Deep Learning and AI frameworks

The Data Science Virtual Machine (DSVM) and the Deep Learning VM supports a number of deep learning frameworks to help build Artificial Intelligence (AI) applications with predictive analytics and cognitive capabilities like image and language understanding.

Here are the details on all the deep learning frameworks available on the DSVM.

Microsoft Cognitive Toolkit

What is it? Deep learning framework
Supported DSVM Editions Windows, Linux
How is it configured / installed on the DSVM? The Microsoft Cognitive Toolkit (CNTK) is installed in Python 2.7, in the root environment, as well as Python 3.5, in the py35 environment.
Links to Samples Sample Jupyter notebooks are included.
Related Tools on the DSVM Keras
How to use / run it? Open Jupyter, then look for the CNTK folder

TensorFlow

What is it? Deep learning framework
Supported DSVM Editions Windows, Linux
How is it configured / installed on the DSVM? On Linux, TensorFlow is installed in Python 2.7 (root), as well as Python 3.5 (py35) environment. On Windows, Tensorflow is installed in Python 3.5(py35) environment.
Links to Samples Sample Jupyter notebooks are included.
Related Tools on the DSVM Keras
How to use / run it? Open Jupyter, then look for the TensorFlow folder.

Keras

What is it? Deep learning framework
Supported DSVM Editions Windows, Linux
How is it configured / installed on the DSVM? Keras is installed in Python 2.7 (root), as well as Python 3.5 (py35) environment.
Links to Samples https://github.com/fchollet/keras/tree/master/examples
Related Tools on the DSVM Microsoft Cognitive Toolkit, TensorLlow, Theano
How to use / run it? Download the samples from the Github location, copy it to a directory under ~/notebooks and open it in Jupyter

Caffe

What is it? Deep learning framework
Supported DSVM Editions Linux
How is it configured / installed on the DSVM? Caffe is installed in /opt/caffe.
Links to Samples Samples are included in /opt/caffe/examples.
Related Tools on the DSVM Caffe2

How to use / run it?

Use X2Go to log in to your VM, then start a new terminal and enter

cd /opt/caffe/examples
jupyter notebook

A new browser window opens with sample notebooks.

Caffe2

What is it? Deep learning framework
Supported DSVM Editions Linux
How is it configured / installed on the DSVM? Caffe2 is installed in /opt/caffe2. It is also available for Python 2.7(root) conda environment.
Links to Samples Sample Jupyter notebooks are included
Related Tools on the DSVM Caffe
How to use / run it? Open Jupyter, then navigate to the Caffe2 directory to find sample notebooks. Some notebooks require the Caffe2 root to be set in the Python code; enter /opt/caffe2.

Chainer

What is it? Deep learning framework
Supported DSVM Editions Windows, Linux
How is it configured / installed on the DSVM? Chainer is installed in Python 2.7 (root), as well as Python 3.5 (py35) environment. ChainerRL and ChainerCV are also installed.
Links to Samples Sample Jupyter notebooks are included.
Related Tools on the DSVM Caffe

How to use / run it?

At a terminal, activate the Python version you want (root or py35), run python, then import Chainer. In Jupyter, select the Python 2.7 or 3.5 kernel, then import Chainer.

Deep Water

What is it? Deep learning framework for H2O
Supported DSVM Editions Linux
How is it configured / installed on the DSVM? Deep Water is installed in /dsvm/tools/deep_water.
Links to Samples Samples are available through the Deep Water server.
Related Tools on the DSVM H2o, Sparkling Water

How to use / run it?

Connect to the VM using X2Go. At a terminal, start the Deep Water server:

java -jar /dsvm/tools/deep_water/h2o.jar

Then open a browser and connect to http://localhost:54321.

MXNet

What is it? Deep learning framework
Supported DSVM Editions Windows, Linux
How is it configured / installed on the DSVM? MXNet is installed in C:\dsvm\tools\mxnet on Windows and /dsvm/tools/mxnet on Linux. Python bindings are installed in Python 2.7 (root), as well as Python 3.5 (py35) environment. R bindings are also installed.
Links to Samples Sample Jupyter notebooks are included.
Related Tools on the DSVM Keras
How to use / run it? Open Jupyter, then look for the mxnet folder

NVIDIA DIGITS

What is it? Deep learning system from NVIDIA for rapidly training deep learning models
Supported DSVM Editions Linux
How is it configured / installed on the DSVM? DIGITS is installed in /dsvm/tools/DIGITS and is available a service called digits.

How to use / run it?

Log in to the VM with X2Go. At a terminal, start the service:

sudo systemctl start digits

The service takes about one minute to start. Start a web browser and navigate to http://localhost:5000.

nvdia-smi

What is it? NVIDIA tool for querying GPU activity
Supported DSVM Editions Windows, Linux
How is it configured / installed on the DSVM? nvidia-smi is available on the system path.
How to use / run it? Start a command prompt (on Windows) or a terminal (on Linux), then run nvidia-smi.

Theano

What is it? Deep learning framework
Supported DSVM Editions Linux
How is it configured / installed on the DSVM? Theano is installed in Python 2.7 (root), as well as Python 3.5 (py35) environment.
Related Tools on the DSVM Keras
How to use / run it? At a terminal, activate the Python version you want (root or py35), run python, then import theano. In Jupyter, select the Python 2.7 or 3.5 kernel, then import theano.

Torch

What is it? Deep learning framework
Supported DSVM Editions Linux
How is it configured / installed on the DSVM? Torch is installed in /dsvm/tools/torch. PyTorch is installed in Python 2.7 (root), as well as Python 3.5 (py35) environment.
Links to Samples Torch samples are located at /dsvm/samples/torch. PyTorch samples are located at /dsvm/samples/pytorch.