Question: About Renet-v1 stride value(looks different with original paper)

System information
Top level dir : tensorflow/models/research/slim/nets/
Have I written custom code (as opposed to using a stock example script provided in TensorFlow):No
OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 16.04
TensorFlow installed from (source or binary): pip3 library install
TensorFlow version : 1.13
Bazel version (if compiling from source): N/A
CUDA/cuDNN version: 10.0
GPU model and memory: N/A
Exact command to reproduce: N/A

problem

It's not an error and just curious about stride value of resnet.
As I know, original paper(Deep Residual Learning for Image Recognition, 2015) use stride "2" on block 2,3,4. but code use stride "2" on block 1,2,3.

• origianal paper's description is on Table 1.
• code is on /models/research/slim/nets/resnet_v1.py(same on "resnet_v2.py "), def resnet_v1_50(...): ...

Is there any specific reason for using stride "2" on block 1,2,3 instead block 2,3,4??

Thank you.