这篇教程pytorch finetuning 自己的图片进行训练操作写得很实用，希望能帮到您。
一、pytorch finetuning 自己的图片进行训练这种读取图片的方式用的是torch自带的 ImageFolder，读取的文件夹必须在一个大的子文件下，按类别归好类。 就像我现在要区分三个类别。 
#perpare data set#train datatrain_data=torchvision.datasets.ImageFolder('F:/eyeDataSet/trainData',transform=transforms.Compose( [ transforms.Scale(256), transforms.CenterCrop(224), transforms.ToTensor() ]))print(len(train_data))train_loader=DataLoader(train_data,batch_size=20,shuffle=True)然后就是fine tuning自己的网络，在torch中可以对整个网络修改后，训练全部的参数也可以只训练其中的一部分，我这里就只训练最后一个全连接层。 torchvision中提供了很多常用的模型，比如resnet ，Vgg，Alexnet等等 # prepare modelmode1_ft_res18=torchvision.models.resnet18(pretrained=True)for param in mode1_ft_res18.parameters(): param.requires_grad=Falsenum_fc=mode1_ft_res18.fc.in_featuresmode1_ft_res18.fc=torch.nn.Linear(num_fc,3) 定义自己的优化器，注意这里的参数只传入最后一层的 #loss function and optimizercriterion=torch.nn.CrossEntropyLoss()#parameters only train the last fc layeroptimizer=torch.optim.Adam(mode1_ft_res18.fc.parameters(),lr=0.001) 然后就可以开始训练了，定义好各种参数。 #start train#label not one-hot encoderEPOCH=1for epoch in range(EPOCH): train_loss=0. train_acc=0. for step,data in enumerate(train_loader): batch_x,batch_y=data batch_x,batch_y=Variable(batch_x),Variable(batch_y) #batch_y not one hot #out is the probability of eatch class # such as one sample[-1.1009 0.1411 0.0320],need to calculate the max index # out shape is batch_size * class out=mode1_ft_res18(batch_x) loss=criterion(out,batch_y) train_loss+=loss.data[0] # pred is the expect class #batch_y is the true label pred=torch.max(out,1)[1] train_correct=(pred==batch_y).sum() train_acc+=train_correct.data[0] optimizer.zero_grad() loss.backward() optimizer.step() if step%14==0: print('Epoch: ',epoch,'Step',step, 'Train_loss: ',train_loss/((step+1)*20),'Train acc: ',train_acc/((step+1)*20))测试部分和训练部分类似这里就不一一说明。 这样就完整了对自己网络的训练测试，完整代码如下： import torchimport numpy as npimport torchvisionfrom torchvision import transforms,utilsfrom torch.utils.data import DataLoaderfrom torch.autograd import Variable#perpare data set#train datatrain_data=torchvision.datasets.ImageFolder('F:/eyeDataSet/trainData',transform=transforms.Compose( [ transforms.Scale(256), transforms.CenterCrop(224), transforms.ToTensor() ]))print(len(train_data))train_loader=DataLoader(train_data,batch_size=20,shuffle=True) #test datatest_data=torchvision.datasets.ImageFolder('F:/eyeDataSet/testData',transform=transforms.Compose( [ transforms.Scale(256), transforms.CenterCrop(224), transforms.ToTensor() ]))test_loader=DataLoader(test_data,batch_size=20,shuffle=True) # prepare modelmode1_ft_res18=torchvision.models.resnet18(pretrained=True)for param in mode1_ft_res18.parameters(): param.requires_grad=Falsenum_fc=mode1_ft_res18.fc.in_featuresmode1_ft_res18.fc=torch.nn.Linear(num_fc,3) #loss function and optimizercriterion=torch.nn.CrossEntropyLoss()#parameters only train the last fc layeroptimizer=torch.optim.Adam(mode1_ft_res18.fc.parameters(),lr=0.001) #start train#label not one-hot encoderEPOCH=1for epoch in range(EPOCH): train_loss=0. train_acc=0. for step,data in enumerate(train_loader): batch_x,batch_y=data batch_x,batch_y=Variable(batch_x),Variable(batch_y) #batch_y not one hot #out is the probability of eatch class # such as one sample[-1.1009 0.1411 0.0320],need to calculate the max index # out shape is batch_size * class out=mode1_ft_res18(batch_x) loss=criterion(out,batch_y) train_loss+=loss.data[0] # pred is the expect class #batch_y is the true label pred=torch.max(out,1)[1] train_correct=(pred==batch_y).sum() train_acc+=train_correct.data[0] optimizer.zero_grad() loss.backward() optimizer.step() if step%14==0: print('Epoch: ',epoch,'Step',step, 'Train_loss: ',train_loss/((step+1)*20),'Train acc: ',train_acc/((step+1)*20)) #print('Epoch: ', epoch, 'Train_loss: ', train_loss / len(train_data), 'Train acc: ', train_acc / len(train_data)) # test modelmode1_ft_res18.eval()eval_loss=0eval_acc=0for step ,data in enumerate(test_loader): batch_x,batch_y=data batch_x,batch_y=Variable(batch_x),Variable(batch_y) out=mode1_ft_res18(batch_x) loss = criterion(out, batch_y) eval_loss += loss.data[0] # pred is the expect class # batch_y is the true label pred = torch.max(out, 1)[1] test_correct = (pred == batch_y).sum() eval_acc += test_correct.data[0] optimizer.zero_grad() loss.backward() optimizer.step()print( 'Test_loss: ', eval_loss / len(test_data), 'Test acc: ', eval_acc / len(test_data))二、PyTorch 利用预训练模型进行Fine-tuning在Deep Learning领域，很多子领域的应用，比如一些动物识别，食物的识别等，公开的可用的数据库相对于ImageNet等数据库而言，其规模太小了，无法利用深度网络模型直接train from scratch，容易引起过拟合，这时就需要把一些在大规模数据库上已经训练完成的模型拿过来，在目标数据库上直接进行Fine-tuning（微调），这个已经经过训练的模型对于目标数据集而言，只是一种相对较好的参数初始化方法而已，尤其是大数据集与目标数据集结构比较相似的话，经过在目标数据集上微调能够得到不错的效果。 Fine-tune预训练网络的步骤：1. 首先更改预训练模型分类层全连接层的数目，因为一般目标数据集的类别数与大规模数据库的类别数不一致，更改为目标数据集上训练集的类别数目即可，一致的话则无需更改； 2. 把分类器前的网络的所有层的参数固定，即不让它们参与学习，不进行反向传播，只训练分类层的网络，这时学习率可以设置的大一点，如是原来初始学习率的10倍或几倍或0.01等，这时候网络训练的比较快，因为除了分类层，其它层不需要进行反向传播，可以多尝试不同的学习率设置。 3.接下来是设置相对较小的学习率，对整个网络进行训练，这时网络训练变慢啦。 下面对利用PyTorch深度学习框架Fine-tune预训练网络的过程中涉及到的固定可学习参数，对不同的层设置不同的学习率等进行详细讲解。 1. PyTorch对某些层固定网络的可学习参数的方法： class Net(nn.Module): def __init__(self, num_classes=546): super(Net, self).__init__() self.features = nn.Sequential( nn.Conv2d(1, 64, kernel_size=3, stride=2, padding=1), nn.BatchNorm2d(64), nn.ReLU(inplace=True), nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1), nn.BatchNorm2d(64), nn.ReLU(inplace=True), ) self.Conv1_1 = nn.Sequential( nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1), nn.BatchNorm2d(64), nn.ReLU(inplace=True), nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1), nn.BatchNorm2d(64), ) for p in self.parameters(): p.requires_grad=False self.Conv1_2 = nn.Sequential( nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1), nn.BatchNorm2d(64), nn.ReLU(inplace=True), nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1), nn.BatchNorm2d(64), ) 如上述代码，则模型Net网络中self.features与self.Conv1_1层中的参数便是固定，不可学习的。这主要看代码： for p in self.parameters(): p.requires_grad=False 插入的位置，这段代码前的所有层的参数是不可学习的，也就没有反向传播过程。也可以指定某一层的参数不可学习，如下： for p in self.features.parameters(): p.requires_grad=False 则 self.features层所有参数均是不可学习的。 注意，上述代码设置若要真正生效，在训练网络时需要在设置优化器如下： optimizer = torch.optim.SGD(filter(lambda p: p.requires_grad, model.parameters()), args.lr, momentum=args.momentum, weight_decay=args.weight_decay) 2. PyTorch之为不同的层设置不同的学习率 model = Net()conv1_2_params = list(map(id, model.Conv1_2.parameters()))base_params = filter(lambda p: id(p) not in conv1_2_params, model.parameters())optimizer = torch.optim.SGD([ {'params': base_params}, {'params': model.Conv1_2.parameters(), 'lr': 10 * args.lr}], args.lr, momentum=args.momentum, weight_decay=args.weight_decay)上述代码表示将模型Net网络的 self.Conv1_2层的学习率设置为传入学习率的10倍，base_params的学习没有明确设置，则默认为传入的学习率args.lr。 注意： [{'params': base_params}, {'params': model.Conv1_2.parameters(), 'lr': 10 * args.lr}]表示为列表中的字典结构。 这种方法设置不同的学习率显得不够灵活，可以为不同的层设置灵活的学习率，可以采用如下方法在adjust_learning_rate函数中设置： def adjust_learning_rate(optimizer, epoch, args): lre = [] lre.extend([0.01] * 10) lre.extend([0.005] * 10) lre.extend([0.0025] * 10) lr = lre[epoch] optimizer.param_groups[0]['lr'] = 0.9 * lr optimizer.param_groups[1]['lr'] = 10 * lr print(param_group[0]['lr']) print(param_group[1]['lr']) 上述代码中的optimizer.param_groups[0]就代表[{'params': base_params}, {'params': model.Conv1_2.parameters(), 'lr': 10 * args.lr}]中的'params': base_params}，optimizer.param_groups[1]代表{'params': model.Conv1_2.parameters(), 'lr': 10 * args.lr}，这里设置的学习率会把args.lr给覆盖掉，个人认为上述代码在设置学习率方面更灵活一些。上述代码也可如下变成实现（注意学习率随便设置的，未与上述代码保持一致）： def adjust_learning_rate(optimizer, epoch, args): lre = np.logspace(-2, -4, 40) lr = lre[epoch] for i in range(len(optimizer.param_groups)): param_group = optimizer.param_groups[i] if i == 0: param_group['lr'] = 0.9 * lr else: param_group['lr'] = 10 * lr print(param_group['lr']) 下面贴出SGD优化器的PyTorch实现，及其每个参数的设置和表示意义，具体如下： import torchfrom .optimizer import Optimizer, required class SGD(Optimizer): r"""Implements stochastic gradient descent (optionally with momentum). Nesterov momentum is based on the formula from `On the importance of initialization and momentum in deep learning`__. Args: params (iterable): iterable of parameters to optimize or dicts defining parameter groups lr (float): learning rate momentum (float, optional): momentum factor (default: 0) weight_decay (float, optional): weight decay (L2 penalty) (default: 0) dampening (float, optional): dampening for momentum (default: 0) nesterov (bool, optional): enables Nesterov momentum (default: False) Example: >>> optimizer = torch.optim.SGD(model.parameters(), lr=0.1, momentum=0.9) >>> optimizer.zero_grad() >>> loss_fn(model(input), target).backward() >>> optimizer.step() __ http://www.cs.toronto.edu/%7Ehinton/absps/momentum.pdf .. note:: The implementation of SGD with Momentum/Nesterov subtly differs from Sutskever et. al. and implementations in some other frameworks. Considering the specific case of Momentum, the update can be written as .. math:: v = /rho * v + g // p = p - lr * v where p, g, v and :math:`/rho` denote the parameters, gradient, velocity, and momentum respectively. This is in contrast to Sutskever et. al. and other frameworks which employ an update of the form .. math:: v = /rho * v + lr * g // p = p - v The Nesterov version is analogously modified. """ def __init__(self, params, lr=required, momentum=0, dampening=0, weight_decay=0, nesterov=False): if lr is not required and lr < 0.0: raise ValueError("Invalid learning rate: {}".format(lr)) if momentum < 0.0: raise ValueError("Invalid momentum value: {}".format(momentum)) if weight_decay < 0.0: raise ValueError("Invalid weight_decay value: {}".format(weight_decay)) defaults = dict(lr=lr, momentum=momentum, dampening=dampening, weight_decay=weight_decay, nesterov=nesterov) if nesterov and (momentum <= 0 or dampening != 0): raise ValueError("Nesterov momentum requires a momentum and zero dampening") super(SGD, self).__init__(params, defaults) def __setstate__(self, state): super(SGD, self).__setstate__(state) for group in self.param_groups: group.setdefault('nesterov', False) def step(self, closure=None): """Performs a single optimization step. Arguments: closure (callable, optional): A closure that reevaluates the model and returns the loss. """ loss = None if closure is not None: loss = closure() for group in self.param_groups: weight_decay = group['weight_decay'] momentum = group['momentum'] dampening = group['dampening'] nesterov = group['nesterov'] for p in group['params']: if p.grad is None: continue d_p = p.grad.data if weight_decay != 0: d_p.add_(weight_decay, p.data) if momentum != 0: param_state = self.state[p] if 'momentum_buffer' not in param_state: buf = param_state['momentum_buffer'] = torch.zeros_like(p.data) buf.mul_(momentum).add_(d_p) else: buf = param_state['momentum_buffer'] buf.mul_(momentum).add_(1 - dampening, d_p) if nesterov: d_p = d_p.add(momentum, buf) else: d_p = buf p.data.add_(-group['lr'], d_p) return loss经验总结：在Fine-tuning时最好不要隔层设置层的参数的可学习与否，这样做一般效果饼不理想，一般准则即可，即先Fine-tuning分类层，学习率设置的大一些，然后在将整个网络设置一个较小的学习率，所有层一起训练。 至于不先经过Fine-tune分类层，而是将整个网络所有层一起训练，只是分类层的学习率相对设置大一些，这样做也可以，至于哪个效果更好，没评估过。当用三元组损失（triplet loss）微调用softmax loss训练的网络时，可以设置阶梯型的较小学习率，整个网络所有层一起训练，效果比较好，而不用先Fine-tune分类层前一层的输出。 以上为个人经验，希望能给大家一个参考，也希望大家多多支持51zixue.net。
Python 如何将integer转化为罗马数(3999以内)
Pytorch中的学习率衰减及其用法详解 |
