MNIST using TensorFlow

By for September 13, 2017

Report Abuse
Introduce situations where Azure Machine Learning (ML) R notebooks can be used. Fit a model using R, then publish the model as a web service.
# Classify MNIST dataset using TensorFlow Run in local conda environment. ``` $ pip install tensorflow $ az ml experiment submit -c local ``` Run in a local Docker container. ``` $ az ml experiment submit -c docker ``` Run in a Docker container in a remote machine. Note you need to create/configure myvm.compute. ``` $ az ml experiment submit -c myvm ``` Run in a Docker container in a remote machine with GPU. - Create a new compute context, nam it _gpu_ (or any arbitary name) - Use _az ml computetarget attach_ to target the GPU equipped VM. - In **conda_dependencies.yml** file, use _tensorflow-gpu_ instead of _tensorflow_. - In **gpu.compute** file, use _microsoft/mmlspark:gpu_ as the base Docker image. - In **gpu.compute** file, add a line _nvidiaDocker: true_ - In **gpu.runconfig** file, set _Framework_ to _Python_ - Now run the script. ``` $ az ml experiment submit -c gpu ```