Datasets:

formospeech
/

yttd_trs

dataset_info:
  config_name: nan_tw
  features:
    - name: id
      dtype: string
    - name: rid
      dtype: string
    - name: audio
      dtype: audio
    - name: duration
      dtype: float64
    - name: hanzi
      dtype: string
    - name: mandarin
      dtype: string
    - name: lang_code
      dtype: string
    - name: lang_name
      dtype: string
    - name: has_romaji
      dtype: bool
  splits:
    - name: train
      num_bytes: 2858884433
      num_examples: 42650
    - name: eval
      num_bytes: 316039452
      num_examples: 4652
  download_size: 3173689503
  dataset_size: 3174923885
configs:
  - config_name: nan_tw
    data_files:
      - split: train
        path: nan_tw/train-*
      - split: eval
        path: nan_tw/eval-*

EVAL

Subset	lang_name	hours	n_utts	n_chars_in_utts	secs/utt	chars/sec	n_sents	n_chars_in_sents
nan_tw	Taigi	2.71	4,652	45,186	2.10	4.63	0	0
Total	-	2.71	4,652	45,186	2.10	4.63	0	0

TRAIN

Subset	lang_name	hours	n_utts	n_chars_in_utts	secs/utt	chars/sec	n_sents	n_chars_in_sents
nan_tw	Taigi	24.77	42,650	408,671	2.09	4.58	0	0
Total	-	24.77	42,650	408,671	2.09	4.58	0	0