vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge branch 'master' into compilade/imatrix-batched-chunks

Browse source
This commit is contained in:
Francis Couture-Harpin 2025-02-09 12:06:15 -05:00
commit db502ddd0e
762 changed files with 149462 additions and 91773 deletions
Download patch file Download diff file Expand all files Collapse all files

34
examples/quantize/quantize.cpp
View file

@ -1,5 +1,6 @@
#include "common.h"
#include "llama.h"
#include "gguf.h"
#include <cstdio>
#include <cstring>
@ -47,9 +48,6 @@ static const std::vector<struct quant_option> QUANT_OPTIONS = {
{ "Q5_K_M", LLAMA_FTYPE_MOSTLY_Q5_K_M, " 5.33G, +0.0569 ppl @ Llama-3-8B", },
{ "Q6_K", LLAMA_FTYPE_MOSTLY_Q6_K, " 6.14G, +0.0217 ppl @ Llama-3-8B", },
{ "Q8_0", LLAMA_FTYPE_MOSTLY_Q8_0, " 7.96G, +0.0026 ppl @ Llama-3-8B", },
{ "Q4_0_4_4", LLAMA_FTYPE_MOSTLY_Q4_0_4_4, " 4.34G, +0.4685 ppl @ Llama-3-8B", },
{ "Q4_0_4_8", LLAMA_FTYPE_MOSTLY_Q4_0_4_8, " 4.34G, +0.4685 ppl @ Llama-3-8B", },
{ "Q4_0_8_8", LLAMA_FTYPE_MOSTLY_Q4_0_8_8, " 4.34G, +0.4685 ppl @ Llama-3-8B", },
{ "F16", LLAMA_FTYPE_MOSTLY_F16, "14.00G, +0.0020 ppl @ Mistral-7B", },
{ "BF16", LLAMA_FTYPE_MOSTLY_BF16, "14.00G, -0.0050 ppl @ Mistral-7B", },
{ "F32", LLAMA_FTYPE_ALL_F32, "26.00G @ 7B", },
@ -67,6 +65,16 @@ static const char * const LLM_KV_IMATRIX_DATASET = "imatrix.dataset";
static const char * const LLM_KV_IMATRIX_CHUNK_COUNT = "imatrix.chunk_count";
static const char * const LLM_KV_IMATRIX_CHUNK_SIZE = "imatrix.chunk_size";
static bool striequals(const char * a, const char * b) {
while (*a && *b) {
if (std::tolower(*a) != std::tolower(*b)) {
return false;
}
a++; b++;
}
return *a == *b;
}
static bool try_parse_ftype(const std::string & ftype_str_in, llama_ftype & ftype, std::string & ftype_str_out) {
std::string ftype_str;
@ -74,7 +82,7 @@ static bool try_parse_ftype(const std::string & ftype_str_in, llama_ftype & ftyp
ftype_str.push_back(std::toupper(ch));
}
for (auto & it : QUANT_OPTIONS) {
if (it.name == ftype_str) {
if (striequals(it.name.c_str(), ftype_str.c_str())) {
ftype = it.ftype;
ftype_str_out = it.name;
return true;
@ -276,15 +284,15 @@ static int prepare_imatrix(const std::string & imatrix_file,
}
static ggml_type parse_ggml_type(const char * arg) {
ggml_type result = GGML_TYPE_COUNT;
for (int j = 0; j < GGML_TYPE_COUNT; ++j) {
auto type = ggml_type(j);
for (int i = 0; i < GGML_TYPE_COUNT; ++i) {
auto type = (ggml_type)i;
const auto * name = ggml_type_name(type);
if (name && strcmp(arg, name) == 0) {
result = type; break;
if (name && striequals(name, arg)) {
return type;
}
}
return result;
fprintf(stderr, "%s: invalid ggml_type '%s'\n", __func__, arg);
return GGML_TYPE_COUNT;
}
int main(int argc, char ** argv) {
@ -305,12 +313,18 @@ int main(int argc, char ** argv) {
} else if (strcmp(argv[arg_idx], "--output-tensor-type") == 0) {
if (arg_idx < argc-1) {
params.output_tensor_type = parse_ggml_type(argv[++arg_idx]);
if (params.output_tensor_type == GGML_TYPE_COUNT) {
usage(argv[0]);
}
} else {
usage(argv[0]);
}
} else if (strcmp(argv[arg_idx], "--token-embedding-type") == 0) {
if (arg_idx < argc-1) {
params.token_embedding_type = parse_ggml_type(argv[++arg_idx]);
if (params.token_embedding_type == GGML_TYPE_COUNT) {
usage(argv[0]);
}
} else {
usage(argv[0]);
}