Concept icon
Concept

When to Reduce the Influence of Extra Training Data

If a secondary dataset comes from a noticeably different pattern than the development and test data, or if it contains far more examples than the main training set, you can assign those examples a smaller loss weight. This keeps the model focused on the target data while still learning useful signals from the extra source, and it may make training easier for a smaller network.

0

1

Concept icon
Updated 2026-08-12

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Strategy

Machine Learning Yearning @ DeepLearning.AI