tokenizer_config.json

{"do_lower_case": false, "tokenize_chinese_chars": false, "max_len": 512}