bert-large-uncased-whole-word-masking-squad2/tokenizer_config.json

{"init_inputs": []}