Running TensorFlow 2.2 in Azure Machine Learning Studio
TensorFlow 2.2.0 was released last month (May 2020) with major improvements and features for the Deep Learning Framework (https://github.com/tensorflow/tensorflow/releases/tag/v2.2.0). One challenge with keeping up with bleeding edge AI and ML frameworks is that the changes can easily outpace APIs and SDKs build on top of these tools. Fortunately, since these tools are largely open-source and made available through industry common package managers, these updates can be used before the frameworks support them, albeit with some workarounds.
Azure Machine Learning Studio (AzureML) is a platform service that provides full end to end management of Machine Learning and Data Science workloads. The tool provides experiment tracking, data set management, model repository, and deployment services to enable data scientists to train/validate/deploy their models in a cloud scale environment. (https://azure.microsoft.com/en-us/services/machine-learning/)
To run TensorFlow 2.2 in AzureML is really pretty trivial, given the existing support for the TensorFlow framework. AzureML provides native support for framework version 2.1, but to get 2.2 to run you simply need to add the requirement for version "==2.2.0". Here's an example of a training experiment:
Note: If you are using the "use_gpu = True" argument, you will need to make sure that the "tensorflow-gpu==2.2.0" is used. If you are using TensorFlow without gpu support, you can use the package "tensorflow==2.2.0". Lastly, just ignore the warning message.
After the training is deployed, you will see the correct version of TensorFlow installed in your training environment.
And lastly, the output from your python file will show the correct version has been imported:
print("Tensorflow version:", tf.__version__)
You can run this example yourself in my fork of the AzureMachineLearning notebooks.