ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-04-29 19:01:47 +00:00
Code Issues Packages Projects Releases Wiki Activity

MSVC fixes (#161)

Browse Source 
Closes #160 

* MSVC fixes

* One more

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2024-12-23 07:57:48 +01:00
committed by GitHub
parent 2ed8f432a4
commit aa2595415a
2 changed files with 15 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
ggml/src/iqk/iqk_quantize.cpp
View File

@@ -5740,7 +5740,7 @@ void iqk_repack_tensor(struct ggml_tensor * tensor) {
// int(tensor->ne[1]), num_chunks, nthread);
std::atomic<int> counter(0);;
auto compute = [&counter, &r, tensor, num_chunks] () {
auto compute = [&counter, &r, tensor, num_chunks, chunkSize = kChunk] () {
int nrows = tensor->ne[1];
int n_per_row = tensor->ne[0];
auto row_size = ggml_row_size(tensor->type, n_per_row);
@@ -5749,8 +5749,8 @@ void iqk_repack_tensor(struct ggml_tensor * tensor) {
while (true) {
int chunk = counter.fetch_add(1);
if (chunk >= num_chunks) break;
int first_row = chunk*kChunk*r.num_rows;
int last_row = std::min(first_row + kChunk*r.num_rows, nrows);
int first_row = chunk*chunkSize*r.num_rows;
int last_row = std::min(first_row + chunkSize*r.num_rows, nrows);
for (int row = first_row; row < last_row; row += r.num_rows) {
std::memcpy(qtmp.data(), data + row*row_size, r.num_rows*row_size);
r.repack(r.num_rows, n_per_row, qtmp.data(), data + row*row_size);