---
title: ImageCaptionTestSpace
emoji: 😻
colorFrom: blue
colorTo: red
sdk: gradio
sdk_version: 5.49.0
app_file: app.py
pinned: false
---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

# Image Captioning (ViT-GPT2) — Hugging Face Space

This Space serves an image captioning model using Hugging Face `VisionEncoderDecoderModel` (ViT + GPT-2).
It runs out-of-the-box with the base model and can optionally load your **fine-tuned** weights.

**Live app entrypoint:** `app.py` (Gradio)

## Quick Start (on Spaces)
1. Click **New Space** → **Gradio** → **Blank** → pick a free CPU or T4 small (GPU) runtime.
2. Upload all files from this repo.
3. (Optional) If you have fine-tuned weights:
   - Upload the saved folder to the Space (e.g., `outputs/caption_finetune/`)
   - Set a Space secret or environment variable: `MODEL_DIR = outputs/caption_finetune`
   - Alternatively push your weights to the Hub and set `MODEL_DIR = your-username/your-model-repo`

If `MODEL_DIR` is not set, the app uses `nlpconnect/vit-gpt2-image-captioning`.

## Local Dev
```bash
pip install -r requirements.txt
python app.py
# then open http://127.0.0.1:7860