Latest version of tensorflow only be used VisualStudio2017 ？ about hello_tf_c_api

[Neargye]

Build on appveyor vs2015 pass

Are you compiling x64? Could you send a error message?

[taotaolin]

Now compiled, how to choose to use CPU or GPU？

[Neargye]

Different build are required with the support of CPU and GPU. My library is build with the support of the Cpu.

[taotaolin]

I can't successfully build tensorflow for GPU..View official README，Does the DEBUG version require python35d.lib ?

[taotaolin]

Different build are required with the support of CPU and GPU. My library is build with the support of the Cpu.

Can you try to build a GPU version tensorflow？ i can't compelete it. thx

[taotaolin]

@Neargye Can you provide Python generated code for PB files, my PB file does not work properly.

[Neargye]

Can you try to build a GPU version tensorflow？ i can't compelete it. thx

@taotaolin i add GPU build https://github.com/Neargye/tensorflow/releases/tag/v1.11.0

[Neargye]

@Neargye Can you provide Python generated code for PB files, my PB file does not work properly.

Could you please clarify which code interests you?

[taotaolin]

@Neargye Can you provide Python generated code for PB files, my PB file does not work properly.

Could you please clarify which code interests you?

I want to load MNIST graph with the C++ API, but follow the link http://jackytung8085.blogspot.com/2016/06/loading-tensorflow-graph-with-c-api-by.html method, it was found that the compiler could not be successfully compiled.
I can't understand how to load my MNIST model using methods like hello_tf_c_api/src/session_run.cpp.

[Neargye]

I use https://blog.metaflow.fr/tensorflow-how-to-freeze-a-model-and-serve-it-with-a-python-api-d4f3596b3adc