lowry02/hs-real-case
收藏官方服务：
资源简介：
该数据集包含句子在模型中的隐藏状态表示，具体特征包括基准行索引、句子文本、层索引、词元位置、词元ID、词元字符串以及隐藏状态向量。数据分片为hidden_states，总大小约为850MB，包含196,704个示例，适用于自然语言处理中的模型分析或特征提取任务。
This dataset contains hidden state representations of sentences in a model, with features including benchmark row index, sentence text, layer index, token position, token ID, token string, and hidden state vectors. The split is named hidden_states, with a total size of approximately 850MB and 196,704 examples, suitable for model analysis or feature extraction tasks in natural language processing.
提供机构：
lowry02
