Update app.py

app.py

@@ -84,7 +84,7 @@ else:
     import pickle
     with open('nn_model.pkl', 'rb') as f:
         nn = pickle.load(f)
-
+@spaces.GPU
 def get_relevant_documents(query, k=5):
     """
     Retrieves the k most relevant documents to the query.
@@ -101,7 +101,7 @@ def get_relevant_documents(query, k=5):
     elapsed_time = time.time() - start_time
     print(f"[PERF] get_relevant_documents took {elapsed_time:.2f} seconds")
     return relevant_docs
-
+@spaces.GPU
 def generate_response(question, history):
     import time
     start_time = time.time()
@@ -115,7 +115,7 @@ def generate_response(question, history):
     elapsed_time = time.time() - start_time
     print(f"[PERF] generate_response took {elapsed_time:.2f} seconds")
     return response
-
+@spaces.GPU
 def _generate_response_gpu(question, history):
     print(f"\n[LOG] Received question: {question}")
     
@@ -201,7 +201,7 @@ Provide the AutoGen v0.4 agent code that fulfills the user's request. Utilize fe
     # Update chat history with new message pair
     history.append((question, response))
     return history
-
+@spaces.GPU
 def _generate_response_cpu(question, history):
     print(f"[LOG] Running on CPU")
     try: