Tensorflow 多 gpu 示例错误：变量 conv1/weights/ExponentialMovingAverage/ 不存在

Question

我正在运行本教程中提到的代码：https://www.tensorflow.org/tutorials/deep_cnn/

我从这里下载了代码：https://github.com/tensorflow/models/tree/master/tutorials/image/cifar10/

我在 ubuntu 14.04 的 AWS 中的 g2.4xlarge 机器上运行代码。单个 gpu 示例运行良好，没有任何错误。

有人可以帮助解决这个问题吗？我正在运行 0.12 版本。

---

ubuntu@ip-xxx-xx-xx-xx:~/pythonworkspace/tensorflowdev/models-master/tutorials/image/cifar10$ python -c 'import tensorflow as tf; print(tf.版本)'

0.12.head

ubuntu@ip-xxx-xx-xx-xx:~/pythonworkspace/tensorflowdev/models-master/tutorials/image/cifar10$ python cifar10_multi_gpu_train.py --num_gpus=2

>> Downloading cifar-10-binary.tar.gz 100.0%
Successfully downloaded cifar-10-binary.tar.gz 170052171 bytes.
Filling queue with 20000 CIFAR images before starting to train. This will take a few minutes.
WARNING:tensorflow:From /home/ubuntu/pythonworkspace/tensorflowdev/models-master/tutorials/image/cifar10/cifar10_input.py:135: image_summary (from tensorflow.python.ops.logging_ops) is deprecated and will be removed after 2016-11-30.
Instructions for updating:
Please switch to tf.summary.image. Note that tf.summary.image uses the node name instead of the tag. This means that TensorFlow will automatically de-duplicate summary names based on the scope they are created in. Also, the max_images argument was renamed to max_outputs.
Filling queue with 20000 CIFAR images before starting to train. This will take a few minutes.
WARNING:tensorflow:From /home/ubuntu/pythonworkspace/tensorflowdev/models-master/tutorials/image/cifar10/cifar10_input.py:135: image_summary (from tensorflow.python.ops.logging_ops) is deprecated and will be removed after 2016-11-30.
Instructions for updating:
Please switch to tf.summary.image. Note that tf.summary.image uses the node name instead of the tag. This means that TensorFlow will automatically de-duplicate summary names based on the scope they are created in. Also, the max_images argument was renamed to max_outputs.
Traceback (most recent call last):
  File "cifar10_multi_gpu_train.py", line 273, in <module>
    tf.app.run()
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/app.py", line 44, in run
    _sys.exit(main(_sys.argv[:1] + flags_passthrough))
  File "cifar10_multi_gpu_train.py", line 269, in main
    train()
  File "cifar10_multi_gpu_train.py", line 210, in train
    variables_averages_op = variable_averages.apply(tf.trainable_variables())
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/moving_averages.py", line 373, in apply
    colocate_with_primary=True)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 110, in create_slot
    return _create_slot_var(primary, val, "")
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 64, in _create_slot_var
    use_resource=_is_resource(primary))
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 1034, in get_variable
    use_resource=use_resource, custom_getter=custom_getter)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 933, in get_variable
    use_resource=use_resource, custom_getter=custom_getter)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 356, in get_variable
    validate_shape=validate_shape, use_resource=use_resource)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 341, in _true_getter
    use_resource=use_resource)
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 671, in _get_single_variable
    "VarScope?" % name)
ValueError: Variable conv1/weights/ExponentialMovingAverage/ does not exist, or was not created with tf.get_variable(). Did you mean to set reuse=None in VarScope?

Answer 1

您可以在这里找到问题的答案： Issue 6220

您需要输入：
with tf.variable_scope(tf.get_variable_scope())
在你的设备上运行的循环前面......

那么，就这样吧：

with tf.variable_scope(tf.get_variable_scope()):
    for i in xrange(FLAGS.num_gpus):
        with tf.device('/gpu:%d' % i):

链接中给出了解释...
这里引用：

当你做 tf.get_variable_scope().reuse_variables() 你设置 当前范围以重用变量。如果您在这样的情况下调用优化器 范围，它试图重用它找不到的槽变量，所以 它抛出一个错误。如果你放一个范围， tf.get_variable_scope().reuse_variables() 只影响那个范围，所以 当你退出它时，你又回到了非重用模式，你 想要。

希望对您有所帮助，如果我需要澄清更多信息，请告诉我。