CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-30 08:42:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml : fix quantize_row_q4_1() ARM_NEON (close #876)

Browse Source
This commit is contained in:
Georgi Gerganov
2023-04-10 19:29:48 +03:00
parent 180b693a47
commit 684da25926
1 changed files with 4 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
ggml.c
View File

@@ -599,10 +599,7 @@ static void quantize_row_q4_0(const float * restrict x, void * restrict vy, int
for (int l = 0; l < 2; l++) amaxv[4*l] = vmaxq_f32(amaxv[4*l], amaxv[4*l+2]); for (int l = 0; l < 2; l++) amaxv[4*l] = vmaxq_f32(amaxv[4*l], amaxv[4*l+2]);
for (int l = 0; l < 1; l++) amaxv[8*l] = vmaxq_f32(amaxv[8*l], amaxv[8*l+4]); for (int l = 0; l < 1; l++) amaxv[8*l] = vmaxq_f32(amaxv[8*l], amaxv[8*l+4]);
// absolute max const float amax = vmaxvq_f32(amaxv[0]);
const float amax = MAX(
MAX(vgetq_lane_f32(amaxv[0], 0), vgetq_lane_f32(amaxv[0], 1)),
MAX(vgetq_lane_f32(amaxv[0], 2), vgetq_lane_f32(amaxv[0], 3)));
const float d = amax / ((1 << 3) - 1); const float d = amax / ((1 << 3) - 1);
const float id = d ? 1.0f/d : 0.0f; const float id = d ? 1.0f/d : 0.0f;
@@ -924,7 +921,7 @@ static void quantize_row_q4_1(const float * restrict x, void * restrict vy, int
float32x4_t minv[8]; float32x4_t minv[8];
float32x4_t maxv[8]; float32x4_t maxv[8];
for (int l = 0; l < 8; l++) srcv[l] = vld1q_f32(x + i*32 + 4*l); for (int l = 0; l < 8; l++) srcv[l] = vld1q_f32(x + i*QK + 4*l);
for (int l = 0; l < 4; l++) minv[2*l] = vminq_f32(srcv[2*l], srcv[2*l + 1]); for (int l = 0; l < 4; l++) minv[2*l] = vminq_f32(srcv[2*l], srcv[2*l + 1]);
for (int l = 0; l < 2; l++) minv[4*l] = vminq_f32(minv[4*l], minv[4*l + 2]); for (int l = 0; l < 2; l++) minv[4*l] = vminq_f32(minv[4*l], minv[4*l + 2]);
@@ -947,7 +944,8 @@ static void quantize_row_q4_1(const float * restrict x, void * restrict vy, int
for (int l = 0; l < 8; l++) { for (int l = 0; l < 8; l++) {
const float32x4_t v = vmulq_n_f32(vsubq_f32(srcv[l], minv0), id); const float32x4_t v = vmulq_n_f32(vsubq_f32(srcv[l], minv0), id);
const int32x4_t vi = vcvtq_s32_f32(v); const float32x4_t vf = vaddq_f32(v, vdupq_n_f32(0.5f)); // needed to round to nearest
const int32x4_t vi = vcvtq_s32_f32(vf);
y[i].qs[2*l + 0] = vgetq_lane_s32(vi, 0) | (vgetq_lane_s32(vi, 1) << 4); y[i].qs[2*l + 0] = vgetq_lane_s32(vi, 0) | (vgetq_lane_s32(vi, 1) << 4);
y[i].qs[2*l + 1] = vgetq_lane_s32(vi, 2) | (vgetq_lane_s32(vi, 3) << 4); y[i].qs[2*l + 1] = vgetq_lane_s32(vi, 2) | (vgetq_lane_s32(vi, 3) << 4);