Florence-2-base-PromptGen-v2.0
| Property | Value |
|---|---|
| Parameter Count | 271M |
| Tensor Type | F32 |
| License | MIT |
| Author | MiaoshouAI |
What is Florence-2-base-PromptGen-v2.0?
Florence-2-base-PromptGen-v2.0 is an advanced image captioning model that builds upon its predecessor (v1.5) with significant improvements in caption generation capabilities. This lightweight model excels at producing high-quality image descriptions while maintaining exceptional efficiency with only 1GB VRAM usage.
Implementation Details
The model implements multiple instruction modes for versatile image analysis and caption generation. It's designed specifically to work with Flux models for both T5XXL CLIP and CLIP_L, enabling efficient single-pass caption generation.
- Supports multiple instruction types including GENERATE_TAGS, CAPTION, DETAILED_CAPTION, and MORE_DETAILED_CAPTION
- New ANALYZE instruction for comprehensive image composition analysis
- Memory-efficient architecture requiring only 1GB VRAM
- Integrated support for Flux model compatibility
Core Capabilities
- Enhanced caption quality across all instruction modes
- Detailed image composition analysis
- Danbooru-style tag generation
- Structured position-aware captioning
- Mixed caption generation combining detailed descriptions with tags
Frequently Asked Questions
Q: What makes this model unique?
The model's standout feature is its ability to deliver high-quality captions while maintaining extremely efficient resource usage. Its versatile instruction set and new ANALYZE capability make it particularly valuable for detailed image understanding tasks.
Q: What are the recommended use cases?
The model is ideal for automated image captioning, tag generation for image databases, detailed image analysis, and integration with Flux models. It's particularly useful in scenarios where resource efficiency is crucial while maintaining high-quality output.
