issues
search
NVlabs
/
edm
Elucidating the Design Space of Diffusion-Based Generative Models (EDM)
Other
1.27k
stars
135
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Using EDM for end-to-end training of latent diffusion models
#28
pandekan
opened
1 month ago
0
Question about Fourier Embedding and EDM Loss calculation
#27
zhengyu-su
closed
2 months ago
1
Is it possible to also share the code for Figure 1b (ideal denoiser outputs) in the paper
#26
tien-th
opened
4 months ago
0
AssertionError: Torch not compiled with CUDA enabled
#25
pierrelouisdelx
closed
5 months ago
0
About hyper-parameters selection
#24
JincanDeng
opened
6 months ago
0
Different Phases of Sampling
#23
coopersvajda
opened
7 months ago
0
On checkpoint for 256*256 images
#22
mengxiangming
opened
7 months ago
0
Question about parameter tuning
#21
xiong-creator
opened
8 months ago
1
Score calculation
#20
rrevoid
closed
8 months ago
0
Fast sampling
#19
daniil-shlenskii
closed
8 months ago
0
ERROR when using multi-gpu training
#18
tsWen0309
closed
9 months ago
7
Question about Augmentation
#17
hojaeklee
opened
9 months ago
1
Can not reproduct the result in Table 2.
#16
Zyriix
opened
9 months ago
5
Could you please provide us imagenet-uncond checkpoint?
#15
huanranchen
opened
11 months ago
0
About the Checkpoint
#14
SYZhang0805
opened
1 year ago
0
VP checkpoints are trained using VE-scaling?
#13
huanranchen
closed
1 year ago
2
Is it possible to also share the code for Figure 3 in the paper
#12
yuanzhi-zhu
closed
1 year ago
1
params[x] in this process with sizes [x, x] appears not to match sizes of the same param in process 0
#11
forever208
closed
1 year ago
2
Fixed modes
#10
KevinRojas1499
closed
1 year ago
0
Is the original loss of VP/VE implemented (i.e. network predict epsilon)?
#9
forever208
closed
1 year ago
1
.
#8
philgzl
closed
1 year ago
0
DataLoader worker (pid xxxx) is killed by signal
#7
Zhendong-Wang
closed
1 year ago
1
How to select options shown in Table2 for training VP/VE models
#6
MingyuKim87
opened
1 year ago
1
Wrong beta_D coefficient for VP during sampling?
#5
qiyan98
opened
1 year ago
1
Deterministic sampling and sampling steps
#4
FutureXiang
opened
1 year ago
0
Zero initialization of convolutions
#3
nicolas-dufour
opened
1 year ago
1
Possible error in the up-sampling function
#2
Msadat97
closed
1 year ago
1
Question about the parameterizations of VP-SDE
#1
FutureXiang
closed
1 year ago
1