// Download Ollama: https://ollama.com/

​

// Step 2:

// Launch Ollama and follow its instructions.

​

// Step 3:

// Find a model of your choice in Ollama library: https://ollama.com/library/

​

// Step 4:

// Copy the code within the model page (such as llama3.1) and run it in your terminal (it may take a while to download): https://ollama.com/library/llama3.1

// ollama run llama3.1

​

// Step 5:

// Change model name in this sketch to match the model you just downloaded in the previous step

// mode: "llama3.1"

​

// Step 6:

// Download this sketch (you might need to save this sketch to your own account before you can download it) and run this sketch locally.

// Instructions on how to run your sketch locally: https://github.com/processing/p5.js/wiki/Local-server

​

// References:

// https://www.jsdelivr.com/package/npm/ollama-js-client

​

const Ollama = window.OllamaJS;

​

const ollama = new Ollama({

  model: "llama3.1",

  url: "http://127.0.0.1:11434/api/",

});

​

let responseContent = "Click canvas for response!";

let responseDone = true;

​

function setup() {

  createCanvas(400, 400);

}

​

function draw() {

  background(220);

  textSize(15);

  text(responseContent, 20, 20, width - 40, height - 40);

}

​

function mouseClicked() {

  if (responseDone) {

    responseDone = false;

    responseContent = "";

    ollama.chat(

      [

        { role: "system", content: "You are a llama AI assistant." },

        { role: "assistant", content: "Hello, I am your AI friend." },

        { role: "user", content: "What is p5.js?" },

      ],

      (error, response) => {

        if (error) {

          console.error(error);

        } else {

          // console.log(response);

          responseContent += response.message.content;

          responseDone = response.done;

        }

      }

    );

  }

}