加载 word2vec 模型时出现数组重塑错误

Question

我有以下代码：from gensim.models import Word2Vecmodel = Word2Vec.load('model2')X = model[model.wv.vocab]这段代码适用于我的一台机器，但不适用于另一台机器。模型文件是一样的。这是怎么回事？我收到的错误消息如下：  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/gensim/models/word2vec.py", line 1330, in load    model = super(Word2Vec, cls).load(*args, **kwargs)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/gensim/models/base_any2vec.py", line 1244, in load    model = super(BaseWordEmbeddingsModel, cls).load(*args, **kwargs)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/gensim/models/base_any2vec.py", line 603, in load    return super(BaseAny2VecModel, cls).load(fname_or_handle, **kwargs)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/gensim/utils.py", line 427, in load    obj._load_specials(fname, mmap, compress, subname)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/gensim/utils.py", line 458, in _load_specials    getattr(self, attrib)._load_specials(cfname, mmap, compress, subname)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/gensim/utils.py", line 469, in _load_specials    val = np.load(subname(fname, attrib), mmap_mode=mmap)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/numpy/lib/npyio.py", line 440, in load    pickle_kwargs=pickle_kwargs)  File "/home/ec2-user/miniconda3/envs/word2vec/lib/python3.7/site-packages/numpy/lib/format.py", line 771, in read_array    array.shape = shapeValueError: cannot reshape array of size 16777184 into shape (134441,128)为了安装 gensim，我使用了conda install -c anaconda gensim

Answer 1

文件大小是错误的。上传过程中一定发生了什么。对于大型模型，word2vec 将模型保存在 3 个文件中。假设您的模型名称是“model2”，您将拥有：

1. 模型2

2. model2.trainables.syn1neg.npy

3. 模型2.wv.vectors.npy

我的.wv.vectors.npy版本比我其他机器上的版本小了几千字节。