# Poplar Basic Model Demo for the MNIST Dataset This demo shows how to use one IPU for a simple Poplar model with MNIST dataset. It contains a complete training program that performs a logistic regression on the MNIST data set, using gradient descent. ## File structure * `regression_demo.cpp` The main script containing the model. * `mnist.h` and `mnist.cpp` Helper files to iterate over the MNIST dataset. * `get_data.sh` A script to download the MNIST dataset. * `tests/` A folder containing tests for this model. * `README.md` This file. ## How to use this application 1) Prepare the Poplar environment. Install the Poplar SDK following the Getting Started guide for your IPU system. Make sure to run the `enable.sh` script for Poplar. 2) Run the ``get_data.sh`` script to download the MNIST data. ./get_data.sh 3) Build the code with the Makefile provided: make 4) Train and test the model: ./regression-demo [-IPU] [number of epochs] [proportion of images to use] Copyright (c) 2018 Graphcore Ltd. All rights reserved.