README.md

metadata

dataset_info:
  features:
    - name: _id
      dtype: string
    - name: url
      dtype: string
    - name: title
      dtype: string
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 20346504450
      num_examples: 41488110
  download_size: 10094783514
  dataset_size: 20346504450
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

This dataset is the English subset of the Cohere/wikipedia-2023-11-embed-multilingual-v3 dataset but without the embeddings (emb) column.