Skip to content

Launching Pixano

From a terminal: locally

You can start the Pixano app with the following command:

pixano your_library_directory/ your_media_directory/

You will then be provided with a URL to open in your browser to use the app.

From a terminal: S3 (Experimental)

Note that you can also connect to an S3 compatible storage by providing an S3 path instead of a local path to your library of datasets.

The following arguments have to be passed:

  • --models_dir: Path to your models.
  • --aws_endpoint: S3 endpoint URL, use 'AWS' if not provided.
  • --aws_region: S3 region name, not always required for private storages.
  • --aws_access_key: S3 AWS access key.
  • --aws_secret_key: S3 AWS secret key.

So the command becomes:

pixano s3://your_library_directory/ s3://your_media_directory/ \
--models_dir="your_local_onnx_models/"
--aws_endpoint="https://your-aws-endpoint.com" \
--aws_region="" \
--aws_access_key="your_access_key" \
--aws_secret_key="your_secret_key" \

From a notebook

If you are in a Jupyter or Google Colab notebook, you can start the app by running the following cells:

from pixano.app import App
app = App("your_library_directory/", "your_media_directory")

You can then use the apps directly from the notebook in another cell with:

app.display()

From Docker

To launch the app you have to mount a library directory, a media directory and a model directory (if you want to use a model).

Here is an example:

docker run -d \
    --name pixano \
    -p 8000:8000 \
    -v ./library:/app/library \
    -v ./media:/app/media \
    -v ./models:/app/models \
    pixano/pixano:stable