This repository contains tutorials and examples to help you learn how to use Graphcore IPUs on the Spell Platform. This feature is currently in limited preview.
If you are interested in finding out more about Graphcore, including getting preview access to IPUs to run these examples, please register your interest here.
If you are actively using this repository and want to report any issues, please reach out to Spell support at [email protected].
The latest version of the documentation for the Poplar software stack, and other developer resources, is available at https://www.graphcore.ai/developer.
The code presented here uses Poplar SDK 2.4 in the default IPU-compatible container image during
spell run. The Spell team will continue to update and support newer SDKs as they are released.
The tutorials/ folder contains simple tutorials to help you get started using the Poplar SDK and Graphcore tools in Spell.
- tutorials/pytorch - A set of tutorials to introduce the PyTorch framework support for the IPU.
- tutorials/tensorflow2 - A set of tutorials to introduce the Tensorflow framework support for the IPU.
- tutorials/poplar - A set of tutorials to introduce the low level C++ Poplar library for programming on the IPU.
For general IPU tutorials, please refer to the Graphcore Github Tutorials repository.
Graphcore maintains sample applications and code examples for the IPU. This limited preview of Graphcore IPUs on Spell supports a smaller set of applications from the examples repository. Go to the Applications README to know more about how you can run IPU application examples in spell and for the list of supported models in this preview.
- For the rest of the IPU applications, please refer to the Graphcore Examples repository.