Short Project aimed to showcase how effortlessly one can integrate wandb into pre-existing accelerator based workflows (both GPU and TPU) using TensorBoard