openai pixel-cnn issues

openai / pixel-cnn

Code for the paper "PixelCNN++: A PixelCNN Implementation with Discretized Logistic Mixture Likelihood and Other Modifications"

https://arxiv.org/abs/1701.05517

Other

1.92k stars 436 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Does the implementation satisfy the causailty constraint?

#58 Weafre opened 2 years ago
0
How many epochs to train for CIFAR-10

#57 erikdao opened 2 years ago
0
How did you use conditional class label h

#56 zhangdan8962 opened 3 years ago
0
How to run pretrained model checkpoint.

#55 CaiShilv opened 3 years ago
0
How to run pretrained model checkpoint.

#54 sharmayush01 opened 4 years ago
0
why your code so weird?

#53 johnzlj opened 4 years ago
0
GPU error & other things

#52 cymqqqq opened 4 years ago
0
Pretrained model not available

#51 vikmary opened 4 years ago
0
Object_detection_image

#50 kriz232121 opened 4 years ago
0
Model builder

#49 slesarev-hub closed 4 years ago
0
Object was never used (type <class 'tensorflow.python.framework.ops.Tensor'>):

#48 kriz232121 opened 4 years ago
1
what do these lines work for

#47 share020 opened 4 years ago
0
Dead link to pretrained model.

#46 AlexanderMath opened 5 years ago
0
about the shift operation

#45 yuffon closed 5 years ago
0
samples

#44 yuffon closed 5 years ago
0
Add Apache License 2.0

#43 atpaino closed 5 years ago
1
Create requirements file for the project

#42 bharatr21 opened 5 years ago
0
How are weights trained?

#41 profpedromoises opened 5 years ago
0
update README with repo status

#40 christopherhesse closed 5 years ago
0
Training fails with other datasets

#39 haihongtran opened 6 years ago
0
why has to change to discretisized loss function?

#38 xfwu opened 6 years ago
1
fix the data dependent initializaiton.

#37 bfs18 closed 5 years ago
0
The output of conv2d should be updated after g and b are updated in data dependent initialization.

#36 bfs18 opened 6 years ago
2
Pretrained model download link broken

#35 huanzhang12 opened 6 years ago
7
Understanding edge cases

#34 ThomasDelteil opened 6 years ago
0
Not understand plus_in = inv_stdv * (centered_x + 1./255.) in discretized_mix_logistic_loss

#33 blues-lin opened 6 years ago
4
Learning curve isn't discontinuous when i continued training it on a pre-trained model

#32 powder21 closed 6 years ago
1
7000 loss value when training cifar10

#31 zoli333 opened 6 years ago
0
Gated unit formulation

#30 ngimel closed 1 year ago
1
unexpected keyword argument 'keepdims'

#29 knauth opened 6 years ago
3
Logistic output layer training slowly

#28 weixsong closed 6 years ago
2
Training results in NaN (commit 2b03725)

#27 bliep opened 6 years ago
3
error while training

#26 ELHoussineT closed 2 years ago
3
not understand "log probability in the center of the bin"

#25 Beronx86 opened 6 years ago
2
How to understand the softmax sampling

#24 weixsong closed 6 years ago
2
how to understand the loss function corner case for 0 and 255

#23 weixsong closed 6 years ago
3
Training pixelCNN on tiny imagenet dataset

#22 ShirinVafaei opened 6 years ago
0
Memory cost too much and do not start to train when using tensorflow1.4

#21 sunume opened 7 years ago
4
Data loader bug fixes

#20 durana opened 7 years ago
0
Training not starting for 64x64x3 images

#19 christopherblum opened 7 years ago
1
Configuration setup for executing the implementation

#18 mrincs closed 7 years ago
1
"Object was never used" errors after upgrading to Tensorflow 1.2

#17 malcolmsharpe opened 7 years ago
39
PixelCNN source code

#16 FuyuWang opened 7 years ago
0
filter sizes

#15 JohannesBrand opened 7 years ago
0
Windows: No one_hot kernel in tensorflow-gpu-v1.0

#14 DawyD opened 7 years ago
1
about pixel-cnn/model.py, line38

#13 Lucas2012 opened 7 years ago
2
Moving to Tensorflow 1.0 API.

#12 prichemond closed 7 years ago
2
update to work with TensorFlow-1.0

#11 liamocn closed 7 years ago
1
Speeding up image generation up to 183x

#10 PrajitR opened 7 years ago
0
is it possible to load the model in cpu and just do a forward pass on an image?

#9 kirk86 opened 7 years ago
6