Comment by ChromaticPanic

When I was looking at your library last week, It didn't look like there was a direct way to use my own embedding model endpoints. For example, I run snowflake arctic embed in vllm and it would be good be able to use it with Chonkie's semantic chunkers.