Large Language Model Foundations

Course overview

Foundations of LLMs and the ecosystem: NLP/Transformers concepts, tokenizers, pipelines, fine‑tuning, datasets, the Hub, Gradio/Argilla tools, training loops, evaluation and RL methods for LLMs. Also covers debugging and model sharing.

Introduces NLP and Transformer basics, how they work and how pipelines solve tasks.

Covers tokenizers, sequence handling, data processing, training loops, fine‑tuning (Trainer API, SFTTrainer, LoRA) and learning curve analysis.

Presents using the Hub for models and datasets, dataset creation/annotation, debugging training pipelines, Gradio for demos and advanced RL techniques applied to LLMs.