vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Re-use existing bytesFromNibbles function

Browse source
This commit is contained in:
Casey Primozic 2023-03-20 15:55:42 -07:00
parent a6598801ad
commit 76af3f5a64
No known key found for this signature in database
GPG key ID: 2A02222DA3425B99
1 changed files with 2 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

17
ggml.c
View file

@ -361,7 +361,7 @@ static const size_t CACHE_LINE_SIZE_F32 = CACHE_LINE_SIZE/sizeof(float);
// AVX routines provided by GH user Const-me
// ref: https://github.com/ggerganov/ggml/pull/27#issuecomment-1464934600
#if __AVX2__
#if __AVX2__ || __AVX512F__
// Unpack 32 4-bit fields into 32 bytes
// The output vector contains 32 bytes, each one in [ 0 .. 15 ] interval
static inline __m256i bytesFromNibbles( const uint8_t* rsi )
@ -1418,21 +1418,6 @@ inline static void ggml_vec_dot_q4_0(const int n, float * restrict s, const void
#error "not implemented for QK"
#endif
#elif defined(__AVX512F__)
inline __m256i bytesFromNibbles( const uint8_t* rsi ){
// Load 16 bytes from memory
__m128i tmp = _mm_loadu_si128( ( const __m128i* )rsi );
// Expand bytes into uint16_t values
__m256i bytes = _mm256_cvtepu8_epi16( tmp );
// Unpack values into individual bytes
const __m256i lowMask = _mm256_set1_epi8( 0xF );
__m256i high = _mm256_andnot_si256( lowMask, bytes );
__m256i low = _mm256_and_si256( lowMask, bytes );
high = _mm256_slli_epi16( high, 4 );
bytes = _mm256_or_si256( low, high );
return bytes;
}
inline __m512 process_one_block(
__m512 acc,