WNM-3D
WNM-3D is a generative world navigation model for continuous
vision-language navigation. It converts monocular egocentric RGB history into
persistent geometry-aware scene tokens and jointly generates future views and
navigation actions for closed-loop control.
License notice: The released checkpoints embed VGGT-Ω/DINOv3-derived
parameters. Use and redistribution are subject to the FAIR Noncommercial
Research License, the DINOv3 License, and other applicable third-party terms.
Released Checkpoints
| Directory | Training stage | Recommended use |
|---|
wnm_3d_stage1_release | Offline A* SFT | Stage-I analysis and initialization |
wnm_3d_stage2_release | Closed-loop DAgger-SFT | Stage-II analysis and initialization |
wnm_3d_stage3_release | Counterfactual DanceGRPO | Evaluation and inference |
The Stage-III checkpoint is the primary released policy. Each directory is a
self-contained inference checkpoint; standalone Wan, UMT5-XXL, and VGGT-Ω
initialization weights are not required for evaluation.
Download
1hf download TeleEmbodied/WNM-3D \
2 --include "wnm_3d_stage3_release/**" \
3 --local-dir checkpoints
Inference
Install WNM-3D by following the
installation guide,
then launch a single policy replica:
1bash scripts/inference/wnm_3d_server.sh \
2 --model-path checkpoints/wnm_3d_stage3_release \
3 --cuda-devices 0 \
4 --num-replicas 1 \
5 --base-port 8000
Run the GN-Bench client from a sibling GN0 checkout:
1cd ../GN0
2bash scripts/evaluation/eval_remote.sh \
3 --exp-config configs/gn_bench/interiorgs/test_unseen.yaml \
4 --enable-stall-recovery \
5 --num-gpus 1 \
6 --result-dir tmp/eval/wnm_3d_test_unseen
The reference single-replica deployment was validated on an NVIDIA H100 80 GB
GPU and used approximately 27 GiB of GPU memory after loading. A released
checkpoint occupies approximately 26 GiB. These figures are observations from
the reference configuration, not strict minimum requirements.
Intended Use and Limitations
WNM-3D is intended for research on embodied navigation, world models, and
closed-loop vision-language navigation. The released policy was developed and
evaluated with the InteriorGS scenes and GN-Matrix task annotations used by
GN-Bench. Performance may not transfer to new simulators, sensors, scene
distributions, languages, or physical robots without additional validation.
Generated actions can fail or behave unexpectedly. Do not use the model as a
safety-critical controller or deploy it around people, property, or physical
systems without appropriate safeguards and human oversight.
License
The released checkpoints contain original WNM-3D parameters together with
VGGT-Ω/DINOv3-derived parameters. They are not licensed solely under
Apache-2.0. Use and redistribution are subject to all applicable terms,
including the FAIR Noncommercial Research License and DINOv3 License.
The model repository provides the complete
Apache-2.0,
FAIR Noncommercial Research,
and
DINOv3
license texts, together with its
Third-Party Notices.
The source repository documents the complete
component-level license boundaries.
Datasets and separately downloaded initialization weights remain subject to
their providers' terms.
Citation
1@article{huang2026wnm_3d,
2 title={WNM-3D: A World Navigation Model with 3D Scene Conditioning for Closed-Loop VLN},
3 author={Huang, Yuehao and Wu, Yunzi and Zhang, Xiaotao and Li, Xinhai and Dong, Jiankun and Lv, Jiajun and Zhang, Chi and Bai, Chenjia and Liu, Yong and Li, Xuelong},
4 journal={arXiv preprint arXiv:2608.07267},
5 year={2026}
6}
