How to make torch.nn.functional.interpolate deterministic?

autograd
2

Any news on this?
torch.nn.ConvTranspose2d does not work as well as torch.nn.functional.interpolate. I want to upscale some feature maps right away (in a non-parametric way) without doing transposed convolution many times!! this last one seems to lose information!!!

  1. torch.nn.functional.interpolate seems to work but it is not deterministic (unstable results)
  2. torch.nn.ConvTranspose2d does not seem to work, but it is deterministic (stable results).