Qdrant/arxiv-titles-instructorxl-embeddings
收藏Hugging Face2023-11-03 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/Qdrant/arxiv-titles-instructorxl-embeddings
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含使用InstructorXL模型从arXiv论文标题生成的768维嵌入向量。每个向量都附带有用于生成它的摘要和DOI(数字对象标识符)。数据集是通过Alexandria Index提供的预计算嵌入生成的。生成嵌入时使用了特定的指令,并提供了生成嵌入的代码示例。
This dataset contains 768-dimensional embedding vectors generated from arXiv paper titles using the InstructorXL model. Each vector is accompanied by the abstract and DOI (Digital Object Identifier) used for its generation. The dataset is generated from pre-computed embeddings provided via the Alexandria Index. Specific instructions were utilized during the embedding generation process, and code examples for generating the embeddings are provided.
提供机构:
Qdrant
原始信息汇总
arxiv-titles-instructorxl-embeddings
概述
- 语言: 英语
- 名称: InstructorXL embeddings of the Arxiv.org titles
- 任务类别:
- 句子相似度
- 特征提取
- 大小类别: 1M<n<10M
数据集描述
- 维度: 768维
- 来源: 使用InstructorXL模型从arxiv论文标题生成的嵌入向量。
- 内容: 每个向量包含用于创建它的摘要以及DOI(数字对象标识符)。
- 生成方式: 使用预计算的嵌入,由Alexandria Index提供。
生成过程
-
指令: text Represent the Research Paper title for retrieval; Input:
-
代码示例: python from InstructorEmbedding import INSTRUCTOR
model = INSTRUCTOR(hkunlp/instructor-xl) sentence = "3D ActionSLAM: wearable person tracking in multi-floor environments" instruction = "Represent the Research Paper title for retrieval; Input:" embeddings = model.encode([[instruction, sentence]])



