You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I noticed that the datasets supported in the code are all multiple-choice and classification types, such as IMDB, QNLI, and BoolQ. Can the code in this repository support free-form types of tasks, such as TriviaQA and NQ? For free-form QA tasks, the possible answer space is infinite. How should the CCS method allocate X+ and X- labels for such tasks?
The text was updated successfully, but these errors were encountered:
Free form tasks, imho, can be reduced to a classification tasks by appending question and answer each time like [Original Question] Answer: "[Sampled Answer]" Is the last answer true or false? True / False (contrast pair). And so on for multi-class
I noticed that the datasets supported in the code are all multiple-choice and classification types, such as IMDB, QNLI, and BoolQ. Can the code in this repository support free-form types of tasks, such as TriviaQA and NQ? For free-form QA tasks, the possible answer space is infinite. How should the CCS method allocate X+ and X- labels for such tasks?
The text was updated successfully, but these errors were encountered: