colab-jupyter-server 0.7.0

Colab/Kaggle as Jupyter a server and kernel

Colab Jupyter Server

I created this library because I want to use Colab/Kaggle as a remote Jupyter server, which I can connect to as a kernel for my local Jupyter notebook in VS Code.

How to Run

First, install the library:

pip install -U colab-jupyter-server

Then, run this command:

colab_jupyter_server \
--ngrok_authtoken=<YOUR_NGROK_AUTHTOKEN> \
--domain=<YOUR_DOMAIN>

Expected Last Output

Waiting for Jupyter server URL... (10s)
Jupyter server URL: 
<YOUR_DOMAIN>?token=<JUPYTER_SERVER_TOKEN>

Or, if domain is not set:

Waiting for Jupyter server URL... (10s)
Jupyter server URL: 
https://xy12-34-567-890-123.ngrok-free.app?token=<JUPYTER_SERVER_TOKEN>

Command Parameters

• ngrok_authtoken: Your ngrok authtoken. You can get it here: https://dashboard.ngrok.com/get-started/your-authtoken

• ngrok_down_url: The default download URL is for ngrok on Linux (x86-64) (https://bin.equinox.io/c/bNyj1mQVY4c/ngrok-v3-stable-linux-amd64.tgz). Find other download URLs here: https://ngrok.com/download

• domain: Your custom domain URL, which will be used as the Jupyter server URL. If not set, ngrok will generate a random URL each time the server starts. To keep the server URL consistent at a low cost, you can create and use a free domain from ngrok here: https://dashboard.ngrok.com/domains

• port: The default port is 8889. Avoid using port 8888, as it is already in use by the notebook you open in Colab/Kaggle.

• wait_time: The duration time to wait for the Jupyter server URL to be retrieved. The default is 10s. A shorter wait time may result in failure to retrieve the Jupyter server token, returning an empty token.