Kowsher/multitask_textqa_benchmark
收藏Hugging Face2025-10-30 更新2025-11-15 收录
下载链接：
https://hf-mirror.com/datasets/Kowsher/multitask_textqa_benchmark
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含问题、提示、答案和相关元数据，分为训练集和多个测试集，用于不同的问答任务，如ARC挑战、简单ARC、BoolQ、HellaSwag、OpenBookQA、PiQA、Social I-QA和Winogrande等。每个测试集都专注于特定类型的问题，这些问题和答案是用于训练和评估机器学习模型在自然语言理解和推理方面的性能。
The dataset includes questions, prompts, answers, and related metadata, divided into a training set and multiple test sets for different question-answering tasks such as ARC Challenge, ARC Easy, BoolQ, HellaSwag, OpenBookQA, PiQA, Social I-QA, and Winogrande. Each test set focuses on a specific type of question, which are used to train and evaluate machine learning models on their performance in natural language understanding and reasoning.
提供机构：
Kowsher
