ytzi's picture
pre-tokenized with gpt2-medium
253dd57 verified
metadata
dataset_info:
  - config_name: julia
    features:
      - name: content
        dtype: string
      - name: input_ids
        sequence: int32
      - name: ratio_char_token
        dtype: float64
      - name: token_count
        dtype: int64
    splits:
      - name: train
        num_bytes: 4488572639
        num_examples: 298672
    download_size: 1340259889
    dataset_size: 4488572639
  - config_name: lua
    features:
      - name: content
        dtype: string
      - name: input_ids
        sequence: int32
      - name: ratio_char_token
        dtype: float64
      - name: token_count
        dtype: int64
    splits:
      - name: train
        num_bytes: 9277991216
        num_examples: 558861
    download_size: 2761679970
    dataset_size: 9277991216
  - config_name: scheme
    features:
      - name: content
        dtype: string
      - name: input_ids
        sequence: int32
      - name: ratio_char_token
        dtype: float64
      - name: token_count
        dtype: int64
    splits:
      - name: train
        num_bytes: 907251818
        num_examples: 44261
    download_size: 216232390
    dataset_size: 907251818
configs:
  - config_name: julia
    data_files:
      - split: train
        path: julia/train-*
  - config_name: lua
    data_files:
      - split: train
        path: lua/train-*
  - config_name: scheme
    data_files:
      - split: train
        path: scheme/train-*