-
The data is from the Zindi Epresso Churn Competition
-
The SQL data is loaded from CSV files in a path with permission, for my sql it is as such "C:/ProgramData/MySQL/MySQL Server 8.0/Uploads/", it will vary depending on what you are using e.g. Postgres etc. Alternatively, you could alter the file priviledges to read from different folders.
-
For Tensorflow serving , use the Docker image on Docker Hub to orchastrate and serve the model locally on your PC. Have it running at the same time as the Flask App, the ports will/should be different.
-
The model was trained initially and saved, for more details check here
-
I run my docker container eith the following command :: docker run -t -p 8501:8501 --mount type=bind,source='/{path to folder}',target=/models/saved_model -e MODEL_NAME=saved_model tensorflow/serving
-
Replace {path to folder} with the actual path to the folder with your model
-
For the employee login, initially add a user ID and Email manually on your SQL server(simulate an admin situation), I avoided adding the whole idea of acces levels, atleast for now, but if you are familiar with it feel free to do so.