Update modeling_qwen.py

modeling_qwen.py

@@ -646,7 +646,7 @@ class Qwen2SdpaAttention(Qwen2Attention):
         past_key_value: Optional[Cache] = None,
         output_attentions: bool = False,
         use_cache: bool = False,
-        is_causal: bool = True,
+        is_causal: bool = False,
     ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
         if output_attentions:
             # TODO: Improve this warning with e.g. `model.config.attn_implementation = "manual"` once this is implemented.