mapto, 1 month ago @j3j5 @blogdiva @KathyReid excuse me, actually at a second thought BERT and GPT are exactly doing corrupt-and-train on the dataset. This certainly weakens my interpretation. Apologies for the confusion.
@j3j5 @blogdiva @KathyReid excuse me, actually at a second thought BERT and GPT are exactly doing corrupt-and-train on the dataset. This certainly weakens my interpretation. Apologies for the confusion.