---
tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:
- text: '1girl, raiden, genshin impact, purple eyes, purple lightning magic, castle background, hair ornaments, bracelet, night, moon, smiling, blushing, hair ornament, purple kimono, sitting, tatami'
  output:
    url: >-
      images/1FA71F4AA70D9BD53DB82F20A64D7206A6FEA666F90212ADA96D96381DF7BFFF.jfif
- text: 'blonde, blue eyes, blue bikini, swimsuit, tall, sunhat, beach, day, sun, ocean, long hair, 1girl, blue hat ribbon, smiling, blushing, bracelet, hair ornament'
  output:
    url: >-
      images/AA3E569223238048143023C89943FA26BC5AD9CCA231A81182DAA5763DF85B0B.jfif
- text: 'school girl uniform, coffee, drink, cafe, shopping mall, day, uniform, white hair, silver eyes, shopping bag, hair ornament, black chocker'
  output:
    url: >-
      images/07340660891E2D3AFAE4FA09E0C24E2C6256703984619FB3BC2695893651C63C.jfif
- text: '1girl, hu tao, genshin impact, forest background, lush greens, sunhat, hair ornaments, bracelet, night, moon, smiling, blushing, hair ornament,'
  output:
    url: >-
      images/ED680D12318D5FC8C2C660A0F01689658D73E3CF6119FCAE4E4921FA456F9E69.jfif
- text: '-'
  output:
    url: >-
      images/0053-1girl, hoshimachi suisei, hololive, alte-animagineanimagine-xl-31-208967273-1.png
- text: '1girl, columbina, genshin impact, night, outdoors, moon light, purple skies, stars, galaxy windows, red moon, closed eyes, looking at viewer, jewelry, accessories, earrings, white dress, white roses, gradient, facing front'
  output:
    url: >-
      images/1651-1girl, columbina, genshin impact, night,-animagineanimagine-xl-31-1134818894.png
      
base_model: cagliostrolab/animagine-xl-3.1
instance_prompt: null
license: apache-2.0
---
# TorinoAqua Anime Style v2

<Gallery />

## Model description 

SDXL based Torino Aqua Anime Drawing style.

Use this lora with Animagine v3.1 base model.

Adjust CFG scale (2-4) for best result. Any higher and the image will look oversaturated and lose details.

Adjust lora strength (0.6-1.0) if the images look deep fried &#x2F; incoherent.

The higher the step count while generating the more detailed and coherent the stylized images get. 

Trained on 40 Epoch.

All credit goes to TorinoAqua for their drawing style.


## Download model

Weights for this model are available in Safetensors format.

[Download](/moemoe101/TorinoAquaAnimeStyleV2/tree/main) them in the Files & versions tab.