BentoML is Like Lego for ML engineers — In this post, we will guide you through the process of building a voice-based ChatGPT clone that relies on the OpenAI API and uses Wikipedia as an additional data source. To build and deploy this app, we’ll be using BentoML: a Python framework for model serving and deployment. BentoML not…

An end-to-end use-case combining PyTorch Lightning and BentoML — As an ML engineer, I often encounter two main challenges when working with deep learning models in production. The first one is the need to rewrite boilerplate code for each project to handle tasks like training loops, data loading, or metric computation. …

Do more things with less code without compromising on quality — Python decorators are powerful tools that help you produce clean, reusable, and maintainable code. I’ve long waited to learn about these abstractions and now that I’ve acquired a solid understanding, I’m writing this story as a practical guide to help you, too, grasp the concepts behind these objects. No big…

This goes from code review, unit testing, performance optimization to code understanding — ChatGPT is the most impressive AI project I have ever seen. I was legitimately lost for words on certain occasions when I was interacting with it. As a developer, I see it as an enormous advantage to embrace to enhance my productivity. In fact, if you use ChatGPT to your…

Usecase: an end-to-end service to summarize Youtube videos — As a huge fan of Ted talks, I have spent countless hours scrolling through YouTube to watch them. However, with my limited free time lately, I began to think about creating a video summarizer that would allow me to get a quick overview of any video before committing to watching…

ChatGPT is like Google, StackOverflow and Readthedocs combined — You’ve probably heard about ChatGPT. It’s a new AI-powered chatbot developed by OpenAI, the same company behind GPT-3, DALL.E2, and other crazy products. ChatGPT is out of this world. It’s nothing like the chatbots I’ve ever used before. Its ability to generate detailed and comprehensive answers on various topics is…

Moving from Jupyter notebooks to production is not that difficult after all — If you’re a data scientist, you probably spend a lot of time developing intricate Jupyter notebooks to perform data analysis, build complex training pipelines, or compute statistics. Jupyter notebooks are great for this and allow us to prototype ideas in no time. But, what happens once you’re done with this…

Just like unit tests, type hints take developer time but pay off in the long run — Python is a dynamically-typed programming language. This means that the interpreter performs type checking only when the code runs and the variable type is allowed to change over its lifetime. Although Python has always remained dynamically typed and was never intended to change this direction, type hints have been introduced…

I’ve never used spaCy beyond simple named entity recognition tasks. Boy was I wrong. — I’ve recently started to revisit the spaCy library and try out many of its core functionalities to perform low-level linguistic tasks. If you’ve never heard about it, spaCy is a modern natural language processing library that handles +66 languages, provides state-of-the-art speed, and has various components for named entity recognition…