arXiv cs.CL· Jannis Vamvas, Rico Sennrich·· 10 小时前AI 评分22
基于参考聚合的线性时间最小贝叶斯风险解码
Linear-time Minimum Bayes Risk Decoding with Reference Aggregation
AI 导读
该论文提出用与聚合参考表示计算的分数近似成对指标分数，将最小贝叶斯风险（MBR）解码的效用估计复杂度从 O(n²) 降至 O(n)，同时实验上保留了 MBR 解码的大部分质量增益。该方法仍基于采样近似，但不再需要成对计算效用指标。源代码已公开，论文发表于 ACL 2024。
正文 · 原文
Abstract:Minimum Bayes Risk (MBR) decoding is a text generation technique that has been shown to improve the quality of machine translations, but is expensive, even if a sampling-based approximation is used. Besides requiring a large number of sampled sequences, it requires the pairwise calculation of a utility metric, which has quadratic complexity. In this paper, we propose to approximate pairwise metric scores with scores calculated against aggregated reference representations. This changes the complexity of utility estimation from $O(n^2)$ to $O(n)$, while empirically preserving most of the quality gains of MBR decoding. We release our source code at this https URL
| Comments: | ACL 2024. v3: More accurate time measurements for ChrF |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2402.04251 [cs.CL] |
| (or arXiv:2402.04251v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2402.04251 arXiv-issued DOI via DataCite |
Submission history
From: Jannis Vamvas [view email]
[v1]
Tue, 6 Feb 2024 18:59:30 UTC (47 KB)
[v2]
Mon, 3 Jun 2024 09:33:14 UTC (49 KB)
[v3]
Tue, 10 Nov 2026 08:05:20 UTC (45 KB)
来源：arXiv cs.CL · arxiv.org