root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-09-22 21:16:20 +00:00
Code Issues Actions 16 Packages Projects Releases Wiki Activity

ggml : do not crash when quantizing q4_x_x with an imatrix (#9192)

Browse Source
This commit is contained in:
slaren 2024-08-26 19:44:43 +02:00 committed by GitHub
parent 06658ad7c3
commit 7d787ed96c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 6 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
ggml/src/ggml-aarch64.c
View File

@ -337,33 +337,18 @@ static size_t quantize_q4_0_nr_bl(const float * restrict src, void * restrict ds
}
size_t quantize_q4_0_4x4(const float * restrict src, void * restrict dst, int64_t nrow, int64_t n_per_row, const float * quant_weights) {
if (!quant_weights) {
return quantize_q4_0_nr_bl(src, dst, nrow, n_per_row, 4, 4);
}
else {
assert(false);
return 0;
}
UNUSED(quant_weights);
return quantize_q4_0_nr_bl(src, dst, nrow, n_per_row, 4, 4);
}
size_t quantize_q4_0_4x8(const float * restrict src, void * restrict dst, int64_t nrow, int64_t n_per_row, const float * quant_weights) {
if (!quant_weights) {
return quantize_q4_0_nr_bl(src, dst, nrow, n_per_row, 4, 8);
}
else {
assert(false);
return 0;
}
UNUSED(quant_weights);
return quantize_q4_0_nr_bl(src, dst, nrow, n_per_row, 4, 8);
}
size_t quantize_q4_0_8x8(const float * restrict src, void * restrict dst, int64_t nrow, int64_t n_per_row, const float * quant_weights) {
if (!quant_weights) {
return quantize_q4_0_nr_bl(src, dst, nrow, n_per_row, 8, 8);
}
else {
assert(false);
return 0;
}
UNUSED(quant_weights);
return quantize_q4_0_nr_bl(src, dst, nrow, n_per_row, 8, 8);
}
void ggml_gemv_q4_0_4x4_q8_0(int n, float * restrict s, size_t bs, const void * restrict vx, const void * restrict vy, int nr, int nc) {