SophiArch
AdvancedML302

Transformer Models & Fine-Tuning

Understand how transformers work at the architectural level, then fine-tune pretrained models from Hugging Face for real tasks — covering attention, tokenisation, full and parameter-efficient fine-tuning, and the systematic skills to catch what AI gets wrong in fine-tuning pipelines.

Lessons are AI-assisted and human-reviewed. Learn more.

Syllabus

Transformer Architecture

01
The Transformer ArchitectureFree preview
30 min
02
Self-Attention and Multi-Head Attention
40 min
03
Tokenisation and Positional Encoding
35 min

Pretrained Models in Practice

04
Working with Hugging Face Transformers
35 min
05
Fine-tuning Strategies
35 min
06
Building a Fine-tuning Pipeline
40 min

Parameter-Efficient Fine-tuning

07
LoRA and Parameter-Efficient Fine-tuning
40 min
08
Debugging Fine-tuning Runs
35 min