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

Still fighting with lambda captures in MSVC

Browse source
This commit is contained in:
Iwan Kawrakow 2023-04-20 18:50:27 +02:00
parent 7fae1c4ee2
commit 0ae02ebb40
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
llama.cpp
View file

@ -1649,7 +1649,7 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
new_data = work.addr;
std::vector<int64_t> hist_cur(1 << 4, 0);
constexpr int chunk_size = 32 * 512;
int chunk_size = 32 * 512;
const int nchunk = (nelements + chunk_size - 1)/chunk_size;
const int nthread_use = nthread > 1 ? std::max(1, std::min(nthread, nchunk)) : 1;
if (nthread_use < 2) {
@ -1657,7 +1657,7 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
} else {
size_t counter = 0;
new_size = 0;
auto compute = [&mutex, &counter, &hist_cur, &new_size, new_type, f32_data, new_data, nelements] () {
auto compute = [&mutex, &counter, &hist_cur, &new_size, new_type, f32_data, new_data, nelements, chunk_size] () {
std::vector<int64_t> local_hist;
size_t local_size = 0;
while (true) {