README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: text
      dtype: string
  download_size: 112131429
  dataset_size: 180048487
configs:
  - config_name: pt
    data_files:
      - split: train
        path: pt/train.jsonl
      - split: test
        path: pt/test.jsonl
  - config_name: es
    data_files:
      - split: train
        path: es/train.jsonl
      - split: test
        path: es/test.jsonl
  - config_name: fr
    data_files:
      - split: train
        path: fr/train.jsonl
      - split: test
        path: fr/test.jsonl
  - config_name: de
    data_files:
      - split: train
        path: de/train.jsonl
      - split: test
        path: de/test.jsonl
  - config_name: en
    data_files:
      - split: train
        path: en/train.jsonl
      - split: test
        path: en/test.jsonl
license: mit
language:
  - pt
  - en
  - de
  - fr
  - es
tags:
  - news
size_categories:
  - 10K<n<100K

Público

This dataset was build by translating a set of 34,157 news from Público, an European Portuguese news paper. The news have been translated using Google Translator. To now more about the data visit the Github repos used to scrape and translate the news.