Update app.py

app.py

@@ -1,17 +1,17 @@
 import streamlit as st
 from PIL import Image
 import easyocr
-from transformers import pipeline, AutoTokenizer, AutoModel
+import torch
+from transformers import CLIPProcessor, CLIPModel, pipeline
 
-# Load CLIP model
+# Load CLIP Model & Processor
 @st.cache_resource
 def load_clip_model():
-    pipe = pipeline("feature-extraction", model="fxmarty/clip-vision-model-tiny")
-    tokenizer = AutoTokenizer.from_pretrained("fxmarty/clip-vision-model-tiny")
-    model = AutoModel.from_pretrained("fxmarty/clip-vision-model-tiny")
-    return pipe, tokenizer, model
+    model = CLIPModel.from_pretrained("fxmarty/clip-vision-model-tiny")
+    processor = CLIPProcessor.from_pretrained("fxmarty/clip-vision-model-tiny")
+    return model, processor
 
-pipe, tokenizer, model = load_clip_model()
+model, processor = load_clip_model()
 
 # Initialize OCR
 @st.cache_resource
@@ -43,8 +43,8 @@ if uploaded_file is not None:
 
     if user_question:
         with st.spinner("Analyzing image and question..."):
-            inputs = tokenizer(user_question, return_tensors="pt")
-            outputs = model(**inputs)
+            inputs = processor(text=[user_question], images=image, return_tensors="pt")
+            outputs = model.get_image_features(**inputs)
         
         st.write("### 🏆 AI Response:")
-        st.write("CLIP Model Processed the Input! (Further improvements coming soon)")
+        st.write("CLIP Model Processed the Image! (Further improvements coming soon)")