Fix GPU worker config: disable GPU persistence, use cluster gateway

- STREAMING_GPU_PERSIST=0 until all primitives support GPU frames
- IPFS_GATEWAY_URL points to cluster's public gateway

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

docker-compose.yml

@@ -172,8 +172,10 @@ services:
       - ARTDAG_CLUSTER_KEY=${ARTDAG_CLUSTER_KEY:-}
       # GPU acceleration settings
       - NVIDIA_VISIBLE_DEVICES=all
-      # Keep frames on GPU between operations for maximum performance
-      - STREAMING_GPU_PERSIST=1
+      # GPU persistence disabled until all primitives support GPU frames
+      - STREAMING_GPU_PERSIST=0
+      # Use our public IPFS gateway for HLS segment URLs
+      - IPFS_GATEWAY_URL=https://celery-artdag.rose-ash.com/ipfs
     volumes:
       # Local cache - ephemeral, just for working files
       - gpu_cache:/data/cache