Commit Graph

10 Commits (bbb648410ec35f335e7e807c07658d0a7ca0d8c6)

Author SHA1 Message Date
powderluv 2b52da951b
Link against libtorch (#955)
This moves torch-mlir to link against libtorch on macOS and linux

TESTS: Tests pass. Tested release builds on linux and macOS
2022-06-30 12:40:17 -07:00
powderluv d872f3e2ca
Build each OSX python version in an venv (#852)
Previously only system default versions were built. Now we build
binaries for both 3.9 and 3.10
2022-05-12 16:39:35 -07:00
powderluv e7f306ec2f
Use delocate to make portable wheels on OSX (#850)
Fix up wheels per python version on OSX
2022-05-12 14:16:32 -07:00
powderluv 0fb7a03ac9
Update build_macos_packages.sh
Set default OSX SDK to 11.0 not 11:0
2022-05-04 08:44:43 -07:00
powderluv fe1237b2a4
Provide a way to override MacOS target and arch (#818)
Useful when we are only building for one architecture.
2022-05-02 09:04:12 -07:00
powderluv 6d09c98b2f
Fix version information in Release builds (#788)
env vars seems to be lost in manylinux docker.
Use a version file like IREE does.
2022-04-25 14:13:17 -07:00
powderluv 7d9138f497
Update build_macos_packages.sh (#787)
Set the environment variable and export it since it doesn't seem to get passed down.
2022-04-22 15:48:03 -07:00
powderluv 4ef61aa27f
Minor buildsystem fixes (#778)
Sets up auto-pinning of latest torch-nightly
2022-04-21 15:53:00 -07:00
powderluv b03eac4224
Enable OSX (Intel, Apple Silicon Builds) (#776)
Update pinned pytorch version. Will submit a follow on PR to bump.
Also update artifacts directory
2022-04-21 10:47:28 -07:00
powderluv cc3a4a58ef
Add oneshot release snapshot for test/ondemand (#768)
* Add oneshot release snapshot for test/ondemand

Add some build scripts to test new release flow based on IREE.
Wont affect current builds, once this works well we can plumb it
in.

Build with manylinux docker

* Fixes a few issues found when debugging powderluv's setup.

* It is optional to link against Python3_LIBRARIES. Check that and don't do it if they don't exist for this config.
* Clean and auditwheel need to operate on sanitized package names. So "torch_mlir" vs "torch-mlir".
* Adds a pyproject.toml file that pins the build dependencies needed to detect both Torch and Python (the MLIR Python build was failing to detect because Numpy wasn't in the pip venv).
* Commented out auditwheel: These wheels are not PyPi compliant since they weak link to libtorch at runtime. However, they should be fine to deploy to users.
* Adds the --extra-index-url to the pip wheel command, allowing PyTorch to be found.
* Hack setup.py to remove the _mlir_libs dir before building. This keeps back-to-back versions from accumulating in the wheels for subsequent versions. IREE has a more principled way of doing this, but what I have here should work.

Co-authored-by: Stella Laurenzo <stellaraccident@gmail.com>
2022-04-21 02:19:12 -07:00