My attempt at replicating the art style of Zenless Zone Zero!
some pretty good results from this
recommended weight is 1.0, but good prompting is required to deviate from the game funky hair styles or clothing
An improved V2, with a new "character screen" tag to replicate the character screen.
this replicates the 3d aspect a bit more than v1 with a larger set of training data.
works well with character LoRa's
buy me a coffee! help support what i love to do
Support me!
works well in most settings
Recommended parameters
Base model: Pony Diffusion XL v6
CFG Scale: 7
Sampler: DPM++ 2M SDE
Scheduler: exponential
Steps: 30
LoRA Strength: 1.0
It works very well at most parameters and strengths