---
license: mit
language:
- ta
pretty_name: solvari
---

# "_solvari_" means line of words

### Data Fields

Each Parquet file contains:
- `text`: Raw Tamil text
- `words`: Word count

### Technical Specifications

- Format: Parquet
- Encoding: UTF-8
- Text Filter: Tamil (Unicode range: \u0b80-\u0bff)
- Tokenizer: AI4Bharat/indic-bert