RNN循环神经网路（数字分类）

RNN循环神经网路（数字分类）

import torch
from torch import nn
from torch.autograd import Variable
import torchvision.datasets as dsets
import torchvision.transforms as transforms
import matplotlib.pyplot as plt
import torch.utils.data as Data

# 定义超参数 hyper parameters
EPOCH = 1  # 训练整批数据多少次, 为了节约时间, 我们只训练一次
BATCH_SIZE = 64  # 做批训练，每次训练的数据个数
TIME_STEP = 28  # rnn 时间步数 / 图片高度 每28步读取一次像素点
INPUT_SIZE = 28  # rnn 每步输入值 / 图片每行像素
LR = 0.01  # learning rate
DOWNLOAD_MNIST = False  # 如果你已经下载好了mnist数据就写上 False

# 读取训练集数据
train_data = dsets.MNIST(
    root='./mnist/',  # 保存或者提取位置
    train=True,
    transform=transforms.ToTensor(),
    # 转换 PIL.Image or numpy.ndarray 成 torch.FloatTensor (C x H x W), 训练的时候 normalize 成 [0.0, 1.0] 区间
    download=DOWNLOAD_MNIST
)
# 读取测试集数据
test_data = dsets.MNIST(
    root='./mnist/',
    train=False,
    transform=transforms.ToTensor()
)
# 批训练 50samples, 1 channel, 28x28 (50, 1, 28, 28)
train_loader = Data.DataLoader(dataset=train_data, batch_size=BATCH_SIZE, shuffle=True)

# 为了节约时间, 我们测试时只测试前2000个
test_x = Variable(test_data.test_data, volatile=True).type(torch.FloatTensor)[
         :2000] / 255  # shape from (2000, 28, 28) to (2000, 1, 28, 28), value in range(0,1)
test_y = test_data.test_labels.numpy().squeeze()[:2000]


# 建立RNN模型
class RNN(nn.Module):
    def __init__(self):
        super(RNN, self).__init__()
        self.rnn = nn.LSTM(
            input_size=INPUT_SIZE,
            hidden_size=64,
            num_layers=1,  # 越大能力越强，时间越久
            batch_first=True
        )
        self.out = nn.Linear(64, 10)

    def forward(self, x):
        # x shape (batch, time_step, input_size)
        # r_out shape (batch, time_step, output_size)
        # h_n shape (n_layers, batch, hidden_size)   LSTM 有两个 hidden states, h_n 是分线, h_c 是主线
        # h_c shape (n_layers, batch, hidden_size)
        # h_n, h_c 分线与主线剧情 r_out放了整个循环过程中产生的结果
        r_out, (h_n, h_c) = self.rnn(x, None)  # None 表示 hidden state 会用全0的 state
        out = self.out(r_out[:, -1, :])
        return out


# 训练模型
rnn = RNN()
optimizer = torch.optim.Adam(rnn.parameters(), lr=LR)
loss_func = nn.CrossEntropyLoss()  # CrossEntropyLoss做多分类任务常用
for epoch in range(EPOCH):
    for step, (x, y) in enumerate(train_loader):
        b_x = Variable(x.view(-1, 28, 28))  # reshape x to (batch, time_step, input_size)
        b_y = Variable(y)
        output = rnn(b_x)
        loss = loss_func(output, b_y)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
        if step % 50 == 0:
            test_output = rnn(test_x)
            pred_y = torch.max(test_output, 1)[1].data.numpy().squeeze()
            accuracy = sum(pred_y == test_y) / test_y.size
            print('Epoch:', epoch, '|train loss:%.4f' % loss.item(), '|test accuracy:%.2f' % accuracy)

# 验证
test_output = rnn(test_x[:10].view(-1,28,28))
pred_y = torch.max(test_output,1)[1].data.numpy().squeeze()
print('prediction number',pred_y)
print('real number',test_y[:10])