通过EOS标记加权控制摘要长度
计算与语言
2025-06-06 v1 机器学习
摘要
控制生成文本的长度在各种文本生成任务中尤为重要，包括摘要生成。现有方法通常需要复杂的模型修改，限制了与预训练模型的兼容性。我们通过增加正确预测EOS标记在交叉熵损失计算中的重要性，开发了一种简单方法来控制自动文本摘要的长度。proposed methodology is agnostic to architecture and decoding algorithms and orthogonal to other inference-time techniques to control the generation length. We tested it with encoder-decoder and modern GPT-style LLMs, and show that this method can control generation length, often without affecting the quality of the summary.
引用
@article{arxiv.2506.05017,
title = {Controlling Summarization Length Through EOS Token Weighting},
author = {Zeno Belligoli and Emmanouil Stergiadis and Eran Fainman and Ilya Gusev},
journal= {arXiv preprint arXiv:2506.05017},
year = {2025}
}