ef7f7369 · AVX2 vpmadd GEMM move constants to functions to params · Aug 05, 2024

23aeef97 · Remove unused templates. · Aug 03, 2024

9e5f6793 · Remove target specific params from `xnn_f32_rnd_params` · Aug 03, 2024

9a03125b · Internal build change · Aug 02, 2024

00b54b62 · Remove target specific versions of rsum, rdsum params · Aug 02, 2024

43b3bf80 · Copybara import of the project: · Aug 02, 2024

c97e447c · Remove f16_f32_vcvt_params struct · Aug 01, 2024

80010005 · Remove f16_vneg_params, f16_vabs_params, bf16_vabs_params · Jul 31, 2024

05848f1c · Fixed unused variable error · Jul 31, 2024

b81f866a · AVX2 and AVX10 vpmaddubsw QD8_F16_QC4W GEMM move xor 0x88 from GEMM to pack functions · Jul 29, 2024

a89ab9a5 · AVX512 vpmaddubsw GEMMS move xor 0x88 from GEMM to pack functions · Jul 26, 2024

dab2a2f9 · xnn_define_multiply2 does not accept int32 tensors · Jul 24, 2024

67e91534 · Enable AVX2 QD8_F32_QC4W GEMM using 1 vpmaddubsw · Jul 23, 2024

409f79ec · s32 multiplication is expected to overflow · Jul 23, 2024

0a60e098 · vpmaddubsw version of avx512 mimic to vnni · Jul 16, 2024

50fc0581 · Remove Scalar FMA library · Jul 16, 2024

64756237 · Add first s16 intrinsics to wrapper · Jul 16, 2024

1b6e58ff · Update the KleidiAI version for the `cmake` and `bazel` builds, enable... · Jul 16, 2024

8974823c · Internal change only. · Jul 10, 2024

f6642f36 · Rename `xnn_shift(l|r)_f32` to `xnn_shift_(left|right)_f32`, add `xnn_shift_right_signed_f32`. · Jul 09, 2024