search

tobegit3hub / simple_tensorflow_serving

Generic and easy-to-use serving service for machine learning models
https://stfs.readthedocs.io
Apache License 2.0
757 stars 195 forks source link

If I want to deploy module on cross-platform #44

Closed yingwei13mei closed 5 years ago

yingwei13mei commented 5 years ago

If I want to deploy module on different-platform. For example I trained the module in python -Tensorflow enviroment. But I expect to use this module file in C++ windows env. I reference some views on TF Serving saved_model_builder() function, it seems like a workable solution. And how could it easily transfer module in different platform within current solution?

Just some thinking about current project.

tobegit3hub commented 5 years ago

I'm not sure if I completely understand what you mean. No mather what languages(Python or C++) you use for TensorFlow, the model file is an protobuf file and we need to use saved_model_builder() to set the model signature. Then you can use this model file in Python, C++ or any other languages which can read TensorFlow SavedModel file.