// files. Make sure to duplicate it, rather than copying

// and pasting code :)

​

let local_sd_api = "https://gpu.gohai.xyz:3000/";

let img;

​

function setup() {

  createCanvas(400, 400);

​

  requestLSD("GET", "sdapi/v1/options", gotOptions);

}

​

function gotOptions(results) {

  console.log("Current model " + results.sd_model_checkpoint);

​

  // for inpainting, we want to use this model

  if (

    results.sd_model_checkpoint != "v1-5-pruned-emaonly.ckpt [cc6cb27103]"

  ) {

    selectModel("v1-5-pruned-emaonly.ckpt [cc6cb27103]");

  } else {

    modelReady();

  }

}

​

function selectModel(sd_model_checkpoint) {

  let options = {

    sd_model_checkpoint: sd_model_checkpoint,

  };

​

  console.log("Changing to " + sd_model_checkpoint);

  requestLSD("POST", "sdapi/v1/options", options, modelReady);

}

​

function modelReady() {

  // 0.5 is the "weight" to apply to the LORA

  // "barbiecore" is the activation word

​

  let modelInput = {

    prompt: " <lora:BarbieCore:0.5> barbiecore sunny day at the beach",

    // for more parameters, see the WebUI and results.parameters

  };

​

  requestLSD(

    "POST",

    "sdapi/v1/txt2img",

    modelInput,

    donePredicting

  );

​

  console.log("Starting prediction, this might take a bit");

}

​

function donePredicting(results) {

  console.log(results);

  if (results && results.images.length > 0) {

    img = loadImage('data:image/png;base64,' + results.images[0]);

  }

}

​

function draw() {

  background(220);

​

  if (img) {

    imageMode(CENTER);

    image(img, mouseX, mouseY, img.width / 2, img.height / 2);

  }

}

​