# Emotion Audio Dataset

This dataset contains audio files with emotional speech samples organized by emotion categories.

## Dataset Structure

```
thucdangvan020999/emotion_audio/
├── angry/          # Angry emotion audio files
├── sad/            # Sad emotion audio files  
├── happy/          # Happy emotion audio files
├── neutral/        # Neutral emotion audio files
├── metadata.json   # Complete dataset metadata
├── angry_metadata.json
├── sad_metadata.json
├── happy_metadata.json
└── neutral_metadata.json
```

## Statistics

- **Total samples**: 400
- **Emotions**: angry, sad, happy, neutral
- **Voices**: echo, fable, onyx, nova, alloy
- **Format**: MP3 audio files

## Emotion Distribution

- **Angry**: 100 samples
- **Happy**: 100 samples
- **Neutral**: 100 samples
- **Sad**: 100 samples

## Voice Distribution

- **alloy**: 80 samples
- **echo**: 80 samples
- **fable**: 80 samples
- **nova**: 80 samples
- **onyx**: 80 samples

## Usage

```python
from datasets import load_dataset

# Load the complete dataset
dataset = load_dataset("thucdangvan020999/emotion_audio")

# Or load specific emotion subset
angry_data = load_dataset("thucdangvan020999/emotion_audio", data_files="angry/*")
```

## License

Please specify your license here.