Published in Towards Data Science·3 days agoMember-only5 Best Python Synthetic Data Generators And How to Use Them When You Lack DataLet's get even more data — In 2021, 2.5 quintillion bytes (2.5 million terabytes) of data were produced daily. Today, it is even more. But apparently, that's not enough because the Python ecosystem has many libraries to produce synthetic data. Maybe some of them are created just for the sake of being able to generate synthetic…Data Science8 min readData Science8 min read
Published in Towards AI·Jan 19Member-onlyBentoml vs. Fastapi: The Best ML Model Deployment Framework and Why It's BentomlA detailed comparison between BentoML and FastAPI for machine learning model deployment. — Motivation What am I even talking about? According to StackOverflow 2022 Developer Survey, FastAPI just became one of the most loved web frameworks in the world, far above its competitors like Django and Flask. How can we compare it to BentoML, a framework that wasn't even included in the survey? Well…Machine Learning9 min readMachine Learning9 min read
Published in Towards AI·Jan 19Member-onlyHow To Create Highly-Organized ML Projects Anyone Can Reproduce With DVC PipelinesWhat is a machine learning pipeline? Imagine a machine learning pipeline as an unbreakable tunnel that stretches from one end of a mountain to the other. At the tunnel entrance is a massive avalanche of raw data, tumbling and cascading from all directions. As the data enters the tunnel, it is cleaned, preprocessed, transformed, and feature-selected…Machine Learning11 min readMachine Learning11 min read
Published in Towards AI·Jan 17Member-onlyHow to Track ML Experiments With DVC Inside VSCode To Boost Your ProductivityManage ML experiments like a pro — Why track experiments in machine learning? Keeping track of machine learning experiments is like keeping FIVE dogs in a bathtub. Without help, at least FOUR of them are bound to slip out of your hands and ruin everything. A total disaster is what’s going to happen if you don’t have a proper experiment management system. First…Machine Learning9 min readMachine Learning9 min read
Jan 12Member-only10 Secret Jupyter Lab Tricks I Wish I Was Given on My BirthdayBut found out the hard way — No, these tricks aren't those fancy ones you learn about to brag to others but never use. These tricks are the ones you brag to others AND use every single day. Jupyter Lab is great. Let's make it awesome! 1. Spellchecker I love my Grammarly Premium, but I like to keep my…Machine Learning6 min readMachine Learning6 min read
Published in Towards Data Science·Jan 9Member-only6 New Booming Data Science Libraries You Must Learn To Boost Your Skill Set in 2023Data science isn't just Pandas, NumPy, and Scikit-learn anymore — Motivation With 2023 just in, it is time to discover new data science and machine learning trends. While the old stuff is still essential, knowing Pandas, NumPy, Matplotlib, and Scikit-learn won't just be enough anymore. Last year's version of this post was more about classic ML, including libraries like CatBoost, LightGBM…Machine Learning7 min readMachine Learning7 min read
Published in Towards Data Science·Dec 15, 2022Member-onlyData Version Control in DVC: Learn What Other Data Scientists Are IgnoringA Complete tutorial to Data Version Control With DVC in Python — The big problem in data science When a dataset is large, it creates an even larger mess. Why? Data scientists and ML engineers perform many experiments on massive datasets and models, and their sheer size creates huge headaches regarding collaboration and software engineering best practices. Traditionally, software engineers collaborate by making copies of the central codebase…Data Science11 min readData Science11 min read
Published in Towards Data Science·Nov 30, 2022Member-onlyHow to (Finally) Install TensorFlow GPU on WSL2Never worry about a failed TensorFlow installation from now on — After my article on installing TensorFlow GPU on Windows took off and became a featured snippet on Google, I decided to write the same tutorial for Windows Subsystem Linux (WSL2). By finishing the article, you will be able to train TensorFlow models with GPU support from your WSL2 installation. 1. Make sure you have TensorFlow GPU installed on the Windows sideData Science5 min readData Science5 min read
Published in Towards Data Science·Nov 29, 2022Member-onlyHow to Create a Perfect Machine Learning Development Environment With WSL2 on Windows 10/11Everything included: terminal, Docker, Anaconda, Git, Jupyter Lab, GPU support, … — What is WSL, and why do you need it? That’s it. I’ve had enough. The failed installations, error messages, memes making fun of Windows, people thinking they are cooler because they are using Linux… That ends today by installing a fully-fledged Linux development environment right on top of the supposedly stupid Windows (10/11) using Windows Subsystem For Linux 2…Programming12 min readProgramming12 min read
Published in Towards Data Science·Oct 18, 2022Member-onlyHow to Boost Pandas Speed And Process 10M-row Datasets in MillisecondsUse Pandas the way it was intended to — Introduction “Great… another article on how to make Pandas n times faster.” I think I have said that countless times for the past two years I have been using Pandas. The most recent one I saw said, “make Pandas 71,803 times faster”. But I won’t give you that kind of promise…Artificial Intelligence6 min readArtificial Intelligence6 min read
