Is there a way to read .pt files in C++ without a python wrapper?
Yes there is
int main(int argc, const char* argv[])
{
if(argc != 2) {
std::cerr << "path to the model not mentioned!!";
return -1;
}
std::shared_ptr<torch::jit::script::Module> module = torch::jit::load(argv[1]);
std::vector<torch::jit::IValue> inputs;
inputs.push_back(torch::ones({1,1,29,22,22}));
for (int i=0; i<COUNT; ++i) {
auto output = at::argmax(module->forward(inputs).toTensor());
}
return 0;
}
1 Like
Thanks for your helpful answer.
- Could you share a link to the C++ torch::jit documentation? I can’t find it on the pytorch website.
- is this supported for pytorch 1.0 as well?
- what would you have to import?
@0xFFFFFFFF It is supported for PyTorch 1.0, and this tutorial https://pytorch.org/tutorials/advanced/cpp_export.html best explains how to load a PyTorch model in C++.
1 Like