My Deep Learning Journey: From Basics to Transformers
Back to PortfolioDeep Learning
June 10, 202610 min read

My Deep Learning Journey: From Basics to Transformers

Sharing my learning path through neural networks, CNNs, RNNs, and finally Transformers — what worked and what didn't.

Where It All Started

My deep learning journey began with a simple question: How do machines learn to see?

The Learning Path

Phase 1: Neural Network Fundamentals

  • Perceptrons, activation functions, backpropagation
  • Built a digit classifier from scratch (no frameworks!)
  • Understood why ReLU beats Sigmoid

Phase 2: Convolutional Neural Networks

  • Image classification with CNNs
  • Transfer learning with ResNet and VGG
  • Data augmentation techniques

Phase 3: Sequence Models

  • RNNs and the vanishing gradient problem
  • LSTMs and GRUs for text generation
  • Attention mechanism — the game changer

Phase 4: Transformers

  • Self-attention and multi-head attention
  • BERT for NLP tasks
  • Vision Transformers (ViT)

Resources That Helped Me Most

  1. Andrew Ng's Deep Learning Specialization
  2. FastAI's Practical Deep Learning course
  3. "Attention Is All You Need" paper
  4. Hugging Face documentation
The best way to learn deep learning is to build things, break things, and build again.
Golam Monayem

Golam Monayem

Machine Learning Engineer & Visual Designer