arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2408.04632cs.CLcs.CV
收藏

Arctic-TILT:亚十亿参数规模的商业文档理解

Arctic-TILT. Business Document Understanding at Sub-Billion Scale

Jagiellonian University(雅盖隆大学);Warsaw University of Technology(华沙理工大学);Adam Mickiewicz University(亚当·密茨凯维奇大学)
Łukasz Borchmann, Michał Pietruszka, Wojciech Jaśkowski, Dawid Jurkiewicz, Piotr Halama, Paweł Józiak, Łukasz Garncarek, Paweł Liskowski, Karolina Szyndler, And… 展开作者

Łukasz Borchmann, Michał Pietruszka, Wojciech Jaśkowski, Dawid Jurkiewicz, Piotr Halama, Paweł Józiak, Łukasz Garncarek, Paweł Liskowski, Karolina Szyndler, Andrzej Gretkowski, Julita Ołtusek, Gabriela Nowakowska, Artur Zawłocki, Łukasz Duhr, Paweł Dyda, Michał Turski

AI总结:

Arctic-TILT 是一个亚十亿参数模型,可在单张 24GB GPU 上微调部署,以极低成本实现与千倍规模模型相当的文档理解准确率,并在七个基准上达到最先进水平。

摘要:

使用大语言模型(LLM)的工作负载中,很大一部分涉及回答基于 PDF 或扫描内容的问题。我们提出了 Arctic-TILT,在这些用例上达到了与规模为其 1000 倍的模型相当的准确率。它可以在单个 24GB GPU 上微调和部署,在降低运营成本的同时,处理最多 400k token 的视觉丰富文档(Visually Rich Documents)。该模型在七个不同的文档理解基准上取得了最先进(state-of-the-art)的结果,并提供可靠的置信度分数和快速推理,这对于在大规模或时间敏感的企业环境中处理文件至关重要。

英文摘要:

The vast portion of workloads employing LLMs involves answering questions grounded on PDF or scan content. We introduce the Arctic-TILT achieving accuracy on par with models 1000$\times$ its size on these use cases. It can be fine-tuned and deployed on a single 24GB GPU, lowering operational costs while processing Visually Rich Documents with up to 400k tokens. The model establishes state-of-the-art results on seven diverse Document Understanding benchmarks, as well as provides reliable confidence scores and quick inference, which are essential for processing files in large-scale or time-sensitive enterprise environments.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑