Baserun NPM | npm.io

Baserun

Baserun is the testing and observability platform for LLM apps.

Quick Start

1. Install Baserun

npm install baserun
# or
yarn add baserun

2. Generate an API key

Create an account at https://baserun.ai. Then generate an API key for your project in the settings tab and set it as an environment variable.

export BASERUN_API_KEY="your_api_key_here"

3. Start monitoring

Baserun will automatically trace all the calls to OpenAI and Anthropic. A simple await baserun.init() is enough to set everything up.

import OpenAI from 'openai';
import { baserun } from 'baserun';

const openai = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

await baserun.init();

const chatCompletion = await openai.chat.completions.create({
  model: 'gpt-3.5-turbo',
  temperature: 0.7,
  messages: [
    {
      role: 'user',
      content: 'What are three activities to do in Paris?',
    },
  ],
});

4. Check out the traces

Now head over to https://baserun.ai/monitoring/traces and have a look at the traces that were just created.

Testing

Baserun comes with built-in jest support, allowing you to run evaluations on your prompts, which get reported to the Baserun dashboard.

How to test

Use our Jest preset and start immediately logging to Baserun. By default all OpenAI completion and chat requests will be logged to Baserun. Logs are aggregated by test.

// test_module.spec.ts

import OpenAI from 'openai';

const openai = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

describe('Baserun end-to-end', () => {
  it('should suggest the Eiffel Tower', async () => {
    const chatCompletion = await openai.chat.completions.create({
      model: 'gpt-3.5-turbo',
      temperature: 0.7,
      messages: [
        {
          role: 'user',
          content: 'What are three activities to do in Paris?',
        },
      ],
    });

    expect(chatCompletion.choices[0].message!.content!).toContain(
      'Eiffel Tower',
    );
  });
});

To run the test and log to baserun:

jest --preset baserun test_module.spec.ts
...
========================Baserun========================
Test results available at: https://baserun.ai/runs/<id>
=======================================================

Existing presets

If you are already using a Jest preset such as ts-jest you will need to merge the presets in a Jest config

// jest.config.js or jest.config.baserun.js

const tsPreset = require('ts-jest/jest-preset');
const baserunPreset = require('baserun/jest-preset');

module.exports = {
  ...tsPreset,
  ...baserunPreset,
  testTimeout: 10000,
};

Then to run a test and log to baserun:

jest test_modules.spec.ts
...
========================Baserun========================
Test results available at: https://baserun.ai/runs/<id>
=======================================================