Run SAM 3

run_sam3.py

import torch
from transformers import Sam3Processor, Sam3Model
from PIL import Image
import requests

# Target your Nvidia GPU
device = "cuda" if torch.cuda.is_available() else "cpu"

# Automatically download and load SAM 3 to the GPU
print("Loading model...")
model = Sam3Model.from_pretrained("facebook/sam3").to(device)
processor = Sam3Processor.from_pretrained("facebook/sam3")

# Load a test image
url = "https://images.unsplash.com/photo-1518791841217-8f162f1e1131"
image = Image.open(requests.get(url, stream=True).raw).convert("RGB")

# Process the image with a text prompt
print("Running inference...")
inputs = processor(images=image, text="cat", return_tensors="pt").to(device)

# Generate the segmentation masks
with torch.no_grad():
    outputs = model(**inputs)

# Post-process the raw outputs into usable masks and bounding boxes
results = processor.post_process_instance_segmentation(
    outputs,
    threshold=0.5,
    mask_threshold=0.5,
    target_sizes=inputs.get("original_sizes").tolist()
)[0]

print(f"Success! Found {len(results['masks'])} objects matching the prompt.")