tokenizer_config.json

{"target_lang": "afr", "source_lang": "spa"}