Learn Before
Rationale for Fine-Tuning in Self-Refinement
A research team is developing a language model to act as a creative writing assistant. They want the model to not only generate story drafts but also to iteratively critique and improve its own writing for plot consistency. The team has a large dataset of story drafts, human-written critiques of those drafts, and the corresponding revised, improved versions. Explain why fine-tuning the model on this specific dataset is a suitable strategy to enhance its self-refinement capabilities for this task.
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
Enhancing a Code-Generating Model's Style Adherence
A development team wants to improve a language model's ability to write concise summaries of long articles. The goal is for the model to generate an initial summary, critique its own work for clarity and relevance, and then revise it. The team has a dataset of thousands of examples, each containing: (1) an initial, verbose summary generated by a model, (2) a human-written critique of that summary, and (3) a final, human-written concise summary. Which of the following fine-tuning strategies would be most effective for improving the model's ability to perform this iterative improvement process?
Rationale for Fine-Tuning in Self-Refinement