python - TensorFlow 崩溃 - 中止(核心转储)

Question

我在 TensorFlow 中收到以下错误:

2017-06-27 03:10:50.310215: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2017-06-27 03:10:50.310485: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties: 
name: GeForce GTX 1060 6GB
major: 6 minor: 1 memoryClockRate (GHz) 1.7715
pciBusID 0000:01:00.0
Total memory: 5.92GiB
Free memory: 5.43GiB
2017-06-27 03:10:50.310498: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 
2017-06-27 03:10:50.310502: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0:   Y 
2017-06-27 03:10:50.310510: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:01:00.0)
2017-06-27 03:10:50.804537: W tensorflow/core/framework/op_def_util.cc:332] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
2017-06-27 03:11:04.026835: E tensorflow/stream_executor/cuda/cuda_dnn.cc:352] Loaded runtime CuDNN library: 6021 (compatibility version 6000) but source was compiled with 5110 (compatibility version 5100).  If using a binary install, upgrade your CuDNN library to match.  If building from sources, make sure the library loaded at runtime matches a compatible version specified during compile configuration.
2017-06-27 03:11:04.027028: F tensorflow/core/kernels/conv_ops.cc:671] Check failed: stream->parent()->GetConvolveAlgorithms(&algorithms) 
Aborted (core dumped)

这是运行示例神经网络后的控制台输出。有任何想法吗？是否是版本不匹配，如果是，我该如何修复？我几乎可以肯定我拥有最新版本的 cuDNN 和 CUDA 工具包。非常感谢任何帮助:)

Answer 1

Tensorflow 已编译为可与 cuDNN 版本 5.1 配合使用，但您为其提供版本 6。一般来说，主要版本号的变化也会伴随着API的变化。

您需要使用tensorflow期望的cuDNN版本，5.1。