Back to Glossary
Few-Shot Learning
NLP & Language Models
Learning from only a few examples.
Few-shot learning enables a model to solve new tasks with just a few examples. The model leverages its prior knowledge and generalization capabilities.
- Examples: Classifying new categories, translating with limited examples.
- Methods: In-context learning, fine-tuning, meta-learning.
- Benefit: Efficient adaptation to small datasets.