|
|
--- |
|
|
library_name: transformers |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- vision |
|
|
- image-captioning |
|
|
- blip |
|
|
- multimodal |
|
|
- fashion |
|
|
datasets: |
|
|
- Marqo/fashion200k |
|
|
base_model: |
|
|
- Salesforce/blip-image-captioning-large |
|
|
--- |
|
|
|
|
|
# Fine-Tuned BLIP Model for Fashion Image Captioning |
|
|
|
|
|
This is a fine-tuned BLIP (Bootstrapped Language-Image Pretraining) model specifically designed for **fashion image captioning**. It was fine-tuned on the **Marqo Fashion Dataset** to generate descriptive and contextually relevant captions for fashion-related images. |
|
|
|
|
|
## Model Details |
|
|
|
|
|
- **Model Type:** BLIP (Vision-Language Pretraining) |
|
|
- **Architecture:** BLIP uses a multimodal transformer architecture to jointly model visual and textual information. |
|
|
- **Fine-Tuning Dataset:** [Marqo Fashion Dataset](https://github.com/marqo-ai/marqo) (a dataset containing fashion images and corresponding captions) |
|
|
- **Task:** Fashion Image Captioning |
|
|
- **License:** Apache 2.0 |
|
|
|
|
|
## Usage |
|
|
|
|
|
You can use this model with the Hugging Face `transformers` library for fashion image captioning tasks. |
|
|
|
|
|
### Installation |
|
|
|
|
|
First, install the required libraries: |
|
|
|
|
|
```bash |
|
|
pip install transformers torch |
