python - 是否可以从 python 中的句子语料库重新训练 word2vec 模型(例如 GoogleNews-vectors-negative300.bin)？

Question

我正在使用预先训练的谷歌新闻数据集，通过在 python 中使用 Gensim 库来获取词向量

model = Word2Vec.load_word2vec_format('GoogleNews-vectors-negative300.bin', binary=True)

加载模型后，我将训练评论句子单词转换为向量

#reading all sentences from training file
with open('restaurantSentences', 'r') as infile:
x_train = infile.readlines()
#cleaning sentences
x_train = [review_to_wordlist(review,remove_stopwords=True) for review in x_train]
train_vecs = np.concatenate([buildWordVector(z, n_dim) for z in x_train])

在 word2Vec 过程中，我的语料库中的单词出现了很多错误，这些错误不在模型中。问题是我如何重新训练已经预训练的模型(例如 GoogleNews-vectors-negative300.bin')，以便为那些缺失的词获取词向量。

以下是我尝试过的:从我的训练句子中训练出一个新模型

# Set values for various parameters
num_features = 300    # Word vector dimensionality                      
min_word_count = 10   # Minimum word count                        
num_workers = 4       # Number of threads to run in parallel
context = 10          # Context window    size                                                                                    
downsampling = 1e-3   # Downsample setting for frequent words

sentences = gensim.models.word2vec.LineSentence("restaurantSentences")
# Initialize and train the model (this will take some time)
print "Training model..."
model = gensim.models.Word2Vec(sentences, workers=num_workers,size=num_features, min_count = min_word_count, 
                      window = context, sample = downsampling)


model.build_vocab(sentences)
model.train(sentences)
model.n_similarity(["food"], ["rice"])

成功了!但问题是我的数据集非常小，而且训练大型模型的资源也很少。

我正在研究的第二种方法是扩展已经训练好的模型，例如 GoogleNews-vectors-negative300.bin。

model = Word2Vec.load_word2vec_format('GoogleNews-vectors-negative300.bin', binary=True)
sentences = gensim.models.word2vec.LineSentence("restaurantSentences")
model.train(sentences)

是否可行，是否是一种好的使用方式，请帮帮我

Answer 1

这就是我在技术上解决问题的方式:

使用来自 Radim Rehurek 的可迭代句子准备数据输入:https://rare-technologies.com/word2vec-tutorial/

sentences = MySentences('newcorpus')  

建立模型

model = gensim.models.Word2Vec(sentences)

将词汇与谷歌词向量相交

model.intersect_word2vec_format('GoogleNews-vectors-negative300.bin',
                                lockf=1.0,
                                binary=True)

最后执行模型并更新

model.train(sentences)

注意事项:从实质性的角度来看，一个可能非常小的语料库是否真的可以“改进”在一个庞大的语料库上训练的谷歌词向量当然是非常值得商榷的......