Views
No views yet
1import ctranslate2
2import pyonmttok
3
4text = "Всем привет, я NLP-инженер"
5ruvi_model_dir = 'ruvi_ctranslate2'
6#viru_model_dir = 'viru_ctranslate2'
7
8ru_tokenizer=pyonmttok.Tokenizer(mode="none", sp_model_path = 'ru_spm.model')
9vi_tokenizer=pyonmttok.Tokenizer(mode="none", sp_model_path = 'vi_spm.model')
10
11tokenized=ru_tokenizer.tokenize(text)
12
13translator = ctranslate2.Translator(ruvi_model_dir)
14translated = translator.translate_batch([tokenized[0]])
15print(vi_tokenizer.detokenize(translated[0][0]['tokens']))
16
17