Releases: rapidsai/pynvjitlink
Releases · rapidsai/pynvjitlink
v0.4.0
- Fix building tests in multi-gpu environment (#98)
- Change cmake.verbose = true to build.verbose = true (#99)
- Use build-system.requires to set scikit-build-core minimum version (#100)
- Set CUDA version in one file (and use everywhere else) (#101)
- Drop Python 3.9 support (#102)
- Use CI workflow branch 'branch-24.10' again (#105)
- Use conda strict channel priority. (#109)
- Update to CUDA 12.6 (#97)
v0.3.0
- Rebuild with CUDA 12.5.1
- Drop CentOS 7
- Fix wheels to exclude
libcuda.so
- Add new
nvjitlink_version
function to determine the nvJitLink version built with - Exclude CMake 3.30.0 when building
- GHA improvements (including publishing directly to PyPI)
- Other minor cleanup
v0.2.4
What's Changed
- Add Python 3.12 by @bdice in #80
- Fix
__all__
by @gmarkall in #83 - Fix pre-commit copyright check by @bdice in #82
- Update to CUDA 12.5 by @gmarkall in #81
- Add xfail_strict to test suite by @vyasr in #84
- Use CI codeowners by @bdice in #85
- Use dependencies.yaml by @bdice in #86
- README: Don't use extra index for pip install by @gmarkall in #78
- use rapids-build-backend by @jameslamb in #88
- Bump version to 0.2.4 by @bdice in #89
Full Changelog: v0.2.3...v0.2.4