get_similarity speed

[quxianjiuguo]

The structure is very clear and worth learning for beginners. Where improvements seem to be possible is in the way vector databases are stored and similarity is calculated