Example of a Creative Writing Task for LLMs
Diverse instructions, such as creative writing prompts, can enhance the generalization capabilities of Large Language Models by defining new tasks. An example is prompting a model to write a short story based on specific constraints, such as: "Write a short story about three characters. Character 1: A boy whose name is Andy. Character 2: Andy's dog. Character 3: Andy's friend, Jane." The model's expected output is a creative narrative incorporating these elements, such as a story about the characters discovering a hidden garden and sharing an adventure.
0
1
Tags
Ch.2 Generative Models - Foundations of Large Language Models
Foundations of Large Language Models
Computing Sciences
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models Course
Related
Multi-Task Capability through Diverse Fine-Tuning Datasets
Modern Focus of Instruction Fine-Tuning Datasets
Using Diverse Data to Steer LLM Specialization
Examples of Instruction-Following Tasks in SFT Datasets
A development team has fine-tuned a large language model to be a helpful assistant. They observe that the model excels at summarizing technical documents and answering direct factual questions, which were the primary tasks in its fine-tuning dataset. However, when users ask it to perform more creative tasks like writing a short poem or brainstorming marketing slogans, the model's performance is poor and generic. Which of the following strategies would be the most effective next step to improve the model's ability to handle this wider range of user requests?
Using Varied Instructions for a Single Task to Enhance Data Diversity
Improving a Customer Service Chatbot's Robustness
Characteristics and Limitations of Early Instruction Fine-Tuning Datasets
Evaluating a Fine-Tuning Strategy for LLMs
Example of a Recipe Generation Task for LLMs
Example of a Creative Writing Task for LLMs
Example of a Math Word Problem Task for LLMs