8.6.rnn concise

nn.RNN

接口很简单。注意隐状态 state 是在 RNN 对象之外另开对象进行存储的。

batch_size, num_steps = 32, 35
train_iter, vocab = d2l.load_data_time_machine(batch_size, num_steps)

num_hiddens = 256
# RNN() 内部含有 W, b 参数之类的
rnn_layer = nn.RNN(len(vocab), num_hiddens)

# state 形状 [隐藏层数，批量大小，隐藏单元数]
state = torch.zeros((1, batch_size, num_hiddens))
# X.shape = [时间步，批量，词典大小(独热)]
X = torch.rand(size=(num_steps, batch_size, len(vocab)))
# Y.shape = [时间步，批量，隐藏单元数], state 形状不变
Y, state = self.rnn(X, state)

封装

class RNNModel(nn.Module):
    """循环神经网络模型"""
    def __init__(self, rnn_layer, vocab_size, **kwargs):
        super(RNNModel, self).__init__(**kwargs)
        self.rnn = rnn_layer
        self.vocab_size = vocab_size
        self.num_hiddens = self.rnn.hidden_size
        # 如果RNN是双向的（之后将介绍），num_directions应该是2，否则应该是1
        if not self.rnn.bidirectional:
            self.num_directions = 1
            self.linear = nn.Linear(self.num_hiddens, self.vocab_size)
        else:
            self.num_directions = 2
            self.linear = nn.Linear(self.num_hiddens * 2, self.vocab_size)

    def forward(self, inputs, state):
        X = F.one_hot(inputs.T.long(), self.vocab_size)
        X = X.to(torch.float32)
        Y, state = self.rnn(X, state)
        # 全连接层首先将Y的形状改为(时间步数*批量大小,隐藏单元数)
        # 它的输出形状是(时间步数*批量大小,词表大小)。
        output = self.linear(Y.reshape((-1, Y.shape[-1])))
        return output, state

    def begin_state(self, device, batch_size=1):
        if not isinstance(self.rnn, nn.LSTM):
            # nn.GRU以张量作为隐状态
            return  torch.zeros((self.num_directions * self.rnn.num_layers,
                                 batch_size, self.num_hiddens),
                                device=device)
        else:
            # nn.LSTM以元组作为隐状态
            return (torch.zeros((
                self.num_directions * self.rnn.num_layers,
                batch_size, self.num_hiddens), device=device),
                    torch.zeros((
                        self.num_directions * self.rnn.num_layers,
                        batch_size, self.num_hiddens), device=device))

32:07