liuhuanyong / Word2Vector

Self complemented word embedding methods using CBOW,skip-Gram,word2doc matrix , word2word matrix ,基于CBOW、skip-gram、词-文档矩阵、词-词矩阵四种方法的词向量生成
180 stars 76 forks source link

疑问 #2

Open haojiepan opened 6 years ago

haojiepan commented 6 years ago

为什么我按照这个代码训练出来的词向量word2word和word2doc都是0,而且运行word_cluster不管输入什么字出来的top 10都是一样的,而且,top 10后面都是nan

lili1234567890 commented 5 years ago

训练skipgram.py 模型全部是nan