semantic-chunking
Semantically create chunks from large texts. Useful for workflows involving large language models (LLMs).
fast-cosine-similarity
A very fast implementation of cosine-similarity for comparing two vectors. Up to 6x faster than the compute-cosine-similarity library.