DialogRPT-updown
Dialog Ranking Pretrained Transformers
How likely a dialog response is upvoted 👍 and/or gets replied 💬?
This is what DialogRPT is learned to predict. It is a set of dialog response ranking models proposed by Microsoft Research NLP Group trained on 100 + millions of human feedback data. It can be used to improve existing dialog generation model (e.g., DialoGPT) by re-ranking the generated response candidates.
Quick Links:
We considered the following tasks and provided corresponding pretrained models.
| Model card | Description |
|---|---|
| Given a context and its two human responses, predict... | |
microsoft/DialogRPT-updown |
... which gets more upvotes? |
microsoft/DialogRPT-width |
... which gets more direct replies? |
microsoft/DialogRPT-depth |
... which gets longer follow-up thread? |
| Given a context and one human response, distinguish it with... | |
microsoft/DialogRPT-human-vs-rand |
... a random human response |
microsoft/DialogRPT-human-vs-machine |
... a machine generated response |
Examples:
The updown score predicts how likely the response is getting upvoted.
| Context | Response | updown score |
|---|---|---|
| I love NLP! | Here’s a free textbook (URL) in case anyone needs it. | 0.613 |
| I love NLP! | Me too! | 0.111 |
Contact:
Please create an issue on our repo
Citation:
@inproceedings{gao2020dialogrpt,
title={Dialogue Response RankingTraining with Large-Scale Human Feedback Data},
author={Xiang Gao and Yizhe Zhang and Michel Galley and Chris Brockett and Bill Dolan},
year={2020},
booktitle={EMNLP}
}