FLORIAN BASTIN
All sessions

Course 03

Large Language Models

This session covers large language models: their architecture, text generation, prompting, and the main techniques used to speed up inference.

One thread runs through the whole session: the cost of compute and memory.

On the programme
  1. 01Definition and architecture
  2. 02Mixture of experts
  3. 03Context length, temperature
  4. 04Sampling strategies
  5. 05Prompting, in-context learning
  6. 06Chain of thought
  7. 07Self-consistency