I used the trained model ,which was saved during training. The code is **_model = torch.load("xxxxx/xxxxx.bin")_** _**aspects = model.get_aspect_words(w2v_model, logger,topn=10)**_ Is this the right way to get the aspects ? thank you very much.