Update app.py

app.py

@@ -26,7 +26,7 @@ def llava(message, history):
         image = message["files"][0]    
     else:
         for hist in history:
-            if type(hist[0])==tuple:
+            if isinstance(hist[0], tuple):
                 image = hist[0][0]
         
     txt = message["text"]
@@ -36,7 +36,10 @@ def llava(message, history):
     prompt = f"<|im_start|>user <image>\n{txt}<|im_end|><|im_start|>assistant"
     
     inputs = processor(prompt, image, return_tensors="pt")
-    return inputs
+    
+    # Return the dictionary format expected by MultimodalTextbox
+    return {"text": txt, "files": [image]}
+
 
 def extract_text_from_webpage(html_content):
     soup = BeautifulSoup(html_content, 'html.parser')