DrBenchmark

DrBenchmark是首个公开可用的法语生物医学语言理解基准，由阿维尼翁大学LIA实验室创建。该数据集包含20个多样化的任务，如命名实体识别、词性标注、问答、语义文本相似性和分类等。数据来源于科学文献、临床试验、临床案例等多种类型，旨在评估和比较不同预训练语言模型在生物医学领域的性能，解决模型评估和选择的问题。

DrBenchmark is the first publicly available French biomedical language understanding benchmark, created by the LIA Laboratory of Avignon University. This dataset comprises 20 diverse tasks, including named entity recognition, part-of-speech tagging, question answering, semantic textual similarity, classification, and more. The data is sourced from multiple types of materials such as scientific literature, clinical trials, and clinical cases. It aims to evaluate and compare the performance of different pre-trained language models in the biomedical domain, addressing the challenges of model evaluation and selection.