Jupyter containers hosted by Coursera cause a lot of trouble. Perhaps more than they are worth.

• They are very limited in terms of lifetime and CPU. The docs say 90 minutes / 0.5-2 CPUs. That's definitely insufficient to train Breakout, for example.
• Updating them is inconvenient. We don't h

I add a function(according the code in the tutorial4) to calculate the bleu score, but i get the vey low score(0.09), could you tell me why?
This is code to calculate bleu:

def translate_sentence(sentence, src_field, trg_field, model, device, max_len = 50):
    
    model.eval()
        
    if isinstance(sentence, str):
        nlp = spacy.load('de')
        tokens = [token.te

e.g. based on the google doc https://docs.google.com/document/d/1ZI1V_2I3tETAeGnAwYZnrYTdeiOfdtWCL4l4DmRimFc/

p.s. I think that using docker for the course is an overhead which is not justified. As people create a designated VM for the course it would have been better to share a VM image so installation is one click and forget about docker. Docker is great when you need multiple environment or

개요

• 공식 튜토리얼 중 pytorch/tutorials@7ba6693 를 참고하여 반영하였습니다.
• 1.3 때(#12)와 다르게, 각 튜토리얼들은 초/중/고급(beginner / intermediate / advanced)으로 나누어 정리하고
  체크박스를 사용하여 번역 여부를 표시합니다.
• 현재 메인 메뉴에서 바로 접근이 불가능한 튜토리얼은 이탤릭 으로 표시하였습니다.

초급 튜토리얼 (beginner)

• [audio_preprocessing_tutorial.py](https://github.com/9bow/PyTorch-tutorials-kr/blo