Update README.md

README.md

@@ -69,30 +69,6 @@ class ParallelTCNBlock(nn.Module):
         return torch.relu_(y + res)
 
 
-class TCNComp(nn.Module):
-    def __init__(self, enc_in, d_model, e_layers, kernel_size=4, dropout=0.0):
-        super().__init__()
-        blocks = []
-        for i in range(e_layers):
-            in_ch = enc_in if i == 0 else d_model
-            dil = 2**i
-            blocks.append(
-                ParallelTCNBlock(
-                    in_ch, d_model, kernel_size=kernel_size, dilation=dil, dropout=dropout
-                )
-            )
-        self.tcn = nn.Sequential(*blocks)
-
-    def forward(self, x: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]:
-        B, T, S, _ = x.shape
-        x = x.permute(0, 3, 2, 1).contiguous()
-        y = self.tcn(x)  # [B, d_model, S, T]
-        tcn_out = y.permute(0, 2, 3, 1).reshape(B * S, T, -1)
-        last = y[:, :, :, -1].transpose(1, 2)  # [B, S, d_model]
-
-        return tcn_out, last
-
-
 class TCN(nn.Module):
     """
     Parallel TCN over [B, T, S, F]: