chenxin061 / pdarts

Codes for our paper "Progressive Differentiable Architecture Search:Bridging the Depth Gap between Search and Evaluation"

chenxin061/pdarts Issues

How to caculate GPU-days(cost)?
Updated 3 years ago
The number of filters in each cell
Closed 3 years ago
Reproducing the Results
Updated 3 years ago
Reproducing the Results
Closed 3 years ago
CIFAR-100 Performance and Hyperparameters
Updated 4 years ago
What does 'operation-level Dropout' mean?
Updated 4 years ago
Question regarding the paper
Updated 4 years ago1
parameter stem_multiplier and module stem
Closed 4 years ago6
why the learning rate and batchsize in "train_imagenet.py" are different from what said in the paper? Which one can reproduce the result？And is the learning rate related to batchsize ?
Closed 4 years ago1
How to use this algorithm to my datasets..
Closed 4 years ago1
How many hours on ImageNet of training?
Updated 4 years ago11
PDARTS on medical images classification
Closed 4 years ago1
How many layers we need in the evaluation stage?
Closed 4 years ago11
lr being passed to train(), but not used
Closed 4 years ago1
semantic segmentation
Closed 4 years ago1
Add more code comment？
Updated 4 years ago3
Is this a normal case?
Closed 4 years ago
Question about Dropout Rate in Paper
Closed 4 years ago4
confusion about depths of network in different stages
Closed 4 years ago1
correct RUN command
Closed 4 years ago1
which genotype is best？
Updated 4 years ago5
关于一阶优化
Closed 4 years ago1
In the original implementation of DARTS, it is input_search, target_search = next(iter(valid_queue), which slows down the training when using PyTorch 0.4 and above.
Closed 4 years ago1
For other datasets, is it better to use pre-searched network from cifar, or search from itself?
Updated 4 years ago13
Is your method only work on CUDA 10?
Updated 5 years ago2
Multi-gpu support
Updated 5 years ago10
use train_search , got genotype like DARTS
Closed 5 years ago2
if i need train it first and than evalutate?
Closed 5 years ago4
Reproducing the Results, and Questions
Updated 5 years ago22
In arch search process , why use SGD for operation weight and Adam for arch_params ?
Closed 5 years ago1
How can resume the train_search.py using weights in log_dir if the code stops due to accidents like power cut ... ?
Closed 5 years ago1
How can i import the found network architecture ?
Updated 5 years ago3
RuntimeError: cuda runtime error (2)
Closed 5 years ago
How to test?
Closed 5 years ago4
Problem while running test.py
Closed 5 years ago4
[BUG] test.py
Closed 5 years ago1
[Bug]CUDA error: out of memory
Closed 5 years ago1
Can i control the flops of the models searched?
Updated 5 years ago3
Missing References
Closed 5 years ago2
如果只有ＧＴＸ１０８０ｔｉ（１１Ｇ内存），如何调整参数，让pdarts也可以跑起来训练？
Closed 5 years ago1