amd/blis
1
0
Fork 0
mirror of https://github.com/amd/blis.git synced 2026-05-04 22:41:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

aocl_softmax_f32 api's for softmax computation as part of lpgemm.

Browse Source 
-Softmax is often used as the last activation function in a neural
network - softmax(xi) = exp(xi)/(exp(x0) + exp(x1) + ... + exp(xn))).
This step happens after the final low precision gemm computation,
and it helps to have the softmax functionality that can be invoked
as part of the lpgemm workflow. In order to support this, a new api,
aocl_softmax_f32 is introduced as part of aocl_gemm. This api
computes element-wise softmax of a matrix/vector of floats. This api
invokes ISA specific vectorized micro-kernels (vectorized only when
incx=1), and a cntx based mechanism (similar to lpgemm_cntx) is used
to dispatch to the appropriate kernel.

AMD-Internal: [CPUPL-3247]
Change-Id: If15880360947435985fa87b6436e475571e4684a
This commit is contained in:
mkadavil
2023-04-20 20:28:28 +05:30
committed by MithunMohan KadavilMadanaMohanan
parent 4aace5f524
commit 3572baa9d3
9 changed files with 488 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
addon/aocl_gemm/config/lpgemm_func_map.h
View File

@@ -67,6 +67,7 @@
#define LPGEMM_UTIL_KERN_FUNC_MAP_AVX512_VNNI_BF16 \
UMACRO(F32_GELU_TANH, lpgemm_util_f32_gelu_tanh_avx512_kernel) \
UMACRO(F32_GELU_ERF, lpgemm_util_f32_gelu_erf_avx512_kernel) \
UMACRO(F32_SOFTMAX, lpgemm_util_f32_softmax_avx512_kernel) \
// Icelake
#define LPGEMM_KERN_FUNC_MAP_AVX512_VNNI \
@@ -91,6 +92,7 @@
#define LPGEMM_UTIL_KERN_FUNC_MAP_AVX512_VNNI \
UMACRO(F32_GELU_TANH, lpgemm_util_f32_gelu_tanh_avx512_kernel) \
UMACRO(F32_GELU_ERF, lpgemm_util_f32_gelu_erf_avx512_kernel) \
UMACRO(F32_SOFTMAX, lpgemm_util_f32_softmax_avx512_kernel) \
// Skylake
#define LPGEMM_KERN_FUNC_MAP_AVX512 \
@@ -115,6 +117,7 @@
#define LPGEMM_UTIL_KERN_FUNC_MAP_AVX512 \
UMACRO(F32_GELU_TANH, lpgemm_util_f32_gelu_tanh_avx512_kernel) \
UMACRO(F32_GELU_ERF, lpgemm_util_f32_gelu_erf_avx512_kernel) \
UMACRO(F32_SOFTMAX, lpgemm_util_f32_softmax_avx512_kernel) \
// Milan, Haswell
#define LPGEMM_KERN_FUNC_MAP_AVX2 \
@@ -139,5 +142,6 @@
#define LPGEMM_UTIL_KERN_FUNC_MAP_AVX2 \
UMACRO(F32_GELU_TANH, lpgemm_util_f32_gelu_tanh_avx2_kernel) \
UMACRO(F32_GELU_ERF, lpgemm_util_f32_gelu_erf_avx2_kernel) \
UMACRO(F32_SOFTMAX, lpgemm_util_f32_softmax_avx2_kernel) \
#endif //LPGEMM_FUNC_MAP_H