tokenizer_config.json

{"model_max_length": 512, "do_lowercase_and_remove_accent": "true"}