Update app.py

app.py

@@ -1,9 +1,12 @@
 import streamlit as st
 from PIL import Image
-from transformers import pipeline
+from transformers import AutoFeatureExtractor, AutoModelForImageCaptioning, AutoTokenizer
 
-# Initialize the image captioning model
-image_captioning_model = pipeline("image-captioning", model="Salesforce/blip-image-captioning-large")
+# Load the image captioning model
+model_name = "Salesforce/blip-image-captioning-large"
+feature_extractor = AutoFeatureExtractor.from_pretrained(model_name)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForImageCaptioning.from_pretrained(model_name)
 
 # Streamlit app title and description
 st.title("Image Information Extractor")
@@ -21,12 +24,14 @@ if image:
     st.write("Extracting information from the image...")
 
     # Use the image captioning model to generate a description
-    captions = image_captioning_model(image)
+    inputs = tokenizer(image, return_tensors="pt")
+    with st.spinner("Generating caption..."):
+        captions = model.generate(**inputs)
+    caption = tokenizer.decode(captions[0], skip_special_tokens=True)
 
     # Display the image description
     st.write("Image Description:")
-    for caption in captions:
-        st.write(caption)
+    st.write(caption)
 
     # You can add more processing or explanations here if needed