Revision history [back]

The method for utilizing at::_upsample_nearest_exact2d in libtorch is as follows:

1. Include the necessary headers:

   #include <torch/torch.h>
   

2. Create a tensor of input:

   torch::Tensor input = torch::rand({1, 1, 2, 2});
   

3. Define the desired output size:

   torch::IntArrayRef output_size = {3, 3};
   

4. Call the at::_upsample_nearest_exact2d function:

   torch::Tensor output = at::_upsample_nearest_exact2d(input, output_size);
   

   The first argument is the input tensor and the second argument is the desired output size.

5. Use the output:

   std::cout << "Input tensor: " << input << std::endl;
   std::cout << "Output tensor: " << output << std::endl;
   

   This will print the input and output tensors to the console.