update type hint for seqsignet path

src/nlpsig/data_preparation.py

@@ -1211,7 +1211,7 @@ def get_torch_path_for_SeqSigNet(
         include_features_in_input: bool,
         include_embedding_in_input: bool,
         reduced_embeddings: bool = False,
-    ) -> dict[str, torch.tensor | int]:
+    ) -> dict[str, dict[str, torch.tensor] | int | None]:
         """
         Returns a `torch.tensor` object that can be passed into `nlpsig_networks.SeqSigNet` model.