You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The mlflow version requirement is <= 1.30.0 now, which restrict pandas' version to be smaller than 2. This will cause bad performance, for example, assigning values to multiindex columns (because by default qlib uses ['feature', 'label'] as the first level of columns, and there will be many assignment operations when preprocessing data. I have tested that for 1,000,000 rows and 100 columns of features, it can have a 50-100x speed up.
The text was updated successfully, but these errors were encountered:
🌟 Feature Description
Update the mlflow version to be higher than 2!!!
Motivation
The mlflow version requirement is <= 1.30.0 now, which restrict pandas' version to be smaller than 2. This will cause bad performance, for example, assigning values to multiindex columns (because by default qlib uses ['feature', 'label'] as the first level of columns, and there will be many assignment operations when preprocessing data. I have tested that for 1,000,000 rows and 100 columns of features, it can have a 50-100x speed up.
The text was updated successfully, but these errors were encountered: