tbwxmu/sampn

sampn's Issues

Weights in the Attention Layer?

The weight matrices W_a and W_b in the attention layer are not mentioned in the paper, but present in the code. I am not so familiar with attention layers, but I was wondering if I missed something in the paper or it is common knowledge to apply those weights in an self attention layer?

提问：网络框架图与代码的对应关系

1.网络框架图中 self attention layer 是否 对应于代码mpn.py 中的self.attention = args.attention语句？
2.fully connect layer 是否对应于代码
if self.attention:
self.W_a = nn.Linear(self.hidden_size, self.hidden_size, bias=self.bias)
self.W_b = nn.Linear(self.hidden_size, self.hidden_size)