用Pycharm调试TensorFlow模型报错解决办法

本人在用Pycharm调试Yolo3模型时，报错，错误如下：

调试

2018-04-18 15:12:10.265921: E C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\stream_executor\cuda\cuda_dnn.cc:385] could not create cudnn handle: CUDNN_STATUS_NOT_INITIALIZED

2018-04-18 15:12:10.266402: E C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\stream_executor\cuda\cuda_dnn.cc:389] error retrieving driver version: Unimplemented: kernel reported driver version not implemented on Windows

2018-04-18 15:12:10.266910: E C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\stream_executor\cuda\cuda_dnn.cc:352] could not destroy cudnn handle: CUDNN_STATUS_BAD_PARAM

2018-04-18 15:12:10.267358: F C:\tf_jenkins\home\workspace\rel-win\M\windows-gpu\PY\36\tensorflow\core\kernels\conv_ops.cc:667] Check failed: stream->parent()->GetConvolveAlgorithms( conv_parameters.ShouldIncludeWinogradNonfusedAlgo<T>(), &algorithms)

解决办法：

重启开机；

报错原因：

非正常停止程序运行，GPU被占用，没有释放；应该正常点击Pycharm界面的停止键，或者加入释放GPU代码；