fe1ixxu / ALMA

State-of-the-art LLM-based translation models.
MIT License
440 stars 35 forks source link

Updates data processing in utils.py [fix bug] #19

Closed sweta20 closed 10 months ago

sweta20 commented 10 months ago

When delta is non-zero, the original implementation returns scores instead of sentences.

fe1ixxu commented 10 months ago

Thanks! It is an error when we clean up the code.