---
task_categories:
- automatic-speech-recognition
- audio-classification
language:
- zh
tags:
- speech-recognition
- entity-tagging
- intent-classifcation
- age-prediction
- gender-prediction
---
# Meta Speech Recognition Mandarin Dataset (AISHELL3)

This dataset contains both metadata and audio files for Mandarin speech recognition samples from the AISHELL3 corpus.

## Dataset Statistics

### Splits and Sample Counts
- **train**: 60098 samples
- **valid**: 3163 samples
- **test**: 24772 samples

## Example Samples
### train
```json
{
  "audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs_train/SSB00430356.wav",
  "text": "她以 ENTITY_PRODUCT 滴鸡精 END 调养身体。 AGE_14_25 GENDER_FEMALE DIALECT_NORTH",
  "duration": 3.0600680272108844
}
```
```json
{
  "audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs_train/SSB00380135.wav",
  "text": "ENTITY_ORGANIZATION 奇乐乐影院 END 。 AGE_14_25 GENDER_FEMALE DIALECT_NORTH",
  "duration": 2.476984126984127
}
```

### valid
```json
{
  "audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs_train/SSB07600351.wav",
  "text": "我觉得她有很大的麻烦。 AGE_14_25 GENDER_FEMALE DIALECT_NORTH",
  "duration": 2.6599092970521543
}
```
```json
{
  "audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs_train/SSB04700221.wav",
  "text": "画的是 ENTITY_LOCATION 纽约 END 地铁西路线。 AGE_14_25 GENDER_FEMALE DIALECT_NORTH",
  "duration": 3.159659863945578
}
```

### test
```json
{
  "audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs/SSB07110415.wav",
  "text": "来首歌吻别。 AGE_14_25 GENDER_FEMALE DIALECT_SOUTH",
  "duration": 1.7899773242630386
}
```
```json
{
  "audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs/SSB10910445.wav",
  "text": "末日重启。 AGE_14_25 GENDER_FEMALE DIALECT_NORTH",
  "duration": 1.9711791383219954
}
```


## Dataset Information

- **Language**: Mandarin Chinese
- **Source**: AISHELL3
- **Format**: Each sample contains:
  - `audio_filepath`: Path to the audio file
  - `text`: Transcription text with speaker metadata
  - `duration`: Duration of the audio in seconds

## Speaker Metadata

The text field includes speaker metadata in the format:
- ENTITY_<type>: Marking begin of an entity
- INTENT_* : Intent of an utterance
- AGE_*: Speaker age range
- GENDER_*: Speaker gender
- DIALECT_*: Speaker dialect

## Usage Notes

1. The dataset includes both metadata and audio files.
2. Audio files are stored in the dataset repository.
3. The dataset is suitable for Mandarin speech recognition tasks.