root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-27 03:44:35 +00:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

imatrix : remove unused n_entries

Browse Source
This commit is contained in:
Francis Couture-Harpin 2024-09-09 22:35:47 -04:00
parent efa9186dc8
commit 2217247051
1 changed files with 0 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
examples/imatrix/imatrix.cpp
View File

@ -235,7 +235,6 @@ void IMatrixCollector::save_imatrix(int32_t n_chunk) const {
// avoid writing imatrix entries that do not have full data // avoid writing imatrix entries that do not have full data
// this can happen with MoE models where some of the experts end up not being exercised by the provided training data // this can happen with MoE models where some of the experts end up not being exercised by the provided training data
int n_entries = 0;
std::vector<std::string> to_store; std::vector<std::string> to_store;
size_t data_size = 0; size_t data_size = 0;
@ -269,7 +268,6 @@ void IMatrixCollector::save_imatrix(int32_t n_chunk) const {
continue; continue;
} }
n_entries++;
to_store.push_back(kv.first); to_store.push_back(kv.first);
data_size += GGML_PAD(ggml_tensor_overhead() + sizeof(float) * kv.second.values.size(), GGML_MEM_ALIGN); data_size += GGML_PAD(ggml_tensor_overhead() + sizeof(float) * kv.second.values.size(), GGML_MEM_ALIGN);
data_size += GGML_PAD(ggml_tensor_overhead() + sizeof(float) * kv.second.counts.size(), GGML_MEM_ALIGN); data_size += GGML_PAD(ggml_tensor_overhead() + sizeof(float) * kv.second.counts.size(), GGML_MEM_ALIGN);