Upload handler.py

handler.py

@@ -11,7 +11,7 @@ class EndpointHandler():
                                            model_id,
                                            model_file=model_file,
                                            model_type="mistral",
-                                           gpu_layers=130,#50 #110
+                                           gpu_layers=50,#50 #110
                                            **config
                                            )
         self.pipeline = llm