Created /c/imageai
I used a LORA and textual inversion, here’s all the info for generating the image (it’s kind of weird, I put random words in sometimes just to see what happens):
parameters
((knitted knit yarn stitched scarecrow)), (warzone combat zone battlefield tracers flying in the night sky explosives in the air), (deer antlers warzone combat fire handstand pentagram demon demonic nightmare dark), (fog foggy mist mud muddy water explosive dark night rain warzone background), (high definition, HD, extreme detail, 4k, 8k, highest crisp resolution, hyper realistic) <lora:add_detail:1>
Negative prompt: badhandv4 verybadimagenegative_v1.3
Steps: 60, Sampler: DPM++ 2M SDE Karras, CFG scale: 9, Seed: 2172657245, Size: 768x768, Model hash: ed989d673d, Model: dreamshaper_7, Lora hashes: "add_detail: 7c6bad76eb54", Version: v1.3.2
Here’s another from the same prompt and stuff lol
There is no reason to be upset. 😂
LOLOLOL thank you again! I feel like I’ve unlocked a new power and have levelled up LOL
LOL the third one is great. I’m going to need to learn how to use adetailer. Thank you for this!
Interesting I’ll try this, thanks.
If you’re asking whether I generated it, yes, I generated it with Dreamshaper v7.
parameters
portait of a young red head freckled woman (confused concerned upset questioning (funny fun amused (cringe (raised eyebrow) shy blush ashamed bruh face) emotional smile expression) emotion), (wearing a pearl necklace and hoop earings), (short hair bun princess), (vibrant neon purple eyes), (8k, 4k, HD, high definition, ultra quality, extreme detail, hyper realistic, UE5) lora:add_detail:1.2
Negative prompt: verybadimagenegative_v1.3, badhandv4
Steps: 60, Sampler: Euler a, CFG scale: 7, Seed: 78986836, Size: 768x768, Model hash: ed989d673d, Model: dreamshaper_7, Lora hashes: “add_detail: 7c6bad76eb54”, Version: v1.3.2
Upscaled version:
Renaissance fair
That aspect ratio is weird, about 16:6.
The giant army robot is amazing looking. The filter, aspect ratio, lighting, and style make me feel like it could have been real. The dancing automaton oil painting is beautiful.
Awesome!
Here’s my submission, I just typed “dieselpunk”, “dieselpunk city”, and “dieselpunk plane” into DALL-E2 image creator.
Also testing to see if imgur albums will fully display in a single embed (imgur album doesn’t show up embed in comment, have to do 1 image at a time)
People are coping so hard with starfield. There have been several major red flags already and I have really low expectations. I’m going to wait a week for reviews and real gameplay, and then make a decision.
My vote is for “Dieselpunk” https://en.wikipedia.org/wiki/Dieselpunk
Actual quotes from Shuji Utsumi: “The action in play-to-earn games is boring,” Utsumi told Bloomberg News. “What’s the point if games are no fun?” “We’re looking into whether this technology is really going to take off in this industry, after all,” he added.
To add from the article: “Sega does plan to let external partners use its lesser-known Three Kingdoms and Virtua Fighter characters for non-fungible tokens, a sort of certificate that confirms ownership of a digital asset.”
“Sega will still offer its lesser franchises to several blockchain games to be announced later this year, and will continue to invest in related projects — to the tune of hundreds of millions of yen in each case, Utsumi said. The technology remains useful, for instance by making it possible to move characters and items between different games, he said. Longer-term, Sega remains open to getting more involved as the technology matures.”
I guess no one reads articles and just wants to bash new technology, makes sense. Next time use a non-editorialized title and use actual quotes.
No limits on post-processing because in that regard it’s very difficult to draw any type of lines and I consider that a part of the process.
Looks like what I imagine some dinosaurs looked like.
Bean meme
Here’s the lemmy.world banner, it’s a strange aspect ratio, but it fits. It’s about 16:6 aspect ratio 800x300.