Insertion-based Decoding with Automatically Inferred Generation Order

Citations Over TimeTop 10% of 2019 papers

Abstract

Conventional neural autoregressive decoding commonly assumes a fixed left-to-right generation order, which may be sub-optimal. In this work, we propose a novel decoding algorithm— InDIGO—which supports flexible sequence generation in arbitrary orders through insertion operations. We extend Transformer, a state-of-the-art sequence generation model, to efficiently implement the proposed approach, enabling it to be trained with either a pre-defined generation order or adaptive orders obtained from beam-search. Experiments on four real-world tasks, including word order recovery, machine translation, image caption, and code generation, demonstrate that our algorithm can generate sequences following arbitrary orders, while achieving competitive or even better performance compared with the conventional left-to-right generation. The generated sequences show that InDIGO adopts adaptive generation orders based on input information.

Related Papers

• → Tied Transformers: Neural Machine Translation with Shared Encoder and Decoder(2019)65 cited
• → Neural Machine Translation with the Transformer and Multi-Source Romance Languages for the Biomedical WMT 2018 task(2018)15 cited
• → Searching Better Architectures for Neural Machine Translation(2020)27 cited
• → Incorporating Pre-trained Model into Neural Machine Translation(2021)2 cited
• → On The Alignment Problem In Multi-Head Attention-Based Neural Machine\n Translation(2018)