Overview
Discover best practices and tips for fine-tuning representation models in this 21-minute conference talk by Trung Nguyen, an applied research scientist at Snorkel AI. Learn how fine-tuning can enhance the performance of generative applications, with a focus on techniques for encoding meaningful features from raw data. Explore the importance of data quality and augmentation in the fine-tuning process. Gain insights into RAG-based pipelines and how multiple representation models can be utilized to optimize added context. Understand how one model can find relevant information while another ranks retrieved chunks to maximize the impact of context in the final prompt. Access accompanying slides and additional resources, including a full summary of Snorkel AI's Enterprise LLM Summit and recordings of related sessions, to deepen your understanding of representation models, data science, and machine learning.
Syllabus
Best Practices and Tips for Fine-Tuning Representation Models
Taught by
Snorkel AI