subprocess llama.cpp server (#401)

* remove c code
* pack llama.cpp
* use request context for llama_cpp
* let llama_cpp decide the number of threads to use
* stop llama runner when app stops
* remove sample count and duration metrics
* use go generate to get libraries
* tmp dir for running llm

app/src/index.ts

@@ -158,7 +158,7 @@ function restart() {
 app.on('before-quit', () => {
   if (proc) {
     proc.off('exit', restart)
-    proc.kill()
+    proc.kill('SIGINT') // send SIGINT signal to the server, which also stops any loaded llms
   }
 })