MT010104 opened this issue 2 years ago · 0 comments
Hi, thanks for the great work. Most of the models are evaluated in a zero-shot manner. I really wanna know whether the dataset can be used for fine-tuning.