Samples

(new, newest, best quality, masterpiece:1.2), 1girl, solo, (upper body:1.2), bikini, short hair, black hair, animal ears, simple background, white background

negative: nsfw, (worst quality, low quality, normal quality:1.2), (bad anatomy, bad hands, missing fingers, extra digit, fewer digits:1.1),

(new, newest, best quality, masterpiece:1.2), 1girl, solo, close-up, bikini, short hair, ocean, wave, water

negative: nsfw, (worst quality, low quality, normal quality:1.2), (bad anatomy, bad hands, missing fingers, extra digit, fewer digits:1.1),

(new, newest, best quality, masterpiece:1.2), 1girl, (solo:1.2), short hair, messy hair, black hair, white hoodie, outdoor, cityscape

negative: nsfw, (worst quality, low quality, normal quality:1.2), embedding:negativeXL_A.safetensors, (bad anatomy, bad hands, missing fingers, extra digit, fewer digits:1.1),

(new, newest, best quality, masterpiece:1.2), 1girl, sitting on bed, sexy pose, looking at viewer,
long hair, beige hair, black eyes, blush

negative: nsfw, (worst quality, low quality, normal quality:1.2), embedding:negativeXL_A.safetensors, (bad anatomy, bad hands, missing fingers, extra digit, fewer digits:1.1),

A simple workflow file for use with comfyui is here.

workflow-swim-in-latent.json

metadata.json

{
    "modelspec.sai_model_spec": "1.0.0.alpha",
    "modelspec.architecture": "stable-diffusion-xl-v1-base",
    "modelspec.implementation": "sgm",
    "modelspec.title": "SwimInLatent",
    "modelspec.author": "ddPn08",
    "modelspec.description": "StableDiffusionXL model fine-tuned for anime.",
    "modelspec.date": "2023-07-29",
    "modelspec.license": "CreativeML Open RAIL++-M"
}