quark0 darts issues - Githubissues

quark0 / darts

Differentiable architecture search for convolutional and recurrent networks

https://arxiv.org/abs/1806.09055

Apache License 2.0

3.92k stars 843 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

framework question: why is the position of 'Relu' before conv rather than after conv?

#125 xingyueye opened 5 years ago
1
The alpha variable's value of "avg_pool" and "skip_connect" is the mostly biggest over the others

#124 shuangjian24 opened 5 years ago
2
Not able to reproduce the cells in the paper

#123 doitslow opened 5 years ago
1
architecture parameters initialization

#122 JingweiZhang12 opened 5 years ago
0
Two questions about code implementation

#121 TianQi-777 closed 5 years ago
1
What is DART

#120 epriwahyu opened 5 years ago
0
Making the code dataset agnostic

#119 dylanrandle closed 5 years ago
1
param_size of searched model is too heavy

#118 xingyueye opened 5 years ago
1
How to implement embedding + mask in torch 1.12?

#117 pingguokiller opened 5 years ago
2
A issue about Figure 2 in paper

#116 jjjjjie opened 5 years ago
3
How to generate a complete cifar network?

#115 zhulishun opened 5 years ago
0
How to stack cells into a complete architect?

#114 zhulishun opened 5 years ago
0
Implementation of approximate architecture gradient is different with paper.

#113 iuhiyuh closed 5 years ago
1
how to get ppl by using customed configrues'ENAS'

#112 hukaidididi opened 5 years ago
0
I can‘t run the train.py directly

#111 shuotian17 opened 5 years ago
0
I can't reproduce the architecture search process following the configuration in paper.

#110 ghost opened 5 years ago
13
one of the variables needed for gradient computation has been modified by an inplace operation

#109 pawopawo opened 5 years ago
2
what's the meaning of function mask2d in rnn.utils?

#108 pauking closed 5 years ago
1
why the hyperparameters, for both the training pipeline (train_search.py )and the final evaluation pipeline (train.py) differ a lot!?

#107 NdaAzr opened 5 years ago
3
"TypeError: Iteration over 0-d tensor"

#106 ilia10000 opened 5 years ago
4
running time when it compares with Resnet from other code

#105 Margrate opened 5 years ago
1
How to get the hidden cell, i.e. the features, in RNN architecture?

#104 QingLi-Chloe opened 5 years ago
0
much information not mentioned in papers

#103 deeptoby opened 5 years ago
0
How to use the RNN cell search on images

#102 QingLi-Chloe opened 5 years ago
0
Doubt about the effectiveness of the method

#101 twangnh opened 5 years ago
28
Why model storage memory takes a huge amount of space?

#100 haithanhp opened 5 years ago
2
first order performs better than second order and much faster

#99 twangnh closed 5 years ago
0
out of memory for different architecture(pytorch0.3)

#98 skx6 opened 5 years ago
7
pytorch0.3 out of memery fordifferent architecture

#97 skx6 closed 5 years ago
1
support latest pytorch

#96 MandyMo opened 5 years ago
5
Bug!DataLoaderIter ConnectionResetError: [Errno 104] Connection reset by peer

#95 zihaozhang9 opened 5 years ago
2
train_search.py ConnectionResetError

#94 zihaozhang9 opened 5 years ago
1
ScheduledDropPath

#93 skmhrk1209 closed 4 years ago
0
ImageNet reproduce results in paper

#92 haithanhp opened 5 years ago
14
What does the "_bns" in model_train.py L40 do?

#91 Catosine opened 5 years ago
0
What changes need to be made to be compatible with pytorch 0.4.0

#90 ghost opened 5 years ago
2
AttributeError: module 'torch' has no attribute 'float32'

#89 FrankYoungchen opened 5 years ago
0
pytorch 1.0+ compatibility

#88 nk0307 opened 5 years ago
1
Train arch_parameters with validation data?

#87 Harry-Up closed 5 years ago
4
Darts for nni

#86 WdBlink opened 5 years ago
0
It seems the global moving average is used as opposed to what the original paper instructed

#85 bombs-kim opened 5 years ago
0
A bug on FactorizedReduce in operations.py?

#84 Dav-Jay opened 5 years ago
3
Reimplementation for rnn using fastai

#83 tinhb92 opened 5 years ago
2
Architecture parameter initialization

#82 JaminFong opened 5 years ago
1
Is it possible to search with larger image size like 224*224?

#81 R00Kie-Liu opened 5 years ago
1
Searched architecture training time.

#80 zhaohui-yang opened 5 years ago
9
How do you learn reduction cell?

#79 imirzadeh opened 5 years ago
1
How to share weights during searching for training to save training time?

#78 stanveer opened 5 years ago
0
after 15minutes training stopped by RuntimeError: CUDNN_STATUS_NOT_INITIALIZED

#77 maoqingsunny opened 5 years ago
1
Problem with channels when using MNIST dataset.

#76 JoseLlorensRipolles closed 4 years ago
1

Previous Next