tokenizer_config.json

{
  "do_lower_case": true,
  "model_max_length": 512,
  "tokenizer_class": "DistilBertTokenizer",
  "pretrained_model_name_or_path": "distilbert-base-uncased"
}