MikoyChinese

Here is my building step. First, the problem : ```Closure Rules requires Bazel >=0.4.5 but was 0.11.1```. If you met this error, just change like this url [ https://github.com/bazelbuild/bazel/issues/4834 ],...

Here is my document to build Tensorflow1.5.0. http://note.youdao.com/noteshare?id=3779eaea012bfec0fd1dc3e6b27a2b62 If you can't visit it and try to visit another web: https://www.jianshu.com/p/d53d231e8cf0

Or if I have mulit ports to listen, such as 20001, 20003, 13889, how can I do?

There is an upsample operation after the feature maps of size (N, C, H/4, W/4).

Hi, @MaybeShewill-CV Can you show the code? I can't reduce my loss on my code which is followed the author thinking. And I don't know what's wrong with my code....

Thanks for your great work.

+1

@Jyurineko OpenGL pipeline need a monitor to run, or you can build a headless opengl library. I guessed that you run this on ssh tunnel without X forward, so you...

@Jyurineko :), Hah, If you want to use opengl pipeline you need a monitor or headless opengl library. But I just think that you can just use CPU pipeline or...

Open the kinect v2 device such like this: [demo](https://github.com/r9y9/pylibfreenect2/blob/97786f5dcf94c994af648d8ec2195cfabab8a792/examples/selective_streams.py).