keras训练期间无效的参数错误

Question

在keras中的word2vec训练期间无效的参数错误 尽管vocab的大小是index+1

请参见下面的网络架构摘要：

__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_11 (InputLayer)           (None, 1)            0                                            
__________________________________________________________________________________________________
input_12 (InputLayer)           (None, 1)            0                                            
__________________________________________________________________________________________________
embedding_11 (Embedding)        (None, 1, 300)       1138500     input_11[0][0]                   
__________________________________________________________________________________________________
embedding_12 (Embedding)        (None, 1, 300)       1138500     input_12[0][0]                   
__________________________________________________________________________________________________
dot_6 (Dot)                     (None, 1, 1)         0           embedding_11[0][0]               
                                                                 embedding_12[0][0]               
__________________________________________________________________________________________________
reshape_6 (Reshape)             (None, 1)            0           dot_6[0][0]                      
__________________________________________________________________________________________________
activation_5 (Activation)       (None, 1)            0           reshape_6[0][0]                  
==================================================================================================
Total params: 2,277,000
Trainable params: 2,277,000
Non-trainable params: 0

这是代码的一部分：

n_epochs=5
for epoch in range(n_epochs):
    loss=0.    
    for i ,doc in enumerate(X_train_tokens):
        data,labels=skipgrams(sequence=doc,vocabulary_size=vocab_size,window_size=4)
        x=[np.array(x) for x in zip(*data)]
        y=np.array(labels,dtype=np.int32)
        if x:
            loss +=model.train_on_batch(x,y)
    print('Epoch:',epoch,'\t loss:',loss)

出现以下错误

从内存中删除基础状态对象，否则它将保留 活着，因为从追溯中可以引用状态 由于InvalidArgumentError：indexs [7,0] = 3795不在[0，3795）