Skip to main content

Prompt + LLM

One of the most foundational Expression Language compositions is taking:

PromptTemplate / ChatPromptTemplate -> LLM / ChatModel -> OutputParser

Almost all other chains you build will use this building block.

Interactive tutorial
The screencast below interactively walks through a simple prompt template + LLM chain. You can update and run the code as it's being written in the video!

PromptTemplate + LLM​

A PromptTemplate -> LLM is a core chain that is used in most other larger chains/systems.

tip

See this section for general instructions on installing integration packages.

npm install @langchain/openai
import { ChatOpenAI } from "@langchain/openai";
import { PromptTemplate } from "@langchain/core/prompts";

const model = new ChatOpenAI({});
const promptTemplate = PromptTemplate.fromTemplate(
  "Tell me a joke about {topic}"
);

const chain = promptTemplate.pipe(model);

const result = await chain.invoke({ topic: "bears" });

console.log(result);

/*
  AIMessage {
    content: "Why don't bears wear shoes?\n\nBecause they have bear feet!",
  }
*/

API Reference:

Often times we want to attach kwargs to the model that's passed in. To do this, runnables contain a .bind method. Here's how you can use it:

Attaching stop sequences​

Interactive tutorial
The screencast below interactively walks through an example. You can update and run the code as it's being written in the video!
import { ChatOpenAI } from "@langchain/openai";
import { PromptTemplate } from "@langchain/core/prompts";

const prompt = PromptTemplate.fromTemplate(`Tell me a joke about {subject}`);

const model = new ChatOpenAI({});

const chain = prompt.pipe(model.bind({ stop: ["\n"] }));

const result = await chain.invoke({ subject: "bears" });

console.log(result);

/*
  AIMessage {
    contents: "Why don't bears use cell phones?"
  }
*/

API Reference:

Attaching function call information​

Interactive tutorial
The screencast below interactively walks through an example. You can update and run the code as it's being written in the video!
import { ChatOpenAI } from "@langchain/openai";
import { PromptTemplate } from "@langchain/core/prompts";

const prompt = PromptTemplate.fromTemplate(`Tell me a joke about {subject}`);

const model = new ChatOpenAI({});

const functionSchema = [
  {
    name: "joke",
    description: "A joke",
    parameters: {
      type: "object",
      properties: {
        setup: {
          type: "string",
          description: "The setup for the joke",
        },
        punchline: {
          type: "string",
          description: "The punchline for the joke",
        },
      },
      required: ["setup", "punchline"],
    },
  },
];

const chain = prompt.pipe(
  model.bind({
    functions: functionSchema,
    function_call: { name: "joke" },
  })
);

const result = await chain.invoke({ subject: "bears" });

console.log(result);

/*
  AIMessage {
    content: "",
    additional_kwargs: {
      function_call: {
        name: "joke",
        arguments: '{\n  "setup": "Why don\'t bears wear shoes?",\n  "punchline": "Because they have bear feet!"\n}'
      }
    }
  }
*/

API Reference:

PromptTemplate + LLM + OutputParser​

Interactive tutorial
The screencast below interactively walks through an example. You can update and run the code as it's being written in the video!

We can also add in an output parser to conveniently transform the raw LLM/ChatModel output into a consistent string format:

import { ChatOpenAI } from "@langchain/openai";
import { PromptTemplate } from "@langchain/core/prompts";
import { RunnableSequence } from "@langchain/core/runnables";
import { StringOutputParser } from "@langchain/core/output_parsers";

const model = new ChatOpenAI({});
const promptTemplate = PromptTemplate.fromTemplate(
  "Tell me a joke about {topic}"
);
const outputParser = new StringOutputParser();

const chain = RunnableSequence.from([promptTemplate, model, outputParser]);

const result = await chain.invoke({ topic: "bears" });

console.log(result);

/*
  "Why don't bears wear shoes?\n\nBecause they have bear feet!"
*/

API Reference:

Help us out by providing feedback on this documentation page: