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 SAM 3 Access for CaPtain Reproduction
Dear Meta team,
I am working on reproducing the paper “CaPtain: A Pipeline for Post-Training Code-as-Policy Robot Agents,” an academic research project on robotic manipulation. The authors’ official implementation uses SAM 3 for visual object segmentation, which is an important part of the perception pipeline.
My previous access request was rejected. I would be grateful if you could reconsider my application and grant me access to the official SAM 3 weights. I intend to use the model only for academic research and reproduction, and I will comply with the applicable license terms and Meta Privacy Policy.
Access to SAM 3 would help me reproduce the authors’ original experimental setup. Thank you for your time and consideration.
Best regards