pablomdata/llm-quickstart

Minimal examples for tinkering around with LLMs.

LLM Quickstart

A few minimal examples for tinkering around with LLMs.

• Docker: Deploys a server to run Llama Index locally.

  • Be sure Docker is running (in Windows, launch Docker Desktop).
  • To deploy: docker compose up -d.
  • To turn down: docker compose down.
  • To recreate (sometimes necessary if you modify the Dockerfile) docker-compose up --build --force-recreate.

• API: Examples for calling to OpenAI API and HuggingFace's InferenceAPI.

Examples and Use Cases

• Examples from Llama Index.
• Use Cases - Llama Index docs.
• Web application starter packs.
• Small Corpora in Czech for quick prototyping/experimenting

Other alternatives to running local models

• Llama Index - Private setup.
• GPT4All.
• PrivateGPT.