llnl/apollo

Apollo: Online Machine Learning for Performance Portability

Apollo

Apollo is a distributed tuning framework for parallel applications. You instrument your code with the Apollo API, tell Apollo what the tuning options are, and Apollo recommends tuning options so that your code runs faster.

Contributing

To contribute to Apollo please send a pull request on the develop branch of this repo. Apollo follows Gitflow for managing development.

Authors

Apollo is currently developed by Giorgis Georgakoudis, [email protected], Chad Wood, [email protected], and other contributors.

Apollo was originally created by David Beckingsale, [email protected]

If you are referencing Apollo in a publication, please cite this repo and the following papers:

• David Beckingsale and Olga Pearce and Ignacio Laguna and Todd Gamblin. Apollo: Reusable Models for Fast, Dynamic Tuning of Input-Dependent Code. In IEEE International Parallel & Distributed Processing Symposium (IPDPS'17), Orlando, FL, May 29-June 2 2017. LLNL-CONF-723337.
• Chad Wood and Giorgis Georgakoudis and David Beckingsale and David Poliakoff and Alfredo Giménez and Kevin A. Huck and Allen D. Mallony and Todd Gamblin. Artemis: Automatic Runtime Tuning of Parallel Execution Parameters Using Machine Learning. In In International Conference on High Performance Computing, pp. 453-472. Springer, Cham, 2021. LLNL-CONF-809192.

License

Apollo is distributed under the terms of both the MIT license. All new contributions must be made under the MIT license.

See LICENSE and NOTICE for details.

SPDX-License-Identifier: MIT

LLNL-CODE-733798

apollo's Issues

[mpi/opencv] Allow user supplied "results" for trials

I think this one would be pretty easy. Just add an "end" method to region that takes in a float/double. Instead of using the time the region took, use that float. It should be a user error if they call end with a value one time and not another in a Region.

Region-specific settings for training and models

Allow for regions to have tailored settings for such components as:

• training frequency
• re-training sensitivity
• tree-depth in the model

Update description

hey apollo team! I think it would be good to come up with a slightly tweaked description for the project, because it technically isn't online (incremental) learning. I was chatting in Mattermost with @davidbeckingsale about it. I don't know the project well enough to make a suggestion - but is it some kind of batch ML server perhaps?

Update RAJA for Apollo to the latest version

Hang when running with HPCToolkit+Apollo

On an x86+V100 system, using CUDA 11.0 and the develop branch of Apollo (currently 917cd5e ), Apollo breaks when used as a Kokkos tool if HPCToolkit is in use (a spack install of it). The app was ExaMiniMD, but it should work with just about anything. There are five configurations:

1. ./the_application
2. KOKKOS_PROFILE_LIBRARY=/path/to/libapollo-tuner.so ./the_application
3. hpcrun ./the_application
4. KOKKOS_PROFILE_LIBRARY=/some/other/kokkos-tool.so hpcrun ./the_application
5. KOKKOS_PROFILE_LIBRARY=/path/to/libapollo-tuner.so hpcrun ./the_application

And every one of them works except number 5. Just bizarre.

Create RAJA tests