retest with 30

gdc_pipeline.py

@@ -128,7 +128,7 @@ def construct_and_execute_api_call(
 
 
 # generate llama model response
-@spaces.GPU(duration=10)
+@spaces.GPU(duration=30)
 def generate_response(modified_query, model, tok):
     set_seed(1042)
     # model = model.to('cuda' if torch.cuda.is_available() else 'cpu')