default redis dump

adding example search, correcting vectors

adding data file

showing multiple cached answers

adding configuration for max similarity score

caching original question as answer, limiting acceptable answer score

using HNSW and COSINE

adding settings, refining prompts

adding example env files

removing unnecessary await

adding more documentation

simplifying the duplicate llm logic, reducing necessary environment variables

semantic caching in use for both Google and OpenAI

actually using semantic answer cache

adding initial readme