V2 is Officially Done!

I will upload the safetensor around 3 days or so, this version is a bit stable compared to V1;

V2 will feature:

• Larger Dataset

• Adjusted Alpha and Dim to cover more details

• Eyes and hands are coherent

zippo (3756c) style :p

notice:

this lora wasn't made to cause harm to the artist, if the artist, in this case wants this lora to be taken down, I will gladly take it down alongside PixAI.

• Lora was trained on 99 images, kinda handpicked.

• Weight can be used to 0.1 to 1

• Best samplers: Euler or Euler A

• Steps: 25 or above

made this lora because no lora of zippo?!?! that's rare