app.py

import streamlit as st
import altair as alt
from transformers import pipeline
from transformers import AutoModelForSeq2SeqLM , AutoTokenizer, TranslationPipeline
from PIL import Image


st.title("Image-Based Question Answering 🕵️‍♂️")

st.subheader("Ask questions directly from images!")

st.write("""
Upload an image (e.g., receipts, documents), type your question, and get precise answers in real-time. 
Powered by the advanced `naver-clova-ix/donut-base-finetuned-docvqa` model.
""")

# context_input = st.text_area("please provice some context", "Many NLP tasks are now benchmarked using datasets like GLUE and SuperGLUE.  Multilingual NLP models like mBERT support multiple languages in a single framework.")

# question_input = st.text_area("enter question about NLP", "what model support multilingual nlp?")
@st.cache_resource
def load_model():
    print("Loading model...")
    return pipeline("document-question-answering", model="naver-clova-ix/donut-base-finetuned-docvqa")


dunno_answerer  = load_model()
# with open('NLP_History_and_Facts.txt', 'r') as file:
#     context = file.read()
uploaded_image = st.file_uploader("Upload an image", type=["png", "jpg", "jpeg"])

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

    question_input = st.text_area("Enter your question", "Any questions ?")

    if st.button("Answer!"):
        if question_input.strip():
            result = dunno_answerer(image=image, question=question_input)

            st.write(f"**Answer:** {result[0]['answer']}")
        else:
            st.write("Please enter a valid question!")