Chapter 5 & 6: More Applications of Computer Vision with fastai
In this blog we look at applications of deep learning applied to computer vision. We will look at multiclass classification, multilabel classification, loss functions, and other techniques applied to computer vision like different types of layers, regularization methods, optimizers, how to put layers together into architectures, labeling techniques, and much more. How do you create an architecture that matches the needs of your particular domain? How do you get the best possible results from the training process? How do you make things faster? What do you have to change as your data‐ sets change? These are questions we will explore here.
Lesson-7: An Introduction to Recommendation Systems and Collaborative Filtering
In this lesson Jeremy takes us through the idea of recommendation systems, collaborative filtering, embeddings, and latent factors. Chapter-8 in the book covers the exact same information as this lesson.
Lesson-4: An Introduction to NLP
In this lesson Jeremy takes us through a simple end-to-end example in an NLP classification task using Huggingface Transformers. Chapter-10 in the book very loosely corresponds to this lecture as it covers the process of NLP with fastai and some of the basic steps of converting text to numbers (which are what our models expect). This blog post will first cover the contents of the chapter as those are fundamental ideas behind NLP and then cover the tutorial from the lecture itself.
Lessons 3 & 5: Building a Neural Network from scratch
In these lessons Jeremy takes us through the foundational ideas of neural networks and how to build a simple neural network from scratch. He creates a model first in excel and then in python. The post will not cover the excel implementation as I think that made it seem harder instead of easier. Chapter-4 loosely corresponds to these two lectures and covers a great deal more about datasets, dataloaders, the training loops, optimizers which are how the model learns and updates weights, as well as more.
Lesson-2: From Model to Production
In this lesson Jeremy takes us through a simple end-to-end example of exploring a dataset, cleaning it, looking at confusion matrix, and more. We also go through an introduction to huggingface spaces and gradio to host our models. Chapter-2 in the book corresponds to this lecture but also adds the ideas of out-of-domain data and data-shift in the context of production systems. Additionally, the book covers model deployment in the notebook itself but the blog post will not cover that since the Huggingface spaces implementation is a much better approach.
Lesson-1: Deep Learning in Practice
In this lesson Jeremy Howard introduces the course, the library, some of the people behind the course, and alumni who have succeeded with the training from the course. This discussion is proceeded by some code to classify birds vs not-birds (or forests). Finally, Jeremy shows us that Deep Learning is not only about images but a lot more, some of which we will cover in later lessons. The chapter-1 in the book that corresponds to the lecture digs into some more details about what Jeremy covers in the lesson and also has a simple example of an image classification code that uses fast.ai built in dataset unlike the lesson where we download a new dataset.