tokenizer_config.json

{"target_lang": "es", "source_lang": "bcl"}