tokenizer_config.json

{"target_lang": "fr", "source_lang": "swc"}