奉献pytorch 搭建 CNN 卷积神经网络训练图像识别的模型，配合numpy 和matplotlib 一起使用调用 cuda GPU进行加速训练

1、Torch构建简单的模型

# coding:utf-8
import torch

class Net(torch.nn.Module):
    def __init__(self,img_rgb=3,img_size=32,img_class=13):
        super(Net, self).__init__()
        self.conv1 = torch.nn.Sequential(
            torch.nn.Conv2d(in_channels=img_rgb, out_channels=img_size, kernel_size=3, stride=1,padding= 1), #
            torch.nn.ReLU(),
            torch.nn.MaxPool2d(2),
            # torch.nn.Dropout(0.5)
        )
        self.conv2 = torch.nn.Sequential(
            torch.nn.Conv2d(28, 64, 3, 1, 1),
            torch.nn.ReLU(),
            torch.nn.MaxPool2d(2)
        )
        self.conv3 = torch.nn.Sequential(
            torch.nn.Conv2d(64, 64, 3, 1, 1),
            torch.nn.ReLU(),
            torch.nn.MaxPool2d(2)
        )
        self.dense = torch.nn.Sequential(
            torch.nn.Linear(64 * 3 * 3, 128),
            torch.nn.ReLU(),
            torch.nn.Linear(128, img_class)
        )

    def forward(self, x):
        conv1_out = self.conv1(x)
        conv2_out = self.conv2(conv1_out)
        conv3_out = self.conv3(conv2_out)
        res = conv3_out.view(conv3_out.size(0), -1)
        out = self.dense(res)
        return out

CUDA = torch.cuda.is_available()

model = Net(1,28,13)
print(model)

optimizer = torch.optim.Adam(model.parameters())
loss_func = torch.nn.MultiLabelSoftMarginLoss()#nn.CrossEntropyLoss()

if CUDA:
    model.cuda()

def batch_training_data(x_train,y_train,batch_size,i):
    n = len(x_train)
    left_limit = batch_size*i
    right_limit = left_limit+batch_size
    if n>=right_limit:
        return x_train[left_limit:right_limit,:,:,:],y_train[left_limit:right_limit,:]
    else:
        return x_train[left_limit:, :, :, :], y_train[left_limit:, :]

　　

2、奉献训练过程的代码

#  coding:utf-8
import time
import os
import torch
import numpy as np
from data_processing import get_DS
from CNN_nework_model import cnn_face_discern_model
from torch.autograd import Variable
from use_torch_creation_model import optimizer, model, loss_func, batch_training_data,CUDA
from sklearn.metrics import accuracy_score

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

st = time.time()
# 获取训练集与测试集以 8:2 分割
x_,y_,y_true,label = get_DS()

label_number = len(label)

x_train,y_train = x_[:960,:,:,:].reshape((960,1,28,28)),y_[:960,:]

x_test,y_test = x_[960:,:,:,:].reshape((340,1,28,28)),y_[960:,:]

y_test_label = y_true[960:]

print(time.time() - st)
print(x_train.shape,x_test.shape)

batch_size = 100
n = int(len(x_train)/batch_size)+1

for epoch in range(100):
    global loss
    for batch in range(n):
        x_training,y_training = batch_training_data(x_train,y_train,batch_size,batch)
        batch_x,batch_y = Variable(torch.from_numpy(x_training)).float(),Variable(torch.from_numpy(y_training)).float()
        if CUDA:
            batch_x=batch_x.cuda()
            batch_y=batch_y.cuda()

        out = model(batch_x)
        loss = loss_func(out, batch_y)

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
    # 测试精确度
    if epoch%9==0:
        global x_test_tst
        if CUDA:
            x_test_tst = Variable(torch.from_numpy(x_test)).float().cuda()
        y_pred = model(x_test_tst)

        y_predict = np.argmax(y_pred.cpu().data.numpy(),axis=1)

        acc = accuracy_score(y_test_label,y_predict)

        print("loss={} aucc={}".format(loss.cpu().data.numpy(),acc))

　　

3、总结

通过博主通过TensorFlow、keras、pytorch进行训练同样的模型同样的图像数据，结果发现，pyTorch快了很多倍，特别是在导入模型的时候比TensorFlow快了很多。合适部署接口和集成在项目中。