Program autotuning has been demonstrated in many domains to achieve better or more portable performance. However, autotuners themselves are often not very portable between projects because using a domain informed search space representation is critical to achieving good results and because no single search technique performs best for all problems.
OpenTuner is a new framework for building domain-specific multi-objective program autotuners. OpenTuner supports fully customizable configuration representations, an extensible technique representation to allow for domain-specific techniques, and an easy to use interface for communicating with the tuned program. A key capability inside OpenTuner is the use of ensembles of disparate search techniques simultaneously, techniques which perform well will receive larger testing budgets and techniques which perform poorly will be disabled.
A list of system dependencies can be found in debian-packages-deps which are primarily python 2.6+ (not 3.x) and sqlite3 (or your supported database backend of choice).
On Ubuntu/Debian there can be installed with:
sudo apt-get install `cat debian-packages-deps | tr '\n' ' '`
OpenTuner (and dependencies) can be installed with
sudo pip install opentuner
or
pip install --user opentuner
This will not install any of the example programs.
For development (running OpenTuner out of a git checkout), a list of python
dependencies can be found in requirements.txt these can be installed
system-wide with pip.
sudo apt-get install python-pip
sudo pip install -r requirements.txt
Or you can use virtual env to create a isolated python environment by running:
python ./venv-bootstrap.py
which will create a ./venv/bin/python (./venv/Scripts/python.exe on windows) with all the required packages installed.
Quickly checking that a successful installation has been made, may be performed by running an example program such as:
./examples/rosenbrock/rosenbrock.py
- OpenTuner: An Extensible Framework for Program Autotuning.
Jason Ansel, Shoaib Kamil, Kalyan Veeramachaneni, Jonathan Ragan-Kelley, Jeffrey Bosboom, Una-May O'Reilly, Saman Amarasinghe.
International Conference on Parallel Architectures and Compilation Techniques.
Edmonton, Canada. August, 2014. Slides. Bibtex.
The preferred way to contribute code to OpenTuner is to fork the project on github and submit a pull request.
OpenTuner is supported in part by the United States Department of Energy X-Stack program as part of D-TEC.