tokenizer_config.json

{
  "model_type": "stanza_tokenizer",
  "language": "kab",
  "vocab_file": "vocab.json",
  "onnx_file": "kab_tokenizer.onnx",
  "pytorch_file": "kab_tatoeba_tokenizer.pt",
  "feature_dim": 5,
  "max_seqlen": 300,
  "num_classes": 3,
  "class_labels": [
    "B",
    "I",
    "O"
  ]
}