A language model is first trained on a large text corpus where some words in each sentence are replaced with a special [MASK] symbol, and the model's goal is to predict the original words. Subsequently, this pre-trained model is adapted for a specific task, such as sentiment analysis, using a new dataset of complete, un-masked sentences. Which of the following statements best analyzes the primary architectural conflict that arises between these two phases?
0
1
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
Related
Impact of Pre-training/Fine-tuning Mismatch on Downstream Tasks
A language model is first trained on a large text corpus where some words in each sentence are replaced with a special
[MASK]symbol, and the model's goal is to predict the original words. Subsequently, this pre-trained model is adapted for a specific task, such as sentiment analysis, using a new dataset of complete, un-masked sentences. Which of the following statements best analyzes the primary architectural conflict that arises between these two phases?Troubleshooting a Pre-trained Model's Performance
Permuted Language Modeling (PLM)