Learn Before
Concept

Disadvantages of Supervised Pre-training

The primary drawback of supervised pre-training is its significant requirement for labeled data. As the complexity of neural networks increases, the volume of labeled data needed for effective pre-training also rises, making the approach challenging and difficult to apply when large-scale labeled datasets are not available.

0

1

Updated 2026-04-14

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