Canopus-Clothing-Flux-LoRA
| Property | Value |
|---|---|
| Base Model | black-forest-labs/FLUX.1-dev |
| License | CreativeML OpenRAIL-M |
| Captioning Model | microsoft/Florence-2-large |
| Network Dimensions | 64 |
| Network Alpha | 32 |
What is Canopus-Clothing-Flux-LoRA?
Canopus-Clothing-Flux-LoRA is a specialized fine-tuned LoRA model designed for generating high-quality clothing images. Built on the FLUX.1-dev base model and utilizing Florence-2-large for natural language processing, it excels at creating detailed clothing visualizations with specific style attributes.
Implementation Details
The model implements a sophisticated training approach using AdamW8bit optimizer with constant LR scheduling. It features a noise offset of 0.03 and employs multires noise iterations with a discount of 0.1 across 10 iterations. Training occurred over 15 epochs with checkpoints saved every epoch.
- Specialized trigger words: Hoodie, Clothes, Clothing, Shirt, Pant
- Supports various aspect ratios for different clothing presentations
- Implements studio-quality lighting and composition parameters
Core Capabilities
- Detailed garment visualization with precise fabric textures
- Multi-style clothing generation including urban, gothic, and minimalist aesthetics
- Professional product photography-style outputs
- Support for various clothing types from casual to formal wear
Frequently Asked Questions
Q: What makes this model unique?
The model combines Florence-2-large's natural language understanding with specialized clothing generation capabilities, offering precise control over garment styles, materials, and presentation.
Q: What are the recommended use cases?
Ideal for e-commerce clothing visualization, fashion design prototyping, and creating detailed clothing mockups with specific style attributes and environmental contexts.
