Update app.py

app.py

@@ -23,7 +23,7 @@ def load_model():
         )
 
     # Load the processor and model using the correct identifier
-    model_id = "google/paligemma2-10b-pt-896"
+    model_id = "google/paligemma2-28b-pt-896"
     processor = PaliGemmaProcessor.from_pretrained(model_id, use_auth_token=token)
     device = "cuda" if torch.cuda.is_available() else "cpu"
     model = PaliGemmaForConditionalGeneration.from_pretrained(
@@ -40,11 +40,8 @@ def process_image_and_text(image_pil, text_input):
     device = "cuda" if torch.cuda.is_available() else "cpu"
 
     # Load the image using load_image
-    # Convert PIL image to bytes
-    buffered = BytesIO()
-    image_pil.save(buffered, format="JPEG")
-    image_bytes = buffered.getvalue()
-    image = load_image(image_bytes)
+    # We can pass the PIL image directly to load_image
+    image = load_image(image_pil)
 
     # Use the provided text input
     model_inputs = processor(text=text_input, images=image, return_tensors="pt").to(