| # export KMP_DUPLICATE_LIB_OK=TRUE | |
| import ctranslate2 | |
| import sentencepiece as spm | |
| from translate import Translator | |
| text="我从税务局打来" | |
| translator = ctranslate2.Translator("zh_en/") | |
| sp = spm.SentencePieceProcessor(model_file="zh_en/sentencepiece.model") | |
| tokens = sp.encode(text, out_type=str) | |
| translated_tokens = translator.translate_batch([tokens]) | |
| output = sp.decode(translated_tokens[0][0]["tokens"]).replace("▁"," " ) | |
| print( f"{output}") |