{
  "tokenizers": [
    {
      "name": "sp_bpe_8000",
      "dataset_suffix": "sp8000",
      "vocab_size": 8000,
      "tokenizer_train_docs": 2000000,
      "trainer_overrides": {
        "input_sentence_size": 4000000,
        "shuffle_input_sentence": true,
        "num_threads": 48,
        "character_coverage": 0.9999
      }
    }
  ]
}