Error: ‘synchronize’ is not a member of ‘torch::cuda’

C++
1

i am using libtorch (1.2),but i can run std::cout << “cuda is used:” << torch::cuda::is_available() and torch::cuda::cudnn_is_available() . i dont know why can’t run torch::cuda::synchronize().

type or paste code here
```int main(int agrc, const char *agrv[]) {
    torch::Tensor tensor = torch::rand({2, 3});
    std::cout << tensor << std::endl;
    std::cout << "cuda is used:" << torch::cuda::is_available() << std::endl;
    std::cout << "cudnn is used:" << torch::cuda::cudnn_is_available() << std::endl;
    torch::cuda::synchronize();
    torch::DeviceType device_type;
    device_type = torch::kCUDA;
    torch::Device device(device_type);
    std::cout << "cuda:" << device_type << std::endl;
  
    torch::jit::script::Module module = torch::jit::load("/home/yu/code/model22.pt", device);
      
      // Deserialize the ScriptModule from a file using torch::jit::load().
    std::vector<torch::jit::IValue> inputs;
    //start = clock();
    inputs.push_back(torch::ones({1, 3, 224, 224}).to(device));
    //end =clock();
    //std::cout << (double)(end-start)/ CLOCKS_PER_SEC << "seconds" << std::endl;
  
    std::cout << "cuda used at:" << torch::cuda::device_count() << std::endl;
  
    
      // Exectute the model
    at::Tensor output = module.forward(inputs).toTensor();
  
    std::cout << output.slice(/*dims=*/1, /*start=*/0, /*end=*/5) << '\n';
    
    std::cout << "ok\n";

  
}
2

i already add
include <torch/torch.h>

//#include <torch/script.h>

#include

#include

#include <ATen/Context.h>

#include <torch/script.h>

#include <torch/cuda.h>

#include

#include but still not work.

type or paste code here
```/home/yuy/code/test12/pytorch/example-app/example-app.cpp: In function ‘int main(int, const char**)’:
/home/yu/code/test12/pytorch/example-app/example-app.cpp:29:16: error: ‘synchronize’ is not a member of ‘torch::cuda’
   torch::cuda::synchronize();
                ^~~~~~~~~~~
CMakeFiles/example-app.dir/build.make:62: recipe for target 'CMakeFiles/example-app.dir/example-app.cpp.o' failed
make[2]: *** [CMakeFiles/example-app.dir/example-app.cpp.o] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/example-app.dir/all' failed
make[1]: *** [CMakeFiles/example-app.dir/all] Error 2
Makefile:83: recipe for ta
3

If you have trouble using this op you could try at::detail::getCUDAHooks().deviceSynchronize(device_index); or

CUDAStream stream = getCurrentCUDAStream();
AT_CUDA_CHECK(cudaStreamSynchronize(stream));

instead.