Hi, say you want to fine tune for Irish vs Australian accents for English. Possibly you have data set striped by accent/speaker. What is the best way to fine tune the model so you do not lose the accent differences and also you do not end up forgetting the original accent used for training for that same language?
Hi, say you want to fine tune for Irish vs Australian accents for English. Possibly you have data set striped by accent/speaker. What is the best way to fine tune the model so you do not lose the accent differences and also you do not end up forgetting the original accent used for training for that same language?