Iam building a docker app that will be

Depending on the processing load, either just a local instance of HuggingFaceEmbedding i.e. with bge or something

Or if theres a ton of processing, a dedicated server with text-embedding-interface would be better

The huggingfaceembedd installa the embedd engine with the package or downloads it on init ?

Nvm just found the docs# https://docs.llamaindex.ai/en/stable/examples/embeddings/huggingface.html

It will download and run a local embedding model yea 👍

text-embedding-interface is a dedicated + optimized server for embeddings

two different setups 👍