llama.cpp/tests/test-autorelease.cpp

25 lines
724 B
C++
Raw Normal View History

// ref: https://github.com/ggerganov/llama.cpp/issues/4952#issuecomment-1892864763
#include <cstdio>
#include <string>
#include <thread>
#include "llama.h"
#include "get-model.h"
// This creates a new context inside a pthread and then tries to exit cleanly.
int main(int argc, char ** argv) {
auto * model_path = get_model_or_exit(argc, argv);
std::thread([&model_path]() {
llama_backend_init(false);
auto * model = llama_load_model_from_file(model_path, llama_model_default_params());
auto * ctx = llama_new_context_with_model(model, llama_context_default_params());
llama_free(ctx);
llama_free_model(model);
llama_backend_free();
}).join();
return 0;
}