AgriEval
收藏资源简介:
AgriEval 是一个全面的中文农业领域语言模型基准,包含 16664 条数据,覆盖了农业六大类和 29 个子类别。数据来源于大学级别的考试和作业,旨在评估大型语言模型在农业领域的应用能力,包括记忆、理解、推理和生成等方面。该数据集具有高质量的数据,多样化的格式和广泛的规模,是迄今为止最广泛的农业基准。AgriEval 可用于评估语言模型在农业领域的性能,为开发农业特定的大型语言模型提供有价值的见解。
AgriEval is a comprehensive Chinese language model benchmark focused on the agricultural domain. It contains 16,664 data entries, covering 6 major agricultural categories and 29 subcategories. Sourced from university-level examinations and assignments, this benchmark aims to evaluate the application capabilities of large language models in the agricultural domain, including memory, comprehension, reasoning, and generation capabilities. Boasting high-quality data, diverse formats, and a substantial scale, AgriEval stands as the most extensive agricultural benchmark to date. It can be used to assess the performance of language models in the agricultural domain, providing valuable insights for the development of agricultural-specific large language models.

- 1AgriEval: A Comprehensive Chinese Agricultural Benchmark for Large Language Models哈尔滨工业大学 · 2025年



