| Property | Value |
|---|---|
| Base Model | black-forest-labs/FLUX.1-dev |
| License | CreativeML OpenRAIL-M |
| Captioning Model | microsoft/Florence-2-large |
| Training Epochs | 15 |
What is Canopus-Clothing-Flux-LoRA?
Canopus-Clothing-Flux-LoRA is a specialized fine-tuned LoRA model designed for generating high-quality clothing and fashion item images. Built on the FLUX.1-dev base model, it incorporates advanced training parameters and the Florence-2-large captioning model to produce detailed clothing visualizations.
Implementation Details
The model utilizes a sophisticated training setup with AdamW8bit optimizer, constant LR scheduler, and specific noise parameters (offset: 0.03, discount: 0.1). It features a network dimension of 64 and alpha of 32, trained over 15 epochs with 3K+ steps.
- Optimized for clothing items including hoodies, shirts, pants, and general apparel
- Implements natural language processing through florence2-en
- Supports various aspect ratios and composition styles
- Trained with specific trigger words for optimal generation
Core Capabilities
- Detailed clothing item generation with precise fabric rendering
- Multiple style support from gothic to minimalist
- Studio-quality clothing visualization
- Versatile composition handling for various garment types
Frequently Asked Questions
Q: What makes this model unique?
The model combines FLUX.1-dev's capabilities with specialized clothing-focused training, using Florence-2-large for enhanced natural language understanding and precise garment generation.
Q: What are the recommended use cases?
Ideal for fashion design visualization, e-commerce product imagery, and clothing mockups. The model excels at generating various garment types with detailed fabric textures and styling.