mariosasko/glue

GLUE（通用语言理解评估基准）是一个用于训练、评估和分析自然语言理解系统的资源集合。它包含多个任务，如文本分类、自然语言推理和语义相似性评分。数据集为单语（英语），并遵循CC-BY-4.0许可。每个任务都有详细的配置，包括数据字段、分割和示例。

原始信息汇总

数据集概述

名称: GLUE (General Language Understanding Evaluation benchmark)

语言: 英语 (en)

许可证: CC-BY-4.0

多语言性: 单语种

大小类别: 10K<n<100K

源数据集: 原始

任务类别: 文本分类

任务ID:

• acceptability-classification
• natural-language-inference
• semantic-similarity-scoring
• sentiment-classification
• text-scoring

配置:

• ax
• cola
• mnli
• mnli_matched
• mnli_mismatched
• mrpc
• qnli
• qqp
• rte
• sst2
• stsb
• wnli

数据集结构

数据实例

• ax:

  • 特征: premise, hypothesis, label, idx
  • 示例: {"premise": "The cat sat on the mat.", "hypothesis": "The cat did not sit on the mat.", "label": -1, "idx": 0}

• cola:

  • 特征: sentence, label, idx
  • 示例: {"sentence": "Our friends wont buy this analysis, let alone the next one we propose.", "label": 1, "id": 0}

• mnli:

  • 特征: premise, hypothesis, label, idx
  • 示例: {"premise": "Conceptually cream skimming has two basic dimensions - product and geography.", "hypothesis": "Product and geography are what make cream skimming work.", "label": 1, "idx": 0}

• mnli_matched:

  • 特征: premise, hypothesis, label, idx
  • 示例: {"premise": "Hierbas, ans seco, ans dulce, and frigola are just a few names worth keeping a look-out for.", "hypothesis": "Hierbas is a name worth looking out for.", "label": -1, "idx": 0}

• mnli_mismatched:

  • 特征: premise, hypothesis, label, idx
  • 示例: {"premise": "What have you decided, what are you going to do?", "hypothesis": "So whats your decision?"}

数据字段

• ax:

  • premise: 字符串
  • hypothesis: 字符串
  • label: 类别标签
  • idx: int32

• cola:

  • sentence: 字符串
  • label: 类别标签
  • idx: int32

• mnli:

  • premise: 字符串
  • hypothesis: 字符串
  • label: 类别标签
  • idx: int32

• mnli_matched:

  • premise: 字符串
  • hypothesis: 字符串
  • label: 类别标签
  • idx: int32

• mnli_mismatched:

  • premise: 字符串
  • hypothesis: 字符串
  • label: 类别标签
  • idx: int32

数据分割

• ax:

  • 测试: 1104个示例

• cola:

  • 测试: 1063个示例
  • 训练: 8551个示例
  • 验证: 1043个示例

• mnli:

  • 测试_匹配: 9796个示例
  • 测试_不匹配: 9847个示例
  • 训练: 392702个示例
  • 验证_匹配: 9815个示例
  • 验证_不匹配: 9832个示例

• mnli_matched:

  • 测试: 9796个示例
  • 验证: 9815个示例

• mnli_mismatched:

  • 测试: 9847个示例
  • 验证: 9832个示例