Skip to content

Latest commit

 

History

History
33 lines (22 loc) · 1.23 KB

README.md

File metadata and controls

33 lines (22 loc) · 1.23 KB
Raw

LLMitless

Build, test, and deploy LLM chains

Simple scaffolding, testbed, and API endpoints for building, testing, and deploying LLM chains.
Supports local files, as well as those passed in directly or via API calls.

** consider experimental and unstable **

Setup

Devcontainer

For fastest setup, start the devcontainer in VS code.

Local Setup

  • Install dependencies via poetry.
  • Add a .env file with your api keys (wandb, openai, etc)
  • Use datasets_sample as a template for the hierarchy of local files for use with chains. Best numbered and laid out in the format shown in datasets_sample
  • Set a port in serve.py, open the port (using ngrok, for example), and run serve.py.
  • LLMitless uses FastAPI: once running, you can find documentation and make test calls by visiting /docs

CI/CD

  • Tests run automatically on pushes to main
  • The manual Deploy action will dockerize the app, store the image in Google Artifact Registry, and deploy from there to Cloud Run.

Key modules

parsers/ contains modules that parse formats for use with LLMs

services/summarize.py collects docs, prepares them for use in a chain, and calls the LLM

chains/ contains core langauge processing chains

serve.py Fast API endpoints