표 5. | Table 5. 파인튜닝, 특징추출 비교 | Fine-tunning vs Feature Extraction

Feature	Fine-tuning	Feature Extraction
Model Size	May require a larger model	Can use a relatively smaller model
Generalization	Task-specific model	Generalizable features can be extracted
Resource Requirements	High computation and memory requirements	Lower computation and memory requirements
Lower computation and memory requirements	When ample labeled data is available	For small datasets or tasks with limited labels