Lots-of-LoRAs/task828_copa_commonsense_cause_effect

该数据集名为task828_copa_commonsense_cause_effect，属于文本生成任务类别，主要用于常识推理中的因果关系分析。数据集包含796个训练样本、100个验证样本和100个测试样本。每个样本包含输入、输出和ID三个特征。数据集的创建者和语言创建者均为众包，语言为英语，许可证为Apache-2.0。数据集的相关论文包括《Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks》和《Compress then Serve: Serving Thousands of LoRA Adapters with Little Overhead》。

The dataset, named task828_copa_commonsense_cause_effect, belongs to the text generation task category and is primarily used for causal reasoning in commonsense analysis. It contains 796 training samples, 100 validation samples, and 100 test samples. Each sample includes three features: input, output, and ID. The datasets creators and language creators are crowdsourced, the language is English, and the license is Apache-2.0. Relevant papers for the dataset include Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks and Compress then Serve: Serving Thousands of LoRA Adapters with Little Overhead.

原始信息汇总

数据集概述

基本信息

• 数据集名称: task828_copa_commonsense_cause_effect
• 任务类别: 文本生成
• 语言: 英语
• 许可证: Apache 2.0
• 数据来源: 众包

数据集结构

• 配置名称: plain_text
• 特征:
  • 输入: 字符串
  • 输出: 字符串
  • ID: 字符串

数据分割

• 训练集: 796个样本
• 验证集: 100个样本
• 测试集: 100个样本