feat(b3): ml-worker becomes optional — embed-only role, decoupled GPU coordination, cpu-embed switch

The ml-worker's ONLY processing role is now the CPU whole-image embed fallback
(tag_and_embed renamed embed_image — Camie tagging was retired #1189 and the
name kept implying otherwise; videos were already handled agent-style: frame
sampling + mean-pool). Detection/cropping/CCIP stay GPU-agent-only, and their
completion is judged per-pipeline: ccip by gpu_job rows, siglip by concept
regions at the current model version — never by image_record.siglip_embedding.
A CPU embed therefore can NEVER close crop work for the agent (regression test
pins this; only the whole-image 'embed' job, the same artifact, is satisfied).

Making removal actually safe (operator will drop the container):
- GPU-queue coordination (enqueue_gpu_backfill, recover_orphaned_gpu_jobs,
  reprocess_gpu_jobs) moved verbatim to tasks/gpu_queue.py on the maintenance
  quick lane — it lived on the 'ml' queue only by module colocation, which made
  the ml-worker a hard dependency of the whole agent pipeline.
- New ml_settings.cpu_embed_enabled (migration 0074, default ON so agent-less
  installs keep working): OFF stops the four import hooks queueing embed work
  nothing will consume and no-ops the manual backfill; switch lives on the
  renamed 'CPU embedding backfill' card.
- NB heads training / auto-apply still run on the ml image (sklearn) — a stack
  that removes the container gives those up too.

Deploy note: in-flight messages under the old task names are dropped by the
new workers; the 60s orphan sweep + hourly backfill re-fire under the new
names immediately.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM

frontend/src/components/settings/MLBackfillCard.vue

@@ -1,16 +1,33 @@
 <template>
   <MaintenanceTile
     icon="mdi-refresh"
-    title="ML backfill"
-    blurb="Compute SigLIP embeddings on images missing them."
+    title="CPU embedding backfill"
+    blurb="Whole-image embeddings without a GPU agent — the built-in fallback."
     :open="busy"
   >
       <p class="text-body-2 mb-3">
-        Compute the SigLIP embedding for any image that doesn't have one yet
-        (CPU). Safe to re-run. To re-embed under a NEW model, use the GPU
-        agent's "Re-embed library" instead.
+        Computes the whole-image SigLIP embedding for anything missing one —
+        images directly, videos by sampling frames (the same approach as the
+        GPU agent). Runs on the ml-worker's CPU, so search, similarity and
+        head suggestions work <strong>without</strong> a GPU agent; new imports
+        are embedded this way automatically. Detection, cropping and character
+        (CCIP) embeddings are GPU-agent-only. Safe to re-run. To re-embed under
+        a NEW model, use the GPU agent's "Re-embed library" instead.
       </p>
-      <v-btn color="primary" rounded="pill" :loading="busy" @click="run">
+      <v-switch
+        v-model="enabled" color="accent" hide-details density="compact"
+        :loading="saving" label="CPU embedding enabled"
+        class="mb-1" @update:model-value="onToggle"
+      />
+      <p class="fc-muted text-caption mb-3">
+        Turn OFF if you run the GPU agent and removed the ml-worker container —
+        imports then stop queueing CPU embed work nothing will consume (the
+        daily GPU embed backfill covers those images instead).
+      </p>
+      <v-btn
+        color="primary" rounded="pill" :loading="busy" :disabled="!enabled"
+        @click="run"
+      >
         <v-icon start>mdi-refresh</v-icon> Run backfill now
       </v-btn>
       <span v-if="done" class="ml-3 text-caption">Enqueued.</span>
@@ -20,13 +37,40 @@
 
 <script setup>
 import { toast } from '../../utils/toast.js'
-import { ref } from 'vue'
+import { onMounted, ref } from 'vue'
 import { useMLStore } from '../../stores/ml.js'
 import MaintenanceTile from '../common/MaintenanceTile.vue'
 import QueueStatusBar from './QueueStatusBar.vue'
 const store = useMLStore()
 const busy = ref(false)
 const done = ref(false)
+const enabled = ref(true)
+const saving = ref(false)
+onMounted(async () => {
+  try {
+    await store.loadSettings()
+    if (store.settings?.cpu_embed_enabled != null) {
+      enabled.value = store.settings.cpu_embed_enabled
+    }
+  } catch { /* non-fatal */ }
+})
+async function onToggle() {
+  saving.value = true
+  try {
+    await store.patchSettings({ cpu_embed_enabled: enabled.value })
+    toast({
+      text: enabled.value
+        ? 'CPU embedding on — imports queue embeds for the ml-worker'
+        : 'CPU embedding off — the GPU embed backfill owns whole-image embeds',
+      type: 'success',
+    })
+  } catch (e) {
+    toast({ text: `Could not save: ${e.message}`, type: 'error' })
+    enabled.value = !enabled.value
+  } finally {
+    saving.value = false
+  }
+}
 async function run() {
   busy.value = true
   try { await store.triggerBackfill(); done.value = true }
@@ -34,3 +78,7 @@ async function run() {
   finally { busy.value = false }
 }
 </script>
+
+<style scoped>
+.fc-muted { color: rgb(var(--v-theme-on-surface-variant)); }
+</style>