Hello First time uploading a LoRA (ive made a few before this is my first time uploading) So a bit of backstory behind this model, I was watching Filian and Snuffy's first VR Chat colab vod and thought to my self "Damn love snuffys model let me check if there is art of it" and i didnt find ANY art based on this model... So i desided to take it into my own hands and create this LoRA for you fine folks!
NOTICE
This LoRA ain't perfect and ONLY has training data from that 1 stream (with 37 hand picked screenshots) so its quite limited what the LoRA can do... I haven't test NSFW but i'd guess it would work...
Due to only being trained on 1 stream and no fan-art the LoRA HEAVILY enforces a VR Chat 3d model style if you like that keep the strength around 0.7 - 0.9 if you'd like to control the style a bit more id say go for 0.5 - 0.6
Refer to the trigger words to fine tune the outfit!
With that being said I hope you enjoy this model and get good use out off it and I hope to make more in the future!
Description
v1.0 Trained on 37 VRChat images
FAQ
Comments (2)
if want change style away from vrc3d, make generation with diffusion and pick ones from away from style, use to train again with original images too
Im a little new to stable diffusion so I don't know what to do about this, everytime I generate something its blurry (like duh its copying off a vrchat screenshot) so what do i do? Do I need another vae? I already lowered the LORA down used different checkpoints so idk