Skip to content

Latest commit

 

History

History
89 lines (60 loc) · 4.11 KB

README.md

File metadata and controls

89 lines (60 loc) · 4.11 KB
Raw

docu-rag-ami

This template contains a reference architecture for Retrieval Augmented Generation against a set of documents using Docugami's XML Knowledge Graph (KG-RAG).

Setup

You need to set the following environment variables before using your new app based on this template:

  1. OPENAI_API_KEY: you can request one in the openai site.
  2. DOCUGAMI_API_KEY: from the Docugami Developer Playground
export OPENAI_API_KEY=<openapi key>
export DOCUGAMI_API_KEY=<docugami api key>

Process Documents in Docugami (before you use this template)

Before you use this template, you must have some documents already processed in Docugami. Here's what you need to get started:

  1. Create a Docugami workspace (free trials available)
  2. Add your documents to Docugami for processing:
    • Upload via the simple Docugami web experience. Detailed instructions.
    • Upload via the Docugami API, specifically the documents endpoint. In this case you need to get an access token via the Developer Playground for your workspace. Detailed instructions. You can find code samples for python and JavaScript or use the docugami python library.
  3. Organize documents in DocSets: Docugami organizes documents in sets, you can either create a DocSet (Document Set) manually or let Docugami group them by similarity of content. Note: Docugami KG-RAG is currently limited to 1 DocSet so you should add all the documents that you want to query in the same Document Set.
  4. Monitor Processing status: before you can star querying documents, they all need to be fully processed by Docugami. You can monitor status in the Docugami webapp, or use a webhook.

You can also find detailed instruction on how to manage the upload and 'documents in ready state' process in the Docugami RAG over XML Knowledge Graphs (KG-RAG) Cookbook.

Usage

Indexing

  1. build your index in Chroma. See index.py which you can run via poetry run python index.py. The CLI will query docsets in the workspace corresponding to your DOCUGAMI_API_KEY and let you pick which Document Sets you want to index. You can find more details on how the indexing process works in this [documentation](https://python.langchain.com/docs/integrations/document_loaders/do cugami) for details.

Create an App

  1. Install LangChain CLI :
pip install -U langchain-cli
  1. Create a new LangChain project:
langchain app new my-app --package git+https://github.com/docugami/langchain-template-docugami-kg-rag.git

If you want to add this to an existing project:

langchain app add git+https://github.com/docugami/langchain-template-docugami-kg-rag.git
  1. Add the following code to your my-app/app/server.py file:
from docugami_kg_rag import chain as docugami_kg_rag_chain

add_routes(app, docugami_kg_rag, path="/docugami-kg-rag")

[Optional] configure LangSmith to trace, monitor and debug LangChain applications. LangSmith is currently in private beta, you can sign up here.

export LANGCHAIN_TRACING_V2=true
export LANGCHAIN_API_KEY=<your-api-key>
export LANGCHAIN_PROJECT=<your-project>  # if not specified, defaults to "default"

Running app

  1. From inside the app directory, spin up a LangServe instance:
langchain serve

You can access the template from code with:

from langserve.client import RemoteRunnable

runnable = RemoteRunnable("http://localhost:8000/docugami-kg-rag")