---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: url
    dtype: string
  - name: tokens
    sequence: int64
  - name: length
    dtype: int64
  splits:
  - name: train
    num_bytes: 11118787571
    num_examples: 1845941
  download_size: 4412590563
  dataset_size: 11118787571
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

#### Tokenized with aliarda/turkish-news-32k-tokenizer

The original dataset is "habanoz/news-tr-1.8M".   
This is the tokenized version of same dataset using "aliarda/turkish-news-32k-tokenizer" for experimental usage.