error in beam_decode

[Tangzy7]

hi I met a bug when running evaluation with beam_size = 3. I think it's because self.sentence_scores, line 17, in evaluate.py, is always [].

File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/main.py", line 49, in run
runTest(n_layers, hidden_size, reverse, args.test, beam_size, inp, args.corpus)
File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/evaluate.py", line 179, in runTest
evaluateRandomly(encoder, decoder, voc, pairs, reverse, beam_size, 20)
File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/evaluate.py", line 132, in evaluateRandomly
output_words_list = evaluate(encoder, decoder, voc, pair[0], beam_size)
File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/evaluate.py", line 116, in evaluate
return beam_decode(decoder, decoder_hidden, encoder_outputs, voc, beam_size)
File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/evaluate.py", line 63, in beam_decode
term, top = sentence.addTopk(topi, topv, decoder_hidden, beam_size, voc)
File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/evaluate.py", line 31, in addTopk
self.avgScore())) # tuple(word_list, score_float
File "/Users/zhiyao/PycharmProjects/pytorch-chatbot-master/evaluate.py", line 21, in avgScore
raise ValueError("Calculate average score of sentence, but got no word")
ValueError: Calculate average score of sentence, but got no word

[Tangzy7]

Sorry, it's just due to the problem of my model