search

titu1994 / Keras-NASNet

"NASNet" models in Keras 2.0+ with weights
MIT License
200 stars 64 forks source link

Pooling kernel size #6

Closed uhfband closed 6 years ago

uhfband commented 6 years ago

Сould you please clarify, It's ok to have pool size less than stride size? eg nasnet.py line 620 p1 = AveragePooling2D((1, 1), strides=(2, 2), padding='valid', name='adjust_avg_pool1%s' % id)(p)

titu1994 commented 6 years ago

Yes. It's a point wise averaging which simply reduces the spacial dimension by half.

Since the weights load correctly, and predictions are correct, such an operation is correct.