Torch::save and torch::load compilation errors - no overloaded operator definition

[cdcdcd (Cathal Dillon)]

The solution is here from mhubi (builds but I haven’t tested it):

(libtorch) How to save model in MNIST cpp example?

I think, the example was written prior to the stable release of libtorch. The way you would implement the torch::nn::Module now is as follows struct NetImpl : torch::nn::Module { // replaced Net by NetImpl NetImpl() // replaced Net by NetImpl : conv1(torch::nn::Conv2dOptions(1, 10, /*kernel_size=*/5)), conv2(torch::nn::Conv2dOptions(10, 20, /*kernel_size=*/5)), fc1(320, 50), fc2(50, 10) { register_module("conv1", conv1); …

The key here is inclusion of TORCH_MODULE(Net). See example.