Google Colab offers free, easy access to Tesla K80 GPUs to anyone with a Google account.

We can run our projects on Google Colab instances with couple of manual steps, allowing us to automate our dependency setup, source and datasets download and remote Terminal, Jupyter, Tensorboard, Serving access. Below are the steps for this

  1. Login to your Clouderizer console and create new project or edit an existing project.
    You can also clone a new project from one of our many cool Community Templates like Fast.ai, Tensorflow Object Detection, Kaggle Competition, etc
  2. On Machine setup, select Local Machine option.
  1. Press Next and fill up details on Setup and Workspace tab as per your project requirements. Finish the wizard.
  2. On the last screen, you will see script to run this project on any Linux machine. Copy this string.
  1. Now login to your Google Colab console and create a new Python 3 notebook.
  1. Once new notebook is launched, from top menu, go to Runtime -> Change runtime type, and select GPU as Hardware accelerator and press save.
  1. Now go to the first code block on your colab sheet and type ! followed by Clouderizer project startup script from step above. Press Run button to execute the script

Don't forget to put ! before the startup command

This should trigger automated Clouderizer project deployment on this instance. You can now switch back to Clouderizer console and see the progress of your project startup from here.

Once project status changes to Running, all your project dependencies, source code, datasets and custom startup scripts will be setup and you can start working.

Did this answer your question?