tokenizer_config.json

{"do_lower_case": true, "model_max_length": 512, "tokenizer_class": "BertTokenizer"}