README.md

metadata

dataset_info:
  config_name: hak_hl
  features:
    - name: id
      dtype: string
    - name: stem
      dtype: string
    - name: audio
      dtype: audio
    - name: duration
      dtype: float64
    - name: hanzi
      dtype: string
    - name: pinyin
      dtype: string
    - name: ipa
      dtype: string
    - name: mandarin
      dtype: string
    - name: lang_code
      dtype: string
    - name: lang_name
      dtype: string
    - name: speaker
      dtype: string
  splits:
    - name: train
      num_bytes: 9413972794
      num_examples: 24131
  download_size: 9323544625
  dataset_size: 9413972794
configs:
  - config_name: hak_hl
    data_files:
      - split: train
        path: hak_hl/train-*

TRAIN

Subset	lang_name	hours	n_utts	n_chars_in_utts	secs/utt	chars/sec	n_sents	n_chars_in_sents
hak_hl	Hakka_Hailu	53.90	24,131	644,279	8.04	3.32	0	0
Total	-	53.90	24,131	644,279	8.04	3.32	0	0