WangYueFt/dgcnn

Undefined "weight_decay" in tensorflow/sem_seg/model.py

MrJimm opened this issue · 3 comments

MrJimm commented

Hi!
In tensorflow/sem_seg/model.py there seems to be undefined "weight_decay" variable used in tf_util.conv2d(...) layers while building model.

For example here:
out1 = tf_util.conv2d(edge_feature, 64, [1,1], padding='VALID', stride=[1,1], bn=True, is_training=is_training, weight_decay=weight_decay, scope='adj_conv1', bn_decay=bn_decay, is_dist=True)

niallomahony93 commented

I came across the same issue, what value should it be?

WangYueFt commented

Hi!
In tensorflow/sem_seg/model.py there seems to be undefined "weight_decay" variable used in tf_util.conv2d(...) layers while building model.

For example here:
out1 = tf_util.conv2d(edge_feature, 64, [1,1], padding='VALID', stride=[1,1], bn=True, is_training=is_training, weight_decay=weight_decay, scope='adj_conv1', bn_decay=bn_decay, is_dist=True)

That could be caused by the tensor flow update. Can you use the old version of tensor flow?

Dinghow commented

You can just use the default value weight_decay=0.0, I don't think this issue is attributed to the version of tensorflow.