(reference-only) Multi backend refactor -> main (full diff of all already merged PRs)#1220

Open

Titus-von-Koeller wants to merge 283 commits intomainfrom

multi-backend-refactor

This pull request is big! We're only showing the most recent 250 commits

Commits on Dec 5, 2023

Commits on Jan 12, 2024

Commits on Jan 17, 2024

Merge pull request #3 from pnunna93/fix_igemmlt_int
amathews-amd
authored

Commits on Jan 26, 2024

Commits on Jan 27, 2024

Skip failing tests for now
pnunna93
committed

Commits on Jan 30, 2024

Merge pull request #4 from ROCm/IFU-master-2024-01-24
amathews-amd
authored

Commits on Feb 2, 2024

ops.hip: adapt to enum naming changes in ROCm/hipBLASLt@95131d6 and ROCm/hipBLASLt@3aad0d8
iiisak
committed

Commits on Feb 5, 2024

Merge remote-tracking branch 'main/main' into upstream_device_abstraction
jianan-gu
committed

Commits on Feb 8, 2024

Commits on Feb 12, 2024

Enable estimate quantile tests
pnunna93
committed

Commits on Feb 13, 2024

Commits on Feb 15, 2024

fix py38 vers incompatibility from other PR
Titus-von-Koeller
committed

Commits on Feb 17, 2024

factor out ensure_backend_is_available, exc instead of assert
Titus-von-Koeller
committed

Commits on Feb 19, 2024

Merge pull request #6 from ROCm/rocwmma_merge
Lzy17
authored

Commits on Feb 20, 2024

Enable transpose flag for row to col transform
pnunna93
committed
Update descriptors for transpose flag
pnunna93
committed
revert nvidia_transform to transform
pnunna93
committed
update changes
$U-AMD\zhaoyili$ $U-AMD\zhaoyili$
U-AMD\zhaoyili
authored and
U-AMD\zhaoyili
committed

Commits on Feb 21, 2024

Remove minor device filter to avoid confusion
jianan-gu
authored

Commits on Feb 23, 2024

Merge pull request #8 from ROCm/enable_transform_with_transpose
pnunna93
authored
fixed minor mistakes
$U-AMD\zhaoyili$ $U-AMD\zhaoyili$
U-AMD\zhaoyili
authored and
U-AMD\zhaoyili
committed
Merge pull request #9 from ROCm/rocm_enabled_fix_bfloat16
pnunna93
authored

Commits on Mar 6, 2024

Commits on Mar 12, 2024

Commits on Mar 18, 2024

improve the gemv 4bit accuracy by forcing the hipcub to 32
$U-AMD\zhaoyili$ $U-AMD\zhaoyili$
U-AMD\zhaoyili
authored and
U-AMD\zhaoyili
committed

Commits on Mar 19, 2024

Commits on Apr 3, 2024

Commits on Apr 4, 2024

Merge remote-tracking branch 'upstream/main' into IFU-master-2024-03-28
pnunna93
committed

Commits on Apr 9, 2024

update instructions
$U-AMD\zhaoyili$ $U-AMD\zhaoyili$
U-AMD\zhaoyili
authored and
U-AMD\zhaoyili
committed
Merge pull request #19 from ROCm/updated_readme
amathews-amd
authored
Update README.md
pnunna93
committed
Merge branch 'rocm_enabled' into IFU-master-2024-03-28
pnunna93
authored
fix PEP errors
pnunna93
committed
Fix typos
pnunna93
committed
Merge branch 'IFU-master-2024-03-28' of https://github.com/ROCm/bitsandbytes into IFU-master-2024-03-28
pnunna93
committed

Commits on Apr 10, 2024

Fix formatting in README file
pnunna93
authored

Commits on Apr 11, 2024

Commits on Apr 15, 2024

Commits on Apr 25, 2024

Fix lint issues
Xia-Weiwen
committed

Commits on Apr 26, 2024

Commits on Apr 29, 2024

implement get_rocm_gpu_arch
root
committed

Commits on Apr 30, 2024

fixing lint
root
committed
fixing lint
root
committed
correct lint error
root
committed
Merge pull request #21 from ROCm/rocm_enabled_arch_detect
pnunna93
authored

Commits on May 3, 2024

merge changes from main
Titus-von-Koeller
committed

Commits on May 7, 2024

Commits on May 8, 2024

Commits on May 11, 2024

Commits on May 15, 2024

Add comments for HIP changes
pnunna93
committed

Commits on May 24, 2024

Commits on May 27, 2024

Add empty stubs for Ascend NPU
ji-huazhong
committed

Commits on May 28, 2024

Commits on May 29, 2024

fix blocksize
jiqing-feng
committed

Commits on May 30, 2024

Merge pull request #1228 from jiqing-feng/4bit
Titus-von-Koeller
authored

Commits on Jun 6, 2024

CPU: add torch.compile for F.double_quant and F.quantize_4bit (#1238)
Xia-Weiwen
authored

Commits on Jun 21, 2024

Commits on Jul 10, 2024

CPU/XPU: disable torch.compile if g++ is not available (#1251)
Xia-Weiwen
authored

Commits on Jul 12, 2024

Create build job for ROCm (#1255)
pnunna93
authored

Commits on Jul 24, 2024

Changelog: add explanation r. QLoRA mem savings
Titus-von-Koeller
committed

Commits on Jul 26, 2024

merge `main` into `multi-backend-refactor`
Titus-von-Koeller
committed

Commits on Jul 30, 2024

allow features flags on bnb
Titus-von-Koeller
committed

Commits on Aug 1, 2024

Fix dequant 4bit (#1300)
jiqing-feng
authored

Commits on Aug 2, 2024

Commits on Aug 14, 2024

add bnb attribute to expose supported devices
Titus-von-Koeller
committed

Commits on Aug 20, 2024

fix 4bit dtype (#1325)
jiqing-feng
authored

Commits on Aug 26, 2024

docs: tweaks for multi-backend preview release prep
Titus-von-Koeller
committed

Commits on Aug 29, 2024

Commits on Aug 30, 2024

actions: update permissions for pr docs publishing
Titus-von-Koeller
committed

Commits on Sep 13, 2024

fix nf4 memory issue by init op_context in forward (#1349)
jiqing-feng
authored

Commits on Sep 16, 2024

AMD: Clarify diagnostic messages; free up disk space for CI build

authored

Commits on Sep 19, 2024

check grad before using ipex (#1358)
jiqing-feng
authored

Commits on Sep 20, 2024

Enable packaging for ROCm 6.2 (#1367)
pnunna93
authored

Commits on Sep 25, 2024

Update for VS2022 17.11 compatibility with CUDA < 12.4 (#1341)

matthewdouglas
authored and
Titus-von-Koeller
committed

Commits on Sep 26, 2024

Commits on Oct 1, 2024

docs: remove 2 obsolete lines
Titus-von-Koeller
committed

Commits on Oct 16, 2024

Remove depth option in installation steps (#1395)

authored

Commits on Nov 20, 2024

Fix issue that no valid semantic version tag found when installing bitsandbytes from source in personal repo (#1419)
ji-huazhong
authored

Commits on Nov 29, 2024

Enable XPU and optimize cpu/xpu op (#1418)

authored

Commits on Dec 2, 2024

fix cpu nf4 (#1432)
jiqing-feng
authored

Commits on Dec 6, 2024

Add Ascend NPU support for nf4 quant (#1422)

authored

Commits on Dec 17, 2024

fix device check (#1453)
jiqing-feng
authored

Commits on Jan 22, 2025

Enable double quant on Intel CPU and XPU (#1472)
jiqing-feng
authored

Commits on Jan 28, 2025

Commits on Feb 7, 2025

Commits on Feb 11, 2025

fix xpu woq linear dtype (#1506)
jiqing-feng
authored

Commits on Feb 20, 2025

fix version (#1532)
jiqing-feng
authored

Commits on Mar 4, 2025

enable benchmark script (#1554)

jiqing-feng
and
Titus-von-Koeller
authored

Commits on Mar 18, 2025

Fix XPU 4bit (#1567)
jiqing-feng
authored

Commits on Mar 24, 2025

Fix xpu to cpu (#1570)
jiqing-feng
authored

Commits on Apr 2, 2025

fix double compress 8bit precision (#1582)
jiqing-feng
authored

Commits on Apr 8, 2025

Remove error log for Intel CPU/XPU (#1503)
jiqing-feng
authored

Commits on Apr 15, 2025

Commits on Apr 21, 2025

fix log (#1604)
jiqing-feng
authored

Commits on Apr 29, 2025

fix xpu ipex linear in torch2.7 (#1618)
jiqing-feng
authored

Commits on May 5, 2025

update compute_type_is_set attr (#1623)
ckvermaAI
authored

Commits on May 20, 2025

supports HPU double quant (#1630)
rsshaik1
authored