Accessing and using the SDumont infrastructure for Deep Learning research. – Aug 30, 2021
This manual details the engagement, access and usage of the Santos Dumont Super Computer. This document is profoundly based on the official support manual provided by LNCC, as well as my personal experience in the program (hence, it may not perfectly represent all cases). Its goal is to present information in a more directed manner for users that share my own profile (Deep Learning researchers that prefer the TensorFlow framework and who are familiar with Docker.
DevOps MLThe basics on experimenting with sacred: executing, logging and reproducing – Oct 2, 2017
When studying machine learning, you'll often see yourself training multiple models, testing different features or tweaking parameters. While a tab sheet might help you, it's not perfect for keeping logs, results and code. Believe me, I tried. An alternative is Sacred. This package allows you to execute multiple experiments and record everything in an organized fashion.
DevOpsDigging Deep in Code Versioning with GIT – Oct 2, 2017
Some weeks ago, I've started this blog with this very subject. Today, as requested by a reader, I'll go back and try to talk a little bit more about versioning focusing on GIT, it's problems and how to avoid/fix them.
DevOpsThe basics project's dependency management using virtualenv – Oct 2, 2017
If you like Java or C#, you most likely have had contact with Maven or NuGet. These are both amazing utilities that can help you managing your projects and their dependencies. These tools will basically retrieve packages from a repository and install them somewhere, making them easily [imported|used|updated|removed]
(read this as a regular expression). They can also modify or create structures or files inside your project, providing you a starting point for doing something. There is much stuff here, check out their web-pages to see what I'm talking about.
The basics on GIT and code storage – Oct 2, 2017
Versioning is one of the most basic ideas in software development, but is treated with such contempt that lots of people only start to actually use it after too much manual effort. Someone somewhere at some point said: "the idea of a human doing a task that could be automated by a machine just makes me sad". I think this sentence sums up pretty much everything about versioning: you can develop without it, sure, but it would be just painful and, considering the formality of computers, it would just result in a worst job, most likely.
DevOps