Skip to content

Playing around with Retrieval Augmented Generation.

Notifications You must be signed in to change notification settings

w-k-s/LLM-RAG-Experiments

Repository files navigation

LLM RAG Experiments

Playing around with Retrieval Augmented Generation.

Next Steps

  • Go through first tutorial
  • Try this with my own dataset (e.g expenses csv)
  • Repeat with my own dataset but this time:
    • See if there's a converter for CSV to Haystack Document
    • Split and clean the document
    • Then see if it provides better answers
  • Understand the pipeline connections (draw a diagram)
  • Make it interactive, with the abiliy to add documents and add questions about newly added documents
  • What are the different configurations and what do they mean?

## Questions

  • In Lesson 2, Why did we initialise the InMemoryDocumentStore with BM25Plus and use InMemoryBM25Retriever?

Notes

Actually, Hugging Face gave a funny answer in lesson 3. Just noting it here.

** Which of the following recipes would be meet with the most disapproval from Ron Swanson? Explain why in Ron Swanson's style of speech**

Ron Swanson would most likely disapprove of the vegan persimmon flan recipe due to its use of non-traditional ingredients such as persimmon pulp, agar agar, and coconut creme.

In Ron Swanson's style of speech, he might say, "This flan is an abomination! Persimmon pulp in a flan? That's like putting a turkey in a lasagna! It's an affront to all that is holy and delicious in this world.

I'll have my men round up every copy of this recipe and burn it in a ceremonial bonfire. From this day forward, flan will be made the way it was meant to be made: with eggs, sugar, and a healthy dose of manliness."

About

Playing around with Retrieval Augmented Generation.

Topics

Resources

Stars

Watchers

Forks

Languages