nihui
4c3debae2d
multiheadattention scale param ( #5526 )
* update swiftshader
* skip vs2017 swiftshader
1 year ago
nihui
f2a34ee7ae
update pybind11 to 2.12, support numpy 2 ( #5525 )
1 year ago
nihui
f56b18aaf0
pnnx convert onnx resize upsample ( #5522 )
1 year ago
Xyzhao
fbd6690d6c
fix: add NCNN_PLATFORM_API macro for VkAndroidHardwareBufferImageAllocator ( #5521 )
1 year ago
nihui
8235cad999
mha allow qdim differs from embed_dim ( #5519 )
* test mha oom
1 year ago
nihui
2828e7ae96
pnnx reset onnx input shape, convert torch.tile torch.where ( #5517 )
* pnnx reset onnx input shape
* eliminate noop cast
1 year ago
CharlieYu
b786af56f8
benchmark: add raspberry pi 5 benchmark after GPU overclock ( #5518 )
1 year ago
nihui
ffb2fe60ee
pnnx convert onnx pad linear sigmoid softmax relu ( #5516 )
* pnnx convert onnx pad
* pnnx convert onnx linear sigmoid softmax relu
* old onnx softmax
1 year ago
nihui
21babb7eed
pnnx convert onnx conv convtranspose ( #5515 )
1 year ago
nihui
e389a15846
update qcom855plus benchmark
1 year ago
nihui
39c27de47b
test concat oom ( #5502 )
1 year ago
nihui
093c516898
test slice oom ( #5501 )
1 year ago
Wei Wu
bb54d575a0
Update ruapu.h to the latest version. ( #5499 )
The updated ruapu adds support for multiple architectures such as RISC-V, MIPS, and Loongson, and can detect more Arm features.
The latest version is 10b02b3755 .
1 year ago
nihui
da7d1a10f7
test x86 arm convolution oom ( #5492 )
* skip mips loongarch riscv oom test atm
* test softmax oom
1 year ago
nihui
03ca9053c1
Update linux-x64-cpu-gcc.yml
1 year ago
nihui
102e98970f
fix unexpected abs error on powerpc vsx ( #5498 )
1 year ago
nihui
19ea54f266
more x86 vnni optimization for lstm ( #5496 )
* workaround vs2019 crash
1 year ago
nihui
debc33fee2
arm handle allocation failures ( #5490 )
2 years ago
nihui
b4379630fb
x86 handle allocation failures ( #5489 )
2 years ago
nihui
75ba2a01c9
pnnx fill tensor data for ort inference, fix output shape2 mismatch ( #5488 )
2 years ago
TianZer
b0de947b32
fix mingw64 avx crash and termux build issue ( #5464 )
* Remove two potential warnings for VisualStudio
* fix mingw64 avx crash
* fix build issue in termux
2 years ago
Asd-g
bbf2e5d533
create_gpu_instance: do not perform destroy_gpu_instance() ( #5437 )
When performing destroy_gpu_instance(), g_instance.created is always 0.
2 years ago
nihui
5b49d0aa0a
pnnx onnx fold partial shape constant, fix dce node lost bug ( #5486 )
2 years ago
nihui
740f437462
pnnx fuse simple onnx shape slice chain, fix squeeze dim list ( #5485 )
2 years ago
nihui
bc4dce7a84
Update README.md, add harmonyos build link
2 years ago
nihui
e009c36155
fix onnx bn silu ( #5483 )
2 years ago
nihui
5e9bf90455
pnnx onnx with inputshape ( #5480 )
2 years ago
村长大人
1e75a2df21
add harmonyos how to build with vulkan ( #5475 )
2 years ago
nihui
f40d5414d5
pnnx build onnx2pnnx ( #5461 )
2 years ago
nihui
86e1ed067a
add harmonyos ci ( #5474 )
2 years ago
RoachZhao
d4292e9a65
Update vulkan-notes.md ( #5472 )
`compute_queue_count` is a function.
2 years ago
nihui
c4a007406d
windows clang ci ( #5469 )
* windows clang ci
* clang msvc use x86intrin.h for xop
* test arm64 compiler features
2 years ago
nihui
1b7e6358c9
pnnx load dynamo onnx of segmentation models ( #5458 )
2 years ago
nihui
432a8d9aa1
pnnx replace ncnn input long type with int ( #5457 )
2 years ago
nihui
ab088e05b8
pnnx do not fold aliased tensors with inplace op ( #5455 )
2 years ago
nihui
6110d41b19
pnnx bf16 data type ( #5456 )
2 years ago
nihui
96e0e8f203
add pnnx torch 2.3 ci ( #5454 )
2 years ago
nihui
08b7d99a75
rnn/lstm/gru dynamic quantization ( #5435 )
2 years ago
Tabbleman
be15dbe421
add riscv-gnu-toolchain build guide;-) ( #5446 )
2 years ago
Tabbleman
b8fefb977d
clear warning: unused variable while building on x86-wsl platform ( #5444 )
2 years ago
quink
e31be492d5
c_api: Fix function prototypes with no argument ( #5436 )
Here is a big difference between C and C++. foo() in C means that
the funtion takes an unspecified number of arguments, while
foo(void) means the function takes no argument.
Fix -Wstrict-prototypes warning.
Signed-off-by: Zhao Zhili <zhilizhao@tencent.com>
2 years ago
nihui
6cdd7110be
fix instruction extension dispatch ( #5427 )
2 years ago
nihui
9ce7930413
x86 optimization for convolution tiled gemm ( #5426 )
2 years ago
nihui
f3041b7c0d
Update README.md
2 years ago
nihui
157d97c22a
unified windows ci ( #5424 )
2 years ago
nihui
0fd25d6c70
fix arm riscv build with NCNN_BF16=OFF ( #5422 )
2 years ago
nihui
110d2ba342
Update README.md
2 years ago
nihuini
802a625ae2
fix pnnx windows build
2 years ago
nihui
56775de509
update readme download matrix, release 20240410 ( #5418 )
2 years ago
nihui
1e4daff075
fix ncnn tools crash with compound layer creation, write gemm weight data ( #5416 )
* fix ncnn tools crash with compound layer creation
* write gemm weight data
2 years ago