The issue has been solved!
My solution was to download the model on my local machine, unpack and include my custom inference script, and then call the .model() and .deploy()` without including the entry scripts and directories.
1 Like