models/demo_tokenizer/config.json

{
  "vocab_size": 30000,
  "min_frequency": 2,
  "special_tokens": [
    "<pad>",
    "<unk>",
    "<s>",
    "</s>",
    "<mask>"
  ],
  "lowercase": false,
  "add_prefix_space": false,
  "trim_offsets": true,
  "show_progress": true,
  "initial_alphabet": [],
  "data_dir": "data",
  "model_save_dir": "models",
  "model_name": "demo_tokenizer"
}