Skip to content

Latest commit

 

History

History
73 lines (50 loc) · 2.12 KB

README.md

File metadata and controls

73 lines (50 loc) · 2.12 KB

Search with Lepton

Build your own conversational search engine using less than 500 lines of code.
Live Demo

Features

  • Built-in support for LLM
  • Built-in support for search engine
  • Customizable pretty UI interface
  • Shareable, cached search results

Setup Search Engine API

There are two default supported search engines: Bing and Google.

Bing Search

To use the Bing Web Search API, please visit this link to obtain your Bing subscription key.

Google Search

You have three options for Google Search: you can use the SearchApi Google Search API from SearchApi, Serper Google Search API from Serper, or opt for the Programmable Search Engine provided by Google.

Setup LLM and KV

The original repo uses Lepton AI cloud service's LLM and key-value storage services. Here, a locally deployed service compatible with the OpenAI API is used, temporarily not using the key-value storage service (TODO, hence there is no functionality for caching queries and results yet).

Build

  1. Set Bing subscription key
export BING_SEARCH_V7_SUBSCRIPTION_KEY=YOUR_BING_SUBSCRIPTION_KEY
  1. Build web
cd web && npm install && npm run build
  1. Run server
BACKEND=BING python search_with_lepton.py

For Google Search using SearchApi:

export SEARCHAPI_API_KEY=YOUR_SEARCHAPI_API_KEY
BACKEND=SEARCHAPI python search_with_lepton.py

For Google Search using Serper:

export SERPER_SEARCH_API_KEY=YOUR_SERPER_API_KEY
BACKEND=SERPER python search_with_lepton.py

For Google Search using Programmable Search Engine:

export GOOGLE_SEARCH_API_KEY=YOUR_GOOGLE_SEARCH_API_KEY
export GOOGLE_SEARCH_CX=YOUR_GOOGLE_SEARCH_ENGINE_ID
BACKEND=GOOGLE python search_with_lepton.py