Spaces:

qgyd2021
/

nx_denoise

Running

HoneyTian commited on Feb 13

Commit

0be4793

1 Parent(s): 3a9b74d

update

Files changed (1) hide show

toolbox/torchaudio/models/spectrum_dfnet/modeling_spectrum_dfnet.py CHANGED Viewed

@@ -829,8 +829,7 @@ class SpectrumDfNet(nn.Module):
         # # spec shape: [batch_size, 1, time_steps, spec_bins, 2]
         # spec = spec.detach()
-        # e0, e1, e2, e3, emb, c0, lsnr, h = self.encoder.forward(feat_power, feat_spec)
-        e0, e1, e2, e3, emb, lsnr, h = self.encoder.forward(feat_power, feat_spec)
         mask = self.decoder.forward(emb, e3, e2, e1, e0)
         # mask shape: [batch_size, 1, time_steps, spec_bins]

         # # spec shape: [batch_size, 1, time_steps, spec_bins, 2]
         # spec = spec.detach()
+        e0, e1, e2, e3, emb, c0, lsnr, h = self.encoder.forward(feat_power, feat_spec)
         mask = self.decoder.forward(emb, e3, e2, e1, e0)
         # mask shape: [batch_size, 1, time_steps, spec_bins]