Shulou(Shulou.com)06/02 Report--

This article is to share with you about how to achieve softmax regression in Pytorch, the editor thinks it is very practical, so I share it with you to learn. I hope you can get something after reading this article.

If you build the network in a custom way:

Import torchimport torchvisionimport torchvision.transforms as transformsimport matplotlib.pyplot as pltimport timeimport sysimport numpy as npimport requests# set up training set and test set Download locally mnist_train = torchvision.datasets.FashionMNIST (root='~/Datasets/FashionMNIST', train=True, download=True, transform=transforms.ToTensor ()) mnist_test = torchvision.datasets.FashionMNIST (root='~/Datasets/FashionMNIST', train=False, download=True, transform=transforms.ToTensor ()) batch_size = 256if sys.platform.startswith ('win'): num_workers = 0 # 0 means no additional processes are needed to speed up reading data else: num_workers = 4train_iter = torch.utils.data.DataLoader (mnist_train Batch_size=batch_size, shuffle=True, num_workers=num_workers) test_iter = torch.utils.data.DataLoader (mnist_test, batch_size=batch_size, shuffle=False, num_workers=num_workers) num_inputs=784# picture is 1 "28" 28 Single channel, so the final category input as 28*28=784num_outputs=10# is 10 categories, so the output is 10 W=torch.tensor (np.random.normal (0jing0.01, (num_inputs,num_outputs), dtype=torch.float) b=torch.zeros (num_outputs,dtype=torch.float) # makes it possible for ww=torch.tensor b to backpropagate. Grad _ (requires_grad=True) b. Def cross_entropy (y_hat, y): return-torch.log (y_hat.gather (1, y.view (- 1,1)) def accuracy (y_hat) Y): return (y_hat.argmax (dim=1) = = y). Float (). Mean (). Item ()''here pay attention to the writing of the correct rate calculation For y_hat, a matrix of batch_size*10 is returned, and argemax (dim=1) is an index that selects the maximum number in a dimension; so y_hat.argmax (dim=1) returns a vector of batch_size*1 Compare the returned vector with y, and average the value. You can get the correct rate of the batch''def sgd (params, lr, batch_size): # this function has been saved in the d2lzh_pytorch package to facilitate future use of for param in params: param.data-= lr * param.grad / batch_size # Note the param.datadef evalute_accuracy (data_iter,net): acc_sum,n=0.0,0 for X used when changing param here Y in data_iter: acc_sum+= (net (X) .argmax (dim=1) = = y) .float () .sum () .item () n+=y.shape [0] return acc_sum/nnum_epochs,lr=30,0.1def softmax (X): X_exp = X.exp () partition = X_exp.sum (dim=1) Keepdim=True) return X_exp / partition # broadcast mechanism def net (X): return softmax (torch.mm (X.view ((- 1, num_inputs)), w) + b) def train (net,train_iter,test_iter,loss,num_epochs,batch_size,params=None,lr=None,optimizer=None): for epoch in range (num_epochs): train_l_sum,train_acc_sum,n=0.0,0.0,0 for X Y in train_iter: y_hat=net (X) l=loss (y_hat Y) .sum () if optimizer is not None: optimizer.zero_grad () elif params is not None and params [0] .grad is not None: for param in params: param.grad.data.zero_ () l.backward () if optimizer is None: sgd (params, lr Batch_size) else: optimizer.step () # the section "concise implementation of softmax regression" will use train_l_sum+=l.item () train_acc_sum+= (y_hat.argmax (dim=1) = = y). Sum (). Item () n+=y.shape [0] test_acc=evalute_accuracy (test_iter,net) Print ('epoch% d, loss% .4f, train acc% .3f, test acc% .3f'% (epoch + 1, train_l_sum / n, train_acc_sum / n, test_acc)) if _ name__ ='_ main__': print (mnist_train) X, y = [] []''for i in range (10): X.append (mnist_ [I] [0]) y.append (mnist_ [I] [1]) show_fashion_mnist (X, get_fashion_mnist_labels (y))' 'train (net, train_iter, test_iter, cross_entropy, num_epochs, batch_size, [w, b], lr)

If you use a typical layer-building function, you can have a more concise implementation version:

Import torchimport torchvisionimport torchvision.transforms as transformsfrom torchimport nnfrom collections import OrderedDictimport matplotlib.pyplot as pltimport timeimport sysimport numpy as npimport requests# set up training set and test set Download locally from torch.nn import initmnist_train = torchvision.datasets.FashionMNIST (root='~/Datasets/FashionMNIST', train=True, download=True, transform=transforms.ToTensor ()) mnist_test = torchvision.datasets.FashionMNIST (root='~/Datasets/FashionMNIST', train=False, download=True, transform=transforms.ToTensor ()) batch_size= 256train_iter = torch.utils.data.DataLoader (mnist_train, batch_size=batch_size, shuffle=True) test_iter = torch.utils.data.DataLoader (mnist_test, batch_size=batch_size) Shuffle=False) num_inputs=784# picture is 1 "28" 28 Single channel, so the final category input as 28*28=784num_outputs=10# is 10 categories, so the output is 10 Class LinearNet (nn.Module): def _ init__ (self,num_inputs,num_outputs): super (LinearNet,self). _ _ init__ () self.linear=nn.Linear (num_inputs,num_outputs) def forward (self,x): y=self.linear (x.view (x.shape [0],-1)) return ydef evalute_accuracy (data_iter,net): acc_sum,n=0.0,0 for X Y in data_iter: acc_sum+= (net (X) .argmax (dim=1) = = y). Float (). Sum () .item () n+=y.shape [0] return acc_sum/nclass FlattenLayer (nn.Module): def _ init__ (self): super (FlattenLayer, self). _ init__ () def forward (self,x): return x.view (x.shape [0]) -1) # convert 1" 28028 to 1*784net=LinearNet (num_inputs,num_outputs) # use orderdict to build the network structure # the first layer flatten converts 1" 28" 28 to 1" 78 "the second layer is the actual working layer net=nn.Sequential (OrderedDict ([('flatten',FlattenLayer (), (' linear',nn.Linear (num_inputs))" )) init.normal_ (net.linear.weight, mean=0, std=0.01) init.constant_ (net.linear.bias, val=0) loss=nn.CrossEntropyLoss () optimizer=torch.optim.SGD (net.parameters (), lr=0.1) num_epochs=10def train (net,train_iter,test_iter,loss,num_epochs,batch_size,params=None,lr=None,optimizer=None): for epoch in range (num_epochs): train_l_sum,train_acc_sum In train_iter: y_hat=net (X) l=loss (y_hat) Y) .sum () if optimizer is not None: optimizer.zero_grad () elif params is not None and params [0] .grad is not None: for param in params: param.grad.data.zero_ () l.backward () optimizer.step () # "simplicity of softmax regression The section "implementation" will use train_l_sum+=l.item () train_acc_sum+= (y_hat.argmax (dim=1) = = y). Sum (). Item () n+=y.shape [0] test_acc=evalute_accuracy (test_iter Net) Print ('epoch% d, loss% .4f, train acc% .3f, test acc% .3f'% (epoch + 1, train_l_sum / n, train_acc_sum / n, test_acc)) if _ name__ ='_ main__': print (mnist_train) X, y = [] []''for i in range (10): X.append (mnist_ [I] [0]) y.append (mnist_ [I] [1]) show_fashion_mnist (X, get_fashion_mnist_labels (y))' 'train (net, train_iter, test_iter, loss, num_epochs, batch_size,None,None,optimizer) above is how to achieve softmax regression in Pytorch The editor believes that there are some knowledge points that we may see or use in our daily work. I hope you can learn more from this article. For more details, please follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.