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

should be good to merge, only thing missing is clblast new quants

Browse source
This commit is contained in:
Concedo 2023-05-13 12:07:29 +08:00
parent 05cf5f7d6e
commit 53e7256a25
1 changed files with 2 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

11
ggml.c
View file

@ -1433,14 +1433,7 @@ bool quants_unshuffled = false; //new GGJT_2 is unshuffled, all old ones are shu
static const quantize_fns_t quantize_fns_v2[GGML_TYPE_COUNT]; //forward decl
static inline quantize_fns_t get_quantize_fn(size_t i)
{
if(quants_unshuffled)
{
return quantize_fns[i];
}
else
{
return quantize_fns_v2[i];
}
return(quants_unshuffled?quantize_fns[i]:quantize_fns_v2[i]);
}
@ -7869,7 +7862,7 @@ static void ggml_compute_forward_mul_mat_q_f32(
}
float * const wdata = params->wdata;
dequantize_row_q_t const dequantize_row_q = quantize_fns[type].dequantize_row_q;
dequantize_row_q_t const dequantize_row_q = get_quantize_fn(type).dequantize_row_q;
for (int64_t i03 = 0; i03 < ne03; i03++) {
for (int64_t i02 = 0; i02 < ne02; i02++) {