Learn Before
Diagnosing Fine-Tuning Performance Issues
A development team is fine-tuning a language model to generate marketing email subject lines. Their goal is to create subject lines that are concise and create a sense of urgency. After an initial fine-tuning process, they observe that the model generates subject lines that are often too long, generic, and lack a compelling tone. Analyze the two instruction sets below and determine which one was more likely used in the initial, unsuccessful fine-tuning process. Justify your choice by explaining how the principles of crafting effective instructions relate to the observed poor performance.
0
1
Tags
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Analysis in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Flexibility in Instruction Formulation for LLMs
A machine learning engineer is preparing a dataset to fine-tune a language model for a specific task: summarizing customer support tickets into a single sentence for a quick-glance dashboard. Which of the following instructions, when included in the training examples, is most likely to result in a high-performing and reliable model for this specific task?
Diagnosing Fine-Tuning Performance Issues
The Importance of Instructional Clarity in Fine-Tuning