Instructions to use facebook/sam3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use facebook/sam3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("mask-generation", model="facebook/sam3")# Load model directly from transformers import AutoProcessor, AutoModel processor = AutoProcessor.from_pretrained("facebook/sam3") model = AutoModel.from_pretrained("facebook/sam3", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Request to review or reset my SAM 3 access application
Hello SAM 3 Team,
My Hugging Face username is cc60c. I am currently a PhD student at City University of Hong Kong.
I recently requested access to SAM 3 because I am studying and reproducing SceneSmith, an open-source system for the agentic generation of simulation-ready indoor scenes.
I have already set up the SceneSmith codebase and its main dependencies. However, the default asset-generation pipeline requires the official SAM 3 checkpoint. SceneSmith uses SAM 3 for object segmentation as part of the SAM 3D Objects pipeline before reconstructing separable 3D assets. Therefore, access to the official sam3.pt checkpoint is necessary for me to run and evaluate the released pipeline correctly.
My research interests include text-to-scene generation, 3D asset generation, physical scene construction, and the reliability of generated simulation environments. This work is for academic research and learning purposes only, not for commercial use.
I understand that my previous application may not have included sufficiently complete or accurate information. Could you please reconsider my access request, or reset its status to pending so that I can submit the access form again with the required information?
I will use the checkpoint only within our research environment, comply with all applicable license terms and policies, and will not redistribute or publicly host the model weights.
Thank you for your time and consideration.
Best regards,
PhD Student
City University of Hong Kong
Hugging Face username: cc60c