Question

我已经安装了TensorFlow。当我下载MNISTdataset时，存在一个错误。谁能告诉我出了什么问题？非常感谢！错误详情如下：

Python 2.7.9 (default, Apr  2 2015, 15:33:21) 
[GCC 4.9.2] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import input_data
>>> mnist = input_data.read_data_sets("MNIST_data/", False, False)
Extracting MNIST_data/train-images-idx3-ubyte.gz
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "input_data.py", line 162, in read_data_sets
    local_file = maybe_download(TRAIN_LABELS, train_dir)
  File "input_data.py", line 22, in maybe_download
    filepath, _ = urllib.request.urlretrieve(SOURCE_URL + filename, filepath)
  File "/usr/lib/python2.7/urllib.py", line 98, in urlretrieve
    return opener.retrieve(url, filename, reporthook, data)
  File "/usr/lib/python2.7/urllib.py", line 245, in retrieve
    fp = self.open(url, data)
  File "/usr/lib/python2.7/urllib.py", line 213, in open
    return getattr(self, name)(url)
  File "/usr/lib/python2.7/urllib.py", line 351, in open_http
    errcode, errmsg, headers = h.getreply()
  File "/usr/lib/python2.7/httplib.py", line 1202, in getreply
    response = self._conn.getresponse()
  File "/usr/lib/python2.7/httplib.py", line 1127, in getresponse
    response.begin()
  File "/usr/lib/python2.7/httplib.py", line 453, in begin
    version, status, reason = self._read_status()
  File "/usr/lib/python2.7/httplib.py", line 409, in _read_status
    line = self.fp.readline(_MAXLINE + 1)
  File "/usr/lib/python2.7/socket.py", line 480, in readline
    data = self._sock.recv(self._rbufsize)
IOError: [Errno socket error] [Errno 104] Connection reset by peer

Answer 1

Python无法从lecun.com下载MNIST数据集。首先，检查以确保您可以从该计算机浏览Yann LeCun's MNIST page。如果不能，则可能是防火墙或Internet连接问题。如果可以，请尝试在几分钟内再次运行下载 - 自Tensorflow发布以来，我已经短暂地看到了这个错误，并且它总是在5分钟内消失。

Answer 2

如果网站即使在数小时后仍然无法访问，您可以修改/usr/local/lib/python2.7/dist-packages/tensorflow/contrib/learn/python/learn/datasets/mnist.py并更改源链接到＆＃39; https://s3.amazonaws.com/lasagne/recipes/datasets/mnist/＆＃39;并再次运行该示例。

Answer 3

作为@ user3851234的回答，您可以通过更改SOURCE_URL来下载数据集，例如：

SOURCE_URL = 'https://s3.amazonaws.com/lasagne/recipes/datasets/mnist/'

您的选择很少：

官方来源：http://yann.lecun.com/exdb/mnist/
Google镜像：https://storage.googleapis.com/cvdf-datasets/mnist/
亚马逊镜像：https://s3.amazonaws.com/lasagne/recipes/datasets/mnist/

中国程序员的特别说明

谷歌和亚马逊提供的镜子在中国大陆经常失败，你应该先试试Yann LeCun's page。

制作自己的MNIST镜像

的制备：将

无论你在哪里获得4个MNIST数据文件：

train-images-idx3-ubyte.gz：训练集图像（9912422字节）
train-labels-idx1-ubyte.gz：训练集标签（28881字节）
t10k-images-idx3-ubyte.gz：测试集图像（1648877字节）
t10k-labels-idx1-ubyte.gz：测试集标签（4542字节）

并将它们放在您的域名下，例如：

https://github.com/Amy/mnistrepo/mnist/train-images-idx3-ubyte.gz

并将SOURCE_URL更改为您的域名：

SOURCE_URL = 'https://github.com/Amy/mnistrepo/mnist

TensorFlow IOError：[Errno套接字错误] [Errno 104]由对等方重置连接

3 个答案: