Canopus-Pencil-Art-LoRA
| Property | Value |
|---|---|
| Base Model | Stable Diffusion XL 1.0 |
| License | Apache 2.0 |
| Network Architecture | LoRA (64 dim, 32 alpha) |
| Training Epochs | 14 |
What is Canopus-Pencil-Art-LoRA?
Canopus-Pencil-Art-LoRA is a specialized fine-tuning model designed to transform images into high-quality pencil art and line drawings. Built on Stable Diffusion XL, it excels at creating clean, black and white illustrations suitable for coloring books and detailed sketches.
Implementation Details
The model utilizes a LoRA architecture with a network dimension of 64 and alpha of 32, trained over 14 epochs. It implements the AdamW optimizer with a constant learning rate scheduler and includes advanced features like noise offset (0.03) and multires noise iterations.
- Optimized for SDXL base 1.0
- Uses EulerAncestralDiscreteScheduler
- Implements specific noise processing parameters
- Supports high-detail line art generation
Core Capabilities
- Clean line art generation with white backgrounds
- Coloring book page creation
- Detailed pencil sketches
- Professional portrait illustrations
- Precise geometric and mechanical drawings
Frequently Asked Questions
Q: What makes this model unique?
The model specializes in creating clean, professional-grade line art and pencil sketches with a focus on detail and composition, making it perfect for coloring book creation and artistic illustrations.
Q: What are the recommended use cases?
Ideal for creating coloring books, technical illustrations, portrait sketches, and clean line art for various artistic and commercial purposes. The model excels at producing black and white drawings with clear lines and professional composition.
