keras中无法下载 https://s3.amazonaws.com/img-datasets/mnist.npz 解决方法

网址：https://s3.amazonaws.com/img-datasets/mnist.npz，由于显而易见的原因，无法访问。
npz实际上是numpy提供的数组存储方式，简单的可看做是一系列npy数据的组合，利用np.load函数读取后得到一个类似字典的对象，可以通过关键字进行值查询，关键字对应的值其实就是一个npy数据。
如果用keras自带的example（from keras.datasets import mnist，在mnist.py下的load_data函数），会使用这种格式。

 

我自己解决方法是在国外的vps机器上下载，然后传到本地，假设保存为mnist.npz，则加载方法：

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

import numpy as np def load_data(path='mnist.npz'):     """Loads the MNIST dataset.     # Arguments         path: path where to cache the dataset locally             (relative to ~/.keras/datasets).     # Returns         Tuple of Numpy arrays: `(x_train, y_train), (x_test, y_test)`.     path = get_file(path,                     origin='https://s3.amazonaws.com/img-datasets/mnist.npz',                     file_hash='8a61469f7ea1b51cbae51d4f78837e45')     """     f = np.load(path)     x_train, y_train = f['x_train'], f['y_train']     x_test, y_test = f['x_test'], f['y_test']     f.close()     return (x_train, y_train), (x_test, y_test)   # the data, split between train and test sets (x_train, y_train), (x_test, y_test) = load_data()

 

原来的是：

?

1	(x_train, y_train), (x_test, y_test) = mnist.load_data()

 替换下OK!