Update modeling_afmoe_scm_liger.py

modeling_afmoe_scm_liger.py

@@ -301,6 +301,7 @@ class AfmoeSCMAttention(nn.Module):
 
         self.scaling = self.head_dim**-0.5
         self.attention_dropout = config.attention_dropout
+        self.is_causal = True
         self.is_local_attention = config.layer_types[layer_idx] == "sliding_attention"
         self.sliding_window = config.sliding_window if self.is_local_attention else None