GitLab

* Add aarch64 cpu tests and CUDA build to nightly workflow

* aarch64: limit CUDA targets to sm75, sm80, sm90, sm100

* aarch64: limit CUDA targets to sm75, sm80, sm90, sm100

* Update build cpu script

* fix

* Update auditwheel for aarch64

* Update python-package.yml

* Update python-package.yml

* Update python-package.yml

* Cleanup

* Matrix update

---------
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>

* Stop building for CUDA toolkit < 11.8

* Simplify

* Drop sm70 from cu128 build targets to align with pytorch

* Drop Python 3.8 support.

* Formatting

* blackwell

* blackwell

* Update python-package.yml

* chore: move configs to pyproject.toml

* fix: drop file from CI workflow

* feat: reorder pytest markers

* chore: retain comments

* chore(build): migrate build data to pyproject
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>

* Apply suggestions from code review
Co-authored-by: Aarni Koskela <akx@iki.fi>

* chore: move configs to pyproject.toml

* Apply suggestions from code review
Co-authored-by: Aarni Koskela <akx@iki.fi>

* bump ruff

---------
Co-authored-by: Matthew Douglas <38992547+matthewdouglas@users.noreply.github.com>
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Start of int8 refactor: remove col32/col_ampere/col_turing transforms in new igemmlt implementation

* Fix unintended change

* New naive mm_dequant kernel for row-major; cleanup

* fix

* int8 refactor: initial sparse decomp, cleanup

* Int8 refactoring: remove separate NO_CUBLASLT build; more cleanup

* int8: inference optimizations, some cleanup

* int8: more tests passing, cleanup

* int8 - more cleanup, most tests passing

* int8: specify CUDA stream for int8 ops

* perf: reduce overhead from getting cudaStream ptr

* Mark some functions for deprecation.

* int8 sparse decomp: small perf improvement

* update setup.py

* Update bitsandbytes/autograd/_functions.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update bitsandbytes/research/autograd/_functions.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* int8 - perf improvement for sparse decomposition inference; deprecate get_tensor_stream() in favor of new private fn

* int8 cleanup

* Ignore ruff rule ISC001 (incompatible with formatter)

* add comment

* int8 more cleanup

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* int8: rename / deprecate old fn signatures

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* type annotation

* format update

* Update bitsandbytes/research/autograd/_functions.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* cleanup

* Add comment to explain division optimization

* more cleanup

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update bitsandbytes/functional.py
Co-authored-by: Aarni Koskela <akx@iki.fi>

* cleanup

* Type annotations, cleanup

* remove unused kernels; improved type annotations

* small perf optimization for single-GPU systems

* small perf optimization for single-GPU systems

* update docstrings

* Improve docs and tests

* Update docstring

* Update test

* add benchmarking script

* test cleanup: add deprecated marker, move benchmarks out

* Add int8 dequant function; misc improvements

* int8 matmul fallback for inner dims not divisible by 4

* improve register usage of kInt8VectorQuant - especially for A100/H100

* disable fail-fast for package build

* maxwell compat

* ptxas verbose

* docs update

* doc update

* backward fix

* Bugfix sparse decomp

* Int8 fix for PEFT OLoRA init

* Fix test for deprecated spmm_coo

* test improvement

* doc update

* typo

* doc cleanup

* docs

* add inference benchmark script

* Add benchmarks, doc update

---------
Co-authored-by: Aarni Koskela <akx@iki.fi>

* [Build] Add CUDA 12.6.2 build; update 12.5.0 to 12.5.1

* bump cuda-toolkit action version

* Update docs for cuda versions

* CI/CD: Add step to publish wheels on tag creation

* Remove file

* Restrict pre-release workflow branches

* Update PyPI publishing

* Update PyPI publishing

* Update package workflow name

* continuous pre-release only on main

* Add CUDA 12.5 builds and enable CUDA 12.4 on Windows

* Update install doc

Closes #1114
Co-authored-by: Titus von Koeller <9048635+Titus-von-Koeller@users.noreply.github.com>

* (ci) build with wider CUDA version matrix

* (ci) build with wider CUDA version matrix

* (ci) skip sm_89 target on CUDA 11.7

* (ci) skip sm_90 target on CUDA 11.8

* modify workflow to publish to test.pypi

* (build) Test for manylinux_2_24 build on GH actions

* (build) got that backwards.

* try fixing manual triggering condition for testpypi

* try if Ubuntu 18.04 is an easy fix to allow for `manylinux_2_24` compatibility

* hardcode publish step to run to test publishing

* set ubuntu to newest supported version

* try statically linking libstdc++ to achieve manylinux_2_18

* last commit only brought us to manylinux_2_34, reverse

* add misssing permission for publishing to pypi

* snake case deprecated in favor of kebab

* downgrade cuda ubuntu aiming for manylinux_2_24

* add step to upgrade cmake due to old Ubuntu for CUDA build

* adjust path to prefer pip installed cmake

* (cmake) set CMAKE_BUILD_TYPE=Release if unspecified

* default to CMAKE_BUILD_TYPE Release for optimized releases and better many_linux compatibility

* (build) back to ubuntu22.04 docker images

* verify Cmake in separte step

* add clarifying comment about Python version compatibility

* (build) we don't need cmake for wheel step

* fixup testpypi publish to run in PR for testing

* add pypi publishing when tagged on main

* add functionality to rewrite platform tags

* (ci) adjust platform tags for wheels

* fix for windows, get order right.

* fix for windows, get order right.

* (build) slim down those fatbins on windows cuda

* sloppy

* remove broken PyPi upload for now

---------
Co-authored-by: Titus von Koeller <9048635+Titus-von-Koeller@users.noreply.github.com>

Co-authored-by: wkpark <wkpark@gmail.com>

* (cmake) Fix generation of targets for nvcc

* Typo

* (ci) linux + CUDA workflow: make sure we specify target architectures

* fix

* fix one more time

* (cmake) Default in CMAKE_CUDA_ARCHITECTURES_ALL when cmake<3.23, make sure we build only selected cubins and only ptx for latest capability

* Fix static lookup for CMAKE_CUDA_ARCHITECTURES_ALL on cmake<3.23

* Remove debug setting

* clarification

* CI: fix cuda-toolkit speed issue

* CI: use MSVC instead msbuild to remove 'visual_stuido_integration' dependency

 * use Ninja to compile without MS toolset

* use 'network', install 'ninja' only
Co-authored-by: Rickard <rickardp@users.noreply.github.com>

---------
Co-authored-by: Rickard <rickardp@users.noreply.github.com>

* Make native code portable and add GitHub workflow for building

* Removed deprecated Python versions

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml
Co-authored-by: Aarni Koskela <akx@iki.fi>

* Update python-package.yml

* Do not test on Python 3.13 until released

* Update python-package.yml

* Update python-package.yml

* Update python-package.yml

* Update python-package.yml

* Refactor build stage

* Fixed breaking actions change

* Slim down Windows cuda

* Create dependabot.yml

* Bespoke local dev requirements.txt

* Enable VS integration

* Group Dependabot updates

* Cleanup

* Update python-package.yml

* Reinstate file that was wrongly merged

* Fixed regression caused by new version of download-artifact

* Update python-package.yml

* Update python-package.yml

* Fix matrix

* Update python-package.yml

* Merge

* Pipeline

* Fixed conflict

* Fixed conflict

* Update CMakeLists.txt

* Fixed merge error

* cleanup

* cleanup

* Find CUDA

* Fix

* Fixing merge error from latest merge from main

* Fix setup.py

* Fixed typo in artifact name

* Remove linker flags

* Build nocublaslt versions

* Fixed formatting

* Fixed VS Code format on save

* Ran format on save from VScode

* Re-saved the json files using the new settings

* Re-saved CMakeLists.txt to get formatting right

* Add path filter

* Formatting

---------
Co-authored-by: Aarni Koskela <akx@iki.fi>