LLM inference in C/C++
C++ ggml
server: remove all internal mentions about "webui" (#24817)
Xuan-Son Nguyen175147e8f612
diff --git a/tools/server/server-context.cpp b/tools/server/server-context.cppindex 00ab31340b..791188b1e7 100644--- a/tools/server/server-context.cpp+++ b/tools/server/server-context.cpp@@ -825,8 +825,7 @@ private:server_metrics metrics;- json json_ui_settings = json::object(); // Primary: new name- json json_webui_settings = json::object(); // Deprecated: use json_ui_settings instead (kept for compat)+ json json_ui_settings = json::object();// Necessary similarity of prompt for slot selectionfloat slot_prompt_similarity = 0.0f;@@ -1308,7 +1307,6 @@ private:try {json json_settings = json::parse(cfg);json_ui_settings = json_settings;- json_webui_settings = json_settings; // deprecated: keep in sync} catch (const std::exception & e) {SRV_ERR("%s: failed to parse UI config: %s\n", __func__, e.what());return false;@@ -3687,7 +3685,6 @@ server_context_meta server_context::get_meta() const {/* has_inp_audio */ impl->chat_params.allow_audio,/* has_inp_video */ impl->chat_params.allow_video,/* json_ui_settings */ impl->json_ui_settings,- /* json_webui_settings */ impl->json_webui_settings, // Deprecated/* slot_n_ctx */ impl->get_slot_n_ctx(),/* pooling_type */ llama_pooling_type(impl->ctx_tgt),@@ -4300,12 +4297,8 @@ void server_routes::init_routes() {{ "endpoint_slots", params.endpoint_slots },{ "endpoint_props", params.endpoint_props },{ "endpoint_metrics", params.endpoint_metrics },- // New keys{ "ui", params.ui },{ "ui_settings", meta->json_ui_settings },- // Deprecated: use ui/ui_settings instead (kept for backward compat)- { "webui", params.ui },- { "webui_settings", meta->json_ui_settings },{ "chat_template", tmpl_default },{ "chat_template_caps", meta->chat_template_caps },{ "bos_token", meta->bos_token_str },diff --git a/tools/server/server-context.h b/tools/server/server-context.hindex 0e84785af4..07afabb926 100644--- a/tools/server/server-context.h+++ b/tools/server/server-context.h@@ -22,8 +22,7 @@ struct server_context_meta {bool has_inp_image;bool has_inp_audio;bool has_inp_video;- json json_ui_settings; // Primary: new name- json json_webui_settings; // Deprecated: use json_ui_settings instead (kept for backward compat)+ json json_ui_settings;int slot_n_ctx;enum llama_pooling_type pooling_type;diff --git a/tools/server/server-models.cpp b/tools/server/server-models.cppindex 23c1f16689..1fffa6b6e5 100644--- a/tools/server/server-models.cpp+++ b/tools/server/server-models.cpp@@ -1474,7 +1474,6 @@ void server_models_routes::init_routes() {}},// New key{"ui_settings", ui_settings},- {"webui_settings", webui_settings},{"build_info", std::string(llama_build_info())},{"cors_proxy_enabled", params.ui_mcp_proxy},});diff --git a/tools/server/server-models.h b/tools/server/server-models.hindex aeb0e874de..98872b0461 100644--- a/tools/server/server-models.h+++ b/tools/server/server-models.h@@ -207,7 +207,6 @@ public:struct server_models_routes {common_params params;json ui_settings = json::object(); // Primary: new name- json webui_settings = json::object(); // Deprecated: use ui_settings (kept for compat)std::atomic<bool> stopping = false; // for graceful disconnecting SSE clients during shutdownserver_models models;server_models_routes(const common_params & params, int argc, char ** argv)@@ -217,7 +216,6 @@ struct server_models_routes {try {json json_settings = json::parse(cfg);ui_settings = json_settings;- webui_settings = json_settings; // Deprecated: keep in sync} catch (const std::exception & e) {LOG_ERR("%s: failed to parse UI config: %s\n", __func__, e.what());throw;diff --git a/tools/server/tests/unit/test_basic.py b/tools/server/tests/unit/test_basic.pyindex d1b89cf1a9..285726abf4 100644--- a/tools/server/tests/unit/test_basic.py+++ b/tools/server/tests/unit/test_basic.py@@ -79,9 +79,9 @@ def test_load_split_model():assert match_regex("(little|girl)+", res.body["content"])-def test_no_webui():+def test_no_ui():global server- # default: webui enabled+ # default: UI enabledserver.start()url = f"http://{server.server_host}:{server.server_port}"res = requests.get(url)@@ -89,8 +89,8 @@ def test_no_webui():assert "<!doctype html>" in res.textserver.stop()- # with --no-webui- server.no_webui = True+ # with --no-ui, the UI should be disabled+ server.no_ui = Trueserver.start()res = requests.get(url)assert res.status_code == 404diff --git a/tools/server/tests/unit/test_proxy.py b/tools/server/tests/unit/test_proxy.pyindex b7c3326187..3b86d80473 100644--- a/tools/server/tests/unit/test_proxy.py+++ b/tools/server/tests/unit/test_proxy.py@@ -12,7 +12,7 @@ def create_server():def test_mcp_no_proxy():global server- server.webui_mcp_proxy = False+ server.ui_mcp_proxy = Falseserver.start()res = server.make_request("GET", "/cors-proxy")@@ -21,7 +21,7 @@ def test_mcp_no_proxy():def test_mcp_proxy():global server- server.webui_mcp_proxy = True+ server.ui_mcp_proxy = Trueserver.start()url = f"http://{server.server_host}:{server.server_port}/cors-proxy?url=http://example.com"@@ -32,7 +32,7 @@ def test_mcp_proxy():def test_mcp_proxy_custom_port():global server- server.webui_mcp_proxy = True+ server.ui_mcp_proxy = Trueserver.start()# try getting the server's models API via the proxydiff --git a/tools/server/tests/utils.py b/tools/server/tests/utils.pyindex c50c9a0f5a..63a959449e 100644--- a/tools/server/tests/utils.py+++ b/tools/server/tests/utils.py@@ -94,7 +94,7 @@ class ServerProcess:enable_ctx_shift: int | None = Falsespec_draft_n_min: int | None = Nonespec_draft_n_max: int | None = None- no_webui: bool | None = None+ no_ui: bool | None = Nonejinja: bool | None = Nonereasoning_format: Literal['deepseek', 'none', 'nothink'] | None = Nonereasoning: Literal['on', 'off', 'auto'] | None = None@@ -107,7 +107,7 @@ class ServerProcess:cache_ram: int | None = Noneno_cache_idle_slots: bool = Falselog_path: str | None = None- webui_mcp_proxy: bool = False+ ui_mcp_proxy: bool = Falsebackend_sampling: bool = Falsegcp_compat: bool = False@@ -225,8 +225,8 @@ class ServerProcess:server_args.extend(["--spec-draft-n-max", self.spec_draft_n_max])if self.spec_draft_n_min:server_args.extend(["--spec-draft-n-min", self.spec_draft_n_min])- if self.no_webui:- server_args.append("--no-webui")+ if self.no_ui:+ server_args.append("--no-ui")if self.no_models_autoload:server_args.append("--no-models-autoload")if self.jinja:@@ -251,8 +251,8 @@ class ServerProcess:server_args.extend(["--cache-ram", self.cache_ram])if self.no_cache_idle_slots:server_args.append("--no-cache-idle-slots")- if self.webui_mcp_proxy:- server_args.append("--webui-mcp-proxy")+ if self.ui_mcp_proxy:+ server_args.append("--ui-mcp-proxy")if self.backend_sampling:server_args.append("--backend_sampling")if self.gcp_compat: