Hi,
If I currently use joblib or pickle to seralize my pipeline , it is only saving my ARIMA model. But in my needs, I want to load the model again and update with new data which gives me errors because different shape size due to missing transformations. I have some models using transforms and some dont, so I cannot use the transform() method because I don't know which model has which steps. Is it possible to save the entire pipeline here or am I missing something? TIA
Describe the question you have
Hi, If I currently use joblib or pickle to seralize my pipeline , it is only saving my ARIMA model. But in my needs, I want to load the model again and update with new data which gives me errors because different shape size due to missing transformations. I have some models using transforms and some dont, so I cannot use the transform() method because I don't know which model has which steps. Is it possible to save the entire pipeline here or am I missing something? TIA
Versions (if necessary)
No response