gru进行时间序列预测pytorch代码

以下是一个使用 PyTorch 实现 Gated Recurrent Unit (GRU) 进行时间序列预测的代码示例：

import torch
import torch.nn as nn

class GRU(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(GRU, self).__init__()
        self.hidden_size = hidden_size
        self.gru = nn.GRU(input_size, hidden_size)
        self.fc = nn.Linear(hidden_size, output_size)

    def forward(self, input, hidden):
        output, hidden = self.gru(input, hidden)
        output = self.fc(output)
        return output, hidden

    def init_hidden(self):
        return torch.zeros(1, 1, self.hidden_size)

input_size = 1
hidden_size = 16
output_size = 1
seq_len = 10

model = GRU(input_size, hidden_size, output_size)

criterion = nn.MSELoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.01)

for i in range(1000):
    seq = torch.randn(seq_len, input_size)
    target = torch.randn(seq_len, output_size)

    hidden = model.init_hidden()
    optimizer.zero_grad()

    loss = 0
    for j in range(seq_len):
        output, hidden = model(seq[j].unsqueeze(0), hidden)
        loss += criterion(output, target[j])

    loss.backward()
    optimizer.step()

    if i % 100 == 0:
        print('Epoch: {}, Loss: {:.4f}'.format(i, loss.item()))

这段代码实现了一个 GRU 模型，用于预测长度为 10 的时间序列。模型的输入维度为 1，输出维度为 1，隐藏层维度为 16。训练过程中使用均方误差损失函数和 Adam 优化器。