Category: Evaluation of Generative Models

Welcome to our blog post series: “Evaluation of Generative Models.” In this series, we will address the challenges of evaluating systems designed for text generation. Be that machine translation, conversational dialogue systems, automated text summarization systems, or classical NLG (data-to-text).

Our goal for this blog post series is twofold:

• firstly, we aim to inspire and motivate our readers on the importance of the evaluation of generative models.
• secondly, we want to take our readers on this journey with us, inviting them to join in and contribute to the conversation. We hope to spark interesting discussions, as well as feedback and critiques, as we believe it will lead to further improvements in the field.

By following this blog post series, you will firstly gain a deeper understanding of the challenges of evaluating text generation models. Later you will learn about an effective statistical framework that can be used to address these challenges. So, come along with us on this journey and let’s explore the fascinating world of generative models together!