Machine Translation Metric - BLEU Score

BLEU Score

from nltk.translate.bleu_score import sentence_bleu
reference = [['the', 'cat',"is","sitting","on","the","mat"]]
candidate = ["on",'the',"mat","is","a","cat"]
score = sentence_bleu(  reference, candidate)
print(score)


from nltk.translate.bleu_score import sentence_bleu
reference = [['the', 'cat',"is","sitting","on","the","mat"]]
candidate = ["there",'is',"cat","sitting","cat"]
score = sentence_bleu(  reference, candidate)
print(score)

Thank you very much my dear. you minimized my terror how can I test my Model trained on NMT for MSc Thesis
But how can I do the average BLEU score shall I take sample or run code for how many numbers of times?