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

@aiff22 Hi，
I try to train with my own data, but when level1 and level0, the MS-SSIM loss be Nan (actually MS-SSIM loss is not used at level1).
So I run the training code in debug mode, I found that, in some case, the SSIM could be negative.
I found the code of MS-SSIM implements at jorge-pessoa/pytorch-msssim and discussions about this problem