Abschnitt: Ressources | Modélisation des langues

CM 01 - Static Semantic Representation Datei

Lab: play with short, static and dense word representations Link/URL

CM 02 - Transformer architecture Datei

Lab - comprendre l'architecture du Transformer Textseite

CM 03 - Deeper with Transformers Datei

Lab: Model size Link/URL

Lab: Generative sampling techniques Link/URL

CM 04 - Pre-training, Prompting or Fine-tuning LLMs Datei

Lab: Instructed model and prompting (zero and few shots) Link/URL

CM 05 - Reducing learning and inference costs Datei

Lab: full vs parameter efficient fine-tuning Link/URL

CM 06 - Aligning language models with human preferences Datei

Lab: Direct Preference Optimization (DPO) Using Hugging Face Link/URL

Lab: Reinforcement Learning from Human Feedback Using PPO Link/URL

Abschnittsübersicht