Lots-of-LoRAs/task020_mctaco_span_based_question
收藏Hugging Face2025-01-01 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/Lots-of-LoRAs/task020_mctaco_span_based_question
下载链接
链接失效反馈官方服务：
资源简介：
task020_mctaco_span_based_question数据集是一个文本生成任务的数据集，包含输入文本、输出文本和唯一标识符三个特征。数据集由众包方式创建，语言为英语。数据集分为训练集、验证集和测试集，分别包含304、38和39个样本。该数据集用于训练和评估模型在基于span的问答任务上的性能。
The task020_mctaco_span_based_question dataset is a text generation task dataset, which includes three features: input text, output text, and unique identifier. The dataset is created by crowdsourcing, with English as the language. The dataset is split into training, validation, and test sets, containing 304, 38, and 39 samples respectively. This dataset is used to train and evaluate the performance of models on span-based question answering tasks.
提供机构：
Lots-of-LoRAs
