README.md

metadata

dataset_info:
  features:
    - name: chars
      sequence: string
    - name: labels
      sequence:
        class_label:
          names:
            '0': D
            '1': I
            '2': P
            '3': S
    - name: logits
      sequence:
        sequence: float32
        length: 4
  splits:
    - name: train
      num_bytes: 175352249
      num_examples: 46033
    - name: validation
      num_bytes: 9697876
      num_examples: 2557
    - name: test
      num_bytes: 9693002
      num_examples: 2558
  download_size: 144555783
  dataset_size: 194743127
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
language:
  - yue
license: mit
task_categories:
  - token-classification

This segmentation dataset was generated with AlienKevin/electra-hongkongese-base-hkcancor-multi on R5dwMg/zh-wiki-yue-long.

See https://github.com/AlienKevin/dips for details.