add unexpected keys

ml-agents/mlagents/trainers/model_saver/torch_model_saver.py

 
         for name, mod in modules.items():
             try:
-                missing_keys, _ = mod.load_state_dict(
+                missing_keys, unexpected_keys = mod.load_state_dict(
                     saved_state_dict[name], strict=False
                 )
                 if missing_keys:
+                if unexpected_keys:
+                    logger.warning(
+                        f"Did not expect these keys {unexpected_keys} in checkpoint. Initializing"
+                    )
+
             except (KeyError, TypeError):
                 logger.warning(f"Failed to load for module {name}. Initializing")