GithubStar/ncnn
1
0
Fork 0
mirror of https://github.com/Tencent/ncnn.git synced 2026-10-22 00:43:53 +08:00
Code Issues Packages Projects Releases Wiki Activity
3,977 Commits 4 Branches 50 Tags
master
Commit Graph

4 Commits

Author SHA1 Message Date
nihui 10cee2a7bc Fix x86 bf16 GEMM packing on i386 (#6708) 
Correct the i386 AVX512BF16 B-tile packing order so dpbf16 consumes k/k+1 pairs instead of adjacent columns. This fixes intermittent x86 32-bit bf16s GEMM failures on AVX512BF16 hosts without disabling AVX512 paths.

Also make the MultiHeadAttention int8 test inputs follow the same dynamic-quantization-safe randomization used by GEMM int8 tests, avoiding values near .5 rounding boundaries while keeping the tensors random and reproducible.
 2026-05-09 13:22:23 +08:00
nihui 171b9d1bba use spdx license header, copyright Tencent (#6152) 2025-07-08 11:01:29 +08:00
nihui e7602a206b fix gemm arm int8 scales descales offset (#5750) 2024-10-21 19:09:43 +08:00
nihui 66b54cbea2 multiheadattention int8 quantization (#5733) 
* x86 vulkan fallback
* comment about bf16s
 2024-10-15 16:38:00 +08:00