Saved in:
Bibliographic Details
Main Authors: Ren, Houxing, Zhan, Mingjie, Wu, Zhongyuan, Li, Hongsheng
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.17103
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913391100035072
author Ren, Houxing
Zhan, Mingjie
Wu, Zhongyuan
Li, Hongsheng
author_facet Ren, Houxing
Zhan, Mingjie
Wu, Zhongyuan
Li, Hongsheng
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2405_17103
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Empowering Character-level Text Infilling by Eliminating Sub-Tokens
Ren, Houxing
Zhan, Mingjie
Wu, Zhongyuan
Li, Hongsheng
Computation and Language
Artificial Intelligence
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.
title Empowering Character-level Text Infilling by Eliminating Sub-Tokens
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2405.17103