Add dev infrastructure improvements

- Central config with logging on startup
- Hot reload support for GPU worker (docker-compose.gpu-dev.yml)
- Quick deploy script (scripts/gpu-dev-deploy.sh)
- GPU/CPU frame compatibility tests
- CI/CD pipeline for GPU worker (.gitea/workflows/gpu-worker.yml)
- Standardize GPU_PERSIST default to 0 across all modules

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

docker-compose.gpu-dev.yml

@@ -0,0 +1,36 @@
+# GPU Worker Development Override
+#
+# Usage: docker stack deploy -c docker-compose.yml -c docker-compose.gpu-dev.yml celery
+# Or for quick testing: docker-compose -f docker-compose.yml -f docker-compose.gpu-dev.yml up l1-gpu-worker
+#
+# Features:
+# - Mounts source code for instant changes (no rebuild needed)
+# - Uses watchmedo for auto-reload on file changes
+# - Shows config on startup
+
+version: '3.8'
+
+services:
+  l1-gpu-worker:
+    # Override command to use watchmedo for auto-reload
+    command: >
+      sh -c "
+        pip install -q watchdog[watchmedo] 2>/dev/null || true;
+        echo '=== GPU WORKER DEV MODE ===';
+        echo 'Source mounted - changes take effect on restart';
+        echo 'Auto-reload enabled via watchmedo';
+        env | grep -E 'STREAMING_GPU|IPFS_GATEWAY|REDIS|DATABASE' | sort;
+        echo '===========================';
+        watchmedo auto-restart --directory=/app --pattern='*.py' --recursive -- \
+          celery -A celery_app worker --loglevel=info -E -Q gpu,celery
+      "
+    environment:
+      # Development defaults (can override with .env)
+      - STREAMING_GPU_PERSIST=0
+      - IPFS_GATEWAY_URL=https://celery-artdag.rose-ash.com/ipfs
+      - SHOW_CONFIG=1
+    volumes:
+      # Mount source code for hot reload
+      - ./:/app:ro
+      # Keep cache local
+      - gpu_cache:/data/cache