I have complete pipeline in place from data inputs to final predictions. I have 10 models in place (wrote function to automate the whole process) then , Choose 1 model based on accuracy. Now, I need to store all the model(because any model can have better accuracy as data changes) and reuse it with new values of inputs from my train features.
As all process is very dynamic, for each step job trigger and does the job. Need help in how I can save and reuse this model which aligned with this flow.
Thank you in advance. an example would be a great help
did you try ?
mlflow.pyfunc.log_model for saving the model
mlflow.pyfunc.load_model(f"models:/{model_name}/production") for loading the model