dclm-dedup-25B / README.md
Kyle1668's picture
Update README.md
588a7d3 verified
metadata
dataset_info:
  features:
    - name: fasttext_score
      dtype: float32
    - name: id
      dtype: string
    - name: language
      dtype: string
    - name: language_score
      dtype: float32
    - name: text
      dtype: string
    - name: url
      dtype: string
    - name: nemo_id
      dtype: string
  splits:
    - name: train
      num_bytes: 111099711518
      num_examples: 20462594
  download_size: 68540242048
  dataset_size: 111099711518
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

The first 25,000,003,108 tokens of Zyphra/dclm-dedup. Tokenized using the NeoX tokenizer (EleutherAI/gpt-neox-20b).