Instructions to use datasolin/sam-3d-objects-mirror with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- SAM 3D Objects
How to use datasolin/sam-3d-objects-mirror with SAM 3D Objects:
from inference import Inference, load_image, load_single_mask from huggingface_hub import hf_hub_download path = hf_hub_download("datasolin/sam-3d-objects-mirror", "pipeline.yaml") inference = Inference(path, compile=False) image = load_image("path_to_image.png") mask = load_single_mask("path_to_mask.png", index=14) output = inference(image, mask) - Notebooks
- Google Colab
- Kaggle
SAM 3D Objects β mesh checkpoints (unofficial mirror)
This repository is an unmodified copy of part of Meta's
SAM 3D Objects checkpoints, taken from revision
2e73555018d2741ccd486e56c24fac41155a1dc6 of facebook/sam-3d-objects. Data Solin redistributes it under the
SAM License (copy in LICENSE), as that license allows. It is not Meta's official repository, and
Meta does not endorse or support it.
It holds only the files Solin CAD's image-to-3D feature needs to make a mesh: the pipeline's config, the geometry model and its decoder, the texture-and-refinement model, and the mesh decoder. It does not hold the Gaussian splat decoders or the encoders. The model's code is Meta's: https://github.com/facebookresearch/sam-3d-objects.
License and permitted use
Downloading or using these files means you agree to the SAM License. In particular:
- You may not use them, or let others use them, for any activity subject to the International Traffic in Arms Regulations (ITAR), or for end uses that trade controls prohibit. That includes military or warfare purposes, nuclear industries or applications, espionage, and the development or use of guns or illegal weapons.
- You must not be the target of trade controls (US, UN, EU or UK sanctions), and your use must comply with them.
- If you redistribute these files, or works derived from them, you must do so under the SAM License and include a copy of it.
- If you publish research done with them, you must acknowledge SAM 3D: SAM 3D: 3Dfy Anything in Images, SAM 3D Team et al., 2025 (arXiv:2511.16624).
The files are provided "as is", without warranty of any kind.
Files
Each file is byte-for-byte the same as in facebook/sam-3d-objects at the revision above, with the same SHA-256.
| File | Bytes | SHA-256 |
|---|---|---|
checkpoints/pipeline.yaml |
3,548 | 53c3d226b21df85c0bb3d16e6e4fa63abde0d6167525765eb929d02bfa9d358c |
checkpoints/ss_generator.yaml |
5,076 | 3c265448bca7c057f94e3ef56adea3a895a10bcd9f15f992a41dd03fa35412cd |
checkpoints/ss_generator.ckpt |
6,690,136,964 | 225f40479e4cff4f39d6fa14c55be3abad1475bf55b61af3bec1e19ed2f6c146 |
checkpoints/ss_decoder.yaml |
244 | baacff269b664f84f7aa1896ebd66128065f6568cdb88d419d5c3d2ccb4193ae |
checkpoints/ss_decoder.ckpt |
147,609,242 | 6dac1cd7b7fda5a38e0614fadae441f1794f80e39ea2981f1ac8aff0a7e99340 |
checkpoints/slat_generator.yaml |
1,986 | 53029fadff6fe34a0344381a16d64d65d0567d603484952712e8969319559c4e |
checkpoints/slat_generator.ckpt |
4,906,537,684 | 91529bde8e7daa12d09618a66c319e3a5a6398db6b23b958cedcb1c3f28faabb |
checkpoints/slat_decoder_mesh.yaml |
300 | 8f46952764aa985c50109a56f0b4f07625cdb06457aaded74957d26f3520c69e |
checkpoints/slat_decoder_mesh.ckpt |
363,726,862 | 85907b37b67d8ce5b099a96629bdcfbd873eb407dee6b3aa9a75deb15038db33 |
- Downloads last month
- 12