admin: surface CMF worker threads + GPU toggle in Settings

Expose the two speed knobs in the /admin Settings panel instead of TOML-only:
- [cmf].gpu (new) -> exports CMF_GPU=1 to the managed cortiq serve (Metal/wgpu).
  On Apple Silicon decode is memory-bandwidth bound so this is ~neutral for
  generation (helps prefill); off by default.
- [cmf].threads (added earlier) now has a UI number input.

- config.rs: CmfCfg.gpu bool (serde default false).
- cmf_runtime.rs: spawn_serve sets CMF_GPU=1 when gpu is on.
- admin/mod.rs: put_settings flags needs_restart on threads/gpu change (the
  server is spawned once at startup with those env vars).
- settings.js: threads input + gpu checkbox + save payload + speed hint.
- i18n en/ru: threads/gpu/speedNote keys (other locales fall back to en, as the
  whole cmf section already does there).

src/admin/mod.rs

@@ -655,9 +655,14 @@ async fn put_settings(State(state): State<SharedState>, Json(b): Json<SettingsBo
         cfg.cache = v;
     }
     if let Some(v) = b.cmf {
-        // manage_server / local_model changes apply on restart (the server is
-        // spawned at startup); local_only / router routing apply immediately.
-        if v.manage_server != cfg.cmf.manage_server || v.local_model != cfg.cmf.local_model {
+        // manage_server / local_model / threads / gpu apply on restart (the server
+        // is spawned once at startup with those env vars); local_only / router
+        // routing apply immediately.
+        if v.manage_server != cfg.cmf.manage_server
+            || v.local_model != cfg.cmf.local_model
+            || v.threads != cfg.cmf.threads
+            || v.gpu != cfg.cmf.gpu
+        {
             needs_restart = true;
         }
         cfg.cmf = v;

src/cmf_runtime.rs

@@ -167,6 +167,11 @@ fn spawn_serve(cfg: &CmfCfg) -> Result<Child, String> {
     if cfg.threads > 1 {
         cmd.env("CMF_THREADS", cfg.threads.to_string());
     }
+    // Offload large matvecs to the GPU (Metal on macOS). Memory-bandwidth-bound
+    // on unified memory, so mostly a prefill win; opt-in.
+    if cfg.gpu {
+        cmd.env("CMF_GPU", "1");
+    }
     cmd.stdout(Stdio::null())
         .stderr(Stdio::null())
         .kill_on_drop(true)

src/config.rs

@@ -116,6 +116,11 @@ pub struct CmfCfg {
     /// (measured ~2.8x faster than serial on a 15B model). Default 8.
     #[serde(default = "default_cmf_threads")]
     pub threads: usize,
+    /// Offload large matvecs to the GPU (exports CMF_GPU=1). On Apple Silicon the
+    /// decode loop is memory-bandwidth bound, so this is ~neutral for generation
+    /// (helps prefill of long prompts). Off by default.
+    #[serde(default)]
+    pub gpu: bool,
 }
 impl Default for CmfCfg {
     fn default() -> Self {
@@ -135,6 +140,7 @@ impl Default for CmfCfg {
             model_id: default_cmf_model_id(),
             local_only: false,
             threads: default_cmf_threads(),
+            gpu: false,
         }
     }
 }