README.md

metadata

title: Image Captioner
emoji: 🖼️
colorFrom: indigo
colorTo: blue
sdk: gradio
app_file: app.py
pinned: false
license: mit

🖼️ Image Captioner

A lightweight, CPU-friendly image captioning demo built with 🤗 Transformers + Gradio.
Upload an image and the model will generate a concise caption.

We provide a few test images in samples/.

Run:

# from project root (with venv active)
python -m image_captioner.scripts.run_batch --dir image_captioner/samples --jsonl-out captions.jsonl --csv-out captions.csv

You’ll see progress like:

[1/2] living_room.jpg -> a living room with a couch and chairs [2/2] street.jpg -> a busy street with cars and pedestrians

Outputs: • captions.jsonl → one JSON record per image • captions.csv → tabular view (image, caption, elapsed time)

---