RCZoo
RCZoo copied to clipboard
question answering, reading comprehension toolkit
- change dataset - add rouge function
add RL loss for bert
#36
https://github.com/sogou/SMRCToolkit/
Hello. It seems that the performance of slqa is worse than bidaf(with self-attention?). I wanna know that are there some problems implementation slqa? In my own implementation, it seems also...
how to use the predict script for squad? could you add a readme? and what's the sogou data?
will you update the dataset? or support more dataset?