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

simplify for style

This commit is contained in:
ddh0 2026-02-11 21:41:37 -06:00
parent ae786b862d
commit 1ccd7a49ba
1 changed files with 5 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/llama-quant.cpp
View File

@ -480,14 +480,12 @@ static size_t llama_tensor_quantize_impl(enum ggml_type new_type, const float *
} }
static bool tensor_type_requires_imatrix(const ggml_type dst_type) { static bool tensor_type_requires_imatrix(const ggml_type dst_type) {
if (dst_type == GGML_TYPE_IQ2_XXS || dst_type == GGML_TYPE_IQ2_XS || return (
dst_type == GGML_TYPE_IQ2_XXS || dst_type == GGML_TYPE_IQ2_XS ||
dst_type == GGML_TYPE_IQ3_XXS || dst_type == GGML_TYPE_IQ1_S || dst_type == GGML_TYPE_IQ3_XXS || dst_type == GGML_TYPE_IQ1_S ||
dst_type == GGML_TYPE_IQ2_S || dst_type == GGML_TYPE_IQ1_M || dst_type == GGML_TYPE_IQ2_S || dst_type == GGML_TYPE_IQ1_M ||
dst_type == GGML_TYPE_TQ1_0 || dst_type == GGML_TYPE_TQ2_0) { dst_type == GGML_TYPE_TQ1_0 || dst_type == GGML_TYPE_TQ2_0
return true; );
} else {
return false;
}
} }
static void llama_model_quantize_impl(const std::string & fname_inp, const std::string & fname_out, const llama_model_quantize_params * params) { static void llama_model_quantize_impl(const std::string & fname_inp, const std::string & fname_out, const llama_model_quantize_params * params) {
@ -1066,6 +1064,7 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
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: 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); 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 (!params->imatrix && params->dry_run && will_require_imatrix) { if (!params->imatrix && params->dry_run && will_require_imatrix) {
LLAMA_LOG_WARN("%s: WARNING: dry run completed successfully, but actually completing this quantization will require an imatrix!\n", LLAMA_LOG_WARN("%s: WARNING: dry run completed successfully, but actually completing this quantization will require an imatrix!\n",
__func__ __func__