Evaluating Fine-Tuning Project Feasibility
A research lab has a fixed budget for a single fine-tuning project. They must choose between two options to create a specialized question-answering model. Based on the principle that updating a model's parameters is a primary driver of computational expense, evaluate which project is likely to be more computationally costly and therefore less feasible within their fixed budget. Justify your reasoning.
0
1
Tags
Ch.4 Alignment - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Evaluation in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Parameter-Efficient Methods for Mitigating Fine-Tuning Costs
Evaluating Fine-Tuning Project Feasibility
A machine learning team is fine-tuning a 70-billion parameter language model. They decide to double the size of their high-quality training dataset, from 500,000 examples to 1,000,000 examples. Which of the following best analyzes the primary driver for the substantial increase in computational cost for this project?
Analyzing Fine-Tuning Resource Requirements