happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge 2383eb224d into 18ddaea2ae

This commit is contained in:
Srihari-mcw 2026-01-02 23:47:03 +02:00 committed by GitHub
parent 18ddaea2ae 2383eb224d
commit 1f4e917237
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 4264 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
ggml/src/ggml-cpu/arch-fallback.h
View File

@ -41,6 +41,7 @@
#define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q8_0_4x4_q8_0_generic ggml_gemv_q8_0_4x4_q8_0
@ -51,6 +52,7 @@
#define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
#define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0
@ -61,8 +63,10 @@
#define ggml_quantize_mat_q8_K_4x8_generic ggml_quantize_mat_q8_K_4x8
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#elif defined(__x86_64__) || defined(__i386__) || defined(_M_IX86) || defined(_M_X64)
// repack.cpp
#define ggml_quantize_mat_q8_0_4x4_generic ggml_quantize_mat_q8_0_4x4
@ -97,6 +101,7 @@
#define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q8_0_4x4_q8_0_generic ggml_gemv_q8_0_4x4_q8_0
@ -107,6 +112,7 @@
#define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
#define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0
@ -129,6 +135,7 @@
#define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q8_0_4x4_q8_0_generic ggml_gemv_q8_0_4x4_q8_0
@ -139,6 +146,7 @@
#define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
#define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0
@ -168,6 +176,7 @@
#define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q8_0_4x4_q8_0_generic ggml_gemv_q8_0_4x4_q8_0
@ -177,6 +186,7 @@
#define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
#define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0
@ -205,6 +215,7 @@
#define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q8_0_4x4_q8_0_generic ggml_gemv_q8_0_4x4_q8_0
@ -215,6 +226,7 @@
#define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
#define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0
@ -245,6 +257,7 @@
#define ggml_gemv_q4_K_8x4_q8_K_generic ggml_gemv_q4_K_8x4_q8_K
#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_K_8x8_q8_K
#define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K
#define ggml_gemv_q6_K_8x8_q8_K_generic ggml_gemv_q6_K_8x8_q8_K
#define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0
#define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0
#define ggml_gemv_q8_0_4x4_q8_0_generic ggml_gemv_q8_0_4x4_q8_0
@ -255,6 +268,7 @@
#define ggml_gemm_q4_K_8x4_q8_K_generic ggml_gemm_q4_K_8x4_q8_K
#define ggml_gemm_q4_K_8x8_q8_K_generic ggml_gemm_q4_K_8x8_q8_K
#define ggml_gemm_q2_K_8x8_q8_K_generic ggml_gemm_q2_K_8x8_q8_K
#define ggml_gemm_q6_K_8x8_q8_K_generic ggml_gemm_q6_K_8x8_q8_K
#define ggml_gemm_iq4_nl_4x4_q8_0_generic ggml_gemm_iq4_nl_4x4_q8_0
#define ggml_gemm_iq4_nl_8x8_q8_0_generic ggml_gemm_iq4_nl_8x8_q8_0
#define ggml_gemm_q8_0_4x4_q8_0_generic ggml_gemm_q8_0_4x4_q8_0

3987
ggml/src/ggml-cpu/arch/x86/repack.cpp
View File

File diff suppressed because it is too large Load Diff

251
ggml/src/ggml-cpu/repack.cpp
View File

@ -616,6 +616,78 @@ void ggml_gemv_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs,
}
}
void ggml_gemv_q6_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
const int qk = QK_K;
const int nb = n / qk;
const int ncols_interleaved = 8;
const int blocklen = 8;
assert (n % qk == 0);
assert (nc % ncols_interleaved == 0);
UNUSED(s);
UNUSED(bs);
UNUSED(vx);
UNUSED(vy);
UNUSED(nr);
UNUSED(nc);
UNUSED(nb);
UNUSED(ncols_interleaved);
UNUSED(blocklen);
float sumf[8];
int sumi1,sumi2,sumi3,sumi4;
int sumi;
const block_q8_K * a_ptr = (const block_q8_K *)vy;
for(int x = 0; x < nc / ncols_interleaved; x++) {
const block_q6_Kx8 * b_ptr = (const block_q6_Kx8 *) vx + (x * nb);
for (int j = 0; j < ncols_interleaved; j++) {
sumf[j] = 0.0;
}
for (int l = 0; l < nb; l++) {
for (int k = 0; k < (qk / (4 * blocklen)); k++) {
const int8_t * scales_0 = b_ptr[l].scales + (k / 4) * 64;
const int8_t * scales_1 = b_ptr[l].scales + (k / 4) * 64 + 16;
const int8_t * scales_2 = b_ptr[l].scales + (k / 4) * 64 + 32;
const int8_t * scales_3 = b_ptr[l].scales + (k / 4) * 64 + 48;
for (int j = 0; j < ncols_interleaved; j++) {
sumi1 = 0;
sumi2 = 0;
sumi3 = 0;
sumi4 = 0;
sumi = 0;
int offset = ((k / 2) % 2) + j * 2;
for (int i = 0; i < blocklen; ++i) {
const int hbits_index = k * ncols_interleaved * blocklen + j * blocklen + i;
const int lbits_index = hbits_index + (k/4) * 256;
int8_t v0 = (int8_t)((((b_ptr[l].qh[hbits_index] & 3) << 4) | (b_ptr[l].ql[lbits_index] & 0xF)) - 32);
int8_t v1 = (int8_t)(((((b_ptr[l].qh[hbits_index] >> 2 ) & 3) << 4) | (b_ptr[l].ql[lbits_index + 256] & 0xF)) - 32);
int8_t v2 = (int8_t)(((((b_ptr[l].qh[hbits_index] >> 4 ) & 3) << 4) | ((b_ptr[l].ql[lbits_index] >> 4) & 0xF)) - 32);
int8_t v3 = (int8_t)(((((b_ptr[l].qh[hbits_index] >> 6 ) & 3) << 4) | ((b_ptr[l].ql[lbits_index + 256] >> 4) & 0xF)) - 32);
sumi1 = (v0 * a_ptr[l].qs[(k >> 2) * 128 + (k % 4) * blocklen + i]);
sumi2 = (v1 * a_ptr[l].qs[(k >> 2) * 128 + (k % 4) * blocklen + i + 32]);
sumi3 = (v2 * a_ptr[l].qs[(k >> 2) * 128 + (k % 4) * blocklen + i + 64]);
sumi4 = (v3 * a_ptr[l].qs[(k >> 2) * 128 + (k % 4) * blocklen + i + 96]);
sumi1 = sumi1 * scales_0[offset];
sumi2 = sumi2 * scales_1[offset];
sumi3 = sumi3 * scales_2[offset];
sumi4 = sumi4 * scales_3[offset];
sumi += sumi1 + sumi2 + sumi3 + sumi4;
}
sumf[j] += sumi * GGML_FP16_TO_FP32(b_ptr[l].d[j]) * a_ptr[l].d;
}
}
}
for (int j = 0; j < ncols_interleaved; j++) {
s[x * ncols_interleaved + j] = sumf[j];
}
}
}
void ggml_gemv_iq4_nl_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
const int qk = QK8_0;
const int nb = n / qk;
@ -1212,6 +1284,87 @@ void ggml_gemm_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs,
}
}
void ggml_gemm_q6_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
const int qk = QK_K;
const int nb = n / qk;
const int ncols_interleaved = 8;
const int blocklen = 8;
assert (n % qk == 0);
assert (nr % 4 == 0);
assert (nc % ncols_interleaved == 0);
UNUSED(s);
UNUSED(bs);
UNUSED(vx);
UNUSED(vy);
UNUSED(nr);
UNUSED(nc);
UNUSED(nb);
UNUSED(ncols_interleaved);
UNUSED(blocklen);
float sumf[4][8];
int sumi1, sumi2, sumi3, sumi4;
int sumi;
for (int y = 0; y < nr / 4; y++) {
const block_q8_Kx4 * a_ptr = (const block_q8_Kx4 *) vy + (y * nb);
for (int x = 0; x < nc / ncols_interleaved; x++) {
const block_q6_Kx8 * b_ptr = (const block_q6_Kx8 *) vx + (x * nb);
for (int m = 0; m < 4; m++) {
for (int j = 0; j < ncols_interleaved; j++) {
sumf[m][j] = 0.0;
}
}
for (int l = 0; l < nb; l++) {
for (int k = 0; k < (qk / (4 * blocklen)); k++) {
const int8_t * scales_0 = b_ptr[l].scales + (k / 4) * 64;
const int8_t * scales_1 = b_ptr[l].scales + (k / 4) * 64 + 16;
const int8_t * scales_2 = b_ptr[l].scales + (k / 4) * 64 + 32;
const int8_t * scales_3 = b_ptr[l].scales + (k / 4) * 64 + 48;
for (int m = 0; m < 4; m++) {
for (int j = 0; j < ncols_interleaved; j++) {
sumi1 = 0;
sumi2 = 0;
sumi3 = 0;
sumi4 = 0;
sumi = 0;
int offset = ((k / 2) % 2) + j * 2;
for (int i = 0; i < blocklen; ++i){
const int hbits_index = k * ncols_interleaved * blocklen + j * blocklen + i;
const int lbits_index = hbits_index + (k/4) * 256;
int8_t v0 = (int8_t)((((b_ptr[l].qh[hbits_index] & 3) << 4) | (b_ptr[l].ql[lbits_index] & 0xF)) - 32);
int8_t v1 = (int8_t)(((((b_ptr[l].qh[hbits_index] >> 2 ) & 3) << 4) | (b_ptr[l].ql[lbits_index + 256] & 0xF)) - 32);
int8_t v2 = (int8_t)(((((b_ptr[l].qh[hbits_index] >> 4 ) & 3) << 4) | ((b_ptr[l].ql[lbits_index] >> 4) & 0xF)) - 32);
int8_t v3 = (int8_t)(((((b_ptr[l].qh[hbits_index] >> 6 ) & 3) << 4) | ((b_ptr[l].ql[lbits_index + 256] >> 4) & 0xF)) - 32);
sumi1 = (v0 * a_ptr[l].qs[(k >> 2) * 512 + (k % 4) * 4 * blocklen + m * blocklen + i]);
sumi2 = (v1 * a_ptr[l].qs[(k >> 2) * 512 + (k % 4) * 4 * blocklen + m * blocklen + i + 128]);
sumi3 = (v2 * a_ptr[l].qs[(k >> 2) * 512 + (k % 4) * 4 * blocklen + m * blocklen + i + 256]);
sumi4 = (v3 * a_ptr[l].qs[(k >> 2) * 512 + (k % 4) * 4 * blocklen + m * blocklen + i + 384]);
sumi1 = sumi1 * (scales_0[offset]);
sumi2 = sumi2 * (scales_1[offset]);
sumi3 = sumi3 * (scales_2[offset]);
sumi4 = sumi4 * (scales_3[offset]);
sumi += sumi1 + sumi2 + sumi3 + sumi4;
}
sumf[m][j] += sumi * GGML_FP16_TO_FP32(b_ptr[l].d[j]) * a_ptr[l].d[m];
}
}
}
}
for (int m = 0; m < 4; m++) {
for (int j = 0; j < ncols_interleaved; j++) {
s[(y * 4 + m) * bs + x * ncols_interleaved + j] = sumf[m][j];
}
}
}
}
}
void ggml_gemm_iq4_nl_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc) {
const int qk = QK8_0;
@ -1625,6 +1778,51 @@ static block_q2_Kx8 make_block_q2_Kx8(block_q2_K * in, unsigned int blck_size_in
}
static block_q6_Kx8 make_block_q6_Kx8(block_q6_K * in, unsigned int blck_size_interleave) {
block_q6_Kx8 out;
// Delta(scale) of the eight Q6_K structures are copied onto the output interleaved structure
for (int i = 0; i < 8; i++) {
out.d[i] = in[i].d;
}
const int end = QK_K * 4 / blck_size_interleave;
// Interleave Q6_K quants by taking 8 bytes at a time
for (int i = 0; i < end; ++i) {
int src_id = i % 8;
int src_offset = (i / 8) * blck_size_interleave;
int dst_offset = i * blck_size_interleave;
uint64_t ql_elems;
memcpy(&ql_elems, &in[src_id].ql[src_offset], sizeof(uint64_t));
memcpy(&out.ql[dst_offset], &ql_elems, sizeof(uint64_t));
}
for (int i = 0; i < 64; ++i) {
int qh_src_id = i % 8;
int qh_src_offset = (i / 8) * blck_size_interleave;
int qh_dst_offset = i * blck_size_interleave;
uint64_t qh_elems;
memcpy(&qh_elems, &in[qh_src_id].qh[qh_src_offset], sizeof(uint64_t));
memcpy(&out.qh[qh_dst_offset], &qh_elems, sizeof(uint64_t));
}
for (int i = 0; i < 128; i++) {
// Index for selecting which q6k super block
int src1 = (i % 16) / 2;
// Index for selecting scale
int src2 = ((i / 16) * 2) + (i % 2);
out.scales[i] = in[src1].scales[src2];
}
return out;
}
static int repack_q4_0_to_q4_0_4_bl(struct ggml_tensor * t, int interleave_block, const void * GGML_RESTRICT data, size_t data_size) {
GGML_ASSERT(t->type == GGML_TYPE_Q4_0);
GGML_ASSERT(interleave_block == 4 || interleave_block == 8);
@ -1718,6 +1916,38 @@ static int repack_q2_K_to_q2_K_8_bl(struct ggml_tensor * t, int interleave_block
GGML_UNUSED(data_size);
}
static int repack_q6_K_to_q6_K_8_bl(struct ggml_tensor * t, int interleave_block, const void * GGML_RESTRICT data, size_t data_size) {
GGML_ASSERT(t->type == GGML_TYPE_Q6_K);
GGML_ASSERT(interleave_block == 8);
constexpr int nrows_interleaved = 8;
block_q6_Kx8 * dst = (block_q6_Kx8 *)t->data;
const block_q6_K * src = (const block_q6_K *)data;
block_q6_K dst_tmp[8];
int nrow = ggml_nrows(t);
int nblocks = t->ne[0] / QK_K;
GGML_ASSERT(data_size == nrow * nblocks * sizeof(block_q6_K));
if (t->ne[1] % nrows_interleaved != 0 || t->ne[0] % 8 != 0) {
return -1;
}
for (int b = 0; b < nrow; b += nrows_interleaved) {
for (int64_t x = 0; x < nblocks; x++) {
for (int i = 0; i < nrows_interleaved; i++) {
dst_tmp[i] = src[x + i * nblocks];
}
*dst++ = make_block_q6_Kx8(dst_tmp, interleave_block);
}
src += nrows_interleaved * nblocks;
}
return 0;
GGML_UNUSED(data_size);
}
static int repack_q4_0_to_q4_0_8_bl(struct ggml_tensor * t, int interleave_block, const void * GGML_RESTRICT data, size_t data_size) {
GGML_ASSERT(t->type == GGML_TYPE_Q4_0);
GGML_ASSERT(interleave_block == 8);
@ -1936,6 +2166,10 @@ template <> int repack<block_q2_K, 8, 8>(struct ggml_tensor * t, const void * da
return repack_q2_K_to_q2_K_8_bl(t, 8, data, data_size);
}
template <> int repack<block_q6_K, 8, 8>(struct ggml_tensor * t, const void * data, size_t data_size) {
return repack_q6_K_to_q6_K_8_bl(t, 8, data, data_size);
}
template <> int repack<block_iq4_nl, 4, 4>(struct ggml_tensor * t, const void * data, size_t data_size) {
return repack_iq4_nl_to_iq4_nl_4_bl(t, 4, data, data_size);
}
@ -1985,6 +2219,10 @@ template <> void gemv<block_q2_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t
ggml_gemv_q2_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
}
template <> void gemv<block_q6_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
ggml_gemv_q6_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
}
template <> void gemv<block_iq4_nl, 4, 4, GGML_TYPE_Q8_0>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
ggml_gemv_iq4_nl_4x4_q8_0(n, s, bs, vx, vy, nr, nc);
}
@ -2029,6 +2267,10 @@ template <> void gemm<block_q2_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t
ggml_gemm_q2_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
}
template <> void gemm<block_q6_K, 8, 8, GGML_TYPE_Q8_K>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
ggml_gemm_q6_K_8x8_q8_K(n, s, bs, vx, vy, nr, nc);
}
template <> void gemm<block_iq4_nl, 4, 4, GGML_TYPE_Q8_0>(int n, float * s, size_t bs, const void * vx, const void * vy, int nr, int nc) {
ggml_gemm_iq4_nl_4x4_q8_0(n, s, bs, vx, vy, nr, nc);
}
@ -2435,6 +2677,9 @@ static const ggml::cpu::tensor_traits * ggml_repack_get_optimal_repack_type(cons
// instance for Q2
static const ggml::cpu::repack::tensor_traits<block_q2_K, 8, 8, GGML_TYPE_Q8_K> q2_K_8x8_q8_K;
// instance for Q6
static const ggml::cpu::repack::tensor_traits<block_q6_K, 8, 8, GGML_TYPE_Q8_K> q6_K_8x8_q8_K;
// instance for IQ4
static const ggml::cpu::repack::tensor_traits<block_iq4_nl, 4, 4, GGML_TYPE_Q8_0> iq4_nl_4x4_q8_0;
static const ggml::cpu::repack::tensor_traits<block_iq4_nl, 8, 8, GGML_TYPE_Q8_0> iq4_nl_8x8_q8_0;
@ -2482,6 +2727,12 @@ static const ggml::cpu::tensor_traits * ggml_repack_get_optimal_repack_type(cons
return &q2_K_8x8_q8_K;
}
}
} else if (cur->type == GGML_TYPE_Q6_K) {
if (ggml_cpu_has_avx512()) {
if (cur->ne[1] % 8 == 0) {
return &q6_K_8x8_q8_K;
}
}
} else if (cur->type == GGML_TYPE_IQ4_NL) {
if (ggml_cpu_has_avx2()) {
if (cur->ne[1] % 8 == 0) {

12
ggml/src/ggml-cpu/repack.h
View File

@ -52,6 +52,14 @@ struct block_q2_Kx8 {
};
static_assert(sizeof(block_q2_Kx8) == sizeof(ggml_half) * 16 + QK_K/2 + QK_K * 2, "wrong q2_K block size/padding");
struct block_q6_Kx8 {
uint8_t ql[1024]; // quants, lower 4 bits
uint8_t qh[512]; // quants, upper 2 bits
int8_t scales[128]; // scales, quantized with 8 bits
ggml_half d[8];
};
static_assert(sizeof(block_q6_Kx8) == sizeof(ggml_half)*8 + (QK_K)/2 + 24*QK_K/4, "wrong block_q6_Kx8 size/padding");
struct block_q8_Kx4 {
float d[4]; // delta
int8_t qs[QK_K * 4]; // quants
@ -88,6 +96,7 @@ void ggml_gemv_q4_0_8x8_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const vo
void ggml_gemv_q4_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q4_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q2_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q6_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_iq4_nl_4x4_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_iq4_nl_8x8_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q4_0_4x4_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
@ -96,6 +105,7 @@ void ggml_gemm_q4_0_8x8_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const vo
void ggml_gemm_q4_K_8x4_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q4_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q2_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q6_K_8x8_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_iq4_nl_4x4_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_iq4_nl_8x8_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q8_0_4x4_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
@ -114,6 +124,7 @@ void ggml_gemv_q4_0_8x8_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs,
void ggml_gemv_q4_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q4_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q6_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_iq4_nl_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_iq4_nl_8x8_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q4_0_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
@ -122,6 +133,7 @@ void ggml_gemm_q4_0_8x8_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs,
void ggml_gemm_q4_K_8x4_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q4_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q2_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_q6_K_8x8_q8_K_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_iq4_nl_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemm_iq4_nl_8x8_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);
void ggml_gemv_q8_0_4x4_q8_0_generic(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, const void * GGML_RESTRICT vy, int nr, int nc);