Factual package intelligence from PyPI

mps-sdpa 0.2.0 package intelligence

Fast drop-in scaled_dot_product_attention for Apple silicon (MPS) — wraps Apple's native MPSGraph SDPA with a zero-copy C++ bridge. 5-7x inference, 2-2.5x training.

pip install mps-sdpa

Current version
0.2.0
Python requirement
>=3.10
License
MIT
Distribution type
Pure Python wheel
Release files
2 (1 wheels, 1 source)
Download size
135.9 KiB
Release history
3 releases with files
Median release cadence
1 days

mps-sdpa dependencies

PyPI declares 12 unique dependency rules for this release. Environment markers are shown when supplied by the project.

  1. torch >=2.11.0
  2. numpy >=1.24
  3. psutil >=5.9
  4. ninja >=1.11 when sys_platform == "darwin"
  5. pyobjc-core >=10.0 when sys_platform == "darwin"
  6. pyobjc-framework-Metal >=10.0 when sys_platform == "darwin"
  7. pyobjc-framework-MetalPerformanceShadersGraph >=10.0 when sys_platform == "darwin"

Compatibility and release files

mps-sdpa publishes 1 wheel and 1 source archive for version 0.2.0. Wheel platform tags: any.

Declared Python classifiers: 3.10, 3.11, 3.12.

Security snapshot for mps-sdpa 0.2.0

An OSV query completed on and found 0 known vulnerabilities affecting this version. 0 advisories are classified as critical. 0 advisories appear in the CISA Known Exploited Vulnerabilities catalog.

Release activity and sources

PyPI lists 3 releases with files. The first dated release is ; 3 releases fall within the 365 days preceding the latest dated release. The current release files were uploaded on . The preceding dated release was .