This message was deleted.

Literally worked on that today 😄 So, at current moment, this is what I came to:

from statsforecast import StatsForecast
from statsforecast.models import (
    AutoARIMA,
)
# Create a list of models and instantiation parameters
models = [
    AutoARIMA(season_length=24),
]
# Instantiate StatsForecast class as sf
sf = StatsForecast(
    df=Y_df, 
    models=models,
    freq='H', 
    n_jobs=-1,
    verbose=True
)
sf.fit()
fitted_model = sf.fitted_
# save the models with pickle

Hi @Akmal Soliev, I am trying to load an already trained MLForecast model for reuse. Is there anyway out for this ? will this "fitted_" method work for MLForecast as well ?

Hi @Akmal Soliev, Can you clarify your code please? In step one, do you do something with the "fitted_model"? Do you save it as a pickle? Could you please complete your code if there something missing? In step 2, you put "test_models", where does it come from? I am trying, to save a model in a MLOps pipeline, but I can't figure how to do it. Last question, does it work to save a model that we used with cross_validation? Thank you for your help and clarification.

Hey @Nasreddine D it is an array, hence, you can save it as a pickle if you want

In step 2, you put "test_models", where does it come from?

from the saved file, let's say pickle

Last question, does it work to save a model that we used with cross_validation?

Haven't played with that, hence, cannot answer.