---
license: cc-by-nc-sa-4.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: text
    dtype: string
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 471991350.2039943
    num_examples: 72218
  download_size: 251312380
  dataset_size: 471991350.2039943
---
A document collection for the summer 2025 [EPFL CS-552](https://nlp.epfl.ch/cs-552-modern-nlp/) MNLP project by artdev99.

Contains Wikipedia and CommonCrawl pages. 

1'293'772  chunks from 72'218 documents (chunk size: 512, overlap: 64).