Empowering Character-level Text Infilling by Eliminating Sub-Tokens

TL;DR

This paper introduces FIM-SE, a novel method for character-level text infilling that eliminates sub-token prediction during inference, significantly improving performance over previous approaches.

Contribution

The paper proposes FIM-SE, a line-level infilling approach with special tokens, to effectively perform character-level infilling without sub-token prediction, enhancing accuracy and guidance.

Findings

FIM-SE outperforms previous methods in character-level infilling tasks.

The line-level format reduces the perplexity associated with sub-token prediction.

Incorporating special tokens improves generation guidance and overall performance.

Abstract

In infilling tasks, sub-tokens, representing instances where a complete token is segmented into two parts, often emerge at the boundaries of prefixes, middles, and suffixes. Traditional methods focused on training models at the token level, leading to sub-optimal performance in character-level infilling tasks during the inference stage. Alternately, some approaches considered character-level infilling, but they relied on predicting sub-tokens in inference, yet this strategy diminished ability in character-level infilling tasks due to the large perplexity of the model on sub-tokens. In this paper, we introduce FIM-SE, which stands for Fill-In-the-Middle with both Starting and Ending character constraints. The proposed method addresses character-level infilling tasks by utilizing a line-level format to avoid predicting any sub-token in inference. In addition, we incorporate two special tokens to signify the rest of the incomplete lines, thereby enhancing generation guidance. Extensive experiments demonstrate that our proposed approach surpasses previous methods, offering a significant advantage. Code is available at https://github.com/SenseLLM/FIM-SE.