Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-27 04:23:06 +01:00
Code Issues Packages Projects Releases Wiki Activity

Temporarily add logging of free device memory at the end of main

Browse Source
This commit is contained in:
Nikita Sarychev 2025-01-25 17:17:21 -08:00
parent 7088822e5a
commit cbf779c450
1 changed files with 11 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
examples/main/main.cpp
View File

@ -3,6 +3,10 @@
#include "console.h"
#include "log.h"
#include "sampling.h"
// vvv REMOVE BEFORE MERGING
#include "llama-model.h"
#include "llama-impl.h"
// ^^^ REMOVE BEFORE MERGING
#include "llama.h"
#include "chat-template.hpp"
@ -912,6 +916,13 @@ int main(int argc, char ** argv) {
}
LOG("\n\n");
// vvv REMOVE BEFORE MERGING
for (auto * dev : model->devices) {
size_t free, total; // NOLINT
ggml_backend_dev_memory(dev, &free, &total);
LLAMA_LOG_INFO("%s: using device %s (%s) - %zu MiB free\n", __func__, ggml_backend_dev_name(dev), ggml_backend_dev_description(dev), free/1024/1024);
}
// ^^^ REMOVE BEFORE MERGING
common_perf_print(ctx, smpl);
common_sampler_free(smpl);