如何将预训练的 LDA 模型加载到 Jupiter Notebook？

Question

我在 PC 上训练了 LDA 模型，并使用 model.save() 命令将其保存在本地。我可以在 PyCharm 中加载这个模型并输出主题，但是当我尝试在 Jupiter Notebook 中加载相同的模型时出现错误。

有人遇到同样的问题并解决了吗？ 以下是完整的错误输出：

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-14-043e6d6083e2> in <module>
      1 # Loading saved model
----> 2 model = models.LdaModel.load('information_extraction/optimal_LDA3.model')
      3 # model_topics = model.show_topics(formatted=True)
      4 # pprint.pprint(model.print_topics(num_words=15))

~/anaconda3/lib/python3.7/site-packages/gensim/models/ldamodel.py in load(cls, fname, *args, **kwargs)
   1636         """
   1637         kwargs['mmap'] = kwargs.get('mmap', None)
-> 1638         result = super(LdaModel, cls).load(fname, *args, **kwargs)
   1639 
   1640         # check if `random_state` attribute has been set after main pickle load

~/anaconda3/lib/python3.7/site-packages/gensim/utils.py in load(cls, fname, mmap)
    424         compress, subname = SaveLoad._adapt_by_suffix(fname)
    425 
--> 426         obj = unpickle(fname)
    427         obj._load_specials(fname, mmap, compress, subname)
    428         logger.info("loaded %s", fname)

~/anaconda3/lib/python3.7/site-packages/gensim/utils.py in unpickle(fname)
   1382         # Because of loading from S3 load can't be used (missing readline in smart_open)
   1383         if sys.version_info > (3, 0):
-> 1384             return _pickle.load(f, encoding='latin1')
   1385         else:
   1386             return _pickle.loads(f.read())

ModuleNotFoundError: No module named 'numpy.random._pickle'

Answer 1

已解决： 问题是 numpy 版本的差异。我在 numpy==1.17.0 上训练了 LDA 模型，然后我安装了 Anaconda 并使用 numpy==1.16.4 运行 Jupiter。