Update app.py

app.py

@@ -1,9 +1,39 @@
 import streamlit as st
+import torch
 
+from transformers import Blip2Processor, Blip2ForConditionalGeneration
 
-def answer_question(image, question):
-    # Integrate your model logic here
-    answer = "This is where the answer will appear."
+global device = 'cuda' if torch.cuda.is_available() else 'cpu'
+
+def load_caption_model():
+    # Quantization Config
+    bnb_config = BitsAndBytesConfig(
+    load_in_8bit=True,
+    bnb_8bit_quant_type="nf4",
+    bnb_8bit_compute_dtype=torch.float16,
+    bnb_8bit_use_double_quant=False
+    )
+
+
+
+
+    processor = Blip2Processor.from_pretrained("Salesforce/blip2-opt-2.7b")
+    model = Blip2ForConditionalGeneration.from_pretrained("Salesforce/blip2-opt-2.7b", load_in_8bit=True,torch_dtype=torch.float16, device_map="auto")
+
+    return, model, processor
+
+    
+
+def answer_question(image, question, model, processor):
+    
+    
+    image = Image.open(image).convert('RGB')
+
+    inputs = processor(image, question, return_tensors="pt").to("cuda", torch.float16)
+
+    out = model.generate(**inputs, max_length=200, min_length=20, num_beams=1)
+
+    answer = processor.decode(out[0], skip_special_tokens=True).strip()
     return answer
 
 st.title("Image Question Answering")
@@ -19,7 +49,8 @@ if st.button("Get Answer"):
         # Display the image
         st.image(image, use_column_width=True)
         # Get and display the answer
-        answer = answer_question(image, question)
+        model, processor = load_caption_model()
+        answer = answer_question(image, question, model, processor)
         st.write(answer)
     else:
         st.write("Please upload an image and enter a question.")