minpeter/FAS-v2

Fantasy Anime Style

A Flux LoRA trained on a local computer with ostris/ai-toolkit.

Prompt

anime style, Digital illustration with a glamorous and captivating atmosphere. Attractive figure with stylish short black bob and bangs. Eye-catching composition showing a beautiful side profile with head turned to the right from the back. Wearing large, sparkling, ornate gold earrings and an elegant lace choker that accentuates the neckline. In the background is a huge, brightly shining mysterious orange full moon dominating the night sky. Dramatic yet warm lighting gently envelops the figure and adds a dreamy feel.

Prompt

anime style, A small-sized fruit tart with a perfectly golden, crispy crust on a wooden plate, lightly dusted with powdered sugar. It's topped with intensely vibrant, jewel-like fruits (berries, kiwi, orange slices) bursting with freshness. The fruits and tart are coated in a dazzling, wet-look mirror glaze that reflects bright natural light intensely. A coffee cup and fork sit nearby on the warm wood table. Focus on the extreme glossiness, juicy textures, and appetizing appearance.

Prompt

anime style, In an anime style digital illustration, a girl with a flawless Quake engine in a surreal background wears mysterious black & mustard yellow Adobe vintage clothing, has a slim build, exudes a unique atmosphere, laughs like a siege engineer alongside a Sphinx, features short messy hair, captivating prehistoric coral/peach/sand glowing turquoise eyes, and a slimy earring, set against a peach garden full of plants within the City of Avalon during the golden hour, depicted in a very wide view composition, with a sketch, fantasy, synthwave art style emphasizing reflected light, a 35mm film feel, complementary colors, vignette, and maroon explosion effects, possessing high detail such as skin pores and natural skin tones, high resolution, photolab quality, and an abstract feel.

Fixed certain character style biases that existed in FAS v1. It still has some difficulties generating images that are not people/food, but I hope this will be fixed in the next version so that it can be used as a "style LoRA".

Trigger words

You should use anime style to trigger the image generation.

Creating images using a diffusers

import torch
from diffusers import FluxPipeline

pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16).to("cuda")

pipe.unload_lora_weights()
pipe.load_lora_weights("minpeter/FAS-v2", weight_name="adapter.safetensors", adapter_name="fas-v2")
pipe.set_adapters(["fas-v2"], adapter_weights=[1])

prompt = "anime style, Digital illustration with a glamorous and captivating atmosphere. Attractive figure with stylish short black bob and bangs. Eye-catching composition showing a beautiful side profile with head turned to the right from the back. Wearing large, sparkling, ornate gold earrings and an elegant lace choker that accentuates the neckline. In the background is a huge, brightly shining mysterious orange full moon dominating the night sky. Dramatic yet warm lighting gently envelops the figure and adds a dreamy feel."

image = pipe(
    prompt=prompt,
    guidance_scale=3.5,
    height=1024,
    width=1024,
    num_inference_steps=15,
    # generator=torch.Generator(device="cpu").manual_seed(208070), # for reproducibility
).images[0]

image.save("s4.png")