drop_path_prob=0 or 0.3 in training ImageNet

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

Open miaozhang0525 opened 4 years ago

miaozhang0525 commented 4 years ago

the drop_path_prob is set as 0, but the paper said it follows "Regularized evolution for image classifier architecture search", which set 0.3.

Does it matter??

pingguokiller commented 4 years ago

Have you tested the right hyperparameter "drop_path_prob"? I reproduce with drop_path_prob set as 0 and get a test accuray 70.8%.