LittleB0sS/NavAnywhere

OpenNWM

A Generalist Navigation World Model
with Latent Action Pretraining

OpenSpace Lab

Learn navigation dynamics from action-free video. Adapt them to physical world.

News · Setup · Data · Models · Evaluation · Training · TODO List

News

• 2026-11-03 — The paper is available on arXiv.
• 2026-11-01 — Updated the paper figure and documentation. Code, models, and NavAnywhere data are available.

Environment Setup

Requires Linux, Conda, and an NVIDIA GPU compatible with CUDA 12.4. Run from the repository root; LAM and NWM use separate environments.

git clone https://github.com/OpenSpace-Lab/OpenNWM.git
cd OpenNWM
for module in lam nwm; do
  conda create -n "open-$module" python=3.10 -y
  conda run -n "open-$module" python -m pip install torch==2.6.0 torchvision==0.21.0 \
    --index-url https://download.pytorch.org/whl/cu124
  conda run -n "open-$module" python -m pip install -r "$module/requirements.txt"
done

conda run -n open-nwm python setup_paths.py \
  --data-root /data/navigation --work-root /data/opennwm --nav-profile 15 \
  --lam-python "$(conda run -n open-lam python -c 'import sys; print(sys.executable)')" \
  --nwm-python "$(conda run -n open-nwm python -c 'import sys; print(sys.executable)')"
source .env

Replace /data/navigation and /data/opennwm with your data and work directories. Run source .env in each new terminal. To reuse existing splits, add --split-root /path/to/splits to setup_paths.py.

Data Preparation

Download NavAnywhere from Hugging Face or ModelScope and extract it to $NWM_DATA_ROOT. The archives total 1.93 TB; extraction requires tar, zstd, and additional disk space.

Download and extract with ModelScope

"$NWM_PYTHON" -m pip install modelscope
conda run -n open-nwm ms-hub download LittleBoss/NavAnywhere --repo-type dataset \
  --local-dir "$OPENNWM_WORK_ROOT/downloads/NavAnywhere" --max-workers 8
(
  cd "$OPENNWM_WORK_ROOT/downloads/NavAnywhere" || exit 1
  mkdir -p "$NWM_DATA_ROOT"
  find NavAnywhere -path '*/packs/*.tar.zst' -print0 \
    | xargs -0 -r -I{} tar --use-compress-program=unzstd -xf '{}' -C "$NWM_DATA_ROOT"
  for first in LAM-Data-{recon,scand,sacson}-*.tar.zst.part-00000; do
    [ -f "$first" ] || continue
    cat "${first%.part-00000}.part-"* \
      | tar --use-compress-program=unzstd -xf - -C "$NWM_DATA_ROOT"
  done
)

The action-data archives retain their LAM-Data- filenames. Expected layout:

$NWM_DATA_ROOT/
├── recon/<trajectory>/{0.jpg,1.jpg,...,traj_data.pkl}
├── sacson/<trajectory>/{0.jpg,1.jpg,...,traj_data.pkl}
├── scand/<trajectory>/{0.jpg,1.jpg,...,traj_data.pkl}
├── tartan/<trajectory>/{0.jpg,1.jpg,...,traj_data.pkl}
└── NavAnywhere/<source>/<trajectory>/{0.jpg,1.jpg,...}

The release includes RECON, HuRoN (sacson), SCAND, and NavAnywhere. TartanDrive requires a separate download for world model post-training and default LAM evaluation; follow the NWM data instructions and place it in tartan/ (split name: tartan_drive). setup_paths.py prepares the fixed train/validation/test lists and NavAnywhere recipe.

Pretrained Models

Download from Hugging Face or ModelScope into $OPENNWM_WORK_ROOT/models/OpenNWM, including configs/.

Checkpoint	Use
lam-nav15-pixel-action-step60k.ckpt	Latent action extraction
opennwm-nav15-stage1-pretrain60k.pth.tar	World model pretraining checkpoint
opennwm-nav15-stage2-joint100k.pth.tar	Physical-action prediction and planning

Download and verify with ModelScope

"$NWM_PYTHON" -m pip install modelscope
conda run -n open-nwm ms-hub download LittleBoss/OpenNWM --repo-type model \
  --local-dir "$OPENNWM_WORK_ROOT/models/OpenNWM" --max-workers 4
(cd "$OPENNWM_WORK_ROOT/models/OpenNWM" && sha256sum -c SHA256SUMS)

NWM downloads SD-VAE on first use; set NWM_VAE_PATH in .env to use a local copy. DINOv3 is needed only for LAM variant 3.

Evaluation

After preparing test data, import the Stage 2 checkpoint and run prediction:

"$NWM_PYTHON" run.py nwm import-checkpoint \
  --checkpoint "$OPENNWM_WORK_ROOT/models/OpenNWM/opennwm-nav15-stage2-joint100k.pth.tar" \
  --config "$OPENNWM_WORK_ROOT/models/OpenNWM/configs/opennwm-nav15-stage2.yaml" \
  --run-dir "$OPENNWM_WORK_ROOT/runs/released-stage2"
"$NWM_PYTHON" run.py nwm infer \
  exp_dir="$OPENNWM_WORK_ROOT/runs/released-stage2" ckp=latest eval_type=time \
  prediction_dir="$OPENNWM_WORK_ROOT/eval/stage2/pred"

Metrics, rollouts, and planning: NWM evaluation. Latent action reconstruction: LAM evaluation.

Training

Follow LAM training → cache preparation → world model pretraining and adaptation. Use the NavAnywhere 15-source profile and matching LAM checkpoints for caches. For the released model, set LAM_PIXEL_ACTION_CHECKPOINT in .env to $OPENNWM_WORK_ROOT/models/OpenNWM/lam-nav15-pixel-action-step60k.ckpt.

TODO List

• Release code, data, and pretrained models.
• Publish the paper, project page, and demos.
• Provide training and evaluation guides.
• Support closed-loop navigation on Unitree Go2 (deployment plan).

Acknowledgements & Citation

We thank NWM, CompACT, DreamDojo, and the dataset authors.

If you use OpenNWM, please cite:

@misc{nie2026opennwmgeneralistnavigationworld,
      title={OpenNWM: A Generalist Navigation World Model with Latent Action Pretraining},
      author={Dujun Nie and Ruiqi Song and Zixuan Wang and Hanqing Wang and Yifan Han and Baiyong Ding and Wenzhao Zheng and Hangbin Wu and Long Chen},
      year={2026},
      eprint={2611.01844},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2611.01844},
}

NWM-derived components retain CC BY-NC 4.0; DreamDojo LAM components retain Apache 2.0. Datasets and external models retain their own licenses.

Dataset release details

NavAnywhere 快照

最新完整快照为 NavAnywhere/manifests/snapshot-v3.json，在 v2 基础上新增 The_Great_Outdoors。其包位于 NavAnywhere/The_Great_Outdoors/packs/，解压后为 NavAnywhere/The_Great_Outdoors/<trajectory>/<frame>.jpg。v2 在 v1 的 13 个数据集基础上新增 ego4d；第一次发布为 NavAnywhere/manifests/snapshot-v1.json。全局和数据集级 manifest 记录轨迹、帧、文件、大小、帧范围及 SHA256；NavAnywhere/SHA256SUMS 可直接校验。每个包压缩后严格小于 70,000,000,000 字节，归档内保持 NavAnywhere/<dataset>/<trajectory>/<frame>.jpg。后续版本只增加 pack 和新的 snapshot-vN.json，不会改写旧 pack。

cd /path/to/NavAnywhere
sha256sum -c NavAnywhere/SHA256SUMS
find NavAnywhere -path '*/packs/*.tar.zst' -print0 \
  | xargs -0 -I{} tar --use-compress-program=unzstd -xf "{}" -C /path/to/data

历史 action 分片

同一个 NNN.tar.zst 的所有 .part-MMMMM 必须按编号拼接。当前上传的分片多数只有 part-00000，仍使用下面的通用方式，避免把多个 NNN 错拼成一个归档：

cd /path/to/NavAnywhere
for first in LAM-Data-{recon,scand,sacson}-*.tar.zst.part-00000; do
  archive=${first%.part-00000}
  cat "${archive}.part-"* > "$archive"
  tar --use-compress-program=unzstd -xf "$archive" -C /path/to/data
done

训练只读取 splits/{recon,scand,sacson}/train/traj_names.txt。测试清单一并发布用于评估；哈希和条目数见 splits/manifest.json 与 splits/SHA256SUMS。

来源与许可

本聚合仓库元数据使用 other，不把上游数据重新许可为 Apache。使用者必须分别遵守原始条款、引用要求、隐私限制和下载页面的最新说明；本仓库不授予超出上游条款的权利。

数据源	原始来源 / 条款入口
BotanicGarden	官方仓库，官方说明为学术用途
CASIA-Nav	发布者提供的 CASIA-Nav 数据；以随附或原始发布条款为准，本仓库不追加许可
CityWalker	CVPR 2025 论文 与项目发布条款
DL3DV-10K	官方 Terms of Use
Ego4D	官方数据页与原始数据使用条款
EgoWalk	论文 与原始发布条款
KrishnaCam	项目论文 与原始发布条款
LAVN	数据集卡
ROVER	官方数据概览
RealEstate10K	官方仓库 与其数据使用说明
SANPO	Google Research 介绍 与下载条款
The Great Outdoors	数据提供方的原始发布条款；本仓库不追加许可
Walking Tours	论文；原视频的权利仍归各自来源方
i2Nav-Robot	官方仓库
uB-VisioGeoloc	论文与数据说明
RECON	官方数据页；页面列明 MIT 数据许可及署名要求
SCAND	官方论文 与原始下载条款
SACSoN / HuRoN	官方仓库；遵守其中软件和数据声明