Stack_Tokenized / README.md
Trevor Dohm
Tokenized creole
c4f511c verified
|
raw
history blame
367 Bytes
metadata
dataset_info:
  features:
    - name: content
      dtype: string
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
  splits:
    - name: train
      num_bytes: 3387797
      num_examples: 767
  download_size: 1372299
  dataset_size: 3387797
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/creole/train-*