import streamlit as st from transformers import pipeline from PIL import Image question = pipeline("visual-question-answering" , model="JBDef/finetuned_yelp") def main(): st.title("Visual Question Answering") with st.form("text_field"): file = st.file_uploader("Upload image", type=["jpg", "png"] ) text_input = st.text_input("Enter some question :") # clicked==True only when the button is clicked clicked = st.form_submit_button("Submit") if clicked: img = Image.open(file) image = st.image(img) results = question(dict(image) , str(text_input)) st.json(results) if __name__ == "__main__": main()