if i need train it first and than evalutate?

Question

if i need train it first and than evalutate?

johsnows opened this issue 5 years ago · comments

or evalutate code train_cifar include the search process?

Cytosine · Answer 1 · Tue Aug 06 2019 12:03:27 GMT+0800 (China Standard Time)

Sorry I can't catch you. To evaluate a searched model generated by PDARTS, you would always need to retrain it by provided codes such as train_cifar.py, or implement your own version of retrain code.

GL,

johsnows · Answer 2 · Tue Aug 06 2019 15:26:09 GMT+0800 (China Standard Time)

就是先用train_search.py 再train_cifar.py这样是嘛

chenxin061 · Answer 3 · Thu Aug 08 2019 15:41:07 GMT+0800 (China Standard Time)

Use train_search.py to do architecture search and train_cifar.py/train_imagenet.py to do architecture evaluation.

johsnows · Answer 4 · Sun Sep 15 2019 09:19:34 GMT+0800 (China Standard Time)

after search on the train_search.py, we get many genotype, no restrict skipconnetct genotype and genotypes with number of skip-connect:7,6,5,4,3,2,1,0, which one is best? which one is recommended to evaluate?