root
/
MNN
mirror of https://github.com/alibaba/MNN.git
1
0
Fork 0
Code Issues Actions 1 Packages Projects Releases Wiki Activity
2,219 Commits 20 Branches 109 Tags 396 MiB
Commit Graph

856 Commits

Author SHA1 Message Date
yanzhang 26667576aa fix: x86 and android build failure 
Change-Id: Ic4da99b67bcb4f785ef4493987398d5552b8f552
 2025-08-05 16:17:00 +08:00
yanzhang a52a884d6d fix: failures for other platforms 
Signed-off-by: yanzhang <yanzhang.wang@arm.com>
Change-Id: Id7b3da7f3d0c4ef88f4c48bfa87471eaab764973
 2025-08-05 15:30:22 +08:00
yanzhang f699ca2842 Replace the KleidiAI macro with the hint config 
Signed-off-by: yanzhang <yanzhang.wang@arm.com>
Change-Id: I9dcc4f68e1e67a11266b66589006650b197cde1e
 2025-08-04 15:58:03 +08:00
SixtyWang 380370fcf8 Fix bug in KleidiAIDenseConvolution, KleidiAIConvolution and QI4_SYM_CHNLQT_F32 
- Corrected outputWidth calculation in KleidiAIDenseConvolution
- Fixed use-after-free due to late call to getPostParameters in KleidiAIConvolution
- Resolved SME symmetry quantization kernel problem
 2025-07-31 16:56:51 +08:00
xiaying 433da8a041 MNN:Bugfix: Fix bug for GeometryReverse don't clear origin region 2025-07-30 12:03:40 +08:00
xiaying db0f559f9d MNN:Sync: Sync Internal 3.2.2 2025-07-23 14:33:57 +08:00
jxt1234 80a917a1b0
Merge pull request #3723 from jules-ai/feature/specify-cpu
android / android_build (push) Waiting to run Details
linux / linux_buil_test (push) Waiting to run Details
macos / macos_buil_test (push) Waiting to run Details
windows / windows_build_test (push) Waiting to run Details
ios / ios_build (push) Has been cancelled Details
wiki / sync_wiki (push) Has been cancelled Details 
new feature to allow user to specify cpu
 2025-07-21 10:05:24 +08:00
Jules c018eacc00 refactor cpuids setting from BackendConfig to HintMode 2025-07-18 09:42:59 +00:00
Jules 8907c96b4e convert power to cpuIds 2025-07-17 12:45:54 +00:00
yanzhang 8e7a63d622 Add imatmul fp16 support for DenseConv 
Change-Id: Ifb6972146a9c7013328fc75e30ab27c6e3d92d6a
Signed-off-by: yanzhang <yanzhang.wang@arm.com>
 2025-07-16 18:24:21 +08:00
Jules e69b0b5ce0 add windows preventing macro 2025-07-16 08:30:33 +00:00
Jules 19c97e3d15 remove unused variable initThreadNumber 2025-07-16 06:14:14 +00:00
Jules 110ae7f645 Implement cpuMask-based near-singleton global Thread Pool 2025-07-16 06:14:14 +00:00
Jules e1ad3feeae add MNNGetCPUMask 2025-07-16 05:42:17 +00:00
Jules d87ebbe227 bind cpuIds 2025-07-16 05:41:41 +00:00
Jules 757020fcf3 add cpuIds validation 2025-07-16 05:41:34 +00:00
Jules fa080f49e1 pass cpuIds to CPURuntime 2025-07-16 05:41:23 +00:00
yanzhang 4c9f48b76b Minor fixes after rebase. 
- Enable fp32 1x1 impl after it's right.
- Remove winograd memory test because kleidiai not support.

Change-Id: Ia61ad8c251490e93a2a365020a6183a944e769b2
Signed-off-by: yanzhang <yanzhang.wang@arm.com>
 2025-07-10 15:42:55 +08:00
yanzhang 5e3c8a3c12 Integrate the KleidiAI imatmul with fp32. 
Note,
        - No support for fp16 and int8 currently.

Signed-off-by: yanzhang <yanzhang.wang@arm.com>
Change-Id: If17c911977dd7eb0603f41d64b8ba879f468ab98
 2025-07-10 15:42:19 +08:00
SixtyWang b5b5845787 Refactor Kleidiai code to fix MNN unit test issues 
- Refactor getInstance function
- Add 1x1 convolution check in canAccelerate
- Use NHWC as input/output format for Kleidiai and convert format in onExecute
- Remove KAI_CONV_NCHW_IN_OUT macro
- Fix SME build issue on M4
 2025-07-08 16:22:23 +08:00
SixtyWang da8b7337c4 Refactor the KleidiAI integration ConvInt8 code in MNN 2025-07-08 16:21:26 +08:00
Shuheng Deng 875814bfb9 Refactor the KleidiAI integration convolution code in MNN 
Change-Id: I2e45fb7ec10793afffad8c25c305c1cb1e260753
 2025-07-04 09:54:00 +08:00
王召德 5f35d621e1
Merge pull request #3670 from yanzhang-dev/features/upgrade-kai-1.9.0
android / android_build (push) Has been cancelled Details
ios / ios_build (push) Has been cancelled Details
linux / linux_buil_test (push) Has been cancelled Details
macos / macos_buil_test (push) Has been cancelled Details
windows / windows_build_test (push) Has been cancelled Details 
Upgrade KleidiAI to v1.9.0
 2025-07-02 17:54:46 +08:00
emil fa3a03d6e3 只将OpenCL的context作为全局变量，并且只创建一次。将OpenCLRutime修改为每次创建新对象 2025-06-30 11:13:51 +08:00
yanzhang 9a56fe323e Upgrade KleidiAI to v1.9.0 
Change-Id: Ia910d9c296aa1569e9e4449b56bb7614fe6c85e0
Signed-off-by: yanzhang <yanzhang.wang@arm.com>
 2025-06-26 13:46:40 +08:00
Jules 38c15d853b align OneDNNConvInt8 with updated function signature 2025-06-24 01:46:34 +00:00
xiaying aeac75acbf [MNN:Bugfix] Fix opencl execute llm decode error (issue 3623) 2025-06-20 14:41:23 +08:00
jxt1234 3db3cc904d
Merge pull request #3635 from alibaba/feature/sync
android / android_build (push) Has been cancelled Details
ios / ios_build (push) Has been cancelled Details
linux / linux_buil_test (push) Has been cancelled Details
macos / macos_buil_test (push) Has been cancelled Details
pymnn-linux / pymnn_linux_buil_test (push) Has been cancelled Details
pymnn-macos / pymnn_macos_buil_test (push) Has been cancelled Details
pymnn-windows / pymnn_windows_buil_test (push) Has been cancelled Details
wiki / sync_wiki (push) Has been cancelled Details
windows / windows_build_test (push) Has been cancelled Details 
MNN:Sync: Sync Internal 3.2.1
 2025-06-17 11:30:08 +08:00
xiaying e5e7fccd99 MNN:Sync: Sync Internal 3.2.1 2025-06-17 11:08:21 +08:00
Jules b9a22516bf fix typo of operator 2025-06-17 02:10:14 +00:00
Jules d30f99a2ae remove unused variables 2025-06-17 02:06:28 +00:00
jxt1234 f9720f994e
Merge pull request #3588 from jules-ai/fix_int_offset_overflow 
a propose to fix address offset overflow caused by using int
 2025-06-11 13:54:04 +08:00
xiaying 210e861650 MNN:Sync: Sync some bugfix 2025-06-09 09:51:46 +08:00
xiaying b66ec85018 OpenCL:Bugfix: Fix bug for recordable bug for llm 2025-06-06 08:48:18 +08:00
xiaying 09b6cd9862 MNN:Sync: Sync 3.2.0 2025-06-05 15:15:29 +08:00
Jules 7a08d2e2ee Fix address offset overflow caused by using int 2025-05-30 10:38:23 +00:00
jxt1234 97223628bc
Merge pull request #3569 from linlinxu-1989/sme_fp16_fp32_new 
Bump MNN KleidiAI ukernel to fp16_sme2 fp32_sme2 ukernel
 2025-05-27 15:37:07 +08:00
jxt1234 c0dde12d89
Merge pull request #3564 from tumuyan/patch-1 
fix MNN_SUPPORT_RENDER OFF
 2025-05-27 15:07:56 +08:00
jxt1234 eef4a726cd
Merge pull request #3546 from HenryDen/buf_fix 
Bugs fix for the KleidiAi integration of MNN
 2025-05-27 14:54:37 +08:00
linlin.xu 556acbfea2 Add macro for kleidiai only resource 2025-05-27 14:46:50 +08:00
linlin.xu a8f2d34b15 Bump MNN KleidiAI ukernel to fp16_sme2 fp32_sme2 ukernel 
Add fp16 and fp32 sme2 kernel
1. Add rhs pack nxk kernel lhs pack kernel
2. Add fp16 GEMM kernel and GEMV kernel
3. Add fp32 GEMM kernel and GEMV kernel
 2025-05-27 10:21:24 +08:00
tumuyan caf73d6859
fix MNN_SUPPORT_RENDER OFF 2025-05-24 10:42:48 +08:00
Jules 03db1c2769
fix blockSize==1 assertion for DepthToSpace 2025-05-23 19:26:18 +08:00
zhaode.wzd bd36a3f749 [MNN:Sync] Sync internal: 
1. SmolVLM, FastVLM support.
    2. QNN backend init.
    3. Qwen3 MoE support.
    4. Speculative decodeing init.
    5. Some bugfix.
 2025-05-23 15:24:18 +08:00
Shuheng Deng 986a065fff Bugs fix for the KleidiAi integration of MNN 2025-05-22 17:35:57 +08:00
xiaying d2477b3091 MNN:Bugfix: Fix bug for ios Compile error 2025-05-09 14:19:41 +08:00
xiaying e457f0ce33 OpenCL:Bugfix: Fix bug for llm bench opencl crash 2025-05-09 14:12:58 +08:00
zhaode.wzd a019d971ad [MNN:Sync] Sync Internal 3.1.4. 2025-05-08 12:39:44 +08:00
xiaying ba766b8bea LLM:Feature, OpenCL:Bugfix: Sync internal bugfix( support llm prompt 
dynamic setting, fix bug for opencl softmax in nvidia error)
 2025-04-30 18:00:38 +08:00
jxt1234 f65ff90302
Merge pull request #3397 from xhzheng1895/kai_1.7.0_integration
android / android_build (push) Waiting to run Details
ios / ios_build (push) Waiting to run Details
linux / linux_buil_test (push) Waiting to run Details
macos / macos_buil_test (push) Waiting to run Details
windows / windows_build_test (push) Waiting to run Details 
Integrate Asym int4 ukernels to MNN
 2025-04-30 11:15:41 +08:00