fix: cast mask to bool

modeling_bert.py

@@ -184,7 +184,7 @@ class BertEncoder(nn.Module):
         """
         if key_padding_mask is None or not self.use_flash_attn:
             mixer_kwargs = (
-                {"key_padding_mask": key_padding_mask} if key_padding_mask is not None else None
+                {"key_padding_mask": key_padding_mask.bool()} if key_padding_mask is not None else None
             )
             for layer in self.layers:
                 hidden_states = layer(hidden_states, mixer_kwargs=mixer_kwargs)