Chromadb is slow for doc retrieval of long textual input

[nimahp]

Document retrieval is slow by using Chromadb to save vector representation of input text. Is there a way to make the whole process faster?