Jupyter notebook fails to connect to kernel due to module name conflict

I had the following folder structure for a python ML project:

├── root
   ├── code
   │   ├── foo.py
   │   └── bar.py
   ├── notebook.ipynb
   ├── foo
   └── bar

After writing the two .py files in the code folder, I tried to run the two files but had some errors similar to this:

ModuleNotFoundError: No module named 'code.foo'; 'code' is not a package

This was caused by imports along the lines of this:

from code.foo import x

I tried searching for a solution and found out that adding an empty __init.py__ file would help "having python know my folder was a module" (sorry for the bland description, I can't remember the technicalities). In fact, I had no such errors since. New hierarchy looks like this:

├── root
   ├── code
   │   ├── __init__.py
   │   ├── foo.py
   │   └── bar.py
   ├── notebook.ipynb
   ├── foo
   └── bar

Unfortunately, as I tried opening notebook.ipynb with Jupyter Notebook, this failed to connect to the kernel, showing this error in the command prompt:

i.e.: the module "code" could not be imported because it is shadowed by:
...\root\code\__init__.py
Please rename this file/folder so that the original module from the standard library can be imported.

I could delete/rename the __init.py__ file everytime I need to use the notebook, but that's not very efficient in terms of workflow. Is there any solution to this conflict? As I stated I have no necessity of keeping the __init.py__ file, if that information helps.

Thanks for the attention!