// Gottfried Haider

// based on https://editor.p5js.org/jonathan.ai/sketches/O8mvwJtfp

​

let model; // this variable will contain the model loaded in preload()

let result; // gotResults() will store the most recent result there

​

// asynchronously load transformers.js and instantiate model

async function preload() {

  let transformers = await import(

    "https://cdn.jsdelivr.net/npm/@xenova/transformers@2"

  );

  console.log("transformers.js version " + transformers.env.version);

​

  // the task (e.g. "sentiment-analysis") and specific model to use for it

  // (optional) go into the next line

  model = await transformers.pipeline("sentiment-analysis");

  // see https://github.com/xenova/transformers.js?tab=readme-ov-file#supported-tasksmodels for a list of possibilities

  console.log("Model loaded");

  modelReady();

}

​

async function predict() {

  // everything but the last argument to predict() is a parameter for the model

  let modelArgs = [];

  for (let i = 0; i < arguments.length - 1; i++) {

    modelArgs.push(arguments[i]);

  }

​

  let out = await model(...modelArgs);

  // but the last argument is the callback function to call

  if (typeof arguments[arguments.length - 1] == "function") {

    arguments[arguments.length - 1](out);

  }

}

​

function setup() {

  createCanvas(400, 400);

}

​

function draw() {

  background(255);

  textAlign(CENTER);

  if (result) {

    text(

      "Sentiment is " +

        result[0].label +

        " with confidence " +

        nf(result[0].score * 100, 0, 2) + "%",

      width / 2,

      height / 2

    );

  } else {

    text("Loading...", width / 2, height / 2);

  }

}

​

function modelReady() {

  predict("I hate it when it rains", gotResults);

}

​

function gotResults(data) {

  console.log("Model returned", data);

  result = data;

}

​