Containers at TACC
In the last section, we covered an overview of tools commonly used to build AI and ML environments and use each of these tools to install Pytorch on TACC’s supercomputer, Frontera. In this section we will focus on building dependencies into containers and leveraging these containers in Jupyter notebooks.
Well get started in the Advanced Container Build tutorial where we’ll walk through building a customized GPU-enabled container built to run a BERT classification model. Then in the Containerized Kernel tutorial we’ll be learning how to use container as the kernel (environments that run your code) for a Jupyter Notebook on Frontera.