Evaluation Quick Start

This quick start will get you up and running with our evaluation SDK and Experiments UI.

1. Install LangSmith

Python

pip install -U langsmith

TypeScript

yarn add langsmith

2. Create an API key

To create an API key head to the Settings page. Then click Create API Key.

3. Set up your environment

Shell

export LANGCHAIN_TRACING_V2=true
export LANGCHAIN_API_KEY=<your-api-key>

4. Run your evaluation

Python

from langsmith import evaluate, Client
from langsmith.schemas import Example, Run

# 1. Create and/or select your dataset
client = Client()
dataset = client.clone_public_dataset("https://smith.langchain.com/public/a63525f9-bdf2-4512-83e3-077dc9417f96/d")

# 2. Define an evaluator
# For more info on defining evaluators, see: https://docs.smith.langchain.com/evaluation/how_to_guides/evaluation/evaluate_llm_application#use-custom-evaluators
def is_concise_enough(root_run: Run, example: Example) -> dict:
  score = len(root_run.outputs["output"]) < 3 * len(example.outputs["answer"])
  return {"key": "is_concise", "score": int(score)}

# 3. Run an evaluation
evaluate(
  lambda x: x["question"] + "is a good question. I don't know the answer.",
  data=dataset.name,
  evaluators=[is_concise_enough],
  experiment_prefix="my first experiment "
)

TypeScript

import { Client } from "langsmith";
import { evaluate } from "langsmith/evaluation";
import type { EvaluationResult } from "langsmith/evaluation";
import type { Run, Example } from "langsmith/schemas";

// 1. Define a dataset
const client = new Client();
const datasetName = "my first dataset"
const dataset = await client.clonePublicDataset(
"https://smith.langchain.com/public/a63525f9-bdf2-4512-83e3-077dc9417f96/d",
{ datasetName: datasetName }
)

// 2. Define an evaluator
function isConcise(rootRun: Run, example: Example): EvaluationResult {
const score = rootRun.outputs?.outputs.length < 3 * example.outputs?.answer.length;
return { key: "is_concise", score: score };
}

// 3. Run an evaluation
// For more info on evaluators, see: https://docs.smith.langchain.com/concepts/evaluation#evaluators
await evaluate(
(exampleInput) => {
return {
answer: exampleInput.question + " Good question. I don't know the answer"
};
}, {
data: datasetName,
evaluators: [isConcise],
experimentPrefix: "my first experiment ",
});

5. View Experiments UI

Click the link printed out by your evaluation run to access the LangSmith experiments UI, and explore the results of your evaluation.