Update app.py

app.py

@@ -32,7 +32,8 @@ Provide a concise answer in markdown format with citations [1], [2], etc."""
 @spaces.GPU(duration=120)  # Allow up to 120 seconds of GPU time
 def generate_answer(prompt: str) -> str:
     """Generate a concise research answer using GPU."""
-    response = generator(prompt, max_length=200, num_return_sequences=1, truncation=True)[0]["generated_text"]
+    # Use max_new_tokens instead of max_length to allow new token generation
+    response = generator(prompt, max_new_tokens=150, num_return_sequences=1, truncation=True)[0]["generated_text"]
     answer_start = response.find("Provide a concise") + len("Provide a concise answer in markdown format with citations [1], [2], etc.")
     return response[answer_start:].strip() if answer_start > -1 else "No detailed answer generated."
 
@@ -66,7 +67,7 @@ def process_deep_research(query: str, history: list):
     prompt = format_prompt(query, web_results)
     answer = generate_answer(prompt)
     
-    # Convert tuple history to messages format (role/content)
+    # Convert history to messages format (role/content)
     new_history = history + [{"role": "user", "content": query}, {"role": "assistant", "content": answer}]
     
     return answer, sources_html, new_history
@@ -174,7 +175,6 @@ with gr.Blocks(title="Deep Research Engine - ZeroGPU", css=css) as demo:
         answer, sources, new_history = process_deep_research(query, history)
         return answer, sources, new_history
 
-    # Remove _js parameter from click event
     search_btn.click(
         fn=handle_search,
         inputs=[search_input, history_state],
@@ -185,7 +185,6 @@ with gr.Blocks(title="Deep Research Engine - ZeroGPU", css=css) as demo:
         outputs=[history_state]
     )
 
-    # Remove _js parameter from submit event
     search_input.submit(
         fn=handle_search,
         inputs=[search_input, history_state],