Check failed: cudaSuccess == err (0 vs. 8) [hl_gpu_apply_unary_op failed] CUDA error: invalid device function
Created by: likeqinqin
基础环境：C++动态链接库、p4卡、cuda8.0、driver384.81、cudnn7.0
具体报错信息如下：
F0524 16:04:56.877745 625 hl_gpu_matrix_kernel.cuh:181] Check failed: cudaSuccess == err (0 vs. 8) [hl_gpu_apply_unary_op failed] CUDA error: invalid device function
* Check failure stack trace: *
@ 0x7f5011eefd3d google::LogMessage::Fail()
@ 0x7f5011ef37ec google::LogMessage::SendToLog()
@ 0x7f5011eef863 google::LogMessage::Flush()
@ 0x7f5011ef4cfe google::LogMessageFatal::~LogMessageFatal()
@ 0x7f50121bf3bb hl_gpu_apply_unary_op<>()
@ 0x7f50121bf72d paddle::BaseMatrixT<>::applyUnary<>()
@ 0x7f50121bf973 paddle::BaseMatrixT<>::zero()
@ 0x7f5012096d18 ZNSt17_Function_handlerIFviPN6paddle9ParameterEEZNS0_15GradientMachine6createERKNS0_11ModelConfigEiRKSt6vectorINS0_19enumeration_wrapper13ParameterTypeESaISA_EEEUliS2_E_E9_M_invokeERKSt9_Any_dataiS2
@ 0x7f501206e3a6 paddle::NeuralNetwork::init()
@ 0x7f5012096f9f paddle::GradientMachine::create()
@ 0x7f5011eeb4e5 paddle_gradient_machine_create_for_inference
