README.md

metadata

license: mit
language:
  - ta
pretty_name: solvari

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

Total Tokens - 93,877,412