Tutorials

There are tutorials for Python users and for BrainScript users.

Python Jupyter Notebook Tutorials

The Python Jupyter notebooks in the Tutorials folder cover a range of different application including image classification, language understanding, reinforcement learning and others. Additionally, the folder Tutorials/NumpyInterop contains a simple example of how to use numpy arrays as input for CNTK training and evaluation.

You can also try out the tutorials live with Azure Notebooks.

Getting Started (BrainScript)

• Tutorial: Simple Logistic Regression and Multi-class Classification
• Tutorial II: Convolutional Networks with Batch Normalization for MNIST

Hands-On Labs (BrainScript)

• Image Recognition on CIFAR-10: Image recognition on CIFAR-10 with convolutional and Residual Nets
• Language Understanding with ATIS: Slot tagging and intent classification with recurrent networks
• Object Detection using Fast R-CNN: Object detection in images using the Fast R-CNN algorithm