My neural network has two inputs , and I don’t know if I use the api rightly to put 2 images as inputs.This is the line of my code:
torch::jit::script::Module module = torch::jit::load("/media/usr515/psmnet_256.pt" , torch::kCUDA);
torch::Tensor result = module.forward({tensor_image , tensor_image_1}).toTensor();
tensor_image and tensor_image_1 are the images having converted to tensors.