tokenizer_config.json

{"target_lang": "spa", "source_lang": "nor"}