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

metal : fix mat-vec Q4_K kernel for QK_K == 64

Browse source
This commit is contained in:
Georgi Gerganov 2023-12-31 13:52:34 +02:00
parent a8b9bb4566
commit 5865b18eeb
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
1 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

4
ggml-metal.metal
View file

@ -3028,7 +3028,7 @@ void kernel_mul_mv_q4_K_f32_impl(
const int r0 = tgpig.x;
const int r1 = tgpig.y;
const int im = tgpig.z;
const int first_row = (r0 * N_SIMDGROUP + sgitg) * N_DST;
const int first_row = r0 * N_DST;
const int ib_row = first_row * nb;
const uint i12 = im%ne12;
@ -3094,7 +3094,7 @@ void kernel_mul_mv_q4_K_f32_impl(
for (int row = 0; row < N_DST; ++row) {
all_sum = simd_sum(sumf[row]);
if (tiisg == 0) {
dst[r1*ne0+ im*ne0*ne1 + first_row + row] = all_sum;
dst[r1*ne0 + im*ne0*ne1 + first_row + row] = all_sum;
}
}
}