sdfdsf / README.md
Codyfederer's picture
Upload merged dataset: sdfdsf
4f8f622 verified
# sdfdsf
This is a merged speech dataset containing 345 audio segments.
## Dataset Information
- **Total Segments**: 345
- **Speakers**: 7
- **Languages**: en
- **Emotions**: happy, sad, neutral, angry
- **Original Datasets**: 2
## Dataset Structure
Each example contains:
- `audio`: Audio file (WAV format, 16kHz sampling rate)
- `text`: Transcription of the audio
- `speaker_id`: Unique speaker identifier
- `emotion`: Detected emotion (neutral, happy, sad, etc.)
- `language`: Language code (en, es, fr, etc.)
## Usage
```python
from datasets import load_dataset
dataset = load_dataset("Codyfederer/sdfdsf")
```
## Speaker ID Mapping
Speaker IDs have been made unique across all merged datasets to avoid conflicts.
Original dataset information is preserved in the metadata.
## Citation
This dataset was created using the Vyvo Dataset Builder tool.