https://github.com/nixtla logo
m

Manuel

05/24/2023, 6:52 PM
Is there an easy way to use an already fitted neuralforecast model as "fallback_model" in statsforecast? @Kin Gtz. Olivares
f

fede (nixtla) (they/them)

05/24/2023, 8:10 PM
hey @Manuel! We don’t have a
fallback_model
option for neuralforecast yet. Could you help us open an issue to work on that feature? Thank you! https://github.com/Nixtla/neuralforecast/issues/new?assignees=&labels=&projects=&template=feature_request.md&title=
m

marah othman

05/24/2023, 8:31 PM
what is the fallback?
m

Manuel

05/25/2023, 8:33 AM
Hi @fede! My question was about having a neuralforecast model as a fallback of a statsforecast model (so a mix of statsforecast and neuralforecast). In general I think any combination could be useful (stats/stats, stats/neural, neural/neural, neural/stats).