What is laminar
Laminar is designed to help teams ship reliable AI products by providing tools for tracing, evaluating, and labeling large language model (LLM) applications. It offers visibility into every execution step, collects valuable data for evaluations, and supports few-shot examples and fine-tuning. The platform is backed by Y Combinator and is used by various teams to improve their LLM-based features.
How to Use laminar
To start tracing your LLM application with Laminar, you can use the following code snippet:
import { Laminar, observe } from '@lmnr-ai/lmnr';
// automatically traces common LLM frameworks and SDKs
Laminar.initialize({ projectApiKey: "..." });
// you can also manually trace any function
const myFunction = observe({name: 'myFunc'}, async () => {
...
})
Features of laminar
-
Tracing
Provides visibility into every execution step of your LLM application, collecting data for evaluations, few-shot examples, and fine-tuning.
-
Evaluation
Helps maintain high accuracy in LLM-based features by providing tools for continuous evaluation and monitoring.
-
Labeling
Supports the labeling of data, which is crucial for training and fine-tuning LLMs.