//=time() ?>
"Waifu-Diffusion v1-2: A SD 1.4 model finetuned on 56k Danbooru images for 5 epochs"
#StableDiffusion #AIArt #AIArtwork #DreamStudio @StableDiffusion
My custom finetuned #ai model "Liminal Diffusion" beta version should be released tomorrow.
Created with my own hand curated and custom dataset this model has been trained to create eerie, strange, and surreal liminal spaces.
Very happy with how it has turned out! Free4All.
Very soon I will be releasing the beta version of "Liminal Diffusion" which is my own custom finetuned Text2Image model.
Very excited to see what people can make with it! If you want special access before I release comment below or shoot me a dm :)
#ai #aiart #discodiffusion
"Lost architecture in the clouds induces memories." created with my own finetuned diffusion model.
#ai #aiart #aiartist #architecture #cloud #cloudcity #liminal #liminalspace
GPT-2 finetuned prompt: "Picnic in the beautiful iridescent temple of starlight, matte painting by Ross Tran and Ivan Aivazovsky"
Made with @midjourney.
Prompt generated with #PromptParrot.
GPT-2 finetuned prompt: "Flowers on the moon shine brightly into an undulating pool, inspiring vivid colours and lively lighting, the ethereal poppy abandoned hydrangea palace, matte painting by Felix Kelly, lighting study"
Made with @midjourney and #PromptParrot.
@SeanPaulSan1 I finetuned gpt3 on tens of thousands of prompts. Few examples:
`Cityscape at night under the neon aurora of Tokyo lights, by Shaun tan.`
`Endless labrynth of glowing portals and sickle moon runes. ArtStation Mega Caves by Nikola njorgiewski`
Meet the #NovelAI AI Model family!
Our AI storytellers:
Euterpe: Fairseq 13b (Our latest addition, still experimental)
Sigurd: GPT-J 6b
Calliope: GPT-Neo 2.7b
Experimental AI models:
Genji: GPT-J 6b finetuned on a Japanese storytelling dataset
Snek: GPT-J 6b for python coding
#LookingGlassAI finetuned on portraits found in old public domain magazine. n°3 and 11 are welcome in my comic.
#AIart #aiartcommunity #aiartists
Prompt: "mirror of love, by greg rutkowski and james jean" (CLIP guided diffusion, 512x512 openai finetuned on openimages)
XD This is really just a documentation of my evolution in digital skill, more so than character design. Left was drawn when I just started digital in 2018, right this year. I've just finetuned her design really, the only big change was I only kept the skull in her final design. https://t.co/OlYAGjJmCA
✨💫If you see this tweet, show us your quarantine art progress!💫✨
Feb 2020 to Feb 2021..
I think aside from doing more bg work, I finetuned my style a lot! I'm always tryin new stuff but I think I gained a bigger understanding of me & all the stuff I like.. https://t.co/5lUNQKBnTw
This Anime Does Not Exist
https://t.co/fjaPZm2DbC
The network with pre-trained weights is available to download! If you make anything with it (samples, videos, finetuned models, research), please share, I'd love to check it out!