(self, x, return_feats=False)
| 361 | self.multiplier_proj = Conv1d1x1(self.middle_channels, self.middle_channels * self.multiplier, bias=c.bias) |
| 362 | |
| 363 | def forward(self, x, return_feats=False): |
| 364 | if self.c.latent_dim is not None and self.mode == 'decoder': |
| 365 | x = self.latent_proj(x) |
| 366 | if self.multiplier != 1: |
| 367 | x = self.multiplier_proj(x) |
| 368 | |
| 369 | feats = [] |
| 370 | for block in self.res_stack: |
| 371 | x = block(x) |
| 372 | if return_feats: |
| 373 | feats.append(x) |
| 374 | if self.c.latent_dim is not None and self.mode == 'encoder': |
| 375 | x = self.latent_proj(x) |
| 376 | if return_feats: |
| 377 | feats.append(x) |
| 378 | if return_feats: |
| 379 | return feats |
| 380 | return x |
| 381 | |
| 382 | def inference(self, x): |
| 383 | for block in self.res_stack: |
nothing calls this directly
no outgoing calls
no test coverage detected