Dailydialog was created with incorrect min and max lengths and will therefore be regenerated as discussed in the March meeting (#51.) As it has not been urgent, focus has been on building the metrics extraction pipeline.
Dailydialog will be regenerated at some point before the classifiers have been finalised. NB. Remember to also extract metrics for the dataset again when it has been regenerated for all language models.
Regenerating DailyDialog
Dailydialog was created with incorrect min and max lengths and will therefore be regenerated as discussed in the March meeting (#51.) As it has not been urgent, focus has been on building the metrics extraction pipeline.
Dailydialog will be regenerated at some point before the classifiers have been finalised. NB. Remember to also extract metrics for the dataset again when it has been regenerated for all language models.