Udemy – GenAI Foundation (by 3aayaam) – Maths of Attention Mechanism

In this course I will be explaining the pre-requisites for my GenAI vol 2 course “GenAI Continuum – Application Development using OpenAI & Gemini” .  This course covers mathematical concepts, terminologies, jargons that are needed for deep understanding of GenAI (Foundation) Models.

This course is divided into 4 sections as mentioned below:

• GenAI Prequel – AI, ML & GenAI; Type & Nature of data; Scalars & Vectors; Mathematical equations in AI Models.
• The Original ML Story – Model Training; Gradient Descent; Types of ML Models; Underfitting & Overfitting; Deep Neural Networks; Epoch, Batch etc.
• Arrival of Transformer – Transformer Architecture; Tokenization; Vector Embedding; Vocabulary; Self-Attention & Masked-Attention; Encoder & Decoder.
• Nerd Attack – Mathematics of Self-Attention & Masked-Attention.

NOTE : This is an optional pre-requisite course for Application Development using GenAI. But I would recommend everyone to go through this if you want to understand the mathematics behind transformer architecture, which is the foundation of current GenAI models.