metadata.json

{
  "vocab_size": 50000,
  "corpus_file": "kannada_corpus.txt",
  "min_frequency": 1,
  "language": "Kannada (kn)",
  "pre_tokenizer": "Whitespace"
}