Intelligent Docs

Appearance

Using Ollama ​

Integrating Ollama into your program is simple using intelligent.

Integrating a Locally-Installed Ollama AI Model ​

intelligent provides a simple interface for integrating a locally-installed OLLAMA AI model into your program. This document outlines the steps to get started.

Prerequisites ​

  • Node.js (version 14 or higher)

  • Ollama installed locally on your machine

  • intelligent installed. npm install intelligent@latest

Usage ​

Here's a step-by-step guide to using the OllamaService class:

1. Import the OllamaService class ​

ts
import { OllamaService } from 'intelligent';

2. Create an instance of the OllamaService class ​

Pass the name of the locally-installed OLLAMA model to the constructor. If no model name is provided, it defaults to "llava".

ts
const ollamaService = new OllamaService('llava');

3. Call the response method ​

Pass a prompt to the response method to generate a response using the locally-installed OLLAMA model.

ts
await ollamaService.response('Your prompt here').then((res) => {
  console.log(res);
});;

Supported Models ​

The OllamaService class supports any locally-installed OLLAMA model. You can specify the model name when creating an instance of the OllamaService class.

Example Use Case ​

Here's an example of using the OllamaService class to generate a response to a user's query:

ts
import { OllamaService } from 'intelligent';

const ollamaService = new OllamaService('llava');

async function handleUserQuery(query) {
  try {
    await ollamaService.response(query).then((res) => {
	  console.log(res);
	});
  } catch (error) {
    console.error(error);
  }
}

handleUserQuery('What is the meaning of life?');

By following these steps, you can easily integrate a locally-installed Ollama AI model into your program using the OllamaService class.

Troubleshooting ​

  • Make sure Ollama is installed locally on your machine and the ollama & intelligent package is installed.

  • Verify that the model name passed to the OllamaService constructor matches the name of the locally-installed Ollama model.

  • Check the console for any error messages if the response method fails to generate a response.