tokenizer_config.json

{"do_lower_case": false, "strip_accents": false, "model_max_length": 128}