BatchNorm after pooling layer?

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

BatchNorm after pooling layer? #41

Closed D-X-Y closed 6 years ago

D-X-Y commented 6 years ago

Hi,

During training, you add one batchnorm layer after the pooling layer, as https://github.com/quark0/darts/blob/master/cnn/model_search.py#L18. However, when testing the searched architecture, there is no batchnorm after pooling layer. Would you mind to let me know the reason?

Best Regards,

quark0 commented 6 years ago

Please refer to #10. In the evaluation phase there're no batchnorms after pooling layers for fair comparison with existing models.