wi_locness / README.md
Matej Klemen
Add first version of dataset script
27d7114
|
raw
history blame
2.27 kB
metadata
license: other
dataset_info:
  - config_name: A
    features:
      - name: src_tokens
        sequence: string
      - name: tgt_tokens
        sequence: string
      - name: corrections
        list:
          - name: idx_src
            sequence: int32
          - name: idx_tgt
            sequence: int32
          - name: corr_type
            dtype: string
    splits:
      - name: train
        num_bytes: 3847179
        num_examples: 10493
      - name: validation
        num_bytes: 392622
        num_examples: 1037
    download_size: 6120469
    dataset_size: 4239801
  - config_name: B
    features:
      - name: src_tokens
        sequence: string
      - name: tgt_tokens
        sequence: string
      - name: corrections
        list:
          - name: idx_src
            sequence: int32
          - name: idx_tgt
            sequence: int32
          - name: corr_type
            dtype: string
    splits:
      - name: train
        num_bytes: 4649805
        num_examples: 13032
      - name: validation
        num_bytes: 468078
        num_examples: 1290
    download_size: 6120469
    dataset_size: 5117883
  - config_name: C
    features:
      - name: src_tokens
        sequence: string
      - name: tgt_tokens
        sequence: string
      - name: corrections
        list:
          - name: idx_src
            sequence: int32
          - name: idx_tgt
            sequence: int32
          - name: corr_type
            dtype: string
    splits:
      - name: train
        num_bytes: 3765831
        num_examples: 10783
      - name: validation
        num_bytes: 390439
        num_examples: 1069
    download_size: 6120469
    dataset_size: 4156270
  - config_name: 'N'
    features:
      - name: src_tokens
        sequence: string
      - name: tgt_tokens
        sequence: string
      - name: corrections
        list:
          - name: idx_src
            sequence: int32
          - name: idx_tgt
            sequence: int32
          - name: corr_type
            dtype: string
    splits:
      - name: validation
        num_bytes: 421656
        num_examples: 988
    download_size: 6120469
    dataset_size: 421656
  - config_name: all
    features:
      - name: src_tokens
        sequence: string
      - name: tgt_tokens
        sequence: string
      - name: corrections
        list:
          - name: idx_src
            sequence: int32
          - name: idx_tgt
            sequence: int32
          - name: corr_type
            dtype: string
    splits:
      - name: train
        num_bytes: 12262815
        num_examples: 34308
      - name: validation
        num_bytes: 1672795
        num_examples: 4384
    download_size: 6120469
    dataset_size: 13935610