Alizee/wikiner_fr_mixed_caps

---
language:
- fr
size_categories:
- 100K<n<1M
task_categories:
- token-classification
pretty_name: wikiner_fr
dataset_info:
features:
- name: id
dtype: int64
- name: tokens
sequence: string
- name: ner_tags
sequence:
class_label:
names:
'0': O
'1': LOC
'2': PER
'3': MISC
'4': ORG
splits:
- name: train
num_bytes: 54139057
num_examples: 120060
- name: test
num_bytes: 5952227
num_examples: 13393
download_size: 15572314
dataset_size: 60091284
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
---

# Dataset Card for "wikiner_fr_mixed_caps"

This is an update on the dataset [Jean-Baptiste/wikiner_fr](https://huggingface.co/datasets/Jean-Baptiste/wikiner_fr) with:
- removal of duplicated examples and leakage
- random de-capitalization of words (20%)

You can see the code to create the changes in the script `update_dataset.py` in the repository.

Dataset Description (reproduced from original repo):

- **Homepage:** https://metatext.io/datasets/wikiner
- **Repository:**
- **Paper:** https://www.sciencedirect.com/science/article/pii/S0004370212000276?via%3Dihub
- **Leaderboard:**
- **Point of Contact:**

原始信息汇总

数据集概述

基本信息

• 语言: 法语
• 大小类别: 100K<n<1M
• 任务类别: 词性标注
• 美观名称: wikiner_fr

数据集特征

• 特征:
  • id: 数据类型为int64
  • tokens: 序列类型为字符串
  • ner_tags: 序列类型，包含类别标签
    • 标签名称:
      • 0: O
      • 1: LOC
      • 2: PER
      • 3: MISC
      • 4: ORG

数据集分割

• 训练集:
  • 字节数: 54139057
  • 样本数: 120060
• 测试集:
  • 字节数: 5952227
  • 样本数: 13393

数据集大小

• 下载大小: 15572314
• 数据集大小: 60091284

配置

• 配置名称: default
• 数据文件:
  • 训练集路径: data/train-*
  • 测试集路径: data/test-*