LLAMARA - Large Language Assistant for Model-Augmented Retrieval and Analysis - is an LLM-based assistant for information retrieval from a provided knowledge base. It aims at supporting researchers working with scientific papers, whitepapers and documentation, as well as possibly serving research findings in an accessible way to the public.
NOTE: This repository contains the LLAMARA distribution build and release pipelines only. Source code for LLAMARA backend can be found at llamara-ai/llamara-backend, the source code for LLAMARA frontend is currently not publicly available.
LLAMARA Distribution is a ready-to-use build of LLAMARA backend that includes LLAMARA frontend as well.
It is published as a Docker container and can be deployed using Docker Compose. You can also download the built JAR as artifact from the CI Build action and from individual releases.
The Docker image ghcr.io/llamara-ai/llamara provides the following labels:
latest: points to the latest releasevX.Y.Z: points to a specific releasemain: points to the latest snapshot build from themainbranches of backend, frontend and distribution repositories