---
license: cc-by-nc-4.0
task_categories:
- automatic-speech-recognition
language:
- hi
- mr
---
# ReadNet
## Dataset Description
ReadNet is an audio dataset collected from more than 2,00,000 children in the age group of 5-16 years in Hindi and Marathi language.
The dataset consists of audio files in the wav format where children read out [ASER Samples](https://asercentre.org/wp-content/uploads/2022/12/Hindi_ASER-2018.pdf) which consists of letters, words, stories and paragraphs in their native language.
### Languages
Hindi and Marathi
```python
from datasets import load_dataset

dataset = load_dataset("PrathamOrgAI/ReadNet")
```