Case Study

Debugging a Text-to-Text Model's Training Data

A data science team is training a text-to-text model with the goal of simplifying complex sentences. However, they find the model is generating very short, high-level summaries instead of rephrasing the original sentences in simpler terms. Analyze the following training data sample and identify the specific error in its format that is causing this incorrect behavior. Explain your reasoning and suggest a correction.

0

1

Updated 2025-10-09

Contributors are:

Who are from:

Tags

Ch.1 Pre-training - 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