MenghaoGuo AutoDeeplab issues

MenghaoGuo / AutoDeeplab

Pytorch Implementation the paper Auto-DeepLab Hierarchical Neural Architecture Search for Semantic Image Segmentation

https://arxiv.org/abs/1901.02985

410 stars 97 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

AttributeError: 'Cell' object has no attribute 'cell_ops'

#67 rogressing opened 1 year ago
2
Siyuan dev

#66 JimmyYourHonor closed 2 years ago
1
ModuleNotFoundError: No module named 'modeling'

#65 ZhinanCai closed 3 years ago
0
How can I visualize the best inner cell and outer network after search for my custom dataset?

#64 NdaAzr opened 3 years ago
0
ModuleNotFoundError

#63 zhuxiao490 opened 4 years ago
2
Given groups=1, weight of size 48 160 1 1, expected input[2, 1280, 25, 25] to have 160 channels, but got 1280 channels instead

#62 mx1mx2 opened 4 years ago
0
Still 'CUDA out of memory' with 4 TITAN X (pascal) when training model in PASCAL VOC dataset

#61 rrryan2016 opened 4 years ago
0
what does ABN in "from operations import ABN" in aspp.py ?

#60 NdaAzr closed 4 years ago
0
Not doing an architecture step until epoch > 19

#59 davidhuangal opened 4 years ago
0
missing LICENSE file

#58 maxkazmsft opened 4 years ago
0
train_autodeeplab.py: error: unrecognized arguments: --gpu_ids 0,1,2,3

#57 NanWangAC opened 4 years ago
2
TypeError: 'NoneType' object is not callable

#56 liu-ca opened 5 years ago
9
how to obtain and train the target architecture of searc？

#55 YANGWAGN opened 5 years ago
2
Network Architecture Weights are non zero for connections that do not exist

#54 manuelbre opened 5 years ago
0
Fix error in model_search.py and add architecture paramater to torch paramrters

#53 manuelbre closed 5 years ago
7
I think the decode part is just a single DFS rather than viterbi algorithm ∩_∩, and definitely can be optimized.

#52 Linfengscat opened 5 years ago
0
The network weights and the architecture weights(coefficient) train together?

#51 Linfengscat opened 5 years ago
2
What does the index of alphas_network mean in auto_deeplab.py?

#50 Linfengscat opened 5 years ago
1
Update auto_deeplab.py

#49 Randylcy closed 5 years ago
0
Revert "Edit ASPP as reproducing the work"

#48 Randylcy closed 5 years ago
1
Edit ASPP as reproducing the work

#47 HankKung closed 5 years ago
0
Update train_autodeeplab.py

#46 Randylcy closed 5 years ago
0
Update __init__.py

#45 Randylcy closed 5 years ago
0
Update train_autodeeplab.py

#44 Randylcy closed 5 years ago
0
Update cityscapes.py

#43 Randylcy closed 5 years ago
0
File "../AutoDeeplab-master2/model_search.py", line 68, in forward s = sum(self._ops[offset+j](h, weights[offset+j]) for j, h in enumerate(states)) RuntimeError: The size of tensor a (81) must match the size of tensor b (82) at non-singleton dime

#42 Randylcy closed 5 years ago
0
File "../AutoDeeplab-master/operations.py", line 90, in forward out = torch.cat([self.conv_1(x), self.conv_2(x[:,:,1:,1:])], dim=1)

#41 Randylcy closed 5 years ago
0
why we should multiply softmax result by 5

#40 Randylcy closed 5 years ago
4
The script

#39 HankKung closed 5 years ago
1
Update train_autodeeplab.py

#38 Randylcy closed 5 years ago
0
Update architect.py

#37 Randylcy closed 5 years ago
0
Dataloader(Cityscapes)

#36 HankKung closed 5 years ago
0
Update train_autodeeplab.py

#35 Randylcy closed 5 years ago
0
self.train_loader1, self.train_loader2 in train_autodeeplab.py

#34 Randylcy closed 5 years ago
1
about the crop image size for input

#33 Randylcy closed 5 years ago
7
how many epoches should we set to make the train loss to be convergent?

#32 Randylcy closed 5 years ago
0
Update auto_deeplab.py

#31 Randylcy closed 5 years ago
0
Update model_search.py

#30 Randylcy closed 5 years ago
0
Update auto_deeplab.py

#29 Randylcy closed 5 years ago
0
Update auto_deeplab.py

#28 Randylcy closed 5 years ago
0
Edited Model Forwarding

#27 HankKung closed 5 years ago
0
Statement

#26 MenghaoGuo closed 5 years ago
0
Fix typos and training procedure

#25 HankKung closed 5 years ago
0
Use GPU '0,1,2,3', but only the card 0 is actually used.

#24 Wangzhuoying0716 opened 5 years ago
7
Single GPU with workers=1, but I still met cuda runtime error.

#23 ShunLu91 closed 5 years ago
1
arch_lr value not provided

#22 vinaysingh5898 opened 5 years ago
2
Very low mIOU value

#21 vinaysingh5898 opened 5 years ago
12
ValueError: not enough values to unpack (expected 5, got 4)

#20 Charlie4zc closed 5 years ago
2
Fixed the bug in dataloader

#19 HankKung closed 5 years ago
0
Softmax applied twice to the archtecture parameters?

#18 yucornetto closed 5 years ago
1