启动内核时出错

Question

我已经安装了所有必要的软件(opencv，tensorflow-gpu，matplotlib，scikit-learn，pandas，keras 2(来运行我的代码并验证了它们。我正在使用Spyder作为IDE，并使用Tensorflow后端在Keras中训练CNN。我可以运行我的代码片段，直到我进入训练阶段：

hist = model.fit(X_train, y_train, batch_size=32, nb_epoch=num_epoch, verbose=1, validation_data=(X_test, y_test))

当我运行此行时，训练会开始一些，而不是显示纪元和其他属性(val_acc、training_acc 等(，内核突然死亡，然后重新连接到内核并再次死亡，等等。最后我收到此错误：

2018󈚧󈚭 16:25:49.961500: I C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcoreplatformcpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2018󈚧󈚭 16:25:50.664501: I C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcorecommon_runtimegpugpu_device.cc:1212] Found device 0 with properties: 
name: GeForce GT 740 major: 3 minor: 0 memoryClockRate(GHz): 1.0715
pciBusID: 0000:01:00.0
totalMemory: 1.00GiB freeMemory: 756.79MiB
2018󈚧󈚭 16:25:50.664501: I C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcorecommon_runtimegpugpu_device.cc:1312] Adding visible gpu devices: 0
2018󈚧󈚭 16:25:51.148102: I C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcorecommon_runtimegpugpu_device.cc:993] Creating TensorFlow device (/device:GPU:0 with 501 MB memory) ‑> physical GPU (device: 0, name: GeForce GT 740, pci bus id: 0000:01:00.0, compute capability: 3.0)
2018󈚧󈚭 16:27:22.549779: I C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcorecommon_runtimegpugpu_device.cc:1312] Adding visible gpu devices: 0
2018󈚧󈚭 16:27:22.549779: I C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcorecommon_runtimegpugpu_device.cc:993] Creating TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 224 MB memory) ‑> physical GPU (device: 0, name: GeForce GT 740, pci bus id: 0000:01:00.0, compute capability: 3.0)
2018󈚧󈚭 16:27:43.118021: E C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowstream_executorcudacuda_dnn.cc:378] Loaded runtime CuDNN library: 7101 (compatibility version 7100) but source was compiled with 7003 (compatibility version 7000). 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.
2018󈚧󈚭 16:27:43.164821: F C:tf_jenkinsworkspacerel‑winMwindows‑gpuPY35tensorflowcorekernelsconv_ops.cc:717] Check failed: stream‑>parent()‑>GetConvolveAlgorithms( conv_parameters.ShouldIncludeWinogradNonfusedAlgo(), &algorithms)

我虽然这是一个Spyder问题并在github上发布并收到了与Spyder无关但兼容性问题的回复

我在网上搜索了希望能找到解决方案，但似乎没有完全相同的问题。(至少在我遇到过(

如果有人遇到同样的问题，请帮助我。我应该怎么做？

Answer 1

当我使用 Jupyter 笔记本时，我遇到了同样的问题，对我来说，修复的是更改运行代码的浏览器。
如果使用不同的IDE(例如Jupyter笔记本，pycharm(不起作用，则建议从终端/命令提示符运行脚本。