Update modeling_ddllama.py

modeling_ddllama.py

@@ -1225,10 +1225,6 @@ class DDLlamaForCausalLM(DDLlamaPreTrainedModel, GenerationMixin):
             router_masks = [mask.to(hidden_states.device) for mask in router_masks]
             router_masks = torch.stack(router_masks, dim=-1).float()
 
-            n_layers = torch.sum(router_masks, dim=1).squeeze(dim=0)
-            n_layers_list = n_layers.cpu().tolist()
-            print(n_layers_list, end='')
-
             if self.training and labels is not None:
                 router_weights *= router_masks
                 shift_router_weights = router_weights[:, :-1, :].contiguous()