quark0/darts issues and pull requests

#175 - cifar10_model.pt file not found

Issue - State: open - Opened by cloud-zhoubingye about 2 months ago - 2 comments

#174 - cifar10_model.pt file not found

Issue - State: open - Opened by cloud-zhoubingye about 2 months ago

#173 - Loading the model after training

Issue - State: open - Opened by Tahmid1406 2 months ago

#172 - confuse about the train_search/train/test

Issue - State: open - Opened by whubaichuan about 1 year ago

#171 - Please upadate codes for latest version of PyTorch

Issue - State: open - Opened by LT1st over 1 year ago

#170 - Can only achieve 93% accuracy on CIFAR10

Issue - State: closed - Opened by tianyic over 1 year ago

#169 - Implementation compatible with the recent PyTorch and CUDA version

Issue - State: open - Opened by qianjiangcn almost 2 years ago

#168 - Not much comments in code.

Issue - State: open - Opened by markyen001 almost 2 years ago - 2 comments

#167 - try to visulize the genotype but got error

Issue - State: open - Opened by LinzheCAI about 2 years ago - 1 comment

#166 - Broken pipe issue

Issue - State: closed - Opened by moon4sake over 2 years ago - 1 comment

#165 - About utils.accuracy

Issue - State: closed - Opened by AustinzzZ-1997 about 3 years ago

#164 - RuntimeError: cannot pin 'torch.cuda.DoubleTensor' on GPU on version 0.10.0

Issue - State: open - Opened by novinsh about 3 years ago

#163 - About alpha

Issue - State: open - Opened by andyzgj over 3 years ago - 1 comment

#162 - Training time on colab

Issue - State: open - Opened by pizhn over 3 years ago - 2 comments

#161 - When running train,py, why can't I get the same high accuracy as in train_search.py

Issue - State: open - Opened by liqier over 3 years ago

#160 - Can you tell me how to calculate Search Cost?

Issue - State: open - Opened by liqier over 3 years ago - 1 comment

#159 - Questions about DARTS

Issue - State: open - Opened by buttercutter over 3 years ago

#158 - Why are stem.0.weight different in pretrained and searched CIFAR-10 architectures?

Issue - State: closed - Opened by chengrunyang over 3 years ago

#157 - Indentation bugs

Issue - State: open - Opened by qgh1223 over 3 years ago - 1 comment

#156 - The discrepancy between DARTS and ENAS for RNN cell searching

Issue - State: open - Opened by cocovoc over 3 years ago

#155 - Can not reproduce the search result on PTB dataset (Fig3)?

Issue - State: open - Opened by fortunechen over 3 years ago - 5 comments

#154 - Imagenet mobile setting

Issue - State: open - Opened by johnjaejunlee almost 4 years ago

#153 - Approximate Architecture Gradient

Issue - State: open - Opened by buttercutter almost 4 years ago

#152 - Modify

Pull Request - State: closed - Opened by zhangsiyu1103 almost 4 years ago - 1 comment

#151 - own data set

Issue - State: open - Opened by bianbianbian1202 almost 4 years ago

#150 - Doubt in the paper

Issue - State: closed - Opened by meghbhalerao about 4 years ago - 2 comments

#149 - Sample random genotypes

Issue - State: open - Opened by bknyaz over 4 years ago

#148 - why the learning rate in rnn is always 20?

Issue - State: open - Opened by hjq133 over 4 years ago

#147 - How to convert to ONNX.

Issue - State: closed - Opened by crook52 over 4 years ago - 2 comments

#146 - torch version

Issue - State: open - Opened by amirreza-m95 over 4 years ago - 4 comments

#145 - question about train_data and validation_data

Issue - State: open - Opened by LiuTingWed over 4 years ago

#144 - Getting Index Error

Issue - State: open - Opened by shashi19081994 over 4 years ago - 1 comment

#143 - Shouldn't we drop on the second dimension in the drop_path function?

Issue - State: open - Opened by woodywff over 4 years ago - 1 comment

#142 - Code details

Issue - State: open - Opened by sjhan91 over 4 years ago - 2 comments

#141 - The reproducibility of the code

Issue - State: open - Opened by PipiZong over 4 years ago

#140 - How to control the model parameters when jointly optimize 'alpha' and 'W' over the union of the training and validation sets using coordinate descent.?

Issue - State: open - Opened by larenzhang over 4 years ago

#139 - In your paper, why is the inner optimization of the loss function expensive?

Issue - State: open - Opened by Jugaris89 over 4 years ago - 1 comment

#138 - Framework question: how the alpha be updated? Some different between the code and paper.

Issue - State: open - Opened by awecefil over 4 years ago

#137 - Generate release?

Issue - State: open - Opened by oblute over 4 years ago

#136 - How is this formula derived?

Issue - State: open - Opened by NoOneUST over 4 years ago - 1 comment

#135 - ImageNet training time

Issue - State: open - Opened by thinkInJava33 over 4 years ago - 1 comment

#134 - parameter "stem-multiplier", "multiplier", "steps"

Issue - State: open - Opened by xjhcassy over 4 years ago - 5 comments

#133 - Is this a normal case?

Issue - State: closed - Opened by mrk1992 almost 5 years ago - 2 comments

#132 - drop_path_prob=0 or 0.3 in training ImageNet

Issue - State: open - Opened by miaozhang0525 almost 5 years ago - 1 comment

#131 - How to set the number of kernels in conv?

Issue - State: open - Opened by MrDavidG almost 5 years ago - 1 comment

#130 - Concatenation of nodes in cell

Issue - State: open - Opened by tkanvar almost 5 years ago

#129 - Nan appears in the time series

Issue - State: open - Opened by zzhangruifeng almost 5 years ago

#128 - a little doubt on train_search.py

Issue - State: closed - Opened by YuechengLiu almost 5 years ago - 1 comment

#127 - Xray

Pull Request - State: closed - Opened by dylanrandle almost 5 years ago

#126 - How to extract the result cell structure of train_search.py?

Issue - State: closed - Opened by kandysoso almost 5 years ago - 2 comments

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

Issue - State: open - Opened by xingyueye almost 5 years ago - 1 comment

#124 - The alpha variable's value of "avg_pool" and "skip_connect" is the mostly biggest over the others

Issue - State: open - Opened by shuangjian24 about 5 years ago - 2 comments

#123 - Not able to reproduce the cells in the paper

Issue - State: open - Opened by doitslow about 5 years ago - 1 comment

#122 - architecture parameters initialization

Issue - State: open - Opened by JingweiZhang12 about 5 years ago

#121 - Two questions about code implementation

Issue - State: closed - Opened by TianQi-777 about 5 years ago - 1 comment

#120 - What is DART

Issue - State: open - Opened by epriwahyu about 5 years ago

#119 - Making the code dataset agnostic

Pull Request - State: closed - Opened by dylanrandle about 5 years ago - 1 comment

#118 - param_size of searched model is too heavy

Issue - State: open - Opened by xingyueye about 5 years ago - 1 comment

#117 - How to implement embedding + mask in torch 1.12?

Issue - State: open - Opened by pingguokiller about 5 years ago - 2 comments

#116 - A issue about Figure 2 in paper

Issue - State: open - Opened by jjjjjie about 5 years ago - 3 comments

#115 - How to generate a complete cifar network?

Issue - State: open - Opened by zhulishun about 5 years ago

#114 - How to stack cells into a complete architect?

Issue - State: open - Opened by zhulishun about 5 years ago

#113 - Implementation of approximate architecture gradient is different with paper.

Issue - State: closed - Opened by iuhiyuh about 5 years ago - 1 comment

#112 - how to get ppl by using customed configrues'ENAS'

Issue - State: open - Opened by hukaidididi about 5 years ago

#111 - I can‘t run the train.py directly

Issue - State: open - Opened by shuotian17 about 5 years ago

#110 - I can't reproduce the architecture search process following the configuration in paper.

Issue - State: open - Opened by ghost about 5 years ago - 13 comments

#109 - one of the variables needed for gradient computation has been modified by an inplace operation

Issue - State: open - Opened by pawopawo over 5 years ago - 2 comments

#108 - what's the meaning of function mask2d in rnn.utils?

Issue - State: closed - Opened by pauking over 5 years ago - 1 comment

#107 - why the hyperparameters, for both the training pipeline (train_search.py )and the final evaluation pipeline (train.py) differ a lot!?

Issue - State: open - Opened by NdaAzr over 5 years ago - 3 comments

#106 - "TypeError: Iteration over 0-d tensor"

Issue - State: open - Opened by ilia10000 over 5 years ago - 4 comments

#105 - running time when it compares with Resnet from other code

Issue - State: open - Opened by Margrate over 5 years ago - 1 comment

#104 - How to get the hidden cell, i.e. the features, in RNN architecture?

Issue - State: open - Opened by QingLi-Chloe over 5 years ago

#103 - much information not mentioned in papers

Issue - State: open - Opened by deeptoby over 5 years ago

#102 - How to use the RNN cell search on images

Issue - State: open - Opened by QingLi-Chloe over 5 years ago

#101 - Doubt about the effectiveness of the method

Issue - State: open - Opened by twangnh over 5 years ago - 28 comments

#100 - Why model storage memory takes a huge amount of space?

Issue - State: open - Opened by haithanhp over 5 years ago - 2 comments

#99 - first order performs better than second order and much faster

Issue - State: closed - Opened by twangnh over 5 years ago

#98 - out of memory for different architecture(pytorch0.3)

Issue - State: open - Opened by skx6 over 5 years ago - 7 comments

#97 - pytorch0.3 out of memery fordifferent architecture

Issue - State: closed - Opened by skx6 over 5 years ago - 1 comment

#96 - support latest pytorch

Issue - State: open - Opened by MandyMo over 5 years ago - 5 comments

#95 - Bug!DataLoaderIter ConnectionResetError: [Errno 104] Connection reset by peer

Issue - State: open - Opened by zihaozhang9 over 5 years ago - 2 comments

#94 - train_search.py ConnectionResetError

Issue - State: open - Opened by zihaozhang9 over 5 years ago - 1 comment

#93 - ScheduledDropPath

Issue - State: closed - Opened by skmhrk1209 over 5 years ago

#92 - ImageNet reproduce results in paper

Issue - State: open - Opened by haithanhp over 5 years ago - 14 comments

#91 - What does the "_bns" in model_train.py L40 do?

Issue - State: open - Opened by Catosine over 5 years ago

#90 - What changes need to be made to be compatible with pytorch 0.4.0

Issue - State: open - Opened by ghost over 5 years ago - 2 comments

#89 - AttributeError: module 'torch' has no attribute 'float32'

Issue - State: open - Opened by FrankYoungchen over 5 years ago

#88 - pytorch 1.0+ compatibility

Issue - State: open - Opened by nk0307 over 5 years ago - 1 comment

#87 - Train arch_parameters with validation data?

Issue - State: closed - Opened by Harry-Up over 5 years ago - 4 comments

#86 - Darts for nni

Pull Request - State: open - Opened by WdBlink over 5 years ago

#85 - It seems the global moving average is used as opposed to what the original paper instructed

Issue - State: open - Opened by bombs-kim over 5 years ago

#84 - A bug on FactorizedReduce in operations.py?

Issue - State: open - Opened by Dav-Jay over 5 years ago - 3 comments

#83 - Reimplementation for rnn using fastai

Issue - State: open - Opened by tinhb92 over 5 years ago - 2 comments

#82 - Architecture parameter initialization

Issue - State: open - Opened by JaminFong over 5 years ago - 1 comment

#81 - Is it possible to search with larger image size like 224*224?

Issue - State: open - Opened by R00Kie-Liu over 5 years ago - 1 comment

#80 - Searched architecture training time.

Issue - State: open - Opened by zhaohui-yang over 5 years ago - 9 comments

#79 - How do you learn reduction cell?

Issue - State: open - Opened by imirzadeh over 5 years ago - 1 comment

#78 - How to share weights during searching for training to save training time?

Issue - State: open - Opened by stanveer over 5 years ago

#77 - after 15minutes training stopped by RuntimeError: CUDNN_STATUS_NOT_INITIALIZED

Issue - State: open - Opened by maoqingsunny over 5 years ago - 1 comment

#76 - Problem with channels when using MNIST dataset.

Issue - State: closed - Opened by JoseLlorensRipolles over 5 years ago - 1 comment

GitHub / quark0/darts issues and pull requests