I see that in the dataset the answers are generated by "gpt" while the readme says that they are generated by BLIP. Which one is the actual label generator?
· Sign up or log in to comment