Update modules/pmbl.py

modules/pmbl.py

@@ -104,7 +104,7 @@ class PMBL:
         for chunk in response.result():
             yield chunk
             
-    @spaces.GPU
+    
     def generate_response_task(self, system_prompt, prompt, n_ctx):
         llm = Llama(model_path=self.model_path, n_ctx=n_ctx, n_threads=8, n_gpu_layers=-1, mlock=True)
         llm = llm.to("cuda")  # Move the model to the GPU
@@ -152,7 +152,7 @@ class PMBL:
 
         conn.close()
         
-    @spaces.GPU
+    
     def generate_topic(self, prompt, response):
         llm = Llama(model_path=self.model_path, n_ctx=1690, n_threads=8, n_gpu_layers=-1, mlock=True)
         llm = llm.to("cuda")  # Move the model to the GPU