1. 实现SRN
（1）使用Numpy
import numpy as np
inputs = np.array([[1., 1.],
[1., 1.],
[2., 2.]]) # 初始化输入序列
print('inputs is ', inputs)
state_t = np.zeros(2, ) # 初始化存储器
print('state_t is ', state_t)
w1, w2, w3, w4, w5, w6, w7, w8 = 1., 1., 1., 1., 1., 1., 1., 1.
U1, U2, U3, U4 = 1., 1., 1., 1.
print('--------------------------------------')
for input_t in inputs:
print('inputs is ', input_t)
print('state_t is ', state_t)
in_h1 = np.dot([w1, w3], input_t) + np.dot([U2, U4], state_t)
in_h2 = np.dot([w2, w4], input_t) + np.dot([U1, U3], state_t)
state_t = in_h1, in_h2
output_y1 = np.dot([w5, w7], [in_h1, in_h2])
output_y2 = np.dot([w6, w8], [in_h1, in_h2])
print('output_y is ', output_y1, output_y2)
print('---------------')
（2）在1的基础上，增加激活函数tanh
import numpy as np
inputs = np.array([[1., 1.],
[1., 1.],
[2., 2.]]) # 初始化输入序列
print('inputs is ', inputs)
state_t = np.zeros(2, ) # 初始化存储器
print('state_t is ', state_t)
w1, w2, w3, w4, w5, w6, w7, w8 = 1., 1., 1., 1., 1., 1., 1., 1.
U1, U2, U3, U4 = 1., 1., 1., 1.
print('--------------------------------------')
for input_t in inputs:
print('inputs is ', input_t)
print('state_t is ', state_t)
in_h1 = np.tanh(np.dot([w1, w3], input_t) + np.dot([U2, U4], state_t))
in_h2 = np.tanh(np.dot([w2, w4], input_t) + np.dot([U1, U3], state_t))
state_t = in_h1, in_h2
output_y1 = np.dot([w5, w7], [in_h1, in_h2])
output_y2 = np.dot([w6, w8], [in_h1, in_h2])
print('output_y is ', output_y1, output_y2)
print('---------------')
（3）使用nn.RNNCell实现
import torch
batch_size = 1
seq_len = 3 # 序列长度
input_size = 2 # 输入序列维度
hidden_size = 2 # 隐藏层维度
output_size = 2 # 输出层维度
# RNNCell
cell = torch.nn.RNNCell(input_size=input_size, hidden_size=hidden_size)
# 初始化参数 https://zhuanlan.zhihu.com/p/342012463
for name, param in cell.named_parameters():
if name.startswith("weight"):
torch.nn.init.ones_(param)
else:
torch.nn.init.zeros_(param)
# 线性层
liner = torch.nn.Linear(hidden_size, output_size)
liner.weight.data = torch.Tensor([[1, 1], [1, 1]])
liner.bias.data = torch.Tensor([0.0])
seq = torch.Tensor([[[1, 1]],
[[1, 1]],
[[2, 2]]])
hidden = torch.zeros(batch_size, hidden_size)
output = torch.zeros(batch_size, output_size)
for idx, input in enumerate(seq):
print('=' * 20, idx, '=' * 20)
print('Input :', input)
print('hidden :', hidden)
hidden = cell(input, hidden)
output = liner(hidden)
print('output :', output)
（4）使用nn.RNN实现
import torch
batch_size = 1
seq_len = 3
input_size = 2
hidden_size = 2
num_layers = 1
output_size = 2
cell = torch.nn.RNN(input_size=input_size, hidden_size=hidden_size, num_layers=num_layers)
for name, param in cell.named_parameters(): # 初始化参数
if name.startswith("weight"):
torch.nn.init.ones_(param)
else:
torch.nn.init.zeros_(param)
# 线性层
liner = torch.nn.Linear(hidden_size, output_size)
liner.weight.data = torch.Tensor([[1, 1], [1, 1]])
liner.bias.data = torch.Tensor([0.0])
inputs = torch.Tensor([[[1, 1]],
[[1, 1]],
[[2, 2]]])
hidden = torch.zeros(num_layers, batch_size, hidden_size)
out, hidden = cell(inputs, hidden)
print('Input :', inputs[0])
print('hidden:', 0, 0)
print('Output:', liner(out[0]))
print('--------------------------------------')
print('Input :', inputs[1])
print('hidden:', out[0])
print('Output:', liner(out[1]))
print('--------------------------------------')
print('Input :', inputs[2])
print('hidden:', out[1])
print('Output:', liner(out[2]))
2. 实现“序列到序列”
观看视频，学习RNN原理，并实现视频P12中的教学案例
案例
import torch
batch_size = 1
seq_len = 3
input_size = 4
hidden_size = 2
num_layers = 1
cell = torch.nn.RNN(input_size=input_size, hidden_size=hidden_size,
num_layers=num_layers, batch_first=True)
# (seqLen, batchSize, inputSize)
inputs = torch.randn(batch_size, seq_len, input_size)
hidden = torch.zeros(num_layers, batch_size, hidden_size)
out, hidden = cell(inputs, hidden)
print('Output size:', out.shape)
print('Output:', out)
print('Hidden size: ', hidden.shape)
print('Hidden: ', hidden)
import torch
input_size=4
hidden_size=4
batch_size=1
idx2char=['e','h','l','o']
x_data=[1,0,2,2,3]
y_data=[3,1,2,3,2]
one_hot_lookup=[[1,0,0,0],[0,1,0,0],[0,0,1,0],[0,0,0,1]]
x_one_hot=[one_hot_lookup[x] for x in x_data]
inputs=torch.Tensor(x_one_hot).view(-1,batch_size,input_size)
labels=torch.LongTensor(y_data).view(-1,1)
class Model(torch.nn.Module):
def __init__(self,input_size,hidden_size,batch_size):
super(Model,self).__init__()
self.batch_size=batch_size
self.input_size=input_size
self.hidden_size=hidden_size
self.rnncell=torch.nn.RNNCell(input_size=self.input_size,hidden_size=self.hidden_size)
def forward(self,input,hidden):
hidden=self.rnncell(input,hidden)
return hidden
def init_hidden(self):
return torch.zeros(self.batch_size,self.hidden_size)
net=Model(input_size,hidden_size,batch_size)
criterion=torch.nn.CrossEntropyLoss()
optimizer=torch.optim.Adam(net.parameters(),lr=0.1)
for epoch in range(15):
loss=0
optimizer.zero_grad()
hidden=net.init_hidden()
print('Predicted string:',end='')
for input,label in zip(inputs,labels):
hidden=net(input,hidden)
loss+=criterion(hidden,label)
_,idx=hidden.max(dim=1)
print(idx2char[idx.item()],end='')
loss.backward()
optimizer.step()
print(',Epoch [%d/15] loss=%.4f'%(epoch+1,loss.item()))
3. “编码器-解码器”的简单实现
seq2seq的PyTorch实现_哔哩哔哩_bilibili
import torch
import numpy as np
import torch.nn as nn
import torch.utils.data as Data
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
letter = [c for c in 'SE?abcdefghijklmnopqrstuvwxyz']
letter2idx = {n: i for i, n in enumerate(letter)}
seq_data = [['man', 'women'], ['black', 'white'], ['king', 'queen'], ['girl', 'boy'], ['up', 'down'], ['high', 'low']]
n_step = max([max(len(i), len(j)) for i, j in seq_data]) # max_len(=5)
n_hidden = 128
n_class = len(letter2idx)
batch_size = 3
def make_data(seq_data):
enc_input_all, dec_input_all, dec_output_all = [], [], []
for seq in seq_data:
for i in range(2):
seq[i] = seq[i] + '?' * (n_step - len(seq[i]))
enc_input = [letter2idx[n] for n in (seq[0] + 'E')]
dec_input = [letter2idx[n] for n in ('S' + seq[1])]
dec_output = [letter2idx[n] for n in (seq[1] + 'E')]
enc_input_all.append(np.eye(n_class)[enc_input])
dec_input_all.append(np.eye(n_class)[dec_input])
dec_output_all.append(dec_output)
return torch.Tensor(enc_input_all), torch.Tensor(dec_input_all), torch.LongTensor(dec_output_all)
enc_input_all, dec_input_all, dec_output_all = make_data(seq_data)
class TranslateDataSet(Data.Dataset):
def __init__(self, enc_input_all, dec_input_all, dec_output_all):
self.enc_input_all = enc_input_all
self.dec_input_all = dec_input_all
self.dec_output_all = dec_output_all
def __len__(self):
return len(self.enc_input_all)
def __getitem__(self, idx):
return self.enc_input_all[idx], self.dec_input_all[idx], self.dec_output_all[idx]
loader = Data.DataLoader(TranslateDataSet(enc_input_all, dec_input_all, dec_output_all), batch_size, True)
class Seq2Seq(nn.Module):
def __init__(self):
super(Seq2Seq, self).__init__()
self.encoder = nn.RNN(input_size=n_class, hidden_size=n_hidden, dropout=0.5)
self.decoder = nn.RNN(input_size=n_class, hidden_size=n_hidden, dropout=0.5)
self.fc = nn.Linear(n_hidden, n_class)
def forward(self, enc_input, enc_hidden, dec_input):
enc_input = enc_input.transpose(0, 1)
dec_input = dec_input.transpose(0, 1)
_, h_t = self.encoder(enc_input, enc_hidden)
outputs, _ = self.decoder(dec_input, h_t)
model = self.fc(outputs)
return model
model = Seq2Seq().to(device)
criterion = nn.CrossEntropyLoss().to(device)
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)
for epoch in range(5000):
for enc_input_batch, dec_input_batch, dec_output_batch in loader:
h_0 = torch.zeros(1, batch_size, n_hidden).to(device)
(enc_input_batch, dec_intput_batch, dec_output_batch) = (
enc_input_batch.to(device), dec_input_batch.to(device), dec_output_batch.to(device))
pred = model(enc_input_batch, h_0, dec_intput_batch)
pred = pred.transpose(0, 1)
loss = 0
for i in range(len(dec_output_batch)):
loss += criterion(pred[i], dec_output_batch[i])
if (epoch + 1) % 1000 == 0:
print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.6f}'.format(loss))
optimizer.zero_grad()
loss.backward()
optimizer.step()
def translate(word):
enc_input, dec_input, _ = make_data([[word, '?' * n_step]])
enc_input, dec_input = enc_input.to(device), dec_input.to(device)
hidden = torch.zeros(1, 1, n_hidden).to(device)
output = model(enc_input, hidden, dec_input)
predict = output.data.max(2, keepdim=True)[1]
decoded = [letter[i] for i in predict]
translated = ''.join(decoded[:decoded.index('E')])
return translated.replace('?', '')
print('test')
print('man ->', translate('man'))
print('mans ->', translate('mans'))
print('king ->', translate('king'))
print('black ->', translate('black'))
print('up ->', translate('up'))
4.简单总结nn.RNNCell、nn.RNN
具体来说，nn.RNNCell是RNN的一个单操作，它接受一个输入端的输入和一个隐藏状态作为输入，并输出一个新的隐藏状态。nn.RNNCell可以通过堆叠多个单元来构建多层的RNN。
nn.RNN则是通过重复调用nn.RNNCell来实现整个循环过程。它接受一个序列作为输入，并在序列的每个时间步应用nn.RNNCell。nn.RNN还可以支持双向RNN，它在每个时间步同时应用一个前向和一个反向的RNNCell。
概括起来就是，nn.RNNCell用于定义循环神经网络的一个单元操作，而nn.RNN利用nn.RNNCell来构建整个循环神经网络。
5.谈一谈对“序列”、“序列到序列”的理解
序列是指一系列按照顺序排列的元素或事件。在计算机领域，序列一般都是一组数据、文本或符号号串，按照排序组合成的一种结构。
“序列到序列”（sequence-to-sequence简seq2seq），该模型包括两个主要部分：编码器（encoder）和解码器（decoder）。编码器将输入序列（如一个句子）转换为固定长度的向量表示，然后解码器将这个向量表示转换为输出序列（如另一种语言的句子）。序列到序列模型可以用来训练循环神经网络（RNN）或者长短期记忆网络LSTM，有效地处理输入和输出序列之间的复杂关系。
6.总结本周理论课和作业，写心得体会
理论课上的学习初步了解到了循环的原理，结合视频理解了“时间”在循环神经网络的具体体现，循环神经网络中的“循环”让其内部的一或多个节点保存上一时刻的状态，并将其作为当前时刻的输入，因此网络会“记忆”之前的信息并将其应用于当前的计算。【循环神经网络】5分钟搞懂RNN，3D动画深入浅出_哔哩哔哩_bilibili
本次实验也有很多含金量高的视频，对视频里嵌入大概的理解——举例子是：
有一个向量（a，b），其中，a表示对A的喜爱程度，b表示对B的喜爱程度，其中数值越大，表示喜爱程度越大，负数表示厌恶。现有甲乙丙丁四人，分别表示的向量可以是甲（-1，5）代表甲轻微厌恶A且很喜欢B，乙（1，5），丙（3，-1），丁（-3，-2），通过这样的方式同样定义一组词嵌入，然后我们就可以通过向量的夹角来判断他们的相似度，而当维度过高时，嵌入维度的定义也可以让神经网络去更新，而不需要举例子来探究实际意义，只需要根据夹角来判断语义相似度即可。
扫一扫
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
