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 Reconsider My SAM 3 Access
Hello SAM 3 Team,
I am a Ph.D. student at Shandong University, and my Hugging Face username is yue2123.
I recently requested access to SAM 3 because I am working with MedSAM3 for medical image segmentation research. I have already set up its code and downloaded the released LoRA weights, but I then learned that these weights cannot run independently. MedSAM3 uses SAM 3 as its base model, so the official checkpoint is required to complete the setup and reproduce the method correctly.
My main interest is to study concept-based segmentation in medical images, particularly ultrasound images, and evaluate where the method works well and where it remains unreliable. This is academic research rather than a commercial project.
I understand that my earlier application may not have explained this background clearly enough. If possible, could you please reconsider my request or allow me to submit the access form again?
I will use the checkpoint only within our research environment, follow the applicable license and policies, and will not redistribute or publicly host the model weights.
Thank you for your time.
Best regards,
yue2123
Ph.D. Student, Shandong University