A quick guide for tracking Pytorch Experiments using Neptune.ai

Neptune.ai is an online platform that manages your ML experiments, model registry and model monitoring. This tool doesn't limit to monitoring the performance of a single model, but it also helps to keep track of more models in a single dashboard. Each time you execute the code of your experiment, it will fill a row of the table, called the Runs table since it displays all the experiment runs. In this way, you can compare different models based on hyperparameters, metrics and other measures. Another important aspect is that you can store any type of metadata for each experiment, such as the hyperparameters, the metrics and learning curves, the images and HTML objects of the plots, tables, the files containing the trained model and much more.