Retrain fitted model with new data

[RamiKrispin]

Hello,

I am trying to log a forecast model from the MLforecast library into MLflow using the mlforecast.flavor.log_model function. The goal is to promote this model to production and reuse it when new data is available to generate a new forecast. Ideally, I would log the pre-trained object and use the fit on the new data, but it requires logging the fitted object. What are the best practices to refit a train model (or fitted object) when new data is available?

Thanks,
Rami