add trainer back after save (#1416)

Co-authored-by: Ziqin Xiong <88057852+haha@users.noreply.github.com>

neuralprophet/utils.py

@@ -38,10 +38,17 @@ def save(forecaster, path: str):
         >>> save(forecaster, "test_save_model.np")
     """
     # Remove the Lightning trainer since it does not serialise correcly with torch.save
-    for attr in ["trainer"]:
+    attrs_to_remove = ["trainer"]
+    removed_attrs = {}
+    for attr in attrs_to_remove:
+        removed_attrs[attr] = getattr(forecaster, attr)
         setattr(forecaster, attr, None)
     torch.save(forecaster, path)
 
+    # Restore the Lightning trainer
+    for attr in attrs_to_remove:
+        setattr(forecaster, attr, removed_attrs[attr])
+
 
 def load(path: str):
     """retrieve a fitted model from a .np file that was saved by save.