Linaqruf/pastel-anime-xl-lora

Pastel Anime LoRA for SDXL

TRAINED WITH ANIMAGINE XL

---

---

Overview

Pastel Anime LoRA for SDXL is a high-resolution, Low-Rank Adaptation model for Stable Diffusion XL. The model has been fine-tuned using a learning rate of 1e-5 over 1300 global steps with a batch size of 24 on a curated dataset of superior-quality anime-style images. This model is derived from Animagine XL.

Like other anime-style Stable Diffusion models, it also supports Danbooru tags to generate images.

e.g. face focus, cute, masterpiece, best quality, 1girl, green hair, sweater, looking at viewer, upper body, beanie, outdoors, night, turtleneck

---

Model Details

• Developed by: Linaqruf
• Model type: Low-rank adaptation of diffusion-based text-to-image generative model
• Model Description: This is a small model that should be used with big model and can be used to generate and modify high quality anime-themed images based on text prompts.
• License: CreativeML Open RAIL++-M License
• Finetuned from model: Animagine XL

---

🧨 Diffusers

Make sure to upgrade diffusers to >= 0.18.2:

pip install diffusers --upgrade

In addition make sure to install transformers, safetensors, accelerate as well as the invisible watermark:

pip install invisible_watermark transformers accelerate safetensors

Running the pipeline (if you don't swap the scheduler it will run with the default EulerDiscreteScheduler in this example we are swapping it to EulerAncestralDiscreteScheduler:

import torch
from torch import autocast
from diffusers import StableDiffusionXLPipeline, EulerAncestralDiscreteScheduler

base_model = "Linaqruf/animagine-xl"
lora_model_id = "Linaqruf/pastel-anime-xl-lora"
lora_filename = "pastel-anime-xl.safetensors"

pipe = StableDiffusionXLPipeline.from_pretrained(
    model, 
    torch_dtype=torch.float16, 
    use_safetensors=True, 
    variant="fp16"
    )

pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config)
pipe.to('cuda')

pipe.load_lora_weights(lora_model_id, weight_name=lora_filename)

prompt = "face focus, cute, masterpiece, best quality, 1girl, green hair, sweater, looking at viewer, upper body, beanie, outdoors, night, turtleneck"
negative_prompt = "lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry"

image = pipe(
    prompt, 
    negative_prompt=negative_prompt, 
    width=1024,
    height=1024,
    guidance_scale=12,
    target_size=(1024,1024),
    original_size=(4096,4096),
    num_inference_steps=50
    ).images[0]

image.save("anime_girl.png")

---

Limitation

This model inherit Stable Diffusion XL 1.0 limitation