FLORIAN BASTIN
All sessions

Course 04

LLM training

This session starts with how language models are trained: pretraining, the cost in compute and memory, distributed training and the main optimisations.

The second half covers finetuning, which the following lab builds on.

On the programme
  1. 01Pretraining
  2. 02Quantization
  3. 03Hardware optimization
  4. 04Supervised finetuning (SFT)
  5. 05Parameter-efficient finetuning (LoRA)