baidu/Dialogue
Baidu's research implementations for multi-turn dialogue systems using deep attention matching networks and seq2seq models.

Velocity · 7d
+0.2
★ / day
Trend
→steady
star history
This repository contains Baidu NLP team’s work on open-domain conversational systems, including multi-turn response selection models with deep attention matching networks, knowledge-driven dialogue approaches, and reinforcement learning for life-long learning conversational agents. It incorporates ERNIE pre-trained models and is part of the PaddlePaddle/PaddleNLP ecosystem for dialogue research.