Dive into the second part of a comprehensive video tutorial on embeddings and fine-tuning, focusing on supervised fine-tuning. Learn how to effectively implement fine-tuning techniques, understand the differences between chat and base models, and explore supervised versus unsupervised approaches. Discover methods for converting datasets into Q&A format and practice supervised fine-tuning using Google Colab. Gain valuable pro tips and access scripts and GitHub repositories to enhance your learning experience. Follow along with presentation slides and a Llama 2 Inference Notebook to compare chat and base models.
Embeddings vs Fine-Tuning - Supervised Fine-tuning - Part 2
Overview
Syllabus
Part 2: Supervised Fine-tuning
How to make fine-tuning work?
How NOT to do fine-tuning
Video Overview
Chat vs Base models
Supervised versus Unsupervised fine-tuning
Converting a dataset into Q&A
Supervised fine-tuning in google colab.
Pro tips
Scripts and GitHub Repo Access
Taught by
Trelis Research
Related Courses
-
Embeddings vs Fine-Tuning - Part 3: Unsupervised Fine-Tuning
-
Preparing Fine-tuning Datasets with Open Source LLMs
-
Chat Fine-Tuning for LLMs - Instruction Format, Datasets, and Implementation
-
Fine-tuning Llama 2 for Tone or Style Using Shakespeare Dataset
-
Embeddings vs Fine-Tuning - Part 1: Understanding and Implementing Embeddings
-
Combined Preference and Supervised Fine-Tuning with ORPO
