CAD-Recode: Reverse Engineering CAD Code from Point Clouds

CAD-Recode is trained to transform point cloud to 3D CAD model as a valid Python code. CAD-Recode basically is Qwen2-1.5B + a single linear layer, which means that we share the original tokenizer of Qwen2. For minimal reproducible example please visit our github https://github.com/filaPro/cad-recode. And if you like it, give us a github ๐ŸŒŸ.

Citation

If you find this work useful for your research, please cite our paper:

@misc{rukhovich2024cadrecode,
      title={CAD-Recode: Reverse Engineering CAD Code from Point Clouds}, 
      author={Danila Rukhovich, Elona Dupont, Dimitrios Mallis, Kseniya Cherenkova, Anis Kacem, Djamila Aouada},
      year={2024},
      eprint={2412.14042},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2412.14042}, 
}
Downloads last month
178
Safetensors
Model size
1.54B params
Tensor type
F32
ยท
BF16
ยท
Inference API
Unable to determine this model's library. Check the docs .

Model tree for filapro/cad-recode

Base model

Qwen/Qwen2-1.5B
Finetuned
(30)
this model

Space using filapro/cad-recode 1