Week 1
PyTorch Foundations
Master Tensors, Autograd, and build your first Neural Network from scratch.
Week 2
Transformer Architecture
Visualize Self-Attention and build the Transformer Block.
Week 3
LLM Variants
Explore Tokenization (BPE) and GPT-2 Latent Spaces.
Week 4
Router Networks
Capstone: Routing queries to expert models.
Week 5
Diffusion Math
Understanding the Forward Process (Adding Noise).
Week 6
U-Net Architecture
Visualizing Skip Connections and Downsampling.
Week 7
Latent Diffusion
Interpolating between concepts in Latent Space.
Week 8
Capstone: Noisy Router
A system that chooses between a Diffusion Artist and a Critic.
Week 9
LoRA & PEFT
Efficient Fine-Tuning with Low-Rank Adapters.
Week 10
Mixture of Experts
Sparse Expert models and Token Routing.
Week 11
Optimization
Quantization (FP32 vs INT8) for inference speed.
Week 12
Final Capstone
The Ultimate Assistant: Modular, Efficient, Powerful.