README.md

---
dataset_info:
  config_name: nan_tw
  features:
  - name: id
    dtype: string
  - name: session/stem
    dtype: string
  - name: audio
    dtype: audio
  - name: duration
    dtype: float64
  - name: hanzi
    dtype: string
  - name: lang_code
    dtype: string
  - name: lang_name
    dtype: string
  - name: speaker
    dtype: string
  - name: accent
    dtype: string
  - name: has_romaji
    dtype: bool
  splits:
  - name: train
    num_bytes: 5524580148
    num_examples: 42252
  download_size: 5472521024
  dataset_size: 5524580148
configs:
- config_name: nan_tw
  data_files:
  - split: train
    path: nan_tw/train-*
---
## TRAIN
| Subset | lang_name | hours | n_utts | n_chars_in_utts | secs/utt | chars/sec | n_sents | n_chars_in_sents |
|---|---|---|---|---|---|---|---|---|
| nan_tw | Taigi | 31.64 | 42,252 | 490,323 | 2.70 | 4.31 | 0 | 0 |
| **Total** | **-** | **31.64** | **42,252** | **490,323** | **2.70** | **4.31** | **0** | **0** |