EngineX-Mthreads/enginex-mthreads-llama.cpp
4
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

sync from b7516

Browse Source
This commit is contained in:
xiezhongtao
2026-01-16 11:16:14 +08:00
parent f4ae4cc7da
commit 6ee41dd9e3
380 changed files with 18435 additions and 38806 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
tools/server/server-models.h
View File

@@ -9,7 +9,6 @@
#include <condition_variable>
#include <functional>
#include <memory>
#include <set>
/**
* state diagram:
@@ -57,7 +56,6 @@ struct server_model_meta {
int64_t last_used = 0; // for LRU unloading
std::vector<std::string> args; // args passed to the model instance, will be populated by render_args()
int exit_code = 0; // exit code of the model instance process (only valid if status == FAILED)
int stop_timeout = 0; // seconds to wait before force-killing the model instance during shutdown
bool is_active() const {
return status == SERVER_MODEL_STATUS_LOADED || status == SERVER_MODEL_STATUS_LOADING;
@@ -85,10 +83,6 @@ private:
std::condition_variable cv;
std::map<std::string, instance_t> mapping;
// for stopping models
std::condition_variable cv_stop;
std::set<std::string> stopping_models;
common_preset_context ctx_preset;
common_params base_params;
@@ -105,7 +99,7 @@ private:
void add_model(server_model_meta && meta);
public:
server_models(const common_params & params, int argc, char ** argv);
server_models(const common_params & params, int argc, char ** argv, char ** envp);
void load_models();
@@ -125,7 +119,7 @@ public:
void unload_all();
// update the status of a model instance (thread-safe)
void update_status(const std::string & name, server_model_status status, int exit_code);
void update_status(const std::string & name, server_model_status status);
// wait until the model instance is fully loaded (thread-safe)
// return when the model is loaded or failed to load
@@ -147,8 +141,8 @@ struct server_models_routes {
common_params params;
json webui_settings = json::object();
server_models models;
server_models_routes(const common_params & params, int argc, char ** argv)
: params(params), models(params, argc, argv) {
server_models_routes(const common_params & params, int argc, char ** argv, char ** envp)
: params(params), models(params, argc, argv, envp) {
if (!this->params.webui_config_json.empty()) {
try {
webui_settings = json::parse(this->params.webui_config_json);
@@ -183,10 +177,7 @@ public:
const std::string & path,
const std::map<std::string, std::string> & headers,
const std::string & body,
const std::function<bool()> should_stop,
int32_t timeout_read,
int32_t timeout_write
);
const std::function<bool()> should_stop);
~server_http_proxy() {
if (cleanup) {
cleanup();