sam

sam

CONTAINERS IMAGES RUN BUILD

• Segment Anything from https://github.com/facebookresearch/segment-anything

The sam container has a default run command to launch Jupyter Lab with notebook directory to be /opt/

Use your web browser to access http://HOSTNAME:8888

How to run Jupyter notebooks

Once you are on Jupyter Lab site, navigate to notebooks directory.

Automatic Mask Generator Example notebook

Open automatic_mask_generator_example.ipynb.

Create a cell below the 4th cell, with only the following line and execute.

!wget https://dl.fbaipublicfiles.com/segment_anything/sam_vit_h_4b8939.pth

Then, start executing the following cells (cells below Set-up)

Predictor Example notebook

Open predictor_example.ipynb.

Make sure you have sam_vit_h_4b8939.pth checkpoint file saved under notebooks directory.

Then, start executing the cells below Set-up.

Benchmark script

You can run the following command to run a benchmark script.

python3 benchmark.py --save sam.csv

Or for full options:

python3 benchmark.py \
  --images https://raw.githubusercontent.com/facebookresearch/segment-anything/main/notebooks/images/dog.jpg  https://raw.githubusercontent.com/facebookresearch/segment-anything/main/notebooks/images/groceries.jpg \
  --runs=1 --warmup=0 \
  --save sam.csv

Outputs are:

• sam_benchmark_output.jpg :
• sam.csv (optional) :

CONTAINERS

sam
Builds
Requires	L4T ['>=34.1.0']
Dependencies	build-essential cuda cudnn python numpy cmake onnx pytorch:2.2 torchvision tensorrt onnxruntime opencv rust jupyterlab
Dependants	efficientvit tam
Dockerfile	Dockerfile
Images	dustynv/sam:r35.2.1 (2023-11-05, 6.1GB) dustynv/sam:r35.3.1 (2024-03-07, 6.1GB) dustynv/sam:r35.4.1 (2024-01-13, 6.1GB) dustynv/sam:r36.2.0 (2024-03-07, 7.9GB)

CONTAINER IMAGES

Repository/Tag	Date	Arch	Size
dustynv/sam:r35.2.1	2023-11-05	arm64	6.1GB
dustynv/sam:r35.3.1	2024-03-07	arm64	6.1GB
dustynv/sam:r35.4.1	2024-01-13	arm64	6.1GB
dustynv/sam:r36.2.0	2024-03-07	arm64	7.9GB

_{Container images are compatible with other minor versions of JetPack/L4T:
    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)
    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)}

RUN CONTAINER

To start the container, you can use jetson-containers run and autotag, or manually put together a docker run command:

# automatically pull or build a compatible container image
jetson-containers run $(autotag sam)

# or explicitly specify one of the container images above
jetson-containers run dustynv/sam:r36.2.0

# or if using 'docker run' (specify image and mounts/ect)
sudo docker run --runtime nvidia -it --rm --network=host dustynv/sam:r36.2.0

^{jetson-containers run forwards arguments to docker run with some defaults added (like --runtime nvidia, mounts a /data cache, and detects devices)
autotag finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.}

To mount your own directories into the container, use the -v or --volume flags:

jetson-containers run -v /path/on/host:/path/in/container $(autotag sam)

To launch the container running a command, as opposed to an interactive shell:

jetson-containers run $(autotag sam) my_app --abc xyz

You can pass any options to it that you would to docker run, and it'll print out the full command that it constructs before executing it.

BUILD CONTAINER

If you use autotag as shown above, it'll ask to build the container for you if needed. To manually build it, first do the system setup, then run:

jetson-containers build sam

The dependencies from above will be built into the container, and it'll be tested during. Run it with --help for build options.