Skip to main content
QUICK REVIEW

[论文解读] Towards Building Multilingual Language Model for Medicine

Pengcheng Qiu, Chaoyi Wu|arXiv (Cornell University)|Feb 21, 2024
Biomedical Text Mining and Ontologies被引用 6
一句话总结

论文引入了 MMedC,一个 25.5B 标记的多语种医疗语料库,一个评估基准 MMedBench,以及开源多语种医疗大模型(MMedLM/MMedLM 2),在 MMedBench 上的表现强劲,堪比 GPT-4。

ABSTRACT

The development of open-source, multilingual medical language models can benefit a wide, linguistically diverse audience from different regions. To promote this domain, we present contributions from the following: First, we construct a multilingual medical corpus, containing approximately 25.5B tokens encompassing 6 main languages, termed as MMedC, enabling auto-regressive domain adaptation for general LLMs; Second, to monitor the development of multilingual medical LLMs, we propose a multilingual medical multi-choice question-answering benchmark with rationale, termed as MMedBench; Third, we have assessed a number of open-source large language models (LLMs) on our benchmark, along with those further auto-regressive trained on MMedC. Our final model, MMed-Llama 3, with only 8B parameters, achieves superior performance compared to all other open-source models on both MMedBench and English benchmarks, even rivaling GPT-4. In conclusion, in this work, we present a large-scale corpus, a benchmark and a series of models to support the development of multilingual medical LLMs.

研究动机与目标

  • 开发一个面向语言多样受众的开源多语种医疗语言模型。
  • 构建一个覆盖六种语言的自回归训练大规模多语种医疗语料库(MMedC)。
  • 创建一个多语种医疗问答基准(MMedBench),并给出评估多语种医疗推理的理由。
  • 评估开源大语言模型以及在 MMedC 上训练的模型,以评估多语种医疗问答与推理生成能力。

提出的方法

  • 从四个数据源汇集超过 25.5B 标记,涵盖英文、中文、日文、法语、俄语和西班牙语。
  • 通过聚合多语种医疗选择题 QA 数据集并用 GPT-4 生成的推理来丰富 MMedBench。
  • 在零-shot、PEFT 以及全量微调等设置下,对包括在 MMedC 上训练的模型(MMedLM/MMedLM 2)在内的一系列 LLM 进行微调和/或评估。
  • 以自动化指标(BLEU-1、ROUGE-1、BERT-score)和人工评估来评估推理质量,以确定可靠的评估方法。
Figure 1 : Overview of our contributions. Figure a demonstrates our proposed large-scale multilingual medical corpus (MMedC), containing 25.5B tokens, covering six main language, collected from four data sources. Figure b shows the composition our comprehensive multilingual medical benchmark (MMedBe
Figure 1 : Overview of our contributions. Figure a demonstrates our proposed large-scale multilingual medical corpus (MMedC), containing 25.5B tokens, covering six main language, collected from four data sources. Figure b shows the composition our comprehensive multilingual medical benchmark (MMedBe

实验结果

研究问题

  • RQ1一个多语种、医学聚焦的语料库是否可以提升开源 LLM 对非英语医学查询的表现?
  • RQ2在 MMedC 上进行训练对六种语言的多语种医疗问答和推理生成有何影响?
  • RQ3哪些评估指标能够最好地反映多语种 LLM 的医学推理的人工判断?
  • RQ4开源多语种医疗 LLM 与封闭源模型在 MMedBench 上的相对表现如何?

主要发现

  • 在多种设置下,经 MMedC 训练的模型(MMedLM、MMedLM 2)优于基线并在 MMedBench 上与 GPT-4 相当。
  • 在六种语言中,MMedLM 2 在多语言全量微调下的准确率从英文 58.13 到西班牙语 80.01 不等,平均在多语言轨道上达到 67.30。
  • 推理生成受益于在 MMedC 上的训练,BLEU-1 和 ROUGE-1 分数以及人工评价均对 MMedLM 2 表现有利。
  • ROUGE-1 和 BLEU-1 被认定为评估 MMedBench 推理的可靠自动指标,GPT-4 的评分与人工判断的相关性最高,但并不易于大规模扩展。
Figure 2 : Statistic results on MMedC. Figure (a) shows our collected corpora can cover most main countries worldwide. Figure (b) shows the detail token number for different languages and Figure (c) shows how the four considered data sources contributes for different languages, i.e. , filtering cont
Figure 2 : Statistic results on MMedC. Figure (a) shows our collected corpora can cover most main countries worldwide. Figure (b) shows the detail token number for different languages and Figure (c) shows how the four considered data sources contributes for different languages, i.e. , filtering cont

更好的研究,从现在开始

从论文设计到论文写作,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。