MultiWOZ_Evaluation
MultiWOZ_Evaluation copied to clipboard
Tomiinek
Unified MultiWOZ evaluation scripts for the context-to-response task.
We aim to conduct DST evaluation on the MultiWOZ 2.4 corpus. This PR shows our proposed extension to the existing code to achieve this.
This is more a question than a bug or something else, I have seen the documentation (The `Readme.md` file) and the codes, but I can't understand how to evaluate the...
I am testing the UBAR method in mwoz2.2 and created a parser to convert back the generated sequence into a state dict. First I loaded groundtruth data with datasets and...
