Introduction
The Illustriou LoRA model, also known as Bidens Pilosa / White Silk Cake, is a specialized AI model designed to generate captivating anime-style images. It focuses on creating sweet, healing, and dreamy aesthetics, making it perfect for artists and enthusiasts aiming to produce high-quality, emotionally resonant anime artwork.
Key Features:
- Detailed Line Art: Produces smooth and fine lines, with exceptional detail in hair strands, showcasing clear, layered, and flowing textures.
- Vibrant Color Palette: Utilizes a soft, rich, and bright array of fresh tones. Features natural color transitions and gradients to build a dreamy, therapeutic atmosphere.
- Dynamic Lighting: Accurately captures color variations under different lighting conditions, enhancing the three-dimensional quality of the generated images.
- Overall Aesthetic: Specializes in a sweet, healing, and dreamy two-dimensional (anime) style, ensuring both character designs and environmental elements convey a sense of warmth and charm.
How to Use * Trigger Word: No specific trigger word is required to activate this LoRA model.
- Recommended Settings: * Steps: 20 * CFG Scale: 7 * LoRA Weight: 0.8
- Tips for Optimal Results: * Sampler: Euler is recommended. * Steps (General): For broader generation, consider using 25-30 steps. * LoRA Weight (Standalone): When using this model by itself, a weight between 0.8 and 0.9 can also yield excellent results. * Hires. fix (Upscaling): * Sampler: R-ESRGAN_4x+ Anime6B * Denoising Steps: Approximately 20 * Denoising Strength: Approximately 0.3
Example Prompts Positive Prompts:
1girl, solo, long hair, breasts, looking at viewer, bangs, blue eyes, skirt, shirt, bow, navel, animal ears, very long hair, jacket, pink hair, hair bow, heart, choker, midriff, cat ears, crop top, plaid, thigh strap, fake animal ears, headphones, stuffed toy, white jacket, stuffed animal, hand on own chest, fishnets, purple skirt, teddy bear, purple shirt, animal ear headphones, cat ear headphones Negative Prompts: missing fingers,extra digit,,lowres,bad anatomy,bad hands,text,error,fewer digits,cropped,worst quality,low quality,normal quality,jpeg artifacts,watermark,excess toes,too many fingers,
FAQ
Q: What trigger word should I use?
A: No trigger word is required for this model.
Q: What are the recommended settings?
A: Steps: 20, CFG: 7, LoRA Weight: 0.8. For general use, 25-30 steps and a LoRA weight of 0.8-0.9 are also suggested.
Q: What resolution works best?
A: Optimal resolution is not explicitly stated in the original description. However, for models compatible with XL base models, a starting resolution of 1024x1024 is commonly used. For portrait images, 1024x1536 can also be effective.