AI / ML

Switch track

Deep Learning

An intermediate, project-driven path from neural-network and backprop intuition through CNNs, sequences, and transformers to a deployable capstone, for learners who already know Python and ML basics.

Finish line: Capstone: End-to-End Deep Learning Project

How to use your map

  • • Work top to bottom — each stage builds on the one before it.
  • • Complete a topic to unlock what it leads to.
  • • When unsure, start at your highlighted next best step.
✓ done▢ next best step (ringed) locked — finish its prerequisites first
Your progress0 / 10 done · 0%

~53 hr left · about 106 days at 30 min/day

1Start here

Neural Networks & Backpropagation Intuition

concept

Every deep-learning model is trained by backpropagation, so a mental model of it prevents blind copy-paste later.

~240 min·difficulty 2/5
2Stage 2

Activation Functions & Loss Functions

concept

Activation and loss choices determine whether a network can learn the task at all and whether gradients vanish.

~150 min·difficulty 2/5
3Stage 3

Training Neural Networks with PyTorch

skill

You need a hands-on framework to turn intuition into working, trainable models.

~240 min·difficulty 3/5
4Stage 4

Regularization for Deep Learning

skill

Deep nets overfit easily, and regularization is what makes them generalize to new data.

~180 min·difficulty 3/5
5Stage 5

Convolutional Neural Networks for Images

skill

CNNs are the workhorse for vision and introduce the idea of learned spatial feature hierarchies.

~300 min·difficulty 3/5
6Stage 6

Transfer Learning & Fine-Tuning

skill

Transfer learning is how practitioners get strong results with small datasets and modest compute.

~180 min·difficulty 3/5

RNNs & Sequence Modeling

concept

Sequence models introduce the notions of state and long-range dependency that motivate attention and transformers.

~180 min·difficulty 4/5
7Stage 7

Introduction to Transformers & Attention

concept

Transformers underpin modern NLP and vision, and attention is the key mechanism to understand.

~210 min·difficulty 4/5

Build & Deploy End-to-End with fast.ai

capability

Consolidating skills into a full train-and-deploy workflow turns knowledge into shippable capability.

~600 min·difficulty 4/5
8Stage 8

Capstone: End-to-End Deep Learning Project

project

A self-directed capstone proves you can integrate every prior skill into a real, evaluated deliverable.

~900 min·difficulty 5/5