search

aws / sagemaker-tensorflow-serving-container

A TensorFlow Serving solution for use in SageMaker. This repo is now deprecated.
Apache License 2.0
172 stars 101 forks source link

Does the underlying code of the container is wrote by C++ #143

Closed Ageneinair closed 4 years ago

Ageneinair commented 4 years ago

Is this project for realizing the same function as the TensorFlow Serving, but use a different way? Or just bring the TensorFlow Serving to AWS?

I have this question because I saw some underlying core code is written in python like this one: https://raw.githubusercontent.com/aws/aws-deep-learning-containers-utils/master/deep_learning_container.py, but as I know the whole structure of TensorFlow Serving is written in C++

laurenyu commented 4 years ago

This repository is for bringing TFS support to Amazon SageMaker.

For what versions of TFS are supported, see https://aws.amazon.com/releasenotes/available-deep-learning-containers-images.

The code is used to build Docker images, whose Dockerfiles can be found at https://github.com/aws/deep-learning-containers/tree/master/tensorflow/inference/docker.

Ageneinair commented 4 years ago

I saw all the code based on

TFS_URL=https://tensorflow-aws.s3-us-west-2.amazonaws.com/2.1/Serving/CPU-WITH-MKL/tensorflow_model_server

But I can't get access to this one, is this one written by C++?

laurenyu commented 4 years ago

Yes, it's basically the same. Since it's installed by the Dockerfiles now stored in https://github.com/aws/deep-learning-containers, I would suggest opening an issue in that repo for further questions.