You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
# setup mel filterbank
if self.scale == "mel":
assert sample_rate != None # Must set sample rate to use mel scale
assert n_bins <= fft_size # Must be more FFT bins than Mel bins
fb = librosa.filters.mel(sample_rate, fft_size, n_mels=n_bins)
self.fb = torch.tensor(fb).unsqueeze(0)
elif self.scale == "chroma":
assert sample_rate != None # Must set sample rate to use chroma scale
assert n_bins <= fft_size # Must be more FFT bins than chroma bins
fb = librosa.filters.chroma(sample_rate, fft_size, n_chroma=n_bins)
self.fb = torch.tensor(fb).unsqueeze(0)
if scale is not None and device is not None:
self.fb = self.fb.to(self.device) # move filterbank to device
This causes an issue when trying to compute the loss term on GPU as the self.fb object will not get moved automatically to the correct device. This is simple to resolve and should only require registering the filterbank object as a buffer.
Something like self.register_buffer("fb", fb)
The text was updated successfully, but these errors were encountered:
This causes an issue when trying to compute the loss term on GPU as the
self.fb
object will not get moved automatically to the correct device. This is simple to resolve and should only require registering the filterbank object as a buffer.Something like
self.register_buffer("fb", fb)
The text was updated successfully, but these errors were encountered: