Remove redundant models for bottom level

pratham16cse / AggForecaster

Code for "Coherent Probabilistic Aggregate Queries on Long-horizon Forecasts", IJCAI 2022

https://arxiv.org/pdf/2111.03394v1.pdf

18 stars 4 forks source link

Closed pratham16cse closed 3 years ago

pratham16cse commented 3 years ago

If two different inference models share a bottom level model, do not retrain, use the existing bottom-level model.

pratham16cse commented 3 years ago

Added this in the main file.
In the if condition, need to manually add any future models that are equivalent at the K=1 level.