issues
search
google
/
XNNPACK
High-efficiency floating-point neural network inference operators for mobile, server, and Web
Other
1.89k
stars
376
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Prevent dangling `else` for certain flag combinations.
#7539
copybara-service[bot]
opened
6 hours ago
0
Mark test as unsigned inputs
#7538
copybara-service[bot]
closed
7 hours ago
0
Fix slow installation of `gcc-13` by using `awalsh128/cache-apt-pkgs-action` to cache the packages.
#7537
copybara-service[bot]
opened
11 hours ago
0
Pass correct parameters to create function to fix flaky test
#7536
copybara-service[bot]
closed
9 hours ago
0
Replace the `C++17` `static_assert(bool)` with the `C++11` `static_assert(bool, const char *)`.
#7535
copybara-service[bot]
closed
11 hours ago
0
[PyTorch] XNNPACK `static_assert` not supported on MSVC 2019
#7534
xuhancn
opened
1 day ago
3
Initialize XNNPACK. Why did this not fail before?
#7533
copybara-service[bot]
closed
1 day ago
0
Fix type in setup function
#7532
copybara-service[bot]
closed
1 day ago
0
Fix static_assert not be supported on VS2019.
#7531
xuhancn
closed
1 day ago
0
Move avx vnni kernels into the correct position and uncomment optimized weights packing
#7530
copybara-service[bot]
closed
11 hours ago
0
AMX GEMM/IGEMM microkernels initialize tiles to zero for msan builds
#7529
copybara-service[bot]
opened
1 day ago
0
AMX microkernels msan builds fill tiles with zero
#7528
copybara-service[bot]
opened
1 day ago
0
Copybara import of the project:
#7527
copybara-service[bot]
opened
1 day ago
0
Copybara import of the project:
#7526
copybara-service[bot]
opened
1 day ago
0
Copybara import of the project:
#7525
copybara-service[bot]
closed
1 day ago
0
Remove excessive unrolling from some microkernels
#7524
copybara-service[bot]
closed
1 day ago
0
Fix the handling of edge cases in `xnn_gemm_best_nc`.
#7523
copybara-service[bot]
closed
11 hours ago
0
Copybara import of the project:
#7522
copybara-service[bot]
closed
1 day ago
0
Adding Tiled Scalar Packing Kernels for QB4W
#7521
mcr229
opened
2 days ago
1
Integration of Kleidi F16 SME kernels
#7520
copybara-service[bot]
opened
2 days ago
0
Use `benchmark::utils::WipeCache()` instead of relying on a set of ops with distinct buffers.
#7519
copybara-service[bot]
closed
2 days ago
0
Replace runtime `assert` with a proper unit test.
#7518
copybara-service[bot]
closed
2 days ago
0
MSAN compliant pad bias with 0 in scalar packw microkernels
#7517
copybara-service[bot]
opened
2 days ago
0
Disable QD8 AMX GEMM/IGEMM microkernels
#7516
copybara-service[bot]
opened
2 days ago
0
packw pad bias with 0 for msan
#7515
copybara-service[bot]
closed
2 days ago
0
Replace `assert` with `xnn_log_warning`.
#7514
copybara-service[bot]
closed
3 days ago
0
Flip the order in which the GEMM tiles are computed, i.e. instead of traversing the tiles in `n` direction first, traverse them in `m` direction, and then in `n` direction.
#7513
copybara-service[bot]
opened
3 days ago
0
Don't pack left-hand operand since this is currently only supported for `qc4w` weights, and this benchmark uses `qc8w` weights.
#7512
copybara-service[bot]
closed
3 days ago
0
Initialize test inputs
#7511
copybara-service[bot]
closed
3 days ago
0
Copybara import of the project:
#7510
copybara-service[bot]
closed
3 days ago
0
Copybara import of the project:
#7509
copybara-service[bot]
closed
3 days ago
0
Fix suggestions for Hexagon to accommodate recent changes
#7508
ejparkqc
closed
3 days ago
2
Initialize input, filter and bias for convoluion 2d tests
#7507
copybara-service[bot]
opened
4 days ago
0
Enable 7x16 F32-GEMM for avx512
#7506
copybara-service[bot]
opened
4 days ago
0
Add f16->qu8 vcvt microkernels
#7505
copybara-service[bot]
closed
3 days ago
0
Changes to `batch_matrix_multiply_bench`:
#7504
copybara-service[bot]
closed
4 days ago
0
QS8-PACKW AVX2 using vpmaddubsw
#7503
copybara-service[bot]
closed
3 days ago
0
Fix `bazel-linux-aarch64-gcc13` workflow and resolve accompanying build errors.
#7502
copybara-service[bot]
closed
4 days ago
0
Copybara import of the project:
#7501
copybara-service[bot]
opened
1 week ago
0
Convolution bias does not need to be explicitly converted to `f16`.
#7500
copybara-service[bot]
closed
1 week ago
0
Fix `bazel-linux-aarch64-gcc13` workflow.
#7499
copybara-service[bot]
closed
9 hours ago
0
Handle `f16` GEMM weights and biases when converting to `f16`.
#7498
copybara-service[bot]
closed
1 week ago
0
Optimize QS8 GIO packing using AVXVNNI instruction
#7497
xujuntwt95329
opened
1 week ago
0
Copybara import of the project:
#7496
copybara-service[bot]
closed
1 week ago
0
Copybara import of the project:
#7495
copybara-service[bot]
opened
1 week ago
1
Improve confusing and less accurate computation of quantization parameters
#7494
copybara-service[bot]
closed
1 week ago
0
F32-IGEMM AVX512 generate up to 16x64
#7493
copybara-service[bot]
opened
1 week ago
0
Remove unused header
#7492
copybara-service[bot]
closed
1 week ago
0
Speculative fix for #7489
#7491
copybara-service[bot]
closed
1 week ago
1
Add a builder for aarch64 under gcc-13 in addition to clang-18 (helps address https://github.com/google/XNNPACK/issues/7489)
#7490
copybara-service[bot]
closed
1 week ago
0
Next