README.md

metadata

task_categories:
  - token-classification
language:
  - vi
tags:
  - legal
size_categories:
  - 10K<n<100K

Vi-Ner

Dataset Description

ner_tags: a list of classification labels (int). Full tagset with indices:

{'B-DATETIME': 0,
 'B-LOCATION': 1,
 'B-ORGANIZATION': 2,
 'B-PERSON': 3,
 'I-DATETIME': 4,
 'I-LOCATION': 5,
 'I-ORGANIZATION': 6,
 'I-PERSON': 7,
 'O': 8}

Data Splits

name	train	validation	test
Vi-Ner	19255	2407	2407

Example

An example of 'train' looks as follows.

{
    {'tokens': ['NSƯT', 'Hồng', 'Liên', '(trái)', 'đến', 'chúc', 'mừng', 'Thu', 'Trang..'],
    'ner_tags': ['B-PERSON', 'I-PERSON', 'I-PERSON', 'O', 'O', 'O', 'O', 'B-PERSON', 'I-PERSON'],
    'ner_idx': [3, 7, 7, 8, 8, 8, 8, 3, 7]}
}

Usage

import datasets
vi_ner = datasets.load_dataset('Minggz/Vi-Ner')
vi_ner