This message was deleted.

Hi. After you fit the models they're available on the

MLForecast.models_

dict, so in your particular example I believe you can use

fcst.models_['LGBMRegressor']

and can access the attributes like feature importance

thanks - is this library a wrapper around sklearn library ?

It leverages the sklearn api to train the models and compute predictions

thank you @José Morales 🙂 for your quick messages and support