am I retarded? I'm trying to follow the instructions to generate ai art, but I get really poor results
what I did:
1. download and install easy diffusion
https://easydiffusion.github.io/docs/installation/2. download a Lora (80MB file)
https://civitai.com/models/452014?modelVersionId=503285 and put into easydiffusion/models/lora
the Lora page says "This asset is designed to work best with the Pony Diffusion XL model, it will work with other SDXL models but may not look as intended."
so I went ahead and did that, even though it looks suspicious that it would depends on a "pony" model for anime, but the preview images look like it supports both so whatever I guess?
3. download the model (a 6GB file) and its vae (a 300MB file)
https://civitai.com/models/257749/pony-diffusion-v6-xl and put them into easydiffusion/models/stable-diffusion and easydiffusion/models/vae respectively
4. start EasyDiffusion and open the WebUI, click refresh, all the 3 models I downloaded (lora+sd+vae) show up in the respective dropdown boxes, I select each one accordingly
5. put in the first prompt that the model page recommended (shylily1st, heart ahoge, facial mark... etc) as a pure copy-paste to get started. let it generate
and then... it generates low quality garbage. it does look like the lora work, I guess? since it seems to generate drawings of shylily (blue tail, blue skirt, etc), which is the model I downloaded, but the art style doesn't look anything like the preview images, it looks like some 8 year old fanart
am I too stupid to understand this? which step did I mess up?
I tried the whole thing with an old stable-diffusion-webui install I had and got pretty much the same results with this
