Really loving the work you put into this app.

On the profile header, could we get an option to have the more expanded appearance? I like seeing the banner art when people add that to their profile, but the new gradient overlay and position of the avatar and username obscure a lot of it now.

I tried to leave a tip, but the testflight version indicates the transaction would not be real. Is there a way to do that and still use testflight?

Just got the new 0.4.2 testflight version. That is a very long list of updates and fixes! Thanks for continuing to make improvements.

In the Expanse books, there’s a planet called Auberon that has an 8 hour rotation, so 4 hours of light and 4 hours of dark. They decided that “1 day” would be light-dark-light and “1 night” is dark-light-dark. It’s really interesting how they describe the way society adapts to the cycle of having a midnight sun and both a midmorning and evening sunset.

I thought i had a good system where each outpost was only exporting 1 solid, 1 liquid, and 1 gas. This allowed me to isolate and sort at the receiving outpost.

The problem occurs when each outposts import & export containers get full. At which point materials flow from the export station, go to the import location where they can’t be unloaded, THEN THEY COME BACK to the original outpost where they get offloaded. You wind up with all the same materials filling both the import and export containers. Now the entire material flow is completely borked, nothing is getting imported, and all you have access to is the stuff thats locally produced.

If you look at the ground map when you land on a “coast” biome, you can see where all the elevation dots go flat. That will help direct you to the ocean if you can’t see it over the hills/fog.

I have a bunch of Akila City security guards that wander around in their underwear in my game. I just assumed it was casual Friday.

Oof. Clearly I hadn’t tried it.

Even with that trick, going from 1000 down to 1 is still painful. Probably faster to give them all 1000, then switch and take 999 back.

Stable Diffusion always (? I think anyway) puts prompt metadata into the output image. The problem is that it’s easy to strip it out converting to jpg or other formats. Even just uploading to Lemmy will strip the metadata, That is why I use catbox.moe which preserves all of that info.

Post it again, and screw the downvotes. I thought that image was pretty good and am kicking myself now for not saying so before you took it down. Please don’t lurk, this community needs more diversity.

Normally, you can use a tool like pngchunk.com to read the metadata, but I just realized that I did an SD Upscale on this one and it didn’t preserve the prompt. Sorry about that, I’ll put the entire metadata dump below in a spoiler. I’m not sure it’ll be that helpful though, this image uses a custom LoRA I’m working on that I haven’t released yet, and is complicated by the fact that I started with one model (ToonYou beta 6) for the first 40% of the generation, then switch to a realistic checkpoint merge for the last 60% as a refiner.

Thank you! I use a bunch of different custom merges. See this very large xy comparison grid i posted earlier: https://files.catbox.moe/1k6mmr.jpg

I’d recommend Absolute Reality or LazyMix for an off-the-shelf model.

There’s not much out there on training LoRAs that aren’t anime characters, and that just isn’t my thing. I don’t know a chibi from a booru, and most of those tutorials sound like gibberish to me. So I’m kind of just pushing buttons and seeing what happens over lots of iterations.

For this, I settled on the class of place. I tried location but it gave me strange results, like lots of pictures of maps, and GPS type screens. I didn’t use any regularization images. Like you mentioned, i couldn’t think of what to use. I think the regularization would be more useful in face training anyway.

I read that a batch size of one gave more detailed results, so I set it there and never changed it. I also didn’t use any repeats since I had 161 images.

I did carefully tag each photo with a caption .txt file using Utilities > BLIP Captioning in Kohya_ss. That improved results over the versions I made with no tags. Results improved again dramatically when I went back and manually cleaned up the captions to be more consistent. For instance, consolidating building, structure, barn, church, house all to just cabin.

Epochs was 150, which gave me 24,150 steps. Is that high or low? I have no idea. They say 2000 steps or so for a face, and a full location is way more complex than a single face… It seems to work, but it took me 8 different versions to get a model I was happy with.

Let me know what ends up working for you. I’d love to have more discussions about this stuff. As a reward for reading this far, here’s a sneak peek at my next lora based on RDR2’s Guarma island. https://files.catbox.moe/w1jdya.png. Still a work in progress.

🫡

Nice work!

AckbarItsATrap.gif

Thank you, it still feels like magic to me, so it’s fun to see how SD reacts to different inputs.

I get a lot of half shirts and sweaters with very unconventional cut outs for sure. SD has trouble “bunching” fabric for the lift/reveal shots, so it likes to just cut things off.

I do like that black & blue knit+leather number though. Its unusual, but really cute. I did a double take when that came out of the diffusion soup.

Thank you!

Lovely!

Delta and Lima have been solid with a bunch of prompts. It’s interesting to me that you picked those two based on the one xy spread, they’re at the top for me. Thanks for taking a look!