-
-
Notifications
You must be signed in to change notification settings - Fork 105
Open
Labels
bugSomething isn't workingSomething isn't working
Description
I'm new to this project and while exploring the code in run_forecast, I noticed a few things that might be worth addressing:
The model is loaded within the function every time it's used. This might make processes slow if forecasts are run a lot.
For instance, the line of code below could be placed outside the function so the model is loaded just once.
# load model only once
model = load_model(f"{dir_path}/../models/model-0.3.0.pkl")With respect, as much as it's not exactly a high priority with the lone call in run_eval, importing the model within run_forecast could be a performance issue if the function was repeatedly called, which might be worth optimizing.
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
bugSomething isn't workingSomething isn't working