PLATO: Pre-trained Dialogue Generation Model with Discrete Latent Variable

Code

• github.com/PaddlePaddle/Research
  OfficialIn paperpaddle★ 1,753
• github.com/egillian1/AblationPLATO
  paddle★ 3
• github.com/PaddlePaddle/PaddleNLP/tree/develop/examples/dialogue/plato-2
  paddle★ 0

Abstract

Pre-training models have been proved effective for a wide range of natural language processing tasks. Inspired by this, we propose a novel dialogue generation pre-training framework to support various kinds of conversations, including chit-chat, knowledge grounded dialogues, and conversational question answering. In this framework, we adopt flexible attention mechanisms to fully leverage the bi-directional context and the uni-directional characteristic of language generation. We also introduce discrete latent variables to tackle the inherent one-to-many mapping problem in response generation. Two reciprocal tasks of response generation and latent act recognition are designed and carried out simultaneously within a shared network. Comprehensive experiments on three publicly available datasets verify the effectiveness and superiority of the proposed framework.

Tasks

Reproductions