Fixed format in some files

espnet2/enh/loss/criterions/tf_domain.py

@@ -238,7 +238,7 @@ def forward(self, ref, inf) -> torch.Tensor:
                 mask = reduce(lambda x, y: x * y, flags)
                 mask = mask.int() * i
                 r += mask
-            r = r.contiguous().view(-1,).long()
+            r = r.contiguous().flatten().long()
             re = F.one_hot(r, num_classes=num_spk)
             re = re.contiguous().view(B, -1, num_spk)
         elif self._loss_type == "mdc":
@@ -255,7 +255,11 @@ def forward(self, ref, inf) -> torch.Tensor:
                 )
 
             re = torch.zeros(
-                ref[0].shape[0], ref[0].shape[1], ref[0].shape[2], num_spk, device=inf.device
+                ref[0].shape[0],
+                ref[0].shape[1],
+                ref[0].shape[2],
+                num_spk,
+                device=inf.device,
             )
             for i in range(num_spk):
                 flags = [abs_ref[i] >= n for n in abs_ref]
@@ -270,7 +274,11 @@ def forward(self, ref, inf) -> torch.Tensor:
             )
 
         V2 = torch.matmul(torch.transpose(inf, 2, 1), inf).pow(2).sum(dim=(1, 2))
-        Y2 = torch.matmul(torch.transpose(re, 2, 1).float(), re.float()).pow(2).sum(dim=(1, 2))
+        Y2 = (
+            torch.matmul(torch.transpose(re, 2, 1).float(), re.float())
+            .pow(2)
+            .sum(dim=(1, 2))
+        )
         VY = torch.matmul(torch.transpose(inf, 2, 1), re.float()).pow(2).sum(dim=(1, 2))
 
         return V2 + Y2 - 2 * VY

espnet2/enh/separator/dan_separator.py

@@ -95,7 +95,6 @@ def forward(
                 'mask_spkn': torch.Tensor(Batch, Frames, Freq),
             ]
         """
-
         # if complex spectrum,
         if isinstance(input, ComplexTensor):
             feature = abs(input)
@@ -122,7 +121,7 @@ def forward(
                 Y = reduce(lambda x, y: x * y, flags)
                 Y = Y.int() * i
                 Y_t += Y
-            Y_t = Y_t.contiguous().view(-1,).long()
+            Y_t = Y_t.contiguous().flatten().long()
             Y = Fun.one_hot(Y_t, num_classes=self._num_spk)
             Y = Y.contiguous().view(B, -1, self._num_spk).float()
 

espnet2/enh/separator/dccrn_separator.py

@@ -161,8 +161,8 @@ def __init__(
         self.flatten_parameters()
 
     def forward(
-        self, 
-        input: Union[torch.Tensor, ComplexTensor], 
+        self,
+        input: Union[torch.Tensor, ComplexTensor],
         ilens: torch.Tensor,
         additional: Optional[Dict] = None,
     ) -> Tuple[List[Union[torch.Tensor, ComplexTensor]], torch.Tensor, OrderedDict]:

espnet2/enh/separator/dpcl_separator.py

@@ -93,7 +93,6 @@ def forward(
                 'tf_embedding': learned embedding of all T-F bins (B, T * F, D),
             ]
         """
-
         # if complex spectrum,
         if isinstance(input, ComplexTensor):
             feature = abs(input)

espnet2/enh/separator/dprnn_separator.py

@@ -84,7 +84,7 @@ def forward(
             ilens (torch.Tensor): input lengths [Batch]
             additional (Dict or None): other data included in model
                 NOTE: not used in this model
-                
+
         Returns:
             masked (List[Union(torch.Tensor, ComplexTensor)]): [(B, T, N), ...]
             ilens (torch.Tensor): (B,)

espnet2/enh/separator/neural_beamformer.py

@@ -140,7 +140,7 @@ def forward(
             ilens (torch.Tensor): input lengths [Batch]
             additional (Dict or None): other data included in model
                 NOTE: not used in this model
-   
+
         Returns:
             enhanced speech (single-channel): List[torch.complex64/ComplexTensor]
             output lengths

espnet2/enh/separator/skim_separator.py

@@ -82,8 +82,8 @@ def __init__(
         }[nonlinear]
 
     def forward(
-        self, 
-        input: Union[torch.Tensor, ComplexTensor], 
+        self,
+        input: Union[torch.Tensor, ComplexTensor],
         ilens: torch.Tensor,
         additional: Optional[Dict] = None,
     ) -> Tuple[List[Union[torch.Tensor, ComplexTensor]], torch.Tensor, OrderedDict]: