| dataset_info: | |
| features: | |
| - name: text | |
| dtype: string | |
| - name: description | |
| dtype: string | |
| - name: url | |
| dtype: string | |
| - name: short | |
| dtype: int64 | |
| - name: audio_tokens | |
| sequence: int64 | |
| - name: md5 | |
| dtype: string | |
| - name: start_us | |
| sequence: float64 | |
| - name: end_us | |
| sequence: float64 | |
| - name: words | |
| sequence: string | |
| splits: | |
| - name: train | |
| num_bytes: 43164739 | |
| num_examples: 867 | |
| download_size: 8645119 | |
| dataset_size: 43164739 | |
| configs: | |
| - config_name: default | |
| data_files: | |
| - split: train | |
| path: data/train-* | |
| This dataset contains no audio file. The `audio_tokens` column is generated using WavTokenizer, to be used by OuteTTS 500M v0.2 model. | |
| If you want to re-use this data to train for TTS model, download the full dataset containing WAV file from https://huggingface.co/datasets/ngxson/ThisTTS_dataset | |