Hands-On Deep Learning: From Neural Networks to Generative AI

About this path

This learning path is based on MIT's 15.773 Hands-On Deep Learning course (Spring 2024), taught by Rama Ramakrishnan. It is designed for learners who already know Python and the basics of machine learning (training/validation/testing, overfitting/underfitting, regularization). You will build and train deep neural networks using Keras/TensorFlow, work with convolutional networks for images, explore transformers for natural language, and learn about large language models and text-to-image generation. The emphasis is on practical understanding: how to set up models, train them, and apply them to real problems with unstructured data.

What you will learn

• Understand the basics of deep neural networks and how to train them.
• Build convolutional neural networks for image and video processing.
• Apply transfer learning and fine-tuning to computer vision tasks.
• Understand embeddings and transformers for natural language processing.
• Learn about generative large language models and retrieval augmented generation.
• Explore parameter-efficient fine-tuning and text-to-image models.

Who this is for

Self-paced learners who want a clear free path with creator-credited YouTube lessons.

Curriculum