f1fef359f319377ee0518a13ede38e4a654622ce,models_vggtiny.py,,model,#Any#Any#Any#Any#Any#Any#Any#,92
Before Change
if data_format != "channels_last":
// TODO: support NCHW
print("data_format=%s is ignored" % data_format)
with tf.variable_scope("model", reuse):
//// Feature extraction part
cnn = vgg_network(x)
After Change
//// stage 2 ~ 6
// for i in range(2, 7):
with tf.variable_scope("stage1/branch2"): // TODO: fix indent here and the names in npz
for i in [5, 6]: // only 3 stage in total
b1, b2 = stage2(cnn, b1, b2, n_pos, mask_miss1, mask_miss2, is_train, scope_name="stage%d" % i)
b1_list.append(b1)
b2_list.append(b2)
net = tl.layers.merge_networks([b1, b2])
return cnn, b1_list, b2_list, net
In pattern: SUPERPATTERN
Frequency: 3
Non-data size: 4
Instances
Project Name: tensorlayer/openpose-plus
Commit Name: f1fef359f319377ee0518a13ede38e4a654622ce
Time: 2018-09-20
Author: lgarithm@users.noreply.github.com
File Name: models_vggtiny.py
Class Name:
Method Name: model
Project Name: NVIDIA/OpenSeq2Seq
Commit Name: f791b89909d8ac69cac38610a4b1ed31d219920d
Time: 2018-08-01
Author: chipn@nvidia.com
File Name: test.py
Class Name:
Method Name:
Project Name: NVIDIA/OpenSeq2Seq
Commit Name: f791b89909d8ac69cac38610a4b1ed31d219920d
Time: 2018-08-01
Author: chipn@nvidia.com
File Name: open_seq2seq/encoders/lm_encoders.py
Class Name: AWDLSTMEncoder
Method Name: _encode