Awesome! Now that I know how to use Colab on Zindi, what exactly can Colab do that my local machine/environment can’t? Well, each Colab runtime comes with at least 12GB of RAM and up to 25GB of hard disk space to upload your files and save your output. For models that require GPU acceleration, such as the famous deep learning models, Colab offers a GPU runtime that comes with up to 12GB of video memory(VRAM). An interesting and perhaps more useful feature is Colab’s ability to connect directly to your Google Drive and access, as well as save, files such as the data used to train your models. This feature also allows you to work collaboratively with your teammates and ensures your files are backed up and accessible from anywhere as you continue working on your model. Here’s a video showing how you can mount your google drive onto Colab and access files uploaded and saved in the Drive.
Hide player controls
Hide resume playing