wassname/denoising-diffusion-pytorch
1
0
Fork 0
mirror of https://github.com/wassname/denoising-diffusion-pytorch.git synced 2026-09-13 12:31:15 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update denoising_diffusion_pytorch.py

Browse Source 
Shouldn't eta be 0 for DDIM sampling? In the DDIM paper in section 5.1 they mention the eta=0 is DDIM, and eta=1 approximates the normal DDPM. They also use [eta=0](https://github.com/openai/improved-diffusion/blob/e94489283bb876ac1477d5dd7709bbbd2d9902ce/improved_diffusion/gaussian_diffusion.py#L487) in the openai implementation of DDIM.
This commit is contained in:
wassname
2022-12-22 10:57:30 +08:00
committed by GitHub
parent cf4f44909f
commit 207d23af86
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
denoising_diffusion_pytorch/denoising_diffusion_pytorch.py
+1 -1
View File
@@ -431,7 +431,7 @@ class GaussianDiffusion(nn.Module):
beta_schedule = 'cosine',
p2_loss_weight_gamma = 0., # p2 loss weight, from https://arxiv.org/abs/2204.00227 - 0 is equivalent to weight of 1 across time - 1. is recommended
p2_loss_weight_k = 1,
ddim_sampling_eta = 1.
ddim_sampling_eta = 0.
):
super().__init__()
assert not (type(self) == GaussianDiffusion and model.channels != model.out_dim)