The latest neurec container- nvcr.io/nvidia/nre/nre-ga:latest has a bug in rendering when even launching the gRPC server with the following command-
docker run \
–shm-size=64g –env CUDA_VISIBLE_DEVICES=0 –name NuRec_manual –gpus all –privileged –rm –net=host
/path/to/output/folder:/workdir/output \
nvcr.io/nvidia/nre/nre-ga:latest \
serve-grpc \
–artifact-glob “NeuralSim/Neurec_dataset/048b974e-1546-488a-b8f9-d32bff77f5aa/048b974e-1546-488a-b8f9-d32bff77f5aa.usdz” \
–port=46435 –host=localhost –test-scenes-are-valid –enable-editing-actors
[2026-04-03 23:16:20,772][ERROR] Failed to render camera camera_front_wide_120fov_1: INVALID_ARGUMENT: Got render request with DynamicObject updates but actor editing is disabled on the server. Restart serve-grpc with --enable-editing-actors.
[2026-04-03 23:16:20,772][ERROR] Failed to render camera camera_front_wide_120fov_2: INVALID_ARGUMENT: Got render request with DynamicObject updates but actor editing is disabled on the server. Restart serve-grpc with --enable-editing-actors.
[2026-04-03 23:16:20,772][ERROR] Failed to render camera camera_front_wide_120fov_3: INVALID_ARGUMENT: Got render request with DynamicObject updates but actor editing is disabled on the server. Restart serve-grpc with --enable-editing-actors.
[2026-04-03 23:16:20,826][INFO] batch_render_rgb: 4 cameras
Even when the serve-grpc with –enable-editing-actors are enabled during the server launch.
Am I missing something as a prerequisite ? Or this is an actual bug?
The data set used is nvidia/PhysicalAI-Autonomous-Vehicles-NuRec at 26.02