Skip to content

Latest commit

 

History

History
36 lines (23 loc) · 626 Bytes

README.md

File metadata and controls

36 lines (23 loc) · 626 Bytes

Model Evaluation

After training and predicting the models, you must copy the following files before evaluating the model performance on the validation set(s).

1. Ground-Truth Files

Copy files from

data/processed/[split]/data.csv

and paste them like this.

src/eval/ground_truths/[split].csv

Do this for the val and val_subset splits.

2. Prediction Files

Copy files from

data/processed/[split]/predictions/[model_prediction].csv

and paste them like this.

src/eval/predictions/[split]/[model_prediction].csv

Do this for all predictions in the val and val_subset splits.