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

Updating files with upstream changes prior to rebase

This commit is contained in:
Colin Kealty 2026-03-10 10:41:22 -04:00
parent 86273028d1
commit 6e414fc2d6
2 changed files with 13 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
src/llama-quant.h
View File

@ -6,11 +6,19 @@
#include "llama-arch.h"
#include <regex>
#include <string>
#include <vector>
struct llama_model;
// result of parsing --tensor-type option
// (changes to this struct must be reflected in tools/quantize/quantize.cpp)
struct tensor_type_option {
std::string name;
ggml_type type = GGML_TYPE_COUNT;
};
struct quantize_state_impl {
const llama_model & model;
const llama_model_quantize_params * params;
@ -30,7 +38,7 @@ struct quantize_state_impl {
bool has_imatrix = false;
// used to figure out if a model has tied embeddings (tok_embd shares weights with output)
bool has_tied_embeddings = false; // assume tied until we see output.weight
bool has_tied_embeddings = true; // assume tied until we see output.weight
// tensor type override patterns (compiled once, used twice)
std::vector<std::pair<std::regex, ggml_type>> tensor_type_patterns;

8
tests/test-quant-type-selection.cpp
View File

@ -309,7 +309,7 @@ static std::string generate_snapshot(const std::string & name,
for (int i = 0; i < LLAMA_FTYPE_GUESSED; i++) {
llama_ftype ft = (llama_ftype) i;
ggml_type default_type = llama_ftype_default_type(ft);
ggml_type default_type = llama_ftype_get_default_type(ft);
if (default_type == GGML_TYPE_COUNT) {
continue;
}
@ -384,8 +384,8 @@ static int run_generate(const std::string & snapshot_dir) {
static bool run_test_section(llama_model & mdl,
const std::vector<mock_tensor> & tensors,
const snapshot_section & section) {
// verify default_type matches what llama_ftype_default_type returns
ggml_type computed_default = llama_ftype_default_type(section.ftype);
// verify default_type matches what llama_ftype_get_default_type returns
ggml_type computed_default = llama_ftype_get_default_type(section.ftype);
if (computed_default != section.default_type) {
printf(" FAIL [%s] default type mismatch: file says %s, code says %s\n", llama_ftype_to_name(section.ftype),
ggml_type_name(section.default_type), ggml_type_name(computed_default));
@ -408,7 +408,7 @@ static bool run_test_section(llama_model & mdl,
}
if (got != expected) {
printf(" FAIL %-50s expected %s, got %s\n", name.c_str(), ggml_type_name(expected), ggml_type_name(got));
printf(" FAIL [%s] %-50s expected %s, got %s\n", llama_ftype_to_name(section.ftype), name.c_str(), ggml_type_name(expected), ggml_type_name(got));
all_pass = false;
}
}