GitHub - aws/sagemaker-tensorflow-training-toolkit: Toolkit for running TensorFlow training scripts on SageMaker. Dockerfiles used for building SageMaker TensorFlow Containers are at https://github.com/aws/deep-learning-containers. (original) (raw)

SageMaker TensorFlow Training Toolkit

The SageMaker TensorFlow Training Toolkit is an open source library for making the TensorFlow framework run on Amazon SageMaker.

This repository also contains Dockerfiles which install this library, TensorFlow, and dependencies for building SageMaker TensorFlow images.

For information on running TensorFlow jobs on SageMaker:

• SageMaker Python SDK documentation
• SageMaker Notebook Examples

Table of Contents

1. Getting Started
2. Building your Image
3. Running the tests

Getting Started

Prerequisites

Make sure you have installed all of the following prerequisites on your development machine:

• Docker

For Testing on GPU

• Nvidia-Docker

Recommended

• A Python environment management tool. (e.g.PyEnv,VirtualEnv)

Building your Image

Amazon SageMakerutilizes Docker containers to run all training jobs & inference endpoints.

The Docker images are built from the Dockerfiles specified indocker/.

The Dockerfiles are grouped based on TensorFlow version and separated based on Python version and processor type.

The Dockerfiles for TensorFlow 2.0+ are available in thetf-2 branch.

To build the images, first copy the files underdocker/build_artifacts/to the folder container the Dockerfile you wish to build.

Example for building a TF 2.1 image with Python 3

cp docker/build_artifacts/* docker/2.1.0/py3/.

After that, go to the directory containing the Dockerfile you wish to build, and run docker build to build the image.

Example for building a TF 2.1 image for CPU with Python 3

cd docker/2.1.0/py3 docker build -t tensorflow-training:2.1.0-cpu-py3 -f Dockerfile.cpu .

Don't forget the period at the end of the docker build command!

Running the tests

Running the tests requires installation of the SageMaker TensorFlow Training Toolkit code and its test dependencies.

git clone https://github.com/aws/sagemaker-tensorflow-container.git cd sagemaker-tensorflow-container pip install -e .[test]

Tests are defined intest/and include unit, integration and functional tests.

Unit Tests

If you want to run unit tests, then use:

All test instructions should be run from the top level directory

pytest test/unit

Integration Tests

Running integration tests require Docker and AWS credentials, as the integration tests make calls to a couple AWS services. The integration and functional tests require configurations specified within their respectiveconftest.py.Make sure to update the account-id and region at a minimum.

Integration tests on GPU require Nvidia-Docker.

Before running integration tests:

1. Build your Docker image.
2. Pass in the correct pytest arguments to run tests against your Docker image.

If you want to run local integration tests, then use:

Required arguments for integration tests are found in test/integ/conftest.py

pytest test/integration --docker-base-name
--tag
--framework-version
--processor

Example

pytest test/integration --docker-base-name preprod-tensorflow
--tag 1.0
--framework-version 1.4.1
--processor cpu

Functional Tests

Functional tests are removed from the current branch, please see them in older branch r1.0.

Contributing

Please readCONTRIBUTING.mdfor details on our code of conduct, and the process for submitting pull requests to us.

License

SageMaker TensorFlow Containers is licensed under the Apache 2.0 License. It is copyright 2018 Amazon.com, Inc. or its affiliates. All Rights Reserved. The license is available at:http://aws.amazon.com/apache2.0/