333-PRO
| Property | Value |
|---|---|
| Base Model | FLUX.1-dev |
| Training Images | 33 |
| Optimal Resolution | 1280 x 832 |
| Model Type | LoRA |
| Author | strangerzonehf |
What is 333-PRO?
333-PRO is a specialized LoRA model designed for creating detailed black and white sketches. Built on the FLUX.1-dev base model, it excels in generating artistic drawings with a particular focus on portrait and figure sketches. The model was trained for 23 epochs using carefully curated 33 images and implements advanced noise processing parameters.
Implementation Details
The model utilizes the AdamW optimizer with a constant learning rate scheduler, featuring a network dimension of 64 and alpha of 32. It incorporates sophisticated noise handling with a 0.03 offset and multi-res noise discount of 0.1 across 10 iterations. The training process involved 2950 steps with 25 repeats.
- Optimized for 3:2 aspect ratio (1280x832)
- Recommended inference steps: 30-35
- Uses "333 Pro Sketch" as trigger phrase
- Implements florence2-en labeling for natural language processing
Core Capabilities
- High-quality black and white sketch generation
- Detailed portrait and figure rendering
- Precise texture and shading control
- Consistent style maintenance across generations
Frequently Asked Questions
Q: What makes this model unique?
The model's specialization in black and white sketches, combined with its optimized parameters for detailed rendering and its specific trigger phrase system, makes it particularly effective for artistic sketch generation.
Q: What are the recommended use cases?
The model excels in creating detailed portraits, figure sketches, and artistic compositions. It's particularly suited for artists and creators looking to generate high-quality black and white illustrations with consistent style and detail.
