app.py

import gradio as gr
from transformers import RAGMultiModalModel  # Importing the ColPali model

# Initialize the ColPali model
model = RAGMultiModalModel.from_pretrained("vidore/colpali-v1.2")

def extract_and_search(image, keyword):
    # Use the model to extract text from the image
    inputs = {"images": [image]}
    extracted_text = model.generate(**inputs)  # Replace with actual prediction method

    # Perform keyword search
    matching_lines = [line for line in extracted_text.splitlines() if keyword.lower() in line.lower()]
    
    return extracted_text, matching_lines

# Create Gradio interface
interface = gr.Interface(
    fn=extract_and_search,
    inputs=[
        gr.Image(type="pil", label="Upload Image"),
        gr.Textbox(label="Enter Keyword")
    ],
    outputs=[
        gr.Textbox(label="Extracted Text"),
        gr.Textbox(label="Matching Lines")
    ],
    title="ColPali OCR with Keyword Search",
    description="Upload an image and enter a keyword to search within the extracted text."
)

# Launch the app
interface.launch(share=True)