barandinho/wikipedia_tr_embedded

该数据集是基于土耳其语维基百科数据集[barandinho/wikipedia_tr](https://huggingface.co/datasets/barandinho/wikipedia_tr)进行分块处理（chunk_size=2048, chunk_overlap=256），然后通过[sentence-transformers/distiluse-base-multilingual-cased-v1](https://huggingface.co/sentence-transformers/distiluse-base-multilingual-cased-v1)嵌入模型进行嵌入处理。嵌入维度为64，值为int8类型。该数据集适用于进行相似性搜索，特别是针对土耳其语文本。

This dataset is an embedded version of the Turkish Wikipedia, originally from barandinho/wikipedia_tr, which was chunked (chunk_size=2048, chunk_overlap=256) and then embedded using the sentence-transformers/distiluse-base-multilingual-cased-v1 model. The embedding dimension is 64 with int8 values. The dataset includes features such as url, title, text, and embed_int8 (embedded int8 sequences), divided into a training set with 762059 samples.

原始信息汇总

数据集概述

数据集信息

• 语言: 土耳其语
• 嵌入模型: sentence-transformers/distiluse-base-multilingual-cased-v1
• 嵌入维度: 64
• 嵌入值类型: int8

数据集特征

• url: 字符串类型
• title: 字符串类型
• text: 字符串类型
• embed_int8: 序列类型，int64

数据集分割

• train:
  • 样本数量: 762059
  • 字节数: 1252396519

数据集大小

• 下载大小: 501996398
• 数据集大小: 1252396519

配置

• config_name: default
• data_files:
  • split: train
  • path: data/train-*