The project aims to provide a convenient method for testing AI-powered web apps before deployment using Google Colab Notebooks, without the need to upgrade hardware or utilize expensive cloud services.
The motivation behind this initiative stemmed from the need to create a straightforward approach for debugging and testing purposes without substantial cost or infrastructure updates.
The existing examples provided fulfill most of the project's requirements, providing a strong foundation for further development.
Utilizing Node.js, with a focus on the npm package "localtunnel," in conjunction with other npm packages such as http-server, enables the creation of a straightforward, zero-configuration command-line HTTP server for serving static files or frameworks like Django or Streamlit. This environment allows users to serve and share web applications globally from Google Colab, a Jupyter notebook environment hosted on the Google Cloud platform, with the added benefit of access to free GPUs for more advanced AI-powered web applications.
Navigate to Samples and select the notebook that aligns with your project type.
After opening any of the sample notebooks and following the outlined instructions, locate the line:
Ensure to copy the IP as depicted in the above screenshot.
Upon clicking on the APP link, you will encounter a preliminary page before being directed to the main page. To proceed to the main page, paste the previously copied Global IP address in the specified location:
Special thanks to FahimFBA for contributing to one of the example projects.
At present, no examples linked to Google Drive have been created, as the current samples primarily focus on basic scenarios.
Addition of a simple Python Django implementation example notebook in Colab.