The note about self.n_ is wrong

[TanTingyi]

(kernel_size, 1)

n = (self.kernelsize - 1) / 2.0 self.n = 2 math.pi torch.arange(-n, 0).view(1, -1) / self.sample_rate

the size of self.n_ is (1, kernel_size/2)

[mravanelli]

Thank you, changed!