PLATO-2: Towards Building an Open-Domain Chatbot via Curriculum Learning
Siqi Bao, He Huang, Fan Wang, Hua Ren Wu, Haifeng Wang, Wenquan Wu, Zhen Guo, Zhibin Liu, Xinchao Xu · 2021
To build a high-quality open-domain chatbot, we introduce the effective training process of PLATO-2 via curriculum learning.There are two stages involved in the learning process.In the first stage, a coarse-grained generation model is trained to learn response generation under the simplified framework of oneto-one mapping.In the second stage, a finegrained generative model augmented with latent variables and an evaluation model are further trained to generate diverse responses and to select the best response, respectively.PLATO-2 was trained on both Chinese and English data, whose effectiveness and superiority are verified through comprehensive evaluations, achieving new state-of-the-art results.* Equal contribution.Stage 1 Coarse-grained Generation One-to-One Mapping General Response Generation Stage 2.1 Fine-grained Generation One-to-Many Mapping Diverse Response Generation Stage 2.2 Evaluation