Using the viewer function of nre-ga:26.02.00 to view the usdz file issues

When I used the viewer function of nre-ga:26.02.00 to view the usdz file I downloaded from the official nurec dataset( nvidia/PhysicalAI-Autonomous-Vehicles-NuRec at main ), I encountered the ValueError: string is not a file: /app/internal/scripts/pycena/runtime/pycena_run.runfiles/nv_assets/nv-car-no-lidar-diy.glb issue.

When I execute the following command:
docker run
–shm-size=8g
–rm
–gpus all
–volume “/data2/xql/liqy/usdz:/artifact:ro”
–publish “9190:9190”
nvcr.io/nvidia/nre/nre-ga:26.02.00
viewer
–artifact-path “/artifact/02e075b9-fd24-426b-971b-7cfcb2074cc9.usdz”
–host “0.0.0.0”
–port “9190”
–enable-nrend

Command line log:
[2026-05-06 05:57:12,007][INFO] Upgrading config from 26.1.112 to 26.2.158…
[2026-05-06 05:57:12,007][INFO] no changes from 26.1.112 to 26.2.80
[2026-05-06 05:57:12,007][INFO] converting from 26.2.80 to 26.2.81
[2026-05-06 05:57:12,648][INFO] no changes from 26.2.81 to 26.2.158 (target)
[2026-05-06 05:57:12,693][INFO] Using renderer: 3dgut-nrend
[2026-05-06 05:57:12,860][INFO] World Size: 1
[2026-05-06 05:57:12,860][INFO] Node Count: 1
[2026-05-06 05:57:12,860][INFO] Device Count per Node: 1
[2026-05-06 05:57:12,860][INFO] Log every 50 steps
[2026-05-06 05:57:12,896][INFO] GaussiansComposite/init
[2026-05-06 05:57:13,262][INFO] BaseCalib: start_global_step=250
[2026-05-06 05:57:13,262][INFO] FreePoseCalib: calib ON for cameras
[2026-05-06 05:57:13,262][INFO] FreePoseCalib: calib OFF for lidars
[2026-05-06 05:57:13,262][INFO] FreePoseCalib: enable_torch_compile=True
[2026-05-06 05:57:13,263][INFO] SkyEnvMapBackground: min_grad_updates=1000
[2026-05-06 05:57:13,284][INFO] LayerTrackIds: Initializing track filtering from 90 available tracks
[2026-05-06 05:57:13,293][INFO] SHGaussianModel/progressive_training: increase_frequency=1000 increase_step=1
[2026-05-06 05:57:13,293][INFO] SHGaussianModel/time_embed: Fourier features enabled with dim=5
[2026-05-06 05:57:13,298][INFO] LayerTrackIds: Initializing track filtering from 90 available tracks
[2026-05-06 05:57:13,306][INFO] SHGaussianModel/progressive_training: increase_frequency=1000 increase_step=1
[2026-05-06 05:57:13,315][INFO] LayerTrackIds: Initializing track filtering from 90 available tracks
[2026-05-06 05:57:13,407][INFO] SHGaussianModel/progressive_training: increase_frequency=1000 increase_step=1
[2026-05-06 05:57:13,407][INFO] SHGaussianModel/time_embed: Fourier features enabled with dim=20
[2026-05-06 05:57:13,415][INFO] BaseTracksCalib: start_global_step=500
[2026-05-06 05:57:13,428][INFO] LayerTrackIds: Initializing track filtering from 90 available tracks
[2026-05-06 05:57:13,433][INFO] call/n_frames_per_camera: [594, 594, 594, 594, 594, 594]
[2026-05-06 05:57:13,433][INFO] call/n_frames_per_lidar:
[2026-05-06 05:57:13,435][INFO] MCMCStrategy/relocate:
[2026-05-06 05:57:13,435][INFO] |─start_iteration=500
[2026-05-06 05:57:13,435][INFO] |─end_iteration=35000
[2026-05-06 05:57:13,435][INFO] |─frequency=100
[2026-05-06 05:57:13,435][INFO] └─max_invisible_steps=disabled
[2026-05-06 05:57:13,435][INFO] MCMCStrategy/add:
[2026-05-06 05:57:13,435][INFO] |─start_iteration=20000
[2026-05-06 05:57:13,435][INFO] |─end_iteration=35000
[2026-05-06 05:57:13,435][INFO] |─frequency=100
[2026-05-06 05:57:13,435][INFO] └─max_n_gaussians=2500000
[2026-05-06 05:57:13,435][INFO] MCMCStrategy/perturb:
[2026-05-06 05:57:13,435][INFO] |─start_iteration=1
[2026-05-06 05:57:13,435][INFO] |─end_iteration=10000000
[2026-05-06 05:57:13,435][INFO] |─frequency=1
[2026-05-06 05:57:13,435][INFO] └─noise_lr=5000.0
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.positions> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.scales> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.rotations> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.densities> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.features_albedo> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.features_specular> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.extra_signal> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.camera_extra_signal> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.lidar_extra_signal> in the state_dict.
[2026-05-06 05:57:13,445][INFO] Gaussian3DNRenderer: enable_ray_based_culling: True
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.positions> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.scales> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.rotations> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.densities> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.features_albedo> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.features_specular> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.extra_signal> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.camera_extra_signal> in the state_dict.
[NuRec::NRend][WARNING] ::: NRESHGaussianModel : missing parameters tensor <.lidar_extra_signal> in the state_dict.
[2026-05-06 05:57:13,453][INFO] Gaussian3DNRenderer: enable_ray_based_culling: True
[2026-05-06 05:57:15,792][INFO] Upgrading config from 26.1.112 to 26.2.158…
[2026-05-06 05:57:15,793][INFO] no changes from 26.1.112 to 26.2.140
[2026-05-06 05:57:15,793][INFO] converting from 26.2.140 to 26.2.141
[2026-05-06 05:57:15,793][INFO] Adding missing gaussians_strategy.invisible_steps.* buffers for MCMC checkpoints
[2026-05-06 05:57:15,793][INFO] adding gaussians_strategy.invisible_steps.background with shape (2106751,)
[2026-05-06 05:57:15,796][INFO] adding gaussians_strategy.invisible_steps.road with shape (153876,)
[2026-05-06 05:57:15,796][INFO] adding gaussians_strategy.invisible_steps.dynamic_rigids with shape (239373,)
[2026-05-06 05:57:15,796][WARNING] skipping layer dynamic_deformables: positions tensor not found
[2026-05-06 05:57:15,796][INFO] Assuming no model changes from version 26.2.141 to 26.2.158
[NuRec::NRend][ERROR] ::: NRERigidSHGaussianModel : missing extra state <.gaussians_nodes.dynamic_deformables._extra_state> in the state_dict.
[NuRec::NRend][ERROR] ::: NREDeformableSHGaussianModel : missing extra state <.gaussians_nodes.dynamic_deformables._extra_state> in the state_dict.
[2026-05-06 05:57:19,767][INFO] server listening on 0.0.0.0:9190
╭──────────────── viser ────────────────╮
│ ╷ │
│ HTTP │ http://localhost:9190
│ Websocket │ ws://localhost:9190 │
│ ╵ │
╰───────────────────────────────────────╯
[2026-05-06 05:57:20,281][INFO] connection open
(viser) Connection opened (0, 1 total), 5 persistent messages
(viser) Connection closed (0, 0 total)
[2026-05-06 05:57:23,292][INFO] connection open
(viser) Connection opened (1, 1 total), 5 persistent messages
Exception in thread RenderStateMachine(client.client_id=1:
Traceback (most recent call last):
File “/app/internal/scripts/pycena/runtime/pycena_run.runfiles/rules_python++python+python_3_11_x86_64-unknown-linux-gnu/lib/python3.11/threading.py”, line 1045, in _bootstrap_inner
self.run()
File “nre/viewer/”, line 0, in run
File “nre/viewer/”, line 0, in _reload_scene
File “nre/viewer/”, line 0, in _build_gui
File “nre/viewer/”, line 0, in _recreate_time_slider
File “nre/viewer/”, line 0, in _on_time_updated
File “nre/viewer/”, line 0, in _adjust_viewpoint
File “nre/viewer/”, line 0, in _update_ego
File “/app/internal/scripts/pycena/runtime/pycena_run.runfiles/.pycena_obfuscated.venv/lib/python3.11/site-packages/trimesh/exchange/load.py”, line 108, in load
) = parse_file_args(file_obj=file_obj, file_type=file_type, resolver=resolver)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “/app/internal/scripts/pycena/runtime/pycena_run.runfiles/.pycena_obfuscated.venv/lib/python3.11/site-packages/trimesh/exchange/load.py”, line 587, in parse_file_args
raise ValueError(f"string is not a file: {file_obj}")
ValueError: string is not a file: /app/internal/scripts/pycena/runtime/pycena_run.runfiles/nv_assets/nv-car-no-lidar-diy.glb

I would like to take a look at the scene effect rendered by USDZ. Is it used in this way?
Additionally, I attempted to directly open the USDZ file in IsaacSim, but it also failed to open.

Thank you for your post. I will reach out to a NuRec expert and have them get back to you.

The ValueError: string is not a file: …/nv_assets/nv-car-no-lidar-diy.glb is a packaging bug in nre-ga:26.02.00 — that ego-car GLB was not included in the image, so the viewer’s _update_ego crashes the moment a client connects. Your USDZ and command are both fine.

You might try render command to take a look at the scene.

docker run --rm --shm-size=8g --gpus all
-v /home/ubuntu/repro/usdz:/artifact:ro
-v /home/ubuntu/repro/render_out:/output
nvcr.io/nvidia/nre/nre-ga:26.02.00 render
–artifact-path /artifact/02e075b9-fd24-426b-971b-7cfcb2074cc9.usdz
–output-dir /output
–camera-id camera_front_wide_120fov
–image-scale 0.5

then encoded to MP4 with ffmpeg (H.264, yuv420p, CRF 20, faststart):

ffmpeg -framerate 30 -i ‘…/%06d.png’ -c:v libx264 -pix_fmt yuv420p -crf 20
-movflags +faststart output.mp4

Thank you very much for your reply. By the way, approximately when this issue will be fixed?