Firstly, I want to express my appreciation for the fantastic work you've done!
I have a couple of questions regarding the data processing part:
In the code, I noticed that the original data labels are not transformed from a single integer to an integer sequence. Could you confirm whether this behavior aligns with the expected processing?
Regarding the training datasets, wow_train_augmented and wow_dev_unseen_augmented: Are these datasets generated by mixing the original dataset with two neg_datasets, followed by a proportional split?
I appreciate your time and assistance in clarifying these points.
Hi there,
Firstly, I want to express my appreciation for the fantastic work you've done!
I have a couple of questions regarding the data processing part:
In the code, I noticed that the original data labels are not transformed from a single integer to an integer sequence. Could you confirm whether this behavior aligns with the expected processing?
Regarding the training datasets, wow_train_augmented and wow_dev_unseen_augmented: Are these datasets generated by mixing the original dataset with two neg_datasets, followed by a proportional split?
I appreciate your time and assistance in clarifying these points.