public_git/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-10-03 13:03:07 +02:00
Code Issues Packages Projects Releases Wiki Activity

talk-llama : sync llama.cpp

Browse Source
This commit is contained in:
Georgi Gerganov
2026-01-30 15:56:40 +02:00
parent b529c0610f
commit 953e503fd9
38 changed files with 2299 additions and 730 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/talk-llama/llama-model-loader.cpp
+20 -6
View File
@@ -2,6 +2,7 @@
#include "ggml.h"
#include <algorithm>
#include <array>
#include <cinttypes>
#include <cstring>
@@ -344,6 +345,7 @@ namespace GGUFMeta {
GGUFMeta::GKV<GGUFMeta::ArrayInfo>::get_kv(ctx, kid);
switch (arr_info.gt) {
case GGUF_TYPE_BOOL:
case GGUF_TYPE_UINT32:
case GGUF_TYPE_INT32: GGML_ASSERT((std::is_same<T, int32_t>::value) ||
(std::is_same<T, uint32_t>::value)); break;
@@ -365,7 +367,13 @@ namespace GGUFMeta {
result[i] = value;
}
} else {
std::copy((const T*)arr_info.data, (const T *)arr_info.data + arr_info.length, result.begin());
if (arr_info.gt == GGUF_TYPE_BOOL) {
std::transform((const bool *)arr_info.data, (const bool *)arr_info.data + arr_info.length, result.begin(), [](bool x) {
return static_cast<T>(x);
});
} else {
std::copy((const T*)arr_info.data, (const T *)arr_info.data + arr_info.length, result.begin());
}
}
return true;
@@ -531,12 +539,18 @@ llama_model_loader::llama_model_loader(
files.emplace_back(new llama_file(fname.c_str(), "rb", use_direct_io));
contexts.emplace_back(ctx);
use_direct_io = use_direct_io && files.back()->has_direct_io();
if (use_mmap && use_direct_io) {
if (files.back()->has_direct_io()) {
LLAMA_LOG_WARN("%s: direct I/O is enabled, disabling mmap\n", __func__);
use_mmap = false;
} else {
LLAMA_LOG_WARN("%s: direct I/O is not available, using mmap\n", __func__);
use_direct_io = false;
// Disable mmap in case Direct I/O is enabled and available
if (use_direct_io && use_mmap) {
use_mmap = false;
LLAMA_LOG_WARN("%s: direct I/O is enabled, disabling mmap\n", __func__);
// reopen file using std::fopen for mmap
files.pop_back();
files.emplace_back(new llama_file(fname.c_str(), "rb", false));
}
}
// Save tensors data offset of the main file.