emre/TARA_Turkish_LLM_Benchmark
收藏Hugging Face2025-04-10 更新2025-04-12 收录
下载链接:
https://hf-mirror.com/datasets/emre/TARA_Turkish_LLM_Benchmark
下载链接
链接失效反馈官方服务:
资源简介:
TARA (土耳其高级推理评估) 数据集是一个用于评估土耳其语大型语言模型(LLM)高级推理能力的基准数据集。它包含10个不同领域的问题,每个问题都有从1到10的详细难度级别。数据集的目的是测试LLM的高级认知技能,如逻辑推理、问题解决、分析、评估和创造性思维。
TARA (Turkish Advanced Reasoning Assessment) is a benchmark dataset for evaluating the advanced reasoning capabilities of Turkish Large Language Models (LLMs). It includes questions across 10 different domains, each with detailed difficulty levels from 1 to 10. The dataset aims to test LLMs higher-order cognitive skills such as logical inference, problem-solving, analysis, evaluation, and creative thinking.
提供机构:
emre



