GitLab

Create upload_pr_documentation.yml

* test_nvidia_transform: fix variable reference

`out_order` is the global parametrization list, not the test fixture argument

* Make `parametrize` use more idiomatic

* Use a more deterministic helper for `dim*` determination

* Convert NO_CUBLASLT errors into skips too

* Mark slow and benchmark tests as such (allows `-k "not benchmark"`)

`out_order` is the global parametrization list, not the test fixture argument

* Adjust Ruff configuration

* do not autofix always
* be less strict around tests and benchmarks
* adjust ignores for now

* Ruff: autofix I and F401

* Apply ruff autofixes

* Fix RUF013 complaint

* Fix mutable default in replace_linear

* Don't use bare except

* Wrap bitsandbytes.__main__ entrypoint in function; fix "sensible" typo

* Fix ruff B008 (function call in arguments)

* Add ruff noqas as suitable

* Fix RUF005 (splat instead of concatenating)

* Fix B018 (useless expression)

* Add pre-commit configuration + GitHub Actions lint workflow

* Fix unused `e` in bitsandbytes/__main__.py

* fix merge conflict resolution error

* run pre-commit hook

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

* Fix erroneous type aliasing

* Fix `Optional` typings (see PEP 484)

* Add Mypy ignores

* Fix Mypy complaints for method tables

* Fix type for get_ptr

* Fix various Mypy errors

* Fix missed call to is_triton_available

* fix library loading
Signed-off-by: Won-Kyu Park <wkpark@gmail.com>

* fixed library loading

 * use os.pathsep

* use glob(), search CUDA_PATH

* call find_file_recursive() without ext

---------
Signed-off-by: Won-Kyu Park <wkpark@gmail.com>
Co-authored-by: James Wyatt <Jamezo97@gmail.com>

* Fix `max_memory` example on README

- The new `max_memory` syntax expects a dictionary
- This change also accounts for multiple devices

* Fix model name in `from_pretrained` on README

* implicitly skip any test that implicitly uses CUDA on a non-CUDA box
* add a `requires_cuda` fixture

* Added install requirements to setup

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

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

Remove redundant key

Bump CUDA 12.2.0 to 12.2.1, fix setup support for Cuda 12.1 (#703), Sort compute capabilities sets to select max

* Add support for CUDA 12.1

* Update README to include CUDA 12.1 version

* Add support for >= 12x
Co-authored-by: Jeongseok Kang <jskang@lablup.com>

* Temporary version of bitsandbytes PR 527: Sort compute capabilities sets to select max

* Modify PR 506 to support C++20

* Add Cuda 12.2

---------
Co-authored-by: PriNova <info@prinova.de>
Co-authored-by: PriNova <31413214+PriNova@users.noreply.github.com>
Co-authored-by: Jeongseok Kang <jskang@lablup.com>

This PR adds initial FSDP support for training QLoRA models. It enables basic FSDP and CPU Offload support, with low memory training via FSDP.sync_module_states option unsupported.

This PR builds off of #840 commit 8278fca and BNB FSDP by @TimDettmers and @Titus-von-Koeller.

An example of using this PR to finetune QLoRA models with FSDP can be found in the demo repo: AnswerDotAi/fsdp_qlora.

* Minimal changes for fp32 4bit storage from BNB commit 8278fca

* Params4bit with selectable storage dtype

* possible fix for double quantizing linear weight & quant storage dtype

* minor fixes in Params4bit for peft tests

* remove redundant

* add float16

* update test

* Remove float16 quant cast as there are fp32, bf16, & fp16 quant kernels

---------
Co-authored-by: Kerem Turgutlu <keremturgutlu@gmail.com>

Delete .github/workflows/delete_doc_commment.yml

fix array index out of bounds in kgetColRowStats

Fix typo "quanitze"

Added scipy to requirements.txt

Make sure bitsandbytes handles permission errors in the right order

Add version attribute as per Python convention

Fix parameter name in error message

Update README.md

Update README.md

Fixed missing `Embedding` export

Fix typo in test_optim.py

improve `make clean` target

doc: Fix typo in how_to_use_nonpytorch_cuda.md