>>50706717I am playing around with the workflow. Seems I get much much better results when I just do img2img with the nina lora over an initial text image.
parameters
, red hair, hair down, short sleeves, shoulder free,
<lora:chibiStyle_v1:1.2> , chibi,1girl
field, tree, blue sky
Negative prompt: (worst quality:1.4), (low quality:1.4),
badhandv4, negative_hand-neg
Steps: 30, Sampler: DPM++ 2M Karras, CFG scale: 10, Seed: 1178455676, Size: 512x768, Model hash: a074b8864e, Model: Counterfeit-V2.5_pruned, Lora hashes: chibiStyle_v1: 009f1dd07b07, Version: v1.3.0