router: Bildgenerierung mit FLUX.2 [klein] 4B Base (GPU-Hotswap)

- POST /v1/images/generations (OpenAI-kompatibel, prompt/size/n/seed/quality)
- quality: standard=30 Steps (Default), high=50 Steps
- Größen: 1024x1024, 1536x1024, 1024x1536, 1920x1088, 1088x1920
- GPU-Hotswap: Qwen stoppen -> FLUX laden -> Bild -> FLUX entladen -> Qwen
  wiederherstellen (exakt vorheriges Profil)
- Zentrales GPU/Modell-Lock (Profilwechsel und Bild teilen sich das Lock)
- Chat-Requests warten während Bild-Job (kein 502), Timeout CHAT_WAIT_TIMEOUT
- Robuste Recovery: try/finally, Worker-Beendigung, VRAM-Check, Qwen-Readiness
- /status: image.phase, image.worker, image.model_loaded, qwen.available,
  qwen.active_chats
- GET /images, GET /images/<datei> (validiert, nur images/-Verzeichnis)
- image_worker.py: FLUX-Worker (eigener Prozess, JSON-Protokoll, bf16 +
  enable_model_cpu_offload)
- deploy: venv (torch/diffusers/transformers/accelerate), Modell-Download,
  Image-Dir, systemd-Unit mit Image-Umgebungsvariablen
- dev: Mock-Worker, fake-systemctl, Benchmarks (GPU-Resident, Offload, Steps,
  Quality-Compare), 32 lokale Tests
- README: Bildgenerierung, Hotswap, Recovery, Benchmarks (RTX 5080),
  Python-Pakete

Benchmarks (RTX 5080, 16 GB, CPU-Offload):
- 512x512 / 10 Steps: ~9.3 s
- 1024x1024 / 30 Steps: ~31.3 s
- 1024x1024 / 50 Steps: ~45.3 s
- 1920x1088 / 50 Steps: ~91 s
- Peak-VRAM: ~8.4-8.9 GB
- Hotswap-Gesamtzeit: ~41-42 s (1024x1024 / 30 Steps)

deploy/deploy.sh

@@ -9,7 +9,7 @@ SSH_KEY="${SSH_KEY:-$HOME/.ssh/lmstudio_unraid}"
 STAGE="/tmp/ai-profile-router-$$"
 
 mkdir -p "$STAGE"
-cp router/ai_profile_router.py deploy/install.sh deploy/mike-ai-profile-router.service "$STAGE/"
+cp router/ai_profile_router.py router/image_worker.py deploy/install.sh deploy/mike-ai-profile-router.service "$STAGE/"
 
 echo "== Übertrage Dateien nach ${TARGET}:/tmp/ai-profile-router/"
 ssh -i "$SSH_KEY" "$TARGET" 'mkdir -p /tmp/ai-profile-router'

deploy/install.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 # AI Profile Router – Installation/Update auf dem Zielsystem.
 # Wird als root auf dem Zielsystem ausgeführt (per SSH, vgl. deploy.sh).
-# Erwartet ai_profile_router.py und mike-ai-profile-router.service
+# Erwartet ai_profile_router.py, image_worker.py und mike-ai-profile-router.service
 # im selben Verzeichnis wie dieses Skript.
 set -euo pipefail
 
@@ -11,6 +11,9 @@ SERVICE=mike-ai-profile-router.service
 OLD_SERVICE=mike-ai-local-llm-router.service
 OLD_DIR=/opt/mike-ai/local-llm-router
 BACKUP_DIR=/opt/mike-ai/.backup-ai-profile-router-$(date +%Y%m%d-%H%M%S)
+VENV="$INSTALL_DIR/venv"
+MODEL_DIR=/opt/mike-ai/models/FLUX.2-klein-base-4B
+IMAGE_DIR="$INSTALL_DIR/images"
 
 echo "== AI Profile Router: Installation/Update =="
 
@@ -29,16 +32,47 @@ else
 fi
 
 # --- 2. Neue Dateien installieren -------------------------------------------
-mkdir -p "$INSTALL_DIR"
+mkdir -p "$INSTALL_DIR" "$IMAGE_DIR"
 install -m 0755 "$DIR/ai_profile_router.py" "$INSTALL_DIR/ai_profile_router.py"
+install -m 0755 "$DIR/image_worker.py" "$INSTALL_DIR/image_worker.py"
 install -m 0644 "$DIR/${SERVICE}" "/etc/systemd/system/${SERVICE}"
 
-# --- 3. Service aktivieren und starten ---------------------------------------
+# --- 3. Python-Venv mit Bild-Abhängigkeiten ---------------------------------
+if [ ! -x "$VENV/bin/python" ]; then
+  echo "-- Erstelle Python-Venv in $VENV"
+  python3 -m venv "$VENV"
+fi
+echo "-- Installiere/aktualisiere Bild-Abhängigkeiten (torch, diffusers, ...)"
+"$VENV/bin/pip" install --quiet --upgrade pip
+"$VENV/bin/pip" install --quiet \
+  torch \
+  diffusers \
+  transformers \
+  accelerate
+
+# --- 4. FLUX-Modell (nur wenn noch nicht vorhanden) --------------------------
+if [ -f "$MODEL_DIR/model_index.json" ]; then
+  echo "-- FLUX-Modell vorhanden: $MODEL_DIR"
+else
+  echo "-- Lade FLUX.2-klein-base-4B nach $MODEL_DIR (kann dauern)"
+  "$VENV/bin/python" - <<'PY'
+import os
+from huggingface_hub import snapshot_download
+snapshot_download(
+    repo_id="black-forest-labs/FLUX.2-klein-base-4B",
+    local_dir="/opt/mike-ai/models/FLUX.2-klein-base-4B",
+    local_dir_use_symlinks=False,
+)
+print("Modell-Download abgeschlossen")
+PY
+fi
+
+# --- 5. Service aktivieren und starten ---------------------------------------
 systemctl daemon-reload
 systemctl enable "$SERVICE"
 systemctl restart "$SERVICE"
 
-# --- 4. Verifikation ----------------------------------------------------------
+# --- 6. Verifikation ----------------------------------------------------------
 sleep 1
 if ! systemctl is-active --quiet "$SERVICE"; then
   echo "-- FEHLER: Service läuft nicht" >&2

deploy/mike-ai-profile-router.service

@@ -1,11 +1,11 @@
 [Unit]
-Description=Mike AI Profile Router (OpenAI-kompatibler Proxy, Port 8081)
+Description=Mike AI Profile Router (OpenAI-kompatibler Proxy + Bild-Orchestrierung, Port 8081)
 After=network-online.target
 Wants=network-online.target
 
 [Service]
 Type=simple
-ExecStart=/usr/bin/python3 /opt/mike-ai/ai-profile-router/ai_profile_router.py
+ExecStart=/opt/mike-ai/ai-profile-router/venv/bin/python /opt/mike-ai/ai-profile-router/ai_profile_router.py
 Restart=on-failure
 RestartSec=3
 Environment=ROUTER_HOST=0.0.0.0
@@ -15,6 +15,13 @@ Environment=PROFILE_SCRIPT=/usr/local/bin/llama-profile
 Environment=PROFILE_DIR=/etc/systemd/system/mike-ai-llama-ui.service.d
 Environment=SWITCH_TIMEOUT=600
 Environment=REQUEST_TIMEOUT=600
+Environment=LLAMA_SERVICE=mike-ai-llama-ui.service
+Environment=IMAGE_WORKER=/opt/mike-ai/ai-profile-router/image_worker.py
+Environment=IMAGE_PYTHON=/opt/mike-ai/ai-profile-router/venv/bin/python
+Environment=IMAGE_DIR=/opt/mike-ai/ai-profile-router/images
+Environment=IMAGE_WORKER_LOG=/opt/mike-ai/ai-profile-router/worker.log
+Environment=IMAGE_GEN_TIMEOUT=600
+Environment=IMAGE_VRAM_FREE_TIMEOUT=120
 NoNewPrivileges=true
 PrivateTmp=true