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
Describe the bug
Functions: make_critical_intent_report(), update_confidence_scores() in test.py requires additional column intent_pred in predictions_df , instead they should retrieve actual model predictions from intent column:
predictions.append(
{
"data_id": row["data_id"],
const.INTENT: intents[0][const.NAME] if intents else "_no_preds_",
const.SCORE: intents[0][const.SCORE] if intents else 0,
}
)
To Reproduce
use s3://plute-data-store/keshav-onboarding/test_small.csv , it has 3 columns alternatives, tag, data_id
Describe the bug
Functions:
make_critical_intent_report()
,update_confidence_scores()
in test.py requires additional columnintent_pred
in predictions_df , instead they should retrieve actual model predictions fromintent
column:To Reproduce
Expected behavior
saves intent error analysis results inside metrics folder
Screenshots
Desktop
The text was updated successfully, but these errors were encountered: