tokenizer_config.json

{"target_lang": "cpp", "source_lang": "cpp"}