Skip to content

Fix model layer for resnet v1. #4230

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 10, 2018
Merged

Fix model layer for resnet v1. #4230

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 6 additions & 2 deletions official/resnet/resnet_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -424,6 +424,7 @@ def __init__(self, resnet_size, bottleneck, num_classes, num_filters,
self.block_strides = block_strides
self.final_size = final_size
self.dtype = dtype
self.pre_activation = resnet_version == 2

def _custom_dtype_getter(self, getter, name, shape=None, dtype=DEFAULT_DTYPE,
*args, **kwargs):
Expand Down Expand Up @@ -518,8 +519,11 @@ def __call__(self, inputs, training):
strides=self.block_strides[i], training=training,
name='block_layer{}'.format(i + 1), data_format=self.data_format)

inputs = batch_norm(inputs, training, self.data_format)
inputs = tf.nn.relu(inputs)
# Only apply the BN and ReLU for model that does pre_activation in each
# building/bottleneck block, eg resnet V2.
if self.pre_activation:
inputs = batch_norm(inputs, training, self.data_format)
inputs = tf.nn.relu(inputs)

# The current top layer has shape
# `batch_size x pool_size x pool_size x final_size`.
Expand Down