动手学PyTorch版RNN报错
RuntimeError: expected scalar type Float but found Long
把源代码中的RNNModel修改：
原代码
class RNNModel(nn.Module):
def __init__(self, rnn_layer, vocab_size):
super(RNNModel, self).__init__()
self.rnn = rnn_layer
self.hidden_size = rnn_layer.hidden_size * (2 if rnn_layer.bidirectional else 1)
self.vocab_size = vocab_size
self.dense = nn.Linear(self.hidden_size, vocab_size)
self.state = None
def forward(self, inputs, state): # inputs: (batch, seq_len)
# 获取one-hot向量表示
X = to_onehot(inputs, self.vocab_size) # X是个list
Y, self.state = self.rnn(torch.stack(X), state)
# 全连接层会首先将Y的形状变成(num_steps * batch_size, num_hiddens)，它的输出
# 形状为(num_steps * batch_size, vocab_size)
output = self.dense(Y.view(-1, Y.shape[-1]))
return output, self.state
更改后的代码
class RNNModel(nn.Module):
def __init__(self, rnn_layer, vocab_size):
super(RNNModel, self).__init__()
self.rnn = rnn_layer
self.hidden_size = rnn_layer.hidden_size * (2 if rnn_layer.bidirectional else 1)
self.vocab_size = vocab_size
self.dense = nn.Linear(self.hidden_size, vocab_size)
self.state = None
def forward(self, inputs, state): # inputs: (batch, seq_len)
# 获取one-hot向量表示
X = F.one_hot(inputs, self.vocab_size) # X是个list
Y, self.state = self.rnn(X.float(), state)
# 全连接层会首先将Y的形状变成(num_steps * batch_size, num_hiddens)，它的输出
# 形状为(num_steps * batch_size, vocab_size)
output = self.dense(Y.view(-1, Y.shape[-1]))
return output, self.state
