How do you modify tensorflow python code so that when you import this into your own python files the modifications are picked up?

Question

I am trying to debug a use-case and the error messages that are given by tensorflow are cryptic. Therefore I want to modify the tensorflow code base to print some values so I see exactly what is going wrong. Unfortunately I am not able to simply modify the python file and run, as the changes are not getting picked up. Is there any extra step to work on? To be concrete

https://github.com/tensorflow/tensorflow/blob/98a2d7d166c6ba19539e6d1fa57c47c4170ee77b/tensorflow/python/keras/engine/data_adapter.py

I have modified line 1233

 "Your input ran out of data; interrupting training." with "Your input ran out of data; interrupting training. TESTING!!!!"

Basically I added some extra log here just to see how this works. However, the changes do not seem to be getting picked up when I run my code which imports tensorflow. I have made the modifications here

 /path_to_anaconda/anaconda3/envs/py39tf26cpu/lib/python3.9/site-packages/tensorflow/python/keras/engine

Any help would be greatly appreciated :)