https://github.com/nat-418/ask-ollama

Query Ollama models with RAG
https://github.com/nat-418/ask-ollama

Last synced: 13 days ago
JSON representation

Query Ollama models with RAG

• Host: GitHub
• URL: https://github.com/nat-418/ask-ollama
• Owner: nat-418
• Created: 2023-10-22T16:13:09.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2023-10-22T16:33:11.000Z (about 1 year ago)
• Last Synced: 2024-11-05T11:36:45.674Z (about 2 months ago)
• Language: JavaScript
• Size: 29.3 KB
• Stars: 6
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# ask-ollama

Query Ollama models with RAG

## Description

This repository contains a program, `ask-ollama`,

that uses [retrieval-automated generation] with

[Ollama]-compatible [large language models] to provide

an advanced web-search command-line interface.

## Usage

This program can be invoked using `npx ask-ollama  `.

If you don't already have Ollama running locally, you can follow

the Docker steps in the Development section of this README to get

started and install some models. This also provides a ChatGPT-style

web UI without this project's RAG capabilities for testing.

```bash

  Usage

    $ ask-ollama  

  Options

    --model, -m  Specify model (defaults to llama2)

  Examples

    $ ask-ollama 'What time is it now in Paris?' 'https://www.timeanddate.com/worldclock/france/paris'

```

## Development

This tool uses the [LangChain] framework for JavaScript.

To hack on the project, follow these steps:

1. Install [Docker] and [Node.js]

2. Clone this repository

3. Run `$ docker-compose up` from inside this repository

4. Install a model using, e.g., `$ docker exec -it ollama-backend-1 ollama pull llama2`

5. Run `$ npm install` from inside this repository

6. Edit `ask-ollama.js`

[retrieval-automated generation]: https://research.ibm.com/blog/retrieval-augmented-generation-RAG

[Ollama]: https://ollama.ai

[large language models]: https://ollama.ai/library

[LangChain]: https://docs.langchain.com/docs/

[Docker]: https://www.docker.com

[Node.js]: https://nodejs.org/