Are we all sure that 87% of data science projects fail?

Presumably every Data Scientist, Machine Learning Engineer or any other person involved in Data Science domain have already heard about MLOps. While it is still relatively new term, more and more people are getting interested in what it is and how to apply MLOps practices and tools in their projects.

I bet you can notice that interest too. MLOps Community is still growing so dynamically (and I’m really glad to be a part of it). MLOps topic and articles are now included in almost every Machine Learning conference. Recently even Andrew Ng and DeepLearning.ai …


Guide to improving your team and project as a newcomer

Figure 1. Hey, hi, hello | Source: UnDraw

People change industries they work in, they change companies but they also simply change teams — either within the same company or when starting work for a new employer. This is a big moment not only for the newbie but also for whole team.

Hey, hi, hello

Joining a new team is a big thing. It can be extremely stressful for the one that has just arrived. New colleagues (often from different countries), new organizational rules and culture, new project (sometimes even with entirely new tech stack).

Let me be the example for this story: I recently became an MLOps Engineer after having…


Deploy and destroy Kubeflow on EKS with one script, no sweat

Figure 1. Kubeflow Dashboard (Source: Kubeflow docs)

Tools, libraries, frameworks are created to make our work easier. They introduce new functionalities, simplify code, reduce boilerplate, automate stuff.

Imagine your project with no dependencies, imagine you need to replace a single function call (e.g. yaml.safe_load) with your own piece of code for that functionality. All these tools make applications easier to build and maintain, develop and deploy. But what if these tools themselves are difficult to be deployed? Ouch.

Introducing Kubeflow

I won’t lie — there’s no coincidence that I introduce Kubeflow right after writing that tools can be difficult to be deployed. …


Forming tech communities and getting into MLOps

Photo by Marvin Meyer on Unsplash

Whenever a new technology, programming language or any other area of interest comes into light and becomes more and more popular, a new group of interested people is formed. Whether it was Blockchain, Machine Learning or even earlier — Internet of Things, it is inevitable that people will start grouping together to learn and exchange observations.

Tech People Have It So Easy

We, computer scientists or tech people in general have it easier. We live in 2021 and the Internet is the only reasonable option to build a community which spans across all countries and allows people to discuss, present their ideas and learn together. …


Comparison of low-cost services for running ML experiments

Photo by Caspar Camille Rubin on Unsplash

I have been using Google Colab for my machine learning experiments for a very long time. Recently, when Colab turned out to be insufficient, I was forced to look for an alternative and I picked Paperspace Gradient.

How are they different in terms of costs, (GPU) power and other features?

Google Colab

I wrote about using Colab for my projects in my previous post. And I still consider it to be one of the best tools for both machine learning beginners as well as experts. Everything seems so trivial there: set up the environment, upload data (e.g. …


Finding good LR for your neural nets using PyTorch Lightning

Finding LR for your neural networks with PyTorch Lightning (Image by Author)

Among all the hyper-parameters used in machine learning algorithms, the learning rate is probably the very first one you learn about. Most likely it is also the first one that you start playing with. There’s a chance to find the optimal value with a bit of hyper parameter optimization, but this requires lots of experimenting. So why not let your tools do that (faster)?

Understanding Learning Rate

The shortest explanation of what learning rate really is that it controls how fast your network (or any algorithm) learns. …


Organize work, track and visualize experiments with Neptune.ai

Neptune.ai dashboard and metrics visualization (Image by Author)

Whenever I start a new project that will require running a hundreds of experiments, integration with experiment tracking tools is the very first thing I add to my code. Such tools can do much more than metrics visualization and storing hyperparameters of each run. In this post I introduce and describe Neptune tool which I used for my recent projects.

Experiment “management” tools

Tools for experiment tracking give you a huge boost for your machine learning (and not only) projects. You can use them to track hyperparameters, visualize graphs (e.g. …

Mateusz Kwaśniak

Software Engineer, Machine Learning, MLOps // Check About page for social links

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store