Open xmshi-trio opened 3 years ago
Hi, I have a question about the data pre-processing. How do you deal with SMM4H-17 task3 dataset? There is no test data in the raw SMM4H-17 dataset.
Hello,
We randomly split the released training data and take the 0.2 of the data as our dev set.
Thanks for your response. Could you provide the TwADR-L and SMM4H-17?
Hi, I have a question about the data pre-processing. How do you deal with SMM4H-17 task3 dataset? There is no test data in the raw SMM4H-17 dataset.