tmnam20/ViGLUE

该数据集包含多个配置，每个配置都有不同的特征和数据分割。例如，cola配置包含句子、标签和索引字段，标签字段表示句子是否可接受；mnli配置包含前提、假设和标签字段，标签字段表示前提和假设之间的关系；mrpc配置包含两个句子和标签字段，标签字段表示两个句子是否等价；qnli配置包含问题、句子和标签字段，标签字段表示问题和句子之间的关系；qqp配置包含两个问题和标签字段，标签字段表示两个问题是否重复；rte配置包含两个句子和标签字段，标签字段表示两个句子之间的关系；sst2配置包含句子和标签字段，标签字段表示句子的情感极性；vnrte配置包含两个句子、主题、来源和标签字段，标签字段表示两个句子之间的关系；vsfc配置包含句子和标签字段，标签字段表示句子的情感极性；vsmec配置包含句子、原始句子、情感和标签字段，标签字段表示句子的情感类别；vtoc配置包含句子和标签字段，标签字段表示句子的主题类别；wnli配置包含两个句子和标签字段，标签字段表示两个句子之间的关系。

The dataset contains multiple configurations, each with different features and data splits. For example, the cola configuration includes sentence, label, and index fields, where the label field indicates whether the sentence is acceptable; the mnli configuration includes premise, hypothesis, and label fields, where the label field indicates the relationship between the premise and hypothesis; the mrpc configuration includes two sentences and a label field, where the label field indicates whether the two sentences are equivalent; the qnli configuration includes question, sentence, and label fields, where the label field indicates the relationship between the question and sentence; the qqp configuration includes two questions and a label field, where the label field indicates whether the two questions are duplicates; the rte configuration includes two sentences and a label field, where the label field indicates the relationship between the two sentences; the sst2 configuration includes sentence and label fields, where the label field indicates the sentiment polarity of the sentence; the vnrte configuration includes two sentences, topic, source, and label fields, where the label field indicates the relationship between the two sentences; the vsfc configuration includes sentence and label fields, where the label field indicates the sentiment polarity of the sentence; the vsmec configuration includes sentence, raw sentence, emotion, and label fields, where the label field indicates the emotion category of the sentence; the vtoc configuration includes sentence and label fields, where the label field indicates the topic category of the sentence; the wnli configuration includes two sentences and a label field, where the label field indicates the relationship between the two sentences.