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

Merge cfb1b2277f into d774ab3acc

Browse source
This commit is contained in:
Herman Semenoff 2025-02-05 23:08:49 +01:00 • committed by GitHub
commit 52d2d92c7e
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/ggml-quants.c
View file

@ -4193,7 +4193,7 @@ static void quantize_row_iq1_s_impl(const float * restrict x, void * restrict vy
float best_score = -FLT_MIN, scale = max;
int besti1 = -1, besti2 = -1, best_shift = 0;
for (int i1 = 0; i1 <= block_size; ++i1) {
for (int i2 = i1; i2 <= block_size; ++i2) {
for (int i2 = i1 + 1; i2 <= block_size; ++i2) {
float sumqx = (sumx[i1] - sumx[0])*x_p[0] + (sumx[i2] - sumx[i1])*x_p[1] + (sumx[block_size] - sumx[i2])*x_p[2];
float sumq2 = (sumw[i1] - sumw[0])*x_p[0]*x_p[0] + (sumw[i2] - sumw[i1])*x_p[1]*x_p[1] + (sumw[block_size] - sumw[i2])*x_p[2]*x_p[2];
if (sumq2 > 0 && sumqx*sumqx > best_score*sumq2) {