Distribute and run LLMs with a single file.
C++ cross-platformggufllama-cpplocal-ailocal-inferencelocal-llmopen-source-aisingle-file-executablespeech-to-text
diff --git a/tools/server/server-http.cpp b/tools/server/server-http.cpp
--- a/llama.cpp/tools/server/server-http.cpp
+++ b/llama.cpp/tools/server/server-http.cpp
@@ -159,7 +159,7 @@ bool server_http_context::init(const common_params & params) {
bool server_http_context::init_listener(const common_params & params) {
auto & srv = pimpl->servers.back();
-#ifdef CPPHTTPLIB_OPENSSL_SUPPORT
+#ifdef CPPHTTPLIB_SSL_ENABLED
if (!params.ssl_file_key.empty() && !params.ssl_file_cert.empty()) {
SRV_TRC("running with SSL: key = %s, cert = %s\n", params.ssl_file_key.c_str(), params.ssl_file_cert.c_str());
srv = std::make_unique<httplib::SSLServer>(
@@ -491,7 +491,11 @@ self.addEventListener('activate', (e) => e.waitUntil((async () => {
return true;
}
-bool server_http_context::start() {
+bool server_http_context::bind_socket() {
+ if (is_bound) {
+ return true;
+ }
+
// Bind and listen
listening_addresses.clear();
@@ -524,6 +528,15 @@ bool server_http_context::start() {
: string_format("%s://%s:%d", is_ssl ? "https" : "http", common_http_format_host(host).c_str(), port));
}
+ is_bound = true;
+ return true;
+}
+
+bool server_http_context::start() {
+ if (!bind_socket()) {
+ return false;
+ }
+
// n_threads_http fixed threads (always alive), plus up to 1024 dynamic threads destroyed after each request
// ref: https://github.com/yhirose/cpp-httplib/pull/2368
pimpl->pool = std::make_unique<httplib::ThreadPool>(pimpl->n_threads_http, pimpl->n_threads_http + 1024);