GRU
RNN存在的问题:梯度较容易出现衰减或爆炸(BPTT)
⻔控循环神经⽹络:捕捉时间序列中时间步距离较⼤的依赖关系
RNN:
H
t
=
ϕ
(
X
t
W
x
h
+
H
t
−
1
W
h
h
+
b
h
)
H_{t} = ϕ(X_{t}W_{xh} + H_{t-1}W_{hh} + b_{h})
Ht=ϕ(XtWxh+Ht−1Whh+bh)
GRU:
R
t
=
σ
(
X
t
W
x
r
+
H
t
−
1
W
h
r
+
b
r
)
Z
t
=
σ
(
X
t
W
x
z
+
H
t
−
1
W
h
z
+
b
z
)
H
~
t
=
t
a
n
h
(
X
t
W
x
h
+
(
R
t
⊙
H
t
−
1
)
W
h
h
+
b
h
)
H
t
=
Z
t
⊙
H
t
−
1
+
(
1
−
Z
t
)
⊙
H
~
t
R_{t} = σ(X_tW_{xr} + H_{t−1}W_{hr} + b_r)\\ Z_{t} = σ(X_tW_{xz} + H_{t−1}W_{hz} + b_z)\\ \widetilde{H}_t = tanh(X_tW_{xh} + (R_t ⊙H_{t−1})W_{hh} + b_h)\\ H_t = Z_t⊙H_{t−1} + (1−Z_t)⊙\widetilde{H}_t
Rt=σ(XtWxr+Ht−1Whr+br)Zt=σ(XtWxz+Ht−1Whz+bz)H
t=tanh(XtWxh+(Rt⊙Ht−1)Whh+bh)Ht=Zt⊙Ht−1+(1−Zt)⊙H
t
• 重置⻔有助于捕捉时间序列⾥短期的依赖关系;
• 更新⻔有助于捕捉时间序列⾥⻓期的依赖关系。
载入数据集
import os
os.listdir('/home/kesci/input')
['d2lzh1981', 'houseprices2807', 'jaychou_lyrics4703', 'd2l_jay9460']
import numpy as np
import torch
from torch import nn, optim
import torch.nn.functional as F
import sys
sys.path.append("../input/")
import d2l_jay9460 as d2l
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
(corpus_indices, char_to_idx, idx_to_char, vocab_size) = d2l.load_data_jay_lyrics()
初始化参数
num_inputs, num_hiddens, num_outputs = vocab_size, 256, vocab_size
print('will use', device)
def get_params():
def _one(shape):
ts = torch.tensor(np.random.normal(0, 0.01, size=shape), device=device, dtype=torch.float32) #正态分布
return torch.nn.Parameter(ts, requires_grad=True)
def _three():
return (_one((num_inputs, num_hiddens)),
_one((num_hiddens, num_hiddens)),
torch.nn.Parameter(torch.zeros(num_hiddens, device=device, dtype=torch.float32), requires_grad=True))
W_xz, W_hz, b_z = _three() # 更新门参数
W_xr, W_hr, b_r = _three() # 重置门参数
W_xh, W_hh, b_h = _three() # 候选隐藏状态参数
# 输出层参数
W_hq = _one((num_hiddens, num_outputs))
b_q = torch.nn.Parameter(torch.zeros(num_outputs, device=device, dtype=torch.float32), requires_grad=True)
return nn.ParameterList([W_xz, W_hz, b_z, W_xr, W_hr, b_r, W_xh, W_hh, b_h, W_hq, b_q])
def init_gru_state(batch_size, num_hiddens, device): #隐藏状态初始化
return (torch.zeros((batch_size, num_hiddens), device=device), )
will use cpu
GRU模型
def gru(inputs, state, params):
W_xz, W_hz, b_z, W_xr, W_hr, b_r, W_xh, W_hh, b_h, W_hq, b_q = params
H, = state
outputs = []
for X in inputs:
Z = torch.sigmoid(torch.matmul(X, W_xz) + torch.matmul(H, W_hz) + b_z)
R = torch.sigmoid(torch.matmul(X, W_xr) + torch.matmul(H, W_hr) + b_r)
H_tilda = torch.tanh(torch.matmul(X, W_xh) + R * torch.matmul(H, W_hh) + b_h)
H = Z * H + (1 - Z) * H_tilda
Y = torch.matmul(H, W_hq) + b_q
outputs.append(Y)
return outputs, (H,)
训练模型
num_epochs, num_steps, batch_size, lr, clipping_theta = 160, 35, 32, 1e2, 1e-2
pred_period, pred_len, prefixes = 40, 50, ['分开', '不分开']
d2l.train_and_predict_rnn(gru, get_params, init_gru_state, num_hiddens,
vocab_size, device, corpus_indices, idx_to_char,
char_to_idx, False, num_epochs, num_steps, lr,
clipping_theta, batch_size, pred_period, pred_len,
prefixes)
简洁实现
num_hiddens=256
num_epochs, num_steps, batch_size, lr, clipping_theta = 160, 35, 32, 1e2, 1e-2
pred_period, pred_len, prefixes = 40, 50, ['分开', '不分开']
lr = 1e-2 # 注意调整学习率
gru_layer = nn.GRU(input_size=vocab_size, hidden_size=num_hiddens)
model = d2l.RNNModel(gru_layer, vocab_size).to(device)
d2l.train_and_predict_rnn_pytorch(model, num_hiddens, vocab_size, device,
corpus_indices, idx_to_char, char_to_idx,
num_epochs, num_steps, lr, clipping_theta,
batch_size, pred_period, pred_len, prefixes)
LSTM
** 长短期记忆long short-term memory **:
遗忘门:控制上一时间步的记忆细胞
输入门:控制当前时间步的输入
输出门:控制从记忆细胞到隐藏状态
记忆细胞:⼀种特殊的隐藏状态的信息的流动
I t = σ ( X t W x i + H t − 1 W h i + b i ) F t = σ ( X t W x f + H t − 1 W h f + b f ) O t = σ ( X t W x o + H t − 1 W h o + b o ) C ~ t = t a n h ( X t W x c + H t − 1 W h c + b c ) C t = F t ⊙ C t − 1 + I t ⊙ C ~ t H t = O t ⊙ t a n h ( C t ) I_t = σ(X_tW_{xi} + H_{t−1}W_{hi} + b_i) \\ F_t = σ(X_tW_{xf} + H_{t−1}W_{hf} + b_f)\\ O_t = σ(X_tW_{xo} + H_{t−1}W_{ho} + b_o)\\ \widetilde{C}_t = tanh(X_tW_{xc} + H_{t−1}W_{hc} + b_c)\\ C_t = F_t ⊙C_{t−1} + I_t ⊙\widetilde{C}_t\\ H_t = O_t⊙tanh(C_t) It=σ(XtWxi+Ht−1Whi+bi)Ft=σ(XtWxf+Ht−1Whf+bf)Ot=σ(XtWxo+Ht−1Who+bo)C t=tanh(XtWxc+Ht−1Whc+bc)Ct=Ft⊙Ct−1+It⊙C tHt=Ot⊙tanh(Ct)
初始化参数
num_inputs, num_hiddens, num_outputs = vocab_size, 256, vocab_size
print('will use', device)
def get_params():
def _one(shape):
ts = torch.tensor(np.random.normal(0, 0.01, size=shape), device=device, dtype=torch.float32)
return torch.nn.Parameter(ts, requires_grad=True)
def _three():
return (_one((num_inputs, num_hiddens)),
_one((num_hiddens, num_hiddens)),
torch.nn.Parameter(torch.zeros(num_hiddens, device=device, dtype=torch.float32), requires_grad=True))
W_xi, W_hi, b_i = _three() # 输入门参数
W_xf, W_hf, b_f = _three() # 遗忘门参数
W_xo, W_ho, b_o = _three() # 输出门参数
W_xc, W_hc, b_c = _three() # 候选记忆细胞参数
# 输出层参数
W_hq = _one((num_hiddens, num_outputs))
b_q = torch.nn.Parameter(torch.zeros(num_outputs, device=device, dtype=torch.float32), requires_grad=True)
return nn.ParameterList([W_xi, W_hi, b_i, W_xf, W_hf, b_f, W_xo, W_ho, b_o, W_xc, W_hc, b_c, W_hq, b_q])
def init_lstm_state(batch_size, num_hiddens, device):
return (torch.zeros((batch_size, num_hiddens), device=device),
torch.zeros((batch_size, num_hiddens), device=device))
LSTM模型
def lstm(inputs, state, params):
[W_xi, W_hi, b_i, W_xf, W_hf, b_f, W_xo, W_ho, b_o, W_xc, W_hc, b_c, W_hq, b_q] = params
(H, C) = state
outputs = []
for X in inputs:
I = torch.sigmoid(torch.matmul(X, W_xi) + torch.matmul(H, W_hi) + b_i)
F = torch.sigmoid(torch.matmul(X, W_xf) + torch.matmul(H, W_hf) + b_f)
O = torch.sigmoid(torch.matmul(X, W_xo) + torch.matmul(H, W_ho) + b_o)
C_tilda = torch.tanh(torch.matmul(X, W_xc) + torch.matmul(H, W_hc) + b_c)
C = F * C + I * C_tilda
H = O * C.tanh()
Y = torch.matmul(H, W_hq) + b_q
outputs.append(Y)
return outputs, (H, C)
训练模型
num_epochs, num_steps, batch_size, lr, clipping_theta = 160, 35, 32, 1e2, 1e-2
pred_period, pred_len, prefixes = 40, 50, ['分开', '不分开']
d2l.train_and_predict_rnn(lstm, get_params, init_lstm_state, num_hiddens,
vocab_size, device, corpus_indices, idx_to_char,
char_to_idx, False, num_epochs, num_steps, lr,
clipping_theta, batch_size, pred_period, pred_len,
prefixes)
简洁实现
num_hiddens=256
num_epochs, num_steps, batch_size, lr, clipping_theta = 160, 35, 32, 1e2, 1e-2
pred_period, pred_len, prefixes = 40, 50, ['分开', '不分开']
lr = 1e-2 # 注意调整学习率
lstm_layer = nn.LSTM(input_size=vocab_size, hidden_size=num_hiddens)
model = d2l.RNNModel(lstm_layer, vocab_size)
d2l.train_and_predict_rnn_pytorch(model, num_hiddens, vocab_size, device,
corpus_indices, idx_to_char, char_to_idx,
num_epochs, num_steps, lr, clipping_theta,
batch_size, pred_period, pred_len, prefixes)
深度循环神经网络
H t ( 1 ) = ϕ ( X t W x h ( 1 ) + H t − 1 ( 1 ) W h h ( 1 ) + b h ( 1 ) ) H t ( ℓ ) = ϕ ( H t ( ℓ − 1 ) W x h ( ℓ ) + H t − 1 ( ℓ ) W h h ( ℓ ) + b h ( ℓ ) ) O t = H t ( L ) W h q + b q \boldsymbol{H}_t^{(1)} = \phi(\boldsymbol{X}_t \boldsymbol{W}_{xh}^{(1)} + \boldsymbol{H}_{t-1}^{(1)} \boldsymbol{W}_{hh}^{(1)} + \boldsymbol{b}_h^{(1)})\\ \boldsymbol{H}_t^{(\ell)} = \phi(\boldsymbol{H}_t^{(\ell-1)} \boldsymbol{W}_{xh}^{(\ell)} + \boldsymbol{H}_{t-1}^{(\ell)} \boldsymbol{W}_{hh}^{(\ell)} + \boldsymbol{b}_h^{(\ell)})\\ \boldsymbol{O}_t = \boldsymbol{H}_t^{(L)} \boldsymbol{W}_{hq} + \boldsymbol{b}_q Ht(1)=ϕ(XtWxh(1)+Ht−1(1)Whh(1)+bh(1))Ht(ℓ)=ϕ(Ht(ℓ−1)Wxh(ℓ)+Ht−1(ℓ)Whh(ℓ)+bh(ℓ))Ot=Ht(L)Whq+bq
num_hiddens=256
num_epochs, num_steps, batch_size, lr, clipping_theta = 160, 35, 32, 1e2, 1e-2
pred_period, pred_len, prefixes = 40, 50, ['分开', '不分开']
lr = 1e-2 # 注意调整学习率
gru_layer = nn.LSTM(input_size=vocab_size, hidden_size=num_hiddens,num_layers=2)
model = d2l.RNNModel(gru_layer, vocab_size).to(device)
d2l.train_and_predict_rnn_pytorch(model, num_hiddens, vocab_size, device,
corpus_indices, idx_to_char, char_to_idx,
num_epochs, num_steps, lr, clipping_theta,
batch_size, pred_period, pred_len, prefixes)
gru_layer = nn.LSTM(input_size=vocab_size, hidden_size=num_hiddens,num_layers=6)
model = d2l.RNNModel(gru_layer, vocab_size).to(device)
d2l.train_and_predict_rnn_pytorch(model, num_hiddens, vocab_size, device,
corpus_indices, idx_to_char, char_to_idx,
num_epochs, num_steps, lr, clipping_theta,
batch_size, pred_period, pred_len, prefixes)
epoch 40, perplexity 276.815235, time 8.50 sec
- 分开
- 不分开
epoch 80, perplexity 276.278550, time 8.51 sec
- 分开
- 不分开
epoch 120, perplexity 276.146710, time 8.53 sec
- 分开
- 不分开
epoch 160, perplexity 275.739864, time 9.04 sec
- 分开
- 不分开