CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-27 08:21:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

nitpick : correct MB to MiB (#15934)

Browse Source 
MB was incorrectly used for 1024 x 1024 bytes instead of MiB
This commit is contained in:
ddh0
2025-09-11 12:12:34 -05:00
committed by GitHub
parent 24a6734daf
commit df082f5630
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/llama-quant.cpp
View File

@@ -920,7 +920,7 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
new_type = tensor->type; new_type = tensor->type;
new_data = tensor->data; new_data = tensor->data;
new_size = ggml_nbytes(tensor); new_size = ggml_nbytes(tensor);
LLAMA_LOG_INFO("size = %8.3f MB\n", ggml_nbytes(tensor)/1024.0/1024.0); LLAMA_LOG_INFO("size = %8.3f MiB\n", ggml_nbytes(tensor)/1024.0/1024.0);
} else { } else {
const int64_t nelements = ggml_nelements(tensor); const int64_t nelements = ggml_nelements(tensor);
@@ -1037,8 +1037,8 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
} }
close_ofstream(); close_ofstream();
LLAMA_LOG_INFO("%s: model size = %8.2f MB\n", __func__, total_size_org/1024.0/1024.0); 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 MB\n", __func__, total_size_new/1024.0/1024.0); LLAMA_LOG_INFO("%s: quant size = %8.2f MiB\n", __func__, total_size_new/1024.0/1024.0);
if (qs.n_fallback > 0) { if (qs.n_fallback > 0) {
LLAMA_LOG_WARN("%s: WARNING: %d of %d tensor(s) required fallback quantization\n", LLAMA_LOG_WARN("%s: WARNING: %d of %d tensor(s) required fallback quantization\n",