What are machine learning images ?

  • Machine learning images are pre-built environments that work out of the box, are supported by Genesis Cloud and suit all kinds of machine learning work. Our machine learning images are based on miniconda which makes expanding, maintaining, and customizing the environment easy.

Why use machine learning images ?

  • Matching GPU driver, CUDA, cuDNN, and the version of your chosen framework is required for it to function and can be time consuming due to the required research, hence we maintain a set images that help you to get start in a matter of minutes instead of spending hours on setting up the environment yourself.

Prerequisite: Install the NVIDIA® driver

  • In order to make use of GPUs, the NVIDIA® driver needs to be installed on the instance. This can be done by simply enabling the corresponding toggle when creating the instance.

Creating an instance based on a machine learning image

  • Login to Console Dashboard & click on 'Create New Instance' button 

  • Then, under 'Select Image' section choose one 'TensorFlow' or 'PyTorch' from 'Images' tab.

  • Simply select your preferred image, complete all other configuration and click 'Create instance'.

  • Alternatively, use our Public API: https://developers.genesiscloud.com/images

  • Additionally, it is possible to combine other pre-configured software from our Catalog such as Docker or JupyterLab with Images such as TensorFlow or PyTorch.

Enjoy training and keep accelerating!