RNN, LSTM和GRU [Pytorch]

SpaceX 猎鹰重型 现代工程的杰作（高燃混剪）

 

 

import numpy as np
import torch
from torch import nn


#RNN
rnn = nn.RNN(10, 7, 2)          #(each_input_size, hidden_state, num_layers)
inputs = torch.randn(5, 3, 10)  #(seq_len, batch, input_size)
h0 = torch.randn(2, 3, 7)       #(num_layers * num_directions, batch, hidden_size)
output, hn = rnn(inputs, h0)
print(output.size(), hn.size())
print(output)
print()

#LSTM
rnn = nn.LSTM(10, 7, 2)        #(each_input_size, hidden_state, num_layers)
inputs = inputs
h0 = h0
c0 = torch.randn(2, 3, 7)      #(num_layers * num_directions, batch, hidden_size)
output, (hn,cn) = rnn(inputs, (h0,c0))   #seq_len x batch x hidden*bi_directiona