README.md

metadata

license: cc-by-nc-4.0
pipeline_tag: image-to-3d
base_model: yyfz233/Pi3X
tags:
  - lux3d
  - rust
  - candle
  - safetensors
  - canonical-weights
  - 3d
  - image-to-3d
  - point-cloud
  - multiview
  - pi3x

Lux3D Pi3X Canonical Weights

This repository contains the canonicalized safetensors export of the Pi3X checkpoint used by the Lux3D runtime.

Model Details

• Model family: pi3x
• Canonical file: model.safetensors
• Tensor count: 1873
• Primary runtime target: Lux3D runtime
• Typical output: point cloud export (.ply)

Included Files

• model.safetensors
• resolved_config.json
• manifest.json
• checksums.json

Usage With Lux3D

The Lux3D CLI can validate and use the package once it is installed into your local model asset layout.

cargo run -p lux3d-cli -- inspect --repo-root <runtime-root> pi3x
cargo run -p lux3d-cli -- run --repo-root <runtime-root> pi3x --source <input-sequence> --conditions <conditions-file> --output <output-file.ply>
cargo run -p lux3d-cli -- run --repo-root <runtime-root> pi3x --source <input-video> --vo --chunk-size 8 --overlap 4 --conf-threshold 0.05 --inject-condition pose,depth,ray --output <output-file.ply>

Provenance

• Upstream source model: yyfz233/Pi3X
• Canonicalization flow: tools/python_baseline/normalize_weights.py

The exact source checksum set is recorded in manifest.json. The integrity of the uploaded package is recorded in checksums.json.

Intended Use

• Pi3X reconstruction workflows with Lux3D
• multiview point cloud generation
• VO-assisted Pi3X inference
• reproducible runtime validation

Limitations

• Verified Lux3D runtime inference currently assumes CUDA.
• This package contains canonical runtime artifacts only.
• The upstream Pi3X weights are non-commercial.

License

• Upstream model page: yyfz233/Pi3X
• License file in this repository: LICENSE

Review upstream terms before redistribution or commercial use.

Citation

@article{wang2025pi,
  title={$\pi^3$: Permutation-Equivariant Visual Geometry Learning},
  author={Wang, Yifan and Zhou, Jianjun and Zhu, Haoyi and Chang, Wenzheng and Zhou, Yang and Li, Zizun and Chen, Junyi and Pang, Jiangmiao and Shen, Chunhua and He, Tong},
  journal={arXiv preprint arXiv:2507.13347},
  year={2025}
}