app.py

import streamlit as st
from transformers import AutoProcessor, VisionEncoderDecoderModel
import requests
from PIL import Image
import torch

# Load processor and model
st.title("Image to Text Captioning App")
st.write("This app converts an image into a text description using the ViT-GPT2 model.")

@st.cache_resource
def load_model():
    processor = AutoProcessor.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
    model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
    return processor, model

processor, model = load_model()

# Upload image
uploaded_file = st.file_uploader("Upload an image", type=["jpg", "png", "jpeg"])

if uploaded_file is not None:
    image = Image.open(uploaded_file).convert("RGB")
    st.image(image, caption="Uploaded Image", use_column_width=True)

    # Preprocessing image
    pixel_values = processor(images=image, return_tensors="pt").pixel_values

    # Generate caption
    generated_ids = model.generate(pixel_values)
    generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]

    st.write("Generated Caption: ")
    st.success(generated_text)