After few training pipeline Ml package giving 100% confidence for wrong field , why this happening
Please help
did you retrain on the base model on the trained model?
ideally we need to retrain all the docs again on base that is preferred to get proper results
cheers
yes i did retraining on base model only
did you happen to train the old documents also or only the new ones?
check the labellign properly again
cheers
i did train old and new both