Shinkawa [ Flux LoRA ]

Lora is trained on 512p artwork images by Yoji Shinkawa. The result is Lora for retro anime art.

Recommended to use Strenght model 0.7-0.8

This is my workflow with FLUX gguf DEV Q8 or Q4. Working with Lora, The Q4 is as fast as the NF4.