ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-03-07 04:20:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

gguf-split: fix the split output files naming (#1336)

Browse Source 
* Fix gguf-split.cpp splits output naming

With this fix, the initial extension of the source .gguf file is not included in the naming of the output file before the numeration of the splits.

ex:

No more model.gguf-00001-of-00200.gguf
Instead, model-00001-of-00200.gguf

* increase ggml_max_context to 2048

* Revert GGML_MAX_CONTEXTS to 64
This commit is contained in:
Nexes the Elder
2026-03-02 08:43:47 +01:00
committed by GitHub
parent d239dabcc6
commit d4ac5f1566
2 changed files with 8 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
examples/gguf-split/gguf-split.cpp
View File

@@ -318,10 +318,16 @@ struct split_strategy {
void write() {
int i_split = 0;
int n_split = ctx_outs.size();
std::string output_prefix = params.output;
const std::string suffix = ".gguf";
if (output_prefix.size() >= suffix.size() &&
output_prefix.compare(output_prefix.size() - suffix.size(), suffix.size(), suffix) == 0) {
output_prefix.resize(output_prefix.size() - suffix.size());
}
for (auto & ctx_out : ctx_outs) {
// construct file path
char split_path[PATH_MAX] = {0};
llama_split_path(split_path, sizeof(split_path), params.output.c_str(), i_split, n_split);
llama_split_path(split_path, sizeof(split_path), output_prefix.c_str(), i_split, n_split);
ensure_output_directory(split_path);

2
ggml/include/ggml.h
View File

@@ -238,7 +238,7 @@
// Maximum number of model contexts (e.g., for model shards).
// Increase this value using -DGGML_MAX_CONTEXTS=<value> in CMake
// if you need to load more than 64 model shards.
#define GGML_MAX_CONTEXTS 64
#define GGML_MAX_CONTEXTS 64
#endif
#define GGML_MAX_SRC 10
#ifndef GGML_MAX_NAME