rag documentation

never expect myself to build something like this. so here i go explain each parts of my rag

content: just a simple file to test run the algo.

qa chain: initializer to load the model and tokenizer. it prepares LangChain prompt template, create the chain (retriever -> formatter -> prompt -> model -> parser). it querying user question through the entire chain and return the generated answer. could say the skeleton :P

documentProcessor: data prep layer and it is the question-answering layer that uses the retriever built. this will pass on to qa chain question-answering layer.

vectorStore: temporary place to store data.

config: config to point out how the algorithm and where it supposed to go and process

main: name says it all, the main. welp it connects everything together, including config. lets just say its the everything above everything.

you know what? see it for yourself! try it! lmk what you think and i would truly appreciate if y'all keep contributing or starred or sth idk. here is the here!