tokenizer_config.json

{"model_max_length": 64, "do_lower_case": true}