from keras.dataset import mnist
mnist.load_data('minist.npz')
用上述代码下载数据集时，会连接到https://s3.amazonaws.com/text-datasets/minist.npz下载，速度极慢，解决办法如下：
网上搜索， 很多给出的办法是先用迅雷等从https://s3.amazonaws.com/text-datasets/minist.npz下载，放到当前目录下或次级目录下，然后改用np.load加载
data = np.load('./datasets/mnist.npz')
train_X, train_y, test_X, test_y = data['x_train'],data['y_train'],data['x_test'],data['y_test']
这个方法可以解决，但不建议
因为后面还有其它的数据集如mdb.load_data('imdb.npz',num_words=10000)
这个还好，还有：word_index = imdb.get_word_index()
最好的办法是放到keras的缓存文件夹
即用户路径的.keras文件夹，win10的keras缓存文件夹路径：C:\Users\Administrator.keras\datasets
其它系统应该也是在类似目录下
建议用迅雷从上述地址下载，然后放到缓存文件夹
则不会从网络上下载mnist.npz文件
c:\Python36\Lib\site-packages\keras\datasets\mnist.py
from ..utils.data_utils import get_file
import numpy as np
def load_data(path='mnist.npz'):
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)
imdb.py cifar.py boston_housing.py等也在此目录下
