Roadmap

[swyxio]

heya, so i know this project isn't open contributions but i do find the need for a GUI and i like the way you do stuff so am just penning my developing thoughts as i start to explore SD and think about what i want in a GUI

everything here is subject to the recognition that you can/should ignore any of it and i wont take any offense

1. more cli options. i would like a CFG parameter setting (which I think means the --scale param?). this one is easy, you'll probably do it eventually. being able to tweak --n_samples would also be nice. and seed (max 4294967295)
2. prevent footguns
   • escape double quotes
   • token length checker - max limit is 75-77 tokens - prevent expensive errors
   • errorboundary? i wrote a bug and the ui went unresponsive until i opened up devtools.
3. track "time elapsed" so we can see how far along we are in the generation...
   • alongside of maybe an "estimated time" thing with heuristics based on past runs and/or "complexity cost" based on the params
4. seeing the results of the image in the app itself. not suuper sure how exactly to do it but its probably not hard?
5. (hardest) being able to choose the image and "refine" it - this to me was the killer UX of midjourney - this probably means storing the metadata of each image alongside the image, including a fixed seed. lets you explore a bunch of stuff in a grid and then refine it

something like that? thoughts? obv if we're aligned then i'd be happy to contribute most of these since i want them

[breadthe]

Hey Shawn, I'm somewhat familiar with your work so I really appreciate you wanting to help, and I know I can trust you with it.

For starters, I moved this to a discussion.

All these suggestions are very sensible. Some of them are on my own roadmap, others are common-sense.

1. already on the wishlist, last item. I can do this myself, unless you're quicker.
2. this one I'll leave to you... I'm not doing a lot of error handling since it's not a mission critical app but I agree these things should he handled better.
3. I was mulling the idea of showing a real-time counter while it's generating. I'll take a shot at it myself after I finish what I'm currently working on.
4. also on the wishlist and the thing I'm currently hacking on. It's quite tricky and my solution is a bit wild, but I'm hoping it'll work. I had a working prototype last night but I found an issue with it so I'm improving it today.
5. this one is a big MAYBE, but if you can pull it off, feel free.

Now, be aware that I'm still refactoring things (extracting components, moving logic) as I see fit. At some point I want to add more structure to the app (multiple pages/sections, etc) so expect the layout and components to change.

My only request is that, if you make a PR and I've changed things in the meantime enough that there are serious merge conflicts, please pull my latest code and re-integrate your changes.

Thanks for helping!

[swyxio]

nice! ya i mean i am not sure how intensely ill be using this myself because my own local copy of stablediffusion started crashing repeatedly (i think it is out of memory but i keep restarting and it doesnt work).

[breadthe]

Are you using --n_samples > 1 by any chance? My SD python command crashes when I try anything > 1.

[swyxio]

yes! that was it! weird, because i initially tried n_samples up to 4 and it was fine.