How would this MNIST data look like in 2D or 3D after dimensionality reduction? Let's figure it out! I am going to write the code in Pytorch. I have to say, Pytorch is so much better than other deep learning libraries, such as Theano or Tensorflow. Of course, it is just my personal opinion, so let's not get into this argument here.
What I want to do is to take Pytorch's MNIST example found here, and make some modifications to reduce the data dimension to 2D and plot scattered data. This will be a very good example that shows how to do all the following in Pytorch:
1. Create a custom network
2. Create a custom layer
3. Transfer learning from an existing model
4. Save and load a model
Here is the plot I get from running the code below.
This code is tested on Pytorch 0.3.1.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import math | |
| import os | |
| import numpy as np | |
| import matplotlib.pyplot as plt | |
| import argparse | |
| from torch.nn.parameter import Parameter | |
| import torch | |
| import torch.nn as nn | |
| import torch.nn.functional as F | |
| import torch.optim as optim | |
| from torchvision import datasets, transforms | |
| from torch.autograd import Variable | |
| # Training settings | |
| parser = argparse.ArgumentParser(description='PyTorch MNIST Example') | |
| parser.add_argument('--batch-size', type=int, default=64, metavar='N', | |
| help='input batch size for training (default: 64)') | |
| parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N', | |
| help='input batch size for testing (default: 1000)') | |
| parser.add_argument('--epochs', type=int, default=10, metavar='N', | |
| help='number of epochs to train (default: 10)') | |
| parser.add_argument('--lr', type=float, default=0.01, metavar='LR', | |
| help='learning rate (default: 0.01)') | |
| parser.add_argument('--momentum', type=float, default=0.5, metavar='M', | |
| help='SGD momentum (default: 0.5)') | |
| parser.add_argument('--no-cuda', action='store_true', default=False, | |
| help='disables CUDA training') | |
| parser.add_argument('--seed', type=int, default=1, metavar='S', | |
| help='random seed (default: 1)') | |
| parser.add_argument('--log-interval', type=int, default=10, metavar='N', | |
| help='how many batches to wait before logging training status') | |
| args = parser.parse_args() | |
| args.cuda = not args.no_cuda and torch.cuda.is_available() | |
| torch.manual_seed(args.seed) | |
| if args.cuda: | |
| torch.cuda.manual_seed(args.seed) | |
| kwargs = {'num_workers': 1, 'pin_memory': True} if args.cuda else {} | |
| train_loader = torch.utils.data.DataLoader( | |
| datasets.MNIST('../data', train=True, download=True, | |
| transform=transforms.Compose([ | |
| transforms.ToTensor(), | |
| transforms.Normalize((0.1307,), (0.3081,)) | |
| ])), | |
| batch_size=args.batch_size, shuffle=True, **kwargs) | |
| test_loader = torch.utils.data.DataLoader( | |
| datasets.MNIST('../data', train=False, transform=transforms.Compose([ | |
| transforms.ToTensor(), | |
| transforms.Normalize((0.1307,), (0.3081,)) | |
| ])), | |
| batch_size=args.test_batch_size, shuffle=True, **kwargs) | |
| class Net(nn.Module): | |
| def __init__(self, ndim=2, last_layer=True): | |
| super(Net, self).__init__() | |
| self.last_layer = last_layer | |
| self.ndim = ndim | |
| self.conv1 = nn.Conv2d(1, 10, kernel_size=3) | |
| self.bn1 = nn.BatchNorm2d(10, affine=True) | |
| self.conv2 = nn.Conv2d(10, 10, kernel_size=3) | |
| self.conv3 = nn.Conv2d(10, 20, kernel_size=3) | |
| self.bn2 = nn.BatchNorm2d(20, affine=True) | |
| self.conv4 = nn.Conv2d(20, 20, kernel_size=3) | |
| self.conv4_drop = nn.Dropout2d() | |
| self.fc1 = nn.Linear(320, 50) | |
| self.fc2= nn.Linear(50, ndim) | |
| self.softmax = AngleSoftmax(10) | |
| def forward(self, x): | |
| x = F.elu(self.conv1(x)) | |
| x = self.bn1(x) | |
| x = F.elu(F.max_pool2d(self.conv2(x), 2)) | |
| x = F.elu(self.conv3(x)) | |
| x = self.bn2(x) | |
| x = F.elu(F.max_pool2d(self.conv4_drop(self.conv4(x)), 2)) | |
| x = x.view(-1, 320) | |
| x = F.elu(self.fc1(x)) | |
| x = self.fc2(x) | |
| if self.last_layer is True: | |
| x = self.softmax(x) | |
| return x | |
| class AngleSoftmax(nn.Module): | |
| def __init__(self, out_feature): | |
| super(AngleSoftmax, self).__init__() | |
| self.out_feature = out_feature | |
| self.weight = Parameter(torch.Tensor(1, out_feature)) | |
| self.reset_parameters() | |
| def reset_parameters(self): | |
| stdv = 1. / math.sqrt(self.weight.size(1)) | |
| self.weight.data.uniform_(-stdv, stdv) | |
| def forward(self, x): | |
| u = torch.cos(self.weight) | |
| v = torch.sin(self.weight) | |
| w = torch.cat([u, v], dim=0) | |
| x = x.mm(w) / torch.sum(x**2, dim=1, keepdim=True) ** 0.5 | |
| return F.log_softmax(x) | |
| def train(epoch): | |
| model.train() | |
| for batch_idx, (data, target) in enumerate(train_loader): | |
| if args.cuda: | |
| data, target = data.cuda(), target.cuda() | |
| data, target = Variable(data), Variable(target) | |
| optimizer.zero_grad() | |
| output = model(data) | |
| loss = F.nll_loss(output, target) | |
| loss.backward() | |
| optimizer.step() | |
| if batch_idx % args.log_interval == 0: | |
| print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format( | |
| epoch, batch_idx * len(data), len(train_loader.dataset), | |
| 100. * batch_idx / len(train_loader), loss.data[0])) | |
| def test(): | |
| model.eval() | |
| test_loss = 0 | |
| correct = 0 | |
| for data, target in test_loader: | |
| if args.cuda: | |
| data, target = data.cuda(), target.cuda() | |
| data, target = Variable(data, volatile=True), Variable(target) | |
| output = model(data) | |
| test_loss += F.nll_loss(output, target, size_average=False).data[0] # sum up batch loss | |
| pred = output.data.max(1, keepdim=True)[1] # get the index of the max log-probability | |
| correct += pred.eq(target.data.view_as(pred)).long().cpu().sum() | |
| test_loss /= len(test_loader.dataset) | |
| print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format( | |
| test_loss, correct, len(test_loader.dataset), | |
| 100. * correct / len(test_loader.dataset))) | |
| def plot_features(model): | |
| model.eval() | |
| for data, target in test_loader: | |
| if args.cuda: | |
| data, target = data.cuda(), target.cuda() | |
| data, target = Variable(data, volatile=True), Variable(target) | |
| output = model(data) | |
| output = output.cpu().data.numpy() | |
| target = target.cpu().data.numpy() | |
| for label in range(10): | |
| idx = target == label | |
| plt.scatter(output[idx,0], output[idx,1]) | |
| plt.legend(np.arange(10, dtype=np.int32)) | |
| plt.show() | |
| break | |
| if __name__ == '__main__': | |
| model = Net(2) | |
| if args.cuda: | |
| model.cuda() | |
| optimizer = optim.SGD(model.parameters(), lr=args.lr, momentum=args.momentum) | |
| model_file = 'model.h5' | |
| if not os.path.isfile(model_file): | |
| for epoch in range(1, args.epochs + 1): | |
| train(epoch) | |
| test() | |
| torch.save(model, model_file) | |
| else: | |
| model = torch.load(model_file) | |
| feature_model = Net(2, last_layer=False) | |
| feature_model.load_state_dict(model.state_dict()) | |
| if args.cuda: | |
| feature_model.cuda() | |
| plot_features(feature_model) |
