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

Update output log

This commit is contained in:
Ed Addario 2026-01-22 23:19:26 +00:00
parent 0b5030d704
commit 220df5f1ff
No known key found for this signature in database
GPG Key ID: E7875815A3230993
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/llama-quant.cpp
View File

@ -2140,24 +2140,26 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
// get more optimal quantization type based on the tensor shape, layer, etc. // get more optimal quantization type based on the tensor shape, layer, etc.
if (!params->pure && (ggml_is_quantized(default_type) || params->target_bpw != -1.0f || params->target_size != -1)) { if (!params->pure && (ggml_is_quantized(default_type) || params->target_bpw != -1.0f || params->target_size != -1)) {
bool manual = false;
// get quantization type overrides targeting a bpw or file size budget // get quantization type overrides targeting a bpw or file size budget
if ((params->target_bpw != -1.0f || params->target_size != -1) && !bpw_overrides.empty()) { if ((params->target_bpw != -1.0f || params->target_size != -1) && !bpw_overrides.empty()) {
const auto override = bpw_overrides.find(name); const auto override = bpw_overrides.find(name);
if (override != bpw_overrides.end() && override->second != new_type) { if (override != bpw_overrides.end() && override->second != new_type) {
LLAMA_LOG_DEBUG("(bpw override %s) ", ggml_type_name(new_type)); LLAMA_LOG_WARN("(target override: %s) ", ggml_type_name(new_type));
new_type = override->second; new_type = override->second;
manual = true;
} }
} }
// if the user provided tensor types - use those // if the user provided tensor types - use those
bool manual = false;
if (params->tensor_types) { if (params->tensor_types) {
const std::vector<tensor_quantization> & tensor_types = *static_cast<const std::vector<tensor_quantization> *>(params->tensor_types); const std::vector<tensor_quantization> & tensor_types = *static_cast<const std::vector<tensor_quantization> *>(params->tensor_types);
const std::string tensor_name(tensor->name); const std::string tensor_name(tensor->name);
for (const auto & [tname, qtype] : tensor_types) { for (const auto & [tname, qtype] : tensor_types) {
if (std::regex pattern(tname); std::regex_search(tensor_name, pattern)) { if (std::regex pattern(tname); std::regex_search(tensor_name, pattern)) {
if (qtype != new_type) { if (qtype != new_type) {
LLAMA_LOG_WARN("(manual override: %s -> %s) ", ggml_type_name(new_type), ggml_type_name(qtype)); LLAMA_LOG_WARN("(manual override: %s) ", ggml_type_name(new_type));
new_type = qtype; // if two or more types are specified for the same tensor, the last match wins new_type = qtype; // if two or more types are specified for the same tensor, the last match wins
manual = true; manual = true;
break; break;