app.py

import streamlit as st
import random
import requests
import io
from PIL import Image
from transformers import AutoProcessor, Blip2ForConditionalGeneration
import torch


model = None
processor = None

st.title("Image Captioner - Caption the images")
st.markdown("Link to the model - [Image-to-Caption-App on 🤗 Spaces](https://huggingface.co/spaces/Shrey23/Image-Captioning)")


class UI:
    def __init__(self):
        mod = Model()
        mod.load_model()

    def displayUI(self):
        image = st.file_uploader(label = "Upload your image here",type=['png','jpg','jpeg'])
        if image is not None:

            input_image = Image.open(image) #read image
            st.image(input_image) #display image

            with st.spinner("🤖 AI is at Work! "):
                device = "cuda" if torch.cuda.is_available() else "cpu"
                print(1, device)
                inputs = processor(images=input_image, return_tensors="pt").to(device, torch.float16)

                print(2)
                pixel_values = inputs.pixel_values
                
                # print(3)
                # pixel_values = pixel_values.to(torch.float32)

                print(3)
                generated_ids = model.generate(pixel_values=pixel_values, max_length=25) #.to(device, torch.float32)
                
                print(4)
                generated_caption = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]

                print(5, generated_caption)
                st.write(generated_caption)
                
            st.success("Here you go!")
            st.balloons()
        else:
            st.write("Upload an Image")

        st.caption("NN Practical Project.")
                        

class Model:
    def load_model(self):
        global model
        global processor
        model = Blip2ForConditionalGeneration.from_pretrained("Shrey23/Image-Captioning", device_map="auto", load_in_8bit=True, torch_dtype=torch.float16) #, 
        processor = AutoProcessor.from_pretrained("Salesforce/blip2-opt-2.7b")

def main():
    ui = UI()
    ui.displayUI()

if __name__ == "__main__":
    main()