voiceoperation / translation.py
Zeimoto
new portuguese model
0536b78
raw
history blame
763 Bytes
from transformers import MarianMTModel, MarianTokenizer
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
def get_model_name(languageCode: str) -> str:
match languageCode:
case "pt":
model_name = "Helsinki-NLP/opus-mt-pt-en"
return model_name
def get_translation(text_to_translate: str, languageCode: str) -> str:
tokenizer = AutoTokenizer.from_pretrained("unicamp-dl/translation-pt-en-t5")
model = AutoModelForSeq2SeqLM.from_pretrained("unicamp-dl/translation-pt-en-t5")
pten_pipeline = pipeline('text2text-generation', model=model, tokenizer=tokenizer)
translated_text = pten_pipeline(text_to_translate)
print("Translated text:", translated_text)
return translated_text