items_review_sample
收藏资源简介:
该数据集是一个结构化文本数据集,包含20,000个样本,划分为训练集(16,000个样本)、验证集(2,000个样本)和测试集(2,000个样本)。每个样本包含8个字段:标题(title,字符串类型)、类别(category,字符串类型)、价格(price,浮点数类型)、完整内容(full,字符串类型)、权重(weight,浮点数类型)、摘要(summary,当前为空值)、提示(prompt,当前为空值)和标识符(id,当前为空值)。数据集总大小约为32.8MB,下载大小约为16.4MB。数据以文件形式组织,支持默认配置下的训练、验证和测试分区加载。
This structured text dataset contains 20,000 total samples, split into three partitions: a training set (16,000 samples), a validation set (2,000 samples), and a test set (2,000 samples). Each sample includes 8 fields: title (string type), category (string type), price (float type), full (string type), weight (float type), summary (currently null), prompt (currently null), and id (string type, currently null). The dataset has an overall size of approximately 32.8 MB and a download size of around 16.4 MB. The dataset is organized in file-based format and supports loading of its training, validation, and test partitions under the default configuration.
- 数据集名称:
items_review_sample - 数据集提供者:
emiliocimino - 数据集语言:未明确指定
- 数据集大小:下载大小约为 15.6 MB,数据集总大小约为 31.3 MB
- 数据集划分:
- 训练集:16,000 个样本,约 25 MB
- 验证集:2,000 个样本,约 3.2 MB
- 测试集:2,000 个样本,约 3.1 MB
- 特征字段:
title(字符串):标题category(字符串):类别price(浮点数):价格full(字符串):完整内容weight(浮点数):重量summary(空):摘要prompt(空):提示id(空):标识符
- 数据文件:默认配置下,训练集、验证集和测试集分别存储在
data/train-*、data/validation-*、data/test-*路径下。





