Skip to content

Fix preprocessing #25

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
treacker wants to merge 2 commits into
base: 2022
Choose a base branch
from
Open

Fix preprocessing #25

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e0a2440
Fix preprocessing bug
treacker Dec 21, 2022
96dc45e
Merge pull request #1 from treacker/treacker-patch-1
treacker Dec 21, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 4 additions & 3 deletions hw4_nv/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,15 +42,15 @@ class MelSpectrogram(nn.Module):
super(MelSpectrogram, self).__init__()

self.config = config

self.mel_spectrogram = torchaudio.transforms.MelSpectrogram(
sample_rate=config.sr,
win_length=config.win_length,
hop_length=config.hop_length,
n_fft=config.n_fft,
f_min=config.f_min,
f_max=config.f_max,
n_mels=config.n_mels
n_mels=config.n_mels,
center=False,
)

# The is no way to set power in constructor in 0.5.0 version.
Expand All @@ -72,7 +72,8 @@ class MelSpectrogram(nn.Module):
:param audio: Expected shape is [B, T]
:return: Shape is [B, n_mels, T']
"""

audio = torch.nn.functional.pad(audio.unsqueeze(1), (int((self.config.n_fft-self.config.hop_length)/2), int((self.config.n_fft-self.config.hop_length)/2)), mode='reflect')
audio = audio.squeeze(1)
mel = self.mel_spectrogram(audio) \
.clamp_(min=1e-5) \
.log_()
Expand Down