happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

show model and quant BPW when quant completes

This commit is contained in:
ddh0 2026-02-11 15:30:12 -06:00
parent 150e1db21d
commit 966b21a981
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
src/llama-quant.cpp
View File

@ -1051,8 +1051,8 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
close_ofstream();
}
LLAMA_LOG_INFO("%s: model size = %8.2f MiB\n", __func__, total_size_org/1024.0/1024.0);
LLAMA_LOG_INFO("%s: quant size = %8.2f MiB\n", __func__, total_size_new/1024.0/1024.0);
LLAMA_LOG_INFO("%s: model size = %8.2f MiB (%.2f BPW)\n", __func__, total_size_org/1024.0/1024.0, total_size_org*8.0/ml.n_elements);
LLAMA_LOG_INFO("%s: quant size = %8.2f MiB (%.2f BPW)\n", __func__, total_size_new/1024.0/1024.0, total_size_new*8.0/ml.n_elements);
if (qs.n_fallback > 0) {
LLAMA_LOG_WARN("%s: WARNING: %d of %d tensor(s) required fallback quantization\n",