diff options
| author | Jordan Gong <jordan.gong@protonmail.com> | 2021-04-06 22:02:00 +0800 |
|---|---|---|
| committer | Jordan Gong <jordan.gong@protonmail.com> | 2021-04-06 22:02:00 +0800 |
| commit | 5e8947fbc90e1d67dadae36d32330a280d057267 (patch) | |
| tree | f32519b4e6564ddfcdeefb697432646fedc82131 | |
| parent | 7f6c65fa954a43c9d248219525cad35fe1b8a046 (diff) | |
Turn off gradient when decoding images
| -rw-r--r-- | models/rgb_part_net.py | 39 |
1 files changed, 20 insertions, 19 deletions
diff --git a/models/rgb_part_net.py b/models/rgb_part_net.py index ecc38c0..b0169e3 100644 --- a/models/rgb_part_net.py +++ b/models/rgb_part_net.py @@ -57,25 +57,26 @@ class RGBPartNet(nn.Module): if self.training: i_a, i_c, i_p = None, None, None if self.image_log_on: - f_a_mean = f_a.mean(1) - i_a = self.ae.decoder( - f_a_mean, - torch.zeros_like(f_c_mean), - torch.zeros_like(f_p[:, 0]) - ) - i_c = self.ae.decoder( - torch.zeros_like(f_a_mean), - f_c_mean, - torch.zeros_like(f_p[:, 0]) - ) - f_p_size = f_p.size() - i_p = self.ae.decoder( - torch.zeros(f_p_size[0] * f_p_size[1], *f_a.shape[2:], - device=f_a.device), - torch.zeros(f_p_size[0] * f_p_size[1], *f_c.shape[2:], - device=f_c.device), - f_p.view(-1, *f_p_size[2:]) - ).view(x_c1.size()) + with torch.no_grad(): + f_a_mean = f_a.mean(1) + i_a = self.ae.decoder( + f_a_mean, + torch.zeros_like(f_c_mean), + torch.zeros_like(f_p[:, 0]) + ) + i_c = self.ae.decoder( + torch.zeros_like(f_a_mean), + f_c_mean, + torch.zeros_like(f_p[:, 0]) + ) + f_p_size = f_p.size() + i_p = self.ae.decoder( + torch.zeros(f_p_size[0] * f_p_size[1], *f_a.shape[2:], + device=f_a.device), + torch.zeros(f_p_size[0] * f_p_size[1], *f_c.shape[2:], + device=f_c.device), + f_p.view(-1, *f_p_size[2:]) + ).view(x_c1.size()) return x_c, x_p, ae_losses, (i_a, i_c, i_p) else: return x_c, x_p |