Update Gradio app with multiple files

app.py

@@ -21,19 +21,6 @@ tokenizer = AutoTokenizer.from_pretrained(
 print("Model loaded successfully!")
 
 
-def filter_think_tags(text):
-    """
-    Remove ... tags and their content from the response.
-    """
-    # Remove ... blocks including the tags
-    text = re.sub(r'.*?', '', text, flags=re.DOTALL)
-    # Remove any remaining opening or closing think tags
-    text = re.sub(r'</?think>', '', text)
-    # Clean up extra whitespace
-    text = re.sub(r'\n\s*\n', '\n', text).strip()
-    return text
-
-
 @spaces.GPU
 def respond(message, history):
     """
@@ -79,12 +66,9 @@ def respond(message, history):
         output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
     ]
     
-    # Decode - skip special tokens to filter out think tokens
+    # Decode - skip special tokens will help but we'll also filter manually
     response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
     
-    # Filter out  tags and their content
-    response = filter_think_tags(response)
-    
     # For streaming effect, yield character by character
     partial_response = ""
     for char in response:
@@ -121,6 +105,7 @@ with gr.Blocks(
             "Explain AI in one sentence",
         ],
         cache_examples=False,
+        chatbot=gr.Chatbot(allow_tags=["think"]),
     )
     
     gr.Markdown(