Ml Flow correctly uses a pathlib Path object when saving a model using mlflow.keras.save_model, but it does not correctly interpret the pathlib object when loading the saved model back in using mlflow.keras.load_model.
A bug should be reported through the ML Flow issues interface. It should be very easy to create a basic code snippet to demonstrate this problem.
Ml Flow correctly uses a pathlib Path object when saving a model using
mlflow.keras.save_model, but it does not correctly interpret the pathlib object when loading the saved model back in usingmlflow.keras.load_model.A bug should be reported through the ML Flow issues interface. It should be very easy to create a basic code snippet to demonstrate this problem.