[BETA] Finetuning Transformer Models

Finetuning is an essential skill in the world of Large Language Models (LLMs), allowing you to customize pre-trained transformer models for specific tasks. This course will guide you through the practical process of finetuning using popular tools like LoRA, quantization, and QLoRA, using Hugging Face libraries and the popular Mistral series of open-weight LLMs. You'll learn about the life cycle of finetuning, GPU usage in deep learning, and parameter-efficient finetuning (PEFT).

Syllabus

Finetuning Transformer Models: Learn the basics of finetuning transformer models as well as the techniques used to efficiently train large models on consumer hardware (PEFT).

Lesson: Finetuning Fundamentals
Lesson: Finetuning with LoRA and QLoRA
Quiz: Introduction to Finetuning Quiz

Finetuning with Hugging Face: Practice finetuning with Hugging Face libraries, performing a full finetune, LoRA, and QLoRA on a BERT sentiment classifier.

Lesson: Finetuning Transformers with Hugging Face
Quiz: Finetuning Transformers with Hugging Face
KanbanProject: Finetuning a Generative Language Model
Informational: Next Steps