Make passing the IP Adapter mask to the attention mechanism optional (huggingface#10346)

Make passing the IP Adapter mask to the attention mechanism optional if there is no need to apply it to a given IP Adapter.

Author: elismasilva

src/diffusers/models/attention_processor.py

@@ -4839,6 +4839,8 @@ def __call__(
                 )
             else:
                 for index, (mask, scale, ip_state) in enumerate(zip(ip_adapter_masks, self.scale, ip_hidden_states)):
+                    if mask is None:
+                        continue
                     if not isinstance(mask, torch.Tensor) or mask.ndim != 4:
                         raise ValueError(
                             "Each element of the ip_adapter_masks array should be a tensor with shape "
@@ -5056,6 +5058,8 @@ def __call__(
                 )
             else:
                 for index, (mask, scale, ip_state) in enumerate(zip(ip_adapter_masks, self.scale, ip_hidden_states)):
+                    if mask is None:
+                        continue
                     if not isinstance(mask, torch.Tensor) or mask.ndim != 4:
                         raise ValueError(
                             "Each element of the ip_adapter_masks array should be a tensor with shape "