fathan/autotrain-data-code-mixed-language-identification
收藏Hugging Face2023-02-15 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/fathan/autotrain-data-code-mixed-language-identification
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是为code-mixed-language-identification项目自动处理的,用于token分类任务。数据集的BCP-47语言代码为unk。数据集包含tokens和tags字段,tokens为字符串序列,tags为ClassLabel序列,标签包括EN、ID、JV等。数据集分为训练集和验证集,训练集有1105个样本,验证集有438个样本。
This dataset was automatically processed for the code-mixed-language-identification project, and is intended for token classification tasks. Its BCP-47 language code is unk. The dataset includes two fields: `tokens` and `tags`. The `tokens` field is a sequence of strings, while the `tags` field is a sequence of ClassLabel objects, with the available labels including EN, ID, JV and others. The dataset is split into training and validation subsets, which contain 1105 and 438 samples respectively.
提供机构:
fathan
原始信息汇总
数据集概述
数据集描述
该数据集是为项目“code-mixed-language-identification”自动处理而生成的。
语言
数据集的语言标识符为BCP-47代码unk。
数据集结构
数据实例
数据集的样本结构如下:
json [ { "feat_Unnamed: 0": 1104, "tokens": [...], "tags": [...] }, { "feat_Unnamed: 0": 239, "tokens": [...], "tags": [...] } ]
数据集字段
数据集包含以下字段:
json { "feat_Unnamed: 0": "Value(dtype=int64, id=None)", "tokens": "Sequence(feature=Value(dtype=string, id=None), length=-1, id=None)", "tags": "Sequence(feature=ClassLabel(names=[EN, ID, JV, MIX_ID_EN, MIX_ID_JV, MIX_JV_EN, OTH], id=None), length=-1, id=None)" }
数据集分割
数据集分为训练集和验证集,分割详情如下:
| 分割名称 | 样本数量 |
|---|---|
| 训练集 | 1105 |
| 验证集 | 438 |



