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")# pip install -U transformers accelerate # 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 for review of rejected SAM3 access request – Adam0492
Hello SAM 3 team,
My name is Yihao Shen, and I’m affiliated with The Chinese University of Hong Kong, Shenzhen. My Hugging Face username is Adam0492. My November 15 access request for the SAM3 collection is marked as rejected, and the model page does not show a reason. My university email is verified and set as my primary email on Hugging Face.
I’m requesting access to reproduce HexaFuture’s HexaShell experiments on the RoboCasa365 robotics benchmark. I plan to use SAM 3 to segment objects in simulated kitchen scenes and combine those masks with camera depth data to estimate where the objects are. This would let me evaluate the robot’s manipulation performance using the perception method described in the HexaShell report.
Could you please review my request? If any information needs to be corrected, I’d appreciate it if you could let me know which field to update and reset the request so I can resubmit it.
Thank you,
Yihao Shen