是否可以在PyTorch中使用LSTMCells模块实现多层LSTM？

import torch
from torch import nn

# 定义多层LSTM
class MultiLayerLSTM(nn.Module):
    def __init__(self, input_size, hidden_size, num_layers):
        super(MultiLayerLSTM, self).__init__()
        self.num_layers = num_layers
        self.lstm_cells = nn.ModuleList([nn.LSTMCell(input_size, hidden_size)])
        self.lstm_cells.extend([nn.LSTMCell(hidden_size, hidden_size) for _ in range(num_layers-1)])

    def forward(self, input):
        outputs = []
        h_t = []
        c_t = []

        for layer in range(self.num_layers):
            if layer == 0:
                h, c = self.lstm_cells[layer](input)
            else:
                h, c = self.lstm_cells[layer](h_t[layer-1])

            h_t.append(h)
            c_t.append(c)
            outputs.append(h)

        return outputs, (h_t, c_t)

# 创建多层LSTM模型实例
input_size = 10
hidden_size = 20
num_layers = 3
model = MultiLayerLSTM(input_size, hidden_size, num_layers)

# 使用多层LSTM进行前向传播
input = torch.randn(5, input_size)
outputs, (h_t, c_t) = model(input)

# 打印输出和隐藏状态
print(outputs)
print(h_t)
print(c_t)