In this new exciting era of impressive AI advancements, Large Language Models (LLMs) like OpenAI’s ChatGPT have captured a huge interest from the public. Whether it’s used for writing emails, debugging code, or answering complex questions, ChatGPT shows an outstanding capability of providing accurate and human-like responses.
But which Machine Learning technologies enable ChatGPT’s capabilities? Both for ML Engineers and AI enthusiasts, understanding the inner working of these models is the key to unlocking their full potential!
In the first part of this article, we go through the evolution of these models from GPT to GPT-4. In the second part, we delve into the technical details, to understand how they can generate such accurate responses.