README.md

metadata

license: mit
task_categories:
  - text-generation
language:
  - bn
dataset_info:
  features:
    - name: content
      dtype: string
  splits:
    - name: train
      num_bytes: 2185999173
      num_examples: 437953
  download_size: 812632015
  dataset_size: 2185999173
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Bangla Newspaper Dataset (Prothom-Alo)

About Dataset

Bangla Newspaper Dataset 400k+ bangla news samples, 25+ categories, Date of Uploading: 8/12/2024

Size

Orizinal - data.json(4.61 GB)

Source

Data collected from https://www.prothomalo.com/archive

Kaggle

https://www.kaggle.com/datasets/furcifer/bangla-newspaper-dataset

HuggingFace

https://huggingface.co/datasets/zabir-nabil/bangla_newspaper_dataset

Inspiration

I have modified JSON to txt with only the content column to easy the text generation task.