# Dicklesworthstone/swiss_army_llama

A FastAPI service for semantic text search using precomputed embeddings and advanced similarity measures, with built-in support for various file types through textract.

Repository: https://github.com/Dicklesworthstone/swiss_army_llama
Canonical: https://ross.abutalabs.com/products/swiss_army_llama
Language: Python
License Family: other
Topics: embedding-similarity, embedding-vectors, embeddings, llama2, llamacpp, semantic-search
Last push: 2025-02-27T11:19:55+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 8, release rhythm 35, longevity 80
- inputs: {"age_days": 1122, "days_push": 552, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1056, forks 66 (observed 2026-08-28T04:03:24.565006+00:00)

## What it is
A FastAPI-based REST service that exposes local LLM capabilities including text embeddings, completions, semantic similarity, and semantic search via llama.cpp and FAISS. It automatically computes and caches embeddings for common document types (PDFs with OCR, Word files, audio transcriptions) in SQLite.

## Use cases
- run a self-hosted embeddings API for local LLMs
- semantic search across my documents
- compute embeddings for PDFs and Word files automatically
- transcribe audio files and embed the text
- compare semantic similarity between text strings with advanced measures
- cache embeddings so I don't recompute them
- expose llama.cpp models behind a REST API

## When to choose
- you want a single self-hosted service combining embeddings, completions, and semantic search
- you need automatic embedding of diverse file types including OCR'd PDFs and audio
- you want advanced similarity measures beyond cosine similarity
- you prefer local LLMs without cloud API dependencies

## When to avoid
- you need a managed or cloud-hosted embedding service
- you require a permissively licensed dependency (no license is specified)
- you need GPU-scale high-throughput inference rather than local convenience
- you want a library to embed in your own code rather than a standalone service

## Facets
- artifact type: service
- maturity: active
- function: llm-inference, rag, search-engine, http-server, ocr, speech-recognition, caching, machine-learning
- domain: large-language-models, apis, self-hosted
- platform: python, self-hosted
- tags: embeddings, semantic-search, fastapi, llama-cpp, faiss, vector-similarity, whisper, textract, swagger-ui, search, natural-language-processing, docker, linux, macos

## Member repositories
- Dicklesworthstone/swiss_army_llama (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:24.565006+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:58:14.250232+00:00, confidence not recorded.
  - readme: https://github.com/Dicklesworthstone/swiss_army_llama (fetched 2026-08-28T04:03:24.565006+00:00, sha bd331cd812cb)
- Data as of 2026-08-30T08:39:29.467469+00:00.