Illustrious-xl-early-release-v0
| Property | Value |
|---|---|
| Base Model | KBlueLeaf/kohaku-xl-beta5 |
| License | fair-ai-public-license-1.0-sd |
| Technical Paper | View Paper |
| Pipeline | Text-to-Image |
What is Illustrious-xl-early-release-v0?
Illustrious XL is a specialized Stable Diffusion XL model developed by OnomaAI Research Team, specifically designed for illustration generation. Built upon Kohaku XL Beta 5, this model has been fine-tuned using the comprehensive Danbooru2023 dataset to excel at creating high-quality anime and manga-style artwork.
Implementation Details
The model comes in two variants: v0.1 (base) and v0.1-GUIDED, with the latter incorporating additional safety controls. It utilizes the StableDiffusionXLPipeline architecture and is optimized for both research and creative applications.
- Recommended sampling method: Euler a
- Optimal sampling steps: 20-28
- Suggested CFG range: 5-7.5
- Supports various quality tags from "worst quality" to "masterpiece"
Core Capabilities
- High-quality illustration generation
- Extensive character design knowledge
- Various artistic style reproduction
- Flexible composition control
- Support for quality-based prompting
Frequently Asked Questions
Q: What makes this model unique?
The model's specialization in illustration generation, combined with its foundation on the comprehensive Danbooru2023 dataset, makes it particularly effective for creating anime and manga-style artwork. The availability of both base and safety-guided variants provides flexibility for different use cases.
Q: What are the recommended use cases?
The model is ideal for illustration generation, character design, and artistic content creation. The base version is suitable for research purposes, while the GUIDED version is recommended for general use cases requiring additional safety controls.
