| Name | pyatf JSON |
| Version |
0.0.2
JSON |
| download |
| home_page | None |
| Summary | Auto-Tuning Framework (ATF) is a generic, general-purpose auto-tuning approach for programs whose tuning parameters may be constrained |
| upload_time | 2024-04-26 08:59:00 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | >=3.9 |
| license | None |
| keywords |
auto-tuning
constraints
optimization
performance
tuning
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# pyATF: The Auto-Tuning Framework (ATF) in Python
Auto-Tuning Framework (ATF) is a generic, general-purpose auto-tuning approach that automatically finds well-performing values of performance-critical parameters (a.k.a. tuning parameters), such as sizes of tiles and numbers of threads.
ATF works for programs written in arbitrary programming languages and belonging to arbitrary application domains, and it allows tuning for arbitrary objectives (e.g., high runtime performance and/or low energy consumption).
A major feature of ATF is that it supports auto-tuning programs whose tuning parameters have *interdependencies* among them, e.g., the value of one tuning parameter has to be smaller than the value of another tuning parameter.
For this, ATF introduces novel process to *generating*, *storing*, and *exploring* the search spaces of interdependent tuning parameters (discussed in detail [here](https://dl.acm.org/doi/abs/10.1145/3427093)).
ATF comes with easy-to-use user interfaces to make auto-tuning appealing also to common application developers.
The Interfaces are based on either:
1. *Domain-Specific Language (DSL)*, for auto-tuning at compile time (a.k.a. offline tuning) (discussed [here](https://onlinelibrary.wiley.com/doi/full/10.1002/cpe.4423?casa_token=FO9i0maAi_MAAAAA%3AwSOYWsoqfLqcbazsprmzKkmI5msUCY4An5A7CCwi-_V8u10VdpgejcWuiTwYhWnZpaCJZ3NmXt86sg));
2. *General Purpose Language (GPL)*, for auto-tuning at runtime (a.k.a. online tuning), e.g., of *C++ programs* (referred to as [cppATF](todo), and discussed [here](https://ieeexplore.ieee.org/abstract/document/8291912)) or *Python programs* (referred to as [pyATF](todo)).
**The full GitHub repository for *pyATF*, i.e., ATF with its GPL-based Python interface can be found [here](https://github.com/atf-tuner/pyATF).**
## Documentation
The full documentation is available [here](https://atf-tuner.org/pyATF/).
## Installation
pyATF requires Python 3.9+ and can be installed using `pip`:
pip install pyatf
pyATF's pre-implemented OpenCL and CUDA cost functions require additional packages to be installed:
- OpenCL cost function:
pip install numpy pyopencl
For the OpenCL cost function, a matching OpenCL runtime is also required, e.g., for Intel CPUs:
pip install intel-opencl-rt
- CUDA cost function:
pip install numpy cuda-python
Raw data
{
"_id": null,
"home_page": null,
"name": "pyatf",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": null,
"keywords": "auto-tuning, constraints, optimization, performance, tuning",
"author": null,
"author_email": "Richard Schulze <r.schulze@uni-muenster.de>, Ari Rasch <a.rasch@uni-muenster.de>",
"download_url": "https://files.pythonhosted.org/packages/c5/b1/10379290a177c44b6a335fa4d47110229d81a2659d9b3ea84bfd6af1160d/pyatf-0.0.2.tar.gz",
"platform": null,
"description": "# pyATF: The Auto-Tuning Framework (ATF) in Python\n\nAuto-Tuning Framework (ATF) is a generic, general-purpose auto-tuning approach that automatically finds well-performing values of performance-critical parameters (a.k.a. tuning parameters), such as sizes of tiles and numbers of threads.\nATF works for programs written in arbitrary programming languages and belonging to arbitrary application domains, and it allows tuning for arbitrary objectives (e.g., high runtime performance and/or low energy consumption).\n\nA major feature of ATF is that it supports auto-tuning programs whose tuning parameters have *interdependencies* among them, e.g., the value of one tuning parameter has to be smaller than the value of another tuning parameter.\nFor this, ATF introduces novel process to *generating*, *storing*, and *exploring* the search spaces of interdependent tuning parameters (discussed in detail [here](https://dl.acm.org/doi/abs/10.1145/3427093)).\n\nATF comes with easy-to-use user interfaces to make auto-tuning appealing also to common application developers.\nThe Interfaces are based on either:\n1. *Domain-Specific Language (DSL)*, for auto-tuning at compile time (a.k.a. offline tuning) (discussed [here](https://onlinelibrary.wiley.com/doi/full/10.1002/cpe.4423?casa_token=FO9i0maAi_MAAAAA%3AwSOYWsoqfLqcbazsprmzKkmI5msUCY4An5A7CCwi-_V8u10VdpgejcWuiTwYhWnZpaCJZ3NmXt86sg));\n2. *General Purpose Language (GPL)*, for auto-tuning at runtime (a.k.a. online tuning), e.g., of *C++ programs* (referred to as [cppATF](todo), and discussed [here](https://ieeexplore.ieee.org/abstract/document/8291912)) or *Python programs* (referred to as [pyATF](todo)).\n\n**The full GitHub repository for *pyATF*, i.e., ATF with its GPL-based Python interface can be found [here](https://github.com/atf-tuner/pyATF).**\n\n## Documentation\n\nThe full documentation is available [here](https://atf-tuner.org/pyATF/).\n\n## Installation\n\npyATF requires Python 3.9+ and can be installed using `pip`:\n\n pip install pyatf\n\npyATF's pre-implemented OpenCL and CUDA cost functions require additional packages to be installed:\n\n- OpenCL cost function:\n\n pip install numpy pyopencl\n\n For the OpenCL cost function, a matching OpenCL runtime is also required, e.g., for Intel CPUs:\n\n pip install intel-opencl-rt\n\n- CUDA cost function:\n\n pip install numpy cuda-python",
"bugtrack_url": null,
"license": null,
"summary": "Auto-Tuning Framework (ATF) is a generic, general-purpose auto-tuning approach for programs whose tuning parameters may be constrained",
"version": "0.0.2",
"project_urls": {
"Homepage": "https://atf-project.org/",
"Issues": "https://github.com/atf-tuner/pyATF/issues"
},
"split_keywords": [
"auto-tuning",
" constraints",
" optimization",
" performance",
" tuning"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "2cedab627987829a0454f7efbcee87ccae509dabc1eff60954aac2a907e1e61e",
"md5": "bae7f0e5a7e07d8c690ed1a425594fcd",
"sha256": "7207930993ec26930e56d11613513881593524f4219b5c8b1db6336d377ad6d3"
},
"downloads": -1,
"filename": "pyatf-0.0.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "bae7f0e5a7e07d8c690ed1a425594fcd",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 39743,
"upload_time": "2024-04-26T08:58:58",
"upload_time_iso_8601": "2024-04-26T08:58:58.753472Z",
"url": "https://files.pythonhosted.org/packages/2c/ed/ab627987829a0454f7efbcee87ccae509dabc1eff60954aac2a907e1e61e/pyatf-0.0.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "c5b110379290a177c44b6a335fa4d47110229d81a2659d9b3ea84bfd6af1160d",
"md5": "effdacb817547ed71cc3a57c6cbc127a",
"sha256": "09d4fe65fe6d01d33e81a5eff7aa6494039fe7dbdaa1217696e7c0bdbcd75db0"
},
"downloads": -1,
"filename": "pyatf-0.0.2.tar.gz",
"has_sig": false,
"md5_digest": "effdacb817547ed71cc3a57c6cbc127a",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 45680,
"upload_time": "2024-04-26T08:59:00",
"upload_time_iso_8601": "2024-04-26T08:59:00.666308Z",
"url": "https://files.pythonhosted.org/packages/c5/b1/10379290a177c44b6a335fa4d47110229d81a2659d9b3ea84bfd6af1160d/pyatf-0.0.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-04-26 08:59:00",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "atf-tuner",
"github_project": "pyATF",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "pyatf"
}
