issues
search
lucidrains
/
recurrent-interface-network-pytorch
Implementation of Recurrent Interface Network (RIN), for highly efficient generation of images and video without cascading networks, in Pytorch
MIT License
194
stars
14
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
follow up paper
#22
lucidrains
opened
2 months ago
8
ValueError: num_samples should be a positive integer value, but got num_samples=0
#21
ad48hp
opened
7 months ago
1
Compute required to get a minimal setup running
#20
kushalj001
closed
1 month ago
0
scale during sampling
#19
zhangdan8962
closed
9 months ago
3
mixed precision training
#18
nicolas-dufour
opened
1 year ago
13
Normalizing the context for cross-attention
#17
yoavhacohen
closed
1 year ago
2
Fix Accelerate Multiprocessing Issues: Model Training/Saving/Loading
#16
leon-w
closed
1 year ago
1
Inconsistent interpretation of model output between self-conditioning step and prediction step
#15
jsternabsci
opened
1 year ago
1
Fix has_int_squareroot
#14
leon-w
closed
1 year ago
1
Naming convention
#13
katop1234
opened
1 year ago
0
Block-wise positional embedding
#12
katop1234
opened
1 year ago
0
Min SNR weighting implemented as max SNR weighting
#11
justinlovelace
closed
1 year ago
4
Handling of the time conditioning
#9
nicolas-dufour
closed
1 year ago
9
Potential Issue with Self-Conditioning
#8
justinlovelace
closed
1 year ago
2
RIN with 100k steps, still noisy images?
#7
Samin100
closed
1 year ago
45
Possible Normalization Bug
#6
MattMcPartlon
closed
1 year ago
4
use maybe_normalized image for calculations
#5
juvekaradheesh
closed
1 year ago
3
small bugfix regarding saving the step
#4
alex-haas
closed
1 year ago
1
Lacks a `ff` layer?
#3
CiaoHe
closed
1 year ago
1
notes
#2
ajabri
closed
1 year ago
39
Could you share more thoughts on RIN?
#1
w-hc
closed
1 year ago
0