| Ressources |
 CM 01 - Static Semantic Representation |
|
|
 Lab: play with short, static and dense word representations |
|
|
CM 02 - Transformer architecture |
|
|
 Lab - comprendre l'architecture du Transformer |
|
|
CM 03 - Deeper with Transformers |
|
|
Lab: Model size |
|
|
Lab: Generative sampling techniques |
|
|
CM 04 - Pre-training, Prompting or Fine-tuning LLMs |
|
|
Lab: Instructed model and prompting (zero and few shots) |
|
|
CM 05 - Reducing learning and inference costs |
|
|
Lab: full vs parameter efficient fine-tuning |
|
|
CM 06 - Aligning language models with human preferences |
|
|
Lab: Direct Preference Optimization (DPO) Using Hugging Face |
|
|
Lab: Reinforcement Learning from Human Feedback Using PPO |
|
