tokenizer_config.json

{"do_lower_case": false, "max_model_length": 512}