it a ID and the TEXT. Then you can Select the package, then Add files to it (this is how you add things to render). Then you can Queue up the files for enhancement, then finally return to the main menu and Execute the queue. Then after it's done, you can generate a preview image of all the image's it's produced. Or go back again, select the package and queue up files for further enhancement.

Missing

I haven't built in spawning yet, which seems like copy the original image, but assign it a new seed number.

Also

Also I converted zoom.sh to zoom.py and there is an extra tool makesome.py, which is a prototype of buddy.py, that just makes a series of images and stitches them together into a single image. And you will need to install magick, for the image editing.

Also python isn't my normal language, so I'm probably making a ton of mistakes with how I formatted everything.

[mgatelabs]

Really thinking to make a discord integration, so you could have discord issues commands to your local machine to make images.