PreSumm
PreSumm copied to clipboard
What was your loss for both the abstractive and extractive models on the validation and test sets?
I couldn't find this in your paper or in the github.
Thanks much!
#41 said that xent is loss