(self, X)
| 1301 | self.layer1.bias = nn.Parameter(torch.FloatTensor(b.flatten())) |
| 1302 | |
| 1303 | def forward(self, X): |
| 1304 | # (N, H, W, C) -> (N, C, H, W) |
| 1305 | self.X = np.moveaxis(X, [0, 1, 2, 3], [0, -2, -1, -3]) |
| 1306 | if not isinstance(self.X, torch.Tensor): |
| 1307 | self.X = torchify(self.X) |
| 1308 | |
| 1309 | self.X.retain_grad() |
| 1310 | |
| 1311 | self.Z = self.layer1(self.X) |
| 1312 | self.Z.retain_grad() |
| 1313 | |
| 1314 | self.Y = self.act_fn(self.Z) |
| 1315 | self.Y.retain_grad() |
| 1316 | return self.Y |
| 1317 | |
| 1318 | def extract_grads(self, X): |
| 1319 | self.forward(X) |
no test coverage detected