Our great sponsors
-
anything-llm
The all-in-one Desktop & Docker AI application with full RAG and AI Agent capabilities.
-
WorkOS
The modern identity platform for B2B SaaS. The APIs are flexible and easy-to-use, supporting authentication, user identity, and complex enterprise features like SSO and SCIM provisioning.
-
BriefGPT
Locally hosted tool that connects documents to LLMs for summarization and querying, with a simple GUI.
I just came across this project which seems to be aiming at streamlining exactly that :
https://github.com/Mintplex-Labs/anything-llm
RAG is very difficult to do right. I am experimenting with various RAG projects from [1]. The main problems are:
- Chunking can interfer with context boundaries
- Content vectors can differ vastly from question vectors, for this you have to use hypothetical embeddings (they generate artificial questions and store them)
- Instead of saving just one embedding per text-chuck you should store various (text chunk, hypothetical embedding questions, meta data)
- RAG will miserably fail with requests like "summarize the whole document"
- to my knowledge, openAI embeddings aren't performing well, use a embedding that is optimized for question answering or information retrieval and supports multi language. Also look into instructor embeddings: https://github.com/embeddings-benchmark/mteb
1 https://github.com/underlines/awesome-marketing-datascience/...
RAG is very difficult to do right. I am experimenting with various RAG projects from [1]. The main problems are:
- Chunking can interfer with context boundaries
- Content vectors can differ vastly from question vectors, for this you have to use hypothetical embeddings (they generate artificial questions and store them)
- Instead of saving just one embedding per text-chuck you should store various (text chunk, hypothetical embedding questions, meta data)
- RAG will miserably fail with requests like "summarize the whole document"
- to my knowledge, openAI embeddings aren't performing well, use a embedding that is optimized for question answering or information retrieval and supports multi language. Also look into instructor embeddings: https://github.com/embeddings-benchmark/mteb
1 https://github.com/underlines/awesome-marketing-datascience/...
Marqo provides automatic, configurable chunking (for example with overlap) and can allow you to bring your own model or choose from a wide range of opensource models. I think e5-large would be a good one to try. https://github.com/marqo-ai/marqo
Very cool. I was planning on working on something very similar myself for a hackathon I attended but my team basically ran out of time (https://github.com/anshumankmr/jovian-genai-hackathon)