这是pytorch初学者的游乐场，其中包含流行数据集上的预定义模型。 目前我们支持 mnist，svhn cifar10，cifar100 stl10 亚历克斯网 vgg16，vgg16_bn，vgg19，vgg19_bn resnet18，resnet34，resnet50，resnet101，resnet152 squeezenet_v0，squeezenet_v1 inception_v3 这是MNIST数据集的示例。 这将自动下载数据集和预先训练的模型。 import torch from torch.autograd import Variable from utee import selector model_raw, ds_fetcher, is_imagenet = selector.select('mnist') ds_val = ds_fetcher(b

The following additional files are required for this exercise: A subset of the STL10 Dataset (stlSubset.zip) Starter Code (cnn_exercise.zip) You will also need: sparseAutoencoderLinear.m or your saved features from Exercise:Learning color features with Sparse Autoencoders feedForwardAutoencoder.m (and related functions) from Exercise:Self-Taught Learning softmaxTrain.m (and related functions) from Exercise:Softmax Regression If you have not completed the exercises listed above, we strongly suggest you complete them first.

用于cnn的 STL10Features.mat 数据集。 subSTL缺少10Features的数据，这个是

In all the exercises so far, you have been working only with grayscale images. In this exercise, you will get to work with RGB color images for the first time. Conveniently, the fact that an image has three color channels (RGB), rather than a single gray channel, presents little difficulty for the sparse autoencoder. You can just combine the intensities from all the color channels for the pixels into one long vector, as if you were working with a grayscale image with 3x the number of pixels as the original image.

直接从下载 STL10 数据集并将它们转换为 Torch 表。 基于 STL-10 格式 写入3个文件：stl10-train.t7, stl10-test.t7, stl10-unlabeled.t7 每个文件都是一个表格形式： th > stl10 = torch. load ( ' stl10-train.t7 ' ) th > print (stl10) { data : ByteTensor - size: 5000 x3x96x96 label : ByteTensor - size: 5000 } Unlabeled 文件只有数据字段